Competitive adsorption of tetracycline, oxytetracycline and chlortetracycline on soils with different pH value and organic matter content

1 Competitive adsorption of tetracycline, oxytetracycline and chlortetracycline on soils with different pH value and organic matter content Manuel Conde-Cid1, Gustavo Ferreira-Coelho2, Avelino Núñez-Delgado2, *, David Fernández-Calviño1, Manuel Arias-Estévez1, Esperanza Álvarez-Rodríguez2, María J. Fernández-Sanjurjo2 1 Department of Plant Biology and Soil Science, Faculty of Sciences, Campus univ. Ourense, 32004 Ourense, Spain. Universidade de Vigo 2 Department of Soil Science and Agricultural Chemistry, Engineering Polytechnic School, campus univ. s/n, 27002 Lugo, Spain. Universidade de Santiago de Compostela * Corresponding author E-mail: avelino.nune[email protected] (A. Núñez-Delgado) Tel: +34-982-823-140; Fax: +34-982-823-001. Declarations of interest: none. 2 Graphical Abstract 3 Abstract Antibiotics spread into the environment can cause soil and water degradation. Specifically, tetracycline antibiotics (TCs) are among those most consumed in veterinary medicine, and near 90% of the doses administered to animals are excreted as original compounds, due to poor absorption. In this study we investigated competitive soil adsorption/desorption for three tetracycline antibiotics (tetracycline: TC, oxytetracycline: OTC, and chlortetracycline: CTC), usually spread on soils by slurry fertilization, affecting to soil degradation due to chemical pollution. The study was carried out on six soils selected according to their pH values (4.49-7.06), and organic matter contents (1.07-10.92%). The competitive experiments were performed in ternary systems (adding all three TCs simultaneously, using five equal and increasing concentrations, from 17 to 200 μmol L-1). The results were compared with those obtained in simple systems (adding individual antibiotics separately), for the same final concentration (in this case, 200 μmol L-1) and for different concentrations (200 μmol L-1 in the simple systems, versus 600 μmol L-1 in the ternary systems, resulting from the sum of 200 μmol L-1 of each of the three antibiotics). In all cases, batch-type adsorption/desorption experiments were carried out, with 24 h as contact time. Those soils with higher organic matter content adsorbed 100% of the TCs, with desorption being always lower than 3%. In soils with less organic matter, adsorption decreased as the dose of added antibiotic increased, due to competition for adsorption sites. CTC was the most retained among the three TCs (up to 20% more than the other when high doses of antibiotic were added). In the simple systems, percentage adsorption was always high (> 85%) for the three TCs; however, percentage adsorption decreased in the ternary systems, reaching just 65% and 40% (for equal and different ionic strength, respectively) in soils with low organic matter contents. These results show the environmental and public health relevance of competition among the three TCs. In fact, the highest risk of entering the food chain takes place when these antibiotics are spread together on soils with low organic matter content, especially in the case of TC and CTC, which are the least adsorbed and the most desorbed molecules. Keywords: Adsorption-desorption, chlortetracycline, oxytetracycline, tetracycline, veterinary antibiotics 4 Funding sources supporting the work described in the manuscript Funding: This work was supported by the Spanish Ministry of Economy and Competitiveness [grant numbers CGL2015-67333-C2-1-R and CGL2015-67333-C2-2-R]. It also received funds from the European Regional Development Fund (ERDF) (FEDER in Spain), being a complement to the previous grants, without additional grant number. M. Conde-Cid holds a pre-doctoral contract (FPU15/0280, Spanish Government). The research of Dr. Gustavo F. Coelho was also supported by the Improving Coordination of Senior Staff (CAPES), Post-Doctoral Program Abroad (PDE) / Process number {88881.172297 / 201801} of the Brazilian Government. 5 1. Introduction Among the various causes of soil degradation, chemical pollution is one of the most relevant (Bridges and Oldeman, 1999). It is the case for emerging pollutants, and specifically for antibiotics spread into the environment, which can cause soil and water degradation (Bastos et al., 2018). In fact, agricultural and livestock productions are continuously increasing (mainly in developed countries) due to the growing demand for food on a global scale. The spreading of organic and inorganic fertilizers (including organic waste and by-products) is a common practice in agriculture, as is the veterinary use of antibiotics for the treatment of diseases, but also to prevent it, and as animal growth promoters (Charuaud et al 2019). Tetracycline antibiotics (TCs) are the most widely used in veterinary medicine, which is due to factors such as their low cost, broad spectrum and high antimicrobial activity (Daghrir and Drogui, 2013). In fact, the European Medicines Agency (2016) indicates that in 2014 the consumption of TCs represented as much as 33.4% of total antibiotics used in veterinary medicine. Due to their poor absorption when administered to animals, 80-90% of the doses of TCs are excreted as original (not metabolized) compounds in feces and urine (Kumar et al., 2005), most of them passing to slurry tanks, and to soil when spread as organic fertilizers (Conde-Cid et al., 2018a; Kivits et al 2018; Charuaud et al., 2019). Specifically, in a recent study carried out in Galicia (NW Spain) (Conde-Cid et al., 2018a) we found concentrations of 0.9, 4.0 and 