VOL. 12 (3) SEPTEMBER 1987 83 AUSTRALIAN BIRD WATCHER 1987, 12, 83-90 Vocal Behaviour and Annual Cycle of the Western Bowerbird Chlamydera guttata by JAMES M. BRADLEY, 2 Benbullen Boulevarde, Kingsley, WA. 6026 Summary Bowers of the Western Bowerbird Chlamydera guttata were observed at three sites in coastal Western Australia over three years (1977-1980) in order to quantify song output, mimicry and bower attendance and to determine the timetable of events in the birds' life cycle. Tape recordings of vocalisations at bowers were obtained at regular intervals in all months, and a scoring system was used to measure the rate of song output. Data were also obtained on breeding parameters. Bower-owning males were present at bowers all year, and maintained song output through the year with a major peak in September to December. Mimicry followed a similar pattern, with a peak in July to October. The rate of bower visitation also peaked in these months (August to December). Courtship and copulation were observed in August, eggs in August-September, nestlings in October and fledglings in September and October. Clutch size (n=4) and brood size (n=2) were invariably two. After the mating period, while females were nesting, groups of birds (presumably sub-adult males) visited bowers. Display and visitation rates were low from December to the next breeding cycle. Factors inducing mimicry included intrusion by raptors and humans. It is concluded that males advertise the location of new bowers by increased song output, and that mimicry serves partly as a defence mechanism. Introduction The Western Bowerbird Chlamydera guttata, now regarded as distinct from the Spotted Bowerbird Chlamydera maculata, is a cryptically coloured, sexually monomorphic species inhabiting wooded rocky country in Central and Western Australia (Schodde & Tidemaqn 1986). Chlamydera bowerbirds are adapted to the arid and more open areas of northern and inland Australia, contrasting with genera such as Ptilonorhynchus which inhabit wet forests on the east coast. Although not as well known as its eastern counterpart, the Western Bowerbird's biology has been documented in general terms. The breeding season (months in which eggs are laid) extends from July to February, varying according to rainfall and inhibited by drought (Marshall 1954, Robinson 1955, Storr 1977, 1984, 1985). Bowers are attended from at least May to December (Slater 1974). The nest has seldom been described (cf_ Kolichis 1979); the species is probably promiscuous and uniparental since the Spotted Bowerbird is known to be so (Schodde & Tidemann 1986). The bower, displays and vocalisations have been described (Serventy 1955, Marshall 1954) and are like those of the Spotted Bowerbird (Schodde & Tidemann 1986). The species' display vocalisations are known to include mimicry of predatory birds (e.g. Serventy & Whittell 1976, Howard 1986), which enhances the aggressive nature of territorial song and also acts as an anti-predator device (Robinson 1974, 1975). The Western Bowerbird is found throughout the Cape Range of Exmouth Peninsula, north-western Australia. I observed bowers of this species over three years, and this paper reports on the pattern of vocalisation in relation to events in the birds' annual cycle (establishment of bowers, courtship and breeding, trends in numbers of birds at the bower). I also report on the factors eliciting mimicry from birds at bowers. Study area and methods Three study sites on the Exmouth Peninsula (22 °00 'S, 114 ooo 'E) were selected. Each site of approximately 75 m square was centred on an active bower. The site on the eastern side of the Cape Range was at Mowbowra Creek; those on the western AUSTRALIAN 84 BRADLEY BIRD WATCHER ---.?'~,!~.~ I~___ r-,;:F ':f X ll'­ e -:1 CAPE<{ ~ X 1 -r~ x tff r-; JV XX ~I _"9 ~ I 'f ':{ ~· j. I RANGE : ~ :( I ~ ~ r--' J. NATIONAL : ~ ;J, I ~ I ~ .:;( ,----~ /~' X::r rff ~ it--' ~ fo. PARK~ 0 ~ ~ J:. 0 ~ J;.. Learmonth I ?rt '11''1';\ J_ I 't' f. I .J. fi. - -~ ~ j. : :\ t:- 1 I I I ------ -' Legend X • Study Bowers X x Other Bowers Figure 1. Map of Exmouth Peninsula, Western Australia, showing location of study bowers (•) and other bowers (x). VOL. 12 (3) SEPTEMBER 1987 Western Bowerbird, Vocal Behaviour 85 side were situated at Mangrove Bay and Tantabiddi Creek (see Figure 1 which shows the location of study bowers and all other bowers found along the Cape Range). Tables 1 and 2 show mean monthly temperature and monthly rainfall recorded at the Learmonth meteorological station (c. 20 km south of the study area) during the study period. Bowers were observed for a total of more than 300 hours. Tape recordings of activity