Constructed wetlands are one of the most appropriate wastewater treatment systems in mountain areas, where altitude, slope and climate constitute major environmental and economic constraints for infrastructure construction and subsequent management. In order to protect mountain natural wetland habitats that are sensitive to ecological equilibrium disruption, instead of the more commonly used macrophytes, plant species native to upland wetlands should be preferentially implemented as a contribution to biodiversity conservation and for the creation of more efficient, more resilient and better-adapted constructed wetlands. Carex paniculata is a key macrophyte in several European mountain aquatic habitats, and one of the few high-biomass producers that can grow at sea level and at altitudes of up to 2,600 m. In this paper, the results of a 2-year investigation demonstrate the efficacy of Carex paniculata for the treatment of the mixed stormwater, sewage and livestock wastewater effluents from a typical rural settlement at 825 m above sea level in the Cantabrian Mountains. The year-round suitability of Carex paniculata for the treatment of wastewater with seasonally variable flow and composition in mountain areas is demonstrated.
According to recent studies, more than 50% of small settlements in Spain, i.e. more than 6,000 small settlements, have no adequate wastewater treatment systems (Aragón et al. 2013). This problem is even more acute in rural and mountainous areas where, according to a recent study conducted by the Economic and Social Council (CES 2018) lack of sanitation systems has been determined to be the current main deficiency in terms of infrastructure provision.
Constructed wetlands (CWs) are one of the most efficient, environmentally sound, low-cost and resilient wastewater treatment technologies. However, a limited number of studies exist on CW implementation in mountain areas (Merlin et al. 2002; Elfanssi et al. 2018), and most of them are based on the more commonly used species, i.e., Phragmites australis and Typha spp., which rarely occur in certain European mountains, including northern Spanish ranges. Furthermore, their invasive behaviour might threaten the long-term survival of native mountain wetlands, bogs and fens, which are amongst the most species-rich and endangered aquatic habitats in Europe (Jiménez-Alfaro et al. 2011). Therefore, the efficacy of wastewater treatment by native species naturally occurring in mountain wetlands must be investigated to provide alternatives which are feasible and ecologically safe, as well as more resistant to extreme weather conditions.
One of the most promising native species in European mountains is Carex paniculata (greater tussock sedge), a long-lived, high biomass producer macrophyte species, which is characterised by its ability to form large tussocks and a dense and deep rhizosphere. Another strength of this species is its altitudinal plasticity, being naturally present in European wetland habitats from sea level up to altitudes of 2,600 m (Luceño & Castroviejo 1988). This makes it suitable for use at almost any altitude, and especially for CWs in mountain areas. Previous studies have reported a preference of Carex paniculata to water with high concentrations of NO3−, NH4+, and SO42− (Wassen et al. 1988; Lucassen et al. 2006). Furthermore, it has been suggested that NO3−, rather than P, is the limiting nutrient for Carex paniculata distribution (House et al. 2015).
In addition, some of the wetland types in which Carex paniculata is common, such as minerogenic mires, are Habitats of European Interest and are becoming rarer, even becoming relict in certain countries (García Rodeja & Fraga 2009). Therefore, Carex paniculata CW implementation for wastewater treatment throughout the territory would contribute not only to water quality improvement, but also to the preservation of those endangered wetland habitats.
However, despite Carex paniculata's apparent strengths for wastewater treatment and ecological conservation, there are a limited number of studies (Bauer et al. 2010), and specific information on Carex paniculata CW performance in mountain areas is non-existent. Therefore, Carex paniculata was selected for present investigation, which was carried out in Cantabria Autonomous Community, a mountainous region located in the north of Spain, which comprises more than 300 dispersed villages, 95% with less than 2,000 inhabitant equivalents (IEs). The area is representative of the natural constraints (Gloersen et al. 2004) and wastewater challenges in European mountain areas.
The overall aim of this research was to investigate Carex paniculata's potential and efficacy for the treatment of mixed wastewater of seasonally variable composition (stormwater, sewage and livestock wastewater inputs) in rural settlements in mountain areas, as well as studying its seasonal behaviour, and the cumulative growth of the persistent parts (roots and tussock) of the plants over two annual cycles under experimental conditions.
To tackle this challenge, a Carex paniculata CW was set up in a small rural settlement in Valdeprado village, in the Cantabrian Mountains, and the results of the CW performance over 2 years are presented in this article.