35.0 mg kg-1 for the tetracycline antibiotics TC, CTC and OTC, respectively, in animal slurries, and up to 0.6 and 0.2 mg kg-1 for TC and CTC, respectively, in agricultural soils. Risks of environmental pollution due to the spreading of animal slurries have been a frequent concern (Núñez-Delgado et al., 2002), however, it is of specific growing relevance all that related to environmental pollution due to emerging pollutant, and specifically to antibiotics, with marked hazards for animal and human health (Kuppusamy et al., 2018; Peng et al 2019). The persistence of these antibiotics in the environment is affected by degradation processes (Conde-Cid et al., 2018b), while transfer to water bodies and plant uptake is also dependent on adsorption/desorption and transport processes (Fernández-Calviño et al., 2015a; Conde-Cid et al 2018a; Zhang et al., 2019). These processes are affected by physicochemical characteristics of antibiotics (such as molecular structure, size, solubility and hydrophobicity), and by soil characteristics, mainly pH and contents of components capable of retaining these pollutants (organic matter, clay, non-crystalline minerals) (Kemper, 2008). Specifically, the role of organic matter is of main importance, due to its large amount of pH-dependent functional groups, with the capability of having highly negative charge, allowing adsorption of positively charged antibiotics through electrostatic interactions. In addition, adsorption can also take 6 place by means of hydrogen bonds or cationic bridges through metal ions (Wang and Wang, 2015; OkaikueWoodi et al., 2018; Wang et al., 2018b; Zhang et al., 2019). The use of sorbent materials is being considered of great interest as a means to remove or retain pollutants (Xu et al., 2016; Anastopoulos et al, 2018; Anjum et al., 2019). Specifically, previous studies have been carried out focusing on antibiotics removal by means of sorbent materials (Chen et al., 2016; Li et al., 2019), even if, in most cases, antibiotics contained in animal slurries are directly spread on soils, using them as organic fertilizers. Some previous works dealt with TCs adsorption on soils, studying each antibiotic individually (simple systems) (Sassman and Lee, 2005; Jia et al., 2008; Figueroa-Diva et al., 2010; Wan et al., 2010; Zhang et al., 2010; Teixidó et al., 2012; Bao et al., 2013; Fernández-Calviño et al., 2015a, b; Li et al., 2015; Peng et al. al., 2019). However, there is a lack of studies focusing on competitive adsorption of antibiotics on soils, and specifically in cases where several TCs are incorporated simultaneously (multiple systems), which is common in real world when fertilizing with slurries (Conde-Cid et al., 2018a). Our research team used stirred flow chamber experiments to carry out preliminary studies on competitive adsorption for three TCs (TC, OTC, and CTC), although limited to two clearly acid soils (pH <4.5) (Fernández Calviño et al., 2015a, b). In addition, we have also verified the frequent simultaneous presence of different TCs in slurries and soils (Conde-Cid et al., 2018a), which indicates the need of performing broader studies on competitive adsorption for TCs in a higher number of agricultural soils with contrasted chemical characteristics. In view of that, the objective of this work is to study the competitive adsorption/desorption of three TCs (TC, OTC, and CTC) in six soils with different characteristics (notably, pH and organic matter content), comparing the results from simple systems (with presence of a single antibiotic) and ternary systems (simultaneous presence of three antibiotics), with equal and different ionic strengths, which as far as we know supposes an original research not previously performed. 2. Materials and methods 2.1 Study area Soil sampling was carried out in two agricultural areas in Galicia (NW Spain) with intensive farming activities: A Limia (Ourense province) (AL samples), and Sarria (Lugo province) (S samples). In the S zone, granite rocks and acid schists are dominant as lithology, while quaternary deposits dominate in the AL zone. Further details on both zones were provided in a previous work (Conde-Cid et al., 2018a), which 7 focused on the content of veterinary antibiotics (TCs and sulfonamides) in animal slurries, soils and crops in these two areas. In fact, six crop soils (of the total of the 65 analyzed in the mentioned previous work) were selected for the present study (three soils from AL, and three from S), which differ in pH (AL soils with pH between 4.5 and 4.8, and S soils with pH between 6.2 and 7.0), in organic matter content (three soils with less than 2% organic carbon, and three with 3.4%, 6.9% and 10.9%, respectively), and also differ in other parameters relevant in adsorption, as the non-crystalline Al and Fe contents (Table 1). To get each of the six soil samples (one per sampling zone), 15 subsamples were taken in each sampling point. The sampling depth was 0-20 cm, restricted to the surface layer of the soil. Sampling was carried out in a zigzag manner, by means of an Edelman probe. For each sampling point, subsamples were mixed to obtain a single (≈2 kg) representative sample. In the laboratory, soil samples were air dried (at ≈20 ºC for 7 days), sieved by 2 mm, homogenized and stored in polyethylene bottles till analyses. Triplicate determinations were carried out in all cases. 