Description, Dueting, Seasonal Variations, and Individual Identification of the Vocalisations of the Brown Kiwi (Apteryx mantelli)
Vocalisations of the endangered Brown Kiwi (Apteryx mantelli) are currently used in a nationwide monitoring program to assess the health of a number of small remnant populations. Only an estimate of population health can be obtained from these surveys. This has to form the basis for a number of critical management decisions. The ability to identify kiwi individually during monitoring would greatly improve the accuracy and quality of information obtained and provide a wider base of knowledge when management decisions are made. This study produces a detailed description of the male and female kiwi call, including duetting and seasonal variations. It then explores whether kiwi demonstrate individually distinctive vocalisations, which could be used in conservation management of kiwi.
Calls from seven male and four female kiwi were recorded from Rarewarewa, Whangarei between the 5th of May 2003 to the 31st of April 2004. Two call variables (call duration, number of syllables), two temporal syllable variables (syllable duration, syllable gap duration) and five syllable spectral variables (frequency with the most amplitude, high frequency, low frequency, start frequency, end frequency) were measured from 48 male and 14 female calls. Variables from each call were used to describe and classify calls using one way repeated measures ANOVAs and Discriminant Function Analysis. Males contained on average 24 syllables and females containing on average 22 syllables, with calls from both sexes lasting on average 28 seconds. Male and female kiwi calls differed in multiple spectral, temporal and structural features. Dueting rates and behaviour differed between pairs. Some pairs almost always called in duets, whereas others never dueted. Dueting behaviour also changed seasonally with fewer duets occurring in the non-breeding season and with females initiating more duets in the breeding season. Call structure did not differ between the breeding and non-breeding season. Although the sample size was not sufficient to test for this.
Statistically significant differences occurred among nearly all variables between individual male and female kiwi calls, despite statistically significant difference occurring between syllables within calls. Frequency variables were the most important variables in discriminating between individuals, but multiple spectral and temporal variables were needed to separate all individuals. Discriminant Function Analysis was able to correctly classify 87.5% of male and 85.7% of female calls correctly when using the means of the seven syllable variables and the two call variables and 68.2% of male and 66.8% when using the values from each syllable. The latter was improved to 85.9% of male calls correctly classified when only 3 syllables from the middle of the calls were used in DFA, reflecting the variation found at the beginning and end of calls.
Brown kiwi show strong signs of individually distinctive vocalisations, which remained consistent over a one year period. The conservation implications of individually distinctive vocalisations and how this feature could be incorporated into the current monitoring of kiwi are discussed.