MATERIALS AND METHODS
Population equivalent seasonality
Valdeprado is a remote rural village located at 825 m above sea level within Liébana Site of Community Importance protected area (north of Spain), whose main economic activity is extensive livestock farming following the traditional management cycle known as transterminancia.
Transterminancia is the seasonal short-distance altitudinal movement of cattle between villages and communal upland grasslands in order to make best use of the pastures in each zone and season. From October to May, the cattle are housed within villages located in the valleys, where they graze in lowland grassland surrounding the villages and/or are fed with hay harvested during the previous summer. From mid-May to the end of September, the livestock are moved to graze freely in upland grasslands in the mountains (Delgado 2010).
Like most municipalities in the Cantabrian Mountains, Valdeprado has a single sewage system. Therefore, in addition to domestic wastewater and stormwater runoff entering the sewage system from the streets, livestock wastewater produced seasonally in the cowsheds according to the transterminancia cycle represent an important wastewater stream to be treated in the single sewage system. From October to May the average population equivalent (PE) is 80–100 IEs, whereas in summer, when the cattle are in the mountains, the average PE decreases to 30–46 IEs, estimated from a permanent population of 17 IEs and a variable visitor population of 10–90 IEs during weekends and summer vacations.
Experimental facilities set up
A wastewater treatment plant was constructed in June 2015, consisting of a pre-treatment, a primary treatment and a CW (Figure 1). Pre-treatment comprised a concrete channel with a bar screen for large material removal and sand and gravel sedimentation which, equipped with a by-pass, also could function as storm water spillway in case of extreme water flow events. The spillway was directly connected to the outlet pit, thus redirecting the excess of stormwater flow in certain scenarios, (i.e., snowmelt from November to March, and in occasional extreme events such as concentrated runoff discharges during rainstorms) preventing the septic tank and CW from overflowing. Primary treatment was undertaken in a 1,800 L capacity prefabricated septic tank with a grit decanter and a grease and oil separator. Next, a flow distribution pit, equipped with a secondary stormwater spillway for by-passing the CW during maintenance operations, divided wastewater flow equally into two basins; each had a 30 m2 surface, 0.60 m floodable depth and 20,000 L capacity, with a theoretical average hydraulic retention time (HRT) of 5 days. The basins were dug into the ground and waterproofed with PVC sheets to create the two CWs, which could work in series or in parallel (the latter being the configuration for this research), enabling the simultaneous development of different investigations.
The present investigation was conducted in one of the basins, where a Carex paniculata constructed wetland was set up with a size ratio of 2 m2 per IE, for an average population of 30 IEs. The size of the constructed wetland was agreed with local farmers, for a maximum number of cattle equivalent to 8 IEs. However, the actual number of cattle housed was equivalent to 50 IEs, increasing the average PE during 9 months of the year to 80 IEs, and the actual size ratio of the CW at 0.75 m2 per IE.
Tussocks of mature Carex paniculata plants were manually taken from a nearby natural wetland and washed to remove the soil from the roots. Each tussock was divided into clumps with a circumference of approximately 70 cm, producing a total of 52 bare-rooted plants that were put into drilled plastic pots at the bottom of the basin, with their root systems submerged and leaves growing above the water surface, thus functioning as a hydroponic system. Planting density was 1.25 plants per m2, reproducing the population density observed in nature.
Wastewater treatment efficacy and Carex paniculata vegetative development monitoring was conducted from March 2016 to December 2017. The Carex paniculata was monitored every 8 weeks during the first year, measuring leaf length and recording the main key stages, i.e., flowering, fruiting and senescence, estimated as the proportion of green and dry leaves. Aboveground biomass was harvested in February. Cumulative growth was estimated using tussock circumference converted into diameter and root length measurements at the start and end of the project.
Water was sampled on a bimonthly basis from March to December each year, with an extra sample taken on the 17th of August, coinciding with maximum visitor population (‘MaxP’ in Figure 2 and ‘Mx’ in Figures 5–8) during local festivities. Samples were taken in the inlet pit (raw wastewater), in the flow distribution pit (after primary treatment) and in outlet pit (after CW) (called 1, 3 and 5 in Figure 1).