2.2 Characterization of soil samples All soil analyses were performed as per standard methods (Tan, 1996; Conde-Cid et al., 2018a). Sand (20.05 mm), silt (0.05-0.002 mm) and clay (<0.002 mm) fractions were quantified by means of the international method of the Robinson pipette. The pH of the soil was determined in water and in 0.1 M KCl (soil : solution = 1 : 5 ratio), using a pH-meter model 2001 (Crison, Spain). Organic carbon and total N contents were determined by elemental analysis, using a ThermoFinnigan 1112 Series NC instrument (ThermoFinnigan, The Netherlands). Exchangeable cations were extracted using 1 M NH4Cl and quantified by absorption/emission atomic spectrometry (AAnalyst 200, Perkin Elmer, USA). The effective cation exchange capacity (eCEC) was calculated as the sum of exchangeable Ca, Mg, Na, K and Al. The noncrystalline Al and Fe (Alox, Feox) were extracted with a solution of ammonium oxalate, acidified at pH 3 with oxalic acid, while the Al and Fe bound to organic matter (Alpir, Fepir) were extracted with sodium pyrophosphate at pH 10; the determination of Al and Fe was made in both cases by atomic absorption spectrophotometry. Available P was determined by UV spectrophotometry (model UV-1201, Shimadzu, Japan) after applying the Olsen method. All determinations were made in triplicate. 2.3 Adsorption and desorption experiments 8 Batch-type experiments were carried out. For the adsorption trials, 1 g of each soil was stirred for 24 h with 40 mL of 0.005 M CaCl2 solutions containing specific concentrations of the antibiotics. In the ternary system experiments (performed to study competitive adsorption), total antibiotic concentrations of 50, 100, 200, 400, and 600 μmol L-1 were used, with each of the three TCs (TC, CTC, and OTC) added to constitute 1/3 of each total concentration. In the simple system experiments (performed to study individual adsorption of each antibiotic), a concentration of 200 μmol L-1 was used for each antibiotic, separately. With these two kinds of experiments, it will be possible to compare the following: a) on the one hand, under conditions of equal ionic strength, adsorption in the simple systems (concentration of individual antibiotic added separately being 200 μmol L-1) versus the ternary systems (in that case where the total concentration added resulting from the sum of the three antibiotics is also 200 μmol L-1); and b), on the other hand, adsorption when the ionic strength of both kinds of systems is different (with concentration being 200 μmol L-1 in the simple systems, compared to the situation in which the concentration is 600 μmol L-1 in the ternary systems, resulting from the sum of concentrations of 200 μmol L-1 for each of the three TCs). After the stirring phase, all samples were centrifuged at 4000 rpm (6167 x g) for 15 min. In all equilibrium solutions, pH values were measured using a glass electrode (Crison, Spain), dissolved organic carbon (DOC) was determined by spectrophotometry (model UV-1201, Shimadzu, Japan), and the concentration of the three TCs was quantified as indicated in section 2.4 (see below). The amount of antibiotic adsorbed was calculated as the difference between the concentration added and that remaining in the equilibrium solution after centrifugation. All trials were performed in triplicate. Regarding desorption, 40 mL of 0.005 M CaCl2 were added to each of the samples used in the previous adsorption phase, then these samples were stirred for 24 hours and centrifuged at 4000 rpm (6167 x g) for 15 min. Subsequently, antibiotics released to the solution were determined. All trials were performed in triplicate. 2.4 Quantification of the three tetracycline antibiotics The procedure previously described by López-Peñalver et al. (2010) and Fernández-Calviño et al. (2015a, b) was used to quantify TCs, after slight modification. Briefly, all suspensions resulting from the adsorption and desorption experiments were subjected to HPLC liquid chromatography, by means of a Dionex apparatus (Dionex Corporation, Sunnyvale, USA), complemented with a P680 quaternary pump, an ASI100 auto-sampler, a TCC-100 thermostatized column compartment, and a UVD170U detector. Also, a Luna 9 C18 column (150 mm long; 4.6 mm internal diameter; 5 µm particle size), from Phenomenex (Madrid, Spain), and a guard column (4 mm long; 2 mm i.d.; 5 µm particle size), packed with the same material as the column, were used to carry out chromatographic separations. The injection volume was 50 µL, and the flow rate was 1.5 mL min-1, with a mobile phase integrated by acetonitrile (phase A) and 0.02 mol L-1 oxalic acid/0.01 mol L-1 triethylamine (phase B). A linear gradient elution program was run from 5 to 32% of phase A (and 95 to 68% of phase B) within 10.5 min. The initial conditions were re-established in 2 min and held for 2.5 min. The total analysis time was 15 min, with a retention time of 8.0 minutes. The wavelength used for TCs detection was 360 nm. More details can be seen in Fernández-Calviño et al. (2015a, b). 2.5 Data analysis and statistical treatment The Freundlich (Eq. 1), Langmuir (Eq. 2), and Temkin (Eq. 3) equations were used to describe data obtained in the adsorption experiments: 𝑞= 𝐾𝐶  (Eq. 1) 𝑞=  (Eq. 2) 𝑞= β ln KT + β ln Ce (Eq. 3) where qa (μmol kg-1) is the amount of each of the three TCs adsorbed onto the soil at equilibrium; Ceq (μmol L-1) is the concentration of each antibiotic present in the solution at equilibrium; KF (Ln µmol1-n kg-1) is the Freundlich affinity coefficient; n (dimensionless) is the Freundlich linearity index. In addition, KL (L μmol1) is a Langmuir parameter related to the adsorption energy, and qm (μmol kg-1) is the Langmuir’s maximum adsorption capacity of the soil. Finally, β = RT/bt with bt being the Temkin isotherm constant; Kt is the Temkin isotherm equilibrium binding constant (L g-1); T is Temperature (25 ºC) (K = 298 º), and R is the universal gas constant (8314 Pa m3/mol K). Furthermore, bearing in mind competition for adsorption sites, adsorption models could be modified, and an initial approach could take into account the total amount of antibiotics adsorbed (Eq. 4) 𝑄  + 𝑄  + 𝑄  = 𝐾𝐶  + 𝐶  + 𝐶   (Eq. 4) where 𝑄 , 𝑄  and 𝑄  are the amounts of TC, OTC and CTC adsorbed; 𝐶 , 𝐶  and 𝐶  are the concentration of TC, OTC and CTC remaining in solution at equilibrium; and 𝐾 and n are Freundlich’s parameters. 