Carbon Sequestration by Roadside Filter Strips and Swales: A Field Study

As legislation continues to focus on greenhouse gas (GHG) and carbon dioxide (CO2) emissions and reductions, the terrestrial biome offers an attractive possibility to sequester carbon. Currently, the terrestrial pool is regarded as a CO2 sink, but scientists are unsure to what extent this trend will continue. Urbanization modifies the existing landscape, and little study has focused on the carbon (C) dynamics of specific urban land uses. In this research, the roadway environment, specifically the grassed right of way (ROW), was studied for carbon sequestration potential, an important ecosystem service. Transportation corridors exist worldwide, and the vegetated filter strip and swale (VFS/VS), two common stormwater control measures (SCMs), often constitute the grassed right of way adjacent to roadways. Carbon pools within roadway VFS/VS soils of North Carolina were specifically examined in this study. This research was conducted in two North Carolina physiographic regions: the Piedmont (characterized by clay-influenced soils) and the Coastal Plain (predominantly sandy soils). Approximately 700 soil samples were collected in VFS/VSs and wetland swales alongside major highways and analyzed for percent total soil C (% total C) and bulk density, which aided in obtaining the C density. Mean soil C densities were 2.55 ± 0.13 kg C m-2 (mean ± standard error, n=160, 0.2 m depth) in the Piedmont and 4.14 ± 0.15 kg C m-2 (n=160, 0.2 m depth) in Coastal Plain highway VFS/VSs. Because grasslands were reported to have similar carbon density values, they could be used as a surrogate land use for roadway VFS/VSs if no specific roadside data were available. Using a 37-year soil chronosequence, the carbon sequestration rate using a simple linear regression within Piedmont VFS/VSs was calculated at 0.053 kg C m-2 yr-1. Utilizing segmented linear regression models, the sequestration rate was calculated to be 0.155 kg C m-2 yr-1 for 15 years and 0.099 kg C m-2 yr-1 for the remaining 21.5 years. The roadside grass sequestration rate assumed by the Federal Highway Administration Carbon Sequestration Pilot Program (0.17 kg C m-2 yr-1) overestimates carbon accumulation by a factor of 3 in the linear model, and by a factor of 1.1 to 1.7 with the segmented linear models. Carbon density did not differ between dry and wetland swales, although % total C was significantly greater in wetland swales. Because C density and % total C in swales were not well defined by age, it appeared more appropriate to assess wetland swales and dry swales using a range of carbon values, rather than a rate of carbon sequestration. The mean VS C density was 3.05 ± 0.13 kg C m-2 (n=40, 0.2 m depth), while that for wetland swales was 5.04 ± 0.73 kg C m-2 (n=44, 0.2 m depth). If promoting C sequestration becomes a factor in ROW management, wetland swales would be more desirable than dry swales. While the VFS/VS sequestration rate is comparable to other grassed systems, the estimated 320-480 tons C per lane-mile expelled during roadway construction (Cass and Mukherjee 2011) is marginally offset through terrestrial sequestration in roadside VFS/VSs. Per kilometer of roadway constructed, Piedmont VFS/VSs would offset between 4% and 7% of C emitted during construction, depending on predictive model of C sequestration rate was used.