Noisy Scrub-bird (Atrichornis clamosus) Recovery Plan

Alan Danks1, Andrew A. Burbidge2, Allan H. Burbidge2 and Graeme T. Smith3
ISSN 0816-9713

1. INTRODUCTION

1.1 Description and history of species

The Noisy Scrub-bird (Atrichornis clamosus (Gould 1844)) is a small solidly built bird with a strong pointed bill, powerful legs, graduated tail and short round wings. They are brown above with dark cross barring extending from the head to the tip of the tail. The dark bars are very fine on the head, broader and more obvious on the back and form irregular bands on the tail feathers. The underparts are paler with a buff coloured abdomen grading to bright rufous around the vent. The species is sexually dimorphic in size and plumage. During the breeding season, females have a mean weight of 34.6 g (n = 42, range = 31.5 g - 39.2 g) while males have a mean weight of 51.8 g (n = 56, range = 47.0 g - 57.0 g). Adult males have a dark grey band of variable width across the off-white throat and prominent white side flashes. Females have cream coloured throats and lack the dark band.

Noisy Scrub-birds have very limited flying abilities, being able to sustain flight for no more than a few metres. However, they frequently use their small wings to assist in rapid manoeuvring and short runs on the ground and in leaping from shrub to shrub. They are also agile climbers moving quickly from shrubs and sedges to the low canopy.

The territorial song of the male easily distinguishes the Noisy Scrub-bird from other birds within its range, but the cryptic colour and behaviour combine with the dense habitat to make them difficult to observe. However, they are not as secretive as casual observations would suggest, but are inquisitive birds, moving quickly to investigate the cause of any disturbance in their territory.

Scrub-birds (the Noisy Scrub-bird and the Rufous Scrub-bird A. rufescens of north-eastern New South Wales and south-eastern Queensland) form a small endemic monogeneric family of song birds, the Atrichornithidae, whose closest relatives are the lyre-birds (Menuridae) (Bock and Clench 1985, Sibley and Ahlquist 1985, 1990). Bock and Clench (op. cit.) place the two families in a superfamily Menuroidea within the sub-order Passeres of the Passeriformes, whereas Sibley and Ahlquist (op. cit.) consider the two groups as subfamilies of the Menuridae, which they place together with the treecreepers (Climacteridae) and the bowerbirds (Ptilonorhynchidae) to comprise the superfamily Menuroidea. A recent review of the taxonomy of Australian birds (Christidis and Boles 1994) supports retention of the family status of Atrichornithidae but treats the Climacteridae and Ptilonorhynchidae as not being closely related to the Menuridae. DNA-DNA hybridisation evidence (Sibley and Ahlquist 1985) suggests that the Menuridae originated 40-45 million years ago and that the scrub-birds differentiated from the lyre-birds 30-35 million years ago. Atrichornis and Menura are considered to be close to the base of one of two evolutionary radiations in the song birds (Fedducia and Olson 1982) and are among the last remaining members of an ancient, probably once more diverse radiation of Australo-Papuan songbirds.

Although the Noisy Scrub-bird with its loud call would have been known to Aboriginal people for many thousands of years, only one Aboriginal name has been recorded: Jeemuluk (pronunciation unknown; may be Djimoolook), from the King George Sound area (Serventy and Whittell 1976). The first European to report the species was John Gilbert, who found it in the Mt William area and at Drakesbrook, near Waroona, in November 1842. Later on the same journey he found it near Augusta and at Cape Leeuwin (Whittell 1951). In 1843 Gilbert travelled from York to Albany along a route similar to today's Albany Highway and noted that he first heard scrub-birds around Mount Barker, with numbers increasing as he approached Albany (Serventy and Whittell 1976, Smith 1977). George Masters collected seven specimens near Albany between 1866 and 1869, and William Webb collected eight specimens, also in the same area in the 1870s. The final nineteenth century specimen was collected by A. J. Campbell at Torbay in October 1889, and he reported having heard scrub-birds at Boojidup Creek, about 25 km north of Karridale, in November of the same year (Whittell 1943, Smith 1977) (Figure 1). Thus, the Noisy Scrub-bird was recorded from three separate areas in the south west of Western Australia in the first 50 years after it was discovered by Europeans: Mt William - Drakesbrook, Augusta - Margaret River and the Albany area. The birds were apparently most numerous in the Albany area which extended to Torbay 15 km to the west and Mt Barker, 50 km to the north. The birds were probably present in the Two Peoples Bay and Mt Manypeaks area at that time and William Webb considered that they might be found as far east as the Pallinup River (100 km east of Albany) (Whittell 1943), although on current knowledge this seems unlikely.

