Introduction

Because of their unique properties, engineered nanoparticles (NPs) are increasingly produced and used in diverse commercial products (agriculture, cosmetics, energy, electronics, paint, medicine…). Toxiceffects of NPs have been observed in numerous in vitro studies; in particular, many NPs, such as silver-, copper-, or zinc oxides-NPs, are known to have antimicrobial properties (Neal 2008; Fabrega et al. 2009; Dinesh et al. 2012). The toxicity mechanism often mentioned is an oxidative stress generated by the production of reactive oxygen species (ROS) from NPs in contact with microbial membranes, causing disruption of membranes, oxidation of proteins, or interruption of energy transduction (Klaine et al. 2008; Neal 2008; Xia et al. 2008).

During the different phases of their life cycle, from production to disposal, NPs can be released to the environment, raising great concerns about potential ecological risks. Direct measurement of NP concentration in the environment remains difficult due to current technical limitations (Cornelis et al. 2014). Currently, the only way to get information on existing levels of NPs in the environment is to model predicted environmental concentrations (Sun et al. 2014). Exposure modeling strongly suggests that soil could be a major sink of NPs compared to air and water ecosystems (Gottschalk et al. 2009; Keller et al. 2013; Sun et al. 2014).

NPs can enter soil through various pathways, such as agricultural amendments of sewage sludge, atmospheric deposition, landfills, or accidental spills during industrial production. The current models estimate that in sewage sludge-treated soil, TiO2-NP concentrations increase between 0.94 and 3.6 mg kg−1 per year, whereas for Ag-NPs and fullerenes, the yearly predicted increases are more than 1000-fold lower. In this case, concentration increases are between 0.09 and 0.65 μg kg−1 and between 0.38 and 1.5 μg kg−1, respectively (Sun et al. 2014). Intentional applications of NPs can also be possible in a context of soil remediation. Nanoscale zero-valent iron (nZVI) particles have the potential to remediate diverse environmental contaminants such as chlorinated organic compounds or inorganic compounds for example. This remediation strategy has been mainly employed for the decontamination of groundwater, and its utility in soil remediation is being increasingly considered (Naja et al. 2009; Satapanajaru et al. 2008). Therefore, there is a demand to assess the risks associated to NPs in soils, in order to preserve the soil capacity to fulfill essential ecosystem services.

Soil microbial communities are both relevant and sensitive indicators of soil perturbations (Brookes 1995; Kandeler et al. 1996; Holden et al. 2014) because of their key role in biogeochemical cycling (carbon, nitrogen, phosphorous, sulfur cycles), biodegradation of pollutants, crop production, and climate. A microbial ecotoxicological approach based on the response of soil microbial communities offers the opportunity to evaluate the impact of pollutants on natural assemblages of populations, contrary to toxicological studies which use single populations in synthetic media. This approach allows a realistic assessment of the response of natural communities to NP contamination and leads consequently to a better environmental risk assessment. Environmental constraints and/or anthropogenic perturbation can trigger a microbial response observable through different variables among the community. The microbial community responses to a disturbance can be monitored using (i) activity rates, (ii) abundances, and (iii) diversity (Brookes 1995; Griffiths and Philippot 2013). Through this review, we provide a synthesis of studies assessing the overall impact of NPs on soil biodiversity and functioning using the descriptors mentioned above.

Among NPs, a wide range of different materials with diverse physical, chemical, and toxicological properties exist. Inorganic and organic NPs can be distinguished based on their core material (Ju-Nam and Lead 2008). Inorganic NPs are divided into metal, metal oxide, and quantum dots NPs, while fullerenes and carbon nanotubes (single and multi-walled CNTs) are defined as organic NPs. Since NPs cannot be considered as a single homogeneous group of contaminants, we will review separately the effects of (i) metal and metal oxides, (ii) carbon nanoparticles, and (iii) nZVI. Due to increasing interest of this latter for in situ soil remediation, a specific attention will be addressed to the nZVI. The direct application to soils of large amounts of nZVI for remediation purposes raise specific concerns about potential consequences on soil microbial communities and their key functions for soil fertility and biodegradation of pollutants.

