Can photocatalytic and magnetic nanoparticles be a threat to aquatic detrital food

23 Freshwaters are likely to serve as reservoirs for engineered nanomaterials (ENMs) due to 24 their accelerated production and usage, increasing the relevance of assessing their impacts on 25 aquatic biota and the ecosystem processes they drive. Stream-dwelling microbes, particularly 26 fungi, and invertebrate shredders play an essential role in the decomposition of organic matter 27 and transfer of energy to higher trophic levels. We assessed the impacts of two photocatalytic 28 (nano-TiO 2 and nano-Er:TiO 2 ) and one magnetic (nano-CoFe 2 O 4 ) ENMs on detrital-based 29 food webs in freshwaters by exposing chestnut leaves, colonized by stream-dwelling 30 microbes, to a series of concentrations (0.25–150 mg L –1 ) of these ENMs. Microbial 31 decomposition and biomass of fungal communities, associated with leaves, were not affected 32 by the ENMs. However, the activities of antioxidant enzymes of microbial decomposers were 33 stimulated by ENMs in a concentration-dependent way, suggesting oxidative stress in stream 34 microbial communities. The stronger responses of these stress biomarkers against nano-TiO 2 35 suggest a higher toxicity of this ENM comparing to the others. To determine whether the 36 effects could be transferred across trophic levels, the invertebrate shredder Sericostoma sp. 37 was exposed to ENMs (1 and 50 mg L –1 ) for 5 days either via contaminated water or 38 contaminated food (leaf litter). Leaf consumption rate by shredders decreased with increasing 39 concentrations of ENMs via food or water; the effects were more pronounced when exposure 40 occurred via contaminated food. Overall, the tested photocatalytic and magnetic ENMs can 41 be harmful to microbes and invertebrates that drive detrital food webs in streams at predicted 42 environmentally relevant concentrations. 43


INTRODUCTION
Recent developments in nanotechnology led to an increased worldwide production and application of engineered nanomaterials (ENMs) (Stark et al. 2015).The TiO2 nanoparticles (nano-TiO2) are among the most extensively used ENMs with a wide range of applications as in supercapacitors, photocatalysis, sensors, personal care products, biomedicine, dyesensitized solar cells, lithium batteries, paints and food products (Chen and Mao 2007;Weir et al. 2012;Tian et al. 2014).In July 2016, the European Commission allowed the application of nano-TiO2 as a UV-filter in sunscreens at a concentration up to 25% (European Commission, 2016), which may further enhance the commercial use of these ENMs in Europe.The estimated global production of nano-TiO2 was about 10 4 tonnes per year, and might even be higher in Europe (Piccinno et al. 2012).
Nanoparticles of TiO2 are often applied in wastewater effluent treatments and chlorine-free disinfection due to their photocatalytic properties (Rickerby 2014).However, the nonporous structure of the bare nano-TiO2 and their aggregation capacity in water may limit the photocatalytic and adsorption of organic contaminants in aquatic environments.Nano-TiO2, doped with rare earth metals, like erbium (nano-Er:TiO2), can enhance the photocatalytic performance because of the vacant f-orbitals of Er 3+ that allow intermediate energy states (reducing the band gap), improving the adsorption of various molecules (e.g.amines, alcohols, aldehydes, amines thiols) from contaminants onto the nanoparticle surface (Gomez et al. 2012;Martins et al. 2014).On the other hand, due to suitable physicochemical and magnetic properties, cobalt-ferrite nanoparticles (nano-CoFe2O4) undergo increasing applications in biomedical engineering, including drug delivery, magnetic separation and purification, biosensor, magnetic resonance imaging, cancer therapy and hyperthermia (Cardoso et al. 2018;Srinivasan et al. 2018).Nanoparticles of CoFe2O4 have potential to remove anionic dyes (Yavari et al. 2016) and to treat metal-rich industrial effluents or wastewaters (Srivastava et al. 2016).
Due to the vast applications and use of these ENMs, they are likely to be present in significant amounts in aquatic environments.Indeed, nanoparticles of TiO2 were detected in groundwater and drinking water as a consequence of their release from house facades into the nearby stream or from urban runoffs (Kaegi et al. 2008;Kiser et al. 2009;Westerhoff et al. 2011).Adverse effects of TiO2 on aquatic organisms including bacteria, microalgae, invertebrates and vertebrates have been reported (Federici et al. 2007;Li et al. 2014b;Schaumann et al. 2015;Girardello et al. 2016).Although bare or Er-doped nano-TiO2 are expected to be biocompatible to humans (Martins et al. 2014;European Commission 2016), the impacts of nano-Er:TiO2 on aquatic organisms are unknown.On the other hand, the few ecotoxicological studies with nano-CoFe2O4 showed toxicity against plant-pathogenic fungi (Sharma et al. 2017) and to freshwater algae and fish (Ahmad et al. 2015a;Ahmad et al. 2015b).
In forest streams, plant litter breakdown is a key ecosystem process driven by microbes, predominantly fungi, and invertebrate shredders that transfer nutrients and energy from plant litter of riparian trees to higher trophic levels (Graça 2001).Invertebrate shredders generally prefer to feed on leaf litter colonized by stream microbial communities because microbial activities and biomass improve leaf litter quality and its palatability (Graça 2001).However, the knowledge on the impacts of nano-TiO2, nano-Er:TiO2 and nano-CoFe2O4 on detritusbased food webs is lacking.
The current study aims to evaluate the effects of nano-TiO2, nano-Er:TiO2 and nano-CoFe2O4 on stream microbial decomposer communities and invertebrate shredders using the detrital model system, which has proven sensitive to various contaminants (Pradhan et al. 2011;Pradhan et al. 2015a;Tlili et al. 2016).We hypothesized that ENMs would i) reduce microbial decomposition and the biomass of leaf-associated fungi; ii) induce oxidative stress in microbial decomposer communities; and iii) decrease leaf litter consumption by invertebrate shredders, mainly when animals were exposed via contaminated food.
The morphology of the primary particles was monitored by transmission electron microscopy (TEM, Tecnai T20, FEI).The ENMs were sonicated for 5 min to achieve a homogeneous dispersion; a drop of the solution was placed on a copper grid and dried at room temperature (RT).The crystallinity of the ENMs was assessed by X-ray powder diffraction (XRD) with Philips X'Pert instrument equipped with Cu Kα radiation (λ = 1.54178Å) at 40 kV/50 mA.
The hydrodynamic diameter and zeta potential (ζ) of the ENMs were determined using Zetasizer (NANO ZS-ZEN3600, Malvern Instruments Limited, UK), in backscatter mode (173°).The analyses were performed at 25°C by dispersing 10 mg of ENMs in 100 mL of ultra-pure water (to avoid multicasting).The suspension was sonicated for 30 min, and aliquots were used to estimate the mean hydrodynamic diameter from the intensity-weighted distributions (Zeta-average), as well as the polydispersity index (PdI), and Zeta-potential values (Zetasizer 6.20 software).

