Effects of silver nanoparticles and silver nitrate in the earthworm reproduction test



The widespread use of silver nanoparticles (Ag-NPs), for example, in textiles and cleaning products, means that they are likely to reach the environment via biosolids or the effluent from wastewater treatment plants. The aim of the present study was to determine the ecotoxicity of Ag-NPs in the earthworm reproduction test using Eisenia andrei. In addition to the usual endpoints, the authors also investigated the uptake and accumulation of Ag by adult earthworms and the concentration of free Ag+ in soil pore water. Silver nanoparticles and Ag nitrate showed similar toxicities in the earthworm reproduction test. The uptake of Ag from Ag-NPs in the earthworm was slightly higher than the uptake of Ag from Ag nitrate. Spiked soils showed a concentration-dependent effect on reproduction, but there was no concentration-dependent increase in the amount of Ag in earthworm tissues. The authors noted a concentration-dependent increase in the levels of free Ag+ in the soil pore water regardless of the Ag source. The number of juveniles is a more suitable endpoint than biomass or mortality. The uptake of Ag does not appear to inhibit reproduction. Instead, inhibition seems to reflect Ag+ released into the soil pore water, which affects cocoons and juveniles in the soil. Analysis of transformed Ag-NPs after purification in wastewater treatment plants would provide additional information. Environ. Toxicol. Chem. 2013;32:181–188. © 2012 SETAC


Silver nanoparticles (Ag-NPs) are advantageous in textiles, microelectronics, inks, medical imaging reagents, and cleaning products 1, 2. According to the Inventory of Nanotechnology-based Consumer Products (Project on Emerging Nanotechnologies), Ag nanotechnology is currently present in more than 313 commercial products. The predominant commercial application of Ag-NPs is in the health and fitness sector, which includes personal care products, clothing, cosmetics, sporting goods, filtration units, and sunscreens (Woodrow Wilson International Center for Scholars, 2011). The rising use of engineered NPs (especially Ag-NPs) increases the potential for environmental contamination and adverse effects.

Large numbers of studies concerning the effects of Ag-NPs in the aquatic environment are already available. Calculations of predicted environmental concentrations (PECs) based on probabilistic material flow analysis from a life-cycle perspective of different engineered NPs revealed that Ag-NP from sewage treatment effluents and surface waters may increase risks to aquatic organisms 3. Studies with various aquatic organisms have demonstrated the hazard of Ag-NPs, which can induce respiratory stress in the Eurasian perch 4, phytotoxicity in Lemna minor 5, and hepatocyte damage in rainbow trout (Oncorhynchus mykiss) 6. Many other studies have considered the fate and behavior of Ag-NPs released from textiles and from nanosilver-coated washing machines into wastewater treatment plants (WWTPs). Physical separation and ion-selective electrode (ISE) analysis suggest that both colloidal and ionic Ag leach from socks 7. The amounts depend on the total Ag content in the products and the stability of the linkage between the silver particles and the product matrix 8. Significant amounts of Ag are released into the environment from nanosilver-coated washing machines and finally reach the effluent of WWTPs 9. A mass balance analysis of Ag in a WWTP indicated that more than 95% of the incoming Ag was sequestered into the wastewater biomass 10. Transmission electron microscopy (TEM) analysis confirmed that nanoscale Ag particles were adsorbed to wastewater biosolids. X-ray absorption spectroscopy (XAS) measurements indicated that most Ag in sludge and effluent was present as Ag2S when Ag-NPs were added to the nonaerated tank of a pilot plant. The addition of Ag-NPs to aerated mixed liquor increased the stability of Ag-NPs 11. Silver from Ag-NPs is therefore likely to reach terrestrial ecosystems via activated sludge when it is applied as agricultural fertilizer.

Despite the large number of studies concerning the influence of Ag-NPs in the aquatic environment, few have considered their impact on terrestrial ecosystems. Engineered NPs may enter soil via biosolids originating from wastewater treatment or the effluent from manufacturing processes and can inhibit organisms in the terrestrial ecosystem 12. There is a growing concern that restrictions on the use of biosolids as agricultural fertilizers may have to be implemented to account for the presence of Ag-NPs 7. In terrestrial ecosystems, Ag-NPs may have adverse effects on soil microflora 13, 14 and (in a limit test) on earthworms 15. Further studies have demonstrated the effect of Ag-NP surface coatings on bioaccumulation and earthworm reproduction 16, and the role of particle size and soil type on the toxicity of Ag-NPs to the earthworm Eisenia fetida 17.

To gain more insight into the effects of Ag-NPs in the terrestrial environment, we used the earthworm reproduction test according to Organisation for Economic Co-operation and Development (OECD) guideline 222 18 and exposed the earthworms to Ag-NPs by spiking the soil. Several aspects were considered to develop a more comprehensive understanding of the ecotoxicity of Ag-NPs, specifically Ag-NP NM-300K from the OECD Sponsorship Program. Silver nitrate was used as a reference material in all tests, because this should indicate where the effects of Ag are caused by particles or ions released from particles.

