Fig. 1.
Fig. 1.

(a) The Baltimore Metropolitan Area showing watershed boundaries for each study site and LIDAR imagery depicting land use and wetland location for: (b) Baisman Run, (c) Minebank Run, and (d) Stony Run study sites. Blue dots = minipiezometers used for push–pull in situ denitrification measurements and sampling for groundwater (GW) chemistry; red dots = sampling location for stream surface water (SW) chemistry. Streams are outlined in yellow and yellow arrows depict direction of flow.

 


Fig. 2.
Fig. 2.

Example cross-sectional profiles of the oxbow 1 (a) and oxbow 2 (b) wetlands relative to the Minebank Run stream reach derived from a digital elevation model. Te indicates the threshold elevation required to initiate surface water exchange, i.e., overbank flooding, between the oxbow and adjacent stream reach.

 


Fig. 3.
Fig. 3.

Summer and winter groundwater NO3 concentrations for each wetland study site: MBR = Minebank Run, SR = Stony Run, and BR = Baisman Run wetlands. Values are mean ± standard deviation. The SR–CW2 site had significantly higher concentrations than the other sites. Concentrations were significantly higher in winter than summer at the SR–CW2 site.

 


Fig. 4.
Fig. 4.

Relationships between denitrification rate and groundwater (a) NO3 (b) temperature (C°), and (c) dissolved oxygen (mg L−1) over all study sites and sampling dates (n = 37).

 


Fig. 5.
Fig. 5.

Example relative concentration profiles of the conservative tracer SF6 from (a) oxbow 1 and (b) oxbow 2 during summer and winter 4-h push–pull incubations at the Minebank Run study site. Rectangle = the core of the introduced plume (e.g., the first two pulled samples collected) after the 4-hr incubation period was complete.

 


Fig. 6.
Fig. 6.

Summer and winter in situ denitrification rates for each study wetland site: MBR = Minebank Run, SR = Stony Run, and BR = Baisman Run wetlands. Values are mean ± standard error of minipiezometers sampled in each site between June 2008 and December 2008. Comparison of means by site showed that the oxbow 2 wetland had significantly (P < 0.05) higher denitrification rates in winter than in summer, whereas the constructed stormwater wetlands (CW) 2 had higher (P < 0.05) denitrification rates in summer than winter (Fig. 8). Across all wetland sites, the CW2 wetland had highest (P < 0.05) denitrification rates.

 


Fig. 7.
Fig. 7.

Longitudinal profile of in situ denitrification rates along surface water flow paths (inlet to outlet) in the: (a) oxbow 1 and (b) oxbow 2 Minebank Run study sites during summer and winter 2008. Values are mean ± standard error. Although seven minipiezometers were installed in oxbow 2, only five were included when data were pooled across seasons due to only one push–pull measurement from oxbow 2 #6 in winter and no available 15N-N2O and N2 data for #6 and #7 in the summer.

 


Fig. 8.
Fig. 8.

The masses of N2O-N and N2 (μg) produced by each wetland pooled across seasons (a) and the mean ratio of N2O:N2 evolved from the oxbow, constructed, and forested wetland types (b). Denitrification rates in this study reflect the sum production of N2O + N2 via denitrification expressed as a mass in units (μg).