Stormwater Wetlands Research Articles

Particulate and Dissolved Organic Matter in Stormwater Runoff Influences Oxygen Demand in Urbanized Headwater Catchments.

Increasing inputs of organic matter (OM) are driving declining dissolved oxygen (DO) concentrations in coastal ecosystems worldwide. The quantity, source, and composition of OM transported to coastal ecosystems via stormwater runoff have been altered by land use changes associated with urbanization and subsequent hydrologic flows that accompany urban stormwater management. To elucidate the role of stormwater in the decline of coastal DO, rain event sampling of biochemical oxygen demand (BOD) in samples collected from the outfall of stormwater ponds and wetlands, as well as samples of largely untreated runoff carried by stormwater ditches, was conducted across a range of urban and suburban development densities. Sampling also included measurements of particulate and dissolved carbon and nitrogen, carbon and nitrogen stable isotopes, and chlorophyll-a. Results suggest stormwater may be a significant source of labile OM to receiving waters, especially during the first flush of runoff, even though BOD concentrations vary both among and within sites in response to rain events. BOD variability was best predicted by particulate OM (POM) and chlorophyll-a, rather than the larger pool of dissolved OM. These findings demonstrate the importance of managing episodic stormwater discharge, especially POM, from urbanized areas to mitigate DO impairment in larger downstream systems.

Reduction in Nitrogen Exports from Stormflow after Conversion of a Dry Detention Basin to a Stormwater Wetland

Stormwater control measures such as dry detention basins and wetlands are often used to reduce the discharge of urban runoff and nutrients to streams, but differences in nutrient treatment may vary between practices. The goal of this study was to compare the nitrogen treatment efficiency of a dry detention basin before and after it was converted into a stormwater wetland. Inflow and outflow from a detention basin in Greenville, North Carolina was sampled during 13 storms and the stormwater wetland was sampled during 10 storms. Total dissolved nitrogen (TDN), NO3−, NH4+, chloride, dissolved organic carbon (DOC), and physicochemical properties were evaluated. Inflow and outflow from the detention basin had identical median concentrations of TDN (0.47 mg L−1). The median TDN concentration for wetland outflow (0.18 mg L−1) was 63% lower relative to inflow (0.49 mg L−1). The hydraulic residence time of stormwater in the wetland was more than 10 times greater relative to the dry basin. There was a significant (p < 0.001) reduction in dissolved oxygen and oxidation reduction potential and an increase in median DOC concentrations in wetland outflow relative to inflow. Most of the reduction in TDN within the wetland was attributed to loss of NO3− (80% reduction), possibly due to denitrification. Conversion of dry detention basins to wetlands may provide significant benefits with regards to reducing TDN transport associated with urban runoff.

Long-term sustainable phosphorus (P) retention in a low-P stormwater wetland for Everglades restoration

The Stormwater Treatment Areas (STAs) are large wetlands constructed for phosphorus (P) retention for Everglades restoration in south Florida (USA), and include areas of submerged aquatic vegetation (SAV) at a globally unprecedented scale (~12,000 ha). The goal of this study was to elucidate the fate of P retained in large-scale SAV wetlands, and the associated temporal trends in P removal and retention. In a well-performing, 929-ha SAV-dominated STA surface water flow-through treatment wetland, measurements of accrued soil depth and soil P storage performed every ~4–6 years revealed a steady-state longitudinal soil P enrichment profile established within the first ~4 years of flow-through operation. Subsequently, the SAV soils accrued P at a relatively steady rate (1.13 g P m−2 yr−1 for the entire 17-year period) without indication of temporal P enrichment, spatial expansion of soil P enrichment in the inflow region, or impairment of water column P removal efficiency. Phosphorus sequestration occurred via accumulation of new sedimentary material (0.9–1.5 cm yr−1), rather than enrichment of existing soil. These soil surveys were accompanied by measurements of porewater SRP concentrations, soil P release under anoxia, and soil P fractions, which demonstrated that soil P release potential and concentrations of highly labile soil P generally decreased over time. These findings demonstrate that the P retention mechanisms operating within this large SAV wetland can be sustainable under managed steady-state conditions. Susceptibility of SAV to extreme environmental perturbations in this and other wetlands, however, remains a research priority.