16 R2 ranging between 0.726 and 0.991 for Langmuir, and between 0.726 and 0.998 for Freundlich. Other authors have previously reported that TCs adsorption fits well with the Freundlich model (Zhang et al., 2010; Teixidó et al., 2012; Fernández-Calviño et al., 2015a, b; Li et al., 2015). Regarding the Langmuir model, the Qm parameter is clearly lower in soil 19AL (with low pH and low organic matter content) and, in general, higher in soils with higher organic C content and/or high pH values (soils 71S and 51S), indicating low probability of saturating TCs adsorption capacity for the latter (Febrianto et al., 2009). The Qm value follows the sequence TC < OTC < CTC, indicating higher affinity of CTC for the adsorption sites. These Qm values are between 2680.51 and 10627.91 μmol kg-1 (Table 3), being much higher than those previously reported by Teixidó et al. (2012) (between 587 and 4458 μmol kg-1), and somewhat lower than those indicated by Li et al. (2010) (between 12310.25 and 21339.16 μmol kg-1). In the Langmuir model, the KL parameter (related to the strength of interaction adsorbent/adsorbate), ranged from 0.032 to 0.47 L μmol-1. In previous studies, reported KL values were between 0.18 and 1.67 L μmol-1 (Teixidó et al., 2012), and between 0.038 and 0.067 L μmol-1 (Li et al., 2010). In the present study, KL values are 10 times higher in soil 71S (the one with the highest content in low crystallinity components, in addition to having a high organic matter content) than in the other five. Therefore, TCs bonds are stronger in soils with high contents in components with variable charge, which can be negatively charged, interacting through cationic bridges with the anionic groups of the CTs. The relevance of organic matter in these strong bonds is supported by the significant correlations (p <0.001) of the KL parameter with C (r = 0.907) for TC, as well as with C (r = 0.972) and N (r = 0.942) for OTC, and again with C (r = 0.975) and N (r = 0.941) for CTC. Other correlations were obtained with N (r = 0.865, p <0.05) for TC. Finally, correlations were also found with the C/N ratio for all three TCs (r = 0.791, 0.793, and 0.768, p <0.10, for TC, OTC and CTC respectively). Table 3: Parameters of the Langmuir model relating to adsorption process for the three TCs (TC, OTC, and CTC) and the A Limia (AL) and Sarria (S) soils studied Soil Antibiotic ------------------- Langmuir parameters ---------------------- Qm (µmol kg -1 ) Error KL (L µmol -1 ) Error R2 19AL TC 2412.33 246.67 0.27 0.15 0.894 OTC 3016.21 284.45 0.07 0.03 0.953 CTC 3572.10 443.77 0.07 0.04 0.936 3AL TC 6187.22 1163.39 0.08 0.05 0.991 17 OTC 8784.94 1465.22 0.04 0.02 0.968 CTC 7266.55 1332.28 0.07 0.04 0.946 50AL TC 6821.05 1765.99 0.78 1.12 0.726 OTC 7921.75 1756.07 0.68 0.76 0.813 CTC 8322.30 2131.18 0.69 0.70 0.825 51S TC 6094.51 583.70 0.08 0.02 0.974 OTC 9602.68 1189.57 0.04 0.01 0.983 CTC 10627.91 1064.82 0.06 0.01 0.991 6S TC 6283.88 1991.41 0.03 0.02 0.901 OTC 8571.64 1609.59 0.03 0.01 0.970 CTC 6178.86 799.17 0.03 0.01 0.972 71S TC 8175.84 589.99 0.44 0.09 0.989 OTC 8588.18 1108.38 0.43 0.18 0.966 CTC 9217.52 2314.47 0.47 0.32 0.935 Qm: maximum adsorption capacity; KL: parameter related to the strength of interaction adsorbent/adsorbate; R2: coefficient of determination In the Freundlich model, the n parameter indicates the reactivity of the active sites in the adsorbent. Table 4 shows that, for the present study, n values are lower in AL soils than in S soils, especially for CT. In all cases, n value is lower than 1 (between 0.14 and 0.58), coincident with results previously reported by Teixidó et al. (2012). This is related to non-linear and concave adsorption curves, and therefore to a decrease in adsorption sites as the concentration of TCs added increases, and could be associated to heterogeneous adsorption surfaces, with high-energy sites being occupied first (Sukul et al., 2008). Although not found in this study, values of n > 1 would correspond to high energy adsorption sites (Khezami and Capart, 2005; Foo and Hameed, 2010), with high accessibility of sorbates to the surface of the adsorbents (Skopp, 2009). In the present work, the n parameter was correlated with different soil characteristics for each antibiotic. In relation to CT, significant correlations were obtained with Feox and exchangeable Al (r = 0.846, and r = -0.813, respectively, p <0.05). It also correlated with pH(H2O) and pH(KCL) (r = 0.791, and r = 0.787, respectively, p <0.10), and with Alox and Fepir (r = 0.721, and r= 0.743, respectively, p <0.10). With respect to CTC, the n parameter also correlated significantly (p <0.05) with pH(H2O) (r = 0.896), pH(KCL) (r = 0.898) and exchangeable Al (r = -0.816). Finally, for OTC, correlations were found with pH(H2O), pH(KCL) and exchangeable Al (r = 0.748, r= 0.779 and r= -0.782, respectively, p <0.10). Thus, those soil parameters having the highest relation with the n value are pH and non-crystalline Fe and Al contents. 