On-site analyses were conducted through the use of portable measurers for temperature (T) and pH (multimeter Crisson-MM40), dissolved oxygen (DO) (oxygen measurer Hanna-HI 9146) and turbidity (turbidimeter Eutech-TN_100). The analyses for conductivity (cond.), biological oxygen demand (BOD5), chemical oxygen demand (COD), total suspended solids (TSS), total nitrogen (TN), ammonia (NH3), nitrates (N-NO3), total phosphorus (TP), phosphates (P-PO4), total coliforms, Escherichia coli and grease and oil (G&O) were conducted by the Centre of Environmental Research of the Government of Cantabria (CIMA) Certified Laboratories following standard methods (Rice et al. 2017).
PE was calculated as the sum of people and housed cattle (bearing in mind that one cow is equivalent to 3 IEs).
Annual average concentrations of T, pH, cond., DO, turbidity, TSS, G&O, BOD5, COD, TN, NH3, N-NO3, TP, P-O4, total coliforms and E. coli in raw wastewater (RW), after pre and primary treatments (PT), and after CW (CW), were calculated from a total of seven sampling campaigns per year, six of which were conducted on a bimonthly basis.
Total bimonthly precipitation (TBP) was calculated as the sum of total monthly precipitation using official data recorded by the State Meteorological Agency (AEMET) at the weather station located in Valdeprado village. This provided an indicator for the stormwater in raw wastewater.
RESULTS AND DISCUSSION
The influence of total bimonthly precipitation (or stormwater) and population equivalent seasonality on the quality of the wastewater was investigated (Figure 2).
While maximum PE occurred when livestock was housed in the village during the winter months, selected pollution indicators such as BOD5, COD and TSS did not reach their maximum values during these months.
This was because, despite the existence of a spillway in the pre-treatment to prevent overflow and system collapse, a substantial part of the high stormwater flow still passed through the system, leading to a dilution of domestic and livestock wastewater constituents, and to the reduction of pollution loads, especially in winter campaigns, when stormwater flow notably increased due to rain and snowmelt. In contrast, maximum annual contamination loads were reached in October, when the presence of the cattle in the village coincided with minimum total bimonthly precipitation and stormwater inflow.
Visual observations (see Figure S1 in Supplementary Information, available with the online version of this paper) revealed that livestock wastewater from the cowsheds was a mixture of urine, liquid slurry from the washing down of the cowsheds and dunghill leachates. It is suggested that the bad practice of frequently dumping manure in the general stormwater collection system lead to certain high pollution load peaks, even coinciding with high TBP values, as would be the case in, for example, March 2017.
During summer, when the cattle were absent and the total bimonthly precipitation was lower, the contamination load was mainly from the domestic wastewater input. Noticeable peaks were recorded at MaxP_17: 68 mg/L (BOD5), 230 mg/L (COD) and 103 mg/L (TSS), when the maximum human population occurred (Figure 2), which coincided with the highest visitor population period in Valdeprado during local festivities.
Carex paniculata annual cycle and cumulative growth
The annual cycle of Carex paniculata is illustrated in Figure 3. Following the pruning of leaves in February, continuous and vigorous growth of aboveground biomass occurred until October, when leaves reached the maximum annual length of 2.21 m and the first signs of senescence were noticed, starting from leaf's apexes then progressively extending to the whole leaf towards December.
By December, completely senescent leaves reached a 30% of the total aboveground biomass, whereas in the following winter months the proportion of senescent leaves increased, but never exceeded 50% of the total aboveground biomass, demonstrating Carex paniculata's semi-evergreen character. Flowering occurred from April to May, and fruiting from June to July.
A noticeable growth of the perennial parts of the CW plants occurred during the 2 years of continuous operation. The cumulative increase of the perennial parts (tussocks and roots) from a 12% representative sample from Valdeprado CW plants was recorded. The cumulative development, monitored as the diameter of the tussock, the average length of the roots and the maximum length of the roots changed from 22.31 cm ±2.51 to 35.00 cm ±5.36; from 34.49 cm ±4.47 to 40.49 cm ±8.61; and from 47.17 cm ±10.80 to 59.51 cm ±10.94, respectively from June 2015 to July 2017. This represented a cumulative increase of 56.88%, 17.39%, 26.16% in tussock diameter, average root length, and maximum root length from June 2015 to July 2017 respectively.
These results suggested a greater investment in lateral growth than deep root development. Various Carex species have been reported to be characterized by highly developed surface and underground zones which have an impact on the retention of pollutants (Parzych et al. 2017).