The first objective was to provide information about the ecotoxicity of Ag-NPs in the earthworm reproduction test according to OECD guideline 222 18 by addressing the following endpoints: mortality, biomass increase, and number of offspring. In addition to the usual endpoints in the standardized reproduction test, we also investigated the uptake and bioaccumulation of Ag in adult earthworms. According to the guideline, adult earthworms were removed after 28 d and used to determine the concentration of Ag after their gut had been purged. It was important to determine the toxic impact of the particles and of released silver ions (Ag+) in comparison with a pure Ag+ source (Ag nitrate), so the amount of Ag+ in pore water was determined with diffusive gradients in thin films (DGT) in addition to the total Ag concentration.


Test soil

The test and carrier soil for the application was RefeSol 01A, a loamy, medium-acidic, and lightly humic sand (pH 5.67; Corg 0.93%, sand 71%, silt 24%, clay 5%), which has been described in detail elsewhere 19. RefeSol 01A matches the properties as stated in various OECD terrestrial ecotoxicological guidelines (e.g., tests with plants and soil microflora). The soils were sampled in the field and were stored and treated according to good agricultural practice. They were stored outdoors in stainless-steel containers, red clover was sown on the stored soils, and no pesticides were used. Appropriate amounts of soil were sampled from the outdoor boxes one to four weeks before the test. If the soil was too wet for sieving, it was dried at room temperature to 20 to 30% of the maximum water-holding capacity (WHCmax), with periodic turning to avoid surface drying. If the tests did not start immediately after sieving, the soil was stored in the dark at 4°C under aerobic conditions 20.

Materials and methods

We used the Ag-NP NM-300K from the OECD Sponsorship Program. This is from the same initial material batch of NM-300, which is one of the best-characterized nanomaterials from the official NM materials. It is a colloidal Ag dispersion with a nominal Ag content of 10% (w/w) and a particle size of approximately 15 nm, with a narrow size distribution (99%). A second particle size of 5 nm, which is much less abundant (1%), was identified by TEM. The NM-300K is a mixture of a stabilizing agent (NM-300K DIS) comprising 4% (w/w) each of polyoxyethylene glycerol trioleate and polyoxyethylene sorbitan monolaurate (Tween-20) and the Ag-NPs. The release of ions from NM-300K particles into the matrix under storage conditions was estimated to be less than 0.01% (w/w) 21. The Ag nitrate was purchased from Merck KGaA.

Test organism

We used Eisenia andrei, which has been cultured in our laboratory for more than 15 years. The earthworms were bred under defined conditions (1:1 mixture of cow manure and sphagnum peat at 20 ± 2°C and a light:dark cycle of 16:8 h). Three days prior to the test, the earthworms were adapted to the test soil under experimental conditions (WHCmax = 55%), with free access to food. In the test, we only used earthworms with a clitellum and a wet mass of 300 to 500 mg.

Application and test concentration

Silver nanoparticles were applied by mixing the test material and air-dried carrier soil with the same physicochemical properties as the test soil. Enough Ag-NPs were added to the carrier to achieve the final test concentration when 5% carrier soil and 95% test soil were mixed to homogeneity. The soil was mixed with a spoon instead of a pestle to avoid modifying the Ag-NPs. Uncontaminated soil (at 20–30% WHCmax) was spread on a plate, and the spiked carrier soil was evenly distributed over the test soil before mixing. The mixed soil was adjusted to 55% WHCmax using deionized water. This procedure was preformed for every test concentration.

All concentrations refer to the dry matter of soil. The first test was carried out using concentrations of 60, 120, and 200 mg Ag-NPs per kilogram dry matter soil, and the second test was carried out using concentrations of 15, 30, 60, 120, and 200 mg/kg soil. Silver nitrate was applied at concentrations of 15, 30, 60, 120, and 200 mg/kg soil in both tests. In all tests with NM-300 K, the stabilizing agent NM-300 K DIS was used as a dispersant control. The amount of dispersant corresponded to the highest test concentration with Ag-NPs.

Ecotoxicological test

All tests were carried out as described in OECD guideline 222, “Earthworm reproduction test with Eisenia fetida18, which allows the use of E. fetida and E. andrei as test organisms. We filled polypropylene containers (Bellaplast GmbH) to a depth of approximately 5 cm with 625 g dry mass of soil (55% WHCmax) and then spread 40 g wet mass of cow dung (air dried, ground, and moistened before application) onto the surface. The cows were kept in an ethical husbandry. The tests were performed in eight replicates for the negative control and four replicates for the dispersant control and each Ag-NP and Ag nitrate concentration. For chemical analysis with DGT devices, we prepared single analytical test replicates for each concentration to avoid exposing the earthworms to stress.