Stream microbial colonization on leaves
Leaves of Castanea sativa (L.) (chestnut) were collected during autumn and air-dried at RT. Chestnut is one of the dominant riparian plant species in Northwest Portugal.The chestnut leaves were cut into discs (12-mm) and placed into fine-mesh (0.5-mm) bags (to minimize the access of benthic macroinvertebrates), and immersed for 12 days in Algeriz Stream (41°35'24.56"N,8°22'36.96"W) to allow colonization by stream-dwelling microbes.The stream was situated in a low populated area.At the sampling site, the width and depth of the stream were 0.5-0.8m and 0.3-0.4m, respectively; the geological substratum was composed mostly of sand and pebbles.

Exposure in microcosms
After 12 days, leaf bags were retrieved from the stream and taken to the laboratory where the leaf discs were carefully washed and allocated to 150-mL Erlenmeyer flasks (microcosms).

Loss of leaf mass
The mass loss of chestnut leaves was estimated by weighing (up to 0.001 mg) lyophilized (Christ alpha 2-4, B. Braun, Germany) leaf discs before and after the microbial colonization in Algeriz Stream, and after the microcosm experiment.Initial leaf mass was determined by immersing 3 leaf bags in the stream for 30 min, and the leaf discs were subsequently lyophilized and weighed.

Fungal biomass
To determine fungal biomass, ergosterol, a sterol present in fungal cell membranes, was quantified by ultra-high-performance liquid chromatography (UltiMate 3000, Thermo Scientific UHPLC system) using a LiChrospher 100 RP18 (5 μm) column (Merck) in 6 lyophilized chestnut leaf discs per replicate.Lipid extraction was carried out from the chestnut leaf discs by heating (80°C, 45 min) in KOH-methanol (0.8%), before purified by solid-phase extraction and eluted in isopropanol (Sigma-Aldrich, analytical grade).Ergosterol peaks were monitored at 282 nm and eluted (at 1.4 mL min −1 ) with methanol (Sigma-Aldrich, HPLC-grade).The concentrations of ergosterol from the samples were computed using a standard curve (Sigma-Aldrich) in isopropanol.The extracted ergosterol was converted to fungal biomass considering the factor of 5.5 μg of ergosterol per mg dry biomass (Gessner and Chauvet 1993).