A microbial perspective on balancing trade-offs in ecosystem functions in a constructed stormwater wetland

Green stormwater infrastructure, such as constructed wetlands (CWs), is a type of stormwater control measure that can decrease nutrient and pollutant loads from urban stormwater runoff. Wetland soil microorganisms provide nutrient and pollutant removal benefits which can also result in ecosystem disservices such as greenhouse gas (GHG) emissions and can inadvertently exacerbate climate change. Microbial respiration by facultative anaerobes in anoxic conditions is the primary pathway for nitrogen removal (benefit). Similar anoxic conditions that support denitrifying microorganisms can also support obligate anaerobes that produce methane (CH4) via methanogenesis (disservice). We examined nitrogen removal potential, GHG production, and microbial community structure within permanently flooded and shallow land or temporarily-flooded areas of a stormwater CW to identify zones for CW design optimization. Results indicate that permanently flooded zones compared to shallow land zones are greater sources of CH4 emissions (80.80 ± 118.31, 2.32 ± 9.33 mg CH4-C m−2 h−1, respectively) and emit more carbon to the atmosphere (7161.27 kg CO2, 93.20 kg CO2 equivalents, respectively). However, nitrogen removal potential rates were similar across both flooded and shallow land zones (24.45 ± 20.18, 20.29 ± 15.14 ng N2O-N hr−1 g−1 dry soil, respectively). At this particular CW, reduction of permanently flooded zones within the wetland could decrease GHG emissions (disservice) without limiting nitrogen removal (benefit) potential of the wetland. Holistic development and design of stormwater control measures, which account for microbial activity, provides the opportunity to maximize benefits (i.e., nutrient and pollutant removal) and reduce disservices (i.e., GHG emissions) of green stormwater infrastructure.

Carbon dioxide and methane fluxes from different surface types in a created urban wetland

Abstract. Many wetlands have been drained due to urbanization, agriculture, forestry or other purposes, which has resulted in a loss of their ecosystem services. To protect receiving waters and to achieve services such as flood control and storm water quality mitigation, new wetlands are created in urbanized areas. However, our knowledge of greenhouse gas exchange in newly created wetlands in urban areas is currently limited. In this paper we present measurements carried out at a created urban wetland in Southern Finland in the boreal climate. We conducted measurements of ecosystem CO2 flux and CH4 flux (FCH4) at the created storm water wetland Gateway in Nummela, Vihti, Southern Finland, using the eddy covariance (EC) technique. The measurements were commenced the fourth year after construction and lasted for 1 full year and two subsequent growing seasons. Besides ecosystem-scale fluxes measured by the EC tower, the diffusive CO2 and CH4 fluxes from the open-water areas (FwCO2 and FwCH4, respectively) were modelled based on measurements of CO2 and CH4 concentration in the water. Fluxes from the vegetated areas were estimated by applying a simple mixing model using the above-mentioned fluxes and the footprint-weighted fractional area. The half-hourly footprint-weighted contribution of diffusive fluxes from open water ranged from 0 % to 25.5 % in 2013. The annual net ecosystem exchange (NEE) of the studied wetland was 8.0 g C-CO2 m−2 yr−1, with the 95 % confidence interval between −18.9 and 34.9 g C-CO2 m−2 yr−1, and FCH4 was 3.9 g C-CH4 m−2 yr−1, with the 95 % confidence interval between 3.75 and 4.07 g C-CH4 m−2 yr−1. The ecosystem sequestered CO2 during summer months (June–August), while the rest of the year it was a CO2 source. CH4 displayed strong seasonal dynamics, higher in summer and lower in winter, with a sporadic emission episode in the end of May 2013. Both CH4 and CO2 fluxes, especially those obtained from vegetated areas, exhibited strong diurnal cycles during summer with synchronized peaks around noon. The annual FwCO2 was 297.5 g C-CO2 m−2 yr−1 and FwCH4 was 1.73 g C-CH4 m−2 yr−1. The peak diffusive CH4 flux was 137.6 nmol C-CH4 m−2 s−1, which was synchronized with the FCH4. Overall, during the monitored time period, the established storm water wetland had a climate-warming effect with 0.263 kg CO2-eq m−2 yr−1 of which 89 % was contributed by CH4. The radiative forcing of the open-water areas exceeded that of the vegetation areas (1.194 and 0.111 kg CO2-eq m−2 yr−1, respectively), which implies that, when considering solely the climate impact of a created wetland over a 100-year horizon, it would be more beneficial to design and establish wetlands with large patches of emergent vegetation and to limit the areas of open water to the minimum necessitated by other desired ecosystem services.