After overnight equilibration of the treated soils, 10 earthworms were added to each container and were incubated at 20 ± 2°C with a light:dark cycle of 16:8 h (∼700 lux). Once per week, the water content was checked gravimetrically, the evaporated water was replaced, and 20 g wet mass (corresponding to 5 g dry mass) of uncontaminated food was spread onto the soil surface in each container. After 28 d, the adult earthworms were removed and weighed, and after 56 d, the number of juveniles in each test container was counted. Statistical calculations were performed with ToxRat Pro 2.10 software for ecotoxicity response analysis (ToxRat Solutions GmbH). For the median effective concentration (EC50) calculation, we use da probit analysis, assuming log-normal distribution of the values. For each concentration, we determined the percentage of mortality, the percentage of loss or gain in biomass of the adults, and the number of offspring produced in the test.

Silver content in earthworms and soil

Earthworms were incubated for 24 h on wet filter paper to purge their gut and then frozen at −20°C. Homogeneous material was prepared by freezing the earthworms in liquid N, followed by coarse and then fine grinding using mortar and pestles of different grades under a laminar-flow hood, transferring the powder to cryoproof vials stored directly above liquid N (−150°C), and lyophilizing until the samples reached constant weights in a Christ Alpha 1–2 freeze dryer (Martin Christ GmbH). Approximately 200 mg homogenized and lyophilized powder was digested in a quartz vessel with 5 ml concentrated nitric acid, heated in a microwave for 25 min at 220°C, held at 220°C for 30 min, and allowed to cool to room temperature for 1 h. The initial pressure was 40 bar. After digestion, the vessels were filled to an exact volume of 20 ml with ultrapure water. This digestion procedure is in accordance with the guidelines for chemical analysis published by the German Federal Environment Agency (http://www.umweltprobenbank.de/en/documents/publications/15363).

The Ag content of the test soil was measured for representative 15 and 120 mg/kg soil samples. Six samples were obtained from the master batch after the application of Ag-NPs. The digestion procedure was carried out according to DIN ISO 11466 22 and DIN EN 13346 23. Prior to digestion, the soil was dried at 105°C until the weight was constant for at least 12 h, and 3 g of the homogenized material was mixed with 28 g aqua regia and incubated at room temperature for 16 h without agitation. The mixture was then heated under reflux for 2 h, with glass chips and several drops of 1-octanole to prevent overboiling and foaming. The mixture was cooled to room temperature and then carefully made up to 100 ml and filtered prior to analysis (0.45-µm syringe filter, Supor membrane; Pall Corporation).

Silver concentrations in aqueous samples of digested soil and earthworms were measured by inductively coupled plasma–optical emission spectrometry (ICP–OES) using an IRIS Intrepid II ICP-OES (Thermo Electron) at wavelengths of 328.068 and 338.289 nm. The instrument was calibrated before each measurement using blank, 1.0, 2.5, 5.0, 10, 25, 50, 100, or 250 µg/L solutions depending on the concentration range in the samples. The calibration formula was calculated using the linear regression algorithm of the ICP–OES instrument software. Spectral interference was observed at 338.289 nm, so the 328.068-nm readings were used to calculate the concentrations. The correlation coefficient was at least 0.99995. For each sample, at least three measurements were taken, and the mean was determined by the instrument software.

Nitric acid was of Suprapur quality (supplied by Carl Roth). The water was purified using an ELGA Pure Lab Ultra water purification system (purified water resistivity >18 MΩ × cm). A commercially available Ag ICP standard containing 1,000 mg/L Ag in 2 to 3% nitric acid was used to prepare appropriate stock and calibration solutions. All prepared standard solutions had a final HNO3 concentration of 3%. The reference material NIST 2977 Mussel Tissue (info value of 4.58 mg Ag/kg; purchased from LGC Standards) was digested and analyzed along with the earthworm samples to verify the procedures. Furthermore, aqueous certified reference material TMDA-70 (Environment Canada; certified concentration 10.9 µg Ag/L) was analyzed along with all samples. The analytical method was also verified using a multielement Merck IV Standard, appropriately diluted to fit in the range of samples.

The bioaccumulation factors (BAFs) were determined as described in OECD guideline 317 24. Therefore, total Ag concentrations measured in the earthworms were divided by the nominal concentrations of Ag in the soil. Silver concentrations in the worms were compared using the Williams multiple sequential t test procedure (α = 0.05).