Annual average treatment system efficacy and treatment stages contributions
As annual average results presented in Figure 4 (see Supplementary Information for results of main quality parameters, available online) indicate, the complete wastewater treatment system (CWTS) substantially reduced all the parameters of concern during the experiment. A clear increase of CWTS efficacy was observed in the second year, which might be related to the system maturity and to the specific cumulative growth of the tussocks and rhizospheres in the CW, as in general terms CW performance was revealed as the main treatment stage determining CWTS efficacy.
CWTS highest removal efficiencies were achieved for turbidity, TSS, BOD5, COD and G&O, reaching annual average removal efficiency values from 70% to 97% in both years, mainly attributed to CW as Figure 4 indicates, even though pre-treatment and septic tank performances increased noticeably and were also remarkable in year two, especially related to BOD5 and COD degradation.
Lower CWTS performance for nutrient and microorganism removal observed in the first year increased noticeably in 2017, reaching annual average removal efficiencies around 50% for TP, P-PO4, TN, NH3, N-NO3, total coliforms and E. coli. These results suggested that TP, P-PO4, total coliform and E. coli removal mainly occurred in the Carex paniculata CW, while the removal of TN, NH3 and N-NO3 occurred in both the septic tank and the CW.
In 2016 no oxygenation efficacy was observed from raw wastewater to the final effluent. As Table S1 (available online) indicates, this occurred as a result of a higher consumption of DO in the septic tank than the wastewater oxygenation performed by the CW, which led to a negative DO concentration balance throughout the complete system, even though an annual average of 8.23% DO increase took place in the CW. By contrast, in 2017 a remarkable increase of the complete system performance up to a 65% DO increase was recorded, and, as can be observed in Figure 4, this was mainly attributed to CW oxygenation, which was estimated as an annual average of 44% DO increase.
Biological oxygen demand, chemical oxygen demand and total suspended solids seasonal removal
Snowmelt led to lower BOD5, COD and TSS loads in March 2016 and December 2017, and despite the HRT being limited by high stormwater flow, removal efficiencies reached to 87% and 50% for TSS, and 41% and 15% for COD degradation, whereas BOD5 values were under the detection limit (Figure 5).
Regardless of the high influent contamination by livestock wastewater and lower HRT in March 2017, 96%, 91%, 83% removal efficiencies were achieved for BOD5, COD and TSS respectively, suggesting that the CW system efficacy persisted in winter even under high contamination loads and lower HRT. By contrast, higher removal efficiencies were recorded in summer campaigns, with maximum annual degradation efficiencies of 88–96% for BOD5, 88–93% for COD, and 94–97% for TSS in October each year, coinciding with maximum contamination loads, lower precipitation and maximum HRTs, demonstrating Carex paniculata CW resilience under extreme pollution load events.
Total nitrogen, ammonia and nitrate seasonal removal
TN and NH3 loads in raw wastewater followed similar fluctuating trends not clearly related to PE or precipitation during this research (Figure 6).
After treatment, TN and NH3 concentrations followed the same trend and similar concentrations as in raw wastewater, showing that the majority of TN corresponded to NH3. Concentrations in the final effluent were more constant than in raw wastewater, suggesting CWTS efficacy in mitigating the peaks of influent TN and NH3. Average removal efficiencies aligned with general CW performance efficiencies (Vymazal 2007): 53% for TN and 43% for NH3 were achieved, with these parameter loads always below 43 mg/L in the final effluent.
A very different trend was observed for N-NO3 in raw wastewater, the concentration of which was generally low, with main peaks coinciding with high precipitation, i.e., March and May 2016 and March, June and December 2017, so N-NO3 influent load might be related to dunghill leachates and resuspension from sludge and sediments previously accumulated in the sewage collection system by stormwater and snowmelt runoff.
Total phosphorus and phosphate seasonal removal
TP and P-PO4 load in raw wastewater followed the same trend during the experiment, with maximum concentrations in summer, coinciding with maximum human PE, reaching maximum annual loads in October, with livestock present and minimum precipitation (Figure 7).
Except for March and December 2016, in which higher stormwater flows and lower HRT might be the cause of removal inhibition, TP was notably reduced in all campaigns in which a relevant load (>1.1 mg/L) was detected in raw wastewater, with removal efficiencies between 12% and 79%.
Phosphate removal followed the same seasonal trend, but performance efficiencies were slightly lower than for TP, commonly by 10 percentage points in each bimonthly campaign. As phosphorus removal through plant uptake is important but fluctuates seasonally in CWs, with limited efficacy in senescence (Vymazal 2007), the semi-evergreen character of Carex paniculata might be an advantage for phosphorus removal during the fall due to the persistent production of photosynthetic tissues and litter release minimisation.
