Australian Field Ornithology 2017, 34, 12–17 http://dx.doi.org/10.20938/afo34012017

The range and habitat of the Kalkadoon Amytornis ballarae

Graham Harrington1,2*, Adam McKeown1 and Brian Venables3

1CSIRO Land & Water, PO Box 780, Atherton QLD 4883, Australia 2BirdLife Northern Queensland, PO Box 680, Malanda QLD 4885, Australia 3527 McCormack Street, Edge Hill QLD 4870, Australia. Email: [email protected] *Email: [email protected]

Abstract. In this paper we discuss the results of the first comprehensive survey of the range and habitat of the Amytornis ballarae. This grasswren is of particular interest because (1) It is not classed as a Restricted Range Species by BirdLife International although we show that it occupies <50 000 km2; (2) This species lives in a fire- prone environment, and the habitat is unsuitable for at least 3 years after fire. The movement and fate of the in these circumstances is unknown; (3) It is sympatric with the A. dorotheae over part of its range, raising questions about competition between the species; and (4) The spatial and behavioural separation from the closely related A. purnelli is not well defined. In 2008 and 2009, we visited 195 locations in firescars of known age within the known range of the Kalkadoon Grasswren and played calls of Kalkadoon and Carpentarian at 6–10 sites in each, totalling 1491 survey sites. We encountered 81 groups of Kalkadoon Grasswrens. They showed a marked preference for metamorphosed sandstone and marble. They were almost always found on rocky hills, but if on flat areas they were within 300 m of hills. Spinifex spp. dominated the ground-layer at all the sites where Kalkadoon Grasswrens were located. Ground-cover by spinifex increased annually after fire and plateaued after 6 years. Kalkadoon Grasswrens had returned to 30% of 3- and 4-year-old firescars, but occurred in 90% of locations surveyed which had not been burnt for ≥10 years. No grasswrens were encountered in surveys between the currently published ranges of the Dusky Grasswren in the Northern Territory and the Kalkadoon Grasswren in Queensland. This paper provides a distribution map, and defines the habitat variables for the Kalkadoon Grasswren. Our mapping shows that the range of the Kalkadoon Grasswren is ~40 000 km2, which qualifies it as a BirdLife International Restricted Range Species. We argue that an Important Area should be declared for conserving and monitoring this species alone, where it is not sympatric with the Carpentarian Grasswren.

