Mesocosm experiments were designed following the example of Carbonell et al. (2009). They consist of plastic pots of a height of 10 cm and an internal diameter of about 12 cm; a hole for drainage was made at the lower part of each pot where a nylon mesh was placed to avoid soil loss. Each pot had a system for the collection of leachates (a large beaker (600 mL) covered with tin foil to prevent light exposure). The pots were filled with unpolluted agricultural soil and lettuce seedlings were planted in the soil (one plant per pot; Fig. 1).
In this type of mesocosm systems, the soil-air interface, the water transport, and the kinetics of absorption/degradation are reproduced better than in standard soil bioassays. The system allows a realistic incorporation of CECs such as PPCPs, which resembles the agricultural practices expected from the use of reclaimed water.
A total of 24 pots were settled in a growth chamber using a completely randomized design with 6 replicate pots per treatment. The treatments consisted in irrigation with distilled water (CW), reclaimed water (RW), spiked water with 5000 ng L− 1 of PPCPs (SW) and spiked water with 5000 ng L− 1 of PPCPs but without plant (SW-P).
2.1. Chemicals, soil, plants, and waters
A total of 5 pharmaceutical compounds (1 psychiatric drug and 4 Nonsteroidal Anti-Inflammatory Drugs (NSAIDs)) and a cosmetic preservative were included in the spiked water (Table S.I.1). The selection was based primarily on previous studies of the research group and their occurrence in treated wastewater (Martínez-Alcalá et al. 2017). The pharmaceutical compounds selected for the experiment were the anticonvulsant and mood stabilizing drug carbamazepine, four non-steroidal anti-inflammatory drugs (NSAIDs) diclofenac, ibuprofen, ketoprofen and naproxen and the cosmetic preservative triclosan (CBZ, DCF, IBU, KTP, NPX and TCS, respectively).
The spiked concentration of each compound was 5000 ng L− 1 to facilitate measurements although it can be considered a severe contamination in treated water. Lower concentrations of PPCPs such as carbamazepine, diclofenac and naproxen have been previously detected (around 500 ng L− 1) in treated wastewater (Kim et al. 2007). However, Shenker et al. (2011) reported that in Israel the level of carbamazepine in treated wastewater used for irrigation could be as high as 3000 ng L− 1. The sources of the chemicals and the deuterated standards are described in the Supplementary Information (S.I.1). Individual stock solutions were prepared in methanol at 500 mg L− 1 and stored at -20 ºC in the dark. An intermediate combined solution was prepared in methanol with all the chemicals at a concentration of 2500 µg L− 1. The spiked water (SW) consisted on distilled water with5000 (ng L− 1) of each selected PPCPs (CBZ, DCF, IBU, KTP, NPX and TCS).
The reclaimed water (RW) was collected from the Murcia-East WWTP (UTMX (ETRS89): 670034, UTMY (ETRS89): 4207420), where an active sludge with modified A2O and disinfection treatment plant is running to 553451 population equivalents (data from 2019). The modified A2O process consisted of one anaerobic stage, two anoxic stages and two oxic stages. The reclaimed water contained 25.6, 27.3 and 134 (ng L− 1) of CBZ, DCF and KTP, respectively.
An uncontaminated calcaric fluvisol (IUSS 2015) was taken from the facilities of the Catholic University of Murcia (UTM coordinates: X: 659338; Y: 4206255, SE Spain) for this study. The soil (with clay loam texture) was collected from the surface (top 20 cm), air-dried and passed through a 2 mm sieve before its use in the experiment and the corresponding analyses. The main physicochemical characteristics of the soil were: silt loam texture (29.0% sand, 55.8% silt, 15.2% clay); pH 7.63; Eh 226 mV; 3.1% organic matter content; 0.22% total nitrogen; and 33.0% CaCO3 concentration. Pots were filled with 500 g (dry weight) of soil and introduced in a growth chamber under controlled conditions (22/18 ± 2 ºC, environmental humidity of 60%, photoperiod of 16/8 h light/dark and illumination between 250 and 300 µmol m− 2s− 1). Deionized water was added to the pots to adjust the overall moisture content to 35% of the water holding capacity (WHC) of the soil. The pots were irrigated with distilled water every 2 days before planting.
After 14 days of stabilization, the pots were wetted to 80% of their WHC with distilled water and then a mini romaine lettuce (Lactuca sativa var. Jabera) seedling of 1 month old, acquired from Deitana nursery (Murcia, Spain), was placed in each pot. From that moment, the pots were irrigated everyday with 100 mL of the corresponding treatment water. The experiment was carried out for 21 days under the same conditions in the growth chamber. Then, the plants were harvested and the different soil and leachate samples collected.
2.2. Sampling procedure and PPCPs analysis.
After plant harvesting (21 days), soil samples were taken from the pots, mixed to homogeneity, freeze dried (Freeze Dryer Christ alpha 1–2/LD plus), ground to a fine powder with a mortar, and then stored at -20°C until PPCPs extraction. Lettuce plants were split into roots and aerial part and rinsed with distilled water; half of the aerial part was weighted (fresh weight and dry weight after 24 h at 60 ºC in an aerated oven) and the other half was transferred to liquid nitrogen, freeze dried and ground to a fine powder, and then stored at -20°C until PPCPs extraction. The leachate samples (one sample per pot) were divided in two subsamples, one to analyze the PPCPs and another to perform an ecotoxicity bioassay with the aquatic crustacean Thamnocephalus platyurus.