Deeper investigations into the implications for nutrient removal/release of the fact that Carex paniculata is semi-evergreen should be conducted in the future.
Carex paniculata seasonal oxygenation performance
The DO content in raw wastewater was clearly related to wastewater temperature in March 2016, when high stormwater flows at 3.9 °C was due to snowmelt, leading to maximum DO concentrations in the influent, as well as in December 2017. During the rest of the experiment, the wastewater temperatures remained between 11.5 °C and 20.6 °C, and the seasonal influence on DO content was not as relevant as PE and TBP were (Figure 8).
The higher PEs coinciding with lower precipitation led to lower DO content in raw wastewater. Pre and primary treatments did not play an oxygenation role and DO concentration after primary treatment remained similar or was lower than in raw wastewater, confirming that oxygenation only occurred in the CW.
From August 2016 and during the whole of the study, a DO concentration increase was consistently observed in the final effluent compared to wastewater coming from the pre and primary treatments, with oxygenation efficiencies performed by the Carex paniculata CW varying between a minimum of 0.5 mg/L in June 2017 and a maximum increase rate of 2.4 mg/L in December 2016, demonstrating Carex paniculata's oxygenation capacity.
There is a very limited number of studies in the literature that have focused on the use of Carex paniculata in constructed wetlands, and the results of the present investigation represent the first relevant scientific contribution to Carex paniculata's biological annual cycle and potential for wastewater treatment assessment.
When the Carex paniculata CW efficacy for stormwater, sewage and livestock wastewater treatment in rural settlements in mountain areas was investigated, seasonal fluctuations and higher loads than those for which the wastewater treatment plant was designed for led to CW under sizing, establishing the real CW size ratio of 0.75 m2 per IE, which could have had a negative impact on the wastewater treatment plant performance. Nevertheless, high removal efficiencies for BOD5, COD and TSS (96%, 91%, 83% respectively) were achieved under the most unfavourable conditions (high influent contamination by livestock wastewater and lower HRT), and a particularly high removal efficiency for G&O, with annual average removal efficiencies of 77% and 93% in 2016 and 2017 respectively, were observed. A high disinfection potential (i.e., E. coli annual average removal of more than 80% in 2017) was also observed. A high oxygenation capacity by Carex paniculata in the range of 0.5 mg/L in June 2017 and 2.4 mg/L in December 2016 was observed, suggesting Carex paniculata CWs have space-saving competitiveness.
While no clear trend was observed in the removal of TN, NH3 and N-NO3, the removal of total P was of interest, reaching removal efficiencies of about 80%, which suggests further research to identify optimal conditions for nutrient removal and potentially recovery.
The semi-evergreen nature of Carex paniculata was observed during the investigation. Since the most frequently used macrophyte species in CWs are deciduous, leading to total P removal efficiency decreases during the fall, further investigation should be done on the implications of Carex paniculata's semi-evergreen character on nutrient removal efficiency maintenance as well as on the minimisation of nutrient release through litter decomposition in the cold season.
Comparative assessment of pre and primary treatments and CW performances on the main wastewater constituents' removal revealed that wastewater purification occurred mainly in the CW. However, the role of pre and primary treatments must not be underestimated as their contribution to the removal of BOD5, TN and NH3 was above 50% of the complete system performance in the second year of investigation.
These results suggest that Carex paniculata CWs show great potential for wastewater treatment, especially for rural settlements in mountain areas, where this plant species might constitute one of the best natural wastewater treatment alternatives, thus synergistically contributing economically and technically affordable effective sanitation infrastructure to rural areas, while at the same time providing some biodiversity protection to European upland wetlands, bogs and fens.
This research was undertaken as part of the Lamizal Research and Development Programme, with support given by the Department of Environment of the Cantabria Autonomous Community Government (Spain). The authors would also like to acknowledge the Centre of Environmental Research of Cantabria (CIMA) for their technical support, the Local Council of Pesaguero, the Neighborhood Council of Valdeprado and all the inhabitants of Valdeprado village, for hosting and supporting this project, especially to the Prieto Salceda family, for the free use of the land where the experimental wastewater treatment plant was built, and to Grandma Evangelina Gomez Caloca, for accommodating us in her home every time we needed her to during this 2-year investigation.