Grasswren are recorded here to enable a comparison with the description in Higgins et al. (2001). Introduction the Kalkadoon Grasswren. Habitat parameters for the The most important influence on the population size Kalkadoon Grasswren are recorded here to enable a of Kalkadoon Grasswrens is probably fire. Burning A proposal to declare an Important Bird Area (IBA) for the comparison with the description in Higgins et al. (2001). temporarily reduces the area suitable for the birds and Kalkadoon Grasswren Amytornis ballarae was rejected The most important influence on the population size of probably causes their death. It is not known to what extent by Dutson et al. (2009). The reasons given were that the Kalkadoon Grasswrens is probably fire. The area suitable 2 a) birds die by perishing in the flames, b) subsequently population was stable and occupied an area >50 000 km , for the birds is temporarily reduced by fire and many starve due to decrease and/or c) are more susceptible which meant that this grasswren could not be described probably die. It is not known to what extent birds (a) die by to increased predation due to reduced protective cover. as a Restricted Range Species. This decision was based perishing in the flames, (b) subsequently starve because of Felderhof (2007) found Kalkadoon Grasswrens in a few upon a mapped range (Stattersfield et al. 1998) that is decreased food availablility and/or (c) are more susceptible 3-year-old firescars but none in younger habitat. This larger than the area attributed to the Kalkadoon Grasswren to increased predation because of reduced protective paper seeks to establish the probability of the Kalkadoon by Schodde & Mason (1999) and Higgins et al. (2001). Our cover. Felderhof (2007) found Kalkadoon Grasswrens in a Grasswren occurring in fire scars of different ages as the decision to survey this species was partially based upon few 3-year-old firescars but none in younger habitat. This spinifex regrows and to define its range and habitat. the concern that this species is vulnerable to fire and the paper seeks to establish the probability of the Kalkadoon area of its distribution was not well defined. Grasswren occurring in firescars of different ages as the spinifex regrows and to define the range and habitat of the Methods Another reason for undertaking this survey was to test the published description of the Kalkadoon Grasswren’s species. habitat—that it is confined to areas where the ground- The area of study was chosen according to the reported layer is dominated by the hummock grass spinifex Triodia range of this species, which equated approximately to the Methods spp.; is usually found on rocky hillsides, ridges and gullies; Mount Isa Inlier bioregion (Schodde 1982; Higgins et al. occasionally in less-rugged terrain and with short spinifex; 2001). As we were not certain how far north or south this The area of study was chosen according to the reported but is never found on flat plains or river valleys (Higgins et species extended, we surveyed beyond the recorded range of this species, which equated approximately to the al. 2001). range.in a northerly and southerly direction. To the north Mount Isa Inlier bioregion (Schodde 1982; Higgins et al. we surveyed up to and including Boodjamulla (Lawn Hill) We were also interested in the spatial separation from 2001). As we were not certain how far north or south this National Park (latitude 18°3′S) and to the south as far as the closely related Dusky Grasswren A. purnelli, which is species extended, we surveyed beyond the recorded range Mount Unbunmaroo (latitude 22°5′S) (Figure 1). The main not well defined. When the Kalkadoon Grasswren was first in both northerly and southerly directions: to the north, survey occurred in May 2008, with some supplementary recorded in the vicinity of Mount Isa as late as 1966, it was we surveyed up to and including Boodjamulla (Lawn Hill) surveys in May 2009. We made two visits to Mt. named as a subspecies of the Dusky Grasswren (Condon National Park (latitude 18°18′S) and to the south as far as Unbunmaroo in March and July 2010 (Figure 1). In August- 1969; Schodde 1982), but later was given species status Mount Unbunmaroo (latitude 22°30′S) (Figure 1). The main September 2014 we extended the southern part of the based on plumage characteristics (Schodde & Mason survey occurred in May 2008, with supplementary surveys 1999). Horton (2011) indicated that the behaviour and in May 2009. We made two visits to Mount Unbunmaroo ‘jizz’ of the Dusky Grasswren also differ from that of in March and July 2010. In August–September 2014, Kalkadoon Grasswren range and habitat 13