The apparently sudden disappearance of an unusual and unstudied bird was viewed with concern by ornithologists of the day and, beginning in 1904, extensive searches were mounted for the scrub-bird in its former areas. Despite these efforts no further records of the Noisy Scrub-bird were confirmed and the species was considered by many to be extinct (eg, Campbell 1920, Whittell 1943, 1951, Chisholm 1951) until it was rediscovered at Two Peoples Bay in 1961 (Webster 1962a; see Chatfield in prep. for a history of the rediscovery). Since the rediscovery of the Noisy Scrub-bird, searches elsewhere in its former range have failed to locate further populations.

1.2 Population trends, distribution and abundance

At the time of its rediscovery, the Noisy Scrub-bird population was very small and confined to the Mt Gardner area and near the current picnic area at Two Peoples Bay east of Albany. Between 1962 and 1966 irregular counts confirmed the existence of 40-45 singing males (Webster 1962b, Smith and Forrester 1981), indicating a possible total population of around 100 individuals. Examination of aerial photographs of the Two Peoples Bay area taken in 1946 indicated that, due to extensive fires in the Mt Gardner headland, there was only enough habitat at that time for between 21 and 31 scrub-bird territories (Smith 1985d). It is probable that the scrub-bird population has been as low as about 50 individuals in historical times.

Only singing males can be censused and this census provides an index to the actual population (Smith and Forrester 1981). Since 1970 the number of singing male scrub-birds has been counted annually, except 1978 and 1981. After a small decline the number of singing males began to show a steady increase from 1971. The Mt Gardner population has continued to increase since then showing only minor fluctuations over a 24 year period (Smith and Forrester 1981, Smith 1985a, in prep. c). The annual counts of singing males suggest that the population on Mt Gardner has not yet reached its maximum. In 1994 a total of 179 singing males were counted in this area.

Dispersal of individuals from the growing Mt Gardner population lead to the beginning of a small population around Lake Gardner, 3 km to the west, first recorded in 1979. By this time one male had also reached the west end of Moates Lake. The number of singing males recorded in this area increased steadily, reaching a total of 64 in 1987, and expanded into the area between Gardner and Moates Lakes. In 1988, however, growth in this sub-population was checked by a sudden decline, which, after an initial recovery, was repeated in 1991. The downward trend in the Lakes area has continued and by 1994 only 12 singing males were recorded. The initial decline occurred in the same year as unusually heavy rainfall and flooding of scrub-bird habitat in the lake margins. The subsequent decline, however, has not correlated with rainfall or flooding and the actual cause of the decrease in numbers of scrub-birds remains unknown.

Figure 1. Former and existing Noisy Scrub-bird localities

Figure 1 Former and existing Noisy Scrub-bird localities

Noisy Scrub-birds have also dispersed into the Angove River area up to 8 km north of Lake Gardner. One male was heard calling a few times from Angove Lake in 1982. The first male to persist in this area, however, was heard singing in swampy riparian habitat in the upper Angove River in 1987. Numbers have increased since then (Table 1), aided perhaps by continuing immigration, but local breeding has also occurred (Danks 1991). A small sub-population is now present in this area which, for management purposes, is combined with scattered singing males in adjoining country to the east as far as Normans Inlet.

Table 1.

Numbers of singing male Noisy Scrub-birds in each subpopulation in the Albany Management Zone, plus numbers at recent release sites

 

Mt Gardner

Lakes

Many-peaks

Angove-Normans

Mt Taylor

Mermaid

Bald Is.

