VOL. 8 (3) SEPTEMBER, 1979 BOEKEL: Purple-crowned Fairy-wren 91

Notes on the Status and Behaviour of the Purple-crowned Fairy-wren Malurus coronatus in the Victoria Downs area, . By CON BOEKEL, Ngukurr, , via Darwin, Northern Territory, 5791.

Introduction The study area's climate, topography, etc., are described elsewhere (Boekel, 1979) . This paper summarises data gathered from February 1976 until November 1977. Many data were gathered incidentally. How­ ever, systematic attempts were made to precisly define the local limits of distribution, time of breeding and habitat requirements. Local Distribution The Purple-crowned Fairy-wren is found in only two localities in the study area: the first around the Station Homestead Pool on the Wickham River, the second along the Victoria River near Pompey's Knob. In August, 1976, the Wickham population consisted of four breeding pairs; the Victoria population was six breeding pairs. There were also 13 de pen· dent young and 3 eggs. The Victoria population was spread along 5 km of river; the Wickham population along 3 km of river. In addition to these localities, two others were reported. Bob Verberg gave accurate descriptions of the species for Pigeonhole, and Kim McKenzie reported seeing it in pandanus at Wattie Creek, further to the south. The Victoria and Wickham birds are separated by over 20 km of . In between are some areas which would seem to contain suitable habitat for the birds. The question thus arose: "Why did they lack Fairy-wrens?" Habitat In an effort to answer this question, various elements in the habitats were isolated. The features in the immediate vicinity of each group were noted. They were Pandanus arquaticus, cane grass, paperbark trees, riverine shrub layer, flood debris and width of water (narrow equals less than 15 m). Table 1 gives the results for the eight territories surveyed: TABLE 1: River River Pandanus Paperbark Cane grass shrubs debris wide narrow I. * * * * 2. * * * * * 3. * * * * 4. * * * * * * 5. * * * * 6. * * * * * 7. * * * * 8. * * * * * 7 7 6 3 7 6 On the basis of this Table, five elements would seem to be important to the Fairy-wrens, a narrow section of the river, stands of pandanus, shrubs, debris and paperbarks. Observations on how the birds utilised their habitat correlate well with these factors. Paperbarks did not seem to figure directly in the activities of the birds but may have been AUSTRALIAN 92 BOEKEL BIRD WATCHER important in keeping the riverbed moist through their root systems and their shade. Several sections of both the Wickham and Victoria Rivers in the study area contained varying amounts of these five elements. The Yarralin/ Deep Creek area on the Wickham River combined all five elements along with stands of cane grass stretching for over a kilometre. The Wave Hill Crossing area, less than a kilometre from the Homestead Pool population, contains extensive stands of pandanus, narrow reaches, shrubs a nd paperbarks as well as much debris. The Dashwood Crossing area on the Victoria River contains all elements except that of a narrow stretch of water. It is marked by one of the largest tracts of cane grass in the study area. Yet not one of these contains Fairy-wrens. The anomaly is accentuated when the distribution of the Fairy-wrens is compared with those of the Shining Flycatcher Myiagra alecto, Northern Fantail Rhipidura rufiventris and Lemon-bellied Flycatcher Microeca flavi­ gaster. Table 2 shows their distribution for various localities:

TABLE 2:

= Homestead Pool; 2 = Pompey's Nob; 3 = Wave Hill Crossing; 4 = Yarralin/ Deep Creek.

1 2 3 4 Purple-crowned Fairy-wren ... * Northern Fantail ... * * * Shi ning Flycatcher * * Lemon-bellied Flycatcher * *

All three birds in the comparison are insectivorous and live primarily in th e shrub layer by the waterside. Shining Flycatchers and Lemon­ bellied Fycatchers feed mainly off insects caught on the ground, although they use different methods to those of the Purple·-crowned Fairy-wrens. Given the distribution of these birds, it would seem that Deep Creek ought to be suitable for Purple-crowned Fairy-wrens. If the assumptions made are correct, then the birds are absent from some areas of habitat which are suitable. A possible explanation for this state of affairs is now suggested. It is believed that the di stribution of the Fairy-wrens was more wide­ spread before the introduction of cattle and that a process similar to that in the Kimberleys (Smith & Johnstone, 1977) has also taken place in the study area. The effect of cattle, particularly on cane grass and the riverine understorey was to reduce the Fairy-wrens to a few refuge areas. How­ ever, the effect of cattle is not constant. In times of very bad drought they concentrate about the more permanent water, eating the cane grass and breaking down or eating the shrub layer (1975, 1976, and 1977 were good years and very little damage was done to either the cane grass or the shrub layer). Fred Schull, a long-time resident of the area, reports times when great stretches of the riverbanks were reduced to dust with only pandanus and the larger trees surviving. The introduced cats would VOL. 8 (3) SEPTEMBER, 1979 Purple-crowned Fairy-wren 93

Habitat: Wickham River, N.T. Pla te 35 P hoto by Con Boeke!

