| Front page | | Contents | | Previous | | Next |

Fate and Effects of Triclosan

4 Terrestrial risk assessment of Triclosan

The main sources of data on fate and effects of Triclosan in terrestrial environments are Reiss et al. (2001) and a manuscript/report, reviewing a series of studies (Mones not dated).

4.1 Fate of Triclosan in soil

The main route of exposure to soil is expected to be via the application of sewage sludge to agricultural soil. The studies reviewed by Mones include sorption and biodegradation in soil.

4.1.1 Sorption and mobility in soil

In the review made by Mones, the water solubility of Triclosan is given as 10 mg/L while Reiss et al. (2001) state that it is 12 mg/L. Both are in agreement

as regards the octanol water partition coefficient, which is log K_ow = 4.8. Based on this information, Triclosan is considered to have a potential for sorbing to organic matter and particles in sludge and soil (Mones).

In a recent (1997) study based on American standardised sorption/desorption methods, the sorption of Triclosan to sludge was measured. Furthermore, results from an early (1988) study of the mobility in soil, based on German guidelines, are available (Mones).

The sorption study resulted in a K_d value of 21,529 (log K_d = 4.3) and a K_oc value of 47,454 (log K_oc = 4.7) in sludge (Mones).

In the mobility study, sludge spiked with ¹⁴C-labelled Triclosan was placed on top of columns of soil (German standard agricultural (LUFA) soil) and for 48 hours, distilled water was allowed to leach through the columns from the top. Samples of eluate collected at the bottom of the columns as well as soil samples from different levels of the columns were analysed by liquid scintillation counting. The study showed no leaching (i.e. no radioactivity) of Triclosan below a level of 15 cm in the columns during 48 hours (Mones).

Based on this information, it is considered to be a reasonably realistic worst case assumption for the effect assessment that Triclosan, which is entering soil, will not be removed by leaching.

4.1.2 Biodegradation in soil

In a contract laboratory study from 1994, the biological degradation of ¹⁴C-labelled Triclosan was investigated in agricultural soil spiked with the test substance at concentrations of 40 and 600 µg/kg (soil dry weight) for 577 days. The agricultural soil originated from a sludge-amended field. A series of controls with sterilised soil was included for control of laboratory conditions and photolysis. The latter was being kept outside in daylight. Degradation was measured as CO₂ production.

The results of this study indicated that approximately half of the parent substance was mineralised during the 577 days while the identity of the labelled substance remaining in the soil could not be unequivocally demonstrated to be different from the parent substance (Mones).

A 64-day study (conducted by another contract laboratory in 1994) included measurement of CO₂ evolution from three different types of agricultural soil, to which sludge was added in the laboratory, and the matrix was spiked with 200 µg/kg of labelled Triclosan. The sludge used for the experiments was adapted as its origin was an industrial sewage treatment facility receiving Triclosan in the waste stream. Therefore, the results of this study cannot be taken as a general indication of the biodegradability of Triclosan in sludge-amended soil.

After 64 days, between 12 and 20% of the Triclosan was mineralised as measured by CO₂ evolution. However, characterisation of the substances remaining in the soil demonstrated that most of the Triclosan had been transformed to extractable metabolites in the soil. A major metabolite was identified as 2’methoxy 2,4,4’-trichloro-diphenyl ether. Only between 4.4 and 23% of the parent compound was intact after 64 days. The half-life of Triclosan was calculated to be in the range of 17.4-35.2 days for the three experimental soils (Mones).

4.1.3 Concentrations in Swedish soil samples

In an investigation of soil in Sweden, samples were collected from two contaminated areas (an area with a former wood preservation plant in Boro and the area of a plastic production plant in Ystad) and from an unaffected forest area at Gårdsjön (Remberger et al. 2002). The results showed Triclosan concentrations between <3 and 15 µg/kg d.w. in the contaminated sites while concentratoins in the forest soil were below the detection limit of 3 µg/kg d.w. soil.

4.1.4 PEC for the soil compartment

As sewage sludge is the major source of Triclosan contamination of soil, the concentrations in sewage sludge (Table 2.7) were used for the estimation of a PEC value for the soil compartment. Concentrations in U.S. and Swedish sewage sludge from plants with activated sludge are reported to range from 0.028-4.2 µg/g (= mg/kg), based on sludge dry weight. The Triclosan concentrations in sludge from trickling filter or other bio-filter plants from the same countries were in the range of 0.38-15.6 mg/kg, all based on sludge dry weight.

In Denmark, the “realistic worst-case” conditions for sewage sludge application to agricultural soil, which are used for risk assessment, are maximum application rate = 3 tonnes dry weight per hectare once every three years, mixing depth = 15 cm, soil density = 1.5 kg/L. This results in a dilution factor of 750.