Activities of antioxidant enzymes
For determining the activities of antioxidant enzymes (glutathione peroxidase: GPx, glutathione S-transferase: GST, and catalase: CAT) in microbial communities on chestnut leaves, 15 leaf discs from each microcosm were retrieved, washed thrice with ultrapure water, and frozen in liquid nitrogen (to prevent biological activities).Leaf discs were homogenised (Utratratrax T 25, IKA, Staufen, Germany) using potassium phosphate (K-phosphate, 0.1 M, pH 7.4) buffer (1:10 w:v) and PMSF (phenylmethylsulfonyl fluoride as protease inhibitor, 1 mM) at 4°C.The leaf homogenates were centrifuged (10,000 × g, 20 min, 4°C) and the supernatants (cell-free extract: CFE) were separated and frozen at -80°C in several aliquots till the measurement of the activities of antioxidant enzymes.
Protein concentration was measured in the CFE according to Bradford (1976) in 96-well flatbottomed microplates and expressed per unit mass of leaves.The activities of the antioxidant enzymes were measured in CFE using a spectrophotometer (SpectraMax Plus 384 Microplate Reader, Molecular Devices) and normalized to the protein concentration.The activity of GST was determined by measuring the formation of 1-glutathione-2,4-dinitrobenzene resulting from the conjugation of GSH with the substrate 1-chloro-2,4-dinitrobenzene (CDNB) (Habig et al.;Barros et al. 2019a).The cell-free extract was added to the reaction mixture (1:3 v:v) containing K-phosphate (0.1 M, pH 6.5) buffer, GSH (1.5 mM) and CDNB (1.5 mM).The GST activity was computed from the slope of absorbance curve (at 340 nm, ԑ = 9.6 mM -1 cm -1 ).
When GR reduced the GSSG (oxidized glutathione) to GSH, the oxidation of NADPH was monitored from absorbance (at 340 nm, ԑ = 6.2 mM -1 cm -1 ) and the GPx activity was computed from the slope (Flohé and Günzler 1984;Barros et al. 2019a).

Invertebrate collection and exposure to nanoparticles
Sericostoma sp.(Latreville) is an invertebrate shredder (Trichoptera, Sericostomatidae) common in low-order streams in Southwest Europe (Bonada et al. 2008;Varandas and Cortes 2010) with good water quality.Early-stage larvae (1.1 ± 0.1 cm) of Sericostoma sp. were collected in upstream of the Cávado River (Northwest Portugal) and brought to the laboratory in a cold box.Shredders were placed in aquaria with mineral water (Fastio®, Gerês Mountain, Portugal) and sterilized (121°C, 20 min) sand and maintained under aeration at 16°C, with a photoperiod (12h/12h: light/dark).Shredders were allowed to feed on chestnut leaves for 28 days before the feeding experiment.To assess the potential effects of nano-TiO2, nano-Er:TiO2 and nano-CoFe2O4, the shredders were exposed to contaminated water or contaminated chestnut leaves for 5 days in microcosms.For exposure via water, the microcosms with mineral water (Fastio®) were supplemented with nano-TiO2, nano-Er:TiO2 or nano-CoFe2O4 at 1 mg L -1 or 50 mg L -1 and shredders were allowed to feed on microbially-colonized leaf discs not exposed to ENMs.For exposure via food (leaves), the microcosms were supplemented with mineral water (Fastio®) and shredders were allowed to feed on microbially-colonized leaf discs previously exposed (for 21 days) to the same concentrations of ENMs (see section 2.3).Same number of pre-exposed or unexposed microbially-colonized leaf discs enclosed in fine-mesh bags was also placed in each microcosm of the respective treatment to determine the contribution of stream-dwelling microorganisms to leaf litter breakdown during 5 days.

Rate of invertebrate feeding
The feeding rate of the shredders on chestnut leaves was determined as Fe / (Sf × t), in which Fe is the leaf consumption by shredders; Sf is the dry mass of shredders at time t (5 days).The leaf consumption by shredders was calculated as Fe = (Fi -Ff) -(Fi × (Di -Df) / Di), where Fi and Ff are the initial and final dry mass of the microbially-colonized chestnut leaves provided to shredders; and Di and Df are the initial and the final dry mass of microbiallycolonized chestnut leaves inaccessible to shredders (Pradhan et al. 2015a).