Stony Hill

Total

                   

1962-1966

40-45*

-

-

-

-

-

-

-

40-45*

1968

50*

-

-

-

-

-

-

-

50*

1970

45

-

-

-

-

-

-

-

45

1971

44

-

-

-

-

-

-

-

44

1972

50

-

-

-

-

-

-

-

50

1973

58

1

-

-

-

-

-

-

59

1974

66

-

-

-

-

-

-

-

66

1975

68

1

-

-

-

-

-

-

69

1976

72

2

-

-

-

-

-

-

74

1977

72

2

-

-

-

-

-

-

74

1978

-#

-#

-

-

-

-

-

-

-#

1979

98

9

-

-

-

-

-

-

107

1980

108

8

-

-

-

-

-

-

116

1981

-#

-#

-

-

-

-

-

-

-#

1982

114

16

-

-

-

-

-

-

130

1983

120

18

4

-

-

-

-

-

142

1984

111

26

4

-

-

-

-

-

141

1985

122

35

12

-

-

-

-

-

169

1986

123

53

10

-

-

-

-

-

186

1987

129

64

15

2

-

-

-

-

210

1988

133

46

26

4

1

-

-

-

210

1989

144

54

33

4

1

-

-

-

236

1990

152

59

60

10

6

-

-

-

287

1991

163

32

81

11

6

-

-

-

293

1992

164

27

100

22

5

4

-

-

322

1993

172

25

156

29

9

3

2

-

396

1994

179

12

223

38

12

1

6

3

474

* estimates, not counts; see Smith and Forrester 1981, Smith 1985a

# no count made, only one partial census (Smith and Forrester 1981)

Translocations of Noisy Scrub-birds from the Two Peoples Bay populations to Mt Manypeaks in 1983 and 1985 resulted in a population of Noisy Scrub-birds in the area between Normans Inlet and the Waychinicup River which by 1994 had attained a population index of 223. The rate of increase in recent years has been very rapid and in some areas the density of singing males is approaching that on Mt Gardner. This population concentration now contains the largest number of singing males and there is sufficient habitat for growth and expansion to continue for some time.

Since 1990 scrub-birds have been translocated to Mt Taylor in the Gull Rock National Park (unvested), to an area near Mermaid Point in Waychinicup National Park, to Bald Island Nature Reserve south east of Cheyne Beach and, in 1994, to Stony Hill in Torndirrup National Park south-east of Albany. This has resulted in a scattering of small population concentrations in the area between Albany and Cheyne Beach.

Overall since 1970, as a result of habitat management and translocation, the Noisy Scrub-bird population has increased and expanded from the original parent population on Mt Gardner and now (1994) occupies an estimated 4330 ha within approximately 50 km of coastal and near coastal country south and east of Albany (Figure 2). In 1994 a total of 474 singing male Noisy Scrub-birds were counted in this area which is referred to as the Albany Management Zone. Based on a rule-of-thumb of 2.5 birds for every singing male, this probably represented around 1100 individual scrub-birds. Six population concentrations were in this area and each is denoted as a subpopulation. Translocated populations (mainly at Mt Manypeaks) accounted for 51.7% of the total number of singing males in 1994.

The total number of singing males and the number present in each of the current population concentrations is given in Table 1 and graphed in Figure 3.

1.3 Habitat

Smith (1985a) studied Noisy Scrub-bird habitat on Mt Gardner during the 1970s. Based on these studies he considers that, historically, scrub-birds in the south west may have been confined to the wetter areas within the distribution of marri (Eucalyptus calophylla) and jarrah (E. marginata); in particular to the ecotone between forest and swamp vegetation.

