Pilot study of biobeds application for the remediation of citrus agro-industrial wastewaters

ABSTRACT Agro-industrial wastewaters from citrus packaging plants are a significant source of pollution. Looking for a solution, lab-scale biobeds were evaluated for the remediation of fungicides from citrus agro-industry wastewaters. They were constructed under controlled conditions, using an ‘in-house’ designed bio-mixture containing soil, peat and cereal bran (1:1:2). The degradation of imazalil, prochloraz, propiconazole and pyrimethanil was studied individually, as a mixture and in real wastewater from a citric packing house facility. The pesticide behaviour was assessed using a novel validated dispersive sample preparation. The extraction was performed with ethylacetate and a salting out step with sodium tetraborate. The clean-up was performed with MgSO4 and Al2O3. The instrumental determination was carried out by LC-MS/MS in MRM acquisition mode. When studied separately, all the pesticides were dissipated. Particularly, propiconazole showed a 90% concentration reduction after 60 days. Two metabolites of prochloraz of environmental concern: 2,4,6-trichlorophenol and N-formil-N’-propil-N’-2(2,4,6 trichlorophenoxy) ethyl urea were detected, and their chemical structures were confirmed through mass spectrometry experiments. The bio-mixture was also efficient in dissipating the four pesticides when applied together in a single biobed. Real wastewaters from a citric agro-industry containing pyrimethanil and imazalil were tested using the same conditions. Both compounds dissipated 50% and 79% respectively, within 70 days.


Introduction
The agro-industry has developed postharvest strategies to expand the shelf life of fruits and vegetables (F&V).Particularly, the postharvest application of pesticides seeking F&V protection from insect and mould attack is a common practice.The use of fungicides in the citrus fruit-packing lines seeks the prevention of fungal proliferation, mainly Penicillium italicum and P. digitatum [1,2].The most used fungicides in citrus postharvest step are: imazalil, orthophenylphenol, pyrimethanil, propiconazole, prochloraz as well as thiabendazole [2][3][4].Each year the industry throws to the environment thousands of litres of contaminated effluents which are of great environmental concern [5][6][7].
Wastewaters produced during the application of postharvest pesticides either in the drencher step, in the packing-line or when washing the packing machinery [8,9] are stored in different reservoirs before their final deposition [10].Therefore, it is relevant to consider in which step of the postharvest process the compounds are used to treat and remediate them properly.
The environmental impact of citrus packing industry wastewaters has been scarcely investigated.Prochloraz, imazalil, propiconazole and pyrimethanil are classified as semi persistent and persistent compounds because their typical soil half-lives are 22.1-936.1;120-190; 15.3-96.3 and 30-55 days, respectively [11][12][13][14].For this reason, if they are not correctly disposed, they could stay in the environment for long periods before they are naturally degraded [14].Recently, thiabendazole, imazalil and pyrimethanil were detected in different agro-industrial wastewaters, some of their natural transformation products structures were elucidated and their toxicological relevances were evaluated [7].It is noteworthy to consider that the main degradation product of prochloraz: 2,4,6-trichlorophenol, is considered carcinogenic to animals and has been classified by NCBI as probable human carcinogenic [15].Furthermore, N-formil-N'-propil-N'-2(2,4,6 trichlorophenoxy) ethylurea is reported as moderately persistent in soil [14].There is a real need to face the remediation of the contaminated wastewaters from citrus production.Different alternatives for their remediation have been investigated in the last decade.There are many biological methodologies, such as biobeds [16], bioreactors or wetlands [17,18]; as well as chemical methods, for example, heterogeneous catalysis [19]; or physical, such as clays for the retention and degradation of polluting compounds [20].The physicochemical alternatives present some disadvantages such as cost-effectiveness, complexity, and huge quantity of sludge generation.Other new alternatives such as the use of egg shell and chitosan for organic pollutants removal in water have been evaluated also [21,22].
Particularly, biobeds are recognised as a good and 'green' bioremediation alternative to avoid punctual contamination in farms [10,23,24] The key of this technology is the bio-mixture composition.The first bio-mixture design from a Swedish investigation group, consisted of straw, soil and peat (50:25:25 v%) [8].The role of each component in the bio-mixture is specific and complementary.Peat provides sorption capacity to the biobed and allows moisture control.Soil also gives sorption capacity, and is the source of pesticide-degrading microorganisms.Straw is the main substrate for pesticide degradation and microbial activity [8].
The model has to be adapted to each situation depending on the characteristics of the different geographical regions, the treatments used in the various agricultural systems and the real situation of each grower.Different biobed designs´ and different bio-mixture compositions have been evaluated worldwide [25][26][27].Some reports show the degradation of imazalil and orthophenylphenol in citrus effluents using biobeds.In this system a biomixture with mushrooms (Pleurotus ostreatus), straw and soil was tested and the enhancement in the depuration capacity of this system was compared to a system without the mushroom.It was concluded that the bio-augmented composition did not ameliorate the degradation of the studied fungicides [28].In line with this observation, Perruchon et al. reported that a bacterial consortium degraded the fungicide thiabendazole under certain conditions [29,30].Moreover, Masis et al. could not obtain shorter half-lives for azoles using biobeds with a novel biomixture composition.For that reason, more research is still needed on the biodegradation of azole compounds like imazalil [31].
The performance of biobeds is evaluated through the decrease of the concentration of the contaminants under study, but the results must be supported by a validated analytical methodology.Although it is very important to study transformation products of pesticides, their presence is seldom evaluated.Being the bio-mixture a particularly complex matrix, fit for purpose protocols of pesticide analysis using chromatography coupled to tandem mass spectrometry were developed and validated in a case-by-case scenario [24,32].
As stated above, biobeds originally designed to hold back punctual contamination caused during their cycle of application in farms, could also be an environmentally friendly tool to treat the wastewaters generated in citrus packings to avoid pesticide contamination [10].The possibility of using biobeds for citrus wastewater remediation is an interesting alternative for Uruguay as citrus is one of its main productive chains.
The present work aimed to develop and evaluate the capability of biobeds, using an inhouse bio-mixture containing rice bran [33] instead of straw, as an environmentally friendly procedure to degrade the main citrus postharvest fungicides in wastewaters at lab scale.Specifically: the degradation of imazalil, pyrimethanil, propiconazole and prochloraz formulations in biobeds i) individually, ii) the mixture of the four fungicides, and iii) its applicability to a real citrus effluent were evaluated.The developed work is the first step to scale it up to real citrus wastewaters.

