National recovery plan for the Golden-shouldered Parrot (Psephotus chrysopterygius) 2003-2007

Recovery Plan
Dr Stephen Garnett and Dr Gabriel Crowley
Queensland Parks and Wildlife Service, Brisbane

1. General Information (continued)

Social and economic impact

This recovery plan aims to contribute positively to communities within the distribution of the golden-shouldered parrots by (1) providing employment to residents and traditional owners, (2) improving ecological sustainability of pastoral activities through the improvement of fire management, vegetation structure and perennial grass composition. The implementation of this recovery plan is unlikely to cause significant adverse social and economic impacts. Unforeseen adverse effects should be avoided through comprehensive consultation processes described above.

Distribution

The golden-shouldered parrot occurs in the headwaters of the Morehead River and adjacent westward flowing streams (Morehead population) and the upper tributaries of the Staaten River (Staaten population). The range of the Morehead population is currently about 1380 km2. Until 1998, it was still contracting, at least along its eastern boundary for which detailed distributional data is available. The Staaten population is currently thought to be contained in an area of about 300 km2 west of the Lynd River in the headwaters of Cockburn, Back and White Horse creeks. Recent reports of parrots in other parts of the species' former range are unconfirmed. Where searches have been made, additional populations were not found. If other populations do exist, they are likely to be threatened by the same processes that are affecting the known populations.

Distribution of golden-shouldered parrot

Habitat critical for survival

Critical and important habitats of the golden-shouldered parrot are described in Table 4. Although the parrots occupy a range of habitats, only a subset of these is thought to be irreplaceable within its life history. In the wet season the parrots appear to require the gravelly slopes of quartzite gravel that occur in association with metamorphic rocks and granites. These areas are refuges early in the wet season when most fallen seed has germinated and no storm-burnt seed is available on the flatter country. Seed on the gravels appears to germinate less readily because it is less vulnerable to early saturation. The other critical habitat is that used by the parrots for breeding. For nesting, the parrots require termite mounds, particularly those of Amitermes scopulus in the Morehead population and A. vitiosus in the Staaten population. They primarily occur along grassy drainage flats fringed by woodland, although they are also present on the gravel slopes. Where these habitats have an open structure, nesting appears to have a higher success rate than where the grass has been invaded by woodland. In the dry season, habitat choice appears to be based on the availability of annual grass seed. Seed is abundant in many different habitat types, so none can be considered critical.

The vegetation of the Morehead distribution has been mapped at 1:100,000 (Neldner and Clarkson 1999; Neldner 1999). Critical habitat occurs in a range of recognized vegetation units (Table 4). However, the habitat of the parrots is a complex and fine mosaic of many vegetation types. Many of these vegetation units have been inadequately mapped, partly because habitat units are of a size that is too small to map, and partly because much of the nesting habitat is difficult to access. Vegetation maps of the area are made up of polygons that represent vegetation mosaics, with up to four units listed at each location. Using the primary vegetation element shown on the maps, it is possible to predict nesting habitat occurring at only 14.8% of 741 known nests. When all vegetation elements of the polygons are used, vegetation maps can be used to predict the occurrence of nesting habitat at 66% of nest sites. Nest sites that were poorly predicted by mapping were predominantly in hills, with15.7% (116 nests) of all known nests being mapped as occurring in Eucalyptus cullenii ± Corymbia clarksoniana woodland, even though no nests occurred in this habitat. Vegetation mapping in the hills is in particular need of refinement, with at least two vegetation mapping units (153 and 159) not being recognized as occurring there. A further 12.3% (91) of nests were mapped as occurring within Eucalyptus tetrodonta ± Corymbia clarksoniana ± Erythrophleum chlorostachys woodland, even though no nests were found in this habitat. Revised mapping of critical habitat in the core study area is under way, with the aim of identifying critical habitat. Critical wet season feeding habitat occurs at an even finer scale than nesting habitat. Much of the critical habitat also occurs in poorly mapped hills. Revised mapping will assist in identifying this habitat, but many sites are too small to be identified at 1:100,000.