The 2008 survey sites were chosen to achieve approximately equal numbers in fire-age classes of 3–10 years. Only firescars which occurred within 1 km of a track navigable by a 4-wheel-drive vehicle were selected. The standard survey consisted of 10 points at intervals of 200 m arranged along three sides of a square 600 m × 600 m. The square was set at least 300 m from the edge of the firescar to increase the chance that the birds encountered in the survey reflected the conditions in that firescar and not the boundary conditions between firescars of different ages. In some locations, it was not possible to set the survey points around a square because of impassable topography or the shape of the firescar; in such cases, the surveyors used their judgment to place the survey points within the firescar at intervals of 200 m, whilst maintaining the protocol of all survey points being at least 300 m from the edge of the firescar. At each survey point, the surveyors stood back to back and one played a standard recording of first Kalkadoon and second Carpentarian Grasswren A. dorotheae calls using an Mp3 player through 3-watt speakers. The calls used were obtained from Plowright (1994). After completion of the play-back procedure, the surveyors watched and listened for response from grasswrens for a further minute. If there was no such response, they switched positions and repeated the exercise. The play-back procedure was not initiated until both surveyors indicated that they Figure 1. The Extent of Occupancy (EOO) of the Kalkadoon were ready to watch and listen for a response from the Grasswren. Locations where Kalkadoon Grasswrens birds because such responses can be very short-lived (KGW) have been recorded in the surveys in this study and could be missed by an inattentive surveyor. If the and from the BirdLife Atlas are shown as black dots. White surveyors obtained a response, either visually or vocally, dots indicate where surveys did not encounter Kalkadoon they attempted to identify the species of grasswren, count Grasswrens. Triangles indicate where surveys were the number of birds in the group, assign the sex of the undertaken in 2014 on spinifex-covered hills to the south of the EOO. The Queensland–Northern Territory border is birds, and note from which part of the landscape the birds along the 138°E longitude. appeared (ridge, upper slope, lower slope, drainage line, rocks). Because the probability of detecting a grasswren when it is present is certainly less than 1, the failure to detect grasswrens at any particular point does not mean we extended the southern part of the survey to include that they were absent. several ranges of hills as well as Mount Unbunmaroo At every survey point, the surveyors completed a (Figure 1). Following that, we surveyed hills covered in vegetation and topography proforma that classified the spinifex adjacent to the Plenty Highway from Boulia to elements of the site, including height of and cover by Dulcie Range National Park, which is the easternmost spinifex. On completion of this proforma (~15 minutes’ record for the Dusky Grasswren. The standard call-back work), they repeated the call-back procedure unless they procedure was used at locations at Dulcie Range National had already encountered grasswrens. Having completed Park, Huckitta and Old Huckitta Stations, Jervois Station, the survey procedure at one point, the surveyors moved hills adjacent to the airstrip near Lucy Springs, Goyders 200 m to the next point, using a GPS unit to measure Pillar on Tarlton Downs Station, Marqua Station, Urlampe distance and direction, and repeated the procedure. Station, Tobermorey, Cravens Peak, Marion Downs and However, if grasswrens had been encountered at the Glen Ormiston Stations. previous survey point, they moved 300 m before playing The Mount Isa Inlier is a bioregion of complex geology, the calls again in order to reduce the chance of the birds shallow soils, broken topography and low eucalypt following the surveyors and being recorded for a second woodland with a spinifex understorey. Similar habitat time. extends north-west through Riversleigh pastoral station into Boodjamulla National Park and the Northern Territory, We surveyed 145 sites and completed habitat proformas but the Kalkadoon Grasswren has not been found there at 1412 survey points in 2008 (at some sites we were (Higgins et al. 2001). Firescar maps were prepared from unable to fit the full number of 10 survey points into the satellite-derived data available on the Northern Australia space available). Fire Information (NAFI) website (www.firenorth.org.au). The methods were adapted in 2009. Along the routes Firescars were classified by age 1–10 years since they last ascribed to the surveyors they were requested to select burned; the rest of the landscape was presumed to have patches of habitat that appeared to be favourable to escaped fire for >10 years. NAFI covered only as far south grasswrens (based on the previous year’s results), without as 21°S; fire age for sites farther south was established by regard to the age of the firescar. They played the calls for asking landholders. three bursts of 10 seconds each followed by 20 seconds 14 Australian Field Ornithology G. Harrington et al.

