Patterns of decline in the native mammal fauna of the north-west slopes of New South Wales

15 Pages·1999·4.4 MB·English

by  David Paull

Patterns of decline in the native mammal fauna of the
north-west slopes of New South Wales
David C. Paulll and Elizabeth M. Date2
'Department of Zoology, University of New England, Armidale, New South Wales 2351
219 Towarri Street, Scone, New South Wales 2337
ABSTRACT
Twenty-one species of non-flying native mammals were positively identified during a
systematic fauna survey of forest, woodland and scrub habitats in State Forests on the north-
west slopes of New South Wales (1993-95). This is about 50% of the known historical diversity
of mammal species from this region (43 recorded species), with 11 considered extinct and
12 species known from the region, but not recorded during the survey. Of the species detected
during the survey, 10 can be considered as common or widespread, while 12 species can be
considered rare or under-recorded. The specific identity of two species detected from hair
and scat samples, Notomys sp. and Rattus sp., remain unclear as do observations made of
rat-kangaroos. This paper presents the results of the mammal survey and analyses the habitat1
resource preferences and weight ranges of the extant, declining and extinct species from the
region. The results of this analysis suggest that grass-dependent species and species between
50 g and 6 kg have suffered most from extinction. Species currently considered under threat
or whose status remains unclear generally have shrubby understorey and mature treeflog
preferences and weigh between 10 g and 1.5 kg. Mammals still common in State Forests of
the region are generally mature woodland (tree and log dependent) or shrubby understorey
species over 1 kg. Patterns of cultural disturbance are compared to these trends.
Key words: Extinction, Species decline, Habitat, Disturbance, Resources.
INTRODUCTION to as art of the North-west Cv~ressa nd
1ronba;k Belt, reflecting biogeogribhical and
A study of the fauna and habitats in State
commercial aspects of the region, mainly
Forests of the north-west slopes of New South
by the presence of the two most important
Wales was undertaken over two spring/summer
timber species, Narrow-leaf Ironbark Eucalyptus
seasons (1993-94 and 1994-95). This was
crebra and White Cypress Pine Callitris
undertaken to assist conservation manange-
glaucophylla. Both species dominate many
ment of the North-west Cypress and Ironbark
areas of State Forest in the study area. This
Management Area, for the State Forests of
region forms a unique biogeographical zone,
New South Wales. Threatened species were
generally regarded as a transitional zone
targeted for the surveys, though during the
between Bassian and Eyrean faunas, with some
course of the study a systematic survey of
Torresian species present. There are some
all vertebrate fauna groups, except fish, was
similarities with the zoogeography of the
undertaken. The objectives of the study were
south-west slopes (Caughley and Gall 1985;
to attempt to investigate forestry impacts upon
Bauer et al. 1996) though the north-west
the vertebrate fauna, as well as investigate
slopes have a higher faunal diversity, as well
their patterns of distribution and relative
as a suite of endemic fauna and unique
abundance. This study is an important part
vegetation characteristics, such as the
of a baseline of biological information that
dominance of White Cypress Pine (Date and
currently exists for the north-west slopes of
Paull 1996).
New South Wales. The data from the study,
with all locations of mammals encountered, There has been a low level of scientific
has been entered in several databases (NPWS scrutiny of the fauna on the north-west slopes
Atlas, Australian Museum, State Forest in historic and recent times though the level
records). of mammal extinction appears to be high (11
spp.) as does the historical mammal diversity
This paper places the results of the non-
when compared with other areas of Australia
flying mammal surveys within a broader
(Date and Paull 1996). Lunney et al. (1997a)
context of native mammal decline on the
found that the most common ecological
north-west slopes of New South Wales. A
attributes of threatened mammals throughout
survey of bats conducted by Coles (1995)
New South Wales were a seedtvegetarian
revealed 19 species present in State Forests
diet, a large body size, or ground-dwelling,
from a known diversity of 22 species for the
burrowing and rocklcave dwelling habits.
region (Date and Paull 1996).
This study will look at these factors with
The woodland, forest and scrub habitats respect to the extinct, declining and common
surveyed during the study are often referred native mammal species on the north-west
210 Australian Zoologist 31(1) June 1999
slopes, particularly their weight class and locations for all study sites assessed during the
habitat resource or sub-habitat preferences. study. The highest number of sites was in the
Pilliga State Forest.
THE STUDY AREA
The study area covers most of the north-west
The North-west Cypress and Ironbark State slopes region of the Great Dividing Range,
Forest Management Area was used as the including the bioregions of the Nandewar
study area for this study. It encompasses an Range, the Brigalow Belt South and the
area of approximately 125 000 km2 from the eastern edge of the Darling River Plains
New South Wales-Queensland border south (Thackway and Cresswell 1995). Elevation
to Dubbo and from Coonamble in the west ranges from 800 m a.s.1. in the east to 200 m
to Gunnedah in the east. Figure 1 gives the a.s.1. in the west. Several major rivers, which
149 E 150 E
Darling River
Plains Moree . .