Chemicals and reagents
The pure analytical standards of imazalil, prochloraz, propiconazole, pyrimethanil and 2,4,6-trichlorophenol were purchased from Dr. Ehrenstorfer.Pesticide stock solutions (2000 mg L −1 ) were prepared in acetonitrile and used for the preparation of serial dilutions with concentrations ranging from 0.1 to 10 mg L −1 .All the solvents were HPLC grade from J. T. Baker (Darmstadt, Germany).A Direct-Q3 Ultrapure Water System from Millipore (Billerica, MA, USA) yielded high purity water for Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS) analysis.Sigma Aldrich (Steinheim, Germany) provided the formic acid (FA).
The salts for the salting out and clean up step: Na 2 B 4 O 7 and NaCl were purchased from Carlo Erba (Milan, Italy), MgSO 4 anhydrous and Al 2 O 3 were from J.T. Mallinckrodt Baker Inc. (Phillipsburg, NJ, USA) respectively.The used instruments included a SL16 centrifuge from Thermo IEC HN-SII (Langenselbold, Germany), a Turbovap® Biotage LV evaporator (Charlotte, NC, USA) a SHIMADZU AUX220 scale, readability 0.1 mg (Philippines), and a SHIMADZU TXB622L scale, readability 0.01 g (Philippine were employed during the sample preparation step.

Experimental design at lab scale
The whole experimental design was performed in the facilities of 'Estación Experimental Mario A. Cassinoni (EEMAC)' Paysandú, Uruguay.