18 Table 4: Parameters of the Freundlich model relating to adsorption of the three TCs (TC, OTC, and CTC) in the A Limia (AL) and Sarria (S) soils studied Soil Antibiotic ------------------- Freundlich parameters -------------------- KF (µmol kg-1) Error n Error R2 19AL TC 756.18 45.94 0.26 0.01 0.994 OTC 551.63 29.78 0.34 0.01 0.997 CTC 642.72 40.75 0.36 0.02 0.997 TC+OTC+CTC* 1334.64 66.86 0.33 0.01 0.999 3AL TC 1125.69 110.78 0.38 0.03 0.991 OTC 906.49 143.84 0.48 0.04 0.987 CTC 1081.75 189.83 0.43 0.05 0.983 TC+OTC+CTC* 1920.67 271.37 0.44 0.03 0.997 50AL TC 4088.29 1826.52 0.146 0.17 0.726 OTC 4420.12 1712.49 0.17 0.15 0.813 CTC 4050.86 1463.13 0.25 0.18 0.825 TC+OTC+CTC* 10249.92 4462.77 0.18 0.13 0.951 51S TC 890.95 35.37 0.44 0.01 0.987 OTC 778.76 147.03 0.54 0.05 0.983 CTC 977.25 81.09 0.59 0.03 0.996 TC+OTC+CTC* 1425.08 167.99 0.52 0.03 0.998 6S TC 494.00 228.01 0.49 0.12 0.925 OTC 593.83 180.54 0.56 0.08 0.967 CTC 541.83 175.57 0.48 0.08 0.960 TC+OTC+CTC* 880.25 387.49 0.52 0.09 0.983 71S TC 2327.80 142.39 0.51 0.03 0.990 OTC 2560.60 172.08 0.45 0.03 0.992 CTC 2956.89 480.47 0.49 0.11 0.955 TC+OTC+CTC* 4596.01 198.83 0.48 0.02 0.999 KF: parameter related to the adsorption capacity; n: parameter related to the heterogeneity of the sorbent; Qd: maximum adsorption capacity; R2: coefficient of determination; * Obtained from Eq. 4 The Freundlich KF parameter, related to the capacity of adsorption in multilayers, showed a wide range of variation (between 494 and 4420 Ln µmol1-n kg-1), much higher than that previously reported by Teixidó et al. (2012) (KF between 240 and 1601 Ln µmol1-n kg-1). These values are higher in those soils showing higher organic matter contents (50AL and 71S); in fact, a significant correlation (p <0.01) was found between C and KF (r = 0.984, r = 0.989, and r= 0.983, for TC, OTC and CTC, respectively). In addition, Table 4 shows 19 that KF values in competition (obtained from Eq. 4) are always higher than those corresponding to individual antibiotics, and KF scores are once again higher for soil 50L, the one with the highest carbon content. In addition, results for the Temkin equation are shown in Table 5. In this model, adsorption is considered to be characterized by uniform distribution binding energies up to the maximum level (Ofomaja and Unuabonah, 2013). The Temkin equation considers adsorption heat, assuming a linear decrease in adsorption energy with surface occupation, which is related to adsorbent-adsorbate interactions. In this work, the Temkin model explains adsorption for all three tetracycline antibiotics in most cases, with R2 ranging from 0.884 to 0.992, even if fitting is not so good for soil 50AL. The Temkin model is considered appropriate for chemical adsorption based on strong electrostatic interactions between positive and negative charges, which makes clear the relevance of chemisorption processes in most of the soils here studied, also taking into account previous comments from Gao et al. (2012) and Rajapaksha et al. (2015). This fact complements the information derived from Freundlich’s n values, which were always <1, indicating a parallel relevance of physical adsorption in all cases. Thus, physical adsorption would be clearly dominant is soil 50AL (the one with the highest carbon content), while both physical and chemical adsorption would be relevant in the other soils here studied. Furthermore, Table 6 shows the good adjustment of CTC-OTC-TC adsorption data to the multiadsorbate model of Murali-Aylmore, taking into account that R2 values ranged from 0.943 to 0.999, with p <0.05. Table 5: Parameters of the Temkin model relating to adsorption of the three TCs (TC, OTC, and CTC) in the A Limia (AL) and Sarria (S) soils studied. Cases of poor fitting indicated by means of lowercase cursive figures Soil Antibiotic ------------------- Temkin parameters -------------------- bt Error Kf (L/g) Error R2 19AL TC 6629.55 39.91 6.82 3.53 0.988 OTC 5234.53 49.48 2.34 0.98 0.986 CTC 5078.94 74.93 4.50 2.97 0.970 3AL TC 3518.43 133.69 10.73 8.86 0.954 OTC 2922.06 193.26 7.14 6.58 0.931 CTC 2610.27 188.91 4.44 3.39 0.962 20 50AL TC 3024.68 941.73 116.54 934.52 0.884 OTC 2336.28 915.51 50.12 264.05 0.922 CTC 1640.45 550.79 12.20 18.26 0.985 51S TC 2527.06 124.68 2.13 0.84 0.978 OTC 2144.59 236.85 2.21 1.40 0.942 CTC 1768.32 228.58 2.34 1.00 0.961 6S TC 2618.27 242.78 0.74 0.57 0.945 OTC 1800.48 246.10 0.71 0.32 0.968 CTC 2048.92 166.20 0.45 0.16 0.982 71S TC 1582.58 111.43 6.63 1.14 0.992 OTC 1658.14 183.50 9.07 3.46 0.984 CTC 1349.50 379.86 6.34 3.57 0.975 bt: Temkin isotherm constant; Kt: Tempkin isotherm equilibrium binding constant (L/g); R2: coefficient of determination. Regarding Table 6, as previously commented, higher a values indicate that lower amounts of antibiotic are adsorbed in situations of competence. The highest a values (and thus the lowest adsorption) were those corresponding to TC (Eq. 5) in all soils. These a values also indicate that TC adsorption in competition is lower in the two soils with the highest carbon contents (soils 50Al and 71S), which would be due to higher competition and higher adsorption of the other two antibiotics. In the opposite side, CTC (Eq. 7) shows the lowest a values, indicative of the fact that CTC competes favorably with the other two antibiotics (and mainly with TC), thus adsorbing more intensively. Finally, OTC (Eq. 6) is in intermediate situation compared to TC and CTC. All that is coherent with results shown in Fig. 1, making clear the relevance of soil organic matter on competitive adsorption of the three tetracycline antibiotics here studied. Table 6. Fitting of adsorption results to Eqs. 5, 6, and 7 (adapted from the Murali and Aylmore (1983) equation), using TC, OTC and CTC solutions in relations 1:1:1, with adsorbed antibiotic