PPCPs in soil were extracted following the method of Martín et al. (2010) with slight modifications. Briefly, aliquots (2 g) of lyophilized soil samples were accurately weighed directly in centrifuge tubes (12 mL). Afterwards, the samples were successively extracted with 5 and 2 mL of methanol and 2 mL of acetone. In each extraction step, the sample was vigorously shaken during 30 s, sonicated for 15 min and centrifuged at 4000 rpm (15.1 g) during 20 min. The supernatants from each extraction step were combined and evaporated to 0.2mL under a N2 stream (TurboVap LV concentrator). The extract was diluted to 250 mL with distilled water acidified to pH 2 with sulfuric acid and subjected to a cleanup procedure, where the aqueous mixture was loaded onto a hydrophilic-lipophilic balance cartridge (HLB, 60 mg, Dublin, Ireland), which was preconditioned with 3 mL of acetone, 3 mL of methanol and 3 mL of deionized water acidified to pH 2 with sulfuric acid. Samples were percolated through the cartridges using a vacuum manifold system (Waters) connected to a vacuum pump. The loaded cartridges were rinsed with 6 mL of water/methanol (95:5 v/v) and 3 mL of n-hexane. The elution was performed with three aliquots of 1 mL of acetone. The combined aliquots were evaporated to dryness by a gentle nitrogen stream and the evaporated extract was dissolved to a final volume of 1 mL with methanol.
Freeze dried and ground plant tissue samples were extracted and analyzed for PPCPs (Wu et al. 2012). Briefly, a 0.2 g (dry weight, DW) aliquot of plant sample was placed in a 50 mL glass centrifuge tube, spiked with deuterated PPCPs as recovery surrogates and then extracted with 20 mL of methyl tert-butyl ether (MTBE) in an ultrasonic water bath (50/60 Hz, Fisher) for 20 min, followed by centrifugation at 3000 rpm (15.1 g) for 20 min. The supernatant was decanted into a 40 mL glass vial and the residue was extracted once more using 20 mL of acetonitrile. The combined extracts were dried under N2 stream at 30°C and re-dissolved in 1 mL methanol, followed by the addition of 20 mL distilled water. The aqueous mixture was then loaded onto a HLB cartridge (60 mg, Dublin, Ireland), which was preconditioned with 7 mL of methanol and 7 mL of distilled water. After the cartridge was dried with N2, the analytes were eluted using 7 mL of methanol. The extract was further condensed under a gentle N2 stream and reconstituted to 1 mL with methanol.
In the leachates, PPCPs extraction was performed according to method described by Martínez-Alcalá et al. (2017). A volume of 500 mL of leachate samples was preserved by adjusting to pH 2 with concentrated sulfuric acid. Then the leachates were loaded onto HLB cartridges (60 mg, Dublin, Ireland), which were preconditioned with 5 mL of MTBE, 5 mL of methanol and 5 mL of distilled water. After that, the cartridges were rinsed with reagent water and eluted with 5 mL of 10/90 (v/v) methanol/MTBE followed by 5 mL of methanol. The resulting extract was evaporated to dryness under vacuum at 40–50 ºC using a TurboVap LV concentrator. Finally, the extracts were brought to a volume of 1 mL using methanol.
All the samples were filtered through polytetrafluoroethylene (PTFE) filters (13 mm, 0.2 µm, Millipore, Carrigtwohill, Cork, Ireland) before instrumental analysis. The final samples were analyzed using an ACQUITY UPLC Waters I-Class system (Waters Corporation, Milford, MA, USA) coupled to a Bruker Daltonics QToF-MS mass spectrometer (maXis impact Series, Bruker Daltonics, Bremen, Germany). Details of the instrumental analysis and quality control are provided in the supplementary information section (S.I.2).
2.3. Estimation of human exposure and acute toxicity tests
The human exposure (HE) to PPCPs through the dietary intake was evaluated according to the corresponding concentrations in the plants produced in the different water treatments. The daily human exposure for the PPCPs accumulated in the aerial part of lettuce was calculated using the formula by Beltrán et al. (2020) (Eq. 1):
-
HE = C·10− 6 × I
Where HE is the human exposure (mg day− 1), C is the concentration of the corresponding PPCP (ng g− 1, Fresh Weight (FW)) in the edible part of the plant (data obtained in the present mesocosm experiment), while I is the romaine lettuce daily intake recommended values (edible portion in grams (uncooked weight) per day) from the U.S. EPA (2018).
The acute toxicity test with the crustacean Thamnocephalus platyurus (Thamnotoxkit F, MicroBioTest Inc., Gent, Belgium) was performed according to the ISO norm 14380 (2011), using larvae hatched from cysts 20–22 h before the assay in diluted (1:8) standard freshwater at 25°C under continuous illumination (3000–4000 lx). So, for the analysis of the leachate samples, a 12 mL aliquot from each treatment was frozen and stored until its use. The tests were performed in 24-well plates, placing 10 crustaceans per well containing 1.0 mL of the corresponding test solution (leachates), in three replicates. Serial dilutions from 0 to 100% of the solutions were tested and the results are presented as a percentage (%) of survival.