Bio-mixture preparation
The bio-mixture employed was prepared by mixing topsoil, peat and cereal bran with the volumetric ratio 1:1:2 using a concrete mixer.Rice bran instead of straw was used in the evaluated bio-mixture according to previous results of our group [33].The rice bran was selected because it is cheap and readily available in the agricultural sector in Uruguay.Soil was collected from the top 20 cm of a field site used for agricultural experiments at the Estación Experimental Mario A. Cassinoni (EEMAC) (32°22ʹ51.46"S-58° 3ʹ13.72"O)belonging to a Vertic Argiudolls type, characterised by cation exchange capacity (CEC) > 25 meq/ 100 g; base saturation (BS) > 60% [34].The soil was sieved (5 mm mesh) and stored at 4°C until biobeds construction.Rice bran was acquired at CALSAL, Salto, Uruguay (local provider) and peat was from Kekkilä Ratatie 11, iVantaa, Finland.The prepared biomixture was analysed to evaluate the possible presence of the fungicides selected for the study.

Commercial formulations
In order to assess the dissipation of pesticides in experimental biobeds at lab scale, commercial formulations were sprayed in the first trial separated and in the second assay with the four pesticides together using a sprayer.
The used formulates were: Fruitgard IS 7.5 (imazalil 75 g L −1 ), Fruitgard PRO (prochloraz 100 g L −1 ), Fruitgard PZ 100 (propiconazole 100 g L −1 ) and Fruitgard PIR 400 (pyrimethanil 400 g L −1 ) from ENZUR S.A.All commercial formulations were analysed by LC-MS/MS and the label doses were verified.The chemical structures and the physicochemical properties of the studied fungicides are presented in Table 1.

Evaluation of the degradation of the four individual fungicides (assay A)
A bulk load (35 kg) of bio-mixture was prepared.For each fungicide, five plastic trays (48x25x8 cm) containing each one 1500 g of the prepared bio-mixture (5 replicates) were prepared and placed in a conditioned area (see supplementary material S1).Each one was individually treated with commercial aqueous solutions of the studied fungicides prepared following label instructions (see Table 1).The final concentrations of the applied fungicides at the application day were around 5 mg kg −1 for imazalil, 50 mg kg −1 for prochloraz, 10 mg kg −1 for propiconazole, and 20 mg kg −1 for pyrimethanil.These values are expressed in mg of active ingredient per kg of bio-mixture.The pesticide concentration to be tested was defined based on the evaluation of the occurrence of the selected fungicides in some citric packing effluents.In addition, two extra trays without any pesticide residue were prepared to be used as control.

Evaluation of the degradation of the four fungicides applied together (assay B)
In the second step, the 4 fungicides were applied as a mix.In this experiment, the whole procedure was the same as assay A.

Evaluation of the degradation in real wastewater (assay C)
The real wastewater for this assay was obtained from a composite sample of 40 L coming from a local citrus packing house.Six trays with 1500 g of bio-mixture were prepared.Five true replicates were treated with 500 mL of citrus wastewater and one sample was used as a control during the whole experiment with the same conditions.

Sampling
A scheduled timeline was defined for the sampling process in order to follow the behaviour of the studied fungicides (see supplementary material Table S2) until day 60 in each assay.The samples were taken using a soil sampler at five different points of each tray.The sub-samples were mixed to have a composite mixture for each tray.Two individual replicates of the composite sample were analysed at each time point.The first sample was taken as the initial concentration in the biobed for data treatment.

Sample preparation and instrumental analysis
The extraction step was adapted from a previous published work [33].It was miniaturised by just using 2 g of lyophilised sample that were extracted with 10 mL of ethyl acetate and sodium tetraborate was added for the salting out step.The dispersive clean up step was a QuEChERS modification [36] performed with MgSO 4 and Al 2 O 3 as sorbents.The final extract obtained was dried under N 2 stream and then re-suspended in 1 mL of acetonitrile and injected in the LC-MS/MS (QqQ) using the conditions described in supplementary material (S3).In particular, the identification of prochloraz degradation products was developed as described below: A high purity standard of 2,4,6-trichlorophenol was used for identification and characterisation of the detected degradation product, through the comparison of its chromatographic and spectrometric behaviour taking into account the legal residue definition for prochloraz [37,38].The identification of the prochloraz unknown degradation product was accomplished combining the information provided by LC-MS/MS operating in QqQ-MRM negative mode and the Ion trap MS full scan and Enhanced Product Ion (EPI) in positive ESI mode.The chromatographic conditions for the unknown degradation product identification were the same used for the targeted analysis of 2,4,6-tricholorophenol.The mobile phase employed throughout the whole study has been reported as the most suitable one for the analysis of this type of molecules [39].

