Estimating the change in regional scale distribution of seagrass and macroalgal beds using discrete local distribution data analyzed from aerial images

Abstract

AbstractThe sensitivity of vegetated habitats to climate change and catastrophic disturbance, as well as their benefits to human beings, motivate efforts to estimate regional changes in distribution. This study examined the regional changes in the distribution of submerged macrophyte vegetation using aerial and satellite images from before and after the 2011 tsunami disturbance along the Pacific coast of northeastern Japan. We conducted an estimation of the change in the distribution of seagrass and macroalgal beds after the tsunami along the whole coastline (1,635 km). This was based on vegetation change in the 13 local areas (each <40 km) spanning 27% of the coastline and included 28% of the macroalgal area and 59% of the seagrass area documented in satellite imagery before the tsunami, using generalized linear modeling (GLM). We found that in the 13 local areas 716.8 and 1,990.3 ha of macroalgal beds and seagrass beds had disappeared, respectively, suggesting that seagrass beds were more impacted by the tsunami than macroalgal beds. Significant predictors in GLM indicated that more seagrass beds disappeared in geographically complex coastal areas, whereas macroalgal beds were only marginally affected in such areas. It was estimated that the total decline in seagrass beds was at least 76% along the whole coastline, whereas macroalgal beds decreased only by about 20%. The magnitude and predictors of losses in these two vegetated habitats were similar to earlier estimates, suggesting that our simplified method with GLM is sufficient to extrapolate monitoring data to a regional scale.