The vegetation of the Staaten distribution has been mapped at 1:1,000,000 (Morgan 1999). Critical habitat also occurs in a range of recognized vegetation units (Table 4). However, because of the scale of current mapping, no attempt has been made to reconcile nest locations with vegetation mapping units. Hence no attempt should be made to identify extent of critical habitat from the vegetation maps that are currently available.

Table 4 Critical (shaded) and important (not shaded) Golden-shouldered parrot habitat
Habitat
Wet season feeding
Breeding
Breeding season feeding
Dry season feeding
Roosting
VMU1
RE2
Biodiversity status
Structural type Dominant species Critical/ important habitat characteristics
Artemis
Mary Valley
Dixie
Kalinga
Killarney
Imooya
Staaten R. NP
Bulimba
Gravel slopes *** ***   *   153 (3.3.47)   Low open-woodland Melaleuca citrolens± Melaleuca foliolosa± Melaleuca viridiflora ±Melaleuca acacioides Termite mounds: Amitermes scopulus Gravel and rocks Alloteropsis semialata, Schizachyrium spp., Eriachne obtusa, Fimbristylis spp. Scleria spp. Hybanthus enneaspermus, Aneilema siliculosum, Heliotropium sp + + + + + 0 0 0
*** ***   * * 159 (3.3.50)   Low open-woodlands Melaleuca viridiflora± Petalostigma pubescens ± Corymbia clarksoniana + + + + + 0 0 0
Planichloa flats ***     *   153 3.3.47   Low open-woodland Melaleuca citrolens± Melaleuca foliolosa± Melaleuca viridiflora ±Melaleuca acacioides Gaps in woody vegetation dominated by Schizachyrium spp±Planichloa nervilemma ± legumes. + + + + + 0 0 0
Broad flats * *** *     158 2.3.28/ 3.3.49   Low open-woodlands Melaleuca viridiflora± Petalostigma banksii Termite mounds: Amitermes scopulus± Amitermes laurensis Alloteropsis semialata, Ischaemum fragile, Storm-burnt Ischaemum fragile Digitaria spp.,±Paspalum scrobiculatum, ± Desmodium sp., ±Echinochloa colona ±Ischaemum decumbens 1 1 1 0 0 0 1 1
* *** *     184 3.3.59 Of concern Closed tussock grassland Sorghum plumosum± Themeda arguens + 0 0 0 0 0 0 0
  ***       180 3.3.56   Closed tussock grassland Eriachne spp. ± Aristida spp.±Eragrostis spp. ± Fimbristylis spp. Termite mounds: Amitermes scopulus± Amitermes laurensis 0 0 0 1 0 0 0 0
Closed flats * ***   *   144 2.3.28/ 3.5.14   Low woodlands Melaleuca viridiflora± Acacia spp.± Asteromyrtus symphyocarpa Termite mounds: Amitermes scopulus± Amitermes laurensis 0 0 0 0 1 0 1 1
+ ***   *   153 3.3.47   Low open-woodland Melaleuca citrolens± Melaleuca foliolosa± Melaleuca viridiflora ± Melaleuca acacioides 0 0 1 1 1 0 0 0