Validation of the pesticide residues analysis methodology
To validate the analytical method for the determination of the targeted analytes; trueness, reproducibility and repeatability, limit of quantification (LOQ), linearity and matrix effects (ME) were evaluated following the EU, SANTE guidelines [40].
The figures of merit of the method were determined at four concentration levels (1, 2, 5 and 10 mg kg −1 ); each analysis was performed by quintuplicate.
Trueness was evaluated in terms of recovery percentage, using the criteria acceptance of the SANTE document (70-120%).Precision was determined by relative standard deviation (RSD% ≤ 20%); reproducibility and repeatability were evaluated.Intra-day variations for spiked levels were chosen to determine the reproducibility of the method (n = 5).For an inter-day precision test, the variation between two analysts was verified by quintuplicate.In all cases, variations were expressed as relative standard deviations (RSD %).The linearity was evaluated through the calculation of the r 2 and back calculated concentration (BCC) that was determined as the deviation of back-calculated concentration from true concentration (≤± 20%).The LOQ of the method was established as the minimum value where the trueness and precision are in accordance with the SANTE guidelines criteria (recoveries percentage from 70 to 120% and RSD lower than 20%).Matrix effects were studied comparing solvent and matrix matched calibration slopes.When the ME value is higher than 20% it is recommended to perform the quantification in matrix matched calibration curves.In this work the quantification was always performed using matrix matched calibration curves.The validation figures are shown in supplementary material Table S4.

Data treatment
Pesticide dissipation data were described by first order kinetics.The evaluation of the behaviour was done using the equation: Where the initial concentration of the pesticide at time 0 after application is C o , the dissipation rate constant (days −1 ) is k and C t is the pesticide concentration (in mg kg −1 ) at time t intervals (days).Then the half-life value was estimated using the typical equation: To evaluate the difference between the assayed bio-mixtures Student-t-tests at the contrasted level of probability (α = 0.05) were developed.

Results and discussion
The present work stems on the evaluation of biobeds performance for the four most used postharvest fungicides in citrus industry, replacing some of the original materials from the original Swedish bio-mixture.The bio-mixture employed was composed by soil, peat and rice bran in the volumetric proportion of 1:1:2.This type of variation changes the composition of the soil microbial community and thus the performance of the system [9].
Previous work from the group showed the capacity of rice bran to substitute straw as a lignocellulosic material [32].Cereal bran mixes better with soil and peat, yielding a more homogenous matrix for microorganisms' growth and an even distribution of the effluents within it.Additionally, the bran is a source of important metabolites such as vitamins and essential fatty acids that can support both, bacterial and fungal growth.Using this biomixture, the degradation rate of chlorpyrifos ethyl was faster than the reported using a classical Swedish model of biobed [8].The degradation rate of chlorpyrifos using the new bio-mixture was higher than that reported by Chilean researchers after varying different parameters while using a typical Swedish bio-mixture.In any case the conversion rate was higher than 50% [31].The chemical structure of the pesticides under study plays a role.Low transformation rates were reported for azole fungicides in biobeds even when different bio-mixture compositions were assayed [41].Reports of the performance of biobedswhen mixtures of compounds are applied in them are scarce.Also, for the case under study, the addition of mixtures of fungicides with different mode of action has not been reported.

