A practical method for quantification of phosphorus- and glycogen-accumulating organism populations in activated sludge systems.

Enhanced biological phosphorus removal (EBPR) from wastewater relies on the enrichment of activated sludge with phosphorus-accumulating organisms (PAOs). The presence and proliferation of glycogen-accumulating organisms (GAOs), which compete for substrate with PAOs, may be detrimental for EBPR systems, leading to deterioration and, in extreme cases, failure of the process. Therefore, from both process evaluation and modeling perspectives, the estimation of PAO and GAO populations in activated sludge systems is a relevant issue. A simple method for the quantification of PAO and GAO population fractions in activated sludge systems is presented in this paper. To develop such a method, the activity observed in anaerobic batch tests executed with different PAO/GAO ratios, by mixing highly enriched PAO and GAO cultures, was studied. Strong correlations between PAO/GAO population ratios and biomass activity were observed (R2 > 0.97). This served as a basis for the proposal of a simple and practical method to quantify the PAO and GAO populations in activated sludge systems, based on commonly measured and reliable analytical parameters (i.e., mixed liquor suspended solids, acetate, and orthophosphate) without requiring molecular techniques. This method relies on the estimation of the total active biomass population under anaerobic conditions (PAO plus GAO populations), by measuring the maximum acetate uptake rate in the presence of excess acetate. Later, the PAO and GAO populations present in the activated sludge system can be estimated, by taking into account the PAO/GAO ratio calculated on the basis of the anaerobic phosphorus release-to-acetate consumed ratio. The proposed method was evaluated using activated sludge from municipal wastewater treatment plants. The results from the quantification performed following the proposed method were compared with direct population estimations carried out with fluorescence in situ hybridization analysis (determining Candidatus Accumulibacter Phosphatis as PAO and Candidatus Competibacter Phosphatis as GAO). The method showed to be potentially suitable to estimate the PAO and GAO populations regarding the total PAO-GAO biomass. It could be used, not only to evaluate the performance of EBPR systems, but also in the calibration of potential activated sludge mathematical models, regarding the PAO-GAO coexistence.     

The long-term effect of initial pH control on the enrichment culture of phosphorus- and glycogen-accumulating organisms with a mixture of propionic and acetic acids as carbon sources

In most studies on phosphorus- and glycogen-accumulating organisms (PAO and GAO), pH was controlled constantly throughout the entire anaerobic and aerobic periods, and acetic acid was used as the carbon source. In this paper, the effect of long-term initial pH values on PAO and GAO was investigated with mixed propionic and acetic acids as carbon sources. It was observed that with pH increasing from 6.4 to 8.0, the anaerobic propionic acid uptake rate by PAO linearly increased but that by GAO proportionally decreased. At pH 6.70 and pH 7.51, PAO and GAO exhibited the same acetic and propionic acid uptake rates, respectively. The acetic acid uptake rate by PAO was greater than that by GAO at pH > 6.70, and the propionic acid uptake rate by PAO was higher than that by GAO at pH > 7.51, which indicated that PAO would take predominance over GAO at pH > 7.51. Poly-3-hydroxybutyrate, poly-3-hydroxyvalerate and poly-3-hydroxy-2-methylvalerate shared 7%, 62% and 31%, respectively in the PAO system, and 11%, 44% and 45% respectively in the GAO system, and these fractions were observed independent of pH either in the PAO or in the GAO system. In the PAO system, with the increase of pH, the phosphorus removal efficiency was improved greatly, and a phosphorus removal efficiency of 100% was achieved at 8.0. Further investigation showed that the higher phosphorus removal efficiency at higher pH was mainly caused by a biological effect instead of chemical one.     

The long-term effect of initial pH control on the enrichment culture of phosphorus- and glycogen-accumulating organisms with a mixture of propionic and acetic acids as carbon sources