* ***   * * 159 2.3.28/ 3.3.42/ 3.3.50   Low open-woodlands Melaleuca viridiflora± Petalostigma pubescens ± Corymbia clarksoniana 1 1 1 1 1 1 1 1
* ***   *   156 3.5.17 Of concern Low open-woodland Melaleuca stenostachya± Melaleuca viridiflora 1 0 1 0 0 0 0 0
Flat edges   ***   * * 57 3.3.15   Woodland Eucalyptus brassiana Termite mounds: Amitermes scopulus± Amitermes laurensis± Schizachyrium spp.± Thaumastochloa spp. + 0 + 0 0 0 0 0
  ***   *   110 3.3.33   Woodland Thryptomene oligandra, Melaleuca viridiflora± Acacia torulosa ± Grevillea pteridifolia ± Neofabricia mjoebergii 1 1 1 1 1 1 0 0
Rocky hills +         69 3.11.7   Woodland Eucalyptus cullenii± Corymbia clarksoniana± Eucalyptus chlorophylla± Corymbia confertiflora Boulders or storm burnt areas with seed of Hyptis suaveolens ± Aneilema siliculosum 0 1 1 1 1 1 0 0
+         70 3.11.8/ 3.12.10   Woodland Eucalyptus cullenii± Corymbia clarksoniana 1 1 1 1 1 0 0 0
Box flats     + *   113 2.3.10/ 3.9.2/ 3.3.36   Open-woodland Eucalyptus chlorophylla Sorghum plumosum, Heteropogon triticeus± Heteropogon contortus 1 1 1 1 1 1 1 1
Sand ridges/ hills +   + * * 63 2.5.5/ 3.3.20   Woodland Corymbia clarksoniana± Erythrophleum chlorostachys± Melaleuca viridiflora± Eucalyptus leptophleba ± Eucalyptus tetrodonta± Eucalyptus chartaboma Eriachne obtusa, Alloteropsis semialata, Sorghum plumosum, Heteropogon triticeus, Eragrostis spp., Panicum spp., Schizachyrium spp., Thaumastochloa spp. ±Digitaria spp. ±Brachyachne convergens ±Brachiaria holosericea ±Phyllanthus spp. 1 0 0 0 0 0 1 1
+   + * * 92 2.5.5/ 3.5.7   Woodland Eucalyptus tetrodonta ± Corymbia clarksoniana± Erythrophleum chlorostachys ( ± Parinari nonda) 1 1 1 1 1 1 1 1
+   + *   98 3.5.9   Woodland Eucalyptus tetrodonta ± Corymbia hylandii subsp. peninsularis ± Erythrophleum chlorostachys± Corymbia setosa subsp. pedicillaris 0 0 1 0 0 1 0 0
+   + * * 101 3.5.10   Woodland Eucalyptus tetrodonta± Corymbia nesophila (± Parinari nonda) 1 0 0 0 0 0 0 0
+   + * * 104 3.5.12   Woodland Eucalyptus tetrodonta± Corymbia nesophila± Corymbia clarksoniana ( ± Parinari nonda) 1 1 1 1 1 0 0 0
Bare areas       *   198 3.2.6   Stream beds     + + + + + + + +
*     *   - -   Roads and bare areas     + + + + + + + +
Riparian forest         * 48 3.3.10   Open forest Melaleuca argentea± Melaleuca fluviatilis± Melaleuca leucadendra   1 1 1 1 1 1 0 0
Swamp edge         * 53 2.2.28/ 3.3.14   Woodland Melaleuca saligna± Melaleuca viridiflora± Asteromyrtus symphyocarpa± Melaleuca spp.   + + + + + + 1 1

