Abstract

Results from experiments designed to test the robustness and reliability of shipboard line transect methods for estimating harbor porpoise densities are presented. These experiments were conducted in the western Bay of Fundy using two teams of three observers which searched independently and simultaneously on the same vessel. Two experiments were conducted: one in which observer teams rotated between the top of the wheel house (4.2m above the water) and the crow’s nest (7.1m) and the other in which they rotated between the bow (1.8m) and the crow’s nest. A total of 434 sightings were made by one team and 490 by the second during 29 half hour transects.

The line transect estimates of density from these experiments tended to be robust to the effects of observer height and team composition. The effect of observer height (within the range of 1.8 to 7.1m) on the detection process for harbor porpoise was small as measured both by estimates of f(0) and comparisons of the frequency distributions of sighting angles, and radial and perpendicular distances. The differences with height that were observed tended to be consistent with what might be expected. However, the proportion of duplicate sightings seen by both teams appears to be small in comparison to the expected proportion given estimates for g(y). These latter results indicate that g(0) was substantially less than one and that direct estimates of g(0) are needed if estimates of absolute abundance are to be obtained from line transect surveys.