Pandanus thicket, Wickham River, N .T. P late 36 Photo by Con Boeke! AUSTRALIAN 94 BOEKEL BIRD WATCHER probably find the Purple-crowned Fairy-wrens easy prey on such occa­ sions. With the return of good seasons, suitable habitat would regenerate in patches isolated by tracts in which erosion and Parkinsonia aculeata would presumably act as barriers to dispersing Fairy-wrens. This failure to recolonize suitable tracts of cane grass has evidently left a vacant niche. Red-backed Fairy-wrens Malurus melanocephalus were recorded in cane grass in the Deep Creek area. Utilization of Habitat In a visualised cross,-section of the habitat of the Purple-crowned Fairy­ wrens, on only one occasion were they found at a point beyond the brow of the riverbank on either side. Nest sites were all below the river bank brow, and all but one would have been flooded at least once a year between 1975 and 1977. Pandanus thickets are used for nesting, preening, foraging and as refuges in times of danger. A Yellow-tinted Honeyeater Lichenostomus fiavescens pursued a male Fairy-wren but turned away as it entered a clump of pandanus. A White-gaped Honeyeater Licenos­ tomus unicolor chased a pair of Fairy-wrens from the ground into a clump of pandanus, but after a furious chase inside the clump it gave up. A Northern Fantail Rhipidura rufiventris chased a female Fairy-wren but stopped immediately the Fairy-wren entered the pandanus. Feral cats are common in the habitat but the sharp spines of the pandanus leaves possibly provide a deterrent to them. Two common goannas, V aranus gouldii and V. mitchelli, are not deterred and are potentially predators of eggs and nestlings. The Fairy-wrens forage both in the crowns and among the tangled boles and exposed roots of pandanus thickets. White­ gaped, Yellow-tinted and Brown Honeyeaters Lichmera indistincta rarely forage in the crowns. Once, preening was noted. The female was brooding and the male was perched in the lowest leaf of a pandanus crown. Aerial predators would have injured themselves on the spines in trying to reach the male. Terrestrial predators climbing the bole would have been easily avoided. Cane grass was used as a nest site. In addition it was used for foraging and as a refuge. Clumps were so dense that a machete would have been needed to penetrate them. Shrubs and flood debris were important sites for food and were also used as refuges. The remaining vegetation, the larger trees, mainly River Red Gums, Coolibahs and Paperbarks were associated with the Fairy-wrens habitat but did not seem to be used much. The width of water probably has some effect on the survival rate of the Fairy-wrens. Territories invariably encompassed both banks of a strip of river, the birds frequently crossing the water and perching in transit on any emergent logs. They often crossed the river when reacting to the observer as potential predator. Fledglings were also led across the river in these circumstances. Flight was strong and swift for up to 30 m. M any approaches to nests were made over water, which must reduce the chances of ground predators finding the nests. Areas of bare ground are invariably present in the territories. Towards the end of the dry season they are covered with a mat of leaves and twigs, moist underneath, which provide a haven for the food on which the birds feed. VOL. 8 (3) SEPTEMBER, 1979 Purple-crowned Fairy-wren 95

This summarises the principal elements of the habitat. However, this must be qualified in that it holds good only for The Dry season and parts of The Wet season. In each of three years all territories were partially inundated. During periods of inundation birds would be forced to leave their accustomed areas and live either in the emergent crowns of trees or above the riverbanks. Intensive efforts were made to find Fairy-wrens during floods but they were almost completely unsuccessfu l. Only once a party was noted above the riverbanks. It was an abnormally wary group. The loud calls normally associated with an encounter were com­ pletely lacking. It is possible that the birds were out of their territory and that their territorial calls were not being made as a consequence, thus rendering them extremely difficult to locate. The habitat in this in stance was a mixture of tangled long grass, low shrubs and annuals. At such times insects were abundant and food would not have been a problem for birds forced out of their normal habitat. The Population In mid-August, 1976, all territories were searched and the number and composition of the groups were noted. Table 3 shows the results:

