AUSTRALIAN 298 WATCHER

AUSTRALIAN BIRD WATCHER 1994, 15, 298-305

Captive Studies Reveal Extensive Vocal Range in the Sacred

by KARIN SITTE and DARRYL JONES, Faculty of Environmental Sciences, Griffith University, Nathan, Queensland 4lll

Summary While rehabilitating Sacred sancta following injury, we recorded many vocalisations previously undescribed and possibly unsuspected. This paper describes eight new calls in two groups: loud calls given primarily during the day, and softer calls produced main!) during the night in a social context. It is suggested that these calls reflect a much greater level of vocal behaviour for this species than previously expected. Introduction The Sacred Kingfisher Halcyon sancta is an abundant species, found throughout , Papua and (Blakers et a!. 1984, Coates 1985, Falla et a!. 1978). It is found in a wide variety of habitats and is one of the common native species apparently little affected by suburban development. Around Brisbane, it is common in houseyards and suburban parks containing mature eucalypts (Vernon 1968). Despite this familiarity, remarkably little has been published concerning the vocalisations of the species. Typically the Sacred Kingfisher is described as being usually silent though becoming noisy at certain times (Lang 1929, Coates 1985). Generally three calls are described (Forshaw & Cooper 1985): during the breeding season a staccato call, often rendered as kek kek kek, is produced (this is frequently termed the 'territorial call'); a soft, squealing kee-kee-ee-ee that is given near the nest; and a harsh, churring note produced when alarmed. Apart from the familiar kek call, attempts to describe other Sacred Kingfisher calls have produced a confusing array of onomatopoeic names with poor consistency and of little use. In addition, the only published sonagrams of calls for this species are of the kek call (see Saunders & Wooller 1988). In this paper we describe a number of vocalisations of the Sacred Kingfisher that have either not been recognised or have not been mentioned in the literature. These calls were produced by wild but captive , held temporarily during a rehabilitation program following injury. A feature of the rehabilitation program is that recovered birds are kept in social groups before release. Although this feature appears to have greatly enhanced recovery success rates (K. Sitte unpublished data), the social setting also allows considerable interaction among the birds. The calls were produced regularly by fully recovered birds as part of their regular vocal behaviour while captive. It is likely that many of these calls are difficult to detect in the wild; they are often produced in or close to nest-holes, during intimate social contact, and many are relatively quiet. Our observations show that Sacred Kingfishers have a much larger array of vocalisations than previously suggested. This paper seeks to provide information obtained from captive birds that may assist in a greater understanding of the social interactions of the species.

Methods During 1990 and 1991, 36 wild Sacred Kingfishers of both sexes were kept temporarily as part of an authorised wildlife rehabilitation program. All were captured by members of the public after the birds had sustained injuries by flying into windows or from mauling by domestic cats and they were subsequently brought to us for the purpose of rehabilitation. The number of birds kept at any one time varied between one and eight. Birds were kept together in an outdoor aviary measuring 1.2 m VOL. 15 (7) SEPTEMBER 1994 Sacred Kingfisher Vocalisations 299

0 Seconds Figure 1. Sonagram of Food-begging Call. The call was produced by a recently fledged juvenile. The three components, A, B, and C, are described in the text.

long, 0.65 m wide and 0.76 m high. The aviary also contained a suitably sized nest-box. The artificial diet used consisted of lean beef mince, canned crab meat, insectivore mix (Wombaroo), wheatgerm, calcium carbonate and vitamin supplement (Vita-Boost). Live food such as grasshoppers and Mosquito Fish Gambusia a./finis were provided when available. Recordings of calls were made opportunistically at all times of the day and night. Large numbers of calls by most birds were obtained and a selection of high-quality recordings representative of each call type was made for subsequent analysis. All recordings were made with a Radio Shack electret condenser stereo microphone, with a frequency response range of 50-15 000 Hz, connected to a Sanyo stereo cassette recorder. The calls were then analysed on a Kay Electronics Sana-Graph (using a filter setting of 150 Hz). From the sonag rams the duration of calls, pulse intervals and frequency range were measured for each call type. Unfortunately, sound reverberation from the walls of the cage or nest-box was common, an inevitable consequence of the recording environment. Although this added considerably to the background sound, the clarity of the call prints themselves remained largely unaffected. In most cases, all of the necessary measurements could still be obtained. For publication, however, traces of the original sonagrams were made. Results Call types The calls of the Sacred Kingfisher may be divided initially into two main groups based on relative loudness. The first group (Group 1) are calls loud enough to be heard clearly at a distance of 10 m or more. All of these calls were produced during the day. The second group (Group 2) of calls are soft calls that were audible to the human ear only if within 5 m. These calls were produced either at night or at any time when birds were inside the nest-box. Although the basis of this subdivision may seem artificial, the social contexts within which the two groups of calls were given were distinctly different: Group 1 calls were audible to other birds present during the day within a relatively wide radius of the caller, whereas Group 2 calls would be heard only by conspecifics very close to the caller typically while roosting or nesting. The calls are described in detail below. The names given to the calls relate mainly to the affinities of the sound produced; only the 'food-begging call' and 'predation scream' have been given a definite functional label.