19
, NET
N.ar n,oi. R ..- - sT
. . - -
* - m Nandewar '.
7r , 11
1
2
16
Coonabarabran
Brigalow Belt South
- - ,
- .
.
Dubbo ,
Sydney Basin
Figure 1. Map of study area showing State Forest sites and bioregional boundaries. Legend: (NET: New England
Tablelands. Dash-dot line shows easternmost extent of deep sands between the Namoi and Castlereagh Rivers).
Key. (State forest): 1. Pilliga west; 2. Pilliga west; 4. Pilliga west; 5. Euligal; 6. Pilliga east; 7.Etoo/Quegobla; 8. Yarrigan;
9. BaradineIOrr; 10. Wittenbra; 11. Pilliga east; 12. Pilliga east; 13. Pilliga east; 14. Yaminbah; 16. Pilliga eastIRutley;
17. Jack's Creek; 18. Killarney; 19. Bobbiwaa; 20. Moema; 21. MontroseICampbell; 22. Irrigappa; 23. Terrie Hie Hie;
24. Warialda; 25. Bingara; 26. Mehi; 27. Sepoy; 28. Bebo; 30. Severn; 31. Kerringle; 32. Leard; 33. Kelvin; 34. Goran;
36. Trinkey; 38. Tallegar; 39. SandgateIBourbah; 40. Yalcogrin; 41. Gilgandra; 42. BreelongIEura; 43. Beni; 44. Momo;
45. Goonoo.
June 1999 Australian Zoologist 31 (1) 21 1
form part of the Murray-Darling System, Floristically mixed woodlands, usually
flow from east to west within the area. They containing several species of eucalypt, were
include, from north to south, the MacIntyre, frequently encountered during the study.
Gwydir, Namoi, Castlereagh and Macquarie However, vegetation types appear to be
Rivers. A western extension of the Great influenced principally by soil type and depth,
Dividing Range, the Liverpool Range, rainfall and soil moisture. Vegetation structure
transects the area from east to west, south of has been influenced by past fire, grazing
Tamworth and Coonabarabran and joins the and logging regimes, and site floristics have
Warrumbungle Range, an ancient volcanic been found to he highly correlated with
outlier in otherwise gently undulating and disturbance history (Date and Paul1 1996).
flat country. A second western extension of
The overstorey vegetation identified during
the Great Dividing Range, the Nandewar
this study includes the following associations,
Range, ends at Mt Kaputar, another tertiary
which were found to support a considerable
volcanic intrusion.
range of understorey and structural variation:
State Forests and Crown timber lands within
(a) IronbarWhite Cypress Pine-Bull Oak
the study area mostly occur away from the
(Eucalyptus crebra, E. fibmsa, E. mehnophloial
major rivers on sands or gravelly soils, sandy
Callitris glaucophylln-Allomuarina leuhmanniz)
loams, duplex clays and yellow earths. Often
forests;
these soils are skeletal as they overlie the
extensive Pilliga Jurassic Sandstones or the (b) Blakely's Red GumiRough-barked Angophod
later volcanic intrusions. Deeper sands are White Cypress Pine (E. blakelyi/Angophora
typical of the western Pilliga and Nyngan floribunda/Callitris glaucophylla) woodlands
district, associated with the Darling Riverine and gully zones;
Plains. This is generally in contrast to the
(c) Box/White Cypress Pine (mostly E. Pofulnea
extensive agricultural areas along the major
and E. pilligaensislCallitris glaucophylla)
rivers, which occur on deep and heavier
woodlands;
black, grey or red clays and rich alluvial soils
(Mitchell and Rundle 1982). (d) Brown BloodwoodIBroad-leaved IronharW
Black Cypress Pine (Corymbia trachyphloial
Psior to European settlement, the vegetation
Eucalyptus fibrosalCallitris endlicfiri) wood-
of the study area was mostly shrubby and
lands; and
grassy woodlands with dry sclerophyll forest,
scrub and grassland also widespread (Forestry (e) Broombush (Melaleuca uncinala), heath,
Commission 1993). The habitats of this region mallee and other types of scrub.
