Introduction

Antimicrobials, which are used to treat diseases, prevent infections, as well as promote growth, make up about 70 % of all used pharmaceuticals in veterinary medicine [1]. These compounds are not completely metabolized in the body and are excreted via manure [2]. Oxytetracycline (OTC) is a broad spectrum antibiotic which enters the microbial cells and binds to ribosome to prevent binding of aminoacyl tRNA, and is widely used to treat livestock animals due to its low cost and low side effects [3]. Its intense use generally results with its occurrence in manure and further in biogas plants when manure is used as a substrate [4]. Although there are several studies regarding effects of OTC on biogas production, studies on the effects of OTC on the microbiology of manure digesters are rather limited [4–6]. Moreover, most of the studies are conducted by adding antibiotics into the digester or using manure from antibiotic-fed animals, and information on the effect after intramuscular medication is almost absent despite intramuscular injection being one of the main medication methods. Therefore, the objective of this study is to determine the effect of OTC excreted in manure from intravascular medicated cows, both in terms of biogas production and microbial groups involved in reactors.

Materials and methods

Chemicals

Oxytetracycline (Mw = 460, CAS no. 79-57-2) was supplied by Acros Organics N.V (NJ, USA). HPLC grade methanol, acetonitrile, oxalic acid, and citric acid were obtained from Merck (NJ, USA). All other chemicals used in this study were of analytical grade. Double distilled water used in this study was obtained using a Millipore water purification system (Millipore Corporation, MA, USA).

Animal medication and manure sampling

Manure samples of a Holstein race (3.5-year-old, 440 kg body mass) dairy cow, which was kept in a barn belonging to the Faculty of Veterinary Medicine, Istanbul University, were used as the substrate for serum bottle tests. The manure in the rectum was collected and stored at 4 °C until future use as the control manure. The dairy cow was then medicated once with Oxytetracycline injection solution (20 mg OTC/kg animal weight) according to the standard dosage in veterinary practice. Equal doses were injected to the right and left body sides between the musculus semitendinosus and musculus semimembranosus muscles. Manure samples (around 1 kg) were collected from the rectum daily for 20 days and used throughout experiments after the OTC concentration was determined by HPLC.

Determination of extraction efficiencies for OTC and method of HPLC analysis

Manure samples were extracted in triplicate by a method modified from a previous study [7]. Briefly, 5 g of wet manure was placed into 50 mL polycarbonate centrifuge tubes with 0.5 g oxalic acid (C2O4H2·2H2O), 4 mL acetic acid, and 7.5 mL of 90 % methanol, and shaken at 100 rpm for 30 min. The tubes were further centrifuged at 11,000 rpm for 10 min. This procedure was repeated for 3 times and the supernatants were collected in 50 mL volumetric flasks. Collected supernatants (30 mL ± 5–10 %) were diluted to a 50 mL volumetric curve with double distilled water and centrifuged again at 14,000 rpm for 3 min and filtrated through 0.2 μm Millipore filters. The extracts were kept in 2 mL amber vials at −20 °C until the day of the HPLC analysis. The HPLC instrument (Schimadzu LC-10 AD) was equipped with an UV detector; (UV VIS Detector, SPD 10-A), an autosampler; (SIL-10 AD), a degasser (DGU-14A), and a system controller (SCL-10A). The column used in this study was Inertsil ODS-3 HPLC column (25 cm × 4.6 mm). Degassing of the solvents was achieved by sonication in a transonic ultrasonic bath (ELMA D-78224, Singen/Htw) prior to use. The mobile phase consisted of 75 % 0.1 M oxalic acid buffer and 25 % Methanol: Acetonitrile (1:1.5) solution which was delivered isocratically at a flow rate of 1 mL/min. The total run time was 30 min. The wavelength for the detection of oxytetracycline was 357 nm at which the retention time was 7.3 ± 0.1 min. The minimum detectable concentration was 0.01 mg/L.

All results were analyzed by the system software; Class VP Schimadzu Scientific Instruments Inc. In order to determine extraction efficiencies, triplicate samples of non-medicated manure were spiked with OTC in concentrations given in Table 1 and incubated for 3 h and extracted as described above. Recovery results shown in Table 1 were calculated as means of triplicate samples at each concentration.