Impact of nanoparticles on microbial activities

Soil microbial activities are good indicators of soil quality, since soil microorganisms control the transformation and mineralization of natural compounds and xenobiotics (Schloter et al. 2003). The impact of perturbations on microbial activity is classically approached through measurements of “broad” processes (Schimel and Schaeffer 2012) of soil respiration or generalist enzyme activities such as β-glucosidase, urease, and dehydrogenase. Some specific transformations or “narrow” processes realized by a phylogenetically constrained group of microorganisms (Schimel and Schaeffer 2012) can be also analyzed, such as nitrification in nitrogen cycle or the ability to degrade specific pollutants.

Impact of metal and metal oxide nanoparticles

The use of microbial activities to assess the effect of metal or metal oxide NPs is often encountered in the literature (Table 1). Contrasted responses can be observed according to the type of NPs, the concentration, the exposure time, and the measured activity.

Table 1 Review of the effects of metal and metal oxide nanoparticles on soil microbial communities
Full size table

Ag-NP has been found to reduce substrate-induced respiration and enzymatic activities after 50 days of exposure to a low concentration of Ag-NPs (0.14 mg kg−1 soil) applied in mesocosms via sewage sludge (Colman et al. 2013). Another study using much lower concentrations of Ag-NPs (0.0032, 0.032, 0.32 mg kg−1 soil) did not show any changes in the different tested enzymatic activities (Hänsch and Emmerling 2010). However, the authors have found a reduction of the net nitrogen mineralization and an increase of the metabolic quotient suggesting a reduction in substrate use efficiency in presence of Ag-NPs. Shin et al. (2012) have also reported adverse effects of Ag-NPs on enzymatic activities, especially on urease, but the larger reductions were observed only for high concentrations (100 and 1000 mg kg−1 soil). Using soluble silver ions as a positive control, these studies show that NPs themselves were responsible for the negative effects on soil microorganisms but not the silver ions dissolved from Ag-NPs (Shin et al. 2012; Colman et al. 2013).

Fe2O3-NPs stimulated urease and invertase activities, whereas Fe3O4-NPs did not induce any modification of enzymatic activities in presence of high concentrations (420, 840, and 1260 mg kg−1 soil) (He et al. 2011). These stimulations were explained as a consequence of changes observed within the bacterial community.

Aged TiO2-NPs (91 mg kg−1 soil) stimulated urease activity but decreased catalase and peroxidase activities after 9 months of incubation in lysimeters (Du et al. 2011). In the same experiment, ZnO-NPs (45 mg kg−1 soil) reduced protease, catalase, and peroxidase activities. Ge et al. (2011) also evaluated the impact of TiO2 and ZnO-NPs on substrate-induced respiration and found a reduction after 15 and 60 days with 500, 1000, and 2000 mg kg−1 soil. TiO2-NPs impact on substrate-induced respiration was also assessed in six contrasted agricultural soils (1 and 500 mg kg−1), and a significant decrease of this activity was noted only in one soil presenting a silty clay texture and a high OM content (Simonin et al. 2014).

These results indicate that metal and metal oxides can induce modifications of microbial activities in soil and consequently on biogeochemical cycles. The decrease of respiration and enzymatic activities in response to very low concentrations of Ag-NPs illustrates the high toxic potential of some metal NPs.

Impact ofcarbon nanoparticles (fullerene and CNTs)

Fullerene (C60) had no adverse effect on soil respiration or enzymatic activities (Table 1, Tong et al. 2007; Johansen et al. 2008) even after 180 days of incubation. In the same range of concentrations, no modification of microbial activities (soil respiration and enzymatic activities) was reported for MWCNT (Shrestha et al. 2013), except for the highest concentrations used (500 and 5000 mg kg−1 soil) (Chung et al. 2011). Likewise, SWCNT caused a reduction of the enzymatic activities only at high concentrations (300 to 1000 mg kg−1 soil) (Jin et al. 2013).

Taken together, these data suggest that carbon NPs have a very low toxic potential to soil microorganisms compared to metal and metal oxide NPs, because modifications of microbial activities were only observed at very high concentrations (>500 mg kg−1 soil) representing the worst case scenario given the present annual production of CNTs (Keller et al. 2013).