Method validation
A fit for purpose analytical method was developed and validated specifically for the four evaluated fungicides with the Uruguayan bio-mixture, in order to obtain precise and accurate results of the pesticide behaviour.The methodology is an adapted miniaturisation of the procedure developed to evaluate chlorpyrifos degradation [33].Particularly, due to the physicochemical properties of the 4 fungicides, the pH was adjusted to higher values to force the deprotonation of the azole fungicides under study.The recoveries of all the studied analytes at the four concentration levels were between 72-107% and the precision of the method was evaluated yielding RSDs ≤ 16% for the 4 compounds.
The intermediate reproducibility for all the analytes was calculated at the 4 concentration levels studied through the RSD %.All figures were under 20%: for imazalil (<15.5%), for prochloraz (<9%), for propiconazole (10%) and for pyrimethanil (8%).The repeatability was evaluated by two operators in different days, and it was below 20% for all the compounds under study.
The linearity range of the whole validated methodology was studied through the calculation of the BCC that were lower than ±20%.The linear ranges were: 1-12 mg kg −1 for prochloraz, propiconazole and pyrimethanil and for imazalil it was 1-10 mg kg −1 .The LOQ of the method was fixed to 1 mg kg −1 , as the lowest concentration in which recoveries and RSD% were between the recommended values.
The ME values showed a small signal suppression for prochloraz (−28%) causing moderate matrix effects, which were negligible for the other compounds (<10%).Despite these results all the calculations were done using matrix matched calibration curves.All the evaluated parameters were according to those required by SANTE guidelines [40].All figures are presented in Table S4.

Degradation studies
The samples taken through all the experiment (see supplementary material S2) were analysed and dissipation curves were built for each pesticide.A summary of the degradation results for the 4 fungicides in the performed experiments is presented in Table 2.
A gradual and continuous concentration decrease was observed for all pesticides.The pesticide dissipation in all the experiments was described by a first-order kinetic.The r 2 showed the adequate fitting of the exponential decay model to the data obtained.

Degradation of the individual fungicides
Figure 1 shows the decay of the 4 fungicides when applied individually to each biobed.From the adjusted exponential curves, the half-life of each one of them could be calculated.Imazalil showed the longest t 1/2 (33d) followed by pyrimethanil (19d).Prochloraz and propiconazole under these conditions presented almost the same and shortest half-life of the studied fungicides of 10 and 9.2 d respectively.
The degradation percentages at 60 days were between 73 and 90% for the assayed analytes.Imazalil had the lowest degradation in 60 days (73%).The result was unexpected, considering that this pesticide is classified as persistent, being a difficult compound to degrade in the field.Furthermore, a wide range of half-lives have been reported for imazalil degradation, both in soils and biobeds, depending on the bio-mixture used and the experimental conditions [9,10,28,31].
Prochloraz and imazalil showed different degradation behaviour, although the two fungicides bear an imidazole structure.The lower Koc of prochloraz makes it more available for its degradation (see Table 1).Prochloraz is a derivative from 1-N (carboxy) imidazol, coupled to a dietil aminol etherified with trichlorophenol (TCP).All the degradation products of prochloraz that contain TCP are of toxicological concern and it is relevant to assess their presence/absence.In the experimental conditions of the present work, prochloraz degradation reached 82%.Trichlorophenol was first detected at 48 days post application.On the other hand, propiconazole is a 1,2,4 triazole coupled to an acetal of dichloroacetophenone that reached similar bioconversion levels.It was demonstrated that propiconazole metabolism in soils goes through the formation of polar and mobile compounds between two cropping seasons but also the acetal hydrolysis was observed [42].It was found a 92% of propiconazole dissipation at 60 days in the biobeds described in this experiment.The other postharvest fungicide, pyrimethanil, had shown high persistence, both in the industrial facilities and the fruit itself [2].However, under the assayed experimental conditions, pyrimethanil was also highly degraded (84%).