Table 1 Recovery rates of OTC in manure
Full size table

Serum bottle tests

Manure samples collected after days 1, 2, 3, 4, 5, 10, 15, 20 of medication were used in the serum bottle tests. Experiments were carried out batch-wise in 120 mL serum bottles for 30 days with a working volume of 40 mL. Manure samples were diluted with tap water to total solids (TS) concentration of 5 % as a usual practice in commercial farm operations [8]. Seed was obtained from a lab-scale manure digester and added to the manure slurry at a ratio of 1:4 (Inoculum: Substrate). On operation days 0, 10, 20, and 30, biogas samples were collected from headspace, and samples for biogas and molecular analyses were collected every 10 days by sacrificing a set of serum bottles appointed to each sampling time. The experiment set was incubated in a temperature-controlled room at 37 ± 1 °C for 30 days on an automatic shaker operated at 120 rpm. All of the serum bottles were run in triplicate.

Total carbon, total nitrogen, Total Kjeldahl nitrogen (TKN), total solids, and total volatile solids (TVS) measurements were carried out according to the American Public Health Association's [9] guidelines. Characteristics of manure samples are given in Table 2. Gas pressures were measured using a manometer (HACH PM-9107) for calculations of biogas production. Gas compositions were determined using a Gas Chromatograph HP Agilent 6850 with a thermal conductivity detector and HP Plot Q Column (30 m, 530 μm). The carrier gas was Helium with an inlet column flow of 5.4 mL/min.

Table 2 Characteristics of manure samples used in serum bottle tests
Full size table

Determination of active microbial populations

Every 10 days of the operation, 5 mL of sample from the serum bottles was transferred into sterile containers, diluted with absolute ethanol (1:1, v/v) and fixed according to a protocol described previously [10]. Samples were taken from the control serum bottle and serum bottles operated with the 2nd, 10th, and 15th day manures as representative to the level of biogas inhibition. Hybridization and visualization of samples were carried out according to a previous study [11], except that 50–75 times dilution of the fixed samples was spotted on Teflon-coated slides. Probes used in this study were EUB338 [12], MB310, MG1200, MS1414, and MSMX860 [13] for the detection of Bacteria, Methanobacteriales, Methanomicrobiales, Methanosarcinaceae, and Methanosarcinales (complete acetoclastic methanogens), respectively. For the detection of non-specific bindings, NON338 probe was used as the negative control [14]. These probes were selected in reference to the most frequently detected groups of methanogens in biogas digesters. Results were reported as the total cell number which gave a positive signal to the specified probe per mL.

Statistical analyses

Significant differences were determined at 0.05 level by one sample t test by means of SPSS 11.5 (SPSS Inc., USA). Redundancy analysis (RDA) was applied in order to investigate the relation between microbial community dynamics and environmental variables (Canoco 4.5, Biometris, the Netherlands).

Results and discussion

Excretion of injected OTC in manure

Approximately 20 % of injected OTC was detected in the manure samples during 20 days of collection (Table 3). The total amount of OTC in the manure was calculated by multiplying the OTC concentration with the total manure production which is 118 g manure/kg body weight [15] daily. A manure sample which was collected on day 1 showed the highest OTC concentration (10.38 ± 0.22 mg/kg). On the 4th day of collection, there was an increase in the OTC concentration, but the level decreased after that rapidly until the 13th day of collection (Fig. 1). Although studies on the fate of OTC in manure after oral administration are plenty, information on intramuscular medication has not been reported. Our results indicated 80 % of OTC was either absorbed and/or degraded to metabolites; theoretically, the absorption rate of Tetracyclines can vary between 10 and 90 % [16]. According to similar studies reported in the literature, a total of 23 % of the oral-fed OTC was recovered from the manure of beef calves [17], and about 10 mg/kg OTC was detected in a fivefold diluted manure slurry of an orally medicated calf [16]. Tetracyclines have been found in manure as low as 0.1 mg/kg [18]. All the differences encountered in the literature can likely be due to administration type of the drug, sampling and storage conditions, the diets, general health of the animal, and type of animal.

