2.4.3. Roxarsone

The EC50 (14 d) of roxarsone on *A. thaliana* seedling rate was 32.18 mg/kg dry soil (95% confidence interval [11.319, 49.795] mg/kg dry soil). On 14 d, the number of plants that emerged in the blank control group was significantly higher than in each treatment group (Table 4). In the treatment groups, only the 5 mg/kg and the 80 mg/kg roxarsone groups showed significant differences between each other in seedling rate on 14 d. This indicated that roxarsone could reduce the growth of *A. thaliana*.

#### *2.5. Daphnia Activity Inhibition Test*

In the control group, the daphnia (*D. magna*) body structures were clearly observed, not immobilized, and did not display unusual behaviours. However, in the high drug concentration groups, the body structure of daphnia appeared blurred, and daphnia immobilization was observed. Table 6 shows the inhibition rate of monensin and roxarsone on daphnia activity. The EC50 (48 h) values were 21.1 mg/<sup>L</sup> (monensin, 95% confidence interval [17.9, 24.7] mg/L) and 292.6 mg/<sup>L</sup> (roxarsone, [283.7, 301.8] mg/L), which suggested that both drugs had low toxicity on daphnia. Acute activity inhibition did not occur in the lincomycin group at all concentrations tested after 48 h. Therefore, the EC50 of lincomycin for daphnia was > 400 mg/<sup>L</sup> (the maximum test concentration).

**Table 6.** Activity inhibition numbers of daphnia in monensin and roxarsone groups at 48 h (n = 8).


#### *2.6. Acute Toxicity Tests of Zebrafish*

The potassium dichromate toxicity test (reference poison) for zebrafish (*D. rerio*) is shown in Table 7. The LC50 of potassium dichromate for zebrafish was 356.6 mg/<sup>L</sup> (95% confidence interval [333.0, 381.8] mg/L). An LC50 value between 200–400 mg/<sup>L</sup> demonstrated that the experimental method and the fish quality met the appropriate standards [39].


**Table 7.** Effect of potassium dichromate on zebrafish mortality (n = 8).

Table 8 shows the toxicity of monensin and roxarsone on zebrafish. There was no significant difference between the methanol control group and the blank control group (*p* > 0.05). The LC50 (96 h) values were 4.76 mg/<sup>L</sup> (monensin, 95% confidence interval [4.670, 4.853] mg/L) and 452.7 mg/<sup>L</sup> (roxarsone, [447.9, 457.5] mg/L), suggesting that monensin and roxarsone had medium and low toxicity, respectively, for zebrafish. The zebrafish in the lincomycin group did not show any abnormal activity at all concentrations tested after 96 h. Therefore, the LC50 (96 h) of lincomycin was >2800 mg/<sup>L</sup> (the maximum test concentration). Previous research showed that roxarsone could cause rapid DNA destruction of *Carassius auratus*, and low doses of roxarsone could cause more sustainable and increased damage than high doses [40]. The experimental LC50 (96 h) values of oxytetracycline and norfloxacin for zebrafish were 1.262 × 10−<sup>3</sup> mol/L and 2.026 × 10−<sup>3</sup> mol/L, respectively [41], suggesting their toxicity may be similar to roxarsone and lower than monensin.


**Table 8.** Effect of monensin and roxarsone on zebrafish mortality (n = 8).

#### *2.7. Acute Toxicity Tests of Earthworm*

Table 9 shows the toxicity of monensin and roxarsone on earthworms (*E. fetida*). The mortality rate of the control group was less than 10%, suggesting the test was valid. The appearance of earthworms in the low concentration groups showed no difference compared to the control group. However, in the 20 g/kg roxarsone group, the earthworms atrophied, the girdles were swollen, and some earthworms died.


**Table 9.** Effect of monensin and roxarsone on earthworm mortality (n = 10).

CK represents control check.

There was no significant difference between the methanol control group for monensin and its blank control group (*p* > 0.05). The LC50 (14 d) values were 346.0 mg/kg dry soil (monensin, 95% confidence interval [309, 387] mg/kg dry soil) and 5.74 g/kg dry soil (roxarsone, [5.14, 6.41] g/kg dry soil), indicating that both drugs had low toxicity for earthworms [30]. No earthworm mortality or abnormal activity was discovered at all concentrations of lincomycin tested after 14 d. Therefore, the LC50 (14 d) of lincomycin was >15 g/kg dry soil (the maximum test concentration). The LC50 of monensin for earthworms was higher than the reported 75.883 mg/kg observed in an artificial soil test [20].