In the area between Oyster Harbour and Cheyne Beach, the core areas of male Noisy Scrub-bird territories are found in dense, long-unburnt vegetation characterised as low forest (5-15 m high), scrub/thicket and (rarely) heath. These vegetation formations occur in the gullies and drainage lines of hills and granite mountains and, in lowland areas, in overgrown swamps, lake margins and beside streams. On hillsides singing males may be found in the taller dense vegetation around springs, in depressions and at the base of rock faces. Most male territories on Mt Gardner occur in scrub/thicket associations and this appears to be the preferred habitat type. A smaller number of territories are found in low forest and only a few in heath. Occupied sites generally have a post-fire age of ten years or more and contain a dense stratum of shrubs and sedges, which provide essential cover for these semi-flightless birds, and a thick layer of leaf litter.

Nesting habitat requires the presence of pliable, long-leaved sedges for construction of the nest and dense clumps of sedges, shrubs or piles of debris as nest sites. The nest is lined with a cardboard-like substance made from rotten wood and the presence of decomposing wood may be an important resource for nest construction. More open areas with thick accumulation of leaf litter and a well developed litter fauna are required for feeding habitat. Good scrub-bird habitat appears to contain very dense cover interspersed with small open areas with suitable nesting areas nearby.

Increasing post-fire age seems to favour scrub-birds probably because successional changes result in more suitable vegetation structure and a richer leaf litter invertebrate fauna. The steady rise in the number of singing males in the Mt Gardner area over the last 25 years is generally correlated with the successful exclusion of wildfire from about 1970. In 1994, in the area between Oyster Harbour and Cheyne beach, scrub-birds were only found in sites where the vegetation was between 10 and 50 or more years old. It is possible that the current distribution of scrub-birds in very old vegetation may reflect the lack of younger vegetation in suitable sites, rather than a preference for older vegetation.

Figure 2. (missing) Noisy Scrub-bird population concentrations in 1994 in the Albany Management Zone

Figure 3. Total number of singing males and number in each subpopulation

Figure 3 Total number of singing males and number in each subpopulation

The maximum post-fire age at which vegetation can support Noisy Scrub-birds is not known and it has been suggested (Smith 1977, in prep. a) that there may be an upper limit to the post-fire age suitability of scrub-bird habitat. If this is so, this point appears not to have been reached on Mt Gardner where the post-fire age of the vegetation is 45 to 50 or more years and the number of scrub-birds is still increasing. Females are also still breeding in habitat that has not been burnt for more than 50 years. Given the exceptional age of the Mt Gardner vegetation it is possible that small scale cyclical changes (initiated, for example, by tree falls) may be mitigating the overall effect of post-fire successional changes.

1.4 Life history and ecological relationships

Various aspects of the biology of Noisy Scrub-birds have been described by Webster (1962a and b), Smith (1976, 1977, 1978, 1985a, b, c and d, in press, in prep. a and b), Smith and Robinson (1976), Robinson and Smith (1976), Smith and Forrester (1981), Smith et al. 1983, Smith and Calver (1984), Danks and Calver (1993), Welbon (1993) and Danks and Smith (in prep.).

Noisy Scrub-birds feed mostly on or near the ground, foraging in leaf litter, the bases of sedge clumps, dense shrubs and decaying wood. The diet of adult scrub-birds has been studied recently by examination of faecal material (Danks and Calver 1993). Although the remains of invertebrates from seven orders were found in faecal samples, the major prey items, in order of abundance, were ants, beetles and spiders. Together these three groups comprised 75.1% by number of all prey in faecal samples. Examination of faecal sacs from nestlings showed that young were being fed a range of invertebrates including spiders, grasshoppers, crickets, cockroaches, earthworms, centipedes, millipedes, various insect larvae and, occasionally, small frogs and lizards. Spiders and orthopterans made up 62% of the individual food items fed to nestlings (Smith and Calver 1984).

Further studies of scrub-bird diet aimed at improving the ability to select suitable translocation sites were undertaken in 1992 by an Honours student from Murdoch University. This project examined the species composition and vegetation structure of scrub-bird habitat and assessed potential food resources in nine occupied territories at Two Peoples Bay. All territories were characterised by a dense lower stratum of shrubs and sedges and had abundant leaf litter. Invertebrates from 17 groups were collected from leaf litter and by pit-trapping. The most abundant were beetles, spiders, woodlice, sand hoppers, flies, ants and springtails. The study also examined aspects of feeding behaviour in captivity and determined the nutritional composition of the invertebrates most commonly found in scrub-bird droppings (Welbon 1993).

