How to assess the impact of herbicide exposure on soil microbial communities

Although the desorption of an herbicide from soil particles into the aqueous phase facilitates its biodegradation, the bioavailable fraction is also a potential risk for non-degrading microbial populations. Microbial-mediated processes in soils are of critical importance to ecosystem functions, including transformation of organic matter, nutrient release and degradation of xenobiotics. Therefore, an active soil microbial population is considered a key component of good soil quality (Parkin et al., 1996; Pell et al., 2006). Several biological parameters have been used to assess soil quality and health as affected by agricultural practices (Anderson, 2003; Benedetti & Dilly, 2006). Microbes are expected to be more effective indicators than physical and chemical parameters as they are able to respond immediately to environmental changes (Nannipieri et al., 2002).

The effects of pesticides on the microbiota can be assessed at the whole community-level (e.g., respiration, enzyme activities, biomass, total bacteria counts, etc.) or at sub-community level (i.e., specific physiological or phylogenetic groups). The use of molecular tools has greatly improved the ability to detect pesticide-induced changes, as they allow better resolution of the microbial community structure. The recommended approach for assessing the effects of pesticides on microbial communities is the simultaneous measurement of multiple ecological, structural and functional end points in soil microcosms or terrestrial model ecosystems, rather than reliance on a single assay (Nannipieri et al., 2002; Burrows & Edwards, 2004; Joergensen & Emmerling, 2006). It should be noted that there is little value in assessing the effects of unrealistically high herbicide concentrations in agricultural soils, as there is no reason to expect that those levels would be reached under normal agricultural use. This section is not intended to be an extensive literature review, but rather show the considerable variation in response among soil microbial communities and the diversity of parameters available to assess potential negative impacts of herbicides on the microbiota.

4.1 Microbial respiration

Besides being a generally accepted measure of total soil microbial activity, respiration has been used as a sensitive indicator of pesticide and heavy metal toxicity (e.g. Anderson (2003); Yao et al. (2006)). Zabaloy & Gomez (2008) observed that metsulfuron methyl at 100 pg kg-1 soil depressed cumulative respiration (measured as evolved CO2 at the end of the 6 weeks incubation) in a Typic Haplustoll [TH] soil while it had no effect in a Petrocalcic Paleustoll [PP] soil, even at a dose of 10 mg kg-1 soil. Similar results have been reported by

Dinelli et al. (1998) and Accinelli et al. (2002) in soils amended with low doses of sulfonylurea (triasulfuron, primisulfuron methyl and rimsulfuron). Zabaloy & Gómez (2008) proposed that the lower tolerance of the microbial community of TH soil was the result of low adsorption and degradation of herbicide due to higher pH in TH soil (7.4) compared to PP (6.1) (Figure 7). Phytotoxic effects of metsulfuron have been reported in soils with high pH (Walker et al., 1989). Higher degradation of metsulfuron methyl in acidic soils compared to alkaline soils is due to the combined actions of chemical hydrolysis and microorganisms (Pons & Barriuso, 1998; Andersen et al., 2001). No mineralization of either metsulfuron methyl or tribenuron methyl was observed in soils of pH > 8, unless the compounds have been pre-hydrolyzed (Andersen et al., 2001). Several studies reported that the effects of glyphosate and 2,4-D on microbial respiration at low rates, equivalent to agronomic doses, are negligible (e.g. Wardle & Parkinson (1990); Busse et al. (2001); Zabaloy & Gómez (2008)).

time after treatment (days)

Fig. 7. Effect of two rates of metsulfuron methyl on cumulative CO2 evolution of Typic Haplustoll (a) and Petrocalcic Paleustoll (b) soils. Symbols: (filled squares) 0.01 mg a.i. kg-1 air-dried soil; (gray diamonds) 0.1 mg a.i. kg-1 air-dried soil; (empty triangle) control (distilled water). Error bars indicate standard deviation. Error bars not shown were smaller than the symbols. From Zabaloy & Gómez (2008)

4.2 Enzyme activities

Many studies have shown that enzyme activities are sensitive enough to detect the effects of soil pollutants, including heavy metals (Avidano et al., 2005), insecticides (Yao et al., 2006) and herbicides (Sannino & Gianfreda, 2001). Dehydrogenases exist as an integral part of intact cells and represent the oxidative activities of soil microbes, whereas fluorescein diacetate (FDA) hydrolysis can be catalyzed by intracellular and extracellular lipases, esterases and proteases produced by microorganisms (Shaw & Burns, 2006). Both are well-established methods to measure the microbial mineralizing capacity in soil and are suitable to assess broad-spectrum biological activity in the short-term (Nannipieri et al., 2002). Zabaloy et al. (2008) reported that metsulfuron-methyl and 2,4-D had transient, relatively small (<25% change from control) effects on soil enzyme activities within two weeks after herbicide addition. While both herbicides induced an early reduction in FDA, 2,4-D also stimulated DHA in the different soils analyzed. In contrast, glyphosate caused a significant reduction (50 %) of intracellular dehydrogenase activity, suggesting a strong influence on bacterial metabolism (Zabaloy et al., 2008). Metsulfuron-methyl at comparable doses inhibited urease, amylase and protease activities in loamy sand and clay loam soils (Ismail et al., 1998). There is general agreement on the lack of significant effects of agricultural rates of metsulfuron (Dinelli et al., 1998; Accinelli et al., 2002) and 2,4-D (Frioni, 1981;Wardle & Parkinson, 1990) on different enzymes. Variable effects of glyphosate and glufosinate on soil enzymatic activities have been reported. In general, literature reports mainly stimulatory effects of glyphosate on enzyme activities for doses within a range of 2-200 mg a.i. kg-1 soil (Sannino & Gianfreda, 2001; Accinelli et al., 2002; Araujo et al., 2003; Lupwayi et al., 2007).