Impact of nanoscale zero valent iron

Tilston et al. (2013) performed an experiment with a sandy loam soil artificially contaminated with the PCB Aroclor-1242 to simulate the utilization of nZVI in a context of remediation of organochlorine-contaminated environments. The authors reported a decrease of the activity of chloroaromatic mineralizing microorganisms (2,4-D mineralization), which biodegradative functions could contribute to contaminant remediation. Nevertheless, this negative effect should be taken with caution due to the possible induction of confounding factors as shown by Cullen et al. (2011). No conclusive evidence for negative effects of nZVI was observed due to the redox properties of these particles that induced a significant abiotic modification of the concentration of either the product or substrate of assays for nitrification or deshydrogenase activities. When examining the impact of redox active particles such as nZVI on microbial oxidation–reduction reactions, sterile controls should be made to take into consideration these potential confounding factors (Cullen et al. 2011).

The influence of environmental physicochemical properties on nZVI toxicity and bioavailability was investigated by comparing their impact in different soils. Pawlett et al. (2013) found that microbial respiration induced by multiple substrates was decreased in a clay soil spiked with 270 mg nZVI kg−1 soil but not in sandy or loam soils for the same dose.

Impact of nanoparticles on microbial biomass and abundance of functional groups

Microbial biomass is a sensitive indicator of pollution in soils (Brookes 1995). It has been used as an indicator of the impact of NPs in the majority of studies assessing their ecotoxicity. Different methodologies are used in NPs ecotoxicology studies to quantify microbial biomass in soils, such as fumigation–extraction or extractable soil DNA (Hänsch and Emmerling 2010; Ge et al. 2011). In addition to these approaches that assess the total biomass including bacteria, fungi, and archaea, the abundance of specific groups of microorganism targeting universal genes (16S or 18SrRNA genes) or functional genes can be assessed using quantitative PCR (Fajardo et al. 2012).

Impact of metal and metal oxide nanoparticles

Different concentrations of Ag-NPs (0.0032, 0.032, 0.32 mg kg−1 soil) induced a decrease in microbial biomass in a dose–response manner (Hänsch and Emmerling 2010). However, soil contamination with iron oxide magnetic NPs (Fe2O3 and Fe3O4) had no effect on microbial biomass with 10 and 100 mg kg−1 soil (Vittori Antisari et al. 2013) nor on bacterial abundance with 420, 840, and 1260 mg kg−1 soil (He et al. 2011). Similarly, Simonin et al. (2014) found no significant effect of TiO2-NPs on bacterial abundances in the six soils tested. TiO2 and ZnO NPs using 500, 1000, and 2000 mg kg−1 soil, both reduced extractable soil DNA (Ge et al. 2011). However, the dose–response curves were linear with TiO2-NPs and exponential with ZnO-NPs (Ge et al. 2011). These differences were explained by contrasting bioavailability and environmental behaviors of these metal oxide NPs.

The simultaneous environmental exposure of soils to different NPs is likely to occur for example after sewage sludge application. The impact of a combination of NPs has been investigated using Ag, Cu, and Si NPs in an arctic soil (Kumar et al. 2011). In this study, a significant decrease of microbial biomass was observed after 6 months of incubation for a mixture concentration of 660 mg kg−1 soil (220 mg kg−1 of each NP).

Impact of carbon nanoparticles (fullerene and CNTs)

Contrasting effects on microbial biomass were reported for carbon nanoparticles. Johansen et al. (2008) found no effect, whereas decreases were observed by other authors but only for concentrations exceeding 250 mg kg−1 of CNTs (Chung et al. 2011; Jin et al. 2013; Rodrigues et al. 2013). These negative effects seemed to be more severe on fungal biomass compared to bacterial biomass (Rodrigues et al. 2013). These effects observed only with high concentrations are consistent with the results previously presented on the microbial activities, suggesting a limited toxicity of carbon NPs toward soil microbial communities.

