*2.1. Adsorption-Desorption Test*

The adsorption rates are shown in Table 1, and the parameters of the Freundlich model fitted are shown in Table 2. The log Kd value of lincomycin and monensin in laterite soil were 1.82 and 2.76. The log Kd value of roxarsone in black soil was 1.29. The corresponding 1/n values of the three equations were less than 2. According to the Kd value, the adsorption of laterite to monensin was greater than to lincomycin, which could be due to the higher molecular weight and lower water solubility of monensin. The 1/n value of lincomycin was 0.0935 in laterite soil. Therefore, the isotherm adsorption line of lincomycin belonged to the "l" type, meaning that at a certain concentration range, the adsorption capacity of soil to lincomycin decreases with increasing drug concentration. This result was consistent with the observed trend of the lincomycin adsorption rate. Another study showed that when the initial drug concentration was 1 mg/L, the adsorption rate of ethanamizuril in podzol soil was approx. 70% and decreased with increasing drug concentration, indicating that podzol soil displayed a stronger adsorption to lincomycin and monensin than ethanamizuril [27]. The lower organic matter contents and higher pH of the laterite and podzol soil are properties that may have contributed the smaller adsorption of roxarsone (pKa1 = 3.49). Rutherford et al. [28] found that when 2 < pH < 8, the adsorption of roxarsone in soils decreased as pH increased. When the pH of the soil solution increased, the negative charge of the soil surface increased, causing an increased repulsive force between roxarsone and soil colloid, and a lower soil adsorption capacity of roxarsone. The zero adsorption of roxarsone was observed in field soil (pH 8.4, organic matter 6.84%) and wasteland soil (pH 8.2, organic matter 4.73%) [29].


**Table 1.** Adsorption rates of drugs in di fferent soils (%).


**Table 2.** Parameters of the Freundlich model.

#### *2.2. Soil Mobility Test*

Soil mobility test results are shown in Table 3. In laterite soil, the ranking of mobility capacity of drugs was monensin > lincomycin > roxarsone, and in black soil, the ranking was monensin > roxarsone > lincomycin. According to the classification in GB/T 31270-2014 (Test guidelines on environmental safety assessment for chemical pesticides) [30], lincomycin and roxarsone were moderately mobile in laterite and black soils, while monensin was highly mobile. Generally, adsorption ability has a negative correlation with water solubility of drugs, and a positive correlation with soil organic content; migration ability has the opposite correlations [31]. A previous study showed that the Rf value of roxarsone was 0.66 in soil (pH 6.5, organic matter 19.08 g/kg), indicating strong mobility [32]. However, in the present study, the Rf value was 0.4993 in laterite soil (pH 6.710, organic matter 3.92 g/kg), and 0.5835 in black soil (pH 5.257, organic matter 263.69 g/kg). These di fferences may have resulted from other physicochemical properties [33]. Florasulam was shown to be primarily distributed in a 9–18 cm soil layer in soil (organic matter 23.57 g/kg), and its Rf value was 0.918 [34]. According to the Rf values in the presents study, the migration ability of the three drugs tested could be weaker than florasulam in a similar soil environment.


**Table 3.** Content distribution of drugs using soil thin layer chromatography.

#### *2.3. Algae Growth Inhibition Test*

The absorbance of algae culture remained unchanged or even decreased with time, indicating that the concentrations of the algae decreased and their growth was restrained. Figure 1A–C show that the inhibition rate of algae (*S. obliquus*) growth by the three drugs was positively correlated with drug concentration. The pH of the treatment groups and controls were 6.9–7.4 at the end of the algae growth inhibition test. As the culture time increased, the percentage inhibition of algae growth rate increased. Based on the 96-h inhibition rate, the calculated EC50 (96h) were 0.813 mg/<sup>L</sup> (lincomycin, 95% confidence interval [0.791, 1.489] mg/L), 1.085 mg/<sup>L</sup> (monensin, [0.554, 1.193] mg/L) and 13.15 mg/<sup>L</sup> (roxarsone, [9.70, 17.83] mg/L). These results suggested that lincomycin, monensin and roxarsone showed low, medium, and low toxicity levels to algae, respectively. The EC50 (96h) of lincomycin and monensin suggested that these two drugs demonstrated higher toxicity to algae than ofloxacin (6.20), sulfamethoxazole (7.20) and sulfamethazine (9.89) [35], which was consistent with the results of Peng et al. [36].

Peng et al. also found that lincomycin and ofloxacin presented a high ecological risk to algae; the risk quotient (RQ) of lincomycin was 1.93, followed by ofloxacin (1.33), tetracycline (0.41) and erythromycin (0.20). According to another study, lincomycin can inhibit the synthesis of the D1 protein in the algal photosynthesis system, and may lead to algal death [37].

#### *2.4. Plant Sensitivity Test*

The effects of drugs on plant growth are shown in Tables 4 and 5. The seedling rates of blank controls and methanol controls (for monensin) were all above 50%, suggesting that experiments were valid.


**Table 4.** Number of plants emerged in drug groups (n = 10 seeds).

1 data represents mean ± standard deviations, n = 3, the same below. \* represents significant difference at *p* < 0.05 compared with the control group.

**Table 5.** Effect of lincomycin and roxsarsone on *A. thaliana* height and biomass (14 d).


**Figure 1.** Effect of drugs on the growth of *S. obliquus.* (**A**) lincomycin, (**B**) monensin, (**C**) roxarsone.