GROUP I Food-begging call This call is composed of three sub-elements (Figure I) designated A, B and C. Component A consists of a main buzzing pulse broadly spread between about 0.8 and AUSTRALIAN 300 SITTE & JONES BIRD WATCHER

6

N' :r: ~4 (;' c: u [ &WP Figure 2. Sonagram of g-~ .. Predation Scream. u: 2 ~ ~ --

0 Seconds

Figure 3. Sonagram of Rolling Screech Call.

0 Seconds

2.6 kHz with the highest energy concentrated at 1.7 kHz. A prominent harmonic is evident within the range 3.7-3.9 kHz. This series is of variable duration though typically from 0.5 to 0.9 seconds. Component B is a low-frequency (between 0.6 and 0.8 kHz) pulse of 0.08-0.1 seconds' duration. Component C is a harsh broad-frequency buzz of 0.3 to 0.4 seconds' duration with the main power being in the 1.4-2.3 kHz range. A wide complex of diffused harmonics was also present from about 2.4 to 5.6 kHz. These three sub-elements are produced sequentially and with virtually no detectable intervals. The most common sequence produced by a hungry juvenile is A-B-C-B-A. This sequence may be repeated continuously but some segments may be omitted (e.g. A-C-B-C-B etc.).

Predation scream A harsh scream (Figure 2) consisting of elements of approximately 0.2-0.3 seconds' duration and a frequency range of 2-7 kHz, the most powerful band being between 3.5 and 6 kHz. These elements may be emitted singly or repeatedly. This call was typically uttered during handling and is often given by birds when captured by cats.

Rolling screech call This is a complex call (Figure 3) and our observations of young birds suggest that it may develop via the food-begging call. It is made up of various components which give it the rolling quality, the most obvious being the array of harmonics of similar power that are evident within a frequency range of up to 4kHz. The harmonics may be described as five or more undulating bands of about 0.2-0.4 kHz in range. These VOL. 15 (7) SEPTEMBER 1994 Sacred Kingfisher Vocalisations 301

Figure 4. Sonagram of Churring Call.

0 Seconds

Figure 5. Sonagram of Scream Whistle.

0 Seconds 0.5 bands coalesce at the end of the element to produce a very wide-frequency (0.5-5.0+ kHz) screech of equal power throughout the range. These calls may be emitted in sequence, each being about 0.4 seconds in duration. The call varies enormously oetween individuals in pitch, duration and structural detail, but the sound produced by each bird appeared to be individualistic, enabling the birds to be recognised by ear.

Churring call This call (Figure 4) is similar to the rolling screech call, but without its individuality. A bird may produce a rolling screech call with a churring element at the end, or the churring call may be produced alone. These configurations were usually repeated several times. (Wild birds have been recorded repeating this call up to 20 times in succession; unpublished data). The call consists of complex rapidly ascending and descending frequencies fluctuating within a range of about 1 kHz. The main frequency rises through the call, from about 1 kHz to about 1.6 kHz. There are five harmonics evident, the first being almost as strong as the fundamental. The call lasts on average about 0.5 seconds.

Scream whistle This is a piercing call (Figure 5) of about 0.4 seconds' duration. There are two high-energy bands, one starting at about 3.5 kHz, decreasing to about 3.0 kHz, before dropping abruptly to 2.6 kHz. A higher band (starting at 6.5 kHz and decreasing to 5.5 kHz) followed the same pattern, before falling to fuse with the lower band. AUSTRALIAN 302 SITTE & JONES BIRD WATCHER

(a) ~' (b)

N J: ~4 g ( (jull I' g 0" 0) 0:: 2 t:'.. ~

0 Seconds 0.5 0

Figure 6. Sonagram of Short Screech Call (a) and Long Screech Call (b).