Dry Wetlands: Nutrient Dynamics in Ephemeral Constructed Stormwater Wetlands.

Constructed stormwater wetlands (CSWs) are used to address contaminants in urban stormwater such as nitrogen (N) and phosphorus (P), but their performance is variable. Ephemeral CSWs tend to be less effective than perennial CSWs at removing N and P. We asked: How does wetland vegetation and sediment affect nutrient cycling/release from sediment and vegetation in ephemeral CSWs? We focused on two ephemeral urban CSWs in Pocatello, ID, USA, one densely vegetated and the other nearly bare. We rewetted intact cores of dry wetland sediments and, separately, senesced vegetation for 1 week at the end of the summer dry period to assess whether wetland sediments and vegetation acted as sources or sinks of N and P. For both CSWs, there was a pulse of nutrients immediately following rewetting, but the magnitude of that pulse and subsequent changes in nutrient concentrations suggest different processes dominate at each wetland, driven by differences in wetland vegetation and associated sediment characteristics. There was evidence of denitrification between and during events at the vegetated wetland, but larger fluxes of P at this site suggests a tradeoff between denitrification and P release. While the experimental results suggested specific biogeochemical controls, CSW nutrient concentrations during three events were more dynamic and suggested more biogeochemical complexity than that represented in our experiment, both within events and seasonally. Ephemeral CSWs may create unique biogeochemical conditions and require careful design to ensure N and P retention. Managers will also need to consider whether perennial water sources would improve CSW function.

Glyphosate and aminomethylphosphonic acid (AMPA) are commonly found in urban streams and wetlands of Melbourne, Australia

Glyphosate is the most widely-used pesticide for weed control in agriculture and in urban and residential areas. This is the first study to quantify glyphosate and AMPA levels in surface water in Australia from different land uses. Glyphosate and AMPA was measured in surface water from 10 rural streams, 30 urban stormwater wetlands and 9 urban streams located in and around the city of Melbourne, Australia on five occasions between October 2017 and February 2018. Glyphosate and AMPA were present in most of the urban surface water samples. The frequency of detection of glyphosate was 77% in wetlands and 79% in urban streams, whereas it was only detected in 4% of the rural streams. Similarly, AMPA detection was 91% in wetlands and 97% in urban streams, whereas it was only present in 6% of the rural stream samples. In both urban streams and wetlands, the highest average glyphosate concentrations occurred in November (1.8 ± 2.2 μg L−1). Overall, wetlands and streams associated with urban land use are vulnerable to glyphosate contamination. These results highlight the importance of screening for contaminants in urban stormwater to identify the source of pollutants that may end up in aquatic ecosystems and the risks therefrom.

Effect of temperature and drying-rewetting of sediments on the partitioning between denitrification and DNRA in constructed urban stormwater wetlands

Constructed wetlands are increasingly used within urban areas to reduce pollutant runoff, including nitrogen. These environments are exposed to frequent wetting and drying events, and also increasing temperatures due to the heat island effect and seasonal variation. In this study, we have investigated the role of drying and rewetting of sediments and temperature on the partitioning of two nitrate reduction pathways: denitrification (DNF) and dissimilatory nitrate reduction to ammonium (DNRA) in four constructed urban stormwater wetlands in Melbourne, Australia. Our results suggest that both DNF and DNRA decreased due to drying of sediments, but DNRA decreased to a greater extent. Rates of DNRA from the drying-rewetting treatment decreased 5–90% compared to the control. Moreover, a concomitant increase in NH4+ fluxes and a slight change in sediment oxygen demand (SOD) suggest that oxic conditions arising from sediment drying was responsible for the decreased rates of DNF and DNRA. Rates of both DNF and DNRA increased with increasing temperature in slurries. The DNF:DNRA ratio suggests that the relative increase in DNRA was more than DNF with increasing temperature. The mean activation energy of DNF ranged from 41 ± 1 to 64 ± 4 kJ mol−1 and the corresponding temperature coefficient (Q10) values ranged from 1.3 to 2.4. In comparison, the mean activation energy of DNRA was higher and ranged from 50 ± 8 to 107 ± 14 kJ mol−1 with the corresponding Q10 values ranging from 1.2 to 3.4. Overall, our results suggest that drying and re-wetting of sediments decreases nitrate reduction and increases nitrogen retention, whereas increased temperature enhances the recycling of bioavailable nitrogen in wet sediments in constructed urban stormwater wetlands.

