2.4.3. Acute and Chronic Tests on *D. rerio*

The acute toxicity test on *D. rerio* for both ELT-dg and ELT-dp suspensions comprised a 96 h exposure to 100.0 mg/L under static conditions [49]. Fish were considered dead if no movements were visible and if no reaction was observable when touching the caudal peduncle; in addition, we checked for visible abnormalities after 2, 24, 48, 72 and 96 h. Juvenile fish were obtained by ChemService Controlli e Ricerche s.r.l. facility and fed up to 24 h before the test. The facilities of ChemService Controlli e Ricerche s.r.l. follow Italian laws, rules and regulations (Legislative Decree No. 116/92; authorization n. 30/2012-A of 25 January 2012). Fish sizes were checked before the exposure by measuring a representative group of fish of the same batch. For ELT-dg exposure, mean length was 2.40 ± 0.26 cm and mean weight was 0.25 ± 0.04 g, whereas for ELT-dp exposure, mean length was 1.80 ± 0.17 cm and mean weight was 0.30 ± 0.04 g. Organisms of 2 months were exposed in 60 L glass aquaria (7 specimens *per* treatment) filled with reconstituted water (see Supplementary Materials for composition), with one replicate *per* concentration at a temperature of 22 ◦C, with 12 h

light (680–827 lx) and 12 h darkness daily photoperiod (2 daily 30 min transition periods; 754–827 lx), oxygenation > 60% and 6.60–7.77 pH range. The chronic toxicity test on *D. rerio*, for both ELT-dg and ELT-dp suspensions, comprised 30 days of exposure to 0.12, 0.37, 1.1, 3.3 and 10.0 mg/L under semi-static conditions, renewing the media 3 times *per* week [52]. The concentrations were chosen according to OECD 210 [52], and the highest concentration was adopted in accordance with both acute (LC50 > 100.0 mg/L) and fish embryo tests (FET; conducted as previous screening; no mortality was observed at 100.0 mg/L). During the test, we evaluated different end points, e.g., hatching, survival, abnormal behavior, weight and length of juvenile fish. For the embryo stage, fish were exposed in glass containers with 50 mL ELT-dg and ELT-dp suspensions, whereas for the juvenile stage, the volume was increased up to 1200 mL. Four replicates *per* treatment (20 eggs *per* well) were used. The exposure was conducted at a temperature between 25.1 and 26.0 ◦C, with 12 h light (602–853 lx) and 12 h darkness daily photoperiod, oxygenation > 60% and 7.75–7.88 pH range. Fish were fed *ad libitum* starting from 2 days after hatching (the amount was incremented during the test period; Gemma Micro 75, Skretting for larvae and Gemma Micro 150 for juveniles).

#### *2.5. Statistical Analyses and Data Elaboration*

To avoid confusion between the specific end points evaluated for each selected ecotoxicological test, we summarized and reported the main results as EC50, no observed effect concentration (NOEC) and lowest observed effect concentration (LOEC). The variance and distribution of data were verified by Bartlett and Shapiro–Wilk tests. For the growth inhibition test on *P. subcapitata*, the EC50 was determined by linear interpolation method for growth rate and by *non*-linear regression analysis for yield. The NOEC and LOEC for growth rate and yield were determined by Bonferroni correction test. For acute and chronic tests on *D. magna*, the EC50 was determined with linear interpolation, and NOEC and LOEC were determined by the Steel many-one rank sum test. For acute and chronic tests on *D. rerio*, the EC50 was measured by linear interpolation, and NOEC and LOEC were calculated, where possible, by Fisher exact test and Dunnett's multiple comparison test. CETIS v.1.8.7.7 software was used to carry out these analyses.

In addition, as further data elaboration, we performed a comparison between treated and control, time *versus* time (where possible) of selected end points. For this purpose, we used one/two-way analysis of variance (ANOVA) or a *non*-parametric test if normality or homoscedasticity were not verified, followed by Bonferroni correction test. STATISTICA 7.1 Software was used in these analyses.