Diffusive gradients in thin films

We used DGT devices (obtained from DGT Research Ltd.) comprising a layer of Chelex resin impregnated in a hydrogel, overlain with a diffusive layer of hydrogel and a filter. Ions must diffuse through the filter and diffusive layer to reach the resin layer, where they are collected for downstream analysis and quantitation. The resulting concentration gradient in the diffusive layer allows the quantitation of metal concentrations in solution without the need for calibration (www.dgtresearch.com). The DGT devices were prepared by equilibrating in 0.01 mol/L NaCl with argon purging for 1 h, then for 24 h in a sealed container under argon. Two DGT devices were carefully introduced into the analytical test replicates with the diffusive layer facing upward and covered with soil and were left in situ for 48 h. The DGT application to the soil began on days 0, 26, and 54. After removal, the devices were rinsed thoroughly with ultrapure water and wrapped in polyethylene bags for storage at 4°C for no longer than 4 d. The resin layer was then extracted and directly transferred into 1.5 ml 1 mol/L nitric acid for the elution of Ag+ for at least 24 h. After adjustment to a precise volume with 1 mol/L nitric acid, the Ag content of the solution was determined by ICP-MS using an Agilent 7500ce ICP-MS instrument (Agilent Technologies). Silver was detected at the isotope 109. Calibrations were performed prior to measurement depending on the concentration range in the samples. The calibration formula was calculated using the linear regression algorithm of the ICP-MS instrument software.

The amount of Ag+ in the DGTs was calculated following the manufacturer's recommendations (www.dgtresearch.com) and was expressed as the mass per liter of pore water. Before exposing the DGTs to soil, the dry matter content was determined (79.2–89.3% in all tests), which allowed the amount of soil per liter of pore water to be calculated. These calculations were combined, yielding the amount of Ag+ per kilogram dry matter soil. The percentage from the nominal concentration was taken into consideration to determine whether increasing concentrations of Ag-NPs increase the concentration of free Ag+.


Ecotoxicological test

The reproduction test results are presented in Table 1, whereas Table 2 shows the calculated ECx, lowest observed effect concentration (LOEC), and no observed effect concentration (NOEC) values. The Ag-NPs and silver nitrate were tested twice, and all tests fulfilled test guideline validity criteria, that is: (1) ≥30 juveniles must be produced in each of the replicate negative control vessels by the end of the test; (2) the reproductive coefficient of variation in the negative control vessels must be ≤30%; and (3) adult mortality in the negative control vessels over the initial four weeks of the test must be ≤10%.

Table 1. Results of the earthworm reproduction test with silver nanoparticles and silver nitrate
Concentration (mg/kg soil)aMortality (%)Changes in biomass (%)bMean number of juveniles per test vessel ± SDSD (%)Inhibition (%) to negative controlInhibition (%) to dispersant control
  • a

    Dry matter soil.

  • b

    Values indicate the increase in biomass.

  • *

    0.05 ≥ p ≥ 0.01.

  • **

    0.01 ≥ p ≥ 0.001.

  • ***

    p ≤ 0.001).

  • SD = standard deviation.

Test 1 (Ag-NP)
 0 (negative control)036347 ± 277.8
 0 (dispersant)039329 ± 41.15.2
 60049*186 ± 2211.646.4***43.5***
 120043*147 ± 2416.457.6***55.3***
 200046*86 ± 1821.475.2***73.8***
Test 2 (Ag-NP)
 0 (negative control)040341 ± 267.6
 0 (dispersant)036268 ± 4516.821.5**
 15062*252 ± 3112.326.3***6.2
 30061*209 ± 4622.038.9***22.1**
 60061*220 ± 3013.635.7***18.1*
 120067*158 ± 1811.453.7***41.0***
 200063*97 ± 2626.871.7***64.0***
Test 3 (silver nitrate)
 0 (negative control)036347 ± 277.8
 15057*261 ± 249.424.9*
 30055*202 ± 178.241.8*
 60059*158 ± 2817.954.4*
 120056*107 ± 1615.269.1*
 200054*64 ± 2132.581.5*
Test 4 (silver nitrate)
 0 (negative control)027240 ± 145.7
 15052*183 ± 179.523.9*
 30053*121 ± 1310.549.6*
 607.543*120 ± 75.449.9*
 120057*69 ± 46.071.5*
 20015.035*18 ± 525.392.5*
Table 2. Median effective concentration (EC50), lowest observed effect concentration (LOEC), and no observed effect concentration (NOEC; mg/kg soil) calculations for the reproduction of Eisenia fetida in all tests
 TestLOECNOECEC50 (mg/kg soil)c
  • a

    Calculation in comparison with the negative control.

  • b

    Calculation in comparison with the dispersant control.

  • c

    Dry matter soil.

