and monensin [89]. Nicarbazin was not toxic to both species at concentrations
between 10 and 1,000 mg/kg dry soil, although monensin was lethal to earthworms
(14d-LC 50 ¼ 31.6 Æ 1.13 mg/kg, mean Æ SD) and significantly decreased the
reproduction rate of collembolans (28d-EC 50 ¼ 95.5 Æ 28 mg/kg). The median
lethal concentration of monensin for earthworms was similar to that reported in a
previous study with E. andrei (28d-LC 50 ¼ 49.3 mg/kg dry soil) [90], although the
incubation time was double than that of the study by Menezes-Oliveira et al. [89].
However, cautions must be taken when extrapolating outcomes from lab-scale
toxicity testing to the field. First, the earthworm ecology and distribution should be
considered in the environmental risk assessment of APIs. Eisenia fetida and
E. andrei are epigeic earthworms, which mean that they live above the mineral
soil surface and feed on plant litter [36]. These species rarely burrow into the soil as
anecic and endogeic earthworms do, so exposure of epigeics to API-contaminated
mineral soils should be lower than that for geophagous earthworm species
[91]. Additionally, because agricultural soils are continually altered by tilling in
successive crop seasons, Eisenia spp. are not abundant in these soils. Conversely,
anecic and endogeic species are well represented in agroecosystems [92–94]. Second, the toxicity tests recommend the use of artificial soils (e.g. OECD soil or LUFA
2.2 soil), which obviously cannot be considered agricultural soils. A myriad of
fluctuating variables of field soils may influence API degradation, bioavailability
and mobility that are not considered in artificial soils, such as quantity and quality of
organic matter content, microbial communities, aggregate distribution, etc. Third,
the risk of species confusion in toxicity testing is another potential disturbing
variable. In the case of E. fetida and E. andrei, both species can be easily confused
with the risk of obtaining non accurate results. They are different species [95], with
probably different responses (ecotoxicological biomarkers) to environmental pollutants [96]. Therefore, caution should be taken when using Eisenia spp. in the
assessment of API toxicity. Finally, species-specific differences in earthworm sensitivity to environmental contaminants should be also considered when assessing
API toxicity. For example, a meta-analysis study revealed that L. terrestris and
A. caliginosa are more sensitive to pesticide toxicity than E. fetida, which questions
the role of the latter for establishing environmental protection limits [97]. Indeed,
earthworm species other than Eisenia spp. are now suggested as model organisms
for standardised soil toxicity testing [91, 98, 99]. Therefore, despite the improvements made by EMA on the original guideline document for the environmental risk
assessment of APIs [81] – discussed in Whomsley et al. [100] – the inclusion of other
earthworm species highly representative of agroecosystems is not considered yet.
Earthworm biomarkers have been also included in toxicity testing as indicators of
API bioavailability and to assess the potential adverse effects of APIs. For example,
signs of oxidative stress (antioxidant enzyme activities and lipid peroxidation) and
genotoxicity (DNA breaks) induced by chlortetracycline were observed in E. fetida
incubated in antibiotic-spiked soils for 28 days, although such responses were not
dose-dependent [101]. The researchers also found neither dead worms nor significant decrease in reproduction rate (number of juveniles and cocoons) at the highest
antibiotic concentrations (100 and 300 mg/kg). Using the contact filter paper test
Vermiremediation of Pharmaceutical-Contaminated Soils and Organic Amendments
349
between 10 and 1,000 mg/kg dry soil, although monensin was lethal to earthworms
(14d-LC 50 ¼ 31.6 Æ 1.13 mg/kg, mean Æ SD) and significantly decreased the
reproduction rate of collembolans (28d-EC 50 ¼ 95.5 Æ 28 mg/kg). The median
lethal concentration of monensin for earthworms was similar to that reported in a
previous study with E. andrei (28d-LC 50 ¼ 49.3 mg/kg dry soil) [90], although the
incubation time was double than that of the study by Menezes-Oliveira et al. [89].
However, cautions must be taken when extrapolating outcomes from lab-scale
toxicity testing to the field. First, the earthworm ecology and distribution should be
considered in the environmental risk assessment of APIs. Eisenia fetida and
E. andrei are epigeic earthworms, which mean that they live above the mineral
soil surface and feed on plant litter [36]. These species rarely burrow into the soil as
anecic and endogeic earthworms do, so exposure of epigeics to API-contaminated
mineral soils should be lower than that for geophagous earthworm species
[91]. Additionally, because agricultural soils are continually altered by tilling in
successive crop seasons, Eisenia spp. are not abundant in these soils. Conversely,
anecic and endogeic species are well represented in agroecosystems [92–94]. Second, the toxicity tests recommend the use of artificial soils (e.g. OECD soil or LUFA
2.2 soil), which obviously cannot be considered agricultural soils. A myriad of
fluctuating variables of field soils may influence API degradation, bioavailability
and mobility that are not considered in artificial soils, such as quantity and quality of
organic matter content, microbial communities, aggregate distribution, etc. Third,
the risk of species confusion in toxicity testing is another potential disturbing
variable. In the case of E. fetida and E. andrei, both species can be easily confused
with the risk of obtaining non accurate results. They are different species [95], with
probably different responses (ecotoxicological biomarkers) to environmental pollutants [96]. Therefore, caution should be taken when using Eisenia spp. in the
assessment of API toxicity. Finally, species-specific differences in earthworm sensitivity to environmental contaminants should be also considered when assessing
API toxicity. For example, a meta-analysis study revealed that L. terrestris and
A. caliginosa are more sensitive to pesticide toxicity than E. fetida, which questions
the role of the latter for establishing environmental protection limits [97]. Indeed,
earthworm species other than Eisenia spp. are now suggested as model organisms
for standardised soil toxicity testing [91, 98, 99]. Therefore, despite the improvements made by EMA on the original guideline document for the environmental risk
assessment of APIs [81] – discussed in Whomsley et al. [100] – the inclusion of other
earthworm species highly representative of agroecosystems is not considered yet.
Earthworm biomarkers have been also included in toxicity testing as indicators of
API bioavailability and to assess the potential adverse effects of APIs. For example,
signs of oxidative stress (antioxidant enzyme activities and lipid peroxidation) and
genotoxicity (DNA breaks) induced by chlortetracycline were observed in E. fetida
incubated in antibiotic-spiked soils for 28 days, although such responses were not
dose-dependent [101]. The researchers also found neither dead worms nor significant decrease in reproduction rate (number of juveniles and cocoons) at the highest
antibiotic concentrations (100 and 300 mg/kg). Using the contact filter paper test
Vermiremediation of Pharmaceutical-Contaminated Soils and Organic Amendments
349