In most studies on phosphorus- and glycogen-accumulating organisms (PAO and GAO), pH was controlled constantly throughout the entire anaerobic and aerobic periods, and acetic acid was used as the carbon source. In this paper, the effect of long-term initial pH values on PAO and GAO was investigated with mixed propionic and acetic acids as carbon sources. It was observed that with pH increasing from 6.4 to 8.0, the anaerobic propionic acid uptake rate by PAO linearly increased but that by GAO proportionally decreased. At pH 6.70 and pH 7.51, PAO and GAO exhibited the same acetic and propionic acid uptake rates, respectively. The acetic acid uptake rate by PAO was greater than that by GAO at pH > 6.70, and the propionic acid uptake rate by PAO was higher than that by GAO at pH > 7.51, which indicated that PAO would take predominance over GAO at pH > 7.51. Poly-3-hydroxybutyrate, poly-3-hydroxyvalerate and poly-3-hydroxy-2-methylvalerate shared 7%, 62% and 31%, respectively in the PAO system, and 11%, 44% and 45% respectively in the GAO system, and these fractions were observed independent of pH either in the PAO or in the GAO system. In the PAO system, with the increase of pH, the phosphorus removal efficiency was improved greatly, and a phosphorus removal efficiency of 100% was achieved at 8.0. Further investigation showed that the higher phosphorus removal efficiency at higher pH was mainly caused by a biological effect instead of chemical one.     

Effects of selected pharmaceutically active compounds on the ammonia oxidizing bacterium Nitrosomonas europaea

Pharmaceutically active compounds (PhACs) are commonly found in wastewater influent. However, little research has focused on determining their impact on fundamental processes in wastewater treatment such as nitrogen removal. In this study, focus was placed on 4 commonly occurring PhACs (ketoprofen, naproxen, carbamazepine and gemfibrozil). Their effect was ascertained in the ammonia oxidizing bacterium (AOB), Nitrosomonaseuropaea in terms of membrane integrity and nitrite production. These PhACs were shown to inhibit nitrite production at concentrations of 1 and 10 μM while no effect was observed at 0.1 μM. The maximum observed nitrification inhibition was 25%, 29%, 22% and 26% for ketoprofen, naproxen, carbamazepine and gemfibrozil, respectively. A decrease in the live/dead ratio ranging from 10% to 16% suggests that these PhACs affect membrane integrity in N.europaea. The difference in nitrite production between PhACs treated cells and non PhAC treated controls was still significant following washing suggesting that inhibition is irreversible. Finally, nitrite production when adjusted to the live fraction of cells was also found to decrease suggesting that PhACs inhibited the activity of surviving cells. These results suggest that the presence of PhACs may affect AOB activity and may impact nitrogen removal, a key function in wastewater treatment. Follow up studies with additional AOB and in mixed culture are needed to further confirm these results.     

利用亚硝酸盐为电子受体反硝化聚磷菌的筛选与富集

依据DPB原理,利用SBR动态反应器和静态释/聚磷装置.以A2/O工艺厌氧段污泥为种泥,研究以亚硝酸盐为电子受体反硝化聚磷菌的筛选与富集,同时对选择、富集污泥的反硝化聚磷性能进行了考察.结果表明:利用亚硝酸盐为电子受体的反硝化聚磷菌存在于A2/O厌氧段污泥中,通过厌氧/好氧和厌氧/缺氧方式运行后,聚磷菌总数由1400个/mL增加到32 000个/mL,其中反硝化聚磷菌占聚磷菌总数的比例也由14.5%提高到81%,磷酸盐和亚硝酸盐去除率分别由最初的8.65%和7.55%上升到91%和95.62%;筛选与富集利用亚硝酸盐为电子受体的反硝化聚磷菌时,缺氧段进水COD的浓度须控制在10 mg/L以下;当体系处于稳定状态,且亚硝酸盐氮浓度高达30 mg/L时,并未对反硝化聚磷菌的生存产生抑制和体系运行产生干扰,此时磷酸盐出水低至1.06 mg/L.     

Identification of Chinese cabbage genotypes with low cadmium accumulation for food safety

The pot-culture experiment and field studies were conducted to screen out and identify cadmium (Cd) excluders from 40 Chinese cabbage genotypes for food safety. The results of the pot-culture experiment indicated that the shoot Cd concentrations under three treatments (1.0, 2.5 and 5.0 mg Cd kg⁻¹ Soil) varied significantly (p < 0.05), with average values of 0.70, 3.07 and 5.83 mg kg⁻¹, respectively. The Cd concentrations in 12 cabbage genotypes were lower than 0.50 mg kg⁻¹. The enrichment factors (EFs) and translocation factors (TFs) in 8 cabbage genotypes were lower than 1.0. The field studies further identified Lvxing 70 as a Cd-excluder genotype (CEG), which is suitable to be planted in low Cd-contaminated soils (Cd concentration should be lower than 1.25 mg kg⁻¹) for food safety.     