#### *2.8. Acute Toxicity Test on French Giant Quail*

Table 10 shows the toxicity of monensin and roxarsone on French giant quails. The mortality rate of the control group was less than 10%, suggesting the test was valid. At high concentrations of roxarsone and monensin, inappetence, listlessness and other symptoms were observed in quails. There was no significant difference between the carboxymethylcellulose sodium control and the blank control (*p* > 0.05), and no significant difference between males and females (*p* > 0.05). The LD50 (7 d) values were 672.8 mg/kg (monensin, 95% confidence interval [451.0, 1003.8] mg/kg) and 103.9 mg/kg (roxarsone, [80.5, 134.2] mg/kg), indicating that monensin and roxarsone displayed low and medium toxicity, respectively. All quails exposed to lincomycin survived after 7 d, and their appetite, activity, and excretion were no different from quails in the control group. Therefore, the LD50 (7 d) of lincomycin was > 2000 mg/kg (the maximum concentration). Another study on the effect of roxarsone on laying hens suggested that roxarsone could cause an increase in aspartate aminotransferase (AST), lactate dehydrogenase (LDH), and creatine kinase (CK) activity, a decrease in the liver weight, and histological evidence of liver damage [42]. Therefore, it is possible that roxarsone causes quail mortality via oxidative stress.


**Table 10.** Effect of monensin and roxarsone on quail mortality (5 male and 5 female).

## *2.9. Summary of Results*

Acute toxicity test data and toxicity category of the three drugs to six species are summarised in Table 11. Lincomycin showed low toxicity for all tested organisms, and medium toxicity for *S. obliquus*. Monensin showed medium toxicity for *S. obliquus* and *D. rerio*, and low toxicity for *E. fetida*, *D. magna*, and *C. coturnix*, and no detectable toxicity for *A. thaliana*. Roxarsone showed medium toxicity for *C. coturnix* and low toxicity for all other tested organisms.

**Table 11.** Summary of acute toxicity data and category of drugs.


As an important component of aquatic ecosystems, algal photosynthesis accounts for a large proportion (up to 50%) of global primary productivity [43]. Due to its toxicity on algae, the clinical dosage and release of lincomycin should be carefully regulated. Monensin showed a significantly higher toxicity risk on some species compared with other commonly used antibiotics, such as tetracyclines and quinolones. This result was consistent with previous research findings showing that ionophores exhibited higher toxicity than other antibiotics [19]. More attention should be given to the use of monensin, its residue and accumulation in the environment. Roxarsone-contaminated soil and its accumulation in rice could present serious problems for human health [44]. Despite the low ecotoxicity of roxarsone found in this study, its use, and the use of other arsenic-containing drugs, requires strict control to avoid arsenic entering the food chain. Possible methods to reduce the concentration of drugs in the environment include the treatment of animal manure before field application, the use of alternative bio-agents for disease treatment, and a well targeted legalized use of antibiotics [45].

## **3. Materials and Methods**

#### *3.1. Chemicals, Test Soils and Organisms*

Roxarsone (purity >98%) was purchased from Shanghai Titan Technology (Shanghai, China), LTD. Sodium morenate (purity >90%) and lincomycin hydrochloride (purity >95%) were purchased from Sangon Biotech LTD (Shanghai, China). The podzol soil was collected from the Shanghai suburban district. The black soil was taken from a forest in Changbai Mountain, Jilin Province, China. The laterite soil was collected from idle farmland in Chuzhou, Anhui Province, China. All test soils were collected from the surface layer (0–10 cm) that had not been farmed for more than 20 years, and were air-dried, grinded and sieved to a 2-mm size. The basic properties of the test soils are shown in Table 12.


**Table 12.** Basic properties of the test soils.

*Scenedesmus obliquus* (green algae) were provided by the Chinese Academy of Sciences Institute of Hydrobiology (Wuhan, Hubei Province, China), and were grown in BG-11 cultures with continuous passage for three times and were tested in the logarithmic growth phase. *Arabidopsis thaliana*, wild type, were provided by School of Life Science and Biotechnology, Shanghai Jiaotong University (Shanghai, China). All seeds were treated at 4 ◦C for more than a week before sowing. *Daphnia magna* was provided by Guangdong Laboratory Animals Monitoring Institute (Guangzhou, Guangdong Province, China). The daphnia was domesticated in laboratory conditions for 7 d. Healthy infant daphnia were selected for experiments. *Danio rerio* was purchased from a pet house in Minhang District, Shanghai, China. The fish were domesticated in laboratory conditions for 7 d and 2 cm-long healthy fish were selected. *Eisenia fetida* were purchased from Wangjun Earthworm Farm in Jiangsu Province, China, and domesticated for 14 d. Healthy adult earthworms weighing 300–600 mg with an obvious girdle band were selected. *Coturnix coturnix* (French giant quails) were purchased from Shanghai Fengxian Quail Farm in Fengxian District, Shanghai, China. They were aged approx. 30 d and weighed approx. 100 g. The quails were domesticated for 7 d under laboratory conditions, and those with no disease were selected. All experimental methods performed were in accordance with National Standards of China [30,39,46] and OECD [47].

#### *3.2. Adsorption-Desorption and Soil Mobility*