1Vegetation Mapping Unit (Neldner and Clarkson 1999). 2 Regional Ecosystem (Neldner 1999; Morgan 1999).

Critical habitat is found throughout the parrot's distribution, and on all of the eight properties on which the parrots are currently found (Table 4). Landholders and traditional owners of each of these properties have been consulted in developing the current recovery plan and each of the two previous recovery plans (Table 5). As most work so far has been undertaken on the Morehead population, consultation with landholders and traditional owners associated with this population's distribution has been most comprehensive. Changes to ownership of some properties have meant that it has been necessary to establish contact with new owners from time to time. Similarly, the nature of consultation with traditional owners, either directly or through representative bodies, is evolving as aspirations of traditional owners groups clarify.

Threats and appropriate management relating to each habitat category are described in detail in the Threats and Management practices sections.

Important populations

Breeding is known from two areas. The Morehead population is known to occur over an area of approximately 1,380 km2 from the Peninsula Development Road at Windmill Creek in the north-east to Dixie Station in the north-west south-west along Eight Mile Creek west of Imooya Station and south-east as far as Tobacco Yard on Kalinga Station. The north-east boundary of this population is known to have retreated by 40 km2 since 1992 when nesting occurred at South Five Mile Creek and Kalkah on Artemis Station. Previously this population was known to have extended at least as far north-east as Violetvale Station, where it was last recorded in 1974, and possibly Marina Plains. Breeding also occurred on the eastern block of Kalinga and probably on Koolburra Station in the 1970s and may also have occurred on Olive Vale and Kalpowar in the 1960s and early 1970s. Repeated surveys on Killarney station have failed to find nests south of Six Mile creek which represents the southern boundary of the population. Limits to and trends in the western distribution of this population are poorly known. The population is at its lowest during the breeding season. During the dry season flocks of immature birds inflate the apparent abundance. Many of these appear to die during the wet season. Population size is therefore estimated on the basis of the number of birds alive during the breeding season. Of about 620 km2 surveyed in detail, 250 km2 had an estimated density of one pair/ km2 with 10% having subsidiary immature males associated with them. The remaining areas had a density of about one pair per 5 km2. Assuming the same ratio of high density to low density pertains over the entire range of the Morehead population (approximately 1,380 km2), the population size during the breeding season before any birds have fledged is 1,070 individuals of which 500 are females. The contraction of breeding distribution is likely to have been associated with the loss of about 40 breeding birds.

Table 5. Distribution of critical golden-shouldered parrot habitat and consultation processes.
N/A not applicable; *** critical habitat; * important habitat; + occasional habitat.
Property Habitat VMU RE Biodiversity status
Wet season feeding
Breeding
Landholders Traditional owners Notes
Consulted
Recovery team representation
Conservation agreement
Property planning
Group(s)
Consulted
Recovery team representation
Morehead population
Artemis Flat edges 57 3.3.15     *** Yes Yes Yes Yes Kuku Thaypan, Ukele Yes Yes Leaseholder involvement since 1992. Traditional owners consulted directly since 1992 and through Balkanu since 1999.
Flat edges 110 3.3.33     ***
Gravel slopes 153 (3.3.47)   *** ***
Planichloa flats 153 3.3.47   ***  
Closed flats 156 3.5.17 Of concern * ***
Broad flats 158 3.3.49   * ***
Gravel slopes 159 (3.3.50)   *** ***
Closed flats 159 3.3.42/ 3.3.50   * ***
Broad flats 184 3.3.59 Of concern * ***
Mary Valley Flat edges 110 3.3.33     *** Yes - - - Kuku Thaypan Yes Yes New leaseholders, invited to recovery team membership October 2002. Traditional owners consulted directly since 1992 and through Balkanu since 1999.
Gravel slopes 153 (3.3.47)   *** ***
Planichloa flats 153 3.3.47   ***  
Broad flats 158 3.3.49   * ***
Gravel slopes 159 (3.3.50)   *** ***
Closed flats 159 3.3.42/ 3.3.50   * ***
Dixie Flat edges 57 3.3.15     *** Yes Yes - - Ukele Yes Yes Involvement since 1992. Current managers representing leaseholders since 1999. Traditional owners consulted through Balkanu since 1999.
Flat edges 110 3.3.33     ***
Gravel slopes 153 (3.3.47)   *** ***
Planichloa flats 153 3.3.47   ***  
Closed flats 153 3.3.47   + ***
Closed flats 156 3.5.17 Of concern * ***
Broad flats 158 3.3.49   * ***
Gravel slopes 159 (3.3.50)   *** ***
Closed flats 159 3.3.42/ 3.3.50   * ***
Imooya Flat edges 110 3.3.33     *** No No No No Ukele Yes Yes Leaseholder consulted 1992-1999. Current absentee leaseholder has not been located. Traditional owners consulted through Balkanu since 1999.
Closed flats 159 3.3.42/ 3.3.50   * ***
Kalinga Flat edges 110 3.3.33     *** Yes Yes - Yes Kuku Thaypan Yes Yes Leaseholders involvement since 1995. Traditional owners consulted directly since 1992 and through Balkanu since 1999.
Gravel slopes 153 (3.3.47)   *** ***
Planichloa flats 153 3.3.47   ***  
Closed flats 153 3.3.47   + ***
Gravel slopes 159 (3.3.50)   *** ***
Closed flats 159 3.3.42/ 3.3.50   * ***
Broad flats 180 3.3.56   * ***
Killarney Flat edges 110 3.3.33     *** Yes Yes - - Kuku Thaypan, Ukele Yes Yes Involvement since 1993, present leaseholders since 1999. Traditional owners consulted directly since 1992 and through Balkanu since 1999.
Closed flats 144 3.3.28/ 3.5.14     ***
Gravel slopes 153 (3.3.47)   *** ***
Planichloa flats 153 3.3.47   ***  
Closed flats 153 3.3.47   + ***
Gravel slopes 159 (3.3.50)   *** ***
Closed flats 159 3.3.42/ 3.3.50   * ***
Staaten population
Bulimba Closed flats 144 2.3.28     *** Yes Yes - - Uwoykand Yes Yes Current landholders and traditional owners consulted directly since 1998.
Broad flats 158 2.3.28   * ***
Closed flats 159 2.3.28   * ***
Staaten River NP Closed flats 144 2.3.28     *** Yes Yes NA NA Uwoykand Yes Yes Traditional owners consulted directly since 1998.
Closed flats 159 2.3.28   * ***
Broad flats 158 2.3.28   * ***