Hydrolysis of di-n-butyl phthalate, butylbenzyl phthalate and di(2-ethylhexyl) phthalate in human liver microsomes

Diester phthalates are industrial chemicals used primarily as plasticizers to import flexibility to polyvinylchloride plastics. In this study, we examined the hydrolysis of di-n-butyl phthalate (DBP), butylbenzyl phthalate (BBzP) and di(2-ethylhexyl) phthalate (DEHP) in human liver microsomes. These diester phthalates were hydrolyzed to monoester phthalates (mono-n-butyl phthalate (MBP) from DBP, mono-n-butyl phthalate (MBP) and monobenzyl phthalate (MBzP) from BBzP, and mono(2-ethylhexyl) phthalate (MEHP)) by human liver microsomes. DBP, BBzP and DEHP hydrolysis showed sigmoidal kinetics in V-[S] plots, and the Hill coefficient (n) ranged 1.37-1.96. The S₅₀, V_max and CL_max values for DBP hydrolysis to MBP were 99.7 μM, 17.2 nmol min⁻¹ mg⁻¹ protein and 85.6 μL min⁻¹ mg⁻¹ protein, respectively. In BBzP hydrolysis, the values of S₅₀ (71.7 μM), V_max (13.0 nmol min⁻¹ mg⁻¹ protein) and CL_max (91.3 μL min⁻¹ mg⁻¹ protein) for MBzP formation were comparable to those of DBP hydrolysis. Although the S₅₀ value for MBP formation was comparable to that of MBzP formation, the V_max and CL_max values were markedly lower (<3%) than those for MBzP formation. The S₅₀, V_max and CL_max values for DEHP hydrolysis were 8.40 μM, 0.43 nmol min⁻¹ mg⁻¹ protein and 27.5 μL min⁻¹ mg⁻¹ protein, respectively. The S₅₀ value was about 10% of DBP and BBzP hydrolysis, and the V_max value was also markedly lower (<3%) than those for DBP hydrolysis and MBzP formation for BBzP hydrolysis. The ranking order of CL_max values for monoester phthalate formation in DBP, BBzP and DEHP hydrolysis was BBzP to MBzP ? DBP to MBP > DEHP to MEHP > BBzP to MBP. These findings suggest that the hydrolysis activities of diester phthalates by human liver microsomes depend on the chemical structure, and that the metabolism profile may relate to diester phthalate toxicities, such as hormone disruption and reproductive effects.     

Modifications of black carbons and their influence on pyrene sorption

Sorption of pyrene on black carbons (BCs) obtained by heating sawdust at two temperatures (400 and 700 °C, denoted as 400BC and 700BC, respectively), as well as on modified BCs (via oxidation, oximation, and hydrolysis) was studied to investigate the role of BC structural characteristics in sorption of hydrophobic organic compounds. Pyrene was bound strongly by 700BC and 400BC, with organic carbon normalized distribution coefficients (K_oc) of 10^5.04-10^5.86 and 10^4.65-10^5.16, respectively, at equilibrium pyrene concentrations of 10-100 μg L⁻¹. Both chemical composition and pore distribution of the two BCs changed after modifications, which led to changes in their sorption characteristics for pyrene. After modifications, the linearity of pyrene sorption isotherm increased for 700BC but decreased for 400BC. For 700BC, both oxidation and oximation reduced pyrene sorption, with K_oc decreasing by 69.1-73.7% and 18.7-33.9%, respectively, whereas hydrolysis did not exert a significant influence. For 400BC, oxidation and hydrolysis reduced K_oc by 2.28-25.9% and 29.2-33.9%, respectively, while oximation increased K_oc. In most cases, the change in sorption capacity could be explained by the changes in C content and type, polarity, surface area, and micropore volume of the BCs; however, the role of conformation (the accessibility to sorption sites) could not be ignored.     

Effect of vegetation in pilot-scale horizontal subsurface flow constructed wetlands treating sulphate rich groundwater contaminated with a low and high chlorinated hydrocarbon