Male Noisy Scrub-birds are territorial, defending their territories with a loud, directional song that can be heard throughout the year, but which is more frequent during the breeding season of May to October. Between 1973 and 1976, the mean male territory size at Mt Gardner was 6 ha (range 4 to 9 ha), and within this area 80% of the calls were made in a core area that averaged 1.25 ha (0.75 to 2.25 ha) (Smith 1985b). Smith (1985b) also distinguished two types of territory: long-term territories in which the males sing throughout the year and nesting probably occurs every year, and short-term territories where the male sings infrequently during the breeding season and breeding may take place only in some years. With increasing numbers of singing males on Mt Gardner this distinction has become less clear-cut. The majority of males on Mt Gardner currently appear to be occupying sites which would have once been considered of low breeding potential.

Removal of birds for translocation has shown that, in most cases, the territorial male is replaced by another male who then proceeds to sing and defend that territory. In some cases replacement may be rapid, and in one territory three singing males were sequentially removed, and replaced by new males over a period of four days. This and the replacement, after removal of a series of territorial males, by a very young male strongly suggests that the replacement males are not immigrants from other areas but are permanently present in or near the territories of the singing males (Danks and Smith, in prep.). The presence of non-territorial males may reflect the quality of the habitat and the length of time over which breeding has been occurring.

The female builds her nest on the outskirts of, or in between male territories. Territorial males seldom visit these areas. In successive years, the female may build within metres of previous nests but often moves to another site to re-nest and re-lay after disturbance or predation of the nest. A nesting area may be used by the female over many years. Females replacing translocated females will often nest in the same area as the original female, suggesting that nest areas are recognised by other females and that nest site selection is not idiosyncratic. On Mt Gardner several nesting areas are known which have been in use since 1974, and perhaps as early as 1972.

The period from the start of nest building to the last chick leaving the nest may extend from May to November, but more commonly lasts from June to October. Egg laying may begin as early as May but the peak of laying usually occurs in late June. Some eggs, possibly replacement clutches, may be laid late in September or early October. The nest is globular, approximately 18 cm in diameter, with a small side entrance. It is usually sited about 20 cm above the ground in a dense clump of sedge or, less commonly, in a dense shrub or pile of debris. The long leaves of pliable sedges such as Anarthria scabra and Lepidosperma spp. are used for the construction of the core with the leaves of Agonis flexuosa, Eucalyptus spp. and Dryandra formosa, twigs and strips of paperbark often being used for the outer layers of the nest. The lower half of the nest interior is lined with a papier mache´-like substance made from decayed wood.

Nest building takes two or three weeks and the single egg is laid one to two weeks after its completion. Nest building, incubation and feeding of the nestling are all carried out by the female. Incubation lasts 36 to 38 days (which is unusually long for a passerine of comparable size) and the chick fledges three to four weeks after hatching. Most young hatch in late July or early August, leaving the nest in late August or early September when invertebrate numbers could be expected to be increasing. Chicks have been observed in the company of an adult female up to six months after fledging.

Scrub-birds moult from their juvenile (pre-basic) plumage into a first basic plumage shortly after fledging. They carry this plumage until their second year when they slowly moult again into full adult plumage between October and March (Smith 1985c). The annual moult of adults apparently also occurs during the summer but in some years scrub-birds showing incomplete moult have been captured in June.

In captivity a 22 to 25 month old male did not successfully fertilise females, although he was in adult plumage and singing territorial song (Smith et al. 1983). However, the presence of other captive scrub-birds and the resulting aggressive encounters between the birds, may have interfered with copulation. Successful fertilisation was achieved in the male's third year when aviary management had separated the birds and controlled the access of the male to the females. The captive females built nests and laid eggs in their first year and females in the wild are known to breed in their first year (Smith 1978).