The range of the Staaten population is less well known. It occurs west of the Lynd River on Bulimba Station and on Staaten River National Park, a particularly remote site with no vehicular access. Surveys in 1997 and 2002 have defined the eastern and northern boundaries with the surveys subscribing an area of approximately 300 km2. The western and southern limits to the population, however, are unknown and may extend beyond the surveyed area. However they are known to have bred on Bulleringa Station to the southeast of the currently known population in the 1960s, but were not be found there in searches during the 1990s. From the number of nests encountered in short surveys and the number of birds seen, including a single flock of over 100 individuals, the density of Staaten population is probably at least twice that of the Morehead population. The population is therefore estimated to be 1,300 individuals, of which about 600 are females.

The total breeding population is estimated to total approximately 2,300 individuals in two populations of approximately equal size. In the non-breeding season, assuming an average productivity of 1.2 fledglings/nest attempt, the population is thought to swell to about 3,600.

While there have been reports of birds away from these areas in the last decade, no nesting has been confirmed. Those reports it has been possible to investigate have proved erroneous. Previously a third breeding population was known north-east of Coen on Mt Croll Station and Mungkan Kandju National Park and to the south-east of Coen on Silver Plains Station. Breeding was last definitively known north of Coen in 1922 (White 1922) and on Silver Plains in 1927 but individuals were seen on Silver Plains in the 1950s and a young bird was caught north of Coen in 1970.

Threats

Predation appears to be the immediate cause of decline, possibly in combination with a shortage of food in the early wet season. Both threats have come about as a result of a change in fire regime, particularly in combination with cattle grazing. One result of the change in fire regime has been an increase in the density of woody plants, notably broad-leaved ti-tree Melaleuca viridiflora. This appears to have increased the vulnerability of birds to predation during the wet season and while nesting. The second effect has been the development of a coarser mosaic of burning histories, which is thought to decrease the chances of dispersing parrots finding suitable habitat in the wet season. Cattle affect the parrot by reducing seed production by wet season grasses and reducing the fuel load, particularly in the habitat used by the parrots when breeding.

Disappearance of golden-shouldered parrots from Lakefield National Parks in the 1970s to 1980s has been attributed to loss of grasslands and thickening of grassy woodlands associated with inappropriate fire regimes (Garnett and Crowley 1999). Similar loss of habitat as a result of inappropriate fire regimes has also occurred on Mungkan Kandju National Park (Neldner et al. 1997), where birds were lasted recorded nesting in the 1920s.