Silver nanoparticlesTest 1a≤60<6074.3 (—)
Test 1b≤60<6083.0 (—)
Test 2a≤15<1580.0 (33.6–413.3)
Test 2b3015146.0 (85.7–741.4)
Silver nitrateTest 3≤15<1546.9 (40.7–53.6)
Test 4≤15<1542.0 (15.2–83.5)

At the beginning of the test, the earthworms tried to escape from the 120 and 200 mg/kg soil test vessels. However, after 24 h, all earthworms went back into the soil, although some tried to stay in the food that had been spread over the top of the soil. This avoidance behavior was observed throughout the tests with Ag-NPs and Ag nitrate, until the adults were removed. There was no significant mortality in either the Ag-NP or the Ag nitrate treatments.

In both Ag-NP tests, the increase in biomass of the adult earthworms was similar in the negative control and dispersant control (36–40%). In replicates with Ag-NPs, the increase in biomass showed a statistically significant increase over the negative controls (43–49% in the first test and 61–67% in the second test). Similar results were obtained with Ag nitrate (Table 1).

Table 1 shows the number of offspring in the tests with Ag-NPs after 56 d. Reproduction was not inhibited by the dispersant in the first test. The replicates with Ag-NPs showed statistically significant differences from both controls and a concentration–effect relationship. In comparison with the negative control, the inhibition of reproduction ranged from 46.4% at the lowest concentration to 75.2% at the highest concentration, yielding an EC50 value of 74.3 mg/kg soil compared with the negative control and 83.0 mg/kg soil compared with the dispersant control (Table 2).

In the second test with Ag-NPs, the dispersant had an impact on reproduction; that is, fewer juveniles were found in the dispersant control (268) than in the negative control (341), representing a statistically significant 21.5% inhibition of reproduction. The standard deviation ranged from 7.6% in the negative control to 16.8% in the dispersant control (Table 1). Both control values were considered with respect to the effect of Ag-NPs, and inhibition was calculated in comparison with both the negative control and the dispersant control, even in the test in which no significant difference between the controls was seen (Test 1).

Even in the second test with Ag-NPs, we found a concentration–effect relationship and observed strong, statistically significant inhibition of earthworm reproduction. In comparison with the negative control, the inhibition of reproduction ranged from 26.3% at the lowest concentration to 71.7% at the highest concentration, yielding an EC50 of 80 mg/kg soil compared with the negative control, which is close to the EC50 value determined in the first test, and an EC50 of 146.0 mg/kg soil for the dispersant control, which is less pronounced than in the first test (Table 2).

Silver nitrate also caused concentration-dependent effects in both tests and was approximately twice as toxic to earthworms as Ag-NPs. For the first test, we observed a reduction in the number of juveniles compared with negative controls from the lowest concentration (261 juveniles, 24.9% inhibition) to the highest concentration (64 juveniles, 81.5% inhibition; Table 1), reflecting an EC50 value of 46.9 mg/kg soil. The second test showed a similar concentration-dependent effect, with a calculated EC50 value of 42.0 mg/kg soil (Table 2).

Silver content in the soil and earthworms

We analyzed the spiked soil at the beginning of the test and achieved 90% recovery of Ag from both the 15 and the 120 mg/kg soil samples (standard deviation 14 and 4%, respectively, confirming that the Ag is distributed homogeneously at both low and high concentrations). All concentrations of Ag in the earthworms and in soil refer to the dry matter content of the sample.

We then determined the silver concentrations in earthworms (Table 3). No silver was detected in the negative control or dispersant control, but silver was detected in all the worms that had been exposed to Ag-NPs or Ag nitrate. The concentrations of Ag in earthworms exposed to NM-300 K ranged from 7.0 mg/kg worm at 15 mg/kg soil to 11.3 mg/kg worm at 120 mg/kg soil. There was a significant difference in the Ag content of adult worms when we compared the 15 mg/kg soil sample with all the higher concentrations. The concentration of Ag in earthworms exposed to Ag nitrate was 8.0 mg/kg worm at 15 mg/kg soil, and this value declined as the Ag content of the soil increased (Table 3). There was a significant difference in the earthworm Ag content when we compared the 15 mg/kg soil and 200 mg/kg soil samples. As shown in Figure 1, all calculated BAF values were below 1.0. There was a concentration-dependent reduction in the BAFs, and Ag-NPs yielded higher BAFs than Ag nitrate from concentrations of 30 mg/kg soil up to 200 mg/kg soil.

Table 3. Silver content in earthworm tissues
Concentration (mg/kg soil)bSilver nanoparticles; mean Ag (mg/kg wormc) ± SDSilver nitrate; mean Ag (mg/kg wormc) ± SD
  • SD = standard deviation.

  • a

    Statistically significant difference shown from the lowest test concentration (α = 0.05).

  • b

    Dry matter soil.

  • c

    Dry matter worm.

157.0 ± <0.018.0 ± 0.9
3010.5 ± 0.4a7.3 ± 1.5
6011.1 ± 0.2a6.8 ± 0.4
12011.3 ± 0.4a6.7 ± 0.2
20011.2 ± 0.1a5.9 ± 0.1a
Figure 1.