4.3 Microbial biomass and abundance

The number and biomass of microorganisms are basic properties of ecological studies, and which can be related to parameters describing microbial activity and soil health (Bolter et al., 2006). Substrate-induce respiration is a commonly-used, sensitive parameter for the observation of pollutant impacts on soil microorganisms (Brohon et al., 2001). Under standardized conditions, the metabolism of glucose added in excess is limited by the amount of active aerobic microbes in soil. Initially, there is no microbial growth and the respiratory response is proportional to glucose-responsive microbial biomass already present in soil (Hoper, 2006). The glucose-responsive and more active part of the microbial community, determined by the SIR biomass, is more sensitive to pollutants than the total microbial biomass, as measured biochemically (Hoper, 2006; Chander et al., 2001; Zabaloy et al., 2008). The number of physiological groups of bacteria has also proved to be useful to measure structural changes in soil due to several anthropogenic factors. Glyphosate is an organophosphonate that can be used as a source of P, C or N by either gram-positive or gram-negative bacteria (van Eerd et al., 2003). Accordingly, increases in bacterial abundance and biomass (Zabaloy et al., 2008) and fungal counts (Araujo et al., 2003; Ratcliff et al., 2006) after glyphosate doses comparable to field rates have been observed. Supporting the hypothesis of a bacterial role in glyphosate dissipation, Gimsing et al. (2004) found a high correlation between glyphosate mineralization rates and Pseudomonas spp. counts in five different Danish soils. Moreover, two soils with high glyphosate mineralization rates also showed high CFU counts (Gimsing et al., 2004). Conversely, low rates of 2,4-D (< 10 mg kg-1 soil) have no effects on heterotrophic bacteria counts (Ka et al., 1995; Merini et al., 2007; Zabaloy et al., 2010). The abundance of cellulose degraders and Azotobater were reported to decrease with 2,4-D treatment (Frioni, 1981), although the dose used was several times higher than the expected concentration in soil after a field rate application.

4.4 Microbial community structure

Community structure could be defined as the abundance and proportion of distinct phylogenetic and functional groups. Functional groups are defined by the substrates used for energy metabolism (Pankhurst et al., 1996). Community-level end points may not be sensitive enough to detect minor shifts in microbial community structure, due to the inherent functional redundancy that is recognized to exist in soil microbial communities. The disappearance of a certain member of the microbial community as a result of herbicide (or other pollutant) exposure may eliminate key ecosystem functions and/or impair the ability of the microbial community to respond to other environmental perturbations (i.e., reducing resilience). Physiological, biochemical or genetic profiling methods give insight of such potential shifts at the subcommunity level. Popular methods include community-level physiological profiles

(CLPP), phospholipids fatty acid analysis (PLFA), and various DNA fingerprint techniques (e.g. denaturing or thermal gradient gel electrophoresis [DGGE/TGGE], terminal restriction fragment length polymorphism [T-RFLP]). These and other methods have been summarized in the excellent reviews by Torsvik et al. (1996), Preston-Mafham et al. (2002), Lynch et al. (2004), Kirk et al. (2004), Ogram et al. (2007) and Garland et al. (2007).