TABLE 3: Group male female dependent young eggs immatures 1 2 2 2 3 4 5 1 6 2 7 2 8 2 9 2 10 3 11 (building) Groups 1-4 constitute the Wickham River population, and groups 5-11 the Pompey's Knob population. There are possible errors in this Table: some of the birds noted as immatures may have still been dependent; and group 4 may have had eggs. Despite this, there are some features which stand out, the first being the size of the groups. In almost every case they consisted of the breeding pair and the progeny of the year. At this stage there was one bachelor male. At times there have been up to three bachelor males, which do not penetrate the territories of the established pairs. Careful observation precluded their having brooding mates. Another feature is the rate of breeding success. Assuming that all clutches numbered three and that the dependent young all survived, the success rate was 58 per cent. Breeding Eggs were laid in July, August and September, at a time when nests were unlikely to be washed out. At the same time there was still sufficient moisture in the riverbed to ensure a good supply of food. On the few occasions building was observed, the female alone did the AUSTRALIAN 96 BOEKEL BIRD WATCHER work. Once during building, the male took on the begging posture used by young birds, with tail lowered, wings spread and slightly quivering. The female fl ew up and fed the male. The female brooded the eggs and nestlings for the first few days. Several times the male fed the female as she sat on the nest. Once a female hopped out of the nest and received some food from the male. When she re-entered the nest the begging calls of the young birds could be heard. When the female ceased brooding the nestlings, the male still did much more of the feed ing of the nestlings. In the first few days cloacal pellets are evidently swallowed. Thereafter they are carried away from the nest. Empty nests were invariably clean. A breakdown of feeding visits shows that the feeding activity was most intense earl y in the morning and the later part of the afternoon. Attempts were made to identify the type of food taken into the nest. T his was possible in some cases because both birds invariably hesitated a "hop" from the nest entrance. Foods identified were moths (12), other insects and arthropods (12), larvae (12), beetle (1), wasps (I), light-green grass­ hopper (2) and caterpillar (2). Identified foods represented about one in four or five of observations. When feeding the nestlings, the adults approached the nest by a series of six or seven hops via adjacent shrubs. On leaving they flew directly away. Both birds defended the nest sites with the loud "cheep, cheepa, cheepa" call. Once the male chased a Brown Honeyeater from two metres away from the nest; the clicks of his bill were plainly a udible. In response to my aproach to a nest, a female adopted the juvenile begging posture, givi ng the begging call " tsee, tsee, tsee". One female, flu shed while brooding eggs, flew a metre to the ground and adopted the "mouse run" attitude as it scurried away, with tail and head lowered and wings held out and down. When aroused males sometimes erected the crest feathers which would change in colour from purple to a flash of fiery-rose. At the nesHite and when protecting dependent fledglings, the male is as aggres­ sive as the female. At other times the male is the more retiring of the pair. Nests we re usually constructed in the crowns of pandanus, most often facing water and a metre or so above a nd away from it. Several nests had old leaves arranged near the entrance, particularly the leaves with only ve in s a nd stems left. These tended to break the outline of the nest. Conclusion Harrison & Parker (1 965) summarise the behavioural characteristics of the Fairy-wrens of the genus Malurus. Several types of behaviour of the Purple-crowned Fairy-wrens are apparently aberrant. Bachelor males, some of them in full breeding plumage, were not incorporated in the groups but were exil ed from the established territories. The rodent-run was performed by a female. On the basis of a limited number of obser­ vations, only the female built. Direct head-scratching was noted on a single occasion. A thorough study needs to be done to establish if these types of behaviour represent isolated or local aberrations or are normal for P urple-crowned Fairy-wrens. The opportunity to do so may not last long. There is ample evidence to show that Malurus coronatus has suffered a drastic decline si nce the establishment of the cattle industry. There is no evidence to VOL. 8 (3) SEPTEMBER, 1979 Purple-crowned Fairy-wren 97

show that the decline has halted. The next big drought may well be the last for this species, unless positive action is taken to prevent cattle from destroying its habitat. Further study may show that its reintro­ duction into isolated pockets of regenerated habitat will help its chances of survival. Acknowledgements I wish to thank Mr. Howard Jarman for his help in finding certain of the references and Dr. Glen Storr and Mr. Laurie Smith for checking the manuscript. Bibliography Harrison, C. J. 0 ., & S. A. Parker, 1965. The Behavioural Affinities of the Blue Wrens of the Genus Malurus. Emu 65: 103-113. Officer, H. R., 1964. Adaptability of the Purple-crowned Wren to loss of habitat. Emu 63: 340-341. Rothwell, V., 1962. Purple-crowned Wren No. 2. The Bird Observer, 370: 3. Sharland, M., 1962. Purple-crowned Wren. The Bird Observer, 370: 3. Smith, L.A., & R . E. Johnstone, 1977. Status of the Purple-crowned Wren (Malurus coronatus) and Buff-sided Robin (Poecilodryas superciliosa) in Western . West Aust. Nat. 13. Wheeler, J. R., 1965. Purple-crowned Wren: Habitat and call-note. Emu 64: 219. ABW