Spatial and temporal variability in greenhouse gas abundance of urban streams: The role of urban infrastructure
Streams and rivers are significant sources of greenhouse gas emissions globally. Water quality and watershed management, are likely to influence GHG production. In urban-impacted watersheds, increased nitrogen loading, organic matter, and warming may contribute to accelerated nitrous oxide (N2O) carbon dioxide (CO2), and methane (CH4) production. While theoretical frameworks describing urban streams have focused on dissolved nutrient export, urban stream ecosystems are also dynamic in terms of GHG production in time and space. In the present study, we hypothesized that nitrogen-loaded streams are likely to produce more N2O than streams with low N, and those draining stormwater management wetlands will produce relatively more CH4. To test these hypotheses, we explored interactions between watershed infrastructure and ecosystem function across spatial and temporal scales. We measured dissolved nitrogen exports, trace gas emissions, and denitrification rates across streams draining a gradient of watershed management. Sampling encompassed seasonal (fall and spring) longitudinal surveys of two 15km2 urbanized watersheds, and one year of bi-weekly monitoring of eight headwater streams within the two larger watersheds. Headwater sites were paired across four distinct infrastructure designs, including 1) complete stream burial, 2) in-line stormwater wetlands, 3) floodplain preservation, and 4) septic systems.
Overall, GHG’s were net sources to the atmosphere throughout the year at all sites. Saturation ratios (measured / equilibrium) were significantly (p<0.01) different between the four headwater management types. N2O saturation ratio varied from 1.1 - 47 across all sites and dates. The highest N2O concentrations were measured in streams draining diffuse septic systems and high total dissolved nitrogen (TDN) concentrations, and the lowest N2O in streams with connected floodplains and low TDN. CO2 was highly correlated with N2O and across all sites and dates (r2=0.84) and CO2 saturation ratio varied from 1.1 - 73. CH4 concentrations were always super-saturated by a factor of 3.0 - 2157. The highest CH4 concentrations were measured in floodplain-connected streams and were negatively correlated with TDN. There was little to no seasonal variation in headwater GHG concentrations, suggesting that spatial variability in nitrogen loading and riparian connectivity may be more important for headwater stream gas emissions. Longitudinal sampling in spring and fall showed consistently super-saturated concentrations of all three gases throughout both watersheds. Ultimately our results show that urban streams can be significant and consistent sources of GHG’s throughout the year, and that watershed management and stream chemistry are key drivers.