6 (a) A

N J: ~4

~ c ,_ g 8" u: 2 " 0 Seconds 0.5

Figure 7. Sonagram of Short Whistle Call (a) and Long Whistle Call (b).

GROUP 2

Screech calls These calls occurred in two obvious patterns, distinguishable by duration. Short screech calls (Figure 6a) consist of a small number (typically two or three) single­ note calls of duration 0.08-0.16 seconds. The notes contained numerous harmonics clearly evident to at least 6kHz. Long screech calls (Figure 6b) lasted up to one second and consisted of simpler notes and screech elements produced in a predictable sequence. These calls had a brief but characteristic ascending commencement and finishing note.

Whistle calls Sacred Kingfishers produce two types of whistles differing distinctly in duration. The short whistle (Figure 7a) is a single brief note of 0.04 seconds' duration. The fundamental is most powerful at about 1.3 kHz. There are at least four clear harmonics. Long whistles (Figure 7b) are 0.7-0. 9 seconds in duration and typically consist of two similar elements. These gradually descend from a wide starting pulse in the range 2.5-3.9 kHz to about 1.5 kHz. There are at least three harmonics. VOL. 15 (7) SEPTEMBER 1994 Sacred Kingfisher Vocalisations 303

6 \ \ lj. '\ ;;: " 84 ' ,, ~ ~ " "<:r 0::" 2 '~

0 Seconds

Figure 8. Sonagram of Chirp Call.

Chirp call Chirps (Figure 8) are a succession of very similar elements, uttered as a phrase of three to six notes. Each element is descending in the range 2.0 to 1.5 kHz and each is of 0.08 to 0.12 seconds' duration. At the start of the phrase the elements are about 0.02 seconds apart and the interval increases to about 0.08 seconds.

Complex calls All of the calls included in Group 2 were frequently incorporated into complex calls. These were produced primarily at night and typically when the birds were roosting in the nest-box. These complex calls were occasionally heard during the day but these were always associated with birds sheltering in the nest-box. Such daytime vocalisations were qualitatively different from those produced at night, consisting entirely of screeches and whistles. The most frequently occurring call pattern observed was a form of duetting. These interactive calls appeared to be synchronised, with one bird providing different components to a vocalisation commenced by another and with virtually no overlap. These duets were performed by both sexes in the company of another bird of either sex and of any age other than young less than one month old. Duetting was also not limited by the presence of other birds; it was commonly heard from a nest-box containing up to four birds. None of the birds that sang together was known to be pair-bonded. (Because the birds were normally hidden from view within the box, it was not possible to identify the bird initiating the calling). When duetting, the initiating bird produced a short screech followed by a long whistle. The responding bird almost always interrupted the first bird's whistle (after precisely the same interval: 1.4 seconds) with a screech. The initiator then completed the call with a modified but stereotypic long whistle. Figure 9 shows this typical sequence clearly: short screech, long whistle (bird 1), screech (bird 2), modified long whistle (bird 1) . This vocal interaction was very common and virtually identical each time. Another common vocal interaction was a short screech followed by a phrase of chirps. Some of these vocal interactions were fairly complex and involved more than two birds and included a wider variety of call types. For example, a typical 'group' song consisted of a short screech, long screech with a whistle quality, a series of chirps, a long whistle and finally another short screech. AUSTRALIAN 304 SITTE & JONES BIRD WATCHER

0 Seconds

Figure 9. Sonagram of a Vocal Interaction. This was a common call, consisting of two modified whistles given by one bird, with a second bird adding the Short Screech clearly seen at about 0.3 seconds.