Degradation of the fungicides mix
The influence of applying the pesticides as a mixture on the degradation rates of individual compounds was evaluated to simulate the real situation.The experiment was conducted under the same conditions described for assay A. Figure 2 shows the degradation of the four evaluated fungicides when applied simultaneously in the biobed system.
The half-lives and degradation percentages for the studied pesticides were different when applied individually or as a mixture of the four compounds (see Figure 3).The fungicides added alone to the biobeds were degraded to a greater extent after two months than when added as a mixture.The fact can be explained on the bio-availability of the microbiota to easily degrade the compounds when they are alone In this case, the microorganisms select and degrade firstly those compounds that can be metabolised by their enzymatic system.Some structures are easier to assimilate than others, depending on the microorganisms present in the systems [43].
The degradation percentage of imazalil and propiconazole was lower when added together in the biobeds with the other fungicides than when studied alone.As shown in Figures 1 and 2 imazalil alone was degraded to a 73% in the biobeds after 60 days, but when assayed together the four fungicides, the dissipation drops to 37% overall degradation in the same period.Masís-Mora et al. reported the study of the remediation of triazines, triazoles and organophosphates using a homemade biobed at low concentrations (4 to 8 mg kg −1 ) resulting outstanding different degradation percentages for the three chemical families, but no degradation for after 128 days was reported [31].Imazalil has an initial degradation step followed by a plateau that was maintained until the end of the trial.Imazalil dissipation pattern is quite different from the other studied pesticides.Different half-lives for imazalil have been reported depending on the biomixture composition.Masis-Mora et al., registered a half-life of 198d using coconut fibre, compost, farm soil (2:1:1) [31] on the other hand, Omirou et al., reported a shorter half-life of 15.5d using grape wine compost [10].Nevertheless, Karas et al. found a similar period for imazalil decay to the half of the initial concentration as in the present experiment but using straw and soil as bio-mixture [28].
Comparing the degradation results in the two sets of experiments (Figure 3) the fungicide dissipation is always higher when they are applied individually compared to their application as a mixture.The fact can be explained based in the different modes of action of the fungicides under study.The combined application of fungicides with different mode of action had proved to be more effective to control pathogenic fungi than a compound alone.A similar mechanism could be operating in biobeds hampering the development of diverse and multiple degrader fungal strains.The fact needs further investigation, beyond the scope of this work.In addition, the biodegradation of the system could be by two factors: firstly, the optimum condition for the survival of the microbial consortium able to degrade pesticide, and the second one is related to the chemical structure of the compounds.The specific combination of the microbiota present in the bio-mixture and the organic compounds (alternative carbon substrates) to be degraded will show a different behaviour in each evaluated condition [44].
Nevertheless, the results show that biobeds under the evaluated conditions are a suitable tool for the mitigation of the residues still present in the effluents of citric industries.The settlement of biobeds at the end of the different processing lines could be an easy, cheap, and environmentally friendly solution to minimise the pesticide contamination in wastewaters from citrus industries.

Degradation of pesticides in the packing effluents
The effluent used for this study was obtained from a citrus packing company.It was analysed for the presence and concentration of these 4 which are the most used in the citric industrial technological package.However, in this case, it only contained imazalil and pyrimethanil and they were applied as such.During lab scale experiments with a real effluent, the half-lives of imazalil and pyrimethanil adjusted using a 1st order kinetic model were 70 and 30 days respectively.
The degradation percentages for these fungicides were 49 and 79% respectively after 70 days.It is interesting to point out that for both fungicides the percentage of degradation obtained for pyrimethanil in the effluent is similar to the one obtained when studying the pesticide mixture alone.The dissipation percentages of imazalil proceeded to values slightly higher than those of Assay B, despite the presence of other organic compounds in the effluent.The degradation profiles for the two fungicides were also different.The degradation curves of these two compounds are shown in Figure 4. Pyrimethanil degradation in real effluent As it can be seen, pyrimethanil concentration diminished continuously during Assay C. Again, imazalil behaviour deserves comment.In a similar sequence as when added as a mixture to biobeds, a first drop of 15% in the concentration of the fungicide was observed, that went into a plateau until day 28.But then, the imazalil dissipation continued, fact that was not observed in Assay B. If a dissipation kinetic is estimated after day 28, the half-life of the compound, assuming a first order decay (r 2 = 0.98) is 53 days.It has been reported that a lag time is needed before the microbiota in the biobed develops the enzymatic systems capable of degrading the compound [45].
Absolute values of half-lives in soils and the calculated ones in the performed experiments were compared.In all cases, the estimated fungicide half-lives were shorter than the reported DT50 for them in soil as shown in Table 1 [14].