can be considered as both warm-temperate
and semi-arid. Since then approximately
METHODS
50% of the natural vegetation of the western
slopes (Reed 1991) and 85% of the Central In order to undertake a comprehensive
Division of New South Wales, also known survey of commercial forests of the study area,
as the wheatbelt (Sivertsen 1993), has been 40 sites were selected from within potentially
cleared. Natural vegetation, in a form largely commercial compartments of State Forests in
undisturbed by Europeans, occurs in the Mt the Baradine, Inverell and Dubbo Districts
Kaputar area, the Warrumbungle National (Fig. 1). In 1993, three transects were selected
Park and the Pilliga Nature Reserve, plus at each site. If topography allowed, transects
adjoining non-commercial parts of the Pilliga were set at creek-line, mid-slope and ridge-top
State Forest and Crown Timber Lands. There positions within five km of each other, but
are also 18 small isolated Nature Reserves. normally one or two kilometres apart. Where
possible, transects of different vegetation
Commercial components of State Forest
type, logging, fire and grazing history were
lands in the study area are covered in natural
included at each site. In 1994, the same sites
vegetation that has been disturbed and
were surveyed, though different transects were
simplified to various degrees by past agri-
selected in an attempt to further quantify
culture, logging, grazing, burning and other
different disturbance patterns. Each transect
human activities. Any remaining areas of
was 200 m in length and ran perpendicular
old-growth or undisturbed forests within
to a forest road or followed a creek-line from
commercial areas of State Forests are small
the road. A 5 km road transect, connecting
and/or fragmented by cultural disturbance
some or all of the fauna transects was also
patterns. In fact few areas can be described as
included at each site.
"undisturbed. Nevertheless, in the North-west
CyptessIIronbark Belt relatively undisturbed For large and medium-sized mammals,
old woodlands and forests of low commercial detection was based primarily on spotlighting
value remain, particularly associated with the along the 240 vegetation and 80 road transects,
larger Pilliga, Goonoo and Bebo State Forests. identification of scats, opportunistic sightings
212 Australian Zoologist 31(1)
and the setting of two hair-tubes per transect Museum, as well as other sources (see Tables
on the ground for 10 or more days. Hair-tubes 2 and 3).
were placed at the roadside and 100 m
RESULTS
from the road along each transect. In 1993,
meat fat, rolled oats and peanut butter were Species detected during survey
used as bait. In 1994, a broader selection or
All mammal species detected during the
mixture of bait ingredients was introduced,
study are presented in Table 1 which also
including sardines, honey, apricots, aniseed
shows the frequency with which each was
and almond oil. In addition, one cage trap
detected, their observed habitat preferences,
per transect was set for three consecutive
their maximum (if available) or average
nights in 1994. Microhabitat features, particu-
recorded weights and their conservation
larly tree species, were recorded if an
status. Frequency counts were tallied from
arboreal mammal was sighted. Playback of
all observations and scat identifications,
Koala Phascolarctos cinereus and Squirrel Glider
trap results and hair-tube samples were
Petaurus norfolcensis calls were used at all sites.
identified for each s~eciesd urinrc the course
For small mammals, small Elliott traps were of the surveys. If c;rrently liste; under the
set for three consecutive nights at each Threatened Species Consewation Act 1995, species
transect. In 1993 eight traps per transect were here are given endangered or vulnerable
set on the ground, but in 1994, for each designations. If recorded five times or less,
transect, eight traps were set on the ground a species was regarded as "rare" though
and eight were mounted on tree brackets. this may also be a reflection of their level
Baits used were similar to those for hair-tubes. of detectability and not of their relative
Predator scats (fox and dog) and any small abundance. However, most species regarded
mammal scats were collected along transects, as rare here are also species which we know
roads and opportunistically. Four dry pit-traps little about, particularly in the Northwestern
were used for three days at all transects in CypressIIronbark Belt. Two species, the Yellow-
1993. They were connected by a Low, plastic footed Antechinus Antechinus flavipes and
drift fence in a cross formation covering an the Sugar Glider Petaurus brewiceps were
area of approximately 100 m2. These pit-traps detected less than twenty times, but were
were a bucket-sized (approximately 26 cm found to be widespread in State Forests of
deep x 23 cm in circumference) and did not the CypressIIronhark Belt.
catch any mammals.
Though listed as a threatened species, the
Analysis of hair from feral mammal scats Koala Phascolarctos cinereus is common and
and hair-tubes was undertaken by Barbara probably secure in the Pilliga but remains
Triggs. Compilation of historic records of vulnerable in the Gunnedah area, mainly
mammal fauna of the region was achieved by due to high levels of habitat fragmentation
consulting the NPWS Atlas, the Australian and destruction. Of the mammals detected on
Table 1. Results of non-flying, native mammal survey (1993-95).