Table 1. Geological substrates of Kalkadoon Grasswren habitat: The number of encounters with groups of Grasswrens on major geological substrates within the Extent of Occupancy (EOO, 39 750 km2) and the expected number of encounters based on the number of surveys in each substrate type (x2= 53.6, P <0.001).

Geological substrate Area of EOO with Survey sites with No. of Grasswren groups detected on this substrate % this substrate % this substrate Observed Expected Regolith (gravel) 30 11 3 9 Sandstone 15 15 12 12 Metamorphosed sandstone 12 12 15 10 Granite 10 9 2 7 Metamorphosed basalt 10 11 7 9 Rhyolite 8 22 15 18 Marble 5 10 15 8 Limestone 5 3 1 2 Argillaceous clay sediments 3 2 9 2 Other 3 6 2 4 Total 100 100 81 81

of silence because we suspected that there was often an Grasswrens and tested them against the observed immediate but short-lived response by the grasswrens, frequency using Chi-square analysis. which could not be heard whilst the recording was being played. We visited 74 survey sites and completed Results 431 survey points in 2009. In April and July 2010, we surveyed the isolated hill We recorded 81 groups of Kalkadoon Grasswrens, and Mount Unbunmaroo, ~100 km south of the nearest record the commonest group size was two (mean 1.9 ± standard of the Kalkadoon Grasswren and 400 km east of the deviation 0.62). Six groups had at least three birds and nearest record for A. purnelli. This was mentioned as a site two groups with four birds were recorded. These figures for A. purnelli by Schodde (1982) before A. ballarae was relate to the number of individuals heard and/or seen, and recognised as a separate species. Unfortunately, it is not probably underestimate the true numbers. known which species was recorded or by whom. The EOO for the Kalkadoon Grasswren, derived from In May 2014, four pairs of surveyors walked Mount encounters in our surveys and records from the BirdLife Unbunmarroo and other hills south and south-west of Australia Bird Atlas (extracted in 2012), has an area of 2 the known range of the Kalkadoon Grasswren and then ~39 750 km (Figure 1). The range of this species may worked their way from Boulia down the Plenty Highway be extended to Mount Unbunmaroo in the future because on our second visit in July 2010 we heard a grasswren as far as the easternmost record of the Dusky Grasswren, alarm call in response to playback of the Kalkadoon surveying any adjacent hills with spinifex ground-cover. Grasswren call, but we failed to see the bird. From the call, We played the Kalkadoon Grasswren call in 20-second we were unable to tell whether it was a Kalkadoon or a bursts at intervals of 100 m where and when it seemed Dusky Grasswren. Subsequent testing demonstrated that that a grasswren might be located. We had previously Dusky Grasswrens react strongly to this recording so this ascertained that the Dusky Grasswren reacted strongly to did not help in the identification of the bird. A party of four the Kalkadoon Grasswren call. pairs of experienced surveyors in 2014 failed to find any In order to assess the Extent of Occurrence (EOO), a trace of any grasswrens at this location, or in nearby hills minimum convex polygon (MCP) was constructed around (Figure 1). No grasswrens were encountered in 2014 on the 148 points where the Kalkadoon Grasswren had been spinifex-covered hills adjacent to the Plenty Highway. recorded in the BirdLife Australia Bird Atlas by 2012. In this Table 1 details the area of the major geological substrates context, the EOO is preferred to the Area of Occupancy within the Kalkadoon Grasswren’s EOO. The difference (AOO) statistic when the recorded locations are related between the observed frequency of occurrence of this to a habitat that is continuous so that the locations do species on different substrates and the expected frequency not have large areas of unsuitable habitat between them based on the number of surveys in each geological type (Gaston & Fuller 2009). was highly significant (P <0.001). Kalkadoon Grasswrens To test whether there is any preference for underlying were found substantially more often than expected on geological substrate