In order to characterize the effect of vegetation on performance of constructed wetlands (CWs) treating low and high chlorinated hydrocarbon, two pilot-scale horizontal subsurface flow (HSSF) CWs (planted with Phragmites australis and unplanted) treating sulphate rich groundwater contaminated with MCB (monochlorobenzene, as a low chlorinated hydrocarbon), (about 10 mg L⁻¹), and PCE (perchloroethylene, as a high chlorinated hydrocarbon), (about 2 mg L⁻¹), were examined. With mean MCB inflow load of 299 mg m⁻² d⁻¹, the removal rate was 58 and 208 mg m⁻² d⁻¹ in the unplanted and planted wetland, respectively, after 4 m from the inlet. PCE was almost completely removed in both wetlands with mean inflow load of 49 mg m⁻² d⁻¹. However, toxic metabolites cis-1,2-DCE (dichloroethene) and VC (vinyl chloride) accumulated in the unplanted wetland; up to 70% and 25% of PCE was dechlorinated to cis-1,2-DCE and VC after 4 m from the inlet, respectively. Because of high sulphate concentration (around 850 mg L⁻¹) in the groundwater, the plant derived organic carbon caused sulphide formation (up to 15 mg L⁻¹) in the planted wetland, which impaired the MCB removal but not statistically significant. The results showed significant enhancement of vegetation on the removal of the low chlorinated hydrocarbon MCB, which is probably due to the fact that aerobic MCB degraders are benefited from the oxygen released by plant roots. Vegetation also stimulated completely dechlorination of PCE due to plant derived organic carbon, which is potentially to provide electron donor for dechlorination process. The plant derived organic carbon also stimulated dissimilatory sulphate reduction, which subsequently have negative effect on MCB removal.     

The role of root hairs in cadmium acquisition by barley

The role of root hairs in Cd acquisition from soil was investigated in three pot experiments using a root hairless mutant (bald root barley, brb) and its wild-type (WT) cultivar of barley (Hordeum vulgare). brb had significantly lower concentrations and lower total amounts of Cd in shoots than WT. The Cd uptake efficiency based on total root length was 8-45% lower in brb than in WT. The difference between brb and WT increased with increasing extractable Cd in soil under the experimental conditions used. Additions of phosphate to soil decreased Cd extractability. Both soil and foliar additions of phosphate decreased root length, and root hair formation in WT. These effects resulted in decreased Cd uptake with increasing P supply. Cd uptake in WT correlated significantly with root length, root hair length and density, and soil extractable Cd. Root hairs contribute significantly to Cd uptake by barley.     

Biodegradation of pentaerythritol tetranitrate (PETN) by anaerobic consortia from a contaminated site

This study examined the role of denitrifying and sulfate-reducing bacteria in biodegradation of pentaerythritol tetranitrate (PETN). Microbial inocula were obtained from a PETN-contaminated soil. PETN degradation was evaluated using nitrate and/or sulfate as electron acceptors and acetate as a carbon source. Results showed that under different electron acceptor conditions tested, PETN was sequentially reduced to pentaerythritol via the intermediary formation of tri-, di- and mononitrate pentaerythritol (PETriN, PEDN and PEMN). The addition of nitrate enhanced the degradation rate of PETN by stimulating greater microbial activity and growth of nitrite reducing bacteria that were responsible for degrading PETN. However, a high concentration of nitrite (350 mg L⁻¹) accumulated from nitrate reduction, consequently caused self-inhibition and temporarily delayed PETN biodegradation. In contrast, PETN degraded at very similar rates in the presence and absence of sulfate, while PETN inhibited sulfate reduction. It is apparent that denitrifying bacteria possessing nitrite reductase were capable of using PETN and its intermediates as terminal electron acceptors in a preferential utilization sequence of PETN, PETriN, PEDN and PEMN, while sulfate-reducing bacteria were not involved in PETN biodegradation. This study demonstrated that under anaerobic conditions and with sufficient carbon source, PETN can be effectively biotransformed by indigenous denitrifying bacteria, providing a viable means of treatment for PETN-containing wastewaters and PETN-contaminated soils.     

Effect of fluorescent silica nanoparticles in embryo and larva of Oryzias latipes: sonic effect in nanoparticle dispersion

We characterized fluorescent silica nanoparticles (FNPs), which had been applied in many biological systems, in fish embryo rearing media (ERM) solution and evaluated the potential toxicity to the early development of Oryzias latipes embryos. Distribution of FNPs in embryos and larvae of O. latipes was studied by fluorescent and confocal laser scanning microscopic studies. Embryos exposed to three different concentrations of FNPs in stirred or sonicated ERM solutions were observed up to 2 d after hatching. FNPs had a negligible effect on the hatchability of O. latipes embryos; however, compared to controls, more than 30% of eggs were abnormal in 10 and 50 mg FNP L⁻¹ solutions. We found that the toxic effect was increased in sonicated FNP solution, which seems to be related with the dissolution of FNPs in ERM solutions that could be accelerated by sonication. Further study found that the CaCl₂ included in ERM solution might enhance the dissolution of the FNPs and the silicate ion released from FNPs partially contributed to larval toxicity. This study showed that some nanoparticles may not be stable in biological fluids even if they are stable in water. Dissolution factors such as sonication and cellular components should be considered in biological application of nanoparticles.     