Determination of the degradation products of prochloraz
The 4 targeted postharvest fungicides were identified using LC-MS/MS (QqQ) in MRM mode using the validated methodology.Also, the presence of 2,4,6-trichlorophenol was investigated qualitatively.It is one of the main metabolites of toxicological concern of prochloraz.Previously, positive and negative electrospray ionisation modes have been used to detect prochloraz and 2,4,6-trichlorophenol, respectively.The ESI (-) mode is preferred because 2,4,6-trichlorophenol is a weak acid and the negative ionisation mode is much more selective for its detection [46,47].
The obtained chromatogram using MRM in the negative mode showed not only the presence of the targeted metabolite: 2,4,6-trichlorophenol, which was characterised through the 4 identification parameters required by the SANTE guidelines, but also other peak showed the same transitions of trichlorophenol.Their retention times were at 8.32 and 8.64 min respectively.The elution pattern followed the already reported one [47].
Höllrigl-Rosta, A. et al. studied the metabolic fate of prochloraz in soil under field and laboratory conditions and reported the presence of two natural degradation products: prochloraz-formylurea and prochloraz-urea [48].The identification of the unknown degradation product was performed combining the information obtained by the QTRAP instrument.LC-ESI (-) QqQ in MRM mode, and the sequence full scan, precursor ion mode and the LC-QTRAP applying the and Enhanced Product Ion (EPI) [M + H] + functionality in ESI (+).
The precursor ion experiment of trichlorophenol allowed the identification of the [326, M + H] + ion.A full scan experiment detected this ion at 8.64 min.Then, an EPI experiment using the linear ion trap yielded the MS2 spectrum of N-formil-N'-propil-N'-2(2,4,6 trichlorophenoxy) ethyl urea (BTS 44596) permitting the selection of appropriate fragments with high sensitivity for MRM in positive ESI mode.
Cleavage of the imidazole ring from prochloraz could explain the appearance of BTS 44596.The fragment 310.6 m/z was detected in the EPI scan, which can be assigned to C12H14Cl3NO2 +.The combination of all these studies allows to assert that under the conditions of the experiment prochloraz degraded to two compounds: 2,4,6-tricholorophenol and N-formil-N'-propil-N'-2(2,4,6 trichlorophenoxy) ethyl urea also named BTS45186 and BTS44596 respectively.The obtained MS2 spectrum is shown in Figure 5.The experiment should be conducted for longer periods to evaluate the degradation of the compounds.

Conclusion
The studied bio-mixture showed the ability to degrade 4 of the most used post-harvest fungicides employed in the citrus agroindustry.The fungicides added alone to the biobeds were degraded after two months to a higher extent than when added as a mixture.The degradation of pyrimethanil and imazalil contained in real effluents reached very good values after 70 days.The results of this work showed the good biomixture performance.The settlement of these bioreactors at the end of the industrial line could be an easy, cheap and environmentally friendly solution.

Figure 1 .
Figure 1.Degradation curves of imazalil, propiconazole, prochloraz and pyrimethanil (Assay A).For each fungicide, five plastic trays with bio-mixture (5 true replicates) were placed in a conditioned area and individually treated with commercial aqueous solutions of the studied compounds.

Figure 2 .
Figure 2. Degradation of imazalil, prochloraz, propiconazole and pyrimethanil in assay B. The 4 fungicides were applied as a mix being the whole procedure the same as assay A.

Figure 3 .
Figure 3. Bioconversion percentage (%) for each fungicide at scale assays A and B.

Figure 4 .
Figure 4. Degradation curves of imazalil and pyrimethanil in a real from a citric production industry.(Assay C).

Figure 5 .
Figure 5. (a) Chemical properties of prochloraz and its metabolites BTS45186 and BTS44596.(b) Identification of the metabolites by LC/MS-MS Multi reaction monitoring (MRM) after negative ionisation.BTS45186 was identified at tR 8.32 min and BTS44596 at 8.64 min.(c) Enhanced product ion spectrum (EPI) at 8.64 min in positive ESI mode.

Table 2 .
First-order kinetics model and estimated half-life (t 1/2 ) for each combination fungicide-assay.Individuals formulate application (assay A), combined formulate application in bio-mixture (assay B) and real effluent application (assay C).