Confirmation of accelerated marsh erosion following the Deepwater Horizon oil spill, improved by planting
We begin by reviewing our field erosion data, collected for 6 years after the spill (7 years in total), to determine the duration of oiling effects and to look for longer-term cleanup treatment influences. We have previously observed increased erosion in our heavily polluted marsh sites for a period of 2 years after the spill, the duration of our study at that time.seven. We also observed no major difference in erosion between soiled sites that were untreated versus those with manual cleaning treatments. In a separate but related experiment, using slightly different methods, there were indirect indications that mechanical cleaning treatments may have further aggravated erosion and direct evidence that planting limited erosion in mechanically treated sites on a one-year study period.seven.
In the present study, our field comparisons of oiling/treatment categories included reference sites, oiled and untreated, oiled and manually treated, and oiled and mechanically treated. Untreated sites underwent no active cleaning (ie.seven for information on the trade-offs of typical oiled marsh treatment tactics). The manual clean-up treatment involved raking, cutting and removing oiled kelp, oiled vegetation mats (laid over oiled and dead vegetation that was still rooted) and underlying thick oil on the substrate by small crews using hand tools to remove surface oil to the ground. where possible and to better expose the remaining oil to natural weathering and degradation processesseven. Manual crews used walking boards to minimize foot traffic on the marsh surface. The mechanical cleaning treatment involved a mechanized grapple to remove soiled kelp and mechanized raking, cutting and scraping to remove or reduce soiled vegetation mats and oil on the substrate. Mechanical treatments were applied using long-reach hydraulic arms mounted on shallow-draft barges and large airboats stationed just off the marsh shore.7.23. Mechanical processing aimed for the same goals as manual cleaning, but with intended increases in speed and scale; however, mechanical processing can also be less precise, resulting in the removal of soils, the mixing of oil into the substrate, etc. Oil conditions were very consistent across all oiled sites, characterized as heavy oiling by Shoreline Cleanup Assessment Technique (SCAT) methodssevenalthough we agree that the oiling in our sites could be considered “very heavy” as suggested by others19. Oiled conditions consisted of a continuous oiled strip 6 to 13 meters (m) wide along the shoreline of the marsh, with heavily oiled kelp and vegetation mats covering a layer 2 to 3 centimeters (cm) thick. oil emulsified on the marsh surface with ~90–100% oil coverageseven. The heavily soiled sites all experienced complete or near-complete vegetation loss, with vegetation recovery taking several years7.23. Nearby reference sites on the same shoreline had light to no oiling, intact vegetation structure, and no clean-up treatmentseven. Further details on oiling conditions, cleaning treatments and vegetation response are included in our previous articles (including photographs)7.23.
Annual rates of shoreline erosion were determined annually by ground surveys using tape measures and differentially corrected GPS (horizontal accuracy of ±10 cm) to measure shoreline position along established transects. In the present study, we observed 147-198% greater marsh bank erosion for oiled sites compared to reference sites over 2 years (2010-2011 and 2011-2012), with no clear differences between sites. oiled with or without cleaning treatment (Fig. 2, Supplementary Table S1). There were indications that mechanical treatment may have worsened erosion at some sites in 2011-2012, which was consistent with our field observations and previous experience.sevenalthough our sample size was small and highly variable in this case.
Potential causes of increased shoreline erosion in oiled marshes may lie most directly in the death of marsh vegetation, which foils wave energy and binds marsh soils. Vegetation death at the edge of the marsh was likely caused by several related factors, including thick and persistent oiling covering all or most of the aboveground vegetation and ground surface, repetitive oiling, and penetration and the mixing of oil in soils, causing fouling and smothering effects on plants such as interference with photosynthesis, gas exchange, thermal regulation, etc., leading to plant death. Our previous work has shown substantial reductions (77-100%) in total above ground vegetation cover (all species) and Spartina alternifloracover (the dominant marsh species) in our oiled sites compared to the reference in 2010-2011 (and reductions of 45 to 99% in 2011-2012)seven. Belowground plant biomass was likely also reduced, although this was not measured in our study. However, belowground biomass has been reduced at similarly oiled sites in studies by other5,9,14,24. Significant relationships have been established between belowground marsh biomass and soil shear strength (a measure of erosion potential)25including in sullied swamps9and belowground biomass is the main feature of vegetation that resists erosion in coastal marshes26. In collaboration with us, Lin et al.9 conducted auxiliary sampling at a subset of our study sites during the 2011-2012 period and provided their unpublished soil shear strength data (0-6 cm) for our use (their sampling did not did not include our planted sites). There was a 42% reduction in soil shear strength in the oiled sites compared to the reference sites (Fig. 3, Supplementary Table S2). The average values of our reference site are very similar to the shear strength values of marsh soil reported by others in our study area.9.27, while the values of our soiled sites are much lower. Thus, there is evidence that oiling affected both marsh vegetation and the erodibility of marsh soils, which likely led to the observed differences in marsh erosion between reference and oiled sites. Oyster beds were not present near the edge of the marsh in our study area; thus, oyster cover impacts did not contribute to the differences in erosion observed between our oiled and reference sites (see Powers et al.17).