Bioaccumulation factors (BAFs) for the earthworm Eisenia fetida exposed to silver nanoparticles (NM-300 K) and silver nitrate compared with the nominal concentrations.

Diffusive gradients in thin films

The amount of free Ag+ in the soil pore water represents the bioavailable part of Ag-NPs and Ag nitrate, which is likely to be predominantly responsible for any adverse effects on earthworms. The free Ag+ was calculated for all negative control tests as well as tests carried out at concentrations of 60, 120, and 200 mg/kg soil, and the data are presented as nanograms Ag+ per kilogram soil and as percentages of nominal concentrations in Table 4.

Table 4. Concentration of silver ions measured by diffusive gradients in thin films in the soil pore water
 Test 1 (silver nanoparticles)Test 2 (silver nanoparticles)Test 3 (silver nitrate)Test 4 (silver nitrate)
  • a

    Dry matter soil.

Concentration (mg/kg soila)Ag+ (ng/kg soila)Percentage from nominal concentration (×10−4)Ag+ (ng/kg soila)Percentage from nominal concentration (×10−4)Ag+ (ng/kg soila)Percentage from nominal concentration (×10−4)Ag+ (ng/kg soila)Percentage from nominal concentration (×10−4)
Day 0
 Negative control0.
Day 26
 Negative control0.
Day 54
 Negative control1.

The results show that only approximately 0.0001% of the nominal concentration of Ag is available as free Ag+ in the soil. Free Ag+ was not present in the negative control, and there was little difference in the amount of free Ag+ in soils spiked with Ag-NPs and Ag nitrate.

The amounts of Ag+ at the first and second measurement points were comparable. At least for tests 1, 2, and 4, there was an increase in free Ag+ levels with increasing total concentrations; for example, in test 1 (NM-300 K), the amount of free Ag+ increased from 22.1 to 73.4 ng/kg soil and, in test 4 (Ag nitrate), it increased from 31.7 to 60.6 ng/kg soil (Table 4). The percentage of the nominal concentration did not increase proportionally but instead declined (tests 2 and 4) or remained the same (test 1). The concentration dependence was less pronounced in test 3. At the third measurement point in all four tests, the highest Ag+ levels were detected at 120 mg/kg soil (151.8 ng/kg soil in test 2 with NM-300 K and 103.8 ng/kg soil in test 4 with Ag nitrate), and the percentage of nominal concentration also increased.


Ecotoxicological effects

We achieved a 90% recovery for both low and high concentrations of Ag, so there was no concern that the Ag concentration in the soil might be reduced by any means during the 56-d test period. According to the guidance document on aquatic ecotoxicology 25, the nominal concentration can be used to express toxicity if the measured concentration has a recovery >80%. Because equivalent guidelines are not available for terrestrial tests, the procedure described for aquatic tests was used.

Our objective was to provide insight into the ecotoxicity of Ag, using the earthworm reproduction test and addressing the following three endpoints: mortality, biomass increase, and number of offspring. Natural soil was used for our tests so that the results were relevant to environmental conditions. The results indicated that Ag-NPs and Ag nitrate do not induce statistically significant mortality in earthworm populations at Ag concentrations up to 200 mg/kg soil. However, we observed a statistically significant increase in the biomass of the adult worms exposed to Ag-NPs and Ag nitrate. The earthworms attempted to avoid the contaminated soil during the first 24 h of the test and then preferred to remain in the food layer spread on top of the soil, a behavior that persisted until the adult worms were removed. Our observations support earlier studies in which earthworms attempted to avoid soil contaminated with Ag-NPs and Ag nitrate at concentrations of 6.92 to 7.42 mg/kg soil, although the particles were larger and were coated with polyvinylpyrrolidone (PVP) 26. No differences between the effects of Ag-NPs and Ag nitrate were observed, confirming that earthworms appear to sense the presence of Ag+ in soil. The increase in biomass can therefore be explained by the tendency of the earthworms to favor the food layer, which leads to the ingestion of more food and an increase in biomass. The avoidance behavior and the resulting increase in biomass can be reduced by spreading the food in a thin layer on the soil surface.

After 56 d, we observed significant differences in the number of offspring at an Ag-NP concentration of 30 mg/kg soil and at a Ag nitrate concentration of 15 mg/kg soil (the lowest concentration we tested). There was only a marginal difference between the toxicity of Ag-NPs and Ag nitrate, yielding EC50 values of 74 to 80 mg/kg in comparison with the negative control for NM-300 K and 42 to 47 mg/kg for Ag nitrate. In one test with Ag-NPs, the dispersant also inhibited reproduction. Several tests with earthworms and other organisms in the presence of the dispersant were carried out, with no evidence of toxic effects (data not shown). We are therefore unable to explain why the dispersant had an inhibitory effect in one of the tests but none of the others. The calculated EC50 value based on the negative control was similar in both tests with Ag-NPs, so we focus on the results compared with the negative control in the following discussion.