Data-driven modeling of phosphorus (P) dynamics in low-P stormwater wetlands

The ability to simulate wetland phosphorus (P) concentration dynamics in the low-P domain (<20 μg/L of total P) has been poor. This paper develops a new model algorithm to describe P removal in Stormwater Treatment Areas (STAs) for Everglades restoration. It includes a performance review of previous low-P models. Analysis of inflow-outflow pulse responses at daily and monthly time steps in the STAs revealed surprisingly counterintuitive dynamics (lowest outflow concentrations occurring during highest flow events). Analysis of internal P concentration and flux data (from within STAs) informed new flow-dependent P removal and zero-order return flux formulations. The resulting parsimonious model is based directly on these two core equations. Calibration and testing are presented, using four independent datasets (6–36 month duration) including outflow P time series and internal water treatment profiles. This data-driven approach significantly improved dynamic simulations (R2 = 0.4–0.7) and identified the determinant role of internal P loading in low-P wetlands.

Factors controlling dissimilatory nitrate reduction processes in constructed stormwater urban wetlands

Water treatment wetlands are increasingly being used to reduce pollutant loads including nitrogen (N) in urban runoff. Processes such as denitrification (DNF) and anaerobic ammonium oxidation (anammox), which remove N, and dissimilatory nitrate reduction to ammonium (DNRA), which recycles N, play an important role in controlling NOx (NO3− + NO2−) removal versus recycling in wetlands. The relative importance of DNF, anammox and DNRA was investigated in four constructed stormwater urban wetlands in Melbourne, Australia. Rates of DNF and DNRA were variable and did not differ significantly among wetlands. However, rates of DNF and DNRA were significantly different (p < 0.05) in different seasons. The relationship between NOx reduction processes and measured concentrations of water column NOx, chlorophyll a (chl-a), sediment organic carbon (OC), porewater ferrous iron (Fe2+) and sulfide (S2−) and water column temperature were examined using multiple regression analysis (MRA). Anammox was an insignificant pathway (< 0.05% of total nitrate reduction). During winter when average water column temperatures were 12 °C, DNRA was consistently higher than DNF averaging 67 ± 23% of total NOx reduction. The MRA revealed that DNF was positively associated with NOx concentration whereas DNRA was negatively associated with temperature, and porewater Fe2+, and positively associated with chl-a. The ratio between DNF and DNF:(DNF + DNRA) showed a positive correlation with both temperature and NOx concentration in the MRA. At higher temperatures and higher NOx concentrations, DNF increased over DNRA. Overall, this study suggests that at low NOx concentrations, N is recycled internally in these urban stormwater wetlands, but the portion of N removed by DNF increases as NOx concentrations increase.

Balancing biodiversity outcomes and pollution management in urban stormwater treatment wetlands

Wetlands are increasingly being constructed to mitigate the effects of urban stormwater, such as altered hydrological regimes and reduced water quality, on downstream aquatic ecosystems. While the primary purpose of these wetlands is to manage stormwater, they also attract animals whose growth, survival and breeding (i.e. ‘fitness’) may be compromised. Such deleterious effects will be exacerbated if animals are caught in ‘ecological traps’, mistakenly preferring wetlands with unsuitable environmental conditions. Alternatively, wetlands that offer suitable habitat conditions for animals could be beneficial, especially in fragmented urban landscapes. Consequently, a thorough understanding of the potential ecological impacts of stormwater treatment wetlands is critical for managing unintended consequences to urban biodiversity.To help facilitate this understanding, we draw upon findings from a four-year research program conducted in the city of Melbourne in south-eastern Australia as a case study. First, we summarise our research demonstrating that some stormwater wetlands can be ecological traps for native frogs and fish in the study region, whilst others likely provide important habitat in areas where few natural waterbodies remain. We use our work to highlight that while stormwater wetlands can be ecological traps, their effects can be properly managed. We propose the need for a better understanding of the ecological consequences of changes to wetland quality and their population-level impacts across the landscape. We hope that this study will generate discussions about how to most effectively manage constructed wetlands in urban landscapes and more research for a better understanding of the issues and opportunities regarding potential ecological traps.

A simple and rapid direct injection method for the determination of glyphosate and AMPA in environmental water samples.