Table 3 Daily OTC concentration in manure
Full size table
Fig. 1
figure 1

Concentrations of OTC for 20 days of manure collection

Full size image

Effects of OTC on anaerobic digestion performance

Serum bottle tests were run for 30 days, during which a total of 255 ± 13 mL biogas was obtained in the control bottles. At the end of 30 days of digestion, inhibitions in terms of biogas production compared to the control serum bottles were 46, 58, 57, 51, and 21 % for the manures collected on days 1, 2, 3, 5, and 10 after treatment, respectively (Fig. 2). Serum bottles operated with manure samples collected on days 15 and 20 were not significantly different from the control bottles in terms of biogas production (p < 0.05). The methane percentages in the biogas were 58 ± 5 for all serum bottles at all sampling times. In this study, 1.0–3.3 mg/L OTC in the slurries (calculated by dividing the OTC concentration of the daily manure samples by the dilution ratio) resulted in 50–60 % decreases in biogas production. In a previous study, 27 % inhibition in cumulative biogas production was reported in which OTC was given to beef calves orally and found in the manure slurry in a concentration of 3.1 mg/L [4]. In another study, tetracycline was found to inhibit methane production by 25 % in a swine manure digester [5]. In this case, although the highest concentration of OTC was detected after the first day of medication, the most severe inhibition on biogas productions was observed in serum bottles operated with manure samples collected on days 2 and 3. In serum bottles operated with manure samples collected after the 5th day of medication, the level of inhibition decreased. Manure samples collected on days 15 and 20 caused almost no inhibition compared to the control bottles. These outcomes can beneficially be utilized in practice in order to prevent possible decreases in biogas production; manures collected for the first 5 days after medication should be mixed with non-medicated manure and/or manures collected after the 10th day of medication prior to transition to biogas plants. However, in commercial barns, generally all of the animals are medicated, instead of just the sick ones, to prevent contamination of diseases. This can cause higher than expected levels of OTC in biogas plants, especially when manure is used as the sole substrate for biogas production. This challenge could be overcome by using other substrates such as agro-crops, bio solids, food wastes, etc. for co-digestion.

Fig. 2
figure 2

Cumulative biogas production in serum bottles after 30 days

Full size image

Changes in active microbial populations in serum bottles

Active microbial groups in serum bottles were characterized using fluorescent rRNA-targeted oligonucleotide probes specific for phylogenetically defined groups of methanogens and total Bacteria. The numbers of active cells detected by the specified probes per mL are given in Table 4. FISH results showed an increase in all samples until the 20th day of digestion after which the number of active cells decreased. Detected methanogens belonged to the groups of Methanobacteriales, Methanomicrobiales, and Methanosarcinales. The cells identified as Methanosarcinacea were almost equal to the number of cells identified as Methanosarcinales. Thereby, it could be assumed that Methanosaetaceae group was nearly non-existent in the serum bottles. The similar community structure of manure digesters have been reported in various sources [20–22].

Table 4 Numbers of active microbial cells in serum bottles (cells/ml)
Full size table

In order to evaluate results, redundancy analysis (RDA) was used as a statistical approach, which is known as the most generally effective ordination method for ecological community data, and was also used in similar studies determining microbial community Dynamics in biogas plants [19]. Figure 3 shows the RDA plot showing the FISH results where Eigenvalue was 0.829. In the case of cumulative percentage variance, 82.9 and 87.8 % of the species–species and species-environment relations, respectively, were explained.

Fig. 3
figure 3

RDA analysis of FISH results with environmental parameters

Full size image

RDA analysis showed that biogas production was significantly negatively correlated with OTC concentration as those arrows directed opposite to each other. Biogas production was found in a positive correlation with total Bacteria and order Methanomicrobiales in the serum bottles. Orders Methanobacteriales and Methanosarcinales explained less variance on biogas production. This indicates that most of the methane production was accomplished through the hydrogenotrophic pathway by Methanomicrobiales spp. The case has also been suggested previously where methanogenesis through the syntrophic association between hydrogenotrophic methanogens and bacteria such as Clostridium spp. [22–24] was reported in manure digesters [25]. The stability of methane percentages in the produced biogas from the serum bottles showed that there was a cease in carbon dioxide production as well. This can be explained by the simultaneous inhibitory effect of OTC on both bacteria and methanogens.

As seen, total bacteria and Methanomicrobiales were negatively correlated with OTC. Consequently, it can be said that effect of OTC was most viable on Methanomicrobiales, whereas Methanosarcinales and Methanobacteriales groups were comparatively less affected. However, it should be kept in mind that this population structure was reserved for batch systems, meaning that a washout of the most susceptible groups to oxytetracycline would have occured if the system was operated in a continuous manner. Future studies focusing on the microbiology in earlier stages of digesters exposed to tetracyclines would be enlightening on these matters.

Conclusion

In this study, a total 10 % of the administrated OTC was excreted in the manure. A discharge pattern could be monitored daily. Around 1–3.3 mg/L OTC caused a 50–60 % decrease in biogas production in which the methane percentage was stable. The collection time of manure was identified as an important factor upon transferring manure to the biogas plants. Redundancy analysis made for the evaluation of microbial dynamics with environmental parameters provided sensible results and revealed that a hydrogentrophic group, Methanomicrobiales, was mostly associated with biogas production and was the most affected group from the OTC. However, bacterial groups that mediate earlier stages should also be examined in future studies in order to reveal the effects on syntrophic associations.