Impact of nanoscale zero valent iron

The effects of nZVI have been investigated on the abundance of microbial functional groups to detect potential modifications that could be detrimental for ecosystem functioning. The presence of 34 and 10 g nZVI kg−1 soil triggered decrease in denitrifying bacteria abundance (Fajardo et al. 2012) and in chloroaromatic mineralizing microorganisms (Tilston et al. 2013), respectively. These results illustrate that a high concentration of nZVI could affect soil nitrogen cycle and the biodegradative potential of a microbial functional group. Pawlett et al. (2013) reported that nZVI addition reduced microbial biomass but only when soil was amended with 5 % straw. These results suggest that the impact of NPs may be dependent of the organic matter content of soil, which has been found to enhance NPs mobility in porous media (Ben-Moshe et al. 2010; Wang et al. 2012; Thio et al. 2011). Organic matter may be an important factor favoring NPs bioavailability for soil microorganisms. However, the number of studies available on the mobility and toxicity of NPs in natural soils with contrasted textures and organic matter contents is too limited to confirm this assumption (Fang et al. 2009; Pawlett et al. 2013; Cornelis et al. 2013; Peyrot et al. 2014; Simonin et al. 2014).

Impact of nanoparticles on microbial diversity

The measurement of the microbial biomass is a black box approach in which the genetic and functional diversity among the microbial community are not considered. Soil hosts an immense diversity of microorganisms (individual taxa commonly described as “operational taxonomic units”; OTUs) of bacteria, fungi, and archaea. Microbial diversity encompasses genetic variability within taxa (species), and the number (richness) and the relative abundance (evenness) of taxa and functional groups (guilds) in communities (Torsvik and Øvreås 2002). That is why, investigating the impact of NPs on microbial diversity is crucial to provide information on how and why soil ecosystem functioning is affected. A panel of techniques, such as PLFA profiles, PCR-DGGE, T-RFLP analysis, or next generation sequencing, has been used to evaluate the impact of NPs on microbial diversity.

Impact of metal and metal oxide nanoparticles

Ag-NPs altered bacterial community structure, after short-term exposure of sewage sludge containing 0.14 mg kg−1 of Ag-NP (Colman et al. 2013). Kumar et al. (2011) also observed that Ag-NPs could modify bacterial community in an arctic soil but with a higher concentration (660 mg kg−1). A field study also reported that zero valent Cu and ZnO-NPs had no effect on PLFA profiles and on microbial community composition as determined by pyrosequencing (Collins et al. 2012). Changes in soil bacterial community composition due to the presence of Fe3O4-NPs were observed after a short incubation (24 or 48 h) as well as after 15 and 30 days (Ben-Moshe et al. 2013; He et al. 2011). Cloning–sequencing of DGGE bands indicated a stimulation of specific groups of Actinobacteria, Duganella, Streptomycetaceae, or Nocardioides (He et al. 2011). These groups facilitate the decomposition of organic matter, which could explain the concomitant soil invertase and urease increases measured during this experiment. Different concentrations of TiO2 and ZnO-NPs decreased soil bacterial diversity after 60 days of incubation (Ge et al. 2011; Ge et al. 2012). The pyrosequencing data indicated that some of the declining taxa are known to be associated to nitrogen fixation (Rhizobiales, Bradyrhizobiaceae, and Bradyrhizobium) and methane oxidation (Methylobacteriaceae), while some positively impacted taxa are known to be associated to the decomposition of recalcitrant organic pollutants (Sphingomonadaceae) and biopolymers including protein (Streptomycetaceae and Streptomyces). The role of these taxa for soil functioning suggest potential consequences on ecosystem-scale processes. Nogueira et al. (2012) assessed the effect of five inorganic nanomaterials (TiO2, TiSiO4, CdSe/ZnS quantum dots, gold nanorods, and Fe/Co magnetic fluid) on soil bacterial community structure using DGGE. After 30 days of soil exposure, TiO2 and gold nanorods induced the highest changes in the structural diversity of bacterial community. The limited effect of TiSiO4, CdSe/ZnS quantum dots, and Fe/Co magnetic fluid NPs on DGGE profiles was attributed to their zeta potential values reflecting an unstable state. Hence, once added to the soil, they may have interacted with soil components, becoming unavailable to exert toxic effects.

Metal or metal oxide NPs may be responsible for a biodiversity loss and a modification of soil microbial community composition. Although many research focused on the effect of these NPs on bacterial communities, more work is still required to assess the impact of NPs on fungal and archaeal communities.