In field arsenic removal from natural water by zero-valent iron assisted by solar radiation

An in situ arsenic removal method applicable to highly contaminated water is presented. The method is based in the use of steel wool, lemon juice and solar radiation. The method was evaluated using water from the Camarones River, Atacama Desert in northern Chile, in which the arsenic concentration ranges between 1000 and 1300 μg L⁻¹. Response surface method analysis was used to optimize the amount of zero-valent iron (steel wool) and the citrate concentration (lemon juice) to be used. The optimal conditions when using solar radiation to remove arsenic from natural water from the Camarones river are: 1.3 g L⁻¹ of steel wool and one drop (ca. 0.04 mL) of lemon juice. Under these conditions, removal percentages are higher than 99.5% and the final arsenic concentration is below 10 μg L⁻¹. This highly effective arsenic removal method is easy to use and inexpensive to implement.     

Performance of photocatalytic lamps on reduction of culturable airborne microorganism concentration

Reduction of viable airborne Staphylococcus epidermidis and Aspergillus niger spore concentrations using two types of photocatalytic fluorescent lamps under controlled environmental conditions (25 vs. 35 °C and 55 vs. 75% relative humidity) were investigated. Visible white-light and UVA black light were in-house spray-coated with TiO₂ and then compared with a commercially coated visible white-light for microbial concentration reduction. The white-light photocatalytic lamps reduced the concentration of culturable S. epidermidis up to 92% independent of temperature or humidity change, while the black light photocatalytic lamps completely inactivated the culturable bacteria at 25 °C, 55% relative humidity. Humidity seemed to alleviate UVA damage since better bacteria survival was found. For A. niger spores, rising humidity or temperature could lower their concentration or drop their culturabilities so that a difference between the natural decay and photocatalytic disinfection could not be distinguished. Reductions of total bacteria and total fungi concentrations using these lamps were also examined under uncontrolled environmental conditions in an office and a waste-storage room. It was found that photocatalytic lamps could reduce total culturable bacteria concentration from 9 to 97% and total culturable fungi concentration from 3 to 95% within irradiation time of 30-480 min, respectively. Insignificant difference in concentration reduction among these photocatalytic lamps was pronounced.     

Stoichiometric and molecular evidence for the enrichment of anaerobic ammonium oxidizing bacteria from wastewater treatment plant sludge samples

Anammox enrichments were readily developed from seven municipal wastewater treatment plants (WWTPs) sludge, but not with methanogenic granular sludge from two agro-industrial WWTPs. Only 50 d was required for the first evidence of anammox activity from a return activated sludge obtained from a WWTP operated for nutrient removal. The molar ratios of nitrite and ammonium consumption of approximately 1.32 as well as nitrate and dinitrogen gas product ratios of approximately 0.095 provided evidence of the anammox reaction. The presence of anammox was confirmed by polymerase chain reaction (PCR) using primer sets (PLA46F and AMX820R) specific for anammox bacteria. The 16S rRNA gene fragment of anammox bacteria was detected in seven enrichment cultures (ECs) with demonstrated anammox activity but not in the original inocula from which the ECs were derived and also not in the two methanogenic sludge samples, which indicates the PCR predicted the anammox activity. Two genera, Brocadia and Kuenenia, were successfully identified as the Planctomycetes occurring in the clone libraries of successful anammox enrichments. Brocadia dominated in cultures that were respiked extensively; whereas Kuenenia predominated in cultures that were less aggressively respiked. These findings indicate that respiking management may play an important role on selecting the genus of anammox bacteria. The batch enrichment results clearly illustrate that anammox can be readily enriched from municipal sludge from a wide variety of process operations at WWTPs.     

Removal of Ni(II) and Cu(II) ions using native and acid treated Ni-hyperaccumulator plant Alyssum discolor from Turkish serpentine soil