Statistical analyses
Two-way ANOVAs were applied to evaluate the effects of the concentration (0.25, 1, 10, 50 and 150 mg L -1 ) and type of ENMs (nano-TiO2, nano-Er:TiO2 or nano-CoFe2O4) on leaf mass loss, fungal biomass and activities of antioxidant enzymes of microbial communities on leaves.Two-way ANOVAs were also applied to analyse the effects of the concentration (1 and 50 mg L -1 ) and type of ENMs on the feeding rate of the shredders.ANOVAs were followed by Tukey's multiple comparisons post-hoc tests.The analyses were performed with Prism 7.0 (GraphPad software Inc., San Diego, CA, USA).

Effects of ENMs on decomposition of chestnut leaves and fungal biomass
After 3 weeks of exposure, the concentration and type of ENMs did not show any significant effect on the leaf mass loss driven by microbes, as the remaining mass under treatments did not differ from control (two-way ANOVA, P>0.05) (Fig. 3A).Also, fungal biomass was not significantly affected by the concentration or type of ENMs (two-way ANOVA, P>0.05) (Fig. 3B).
In the control, the activity of GST in microbial communities was 11.2 nmol min -1 mg -1 protein.GST activity was significantly stimulated by increased concentration and varied with the type of ENMs (two-way ANOVA, P<0.0001; Fig. 4C).The activity of GST increased in a dose-dependent manner under exposure to all ENMs (P<0.05),except at the lowest concentration of nano-Er:TiO2 (P>0.05).The maximum increase in GST activity was observed upon exposure to the highest concentration of nano-TiO2 (1154.6%),followed by nano-CoFe2O4 (814.6%) and nano-Er:TiO2 (539.9%) (Fig. 4C).

Effects of ENMs on the feeding rate of invertebrate shredders
After 5 days, in the absence of ENMs, the feeding rate of invertebrate shredders on microbially-colonized leaves was 0.15 mg leaf mass mg -1 animal mass day -1 (Fig. 5).The shredder feeding rate was affected significantly by the concentration of ENMs (two-way ANOVA, P<0.0001) irrespective of their type (P>0.05),when animals were fed on contaminated leaves (via food,Fig 5A).The exposure to 1 mg L -1 of nano-Er:TiO2, nano-TiO2 and nano-CoFe2O4 via food led to 86.3%, 88.8% and 89.3% inhibition (P<0.05) in the feeding rates, respectively.The exposure at 50 mg L -1 also led to a severe inhibition (P<0.05) in the feeding rate by nano-Er:TiO2 (99.3%), followed by nano-CoFe2O4 (90.7%) and nano-TiO2 (90.3%) (Fig. 5A).When exposure occurred via contaminated water, the feeding rate of shredders was affected by the concentration of ENMs (two-way ANOVA, P<0.0005), but not by the ENM type (P>0.05) (Fig. 5B).The waterborne exposure to ENMs at 1 mg L -1 led to a significant decrease (P<0.05) in the feeding rate of shredders (up to 77.8% for nano-Er:TiO2), and the inhibition was maximum when shredders were exposed to 50 mg L -1 of nano-CoFe2O4 (84%) (Fig. 5B).