Glyphosate is currently the most widely used herbicide in the world, yet screening of environmental waters for this chemical is limited by the need for specialized derivatization and measurement methods that can be tedious and time-consuming. In this work, we present a novel method for the detection and quantification at trace levels of glyphosate and aminomethylphosphonic acid (AMPA) in environmental water samples. The detection and quantification of the analytes was performed by liquid chromatography (LC) coupled with tandem mass spectrometry (MS/MS). Chromatographic separation was achieved with an ion-exchange column and a pH-gradient elution of a solution of ammonium hydroxide and ammonium acetate. The limit of detection for glyphosate and AMPA was 0.25 μg L-1 and the limit of quantification was 0.5 μg L-1with a 20-μL injection. The method was used to investigate the levels of glyphosate and AMPA in surface water samples from the Yarra River catchment area and urban constructed stormwater wetlands. The results indicate that at the time of sampling, no glyphosate or AMPA was present in the samples from the Yarra River catchment area (n = 10). However, glyphosate was detected above the limit of quantification in 33% of the wetland samples (n = 12), with concentrations ranging from 1.95 to 2.96 μg L-1. Similarly, AMPA was quantified in 83% of the wetland samples, with concentrations ranging from 0.55 to 2.42 μg L-1. To our knowledge, this is the first report of a pH-gradient LC-MS/MS method for glyphosate and AMPA analysis at ultratrace levels, with minimal sample processing, avoiding costly, time-consuming derivatization and preconcentration steps. Graphical abstract ᅟ.

Wetland as revitalization pond at urban area based on the eco hydrology concept

Universitas Indonesia (UI) has a campus in Depok with six ponds, five of which are artificial ponds and one of the catchment area that operates in the central Ciliwung Sub-watershed. The ponds (Kenanga, Agathis, Mahoni, Puspa, Ulin and Salam "KAMPUS") are designed as one of the catchment area, referring to Presidential Decree No. 32/1990 about the management of protected areas, ponds as catchment areas, flood controllers, and groundwater runoff. However, due to uncontrolled development in the catchment area of the UI pond system, KAMPUS cascade ponds transformed into a "toilet" which accommodates the liquid and solid waste of its catchment area. The mechanism designed for management in the KAMPUS cascade pond system is to utilize the cascade pond as a stabilization pond (improving water quality). One of them is utilizing Agathis cascade pond as a constructed stormwater wetland system, which is a pretreatment for the inflow to the KAMPUS pond system. This constructed wetland design is planned with several mechanisms: precipitation, filtering, chemical process by utilizing plants for absorption, nutrient transformation and eliminating pathogens. The result shows that this constructed wetland gives the water quality improvement percentage up to 87%. Water quality on effluent conforms with the water quality standards for irrigation and planting (grade 4) referring to Government Regulation No. 82/2001 about Water Quality Management and Pollution Control.

Using conservation behavior to manage ecological traps for a threatened freshwater fish

AbstractUrban stormwater treatment wetlands are constructed to treat stormwater runoff before it enters waterways, but are often inhabited by animals that can suffer impairments. Such effects will be exacerbated if animals mistakenly prefer sites where their fitness is reduced, and are caught in ecological traps. Traps can compromise the persistence of threatened species in urban landscapes, so assessing how animals respond to stormwater wetlands, and whether their fitness varies between stormwater and natural wetlands, can guide management efforts. We examined the habitat selection behavior of dwarf galaxias (Galaxiella pusilla), a threatened freshwater fish inhabiting wetlands and streams around Melbourne, Australia. We tested whether fish respond to cues that could indicate habitat quality, and whether they can differentiate between water from stormwater and non‐stormwater wetlands. We also tested whether fitness (e.g., survival, development) was lower at stormwater wetlands. Fish responded weakly to cues and did not avoid stormwater wetlands. Survival was lower at stormwater wetlands, and fish suffered delayed ovarian maturation, potentially due to a lack of food. Thus, stormwater wetlands can be equal‐preference ecological traps for G. pusilla, a key finding with direct implications for conservation. For highly dispersive species that are not selective about habitats, managing ecological traps requires a whole‐of‐landscape approach, with consideration given to where wetlands are constructed. Given the investment required to remediate existing ecological traps, this approach can also identify priority wetlands for intervention. We discuss our results in the context of how insights from animal behavior can inform threatened species conservation in fragmented urban landscapes.