at the bowers were made from 0800 to 1000 h weekly from August 1CJ77 to August ICJ78 (Mowbowra Creek bower), August 1CJ77 to November 1979 (Mangrove Creek bower) and May 1CJ78 to August 1980 (Tantabiddi Creek bower). A cassette tape recorder (National Panasonic RQ-2133) was left recording at the bower while I retreated, usually 2-3 km away. When present, I was 1-2m away and in full view of the birds (i.e. no hide was used) in order to gauge their reactions. Vocalisation at the bower was quantified using a scoring system: if vocalisation occurred during the first 10 seconds of any minute it was scored as one point, the maximum number of points being 120 for any two-hour recording session (i.e. one point per minute). Each time mimicry was heard, I attempted to determine its cause. Only mimicry inside the study site was recorded. Although male and female Western Bowerbirds are similar in plumage, my familiarity with the species enabled me to tentatively sex the birds at the bower by their behaviour, particularly the displays of the bower-owning male and the responses of visiting birds during the breeding season. If the male adopted an aggressive posture and did not allow a visiting bird within a metre of the bower, the visitor was assumed to be male. If the bower-owning male allowed another bird to approach the bower and he courted it with raised nape patch or display objects in his bill and his song was not harsh, the visitor was assumed to be female. Juveniles were recognisable by their smaller size, lack of a nape patch and the fact that they foraged with, and were fed by, adults. The following terminology was used to categorise mimicry. Induced by bird of prey: a raptor entered the observation area and the bower-owning male or other bowerbirds at the study site broke into mimicry. Induced by observer: upon my intrusion at the bower site, the bowerbird broke into mimicry. Unknown causes: the bowerbird at the bower broke into mimicry when I was outside the immediate observation area and invisible (though possibly audible) to the bird, and no obvious intrusion was made by birds of prey. Results Bower site characteristics Two bowers (including the Mowbowra Creek site) to the east of Cape Range, near the shores of Exmouth Gulf, were in scrub and spinifex near creek beds. They were under Casuarina trees and easily seen. No further attempt was made to locate bowers on the eastern side. On the western side of the range, 12 bower areas (including the other two study sites) were located under native figs Ficus platypoda. Of the 12 bowers, two were on the floor of gorges, four were on the ridges of Cape Range and six were on the coastal flats away from the range. None of these western sites was visible from more than 2 m away. Many of these bowers were approximately 2 km from the nearest known neighbouring bower. Seasonal pattern of vocalisation Figure 2 shows the monthly level of song output at each of the three bowers observed. The data show that the Western Bowerbird maintains song output throughout the year, AUSTRALIAN 86 BRADLEY BIRD WATCHER 250 ············ M owbowra Creek -- Mangrove Bay 200 --- T an tabidd i Creek .."' ·c;c: 150 ~ en c: 0 100 (/) 50 o I A s 0 N D Month Figure 2. Total monthly vocalisation at the three study bowers, August 1977 to July 1980. For explanation of scale see text. "'c: 20 ·;;0 "'~ 10 ~ .0 0 ..... 5 0 ci z 0 J F M A M J J A s 0 N D Month Figure 3. Total monthly mimicry irrespective of cause, all bowers (n =3) and years (1977-1980) combined. 30 25 -o"' 20 .!::: .....CD 0 15 ci z 10 5 0 J F M A M J J A s 0 N D Month Figure 4. Maximum number of birds at the bower per month during the observation period, aU bowers (n = 3) and years (1977-1980) combined. VOL. 12 (3) SEPTEMBER 1987 Western Bowerbird, Vocal Behaviour 87 with a major peak in September to November (extending to December in llJ77, the year in which there was some spring rain). The minor peaks at other times of the year appear to correspond with, or closely follow, months of high rainfall. Conversely, the decline in song in October llJ78 and 1m corresponded with a dry spring. While two bowers were being recorded simultaneously (August llJ77 to August llJ78 and May to November llJ79), high and low song output at one bower generally corresponded with the highs and lows at the other bower (Figure 2). However, there was some variation between years in the period of peak song output, and a suggestion of a minor peak in autumn (perhaps when display activity was renewed after breeding and moulting). Song output was least at bowers that had been allowed to deteriorate (Mangrove Bay site January-July 1m and November 1m onwards: Figure 2).