concentrations expressed in µmol kg-1 and antibiotic concentrations in the equilibrium solution expressed in µmol L-1 Soil Antibiotic a R2 21 19 AL TC (OTC+CTC) ; Eq. 5 3.55 0.996* OTC (TC+CTC); Eq. 6 0.25 0.997* CTC (TC+OTC); Eq. 7 -4.02 0.964* 3 AL TC (OTC+CTC); Eq. 5 8.92 0.996* OTC (TC+CTC); Eq. 6 1.07 0.988* CTC (TC+OTC); Eq. 7 -7.68 0.984* 50 AL TC (OTC+CTC), Eq. 5 36.75 0.997* OTC (TC+CTC); Eq. 6 -1.39 0.999* CTC (TC+OTC); Eq. 7 -75.41 0.977* 51 S TC (OTC+CTC); Eq. 5 28.25 0.992* OTC (TC+CTC); Eq. 6 0.86 0.983* CTC (TC+OTC); Eq. 7 -72.65 0.976* 6 S TC (OTC+CTC); Eq. 5 1.98 0.958* OTC (TC+CTC); Eq. 6 0.12 0.971* CTC (TC+OTC); Eq. 7 -1.13 0.943* 71 S TC (OTC+CTC); Eq. 5 81.35 0.978* OTC (TC+CTC); Eq. 6 6.23 0.999* CTC (TC+OTC); Eq. 7 -104.25 0.969* a: parameter related to competence among all three antibiotics; * Significant at p<0.05 3.4 Desorption of the three tetracycline antibiotics in a ternary system For all soils, desorption values were always negligible for the two lowest TCs concentrations added, increasing as the concentration of added antibiotic increases. The maximum desorption value was 850 μmol kg-1, found for OTC when the highest dose of the antibiotic was added (Fig. 2). Expressing desorption in percentage of antibiotic released, compared to the amount previously adsorbed, values are low (generally less than 10%) for all three TCs in all soils, indicating that TCs adsorption is not easily reversible for the added concentrations. It should be noted that those soils with most organic matter (50AL and 71S) show high adsorption and low desorption, making clear the important role of the organic components in the strong retention affecting TCs, preventing their release to the soil solution, to water bodies, and plant uptake. Fernández Calviño et al. (2015a) also found low desorption (<15%) for these antibiotics from two crop soils from A Limia. In the present study, desorption percentage tends to a slight increase when increasing the dose of antibiotic added. This could be due to the fact that high-energy sites are occupied first when low concentrations of antibiotics are added, so that the resistance to desorption is greater (Zhang et al., 2010). OTC and TC desorb more than CTC (Fig. 2), which is in agreement with previous findings by FernándezCalviño et al. (2015a). Desorption values corresponding to the highest TCs concentrations added, in each soil, were subjected to an analysis of bivariate correlations with soil characteristics. The results showed a positive correlation (p <0.01) with pH(KCl) (r = 0.677), and a negative correlation (p <0.01) with available 22 P (r = -0.62). Therefore, TCs desorption would be favored by higher pH values. The negative correlation with available P could indicate a competition for the adsorption sites positively charged between phosphate anions and negatively charged TCs. Figure 2. Desorption results (in mg kg-1, and in percentage) for the three TCs (OTC, CTC, and TC) added simultaneously, in the six A Limia (AL) and Sarria (S) soils studied (from above, left to right: 19AL, 3AL, 50AL, 51S, 6S, 71S). Average values for three replicates, with error bars indicating that coefficients of variation were <5% 3.5 Comparison of TCs adsorption results in simple and ternary systems Firstly, the adsorption of each of the three TCs was compared between an individual system (for a concentration of 200 μmol L-1 of a single antibiotic), and a ternary system (all three TCs together, at the 23 same concentration of 200 μmol L-1 each, reaching a total sum of 600 μmol L-1). Since the final concentration in the ternary system is 600 μmol L-1, there will be different ionic strength with respect to the simple systems (with only one antibiotic at 200 μmol L-1). In a second phase, the simple and ternary systems were compared in situations where the ionic strength was the same, so that in the simple systems each of the TCs was added separately in a concentration of 200 μmol L-1, and in the ternary systems each antibiotic was added in a concentration of 66.67 μmol L-1, giving a final concentration of 200 μmol L-1. Figure 3 shows the adsorption results of these experiments for the six soils studied. Figure 3. Percentages of adsorption for the three TCs (TC, OTC and CTC) in the six A Limia (AL) and Sarria (S) soils studied (19AL, 3AL, 50AL, 51S, 6S, and 71S) in simple systems (each antibiotic added individually) and in ternary systems (the three antibiotics added simultaneously), with different ionic strength, and with the same ionic strength. Average values for three replicates, with error bars indicating that coefficients of variation were <5% 24 Simple systems show higher adsorption than ternary systems, both for situations with equal and different ionic strength (Fig. 3). For each antibiotic, differences in adsorption between simple and ternary systems are more pronounced when ionic strength is different. In simple systems, 100% adsorption is always reached in those soils having the highest organic matter contents (50AL and 71S), being always > 85% for the other four soils (19AL, 51S, 3AL and 6S) (Fig. 3). In the ternary system, adsorption is close to 100% for soils with high organic matter content, but it decreases for soils having lower organic matter, reaching 47% when the ionic strength is greater than that of the simple systems, and up to 65% when the ionic strength is the same. The decrease in the adsorption of one antibiotic in the presence of another can be due to a competition for adsorption sites (López-Peñalver et al., 2010), which is enhanced in situations where ionic strength is higher (Sun et al., 2010). This would be due to the decrease in the ratio cationic/zwitterionic species, with the latter having lower affinity for negatively charged soil surfaces (Figueroa et al., 2004). 