Contaminant mixtures interact to impair predator-avoidance behaviours and survival in a larval amphibian

Global declines in amphibian populations are a significant conservation concern, and environmental contamination is likely a contributing driver. Although direct toxicity may be partly responsible, contaminants are often present at sub-lethal concentrations in the wild. Behavioural end-points are becoming an increasingly useful method to estimate the impact of contaminants, particularly if the behavioural responses manifest to affect individual fitness (i.e. survival, growth, or reproduction). In the wild, most animals are affected by multiple stressors, and determining how these interact to affect behaviour is critical for understanding the ecological implications of contaminant exposure. Here, we examined the individual and interactive effect of the heavy metal copper and the insecticide imidacloprid on mortality rates and anti-predator behaviours of spotted marsh frog (Limnodynastes tasmaniensis) tadpoles. This common species frequently occupies and breeds in contaminated stormwater and agricultural wetlands, where copper and imidacloprid are often present. These contaminants may alter behaviour via physiological and neurological pathways, as well as affecting how tadpoles respond to chemical cues. Tadpoles suffered unexpectedly high mortality rates when exposed to imidacloprid concentrations well below published LC50 concentrations. Only unexposed tadpoles significantly avoided predator cues. Copper and imidacloprid reduced swimming speed and distance, and escape responses, while increasing erratic swimming. We observed an interactive effect of imidacloprid and copper on erratic swimming, but in general imidacloprid and copper did not act synergistically. Our results suggest that as contaminants enter waterbodies, tadpoles will suffer considerable direct mortality, reduced foraging capacity, and increased susceptibility to predation. Our results provide the first evidence of imidacloprid affecting amphibian behaviour, and highlight both the adverse effects of copper and imidacloprid, and the importance of exploring the effect of multiple contaminants simultaneously.

Hurricane and seasonal effects on hydrology and water quality of a subtropical urban stormwater wetland

Understanding the hydrology of a created or restored wetland is a critical component in understanding the overall functioning of its design, and in determining if the original goals of the wetland are being met. Freedom Park, a 20-ha park constructed in Naples, Florida in 2009, includes a 4.6-ha subtropical stormwater wetland treatment system for improving the water quality of the downstream impaired Gordon River and estuaries that flow into the Gulf of Mexico. Subtropical wetlands often present significant issues of sustaining viable wetland vegetation communities because of long wet and then long dry patterns in the wetlands’ hydroperiods. The treatment wetland’s watershed includes 1766 ha of an urbanized Naples metro area. The goals of the stormwater wetlands are to improve the water quality, reduce peak flows discharging into Gordon River and further to Naples Bay, and provide an aesthetic water park for the public. A monthly hydrologic budget for the period August 1, 2016 to January 31, 2018 showed the following averages: pumped surface water inflow of 50.1 cm week−1; average surface outflow of 38.8 cm week−1; precipitation of 3.76 cm week−1, potential evapotranspiration (PET) of 5.34 cm week−1; and estimated seepage of 14.8 cm week−1. During the wet season, water pathways are largely influenced by pumped surface inflow and surface outflow, balanced to not produce significant changes in water level from month to month. Groundwater seepage and evapotranspiration are the most important outflows during the dry season. A stormwater pulsing study during Hurricane Irma (September 10, 2017) showed that after an initial inflow flush due primarily to rainfall, total phosphorus and total nitrogen concentrations were higher at the outflow than at the inflow for several days during and immediately after the hurricane. Similarly, the mass of nutrients leaving the system was 130% and 37% higher for TN and TP, respectively, than the mass entering the system during this 6-day period during and immediately after the hurricane. Ironically, the design of the pump-only inflow did not allow for major urban runoff during the hurricane to enter the wetlands. Samples taken at the inflow and outflow of the system for four other storm events throughout the 2017 wet season (July through October) were also analyzed for their effects on water quality. While total phosphorus removal during and after storm events is comparable to normal conditions at the Freedom Park wetlands and other stormwater treatment wetlands in the world, total nitrogen removal decreased during and after storm events.

Stormwater wetlands can function as ecological traps for urban frogs.