Impact of carbon nanoparticles (fullerene and CNTs)

Fullerene NPs had no effect on microbial diversity (Tong et al. 2007) or induced only slight modification of Eubacteria and Kinetoplastida (Protozoans) community structure on DGGE profiles (20 to 30 % of dissimilarity, Johansen et al. 2008). Pyrosequencing data indicated that MWCNT (10 g kg−1 soil) induced an enrichment of potential degraders of recalcitrant contaminants (PAH) Rhodococcus, Cellulomonas, Norcardioles, and Pseudomonas, while some bacterial genera like Derxia, Holophaga, Opitutus, and Waddlia were decreased (Shrestha et al. 2013). Using a comparative metagenomic analysis of bacterial communities, Khodakovskaya et al. (2013) found that the diversity and richness of bacterial communities were not affected by MWCNTs, while a significant modification of the bacterial composition was observed. SWCNTs induced a modification of microbial community composition resulting in a decrease of Gram-positive and -negative bacterial biomass and fungal biomass as well (Jin et al. 2014). Rodrigues et al. (2013) reported also a modification of the fungal community structure after 14 days of soil exposure to SWCNT (250 and 500 mg kg−1). Consistent with activity and abundance measurements, carbon NPs can alter soil microbial community structure but only in presence of high concentrations (>250 mg kg−1).

Impact of nanoscale zero valent iron

The impact of nZVI on microbial diversity was investigated using FISH, DGGE, and PLFA analysis. Fajardo et al. (2012) did not report a broad bactericidal effect of nZVI but observed significant shifts in the structure and phylogenetic composition of the soil microbial community after 72 h of incubation. The FISH assays provided evidence that nZVI exerts a selective pressure on the microbial community, promoting the dominance of some microbial groups (Archaea, α-Proteobacteria, and low G + C Gram-positive bacteria) or the decrease of other ones (β- and γ-Proteobacteria and subclasses). DGGE profiles also indicated a significant modification of bacterial community composition after 28 days in the presence of 10 g kg−1 of nZVI (Tilston et al. 2013). Pawlett et al. (2013) observed that nZVI caused a modification of PLFA profiles in all soil texture tested, but that these effects were modulated by the organic matter content of the soil. These studies suggest that nZVI could induce a significant modification of soil microbial community structure, affecting bacteria, archaea, and fungi populations on the short term (<4 months).

Concluding remarks

NPs toxicity toward microorganisms has been demonstrated in numerous in vitro studies (e.g., Simon-Deckers et al. 2009; Jiang et al. 2009); yet, the assessment of NPs environmental impact is still in its early stages. This synthesis on the effects of NPs on soil microbial communities supports different conclusions.

It has been demonstrated using different methodologies, different indicators, and natural soils, which NPs could have an impact on microbial activities, abundances, and diversity. However, this synthesis highlights that the toxic effects on microbial community are highly dependent on both the NPs considered and the soil properties. In fact, inorganic NPs (metal, metal oxide NPs) may have a greater toxic potential than organic NPs (fullerenes and CNTs) to soil microorganisms. An exception could be the iron oxide magnetic NPs, which exhibit limited negative effects on microbial communities even when high concentrations were applied. Alarmingly, the use of nZVI could have detrimental effects on biodegradative functions of microorganisms in a context of soil remediation. Further research is needed to determine the real efficiency of nZVI treatments in soil and their potential consequences on soil functioning using activity and functional gene measurements.

The soil properties seem to play a key role for the bioavailability of NPs, especially the clay and organic matter content. We strongly encourage to take more into consideration the physicochemical characteristics of the soil used in the experiments (texture, organic matter content, pH…) and to compare the ecotoxicity of NPs in a range of different soils. The identification of soil parameters controlling the bioavailability of NPs is fundamental for a better environmental risk assessment (Cornelis et al. 2014).

It should be noted that the effects of numerous NPs have not been investigated yet or only in a single study (Al2O3, CeO2, quantum dots, SiO2, SnO2…) (Fig. 1), whereas a significant amount of these NPs is susceptible to be released to soils (Keller et al. 2013). Some NPs have been more studied than others (Ag, TiO2, ZnO) (Fig. 1), and paradoxically, these are not necessarily the most produced and used NPs (Keller et al. 2013; Sun et al. 2014). The overall number of publications on each class of NP remains still limited to date (≤6) (Fig. 1), and thus, it is still difficult to generalize the results. More research is needed, especially through experiments using more environmentally realistic concentrations of NPs based on the predicted concentrations in modelization studies and using more realistic exposure conditions. All the literature analyzed in this review assessed the acute toxicity of NPs corresponding to a sudden disturbance due to a unique application of NPs and the monitoring of the response of microorganisms over time. These experiments assessed the short-term sensitivity of microbial communities, but to date, no data are available neither on the long-term effect and the chronic toxicity of NPs nor on the ability of microorganisms to be resilient to NP disturbance over time.