Based on the above concentrations and assumptions, the following PEC values for Triclosan can be estimated:

Activated sludge:     PEC_soil = 0.00004-0.0056 mg/kg soil

“Bio-filter” sludge:   PEC_soil = 0.0005-0.021 mg/kg soil

4.2 Effects of Triclosan on terrestrial organisms

4.2.1 Toxicity to soil-living organisms

A 2-week acute toxicity test with compost worms (Eisenia fetida) in soil was carried out according to the OECD TG No. 207 (Mones, Reiss et al. 2001). In this study, the artificial soil was deviating from that described in the TG as 21% of the sand was exchanged for natural soil, giving a higher sorptive capacity. Furthermore, the test substance was mixed with dry soil, which is known to give maximum sorption. Therefore, the test may have underestimated the toxicity of Triclosan to the worms. No significant effects on worm survival or weight were measured at the highest concentration of Triclosan tested (1,026 mg/kg, soil dry weight).

Mones and Reiss et al. (2001) quote a seedling growth test with six plant species (corn, ryegrass, wheat, cucumber, soybean and tomato). Reiss et al. (2001) report that this study showed cucumber to be the most sensitive species with NOEC = 96 µg/kg. Furthermore, both mention an additional test with cucumber (the most sensitive species), which is described by Mones. The result of this study is quoted by Mones as NOEL for all parameters (shoot length, shoot and root weight) of “>424 1000 µg/kg” (presumably: >424-1,000 µg/kg) while Reiss et al. (2001) state that there were no effects up to 1,000 µg/kg. Chemical analysis of the test substrate indicated that only 34% of the test substance was left intact at the end of the study.

In the terrestrial toxicity section, Mones quotes a study, which was carried out with “activated sludge-mixed liquor”, as being relevant to the top layers of soil. Exposure time was 15 minutes and inhibition of bacterial heterotrophic activity was measured, resulting in an EC₅₀ = 239 mg/L.

As the bacterial test mentioned cannot be considered relevant to soil-living organisms, only results of tests with two groups of soil-living organisms are available. Both tests must be considered to be acute tests and no L/EC₅₀ value is reported. Therefore, the calculation of a PNEC for the soil compartment must be based on the only effect concentration available (NOEC = 96 µg/kg) and an assessment factor of 1,000. Thereby, a PNEC_soil can be estimated, which must be considered as preliminary due to lack of data. It is most likely that availability of results from standardised quality studies could lead to the use of an L/EC₅₀, which was lower than the 96 µg/kg. Furthermore, if a large dataset could be used, the assessment factor could be lowered. However, such data are not available at present and the preliminary PNEC value must be used for the terrestrial risk assessment:

PNEC_soil = 0.096 µg/kg.

4.2.2 Toxicity to birds

The available toxicity studies with birds include two 14-day acute oral toxicity studies with mallard duck and bobwhite quail (Mones, Reiss et al. 2001) and an 8-day acute dietary study with bobwhite quail (Mones). All studies seem to have been carried out according to standardised test guidelines in contract laboratories during the 1990s.

The mallard duck study showed no significant effects on body weight, feed consumption or gross pathology at doses up to 2,150 mg/kg body weight. Therefore, the NOEC = 2,150 mg/kg body weight.

The acute oral study with bobwhite quail resulted in a LD₅₀ of 862 mg/kg body weight and diarrhoea was noted in the lowest concentration test group (147 mg/kg body weight) and therefore no NOEL could be established.

The 8-day acute dietary study with bobwhite quail indicated no mortality up to 1,250 mg/kg food. However, at 2,500 mg/kg food, one death and, at 5,000 mg/kg food, 4 deaths (10 birds/group) were recorded. The LC₅₀ was > 5,000 mg/kg food but no conclusion is drawn by Mones regarding the significance and interpretation of the mortality recorded at 2,500 and 5,000 mg/kg food.

It is not possible to apply this information to a risk assessment as there is no effect concentration for dietary intake of Triclosan.

4.3 Risk assessment for the soil compartment

For the soil compartment, the risk quotients based on the above PEC and PNEC values can be calculated as summarised in Table 4.1.

Click on the picture to see the html-version of: Table 4.1

Based on the preliminary PNEC value as discussed above and measurements in US and Swedish sludge, the majority of the risk quotients for the soil compartment are > 1. Therefore, all the sludge concentrations measured in the U.S.A. and Swedish samples, except for the one with the lowest concentration of Triclosan, would be expected to cause effects in the soil immediately after application of the maximum amount used in Denmark.

For a more realistic terrestrial risk assessment, information regarding concentrations of Triclosan in Danish sewage sludge would be needed. Furthermore, if toxicity data of high quality for terrestrial organisms were available, more confidence could be laid on the PNEC value.

| Front page | | Contents | | Previous | | Next | | Top |