Around cities, natural wetlands are rapidly being destroyed and replaced with wetlands constructed to treat stormwater. Although the intended purpose of these wetlands is to manage urban stormwater, they are inhabited by wildlife that might be exposed to contaminants. These effects will be exacerbated if animals are unable to differentiate between stormwater treatment wetlands of varying quality and some function as "ecological traps" (i.e., habitats that animals prefer despite fitness being lower than in other habitats). To examine if urban stormwater wetlands can be ecological traps for frogs, we tested if survival, metamorphosis-related measures, and predator avoidance behaviors of frogs differed within mesocosms that simulated stormwater wetlands with different contaminant levels, and paired this with a natural oviposition experiment to assess breeding-site preferences. We provide the first empirical evidence that these wetlands can function as ecological traps for frogs. Tadpoles had lower survival and were less responsive to predator olfactory cues when raised in more polluted stormwater wetlands, but also reached metamorphosis earlier and at a larger size. A greater size at metamorphosis was likely a result of increased per capita food availability due to higher mortality combined with eutrophication, although other compensatory effects such as selective-mortality removing smaller individuals from low-quality mesocosms may also explain these results. Breeding adults laid comparable numbers of eggs across wetlands with high and low contaminant levels, indicating no avoidance of the former. Since stormwater treatment wetlands are often the only available aquatic habitat in urban landscapes we need to better understand how they perform as habitats to guide management decisions that mitigate their potential ecological costs. This may include improving wetland quality so that fitness is no longer compromised, preventing colonization by animals, altering the cues animals use when selecting habitats, pretreating contaminated water prior to release, providing off-line wetlands nearby, or simply not constructing stormwater treatment wetlands in sensitive areas. Our study confirms the potential for urban stormwater treatment wetlands to function as ecological traps and highlights the need for greater awareness of their prevalence and impact at landscape scales.

Shifts in dissolved organic matter and microbial community composition are associated with enhanced removal of fecal pollutants in urban stormwater wetlands

Constructed stormwater wetlands provide a host of ecosystem services, including potentially pathogen removal. We present results from a multi-wetland study that integrates across weather, chemical, microbiological and engineering design variables in order to identify patterns of microbial contaminant removal from inlet to outlet within wetlands and key drivers of those patterns. One or more microbial contaminants were detected at the inlet of each stormwater wetland (Escherichia coli and Enterococcus > Bacteroides HF183 > adenovirus). Bacteroides HF183 and adenovirus concentrations declined from inlet to outlet at all wetlands. However, co-removal of pathogens and fecal indicator bacteria only occurred at wetlands where microbial assemblages at the inlet (dominated by Proteobacteria and Bacteriodetes) were largely displaced by indigenous autotrophic microbial communities at the outlet (dominated by Cyanobacteria). Microbial community transitions (characterized using pyrosequencing) were well approximated by a combination of two rapid indicators: (1) fluorescent dissolved organic matter, and (2) chlorophyll a or phaeophytin a fluorescence. Within-wetland treatment of fecal markers and indicators was not strongly correlated with the catchment-to-wetland area ratio, but was diminished in older wetlands, which may point to a need for more frequent maintenance.

Identifying critical inundation thresholds to maintain vegetation cover in stormwater treatment wetlands

Vegetation cover, specifically emergent macrophyte cover, is critical to the performance of constructed stormwater treatment wetlands which are widely used to protect receiving waterways. Maintenance of vegetation relies on a suitable water level regime, which in turn is governed by the hydraulic loading rate and efficiency of the outlet control structure. Wetland designers need to take these factors into account to avoid excessive depth and duration of inundation. We investigated the relationship between the water level regime and vegetation cover in 19 stormwater wetlands. We found a strong relationship between the median water level and vegetation cover, where vegetation cover was generally high (>80%) at median depths <0.35 m. Interestingly, poor outlet efficiency, rather than the hydraulic loading rate, was driving the water level regime. While this again highlights the importance of appropriate design and maintenance, we suggest that adjustable outlets in combination with real-time water level monitoring should be used to maintain a suitable water level regime in stormwater wetlands fed by dynamic urban catchments. The results of the study provide firm evidence for threshold depths for emergent macrophytes in stormwater wetlands which can be used to inform design guidelines.

Particle size distribution predicts particulate phosphorus removal

Particulate phosphorus (PP) is often the largest component of the total phosphorus (P) load in stormwater. Fine-resolution measurement of particle sizes allows us to investigate the mechanisms behind the removal of PP in stormwater wetlands, since the diameter of particles influences the settling velocity and the amount of sorbed P on a particle. In this paper, we present a novel method to estimate PP, where we measure and count individual particles in stormwater and use the total surface area as a proxy for PP. Our results show a strong relationship between total particle surface area and PP, which we use to put forth a simple mechanistic model of PP removal via gravitational settling of individual mineral particles, based on a continuous particle size distribution. This information can help improve the design of stormwater Best management practices to reduce PP loading in both urban and agricultural watersheds.