Fig. 1
figure 1

Number of publications studying the impact of metal/metal oxide or carbon nanoparticles on soil microbial communities. Thirty-one publications were available in July 2014

Full size image

Although this review emphasizes that microbial ecotoxicology is a valuable approach for risk assessment of NPs on soil ecosystem functioning, the use of microorganisms as indicators of the NPs impact in soil is still in its infancy. At least three reasons might be evoked to explain this fact; first, this is a recent concern steadily increasing since 2011 (Table 1), second this is a complex multidisciplinary subject requiring both biological and physicochemical approaches, and third, it suffers from current technical limitation that hampered our knowledge on NPs behavior in soil. The following section details some crucial subjects that must be investigated to improve our understanding of the environmental impact of these materials using microbial ecotoxicology.

Future research needs

Characterization and fate of NPs in soil

It is unadvisable to conduct ecotoxicological studies without analytical data of the used pollutant. To date, NPs have been extensively characterized in spiking suspensions usually prepared in ultrapure water, in aquaticenvironments, but not directly in the soil, because of the current technical limitations to detect NPs in complex media (Tourinho et al. 2012; Cornelis et al. 2014). To understand bioavailability and toxicity of NPs, more effort are needed to determine the fate of NPs, i.e., the speciation, the mobility, the homoaggregation and heteroaggregation processes, and all the physicochemical and biological transformations that NPs can undergo in the soil. These issues have been addressed in several reviews (see Lowry et al. 2012; Pan and Xing 2012; Cornelis et al. 2014). The crucial point for the assessment of NP toxicity to soil microorganisms relies on the accurate measurement of the bioavailable fraction of NPs in soil. As discussed by Cornelis et al. (2014), current techniques to estimate the bioavailable fraction of NPs in complex environments are unsatisfactory and need further developments.

Evaluation of a realistic exposure of soil to NPs

Currently, most of the data on the effect of NPs on soil microorganisms were obtained under unrealistic exposure conditions, using NP concentrations 50- to 10,000-fold higher than modeled concentrations in soils (Gottschalk et al. 2009; Sun et al. 2014). Few studies used NPs concentrations ≤1 mg kg−1 soil and most of them concerned Ag-NP (Table 1). More research using realistic concentrations is needed especially since it has been demonstrated that these low concentrations can still have detrimental effects for soil microbial communities (Hänsch and Emmerling 2010; Colman et al. 2013; Simonin et al. 2014).

Microcosm experiments enable assessment of the impact of NPs under controlled conditions (temperature, soil humidity…) on short-term monitoring. Mesocosms are particularly effective tools to assess the impact of pollutants on ecosystem functioning under realistic conditions and long-term monitoring. Few studies performed experiments in mesocosms and/or under field conditions to simulate more realistic exposures of soils to NPs. Moreover, the addition of NP in suspension prepared in ultrapure water used in most experiments is not a realistic mode of soil exposure. A likely route by which NPs enter soil is as aged NPs through sewage sludge applications for field fertilization. Further research in these directions will bring useful information to set regulations for the manufacture, use, and disposal of NPs.

Resilience of soil microbial communities to NPs disturbance

This review highlights that soil microorganisms are sensitive to acute contamination of NPs. However, because of the lack of long-term experiments, the resiliency of soil microbial communities to such disturbance and the return to the predisturbance activity levels after sustainable changes in microbial abundance and diversity along time remains still unknown. Allison and Martiny (2008) demonstrated that the composition of most microbial groups is sensitive and not immediately resilient to disturbance, regardless taxonomic breadth of the group or the type of disturbance. This also appears to be true for NP contamination, because in several studies, a greater effect was observed over time, but in period generally not exceeding 2 months (e.g., Ge et al. 2011; Colman et al. 2013). The response of soil microorganisms to NPs needs to be monitored over a long-term periods to evaluate whether ecosystem functioning is permanently disturbed or not (Fig. 2).