Alyssum discolor biomass was collected from serpentine soil and was used for removal of metal ions. The plant species grown on serpentine soils are known to be rich with metals ions and thus have more capability for accumulating heavy metals. Native and acid-treated biomass of A. discolor (A. discolor) were utilized for the removal of Ni(II) and Cu(II) ions from aqueous solutions. The effects of contact time, initial concentration, and pH on the biosorption of Ni(II) and Cu(II) ions were investigated. Biosorption equilibrium was established in about 60 min. The surface properties of the biomass preparations were varied with pH, and the maximum amounts of Ni(II) and Cu(II) ions on both A. discolor biomass preparations were adsorbed at pH 5.0. The maximum biosorption capacities of the native, and acid-treated biomass preparations for Ni(II) were 13.1 and 34.7 mg g⁻¹ and for Cu(II) 6.15 and 17.8 mg g⁻¹ dry biomass, respectively. The biosorption of Ni(II) and Cu(II) ions from single and binary component systems can be successfully described by Langmuir and Freundlich isotherms. When the heavy metal ions were in competition, the amounts of biosorbed metal ions on the acid treated plant biomass were found to be 0.542 mmol g⁻¹ for Ni(II) and 0.162 mmol g⁻¹ for Cu(II), the A. discolor biomass was significantly selective for Ni(II) ions. The information gained from these studies was expected to indicate whether the native, and acid-treated forms can have the potential to be used for the removal and recovery of Ni(II) ions from wastewaters.     

Field and laboratory studies of the fate and enantiomeric enrichment of venlafaxine and O-desmethylvenlafaxine under aerobic and anaerobic conditions

The stereoselectivity of R,S-venlafaxine and its metabolites R,S-O-desmethylvenlafaxine, N-desmethylvenlafaxine, O,N-didesmethylvenlafaxine, N,N-didesmethylvenlafaxine and tridesmethylvenlafaxine was studied in three processes: (i) anaerobic and aerobic laboratory scale tests; (ii) six wastewater treatment plants (WWTPs) operating under different conditions; and (iii) a variety of wastewater treatments including conventional activated sludge, natural attenuation along a receiving river stream and storage in operational and seasonal reservoirs. In the laboratory and field studies, the degradation of the venlafaxine yielded O-desmethylvenalfaxine as the dominant metabolite under aerobic and anaerobic conditions. Venlafaxine was almost exclusively converted to O-desmethylvenlafaxine under anaerobic conditions, but only a fraction of the drug was transformed to O-desmethylvenlafaxine under aerobic conditions. Degradation of venlafaxine involved only small stereoisomeric selectivity. In contrast, the degradation of O-desmethylvenlafaxine yielded remarkable S to R enrichment under aerobic conditions but none under anaerobic conditions. Determination of venlafaxine and its metabolites in the WWTPs agreed well with the stereoselectivity observed in the laboratory studies. Our results suggest that the levels of the drug and its metabolites and the stereoisomeric enrichment of the metabolite and its parent drug can be used for source tracking and for discrimination between domestic and nondomestic wastewater pollution. This was indeed demonstrated in the investigations carried out at the Jerusalem WWTP.     

Temporal evolution in PPCP removal from urban wastewater by constructed wetlands of different configuration: a medium-term study

Pharmaceuticals and personal care products (PPCPs) are widely distributed in urban wastewaters and can be removed to some extent by constructed wetlands (CWs). The medium-term (3-5 years) behaviour of these systems regarding PPCP removal is still unknown. Seven mesocosm-scale (1 m²) CWs of different configurations were operated outdoors for 39 months under the same conditions to assess their PPCP removal ability and temporal evolution. CWs differed in some design parameters, namely plant presence, species chosen (Typha angustifolia vs Phragmites australis), flow configuration and presence/absence of gravel bed (floating macrophytes surface flow, FM-SF; free-water surface flow, FW-SF; free-water subsurface flow, FW-SSF; or conventional horizontal subsurface flow, SSF). PPCP efficiencies decreased throughout time and performance differences among CWs disappeared with the systems aging. This could be due to a homogenization process in the systems caused by detrimental factors like saturation, clogging and shading. Winter efficiencies were lower than summer ones for salicylic acid, caffeine, methyl dihydrojasmonate, galaxolide and tonalide, and seasonal biological activities seem key factors to explain this fact. Maximal removal efficiencies were achieved in an unplanted-FW-SSF for ketoprofen (47-81%), naproxen (58-81%) and salicylic acid (76-98%); in an unplanted-SSF for caffeine (65-99%); in a Phragmites-FM-SF for ibuprofen (49-96%) and diclofenac (16-68%); in a Typha-FM-SF for carbamazepine (35-71%); and in a Typha-FW-SSF for methyl dihydrojasmonate (71-96%), galaxolide (67-82%) and tonalide (55-74%). Photodegradation could be involved in ketoprofen, naproxen, ibuprofen and diclofenac removal. Carbamazepine and diclofenac were moderately removed by the most efficient CWs studied. Carbamazepine might be eliminated by vegetal uptake.