Evapotranspiration affecting redox conditions in horizontal constructed wetlands under Mediterranean climate: Influence of plant species

Highlights

• Evapotranspiration and redox potential in mesocosms horizontal wetlands was assessed.

• Vegetation was the main design parameter affecting water loss within the wetlands.

• Typha angustifolia spent more water than Phragmites australis but more efficiently.

• Redox increased as evapotranspiration but with a stronger relationship for cattail.

• Evapotranspiration and redox state are plant species-dependent.

Abstract

The aim of this study was to conduct a comparative evaluation of evapotranspiration (ET) rates for eight different mesocosm constructed wetlands (CWs), and the relationship with redox potential (E_H). Inflow, outflow and E_H were measured over 4 years in winter and summer campaigns as well as over 24 h on selected days in summer. Vegetation was the main design parameter which affected water loss in the wetlands (on average, ET in planted wetlands was 4 times higher than in unplanted ones), and Typha angustifolia was more active than Phragmites australis (mean daily ET – expressed as the average of ET rate measured every 2 h in selected days in summer – was 36.8 ± 2.3 mm d⁻¹ and 23.0 ± 1.9 mm d⁻¹ for hydroponic wetlands planted with cattail and common reed, respectively), although P. australis water use efficiency was lower. Positive relationships were found between ET and E_H for planted wetlands. Cattail presented a stronger linear regression than common reed, demonstrating that ET and consequently redox conditions are plant species-dependent.

Introduction

Wastewater disposal constraints largely determine the type of treatment used. Thus, a wetland can be designed to achieve zero discharge if rigorous quality standards for effluent must be adhered to, if soil infiltration is not possible (Gregersen and Brix, 2001), or when water reuse is priority, i.e. for restoring sensitive areas (Greenway, 2005, Guardo, 1999, Paulo et al., 2012). One of the most important factors in designing and dimensioning constructed wetlands (CWs) for wastewater treatment is the water balance, which depends on the hydraulic loading rate (HLR), rainfall and plant-mediated evapotranspiration (ET). At constant inlet flow rates, evapotranspiration will induce changes in the water table (Mann and Wetzel, 1999), which has been related to enhanced removal efficiencies (Tanner et al., 1999). Therefore, ET is a decisive factor in achieving water discharge targets.

Evapotranspiration in a CW is the only means by which wetlands lose water, given that the water basin is insulated to avoid water infiltration of the soil. Evapotranspiration is intimately related to meteorological conditions (Kumar et al., 2012), but is also related to plant growth stage (Borin et al., 2011, Headley et al., 2012). In temperate climates with pronounced seasonality, ET differs greatly from winter to summer, with a maximum ET rate at mid summer, when the vegetation attains its effective full ground cover. Moreover, ET is subjected to a diurnal cycle, with maximum values in early afternoon and minimum values at night (Kadlec and Wallace, 2009), which can lead to a lack of outflow in periods of high ET. Therefore, variations in ET rates should be considered when dimensioning and modelling a constructed wetland (Galvão et al., 2010).

Additionally, ET has an effect on treatment performance, mainly because water loss increases hydraulic retention time (Kadlec and Wallace, 2009). Furthermore, contaminant removal is closely related to the redox state of the system (Ávila et al., 2013, Faulwetter et al., 2009, Pedescoll et al., 2011). Therefore, ET should contribute to the redox conditions within the systems, at least during periods of high ET. Several studies have reported the effect of redox soil conditions on macrophyte growth and activity (Gorai et al., 2011, Pezeshki et al., 1996), indicating that wetland plants (such as cattail, common reed or common rush) present tolerance to hypoxia. However, no information is available on the influence of ET on the redox conditions of the system.

Understanding the main factors that affect evapotranspiration would help to clarify decision-making during the system design process. Vegetation is the most important factor, in as much as plants are primarily responsible for wetland water loss. Vegetation height and leafiness plays an important role in ET rates within a system, as those wetlands with a high leaf area index (LAI) and tall vegetation present higher ET values. Pauliukonis and Schneider (2001) found that ET rates were three times higher for Salix babylonica than for Typha latifolia, and that the LAI for willow was triple that of cattail. Similarly, Eichhornia crassipes presented ET rates two times higher than those of Lemna minor in mesocosm tanks (DeBusk et al., 1983). In fact, species with a high LAI can intercept more dry wind, thus enhancing evapotranspiration. Nevertheless, other design parameters are also involved in water loss, such as flow type. Free water surface (FWS) and horizontal subsurface flow (HSSF) systems perform differently in terms of hydraulics (Pedescoll et al., 2013) and as regards sensitivity to environmental conditions (Kadlec, 2009).

The main aim of this study was to conduct a comparative evaluation of evapotranspiration rates for different constructed wetland designs and to determine the relationship with redox conditions within the systems. To this end, an experimental plant with 8 mesocosm wetlands was constructed and operated over almost 4 years. The plant included the most common wetland designs, such as FWS, SSF and floating macrophytes. This wide variety of configurations enabled us to evaluate different design parameters, e.g. presence/absence of vegetation, plant species, flow type and presence/absence of gravel matrix. As far as we are aware, this is the first time that evapotranspiration and redox conditions have been tested for different configurations under the same climatic conditions.