Fig. 2
figure 2

The response of soil microbial communities to NPs contamination. 1 Soil properties (clay, organic matter content, pH, ionic strength…) can modulate the toxicity of NPs for microorganisms. 2 If the NPs can be toxic, the impact on microbial community will depend on the ability of the community to resist to this perturbation. 3 If the indigenous microorganisms are not resistant to NPs contamination, soil microbial communities could be affected through different nonexclusive mechanisms: (i) effects of NPs on microbial activities due to an alteration of the synthesis and/or functioning of enzymes, (ii) a modification of microbial abundance caused by a mortality of sensitive population in presence of NPs, and (iii) a modification of microbial community structure and/or a diversity loss. 4 Soil microbial communities may become tolerant to NPs contamination and may be resilient to this disturbance. If resilience is observed, the impact of NPs will be on relatively short-term duration. If no or limited resilience is observed, NPs could have long-lasting effects on soil microbial communities with cascading effects on ecosystem functioning, especially biogeochemical cycles, soil fertility, and climate regulation

Full size image

From a more realistic point of view, chronic contamination of NPs should be simulated. To date, no study using chronic addition of NPs is available. It is likely that NPs are added to soil chronically through multiple applications of sewage sludge over time. Microbial communities are likely to respond differently to chronic and acute exposures. Acute exposures may lead to more resistant microbial communities to NP perturbation if sensitive communities are replaced by resistant communities. On the contrary, chronic exposures may have more detrimental effects if microorganisms are not resilient to such repeated contaminations. Studying if resilience of soil microbial community occurs after acute or chronic contaminations with NPs is crucial to evaluate possible consequences on soil functioning over time (Fig. 2).

Effect of mixture toxicity on soil microbial communities

There is a growing interest in the impact of mixture toxicity in soil since most pollutants share some identical routes of entry in soil (wastewater irrigation, sewage sludge application…) and occur simultaneously in polluted sites. With the increasing use of manufactured NPs, it is necessary that NPs are considered in mixture toxicity studies along with heavy metals, hydrocarbons, or antibiotics. Moreover, it is even more crucial because some NPs are ableto mobilize some pollutants. In a context of remediation, this ability may be beneficial, but when not controlled, NPs may enhance the bioavailability of other pollutants and caused detrimental effects.

Assessing the impact of NPs in soil exhibiting different physicochemical properties

Soil is a complex matrix and extrapolation of results from one contaminated soil to another is difficult because of the great variability in soil composition and structure (Ranjard et al. 1997). However, most experiments were conducted on a single model soil (Table 1), which does not permit a comparison of soil sensitivity to NPs. Soil properties, such as pH, texture, or organic matter content, influence microbial community composition (Fierer and Jackson 2006) and bioavailability of pollutants (Giller et al. 1998). Thus, the knowledge of the soil physicochemical properties influencing NPs bioavailability (Fig. 2) will greatly enhance our understanding of NPs impact on soil functioning. In addition, we have little information on the spatialization and mobility of NPs in soil (Fang et al. 2009; Vittori Antisari et al. 2013), but like heavy metals, hot spots of pollutants may occur (Ranjard and Richaume 2001). Microscale approaches are needed to determine if NPs have a heterogeneous impact in soil. Moreover, the use of microbial indicators may be useful to have better insights on NPs spatialization to circumvent the current lack of techniques to characterize NPs in situ (Tables 2 and 3).

Table 2 Review of the effects of carbon nanoparticles on soil microbial communities
Full size table
Table 3 Review of the effects of nanoscale zero valent iron nanoparticles on soil microbial communities
Full size table

Assessing fungal and archaeal response to NPs

More attention is classically given to soil bacterial communities, despite the crucial role of fungal communities in energy flow and nutrient transfer in terrestrial ecosystems. New insights regarding the broad distribution and abundance of archaea in soils and oceans imply that they also contribute to global energy cycles, especially nitrogen cycle where ammonia-oxidizing archaea play a key role in nitrification (Schleper et al. 2005; Robertson et al. 2005; Zhang et al. 2010). What is known on NPs toxicity from in vitro studies using bacterial strains may not be extrapolated to archaeal and fungal communities. Thus, it is important to consider the response of archaeal and fungal communities to NPs contamination, in order to have a better assessment of NPs ecotoxicity in soils.