by the Kalkadoon Grasswren, we metamorphosed sandstone, marble and argillaceous overlaid all the survey points within the MCP on the (clay) detrital sediment. However, the latter substrate does Cloncurry and Normanton 1:1 000 000 geology map not usually support spinifex, and these sites appear to be sheets (www.geoscience.gov.au). From the number of incorrectly mapped. Kalkadoon Grasswrens were found surveys conducted on each substrate type, we calculated at well below the expected frequency on outwash fans of the expected frequency of encounters with Kalkadoon sand and gravel (regolith), rhyolite, and granite. Kalkadoon Grasswren range and habitat 15

45 45

40 40

35 35

30 30

25 25

20 20

% of expected 15 15

10 10 Percentage cover by spinifex by cover Percentage Percentage cover by spinifex 5 5 Groups of Grasswrens encountered as Groups of Grasswrens encountered as Mean height of spinifex hummocks (cm) 0 0 0 1 2 3 4 5 6 7 8 9 10 height ofMean spinifex hummocks (cm) Years since last fire Age of firescar: Years since fire Years since last fire

Ht Cover

Figure 2. Detection of groups of Kalkadoon Grasswrens Figure 3. The mean height (cm) and cover (%) of spinifex in firescars of different ages. The number of groups in firescars 1–10 years old. Data for Years 1–2 are from encountered is expressed as a percentage of the number Felderhof (2007), and for Years 3–10 are from the present expected based on the number of surveys for each firescar study. age. See text for detail on age of firescars.

Figure 2 illustrates the number of groups of Kalkadoon Spinifex recovered rapidly after being burnt and ground- Grasswrens located in firescars of different ages expressed cover achieved near pre-burn coverage by 6 years post- as a percentage of the expected number based on the fire (Figure 3). Data for 1 and 2 years post-fire are from number of surveys undertaken for each firescar age. The Felderhof (2007), with whom we collaborated in a study regression line is forced through the origin as it is assumed of grasswrens and other birds in firescars <4 years old; that no birds are present immediately after fire. Firescars data for 3–10 years post-fire are from the present study. >10 years of age have been treated as 15 years old, the Data from older firescars are not used because their ages midpoint between 10 years (for which we had fire records) were variable and unknown. The mean ground-cover of and 20 years (discussion with landholders indicated that spinifex at survey points where Kalkadoon Grasswrens few areas survive >20 years without being burned). We were located was 37 ± 15.9%, which is similar to the 36% encountered six groups of Kalkadoon Grasswrens in ground-cover achieved after 6 years of growth post-fire 17 locations (35%) in firescars <5 years old, and located (Figure 3). the species in 32 locations (90%) in firescars ≥10 years Trees, shrubs, stones and rock were a feature of all but a old. few survey points, whether or not Kalkadoon Grasswrens The topography within 100 m of the 81 survey points were present. However, boulders were three times more where Kalkadoon Grasswrens were located is compared likely to be present at survey points where Kalkadoon with that within 100 m of the remaining 1412 survey points Grasswrens occurred than where they were not recorded. in Table 2. This table also details the topographic stratum at the point where the Grasswrens were first detected. Chi- Discussion square analysis (χ2 = 45, df = 5, P <0.001) indicates that the birds occurred in significantly hillier topography than Data from this study confirm the habitat description in would be expected by chance. Similarly, these Grasswrens Higgins et al. (2001) that Kalkadoon Grasswrens occur were predominantly located on hillsides when first sighted. over a wide range of topographies but most commonly are They were found on undulating areas with low topography associated with the slopes adjacent to steep, rocky ridges. only twice out of 81 encounters, both times within 300 m In our data, boulders were present three times more often of hills. than at survey sites where no Grasswrens were detected,

