摘要

Different monitoring methodologies are available to quantify the spatial distribution of seagrass beds, but repeated coverage over short time intervals is usually expensive or impossible. Sampling resolution is also a limiting factor, due to costs and manpower. We propose a low-cost alternative that allows for the free definition of sampling frequency and resolution.
Since January 2006, a shallow seagrass meadow (Zostera spp.) (Sado estuary, Portugal), has been monitored using true color large-scale oblique terrestrial photographs taken monthly during low-water spring tides, from a fixed elevated point.
To rectify the images and create an orthogonal view, a set of ground control points (GCPs) were taken in and around the meadow, each coupled with a photograph. The photo-coordinates of each GCP were matched to its geographic coordinates, and a 2nd degree polynomial adjustment was used, to rectify every subsequent monthly photograph (RMS error of 0.7 m and final resolution of 0.5 m). An unsupervised classification was performed on the rectified images to map and quantify the meadow area.
The meadow showed a seasonal cycle typical of temperate areas, in response to the annual cycle of solar radiation, with maximum coverage (similar to 1 ha) in late summer/early fall, and minimum areas (similar to 0.5 ha) in late winter/early spring. A marked shift in the meadow distribution was also found, with similar to 25% average annual change (Index of relative change) and similar to 43% change over the 4-year period studied.
These results suggest that the method performed consistently, over time and over a range of varying environmental conditions, its main limitation being the need for an elevated point overlooking the area of interest. Whenever elevated viewpoints exist overlooking seagrass meadows or other intertidal or shallow subtidal coastal habitats, this methodology, may constitute a viable and economically advantageous alternative to other more costly approaches, e.g. aerial photography.

  • 出版日期2011-8