Figure 4. Video photograph from Sørmannsneset at the Norwegian continental break, 220 m depth (16 May 1998), showing a barren landscape with crushed remains of Lophelia skeleton spread over the area. This is a region subject to considerable bottom trawling. A track can be seen stretching from bottom-left to up-right of the photograph, indicating the path of a trawl.
damaged coral reef areas varies between 5% (R4) and 52% (R3) (Table 3). There is a connection between the size of the coral areas, the number of reporting fishermen in all regions and the extent of the damage, i.e., the more reports from fishermen, the more corals in the region and higher the percentage of damaged corals (Tables 1 and 3).
We have relied heavily on fishermen reports, as they were the only source of information on the largest coral areas. It is thus crucial that the fishermen distin- guish between Lophelia, which they call ‘glass corals’ or ‘white corals’ and other corals such as gorgonians, which they report as ‘red forest’ and ‘bushes’. The chances of confusion between Lophelia, Madrepora and stylasterid hydrozoans was considered minimal as Madrepora and stylasterids don’t build reefs and are by far not as abundant as Lophelia (Dons, 1944; Frederiksen et al., 1992; Mortensen et al., 1995). All ROV-inspections confirmed that the localities pointed out contained Lophelia corals, intact or damaged. We
therefore conclude that fisherman’s reports are a reli- able source of information and also have been of great value for the estimation of coral areas.
The estimation of the size of intact and damaged areas are based on points, lines (stretches) and areas as reported by fishermen. The estimation of areas from lines between co-ordinates were based on an assumed range of width of the trawl ground, and introduced thus, an uncertainty. However, as the number of re- ported stretches comprised a small fraction compared to the number of points and areas, the total estimates exhibited a moderate range (Table 3) and do not affect the main conclusions. Since the methodology used for the estimation of total coral area and damaged area is the same, the estimates are comparable, i.e., the bias is systematic.
Distribution of corals
In general the corals occur on substrate of morainic origin (Hovland & Mortensen, 1999; Mortensen, 2000) and the largest densities are found along the continental break and edges of trenches crossing the shelf. Very few records are from levelled parts of the shelf.