3.4. Biotic Microcosms
As expected, no significant degradation of carbamazepine was observed in any of the microcosms (
Figure A1 in the Supplementary Materials). The concentrations of caffeine, paracetamol, metoprolol, and valsartan during the main experiment are presented in
Figure 1. The concentrations of atenolol acid and valsartan acid are presented in
Figure 2. As expected from the previous study [
11], the biodegradation rate of caffeine increased with increasing concentration of TE in the matrix. Despite the clear evidence for biodegradation, the analyzed caffeine TPs (mono- and dimethylxanthines) were not detected in any of the microcosms. This observation is in agreement with findings of [
32] who observed significant attenuation of caffeine in a karst aquifer but failed to detect mono- and dimethylxanthines. In addition to dilution effects they suggested fast degradation of the TPs itself or different degradation pathways as possible reasons. Valsartan was also much faster degraded in presence of 10% TE compared with the RW and smaller proportions of TE and the evolution of the valsartan TP valsartan acid followed this trend (
Figure 2). However, in contrast to caffeine and valsartan, the degradation rates of paracetamol and metoprolol decreased in the presence of TE. For paracetamol significant differences were observed at 1% and 10% TE compared to the pure RW. However, due to the sampling schedule the difference was only detected at one sampling time and, therefore, this observation should not be overestimated. The differences of metoprolol biodegradation were much clearer: No metoprolol was present in the RW microcosms after 21 days and the highest concentration of atenolol acid was observed. However, the biodegradation was strongly inhibited at presence of ≥0.1% TE. The impact of the lowest proportion was surprising, as the river water already contained wastewater at the specific sampling location. Similar observations regarding metoprolol degradation were reported by Hillebrand
et al. [
11]. However, in their study paracetamol demonstrated a faster degradation in pure TE than in RW.
Figure 1.
Concentrations of caffeine, paracetamol, metoprolol, and valsartan in the microcosms (RW: river water, TE: treated effluent). Mean values of two batches per time point, error bars represent the respective concentration range of duplicates.
Figure 1.
Concentrations of caffeine, paracetamol, metoprolol, and valsartan in the microcosms (RW: river water, TE: treated effluent). Mean values of two batches per time point, error bars represent the respective concentration range of duplicates.
Figure 2.
Concentrations of atenolol acid and valsartan acid in the microcosms (RW: river water, TE: treated effluent). Mean values of two batches per time point, error bars represent the respective concentration range of duplicates.
Figure 2.
Concentrations of atenolol acid and valsartan acid in the microcosms (RW: river water, TE: treated effluent). Mean values of two batches per time point, error bars represent the respective concentration range of duplicates.
An elevated total phosphorus (P) concentration was detected in the TE (384 μg·L
−1 vs. 67.1 μg·L
−1 in the RW) and, consequently, a significantly higher P concentration was present in the 10% TE experimental matrix. The same applies to micro-nutrients such as zinc (Zn) and molybdenum (Mo) (see
Table A1 in the Supplementary Materials). It is well known that P plays a key role in the biodegradation of environmental contaminants (e.g., [
33]), but no significant effects were observed during the additional experiment with native RW and RW spiked with additional PO
4−3 (increase to twofold concentration value) (see
Figure A2 in the Supplementary Materials). The impact of elevated Zn and Mo concentrations was not tested.
The presence of readily degradable matrix components can result in significantly higher transformation rates of pharmaceuticals residues (co-metabolism) but can also decrease the biotransformation rate of other compounds [
14]. From a quantitative point of view this issue is unlikely a reason for the here presented observations, as the TOC concentration of both RW and TE were in the same order of magnitude. Especially for the effects on metoprolol at the very low proportion of 0.1% TE it is very unlikely that co-metabolic processes were responsible. Elevated co-metabolism at 10% TE and thus higher degradation rates of caffeine and valsartan may occur, if a suitable substrate was added via the TE. However, the major task of WWTPs is the removal of the organic load from the wastewater, which is expressed by a decreased biochemical oxygen demand (BOD) of the TE [
34]. Therefore, the TOC of treated effluents usually consists of compounds persistent to biodegradation and thus, higher degradation rates due to co-metabolisms seem unlikely. WWTP final effluents are well-known point-sources of bacteria [
35]. The observed increasing degradation rates and especially the faster onset of biodegradation of caffeine and valsartan at increasing TE-proportions (
Figure 1) may have been induced by the addition of adapted microorganisms via the TE. Furthermore, the inoculated bacteria may have suppressed the autochthonous RW microorganisms, which were responsible for the degradation of paracetamol and especially metoprolol.
Other specific and chemically induced supporting/inhibiting mechanisms are also conceivable. Soluble microbial products (SMPs; humic and fulvic acids, polysaccharides, proteins
etc.), which are generated during wastewater treatment processes, can inhibit microbial growth [
36]. No acute toxicity was observed in the TE by the luminescent bacteria test. However, the SMPs demonstrate also metal chelating properties, playing probably a role in metal bioavailability, as they hinder the consumption of metal micro-nutrients by microorganisms [
36]. Furthermore, the bioavailability of the substrate also may be influenced by humic and/or surface-active substances [
37]. Further micro-contaminants were determined in the RW and TE as described in the literature [
25]. The concentrations of the analyzed compounds did not exceed the spike levels applied in the stabilization study conducted by Hillebrand
et al. [
11]. Therefore, these compounds were not responsible for the observed effects.
3.5. Environmental Relevance of the Study
The direct transferability of results from laboratory to field scale is often not possible. However, laboratory studies are necessary, as they allow for decreasing the complexity typically encountered in real ecosystems. The here applied initial concentration of micro-contaminants (100 μg·L
−1) was higher than typically encountered in surface waters [
38]. This concentration was chosen to achieve the analytical sensitivity for the here applied reporting limit (1 μg·L
−1; 1% of the initial concentration) without any pre-concentration step such as the SPE. Biodegradation rates of micro-contaminants can be concentration-dependent [
39]. However, as the half-lives of caffeine and paracetamol were in the same range as encountered in the preliminary experiments with lower initial concentrations (1 μg·L
−1) [
11], the here applied approach is justified. The here presented test design purposely neglected the impact of natural river sediments and sterilized gravel was offered solely as a substratum and not as a natural reservoir of microorganisms. Radke and Maier [
40] followed the opposite approach. They collected sediments from different rivers and locations within a river and used artificial river water for their experiments on the biotransformation of nine different compounds in water/sediment tests. They observed significant differences in biodegradation rates depending on the sampling location of the sediment. Similar to the here presented results Radke and Maier [
40] also encountered rather unsystematic results for the biodegradation rates of their compounds, e.g., one sediment was very effective in bezafibrate biodegradation while it was rather ineffective regarding ibuprofen and
vice versa for another investigated sediment. Furthermore, it is noticeable that sediment from river A was more efficient in removing micro-contaminants than sediment from river B, while the opposite pattern was observed in previous field studies at the investigated rivers. They concluded that physical boundary conditions are more important than the presence of specific microbial communities in the sediment. However, the here presented laboratory results demonstrate that even slight changes of the water phase by TE alone can result in substantial changes of biotransformation rates. The extent to which the here demonstrated impact is noticeable in water/sediment tests and real ecosystems as well as the identification of triggering factors for the observations needs to be addressed in future studies.