Generally only one active nest is found in a nesting area (Smith 1985b) although, on one occasion during nest searching for the translocation program, two active nests were found which were only 20 m apart. Trapping in two adjacent territories during breeding seasons between 1983 and 1991 resulted in the capture of six non-breeding females and four breeding females (indicated by active brood patches), with only one apparently breeding female being captured during any one trapping season. The male held in captivity, however, mated with three females in one breeding season and some observations suggest that males will mate with more than one female in the wild (Smith 1985a). Males may therefore be opportunistically polygamous.

Of 53 active nests examined during the translocation program one, and possibly one other, were considered to have been predated. Smith and Robinson (1976) report the predation of scrub-bird eggs by the Mardo Antechinus flavipes (a small dasyurid marsupial). Snakes such as Morelia spilota, Notechis scutatus and Pseudonaja affinis, together with the goanna Varanus rosenbergi, which are all common in scrub-bird habitat, could conceivably prey on scrub-birds, especially nestlings and chicks. However, the timing of the breeding season, which generally coincides with the period of reptilian inactivity, would minimise nest predation by these species. The fox might also be expected to prey on scrub-birds. However, the density of vegetation normally inhabited by scrub-birds probably provides sufficient protection against these introduced predators. Prior to a control program begun in 1988, foxes were often seen in Two Peoples Bay Nature Reserve. The steady increase in scrub-bird numbers on Mt Gardner during the 1970s and 1980s, before fox control was carried out at the Reserve (see Section 2.4.2) suggests that predation by foxes, or any other predator, was not seriously affecting the population.

1.5 Effective population size

The relationship between the census size (ie, the number of singing males) of a population of the Noisy Scrub-bird, the effective population size (ie, the number of breeding adults, Ne) and total population size (N) is not known. Attempts to define these relationships in the field would be very difficult because:

  • 1. The species is cryptic and lives in dense habitat, making the counting of individuals by observation extremely difficult.
  • 2. Determining the number of breeding birds would require finding active nests. However, nests are difficult to find so failure to find a nest in any given territory does not necessarily indicate lack of breeding.
  • 3. Within each habitat type the carrying capacity in terms of the number of breeding females and/or sub-dominant males may vary. Hence, estimating carrying capacity, even by sampling of habitat types, would be difficult.
  • 4. Data on breeding potential can not be transferred from one area to another because the number of reproductively active birds may vary with population density; for example, the first subdominant male at Two Peoples Bay was recorded in 1975, but by 1983 all territories where males were captured for translocation appeared to have one. A similar situation probably applies to females.

Thus, when considering the population objectives for this plan, the only reliable population index remains the number of singing males. Estimates of Ne and N require assumptions, the accuracy of which are unknown. Despite these limitations, it is worthwhile to attempt to estimate Ne so as to provide a rough guide to the status of the population.

Smith (in press) has calculated a range of values for Ne using the formula of Ewens et al. (1987) and a range of assumptions about the number of breeding females in a territory in relation to the habitat type. These assumptions are based on limited data and are at best crude estimates. Smith (op. cit.) considers that the following two assumptions (his assumptions 2 and 3) probably span the real productive potential. The first assumption is that all territories except those in heath have one breeding pair and 50% of territories in low forest have one extra breeding pair. The second is that all territories in low forest and tall thicket have one breeding pair, 50% of territories in low thicket have one breeding pair and 25% of territories in low forest have one extra breeding female.

The values of Ne calculated from the population and habitat data at Two Peoples Bay from 1987 and 1988 are provided in Table 2. Based on these assumptions, Smith (op. cit.) suggested that to reach a Ne of 500 (see 1.6 below) would require between 208 and 311 territories, if the proportion of habitat types remained the same. However, at Mt Gardner, most new territories are now in thicket and hence a larger number of territories would be required. There is insufficient habitat data from other population concentrations to allow an estimate of Ne for them.

