Part of this research project established how and when Australian magpies (Gymnorhina tibicen) vocalise and whether these vocalisations have specific patterns and functions.
The results showed that both major forms of vocalisations, carols and warbles, undergo seasonal cycles but not in concordance with each other. Warbling declines by two-thirds during the breeding season (July to September) while carolling increases nearly five-fold in the pre-breeding season (May to July) and then declines during the breeding season. Four functions of carolling, rather than the familiar one function of territorial defence, were identified: including also food calling, nest defence, and bonding. Carolling was found to be more common in groups rather than in pairs and exclusively an adult activity. Carolling is not only an expression of inter-group conflict but also as an intra-group vocalisation of disapproval, although, prior to breeding season, the main function of carolling is territorial defence. Diurnal vocal patterns also exist, showing a peak around midday.
Further, it was shown that there are substantial individual differences in vocalisations and that warbles are often concluded by phrases unique to the individual singer. It was noted also that males and females both sing and no song of any type is associated with the breeding season, in contradistinction to the model species in which male song is tied to the breeding season.
Vocalisations were described and their functions determined by observation of the magpie's behaviour when vocalising. The analysis distinguished between generic alarm and mobbing calls and between expressions of anger, distress and fear. New calls were also discovered; these included approval sounds (or parental coos), food calls and, instead of the two or three alarm call types that had been described in the literature, six different alarm call types were found, many of them with variations.
Of particular interest were the discovery of an 'eagle' alarm call and possibly one pertaining to the monitor lizard. As these calls appeared to signal warning about specific types of danger, they were discussed in the context of referential signals.
This research also described systemic use of mimicry by free-ranging magpies and addressed its possible function. Mimicked species were not just those that are easy to mimic, or closely allied in acoustic parameters or belonging to species with a similar repertoire. The mimicked sounds showed no signs of systematic incorporation in song (e.g. as sequences or embellishments). However, these mimicry examples were considered as an adaptive behaviour.
Physiological and anatomical characteristics of vocal production were considered in order to discover how sound types are produced and in what manner magpies employ their vocal apparatus. Body posture, beak position and laryngeal involvement were found to be important in sound production, in addition to the syrinx; indeed, the entire secondary sound system and even the entire body is essential to producing certain sound types. Mini-pauses were identified in warbling sequence, providing a possible reason for sustained song. Warbling sequences were also analysed frame-by-frame, showing that the beak was either closed completely during this vocal production or opened just a fraction and for very brief periods only. We know that the vocal tract has a filtering role, a closed beak during song causing a reduction in harmonics or even their absence and this is consistent with the warbling sequences, which tend not to have harmonics but may involve both sides of the syrinx (two-voice phenomenon). Further, having the beak closed may lower the fundamental frequency because it effectively lengthens the vocal tract. In fact, warbling sounds can be produced with minimal output of energy and result in non-harmonic, pure tone vocalisations of considerable duration. The duration of warbling follows a binomial frequency distribution (over one third of warbling bouts concluded within ten minutes but some continued for much longer). Warbling, furthermore, was found to have an association with sleep, a finding that raised important questions concerning its possible function in vocal learning.
Vocal development in magpies was studied and expressions of vocalisations were traced as precisely as possible. One specific finding was that the presence of parent birds suppresses vocal practice in their offspring; the latter cease vocalising as soon a parent is within 10- 15 m. The discovery of mimicry in a hand-raised magpie further assisted in clarifying questions of the extent to which magpies learn, practise and remember what they have learned. In mimicry, it was discovered that new sounds were reproduced after a latency of 48 hours and that there was some similarity between vocal development in magpies and human speech development (babbling).
Finally, the project asked and then linked repertoire size to learning styles and concluded that magpies are improvisers and have many vocalisations that are never repeated in quite the same way while other calls, such as specific alarm calls, mobbing, calls and food calls, are stereotyped.