Context in which the calls are used The predation scream was emitted by both sexes and only when a bird was threatened. This call was accompanied by a threat posture where the bird ruffled its feathers, particularly on the head, spread its wings and opened the beak. In situations of increasing threat the screams become more frequent. In captivity, the kingfishers produce this sound when first handled, but the call ceases after a few minutes although the threat posture will be retained. The begging call reported here was recorded from a recently fledged juvenile. This call differs from that produced by a hatchling. It has been observed previously (Lang 1929) that the call changes as the chick grows and develops. Eventually the call is modified to a rolling screech call which may have an important role in allowing recognition of individuals within social groups or families. The capacity for individual recognition has similarly been suggested for the territorial call (Saunders & Wooller 1988). The rolling screech call is produced throughout the day and often just before feeding. It was sometimes emitted during preening or for no obvious reason. Some birds call more frequently than others but it is not known if this is a result of the bird's position in the strict hierarchy formed within the group. The churring call appears to be a modification of the rolling screech call and was most frequently given before feeding. When one bird produced this call before feeding, others in the group typically flew down to feed as well. This phenomenon was observed among two groups of birds of different individuals. The scream whistle was recorded only once and was rarely heard. In one instance, a young male had been introduced to a group of one male and two females. The resident male produced the scream whistle which was answered with the same call by the newcomer. This was repeated twice, after which the new bird became silent. The two males did not fight, and it is possible that this call was used by the resident male to convey information on his dominant position within the group. The calls in Group 2 were produced equally by males and females but rarely by a single bird. Two or more kingfishers roosting together in a nest-box will normally produce all of the sounds in this group of calls. Vocal interactions were the most common type of call produced in social situations, but calls were also produced together in a wide variety of sequences. To date we have identified two main patterns which VVL. D lf) SEPTEMBER 1994 Sacred Kingfisher Vocalisations 305 are used frequently. One is the vocal interaction mentioned above, and this is also the most common. The other is a long screech or a short screech followed by a phrase of chirps. The combination of screech and whistle is the first form of complex call produced regularly by newly fledged young. Sacred Kingfishers sing spontaneously at night, or in the nest-box. A single bird occupying the nest-box will only make the short screech; if another bird joins the first (either in response to the first calling or in order to hide), the two birds will both sing. When placed in a box at night, calling is also triggered by changes in light outside the box or by sudden noises.

Conclusions The detection of this unexpectedly large vocal range in a familiar species is probably related to the social context in which the birds were kept during rehabilitation. Even healthy, wild-caught Sacred Kingfishers do not normally make good captives, being characteristically flighty and easily disturbed (e.g. Goudswaard 1990). The excellent rate of recovery of our birds compared with other rehabilitation programs (see Goudswaard 1990) is very likely to relate to the fact that the birds were kept in groups. Solitary birds only rarely call, and this is always the short screech. If there is no response from another kingfisher the calling stops. Although this study was based on birds in an admittedly artificial social environment, the intraspecific interactions that resulted led to the detection of many vocalisations otherwise undescribed. Although this captive situation precludes offering reliable functional explanations for most of these vocalisations, it does suggest that Sacred Kingfishers, at least those in southern Queensland, have a greater expression of social signals than previously expected. The fact that many of the calls were produced or elicited by the presence of conspecifics points to the possibility of extensive communication between birds other than those associated with courtship. Researchers making field studies of this species may need to be aware of this relatively hidden social dimension.

Acknowledgements We sincerely thank Ken Shultz of the University of Queensland for assistance in sonagram production, Dr Karen Bibo for advice on diet and husbandry, Drs Louie Filippich and David Banks for veterinary advice, and Pearl Symrnons for orthopaedic surgery. This work was carried out under permits from the Queensland Department of Environment and Heritage and with the clearance of the Ethics Committee of Griffith University. Finally we thank David Curl, Penny Slater and Stephen Debus for critical assistance with the manuscript. References Blakers, M., Davies, S.J.J.F. & Reilly, P.N. (1984), The Atlas of Australian Birds, Melbourne University Press, Melbourne. Coates, B.J. (1985), The Birds of Papua New Guinea, vol. I , Dove, Brisbane. Falla, R.A., Sibson, R.B. & Turbott, E.G. (1978), The New Guide to the , Collins, Auckland. Forshaw, J. & Cooper, W. (1985), Kingfishers and Related Birds, Vol. II, Lansdowne, Sydney. Goudswaard, R. (1990), 'Breeding the New Zealand Kingfisher at Wellington Zoo', A vic. Mag. 96, 10-19. Lang, T. (1929), 'Notes on the Sacred Kingfisher' , Vic. Nat. 45, 253-258. Saunders, D.A. & Wooller, R.D. (1988), 'Consistent individuality of voice in birds as a management tool', Emu 88, 25-32. Vernon, D.P. (1968), Birds of Brisbane and Environs, Queensland Museum, Brisbane. Received 23 January 1993, revised 11 May 1994 •