Unintended consequences of herbicide applications may be the reduction of sensitive populations and/or stimulation of a certain microbial group with or without detriment to co-existing microbial populations that may compete for available resources. Several investigations that used culture-independent methods reported only slight, short-lived effects of field levels of glyphosate (Weaver et al., 2007; Accinelli et al., 2007) and 2,4-D (Chinalia & Killham, 2006; Macur et al., 2007; Vieuble-Gonod et al., 2006) on microbial communities. No major changes in community structure, assessed by CLPP and PLFA, occurred with application of field rate concentrations of glyphosate in soils from two pine plantations in California (Ratcliff et al., 2006). Both higher abundance of PLFA biomarkers of gram-negative bacteria (Weaver et al., 2007; Lancaster et al., 2009) and fungal to bacterial biomass ratios (Powell et al., 2009) have been reported in glyphosate-treated soils. In a recent study, Zabaloy et al. (2009) reported minor effects of glyphosate on sole C sources utilization with BDOBS. However, the number of 16S ribosomal gene copies, as determined by quantitative PCR (qPCR), increased in a glyphosate-treated soil relative to the control soil, although T-RFLP analysis did not show consistent selective enrichment for specific bacteria species (i.e., no specific phylotype dominated in glyphosate-treated microcosms) (Zabaloy et al., 2009). Due to the enormous diversity of soil microbial communities, more relevant results could be obtained by targeting specific functional groups that are more likely to be directly affected by the herbicide or indirectly by herbicide-induced changes in the soil environment. Interestingly, no effects of glyphosate on denitrifying bacteria nor rhizosphere fungal abundances (qPCR) or communities composition (T-RFLP) have been reported (Hart et al., 2009). Glyphosate was reported to inhibit growth of mycorrhizal fungi and could favor the growth of less desirable fungal species, like soil-borne pathogens (Johal & Huber, 2009). Krzysko-Lupicka & Sudol (2008) observed a bias towards the selection of autochtonous Fusarium strains after treatment with the herbicide. This could be related with changes in microbial populations that alter the equilibrium and ultimately lead to diminishing biodiversity, as the postulated decrease in the (pseudomonad) antagonists of fungal pathogens observed by Kremer & Means (2009) in long-term field studies. The most noticeable effect of 2,4-D on community structure is the enrichment of degrading populations that use this compound as a source of C and energy. Zabaloy et al., 2010 reported a persistent 2,4-D degrading population able to use the herbicide as C and energy source in an agricultural soil where herbicide applications had ceased 2 years before the study. The number of degraders increased immediately after treatment of soil microcosms with 2,4- D and remained high until the end of the incubation, while culturable aerobic heterotrophic bacteria counts were not affected by the herbicide (Figure 8). The addition of succinate (S) as an alternative source of C to soil microcosms did not stimulate degrader population, which confirmed that 2,4-D degradation in this soil was mainly a metabolic process performed by specific degraders. Similar results have been obtained by a number of researchers that used a range of herbicide concentrations in different agricultural soils (e.g. Ka et al. (1995); Merini et al. (2007); Macur et al. (2007); Lerch et al. (2009). One practical implication of the proliferation of soil microbes able to degrade some herbicides, such as foliar-applied chlorophenoxy acids, is that this phenomenon guarantees self-cleaning of herbicide-impacted agricultural soils, reducing the risk of contamination.

4.5 Pollution-induced community tolerance

The pollution-induced community tolerance (PICT) concept is based on the assumption that long-term exposure of a community to a given toxicant will lead to a higher tolerance for this pollutant (Blanck et al., 1988; Blanck, 2002). PICT is tested by collecting intact communities from polluted and reference sites and exposing these communities to contaminants under controlled conditions. Increased community tolerance resulting from the elimination of sensitive species and addition of tolerant species is considered strong evidence that changes were caused by the pollutant. A fundamental step in the PICT measurements is the selection of an ecologically relevant parameter as endpoint that reflects the toxic effects at the community level (Blanck, 2002).

Control S

Time after amendment (days)

Fig. 8. Effect of combined amendments of 2,4-D and succinate (S) on aerobic heterotrophic bacteria (AHB) counts (a) and most probable number of 2,4-D degraders (MPN2,4^D) for soil microcosms sampled after 0, 4 and 33 days of incubation. AHB data are given as means ±S.E (n=3). MPN2,4^d data are represented as median of three replicates and 95% confidence intervals. LoD, limit of detection. From Zabaloy et al., 2010.

Microbial activity may be affected by soil characteristics as well as other environmental factors other than contamination. However, increased tolerance to a specific contaminant is less sensitive to variation in physicochemical variables, and more likely a direct result of contaminant exposure (Siciliano & Roy, 1999; Gong et al., 2000). The PICT approach has been used to study effects of chemicals on microbial communities with various methods such as BiologTM plates (Schmitt et al., 2004), respirometer (Gong et al., 2000) and methane oxidation assay (Seghers et al., 2003). Zabaloy et al., 2010 used BDOBS to assay mineralization of coumaric acid as an indication of PICT to 2,4-D in an agricultural and a forest soil. This study revealed that past field exposure of the agricultural soil to 2,4-D was enough to develop resistant microbial populations, while the herbicide exerted a more severe inhibitory effect on coumaric acid use in the pristine forest soil (Figure 9). In a similar study, Seghers et al. (2003) reported that long-term use of atrazine and metolachlor selected towards a methanotrophic community more tolerant to the methane oxidation inhibitor 2,4-D in an agricultural soil.



0 50 100 150 200 250 300

Fig. 9. Respiratory index with coumaric acid as C source, in agricultural soil treated with 5 mg kg-1 de 2,4-D (ERL) or untreated (control) and forest soil, exposed to increasing doses of 2,4-D (25-250 mg l-1) in BDOBS. Values represent means ±S.E (n=3); for forest soil is the average of two samples. From Zabaloy et al., 2010

0 0

Post a comment