Table 6.Threats summary. ***, significant current or probable potential threat; * possible potential threat
Location name Coordinates Population size Land tenure Type of threat Status of threat
Wet season feeding
Breeding
Breeding season feeding
Dry season feeding
Roosting Current actions to reduce threats Future actions to reduce threats
A Morehead population 15°15' S 143°30' E 1,500 Pastoral leases Increased success of predators as a result of vegetation thickening Current/ primary   *** ***    
  • Fencing to allow stock control
  • Firebreak network of early dry season fires
  • Storm-burning2 at 2-4 year intervals
  • Monitoring effects of pigs and cattle
  • Fencing to allow stock control
  • Firebreak network of early dry season fires
  • Storm-burning at 2-4 year intervals
  • Monitoring effects of pigs and cattle
  • Conservation agreements (property planning)
Loss of perennial grasses Current/ primary ***   ***    
  • Fencing to allow stock control
  • Monitoring effects of pigs and cattle
  • Control of pig numbers (fencing, poisoning and hunting)
  • Conservation agreements (property planning)
Loss of termite mounds Potential/primary   ***      
  • Monitoring effects of pigs and cattle
  • Monitoring effects of pigs and cattle
  • Actions that will derive from the results of monitoring
Inappropriate fire regime (vegetation thickening; loss of perennial grasses; loss of termite mounds) Current/ secondary *** *** ***    
  • Fencing to allow stock control
  • Firebreak network of early dry season fires
  • Storm-burning at 2-4 year intervals
  • Conservation agreements (property planning)
Overgrazing (vegetation thickening; loss of perennial grasses) Current/ secondary ***   ***    
  • Fencing to allow stock control
  • Destocking of National Parks
  • Monitoring effects of cattle
  • Conservation agreements (property planning)
Pig rooting (vegetation thickening; loss of perennial grasses; loss of termite mounds) Current/ secondary *** *** ***    
  • Monitoring effects of pigs
  • Control of pig numbers (fencing, poisoning and hunting)
  • Conservation agreements (property planning)
Clearance (habitat loss) Potential/ secondary * * * * *
  • Conservation agreement (one area)
  • Conservation agreements (property planning)
Disease Potential/ primary Effect not habitat specific
  • None
  • Action will be taken only if there is evidence that disease is affecting the population. Such actions may include quarantine and limitations to keeping birds in the vicinity
Trapping and trade Potential/ Historical/ primary Effect not habitat specific
  • Restrictions on export (CITES App. I listing)
  • Prohibition of wild take (Nature Conservation Act 1992 Section 88 (1)) except under license
  • Continue existing measures
Predation by feral cats Potential/ primary Effect not habitat specific
  • Feral cats shot when encountered, which is rarely
  • Continue existing actions unless feral cat predation shown to be of greater significance than currently believed
A Staaten Population 16°50' S 143°20' E 500 Pastoral lease/ National Park Increased success of predators as a result of vegetation thickening Current/ primary *** >*** ***    
  • Firebreak network of early dry season fires
  • Storm-burning at 2-4 year intervals
  • Continue existing measures
Loss of perennial grasses Current/ primary ***   ***    
  • Firebreak network of early dry season fires
  • Storm-burning at 2-4 year intervals
  • Continue existing measures
Loss of termite mounds Potential/ primary   ***      
  • Firebreak network of early dry season fires
  • Storm-burning at 2-4 year intervals
  • Continue existing measures
Inappropriate fire regime (vegetation thickening; loss of perennial grasses; loss of termite mounds) Current/ secondary *** *** ***    
  • Firebreak network of early dry season fires
  • Storm-burning at 2-4 year intervals
  • Continue existing measures
Overgrazing (vegetation thickening; loss of perennial grasses) Current/ secondary ***   ***    
  • Monitoring trends in parrot numbers
  • Continue existing measures and take further action if decline evident
Pig rooting (vegetation thickening; loss of perennial grasses; loss of termite mounds) Current/ secondary *** *** ***    
  • Monitoring trends in parrot numbers
  • Continue existing measures and take further action if decline evident
Clearance (habitat loss) Potential/ secondary * * * * *
  • Half the area is National Park
 
Disease Potential/ primary Effect not habitat specific
  • None
 
Trapping and trade Potential/ Historical/ primary Effect not habitat specific
  • Restrictions on export (CITES App. I listing)
  • Prohibition of wild take (Nature Conservation Act 1992 Section 88 (1)) except under license
  • Continue existing measures
Predation by feral cats Potential/ primary Effect not habitat specific
  • Monitoring trends in parrot numbers
  • Continue existing measures and take further action if decline evident

2 Burning within 5 days of the first heavy rains (≥ 50 mm over 72 hours) of the wet season.