Few studies have considered the different issues that influence the toxicity of Ag-NPs to the earthworm E. fetida. A limit test using PVP-coated Ag-NPs and Ag nitrate at a concentration of 1,000 mg/kg soil in a natural soil resulted in 97.5% survival for earthworms exposed to Ag-NP and 2.5% survival for those exposed to Ag nitrate 15. The number of cocoons was used as an indicator of reproduction, but the surviving earthworms produced no cocoons even in the Ag-NP test, in which most of the earthworms survived. Our data also show that reproduction is a sensitive endpoint and is strongly affected by both Ag-NPs and Ag nitrate.

Another study focusing on the influence of surface coatings on the bioaccumulation of Ag-NPs and reproduction toxicity in E. fetida was carried out with artificial soil 16. Silver nanoparticles coated with PVP and oleic acid, with a nominal particle size of 30 to 50 nm, were tested at nominal concentrations of 10, 100, and 1,000 mg/kg against Ag nitrate at nominal concentrations of 10 and 100 mg/kg. In these tests, neither the Ag nitrate nor the Ag-NPs coated with PVP and oleic acid affected growth and mortality. However, there was a significant effect on earthworm reproduction at 773.3 mg/kg for PVP-coated Ag-NPs, 727.6 mg/kg for oleic acid-coated Ag-NPs, and 94.12 mg/kg for Ag nitrate. The coated Ag-NPs were approximately 10 times less toxic than our uncoated Ag-NPs with a primary particle size of 15 nm, whereas the results for Ag nitrate were in a comparable range. However, cocoon production was again used as the parameter to measure reproduction in the investigation discussed above, whereas in the present study the juveniles were counted, which limits direct comparisons. Nevertheless, it can be assumed that particle properties, for example, size and coating, play an important role in the toxicity of Ag-NPs to earthworms.

A further study focused on the role of particle size and soil type in the toxicity of Ag-NPs to earthworms 17. Two types of soil were tested to determine the influence of soil composition, a sandy loam soil comparable to our test soil and an artificial soil. The study also included two types of Ag-NPs, one with a small particle size (10 nm) and another with particles of 30 to 50 nm, both coated with PVP. There were no differences in toxicity between the two types of particles. Growth and reproduction (expressed as “juveniles and worms”) were significantly affected in the natural soil by 7.413 mg/kg Ag nitrate, but the Ag-NPs had no significant effect on any of the tested endpoints. However, only a nominal concentration of 10 mg/kg was tested in the natural soil. The higher toxicity observed in our study may reflect differences in the organic matter content of the soils (0.93% in our case but 1.77% in an earlier study 17). The organic matter content affects the fate of Ag-NPs in soil, and all types of Ag are more mobile in mineral soils compared with soils rich in organic matter 27.

The studies discussed above indicate that the toxicity of Ag-NPs is strongly dependent on the properties of test medium (e.g., organic matter) and the coating. We also used smaller particles compared with previous investigations, which may also increase the toxicity of Ag-NPs.

Silver content of earthworms

In addition to the typical endpoints considered in the standardized reproduction test, we also investigated the uptake and accumulation of Ag in adult earthworms. As specified in OECD guideline 222 18, adult earthworms were removed after 28 d. Worms were used to determine the concentration of Ag after the gut had been purged. A concentration-dependent effect on reproduction above the lowest test concentrations (15 mg/kg) was observed, but, although the lowest and highest test concentrations differed by a factor of 13, the Ag concentrations in the earthworms were comparable (and were higher in earthworms exposed to Ag-NPs than in those exposed to Ag nitrate). We therefore assume that a steady state of Ag uptake is already achieved at 30 mg/kg soil. It is unclear whether the measured Ag is located in the tissues or whether residues remain in the gut because of incomplete purging. The comparable concentrations in earthworms exposed to soil concentrations greater than 30 mg/kg and the concentration-dependent inhibition of reproduction at concentrations of 30 to 200 mg/kg soil indicate that the Ag content in the worms is not responsible for the observed effects. It can be assumed that the fertility of adults is not affected, but the development of cocoons and the survival of juveniles in soil are sensitive life stages.

In a previous study 16, a concentration-dependent increase in the levels of Ag in earthworm tissues was observed, although none of the BAFs exceeded 1, suggesting that there was no bioaccumulation of Ag-NPs. The BAFs for Ag-NPs and Ag nitrate at 10 and 100 mg/kg soil in the study cited above were comparable to our results at 15 and 120 mg/kg soil.

