300 Monitoring Threatened Species and Ecological Communities
population estimate. The last is typically negatively correlated with the size of the
population (Gerrodette 1987). Thus small declining populations become almost
impossible to census for trends as animals become increasingly rare, although
improved statistical methods are being developed in an attempt to improve this
situation. Barlow and Reeves (2001) suggest a useful rule of thumb for cetaceans: at
least 10 annual surveys with high precision (CV <20%) are required to yield a high
probability (>80%) of detecting a 50% change in total population size.
Detection probability is usually more variable for coastal marine mammals
than oceanic species because of the finer spatial scale of environmental variability.
Thus a population could be reduced in size by 50% or more before strong statistical
evidence of a decline becomes available, despite frequent replicate surveys with
high precision. It is also difficult to confirm increases in abundance, making it
difficult to confirm whether management interventions are effective.
Using case studies, Taylor et al. (2007) demonstrate that the ability to detect
declines in marine mammal stocks with current monitoring programs is generally
poor, even when the decline is considerable. They concluded that it would be
impossible to detect even precipitous declines in most marine mammal
populations with present levels of investment, survey technology and design.
Improvement of performance in detecting declines depends on increasing survey
extent and frequency, and developing different methods to detect decline. Such
improvements will generally require a substantial increase in funding. In some
places, numbers are now so low that it may be impossible to detect trends in
abundance at a local scale. The design of future population studies should account
for all biases, including variation in detectability, and temporary and permanent
emigration, and be suitably resourced.
Lessons learned
● (^) Detecting population trends for marine mammals is possible but requires high
technical expertise and access to considerable resources because these animals
generally occur in small numbers, are sparsely distributed and are difficult to
capture at suitable rates.
● (^) The effort expended and results obtained in these cases studies demonstrate
that management intervention should not require the trigger of statistical
evidence of reduction in abundance; indeed such a requirement can be a red
herring that unduly delays conservation action.
Acknowledgements
The Darwin Harbour Coastal Dolphin Monitoring Program was funded through
the INPEX-led Ichthys LNG Project’s Nearshore Environmental Monitoring
Program as a partnership between the Northern Territory Government, INPEX