Table 2. Topography of Kalkadoon Grasswren habitat: Percentage of survey points with the specified topography within 100 m of the survey points for sites with (81 sites) and without (1412 sites) records of Grasswrens, and the percentage of times that the groups of Grasswrens were first located on this topography.

Topography Survey points with Survey points without Birds first located on this Grasswren records (%) Grasswren records (%) topography Flat/undulating 22 53 2 Shallow slope 65 57 23 Medium slope 58 43 29 Steep slope 46 28 27 Ridge 59 41 11 Drainage line 41 35 8 16 Australian Field Ornithology G. Harrington et al.

which suggests that protection afforded by exposed rock is firescars nor is it likely to be a sampling error because the important. This might be to provide protection from avian same people using the same methods were involved. predators and/or insulation from the heat of the sun. We We can offer no explanation. The density of Kalkadoon did not adequately test the proposition that the birds do not Grasswrens in habitat that has not been burnt for appear on large expanses of flat plains and river valleys ≥10 years reveals the importance of maintaining long-term even when covered in old-growth spinifex, but at the four fire refuges (Murphy et al. 2011). Although the population locations that could be described thus no Kalkadoon size of Kalkadoon Grasswrens appears to have been Grasswrens were recorded. relatively stable since it was first located in 1966 (Horton 2011), through to 2009, there is an ever-present threat The non-random occurrence of the Kalkadoon Grasswren of extensive fire in spinifex grassland. Across northern on geological substrates (Table 1) will have significance Australia many bird species are in decline. Commonly, as the environment responds to global warming and the the cause of this is attributed to grazing and/or fire, with species might need to change its territorial range. The introduced grasses often an influential, interactive factor favoured geological substrates do not occur widely outside (Crowley et al. 2004; Garnett et al. 2010). With the climate the current EOO. We did not anticipate that Kalkadoon trending to warmer and wetter in central and northern Grasswrens would occur on argillaceous detrital sediment Australia (VanDerWal et al. 2013), there is an increased (Table 1) because that substrate would be unlikely to carry likelihood of large, hot fires, so greater co-ordination in fire spinifex or provide the hills and rocks that this species management amongst graziers is vital to maintain viable appears to favour. The nine Kalkadoon Grasswren records populations of grasswrens and other spinifex-dependent from our survey on these mapped clayey substrates fauna. occurred close to the mapped boundary of the substrate. Inspection of the satellite imagery of this area indicates that The numbers of individuals within each group of it is not flat but comprises medium–shallow undulations Kalkadoon Grasswrens (mean 1.9, median 2) were low and valley floors. Our surveyors recorded 30% cover of compared with records for the Dusky Grasswren (3–10: rocks and stones and a similar cover by spinifex where Rowley & Russell 1997) and previous records for the Kalkadoon Grasswrens were found. Our conclusion is that Kalkadoon Grasswren (4–10: Carruthers et al. 1970). It the mapping is incorrect and that we do not have evidence seems probable that where we recorded single birds, there that the grasswrens preferentially occur on argillaceous was in fact a pair but we failed to see both birds. detrital sediment. The EOO presented in Figure 1 for the Kalkadoon Forty-five percent of the Kalkadoon Grasswrens’ EOO Grasswren shows substantial overlap with that ascribed to did not burn in the 6-year period 1998–2003 and only 6% the Carpentarian Grasswren (Harrington & Murphy 2016). burned twice in that period (Felderhof 2007). The frequency The size and shape of the body, bill and legs of these of burning is partially linked to the rate at which the cover species are similar, raising questions regarding their diet by spinifex recovers after fire, which is dependent upon and competition for food. Information on the diet of both rainfall and usually takes 4–5 years before it will support species is sparse, but the description in Higgins et al. (2001) the natural spread of fire within the range of the Kalkadoon indicates considerable similarity. This raises the question of Grasswren (Felderhof 2007). Burning in 2004–2007, why Kalkadoon Grasswrens thrive in the overlap zone but preceding our survey, involved only 7% of the EOO, leaving not farther north where the Carpentarian Grasswren occurs >90% of the EOO with spinifex >4 years old (Murphy et al. alone. It is most likely that the two species favour different 2011). Although our measurements showed that spinifex parts of the topographic catena in the overlap zone: the achieved near maximum cover 6 years after burning, the steep rocky hills favoured by Kalkadoon Grasswrens are form of the spinifex hummocks continues to change with absent from Boodjamulla National Park (pers. obs.). This age (pers. obs.). Mature spinifex >10 years old supports the may restrict the Kalkadoon Grasswren from moving farther highest density of Kalkadoon Grasswrens, which indicates north into the Park, but the habitat restrictions on which the that the birds are not simply responding to ground-cover by two species separate are also likely to include a climatic spinifex because cover did not increase significantly after factor because the Kalkadoon Grasswren is derived from the fire age of 6 years. This higher density of Kalkadoon the central Australian Dusky Grasswren whereas the Grasswrens in long-unburnt habitat may be related to the Carpentarian Grasswren is related to the monsoonal Black change in structure of the spinifex with time and/or the Grasswren A. housei and the White-throated Grasswren relatively larger area of contiguous habitat compared with A. woodwardi (Schodde 1982; Rowley & Russell 1997). the mosaic of different ages of spinifex after many small The climatic characteristics of central and northern fires. Such long-term stable habitat may allow Kalkadoon Australia are starkly different, and the influence of climate Grasswrens to achieve a relatively stable density compared on the distribution of these two species warrants further with the disturbed habitat, which is vacated by the birds study. after fire and presumably has to await a surplus of birds We did not encounter any grasswrens in the zone between nearby to be repopulated. The fate of birds that disappear the most easterly record for the Dusky Grasswren and the after their territory is burned is unknown. Habitat of a westerly location for the Kalkadoon Grasswren. Although subpopulation of Kalkadoon Grasswrens identified in 2008 conditions were unusually dry in August–September 2014 was partially burnt in 2009 and there were no grasswrens and may have rendered the birds unresponsive to the call- in the burnt area when surveyed 1, 3 and 4 years later. back, we did locate Kalkadoon Grasswrens to the north The lowest density of Kalkadoon Grasswrens occurred of Mount Isa whilst training surveyors before this survey. in 3-year-old firescars and the highest in 10-year-old Suitable habitat within the gap between the two grasswren firescars. However, the density in 7-, 8- and 9-year- species’ ranges is limited to isolated hills, so it seems old firescars was substantially lower than expected that our survey results tend to confirm the absence of (Figure 2), and is not reflected in the spinifex cover of these grasswrens in the gap. Kalkadoon Grasswren range and habitat 17