Table 2. Effective population size, Ne, using Smith's (in press a) assumptions 2 and 3

(for the Two Peoples Bay population)

  1987  1988 
No of singing males 

193

179 
Ne, Assumption 2  432  395 
Ne, Assumption 3  382  345 

As stated above, the relationship between the number of singing males and total population (N) is also unknown, but we suggest a rule of thumb of N = 2.5 times the number of singing males. At the present time (1994) N is estimated to be 1185.

1.6 Reasons for threatened status

The disappearance of the Noisy Scrub-bird from most of its former range is thought to have been due primarily to a major change in fire regimes following the breakdown of Aboriginal society by the 1880s and the subsequent changes wrought on the environment by Europeans. Formerly, Aboriginal people deliberately burnt certain areas for hunting, ease of access and other purposes, which maintained generally low fuels but prevented any one area being burnt too frequently. Under this regime there must have been considerable areas of dense, long-unburnt vegetation around swamps to allow for the former relative abundance of scrub-birds. The fire regime altered when Aborigines left the land, allowing fuels to build up over large areas, and fires that did occur were probably hot and extensive. Further, as cattle grazing became established in coastal heaths and swamps, the practice developed of burning this country every two or three years to provide new growth suitable for feed. The clearing and draining of swamps for agriculture would also have contributed significantly to the loss of habitat (Smith 1977, 1985d).

The small population which remained on Mount Gardner was protected from the effects of frequent or extensive fires for several reasons. The peninsula on which the mountain is situated juts south-easterly into the Southern Ocean and is unlikely to be burnt by large scale wildfires running before the predominantly easterly winds of summer. Furthermore, the mountain has extensive areas of exposed granite which act as natural fire breaks and a number of deep, wet gullies. In this landscape fires would probably have been confined to relatively small areas and a single fire would have been unlikely to affect all of the gullies which provided refuge for the Noisy Scrub-bird. This fortuitous combination of geographic features is not repeated anywhere else in the scrub-bird's former range.

There has been some disagreement about the status of the Noisy Scrub-bird. In 1990, the Royal Australasian Ornithologists Union (RAOU) listed it as Rare (Brouwer and Garnett 1990), even though the compilers of the text for this species (two of the authors of this Recovery Plan) recommended it be listed as Endangered. In 1991 the Australian and New Zealand Environment and Conservation Council (ANZECC) listed it as Vulnerable (ANPWS 1991). However, more recently, the Action Plan for Australian Birds prepared by RAOU for the Australian National Parks and Wildlife Service (Garnett 1992b) lists it as Endangered, and ranks it equal ninth in priority for conservation action among all Australian bird taxa (including many subspecies). This listing is repeated in Garnett 1992a. These differences have occurred because of the subjective nature of allocating taxa to the IUCN categories of Endangered, Vulnerable, Rare and Indeterminate used in those times.

In 1991, a re-evaluation of the IUCN Threatened Species categories for vertebrates recommended that three categories be employed: Critical, Endangered and Vulnerable, and that categorisation be based on the probability of extinction over time (Mace and Lande 1991). Under this draft system a species meeting two of the following: a total effective population (Ne) of less than 500, having a fragmented population with five or less subpopulations, each with Ne less than 100 and immigration rates of less than one per generation, or subject to periodic catastrophes is categorised as Endangered.

In 1994, the total number of singing males in the whole Albany Management Zone was 474 and Ne for the total population probably now exceeds 500 (see Section 1.5). However, the species still clearly meets the Mace and Lande (op. cit.) criteria for classification as 'Endangered' because the population is fragmented with no known gene flow between populations and it is subject to catastrophic crashes (due to wildfire). Also the total population is well below the figure of 2 500 given by Mace and Lande (1991) as the upper limit for endangered taxa.

In November 1994, IUCN adopted the new Red List Categories of Critically Endangered, Endangered and Vulnerable together with criteria for evaluating taxa (IUCN 1994). During 1995, a Scientific Ranking Panel set up by CALM evaluated all taxa that were listed as threatened using the IUCN criteria as a guide. This Panel recommended that the Noisy Scrub-bird be categorised as Endangered and this recommendation was endorsed by the Minister for the Environment.