PhD Thesis
This thesis investigates the use of bioacoustic technology for monitoring breeding
behaviours in two endangered subspecies of black-cockatoo, the Kangaroo Island glossy
black-cockatoo, Calyptorhynchus lathami halmaturinus, and the south-eastern red-tailed black-cockatoo, C. banksii graptogyne. Over three years, I collected breeding behavioural data, including animal sound data and video footage, at nests of glossy black-cockatoos on Kangaroo Island in South Australia and red-tailed black-cockatoos in the Casterton region of Victoria. This thesis comprises an introduction (Chapter 1) and synthesis and conclusion (Chapter 6) and four research chapters (Chapters 2 to 5).
In Chapter 2, I provide a literature review of the potential use of bioacoustics in
monitoring animal behaviour for conservation. I argue that bioacoustic studies should better incorporate knowledge of species’ vocal behaviours, to improve the resolution of context-specific data. Animal behaviour is often relevant to conservation, and bioacoustics could greatly improve our ability to acquire behavioural data. A necessary first step is to understand the vocal behaviours of a species to the extent required for conservation monitoring. To this end, in Chapter 3, I describe the nest-associated vocal behaviours of the Kangaroo Island glossy black-cockatoo and the south-eastern red-tailed black-cockatoo. Over two breeding seasons, I recorded daily vocal activity at nests using autonomous sound recorders. Combined with behavioural observations, including video footage, I identified vocalisations from six behavioural contexts: birds in flight, while perched, during begging (adult females), during courtship displays (adult males), when entering or sitting near to the nest hollow entrance (adult females), and from nestlings. In total, I describe 14 putative call types for the glossy black-cockatoo and 11 putative call types for the red-tailed black-cockatoo. For both subspecies, the female nest call and nestling calls are the most conspicuous vocal indicators of active nesting.
In Chapter 4, I investigate the utility of bioacoustics for monitoring nest outcomes in the Kangaroo Island glossy black-cockatoo and the south-eastern red-tailed black-cockatoo. First, I provide the vocal signal of a successful fledging event for both subspecies. This unique behaviour is clearly indicated in spectrograms and provides a bioacoustic signal with which to confirm nest outcome. I then investigated automated methods to identify nest success or failure using an open-source call recogniser implemented in the monitoR package in R software. Constructed from templates of nestling calls, the recogniser was tested on 3 x 3- hours of sound data from early, mid and late stages of the recording period for each nest monitored (n = 23 for the glossy black-cockatoo; n = 21 for the red-tailed black-cockatoo). Daily nest activity was correctly assigned in 61.9% of survey days analysed for the red-tailed black-cockatoo, and 68.1% of survey days for the glossy black-cockatoo. Importantly, the recogniser successfully detected the fledging event in almost all cases (five out of six fledging events in the glossy black-cockatoo, and two out of three fledging events in the red-
tailed black-cockatoo). Precision of individual detections was moderate, with many false
positives. Manual verification of outputs was required, making this a semi-automated
method.
In Chapter 5, I examine the ontogeny of vocalisations in nestlings of the Kangaroo Island glossy black-cockatoo and the south-eastern red-tailed black-cockatoo. To determine whether nests can be aged from sound recordings, I examined changes in acoustic structure and daily call rate during nest development. I determined that nestlings vocalised from about 4 weeks of age, but calls were soft and infrequent until about 6 weeks. Daily call rate increased over time, especially in the final week of nesting. Peak amplitude and low frequency of nestling calls increased significantly with development. Call duration increased significantly for the glossy black-cockatoo, but not for the red-tailed black-cockatoo. Average entropy declined significantly for both subspecies. Aggregate entropy declined significantly for the red-tailed black-cockatoo but not the glossy black-cockatoo. Together, these changes in call rate and acoustic structure provide a useful way to broadly categorise nest age from sound recordings
and thereby improve knowledge on nestling survival. This knowledge may be useful in future studies examining the influence of habitat variables, such as food availability, on nest development and nestling survival across landscapes.
This thesis presents novel bioacoustic methods for monitoring breeding in the Kangaroo
Island glossy black-cockatoo and the south-eastern red-tailed black-cockatoo. I demonstrate that these subspecies exhibit diverse repertoires at nests. I describe the vocal signal of fledging, which is a direct measure of breeding success, and provide an open-source call recogniser to aid sound data processing. Finally, I describe the ontogenetic development of nestling vocalisations. I conclude that bioacoustics has potential to greatly improve nest monitoring of the endangered Kangaroo Island glossy black-cockatoo and the south-eastern red-tailed black-cockatoo, to the benefit of conservation.