Sea level rise is reshaping the coasts, allowing coastal habitats such as tidal marshes to migrate inland. To predict where changes will occur, it is critical to understand the factors that influence land cover transition. Here, we test the influence of land cover type on land cover transition. We hypothesized that marsh migration may vary by upland land cover type, due to dominant plant species’ differences in salinity and inundation tolerance. Additionally, the response of people may make specific land cover types more likely to be protected from transition. We measured land cover change in high resolution aerial imagery over the relatively short period of 2009 to 2017 in coastal Somerset County, Maryland, a region with a high rate of relative sea level rise. We used artificial intelligence algorithms to segment and classify imagery and investigated land cover transition probabilities across gradients in elevation and distance from shore.
Results/Conclusions
In logistic models of land cover transition, we found that 'agricultural land' and 'scrub shrub wetland / forested wetland' cover classes were more likely to transition to 'emergent wetland' than 'forest / scrub shrub' or 'urban or built-up land' cover classes, after controlling for elevation and distance to shore, two well-known predictors of marsh migration. Over only 8 years, the loss of upland area in the county totaled 6.1 km2, of which 5.7 km2 was agricultural land. This represents a loss of over 2% of the farmland in the county, the majority of which converted to emergent wetland during the study period. This is one of the first studies to explore coastal land cover transition in agricultural land. A major finding of this work is that agricultural land is readily converting to coastal marsh, at faster rates than other land cover types. These findings have implications for the way sea level rise will affect coastal livelihoods and for policy instruments being developed to protect marsh migration corridors and reduce non-point water pollution.