Click counting : an acoustic censusing method for estimating sperm whale abundance
Douglas, Lesley Anne
Present estimates of sperm whale (Physeter macrocephalus) population size are extraordinarily imprecise. This is largely due to sperm whale behaviour and distribution, which compromises the utility of the visual survey methods presently used. This study investigated an acoustic censusing method - the click counting method - which although thought to have much promise in estimating sperm whale abundance, has received little research effort to date. Between summer 1996 and winter 1998, I spent four seasons in the field off Kaikoura, New Zealand, collecting information on the vocal behaviour of sperm whales. This added to a larger data set that began in 1990. These data were used to estimate two of the three core parameters required for the click counting method - mean regular click rate, and the percentage of time sperm whales spend regular clicking (Figure A). Overall Mean Regular Click Rate 1.272 clicks s·1 (± 0.029 95% Cl) Percentage of Time Spent Regular Clicking 60% in summer (cv = 19%) 62% in winter (cv = 25%) Estimated Effective Range of the Hydrophone 2 nautical miles Jake (1993) Figure A. Three core parameters required for the click counting method of abundance estimation. Data in bold were estimated in this study. The range of the hydrophone (shown in italics) was estimated by Jake (1993). Regular click rate during three dives each of three identified whales was analysed using specially written software (Moby Click 1.0B; Jake, 1996). Mean regular click rate did not differ significantly within dives, among dives of the same whale, or among whales. Thus, it is appropriate to census sperm whales at Kaikoura using one overall mean regular click rate (Figure A). The percentage of time sperm whales spent regular clicking during an entire dive cycle (from fluke-up to fluke-up) was determined from 54 recording 'sections', and was shown to differ between seasons (Figure A). These data were combined with the third core parameter, the estimated effective range of the hydrophone used to make recordings (Figure A), and an equation developed using these core parameters to estimate sperm whale density and abundance. Sperm whale abundance was estimated in five-minute recordings made at six stations along each of four transects at Kaikoura. It was clear that in some cases, noise was mistaken for sperm whale clicks during analysis using Moby Click 1 .OB. Thus, these mean regular click rates and the resulting abundance estimates were probably artificially increased. Including frequency-domain rules (as well as the time-domain rules that Moby Click 1.0B uses), to detect clicks, and the ability to play-back sound should decrease these errors substantially. Despite these problems, results showed that abundance estimates were similar to the number of whales encountered on the same field day. At present, the effective range of the hydrophone used to make recordings is poorly known. An accurate estimate of this range is fundamental to a census using the click counting method, and should be a focus of further study. This study presents the most comprehensive research on sperm whale click rate and the proportion of time sperm whales spend clicking to date. It is the first to present these data for whales at Kaikoura, and to our knowledge, the first application of the click counting method in assessing absolute abundance. Results from this study show that click counting is certainly worth further investigation.
Advisor: Dawson, Steve
Degree Name: Master of Science
Degree Discipline: Marine Science
Publisher: University of Otago
Research Type: Thesis