DISCUSSION
Our study showed for the first time that photocatalytic and magnetic ENMs can affect key players involved in organic matter breakdown in streams, such as the microbial decomposers of plant litter and the invertebrate shredders.Stimulation of antioxidant enzymatic activities in microbial communities was found, but the effects depended on the dose and type of the ENMs.However, unlike hypothesized, the biomass of fungal communities and leaf litter decomposition driven by microbes were not affected by ENMs.The absence of effects on the biomass of fungal communities was found earlier after short-and long-term exposure to metals (Duarte et al. 2004;Duarte et al. 2009) or nanometals (Pradhan et al. 2011).These results might be the consequence of i) triggering physiological acclimation mechanisms in fungi, ii) decreasing the direct contact between fungal mycelia and ENMs with the plant litter tissues acting as a physical barrier, and/or iii) shifting towards a better adapted microbial community (Fernandes et al. 2009;Pradhan et al. 2014).In our study, the absence of effects of ENMs on microbial decomposition might be explained by the non-effects on fungal biomass since fungi are considered the major microbial decomposers of plant litter (Graça 2001;Pascoal and Cássio 2004).
Despite the minimal effects of nano-TiO2, nano-Er:TiO2 and nano-CoFe2O4 on fungal biomass or microbial decomposition, our study clearly unravelled sublethal effects of these ENMs on microbial decomposers of plant litter.The strong responses of enzymatic stress biomarkers in microbial communities suggest that they were under oxidative stress.The activities of antioxidant enzymes from the ascorbate-glutathione cycle play active role in cellular defense against reactive oxygen species (ROS), preventing the cellular damage and maintaining the cellular redox homeostasis (Ayer et al. 2014); hence our results reinforce the role of these enzymes as early warning biomarkers of oxidative stress induced by EMNs (nano-CuO: Pradhan et el.2015b; nano-Ag: Barros et al. 2019a;Barros et al. 2019b).In our study, metal oxide nanoparticles significantly induced the activities of CAT, GPx and GST in microbial decomposer communities at ≥ 0.5 mg L -1 .Microbial decomposers exposed to nano-TiO2 exhibited the highest enzymatic activities, suggesting intense oxidative stress.
Negative impacts of nano-TiO2 on freshwater planktonic and biofilm communities were associated with increased activities of stress biomarkers, and damages in cell-membrane and DNA due to intracellular accumulation of ROS under light (Battin et al. 2009;Wang et al. 2019).ROS can be generated from the surface of the photoexcited nano-TiO2 (Li et al. 2014a).However, that was not the case in our study as we clearly showed that the induced stress by nano-TiO2 to microbial decomposers occurred in the dark, without any photocatalytic interference.Also, nano-TiO2 was able to induce oxidative stress and lipid peroxidation in bacteria in the absence of light (Kumari et al. 2014;Erdem et al. 2015).
In the present study, nano-CoFe2O4 and nano-Er:TiO2 also increased enzymatic biomarker activities in microbial decomposers, denoting oxidative stress.The information on the behaviour of nano-CoFe2O4 and nano-Er:TiO2 in freshwater environments is scarce.
However, the adsorption of nano-CoFe2O4 to the microalgae Chlorella vulgaris caused severe oxidative damage through the production of intracellular ROS leading to accelerated lipid peroxidation and increased activities of CAT and GST (Ahmad et al. 2015b).In our study, the lowest oxidative stress was induced by Er-doped nano-TiO2 as indicated by the level of biomarker activities in freshwater microbes.However, the nano-Er:TiO2 is likely to perform higher photocatalytic activity than non-doped nano-TiO2 (Martins et al. 2014), which in turn may cause severe oxidative damage in the presence of light.
In our study, metal ions released from the surface of ENMs might have played a role in inducing oxidative stress; however, the underlying mechanisms in the absence of light are not clear.Depending on the environmental conditions, nano-TiO2 can release Ti 4+ ions from the nanoparticle surface.In fact, enhanced attachment of nano-TiO2 to the microbial cell surface may occur in dark (Dalai et al. 2012) which may lead to the release of Ti 4+ from nanoparticles outside the cells (Dasari and Hwang 2013).In our study, physicochemical characterization (based on TEM and XRD) showed that the primary particles of nano-Er:TiO2 and nano-TiO2 were smaller than nano-CoFe2O4; whereas the hydrodynamic size, PdI and zeta potential data indicated relatively lower agglomeration and higher dispersity and stability of nano-TiO2 in suspensions, explaining the strongest effects of these nanoparticles among all tested ENMs.In our study, the possible action mechanisms of nano-TiO2 might have involved the following steps: i) interaction and adsorption of nanoparticles to microbes, ii) release of Ti 4+ ions from surface of the outer membrane-localized nanoparticles and internalization of the ions by the cells, iii) partial internalization of the nanoparticles, iv) release of Ti 4+ ions in acidic condition of the lysosome-like organelles, and v) reduction of Ti 4+ to Ti 3+ by peroxides via pseudo-Fenton-type reaction and reoxidation (Ti 4+ + H2O2 → Ti 3+ + OH − + •OH; Ti 3+ + O2 → Ti 4+ + •O2 − ), resulting in ROS generation that induced oxidative stress (Dodd and Jha 2011;Dalai et al. 2012;Pradhan et al. 2015b;Liu et al. 2017).Similar mechanisms are expected for nano-Er:TiO2; but their relatively lesser stability and higher agglomeration compared to the bare nano-TiO2 might have contributed to induce less oxidative stress in microbial decomposer communities.Moreover, the doping with Er might have decreased the surface release of Ti 4+ ions.On the other hand, relatively greater primary particle size and higher agglomeration of nano-CoFe2O4 might have led to the less negative effects of these nanoparticles.These magnetic nanoparticles might have been attached to microbial cells, and Co 2+ and Fe 3+ ions released from the surface of the nanoparticles could be internalized by the cells where the ions might have undergone pseudo-Fenton-type reactions to generate ROS and induce oxidative stress (Novak et al. 2013;Ahmad et al. 2015b;Pradhan et al. 2015b).Co 2+ ions appeared to be more toxic than nano-CoFe2O4, and intracellular accumulation of Co 2+ have been shown while nano-CoFe2O4 were not retained in vivo (Novak et al. 2013).
Our results also showed that photocatalytic and magnetic ENMs can affect stream invertebrate shredder performances.Negative effects of nano-TiO2 on freshwater invertebrates were reported earlier (Menard et al. 2011;Girardello et al. 2016).Changes in the feeding activity of invertebrates may have dramatic ecological consequences and have often been used to assess sublethal effects of nano-metal oxides (Buffet et al. 2011;Pradhan et al. 2012;Pradhan et al. 2015a).In the present study, the feeding rate of Sericostoma sp. on microbially colonized leaves in the absence of ENMs was within the conventional range (0.04-0.5 mg leaf mass mg -1 animal mass day -1 ) documented for invertebrate shredders in streams (Arsuffi and Suberkropp 1989).The feeding rate decreased significantly upon exposure to all ENMs, even at the lowest concentration (1 mg L -1 ) via contaminated food or water.The lowest observed effect concentration on shredder feeding rate in our study was similar to the hazard concentration (HC50: 1.1 mg L -1 ) of nano-TiO2 estimated for freshwater secondary consumers, predominantly invertebrates (Semenzin et al. 2015).
The reduced feeding rate of the shredders probably resulted from the food avoidance behaviour (Wilding and Maltby 2006;Pradhan et al. 2012;Pradhan et al. 2015a).In our study, the effects of ENMs on feeding rate via contaminated leaves were more pronounced than via contaminated water, which was probably due to the decreased quality and palatability of the chestnut leaves after 21 days of exposure to the ENMs.The exposure to ENMs might have led to high adsorption and accumulation of metals and/or nanoparticles to leaves (Pradhan et al. 2012) and aquatic fungi (Barros et al. 2019b).Indeed, an earlier study on trophic transfer of nano-TiO2 in freshwaters demonstrated that, comparing to aqueous exposure, the dietary intake could constitute the main route of ENM exposure to higher trophic levels (Zhu et al. 2010).In addition to the decrease in food quality, the aqueous or dietary exposure of shredders to ENMs probably led to their accumulation in the gut, inducing oxidative stress to the invertebrate shredders (Pradhan et al. 2015a;Girardello et al. 2016).
In our study, the adverse effects of ENMs on microbial decomposers and invertebrate shredders in stream detrital food web were observed even at concentrations predicted to be environmentally relevant (Gottschalk et al. 2013;Xia et al. 2017).On the other hand, the effects of ENMs at higher concentrations may mimic the conditions of wastewaters, mine-drainage streams or accidental spills and, therefore, are also relevant to be considered for environmental safety.