Because the EOO of the Kalkadoon Grasswren (Figure 1) Dutson, G., Garnett, S. & Gole, C. (2009). Australia’s Important is <50 000 km2, this species qualifies as a Restricted Range Bird Areas. Birds Australia Conservation Statement No. 15. Species and as such deserves the special conservation Birds Australia, Melbourne. attention that would be provided by an IBA. Although it is Felderhof, L. (2007). The Fire Patchiness Paradigm: A Case Study in Northwest Queensland. PhD thesis. James Cook present in the Buckley River IBA (which was declared for the University, Cairns, Qld. Carpentarian Grasswren), we believe that the Kalkadoon Garnett, S.T. & Franklin, D. (2014). Climate Change Adaptation Grasswren should also be monitored where it is the only Plan for Australian Birds. CSIRO Publishing, Melbourne. grasswren present. The dynamics of both species may Garnett, S., Szabo, J. & Dutson, G. (2011). The Action Plan for differ in the mono-specific situation compared with where Australian Birds 2010. CSIRO Publishing, Melbourne. they are sympatric. Both are classified as highly exposed Gaston, K.J. & Fuller, R.A. (2009). The sizes of species’ to the onward progression of climate change (Garnett & geographic ranges. Journal of Applied Ecology 46, 1–9. Harrington, G.N. & Murphy, S.A. (2016). The distribution and Franklin 2014). We recommend that an IBA be established conservation status of Carpentarian Grasswrens (Amytornis for the Kalkadoon Grasswren using the bounds of the EOO dorotheae), with reference to prevailing fire patterns. Pacific presented in Figure 1 south of latitude 20°50′S. Conservation Biology 21, 291–297. Higgins, P.J., Peter, J.M. & Steele, W.K. (2001). Handbook of Australian, New Zealand & Antarctic Birds, Volume. 5: Tyrant- Acknowledgements flycatchers to Chats. Oxford University Press, Melbourne. The data used in this paper were collected by 38 BirdLife Horton, H. (2011). Of spinifex and grasswrens, with particular Australia volunteers, who travelled from all states in Australia at reference to Amytornis ballareae, the Kalkadoon Grasswren. their own expense. Mount Isa Water contributed to the expense Queensland Naturalist 49, 1–3. of the survey. Paul Carter provided the records from the Selwyn Murphy, S., Harrington, G. & Felderhof, L. (2011). Preliminary Range. Prof. Jon Nott, James Cook University, assisted with spatial model using fire scar data to monitor Carpentarian interpretation of the geology. The Southern Gulf NRM provided Grasswrens. Report by Map IT, CSIRO Ecosystem Sciences logistical assistance and field maps. Funding for helicopter use and Birds Australia North Queensland. Available online: http:// in Boodjamulla National Park was provided by Xstrata Mount Isa birdlifenq.org/pdfs/Fire%20challenge%20report%20-%20 Mines. Particular thanks go to Bob Forsyth and Mark van Ryt for April%202011.pdf (retrieved 21 July 2016). sharing their expert knowledge of the area. David Stewart (Nature Plowright, H. (1994). A Field Guide to Australian Birdsong, CD 5: Sound) recorded the grasswren calls used in this survey. Dave Tawny Frogmouth to Kalkadoon Grasswren. Bird Observers Westcott generously provided insights into the interpretation of Club of Australia, Melbourne. our data. Justin Perry made useful comments on an earlier draft Rowley, I. & Russell, E. (1997). Fairy-wrens and Grasswrens. of this paper. Guy Dutson was generous with his ideas on how to Oxford University Press, Oxford, UK. improve the submitted manuscript. We wish to acknowledge Kath Schodde, R. (1982). The Fairy-wrens: A Monograph of the Shurcliff, Ceinwen Edwards, Steve Murphy, Chris Armstrong and Maluridae. Lansdowne Editions, Melbourne. Philip Brook for their contribution to this study. Schodde, R. & Mason, I.J. (1999). The Directory of Australian Birds: . CSIRO Publishing, Melbourne. Stattersfield, A.J., Crosby, M.J., Long, A.J. & Wege, D.C. (1998). References Endemic Bird Areas of the World: Priorities for Biodiversity Conservation. Birdlife International, Cambridge, UK. Carruthers, R.K., Horton, W. & Vernon, D.P. (1970). Distribution, VanDerVal, J., Murphy, H.T., Kutt, A.S., Perkins, G.C., Bateman, habits and sexual dimorphism of the , B.L. Perry, J.J. & Reside, A.E. (2013). Focus on polewards Amytornis textilis ballareae in northwestern Queensland. shifts in species’ distribution underestimates the fingerprint of Memoirs of the Queensland Museum 16, 334–341. climate change. Nature Climate Change 3, 239–243. Condon, H.T. (1969). A new subspecies of the Western Grass-Wren, Amytornis textilis (Dumont) in north-western Queensland. Memoirs of the Queensland Museum 15, 205– Received 22 February 2016, accepted 25 October 2016, 206. published online 2 February 2017 Crowley, G.M., Garnett, S.T. & Shephard, S. (2004). Management Guidelines for Golden-shouldered Parrot Conservation. Queensland Parks & Wildlife Service, Brisbane.