As with earthworms, nematodes are also exposed to Ag via soil pore water, so studies considering the uptake of Ag-NPs in the nematode Caenorhabditis elegans are also relevant to this discussion. These studies have shown that Ag-NPs are taken up into cells from the gut lumen and that transgenerational Ag-NP transfer is possible 28. Citrate-coated Ag-NPs with a particle size of 50.6 nm induced epidermal fissuring and serious epidermal burst effects in a concentration-dependent manner at concentrations of 10 and 100 mg/L 29. However, both studies were carried out using aqueous media rather than soil and might not reliably predict the influence of Ag-NPs on earthworm guts and tissues or on juveniles or cocoons.

DGT approach

The concentration of Ag+ in pore water was determined in addition to the total silver concentration using the DGT approach, which showed that only 0.0001% of the nominal Ag concentration exists as freely available Ag+ in the soil pore water. There was no statistically significant difference between Ag-NPs and Ag nitrate in our tests.

In tests with C. elegans 29, the measurement of dissolved Ag+ in K-medium after 24 h of incubation revealed a concentration equivalent to 0.001% of the nominal concentration, which may differ from our results because of the different incubation conditions. In aqueous K-medium, the equilibrium between bulk Ag and Ag+ may differ from that in soil, resulting in the lower concentration of Ag+ observed in this investigation. Other studies have focused on the interactions between Ag+ and other environmentally relevant compounds. The release of Ag+ may also be modified by interactions between Ag-NPs and sulfide ions or dissolved organic matter (DOM). A study with nitrifying bacteria in a wastewater treatment plant showed that sulfide reduces the toxicity of Ag-NPs by promoting the formation of AgxSy complexes or precipitates 30. Furthermore, if the concentration of DOM is higher than that of metal ions in water, then DOM binds to the dissolved metal ions, which may influence the Ag+ content detectable by DGTs 31.

We observed a concentration-dependent increase of Ag+ in the soil pore water at least at the first two measurement points. Furthermore, the comparable concentrations of Ag+ detected via DGT in soil pore water from soils spiked with Ag-NPs and Ag nitrate match the similar effect of these substances on earthworm reproduction. Therefore, it can be assumed that Ag+ is responsible for the inhibition of earthworm reproduction. Because the effect of Ag-NPs and Ag nitrate is in a similar range, we can conclude that the locally increased ion concentration on the surface of NPs compared with evenly distributed Ag nitrate is not relevant for earthworm toxicity. This may not be the case for single-cell organisms, which may be exposed to high ion concentrations in the vicinity of AgNPs. Yang et al. 32 note that Ag-NPs are more toxic than the equivalent mass of dissolved Ag or the generation of reactive oxygen species in studies with single-cell systems (e.g., nitrifying bacteria, human cells). In multicellular organisms (e.g., nematodes, fish), toxicity is caused predominantly by the dissolution of Ag-NPs.


Silver NPs and Ag nitrate display similar toxicity in the earthworm reproduction test, and we observed a concentration–effect relationship in all the tests that we carried out. The number of juveniles was a more suitable endpoint than biomass or mortality, and we recommend that at least this endpoint should be determined. Earthworms can sense low concentrations of uncoated Ag-NPs in the soil, so their food must be distributed in a thin surface layer to prevent avoidance behavior. In further tests, the number of juveniles and cocoons should be counted to achieve better comparability with previous investigations. Studies at the cellular level may help to elucidate the mode of action.

The uptake of Ag-NPs in the earthworm was slightly more efficient than that of Ag nitrate. Spiked soils with a concentration-dependent effect on reproduction did not cause a concentration-dependent increase in the amount of Ag in earthworm tissues. The uptake of Ag appears not to be the factor that inhibits reproduction.

Diffusive gradients in thin films measurements revealed no differences between the free Ag+ content of soils spiked with Ag-NPs and Ag nitrate. We observed a concentration-dependent increase at least during the first two measurements points, suggesting that the released ions affect the cocoons and juveniles and are responsible for the inhibition of reproduction. The results strengthen the assumption that toxicity caused by Ag-NPs is closely related to the content of released Ag+ measured in the soil pore water. Introducing the DGT method for measuring Ag+ ions released from Ag-NPs in the soil pore water is potentially an inexpensive way to gain insight into the concentration of free ions.

We focused on pure Ag-NPs in this investigation, but transformation processes in WWTPs generate modified Ag compounds. Therefore, to increase the environmental relevance of the reported observations, it will be important to study the effects on terrestrial organisms of Ag-NP-containing sewage sludge derived from model or pilot WWTPs. This includes potential Ag transformation products generated during the purification process.


The data reported in the present study are from projects performed on behalf of the German Federal Ministry of Education and Research and the German Environment Agency, which were supported by federal funds. We thank T. Görtz, K. Mock, R. Nöker, R. Schlinkert, J. Tigges, and P. Schulte for counting thousands of juvenile worms; D. Hansknecht and J. Schörmann for their support with the digestion and chemical analysis; and H. Rüdel and B. Knopf for scientific discussion.