3.6 Comparison of TCs desorption results in simple and ternary systems After studying adsorption, desorption experiments were carried out for the three TCs in the same simple and ternary systems, with equal and different ionic strength. The results are shown in Fig. 4. Desorption percentages were generally low, especially in simple systems (always lower than 5%) (Fig. 4). In ternary systems, desorption is higher in situations of higher ionic strength, reaching up to 20% of the amount of antibiotic previously adsorbed. Desorption is favored in soils with less organic matter content and higher pH (soils 6S and 51S). Also, in ternary systems CTC desorption increases in lower degree than that of TC and OTC, confirming the lower affinity of these two antibiotics for adsorption sites. Fernández-Calviño et al. (2015b) compared results corresponding to competitive desorption of the same three tetracycline antibiotics from two A Limia soils with non-competitive experiments previously reported in Fernández-Calviño et al. (2015a), finding that desorption percentages were higher in competitive trials, as in the current work, making clear that competition among TCs affects to bindings onto high energy adsorption sites, favoring desorption and increasing hysteresis. Conde-Cid et al. (2019) found that desorption results for the same three tetracycline antibiotics from soils of the same geographical origin as those here studied were also very low. These authors compared their desorption results with previous studied, such as those performed by Pils and Laird (2007), and by Munira and Farenhorst (2017), who reported values <1% for TC desorption from soils. Comparing with other antibiotics, Białk-Bielińska et al. (2012) reported clearly higher desorption results for sulfadimethoxine (from 13.4 to 96.8%), and for 25 sulfaguanidine (14.3-71.6%); whereas Hu et al. (2019) found desorption ranging from 30.3 to 52.1% for sulfadiazine, and from 43.7 to 48.3% for sulfamethoxazole. In addition, Rabølle and Spliid (2000) reported desorption values from 26 to 69% for tylosin (a macrolide). Regarding norfloxacin, ciprofloxacin and enrofloxacin (fluoroquinolones), Leal et al. (2012) found high retention, associated to desorption results <0.22%. Figure 4. Percentages of desorption for the three TCs (TC, OTC and CTC) in the six A Limia (AL) and Sarria (S) soils studied (19AL, 3AL, 50AL, 51S, 6S, and 71S) in simple systems (each antibiotic added individually), and in ternary systems (the three antibiotics added simultaneously), with different ionic strength, and with the same ionic strength. Average values for three replicates, with error bars indicating that coefficients of variation were <5% 3.7 Slurries, agricultural impacts and territorial management regarding TCs pollution Animal slurries spread on agricultural lands have been for decades a matter of concern regarding diffuse source pollution, mainly as regards some heavy metals, excess of nutrients favoring eutrophication, and 32 Li, C., Zhu, X., He, H., Fang, Y., Dong, H., Lü, J., Li, J., Li, Y. 2019. Adsorption of two antibiotics on biochar prepared in air-containing atmosphere: Influence of biochar porosity and molecular size of antibiotics. J. Mol. Liq. 274, 353-361. https://doi.org/10.1016/j.molliq.2018.10.142. López-Peñalver, J.J., M. Sánchez-Polo, C.V. Gómez-Pacheco, J. Rivera-Utrilla. 2010. Photodegradation of tetracyclines in aqueous solution by using UV and UV/H2O2 oxidation processes. J. Chem. Technol. Biotechnol. 85, 1325-1333. https://doi.org/10.1002/jctb.2435. López-Periago, E. Núñez-Delgado, A. Diaz-Fierros, F. 2000. Groundwater contamination due to cattle slurry: modelling infiltration on the basis of soil column experiments. Water Research 34(3), 10171029. https://doi.org/10.1016/S0043-1354(99)00226-2. López-Periago, E. Núñez-Delgado, A. Diaz-Fierros, F. 2002. Attenuation of groundwater contamination caused by cattle slurry: a plot-scale experimental study. Bioresource Technology 84(2), 105-111. https://doi.org/10.1016/S0960-8524(02)00041-X. MacKay, A.A, Canterbury, B. 2005. Oxytetracycline sorption to organic matter by metal bridging. J. Environ. Qual. 34, 1964-1971. https://doi.org/10.2134/jeq2005.0014. Munira, S., Farenhorst, A., 2017. Sorption and desorption of glyphosate, MCPA and tetracycline and their mixtures in soil as influenced by phosphate. J. Environ. Sci. Health B. 0, 1-9. https://doi.org/10.1080/03601234.2017.1361773. Murali, V., Aylmore, A.G., 1983. Competitive adsorption during solute transport in soils: 1. Mathematical models. Soil Science 135, 143– 150. https://doi.org/10.1097/00010694-198303000-00002. Núñez-Delgado, A., López-Periago, E., Dıaz-Fierros Viqueira, F. 1997. Breakthrough of inorganic ions present in cattle slurry: Soil column trials. Water Research 31(11), 2892-2898. https://doi.org/10.1016/S0043-1354(97)00145-0. Núñez-Delgado, A., López-Periago, E., Dıaz-Fierros Viqueira, F. 2002. Chloride, sodium, potassium and faecal bacteria levels in surface runoff and subsurface percolates from grassland plots amended with cattle slurry. Bioresour. Technol. 82(3), 261-271. https://doi.org/10.1016/S0960-8524(01)00183-3. Núñez-Delgado, A., Pousada-Ferradás, Y., Álvarez-Rodríguez, E., Fernández-Sanjurjo, M.J., Conde-Cid, M., Nóvoa-Muñoz, J.C., Arias-Estévez, M. 2019. Effects of Microbiological and Non-Microbiological Treatments of Sewage Sludge on Antibiotics as Emerging Pollutants Present in Wastewater: A Review. In (Shah, M.P., Rodriguez Couto, S., Eds.): Microbial Wastewater Treatment. Elsevier, Amsterdam, Netherlands. pp 1-18. 