Abstract:
Geographically referenced data on humpback
dolphin (Sousa plumbea) distribution and behaviour were
collected in Richards Bay, South Africa, between 1998 and
2006. Utilization distributions presented a clear pattern in
the distribution of dolphin activities and use of various
locations within the study area. The 50 % kernel density
estimate (core area) for foraging/feeding (2.1 km2) was
clustered around the harbour mouth (including the shark
net installation). The core areas for resting, socialising and
travelling were more widespread ([2.95 km2). A mixed
effects model indicated that location (longitude and latitude),
distance from shore, year and group size are
important variables in predicting foraging/feeding behaviour
as opposed to other behaviours. The resting core was
found in the southern part of study area and socialising was less clustered. This indicates preferential use of certain
areas for specific behaviours. The overlap of foraging/
feeding areas with stationary fishing gear (shark nets) and
boat traffic are a major concern due to the risk of humaninduced
incidental mortalities. Any future alteration of the
coastal area of Richards Bay, particularly development in
the harbour, should be considered carefully, as further
intrusion into areas critical to humpback dolphins, such as
their foraging grounds, will inevitably carry negative
implications for this already severely impacted population.