CONCLUSIONS
Overall, the responses of enzymatic biomarkers revealed that nano-TiO2, nano-Er:TiO2, and nano-CoFe2O4 induced oxidative stress in microbial decomposer communities involved in the decomposition of plant litter in streams.The effects increased in a dose-dependent manner for all ENMs, although the effects of nano-TiO2 were the most pronounced.All three ENMs were able to decrease the feeding rate of the invertebrate shredder Sericostoma sp.via aqueous and dietary exposure.The effects on the feeding rate were stronger when the shredders were exposed to ENMs via contaminated food (leaves).To our knowledge, our study is the first to show the harmful effects of erbium-doped nano-TiO2 and nano-CoFe2O4 on microbial decomposer communities and invertebrate shredders with a key role in detrital food webs in streams.Our study also provided evidence that photocatalytic and magnetic ENMs can induce negative effects even in the absence of light at predicted environmentally relevant concentrations.These findings pinpoint that stream detrital food webs may have potential for ecological risk assessment of emergent contaminants in complex realistic environments.

Figure 3
Figure 3 Dry mass of decomposing chestnut leaves after exposure (21 days) to different

Figure 4
Figure 4 Activities of CAT (A), GPx (B) and GST (C) in microbial decomposer communities

Figure 5
Figure 5Feeding rate of the stream invertebrate shredder Sericostoma sp. after exposure (5