33 Official Journal of the European Union. 2013. Directive 2013/39/EU of the European Parliament and of the Council of 12 August 2013 Amending Directives 2000/60/EC and 2008/105/EC as Regards Priority Substances in the Field of Water Policy Text with EEA Relevance. Official Journal of the European Union. 2015. Commission Implementing Decision (EU) 2015/495 of 20 March 2015 Establishing a Watch List of Substances for Union-wide Monitoring in the Field of Water Policy Pursuant to Directive 2008/105/EC of the European Parliament and of the Council (Notified under Document C(2015) 1756) Text with EEA Relevance. Ofomaja, A.E., Unuabonah, E.I. 2013. Kinetics and time-dependent Langmuir modeling of 4nitrophenoladsorption onto Mansonia sawdust. J. Taiwan Inst. Chem. Eng. 44, 566–567. https://doi.org/10.1016/j.jtice.2012.12.021. Okaikue-Woodi, F.E.K., Kelch, S.E., Schmidt, M.P., Enid-Martinez ,C., Youngman, R.E., Aristilde, L. 2018. Structures and mechanisms in clay nanopore trapping ofstructurally-different fluoroquinolone antimicrobials. J. Colloid. Interface Sci. 513, 367–378. https://doi.org/10. 1016/j.jcis.2017.11.020. Peng, J., Wu, E., Wang, N., Quan, X., Sun, M., Hu, Q. 2019. Removal of sulfonamide antibiotics from water by adsorption and persulfate oxidation process. J. Mol. Liq. 274, 632–638. https://doi.org/10.1016/j.molliq.2018.11.034. Pikkemaat, M.G., Yassin, H., van der Fels-Klerx, H.J., Berendsen, B.J.A. 2016. Antibiotic Residues and Resistance in the Environment. Wageningen, RIKILT report 2016.009. 32 pp. Available at: https://doi.org/10.18174/388253. Pils, J.V., Laird, D.A. 2007. Sorption of tetracycline and chlortetracycline on Kand Ca-saturated soil clays, humic substances, and clay-humic complexes. Environ. Sci. Technol. 41, 1928–1933. https://doi.org/10.1021/es062316y. Rabølle, M., Spliid, N.H., 2000. Sorption and mobility of metronidazole, olaquindox, oxytetracycline and tylosin in soil. Chemosphere 40, 715-722. https://doi.org/10.1016/S0045-6535(99)00442-7. R Core Team, 2015. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/. Rajapaksha, A.U., Vithanage, M., Ahmad, M., Seo, D.C., Cho, J.S., Lee, S.E., Ok, Y.S. 2015. Enhanced sulfamethazine removal by steam-activated invasive plant-derived biochar. Journal of Hazardous Materials 290, 43–50. https://doi.org/10.1016/j.jhazmat.2015.02.046. 34 Sassman, S.A., Lee, L.S. 2005. Sorption of three tetracyclines by several soils: assessing the role of pH and cation exchange. Environ. Sci. Technol. 39, 7452–7459. https://doi.org/10.1021/es0480217. Skopp, J. 2009. Derivation of the Freunlich adsorption isotherm from kinetics. J. Chem. Education 86, 1341-1343. https://doi.org/10.1021/ed086p1341. Sukul, P., Lamshöft, M., Zühlke, S., Spiteller, M. 2008. Sorption and desorption of sulfadiazine in soil and soil-manure systems. Chemosphere 73, 1344-1350. https://doi.org/10.1016/j.chemosphere.2008.06.066. Sun, H., Shi, X., Mao, J., Zhu, D. 2010. Tetracycline sorption to coal and soil humic acids: an examination of humic structural heterogeneity. Environ. Toxicol. Chem. 29, 1934–1942. https://doi.org/10.1002/etc.248. Tan, K.H., 1996. Soil Sampling, Preparation, and Analysis. Marcel Dekker, New York. Teixidó, M., Granados, M., Prat, M.D., Beltrán, J.L., 2012. Sorption of tetracyclines onto natural soils: data analysis and prediction. Environ. Sci. Pollut. Res. 19, 3087–3095. https://doi.org/10.1007/s11356-0120954-5. USEPA. 2019. https://www.epa.gov/agriculture/laws-and-regulations-apply-your-agricultural-operationfarm-activity Wan, Y., Bao, Y., Zhou, Q. 2010. Simultaneous adsorption and desorption of cadmium and tetracycline on cinnamon soil. Chemosphere 80(7), 807–812. https://doi.org/10.1016/j.chemosphere.2010.04.066. Wang, R.Z., Yang, S.K., Fang, J., Wang, Z.Z., Chen, Y.Y., Zhang, D., Yang, C.Y. 2018. Characterizing the interaction between antibiotics and humic acid by fluorescence quenching method. Int. J. Environ. Res. Public Health 15(7), 1458-1471. https://doi.org/10.3390/ijerph15071458. Wang, S., Wang, H. 2015. Adsorption behavior of antibiotic in soil environment: a critical review. Front. Environ. Sci. Eng. 9, 565-574. https://doi.org/10.1007/s11783-015-0801-2. Xu, X., Gao, B., Jin, B., Yue, Q. 2016. Removal of anionic pollutants from liquids by biomass materials: A review. J. Mol. Liq. 215, 565-595. https://doi.org/10.1016/j.molliq.2015.12.101.Bao, Y., Wan, Y., Zhou, Q., Li, W., Liu, Y. 2013. Competitive adsorption and desorption of oxytetracycline and cadmium with different input loadings on cinnamon soil. J. Soils Sed. 13, 364–374. https://doi.org/10.1007/s11368-012-0600-3. Yao, F., Li, Z., Hao, X. 2016. Impacts of soil organic matter, iron-aluminium oxides and pH on adsorption desorption behaviors of oxytetracycline. Res. J. Biotechnol. 11, 121-131. 35 Zhang, G., Liu, X., Sun, K., Zhao, Y., Lin, C. 2010. Sorption of tetracycline to sediments and soils: assessing the roles of pH, the presence of cadmium and properties of sediments and soils. Front. Environ. Sci. Eng. 4, 421-429. https://doi.org/10.1007/s11783-010-0265-3. Zhang, D., Yang, S., Wang, Y., Yang, C., Chen, Y., Wang, R., Wang Z, Yuan, X, Wang, W. 2019. Adsorption characteristics of oxytetracycline by different fractions of organic matter in sedimentary soil. Environ. Sci. Pollut. Res. https://doi.org/10.1007/s11356-018-4028-1. Zhao, Y., Geng, J., Wang, X., Gu, X., Gao, S. 2011. Adsorption of tetracycline onto goethite in the presence of metal cations and humic substances. J. Colloid Interface Sci. 361, 247–251. https://doi.org/10.1016/j.jcis.2011.05.051.