Soecier Freauencv Observed habitat Weieht (kc) Status'
Tochyglosms ncubntus 139 ubiquitous 7 common
Daryunrr mnruhtw 1 mixed eucalyptus. woodland 7 vulnerable
Antechinus Javipir 16 woodlands, rocky complex .05 widespread
Sminthopri+r mu ri~ 5 shrubby woodlands ,025 rare
Macropus 3 grasslandlsavannaledge 65 rare
Macmpw giganteus 277 ubiquitous 60 abundant
Mnrmpw mbusw 40 mcky complex, ridge 45 common
Marmpur mfog.ircur 68 shrubby woodland, suub 19 common
Wallabia bicolor 169 shrubby woodland, scrub 17 common
Mumpus dorrolir 1 tall acacia woodland 16 endangered
Phcolarctos cine- 74 mature, mixed woodlands 6.5 vulnerable
Trichosunrr vulpicula 160 ubiquitous 4 common
Pscudochcirus pcrcgrinus 42 mature woodlands 1 common
Pclaurur norfolcccnr 5 mature woodlands .23 vulnerable
Pclounrr brrvicepr I3 mature woodlands .14 widespread
Cenarlchrr mnus 5 shrubby woodlands ,024 rare
Acmbotlr pypcus 2 mature woodlands ,014 rare
Ps~udarysp illigncnsir 3 shrubby woodlands ,012 vulnerable
Notomyr sp. 1 scrub? ? ?
Rnttur sp. 1 riparian grassy woodland ? ?
Hydromy chryrogarfir 1 permanent river 0.75 rare
Conir familiaris dingo 2 scrubby open woodland 19 rare
'Status as listed under the Threatmid Species Cmcrvation Act, 1995.
June 1999 Australian Zoologist 31(1) 213
the survey, 10 can be regarded as common detected during the study weighing under a
and widespread, while the other 12 can he kilogram that can be regarded as common. Of
regarded as rare or poorly known. the 11 species rarely encountered during the
study, four were over a kilogram; the Tiger
a) Habitat prefirences Quoll Dasyurus maculatus, the Black-striped
Wallaby Macropus dorsalis, the Red Kangaroo
The habitats identified in Table 1 also reflect
Macropw mfus and the Dingo. The rest are
the resource requirements of each species.
small species, weighing between 12 gm and
Most species in the study area seem to
750 gm (see Table 1).
be reliant on either of two main resource
typeslhabitats. Mature woodland species Species not detected during the suruey
primarily require tree and log resources (e.g.,
Table 2 lists native mammals that have
foliage, hollows) and may be found in shrubby
previously been positively identified from the
or grassy woodlands and forest, wherever there
study area, as recorded in current databases,
is adequate resource. Shrubby understorey is
but were not detected during these surveys.
a common habitat component throughout the
This may be because the surveys did not cover
CypressIIronbark Belt and is an important
preferred habitats for these species or these
food and shelter resource for most terrestrial
species are very rare or absent in the study
species encountered during the study. In many
area.
areas of shrubby understorey, trees are often
scattered and stunted, unsuitable for arboreal There are three and perhaps four species of
mammals. Other identified habitat types which dunnart known from the Pilliga. A dunnart
are important for some species are scrub described as Sminthopsis sp. nov. was identified
(vegetation dominated by growth between two by Lim (1992) during small mammal surveys
and four metres high), rocky complex, aquatic in the Pilliga Nature Reserve. He regarded
(for the Water Rat Hydromys chrysogaster) and it as closely resembling S. ooldea. The Stripe-
grasslands. faced Dunnart S. macroura, is known from
the northern parts of the study area with
Three species, the Eastern Grey Kangaroo
another record from Rocky Glen, east of
Macropus giganteus, the Echidna Tachyglossus
Coonabarabran in 1979 (Australian Museum
aculeatus and the Common Brushtail Possum
Records). Old records show the Fat-tailed
Trichosu~vu lpe&, were found to be common
Dunnart S. crassicaudata was once widespread
and widespread in the study area. The Eastern
across the study area, yet there is only one
Grey Kangaroo, a grass dependent species,
recent record from Sandgate State Forest
ranges widely throughout most habitats, but
(Clive Barker, pers. comm. 1998) as well as
tends to avoid rocky areas (the habitat of
other observations of this species near the
the Wallaroo M. robustus) and thick scruh
Macquarie Marshes (pers. obs. 1992). The
(which is the preferred habitat of the smaller
Common Dunnart S. murina was commonly
macropods, the Red-necked Wallaby M.
recorded in old records, however, none was
mfogrisew and the Swamp Wallaby Wallabia
trapped during this study.
bicolor). Both wallabies are also found wherever
there is extensive understorey cover. The No planigales have been recorded from
Echidna is also ubiquitous, although it seems State Forests of the CypressIIronbark Belt to
to avoid areas where there is no rock or log our knowledge, though the Narrow-nosed
cover. The Brushtail Possum is the most Planigale Planigale tenuimstrk and the
common arboreal species in the study area, Paucident Planigale J? gilesi may occur in
occurring widely throughout wooded habitats suitable wetlands or cracking clay habitats.
as well as in town trees and roofs. In the The Brush-tailed Rock-wallaby Petrogale
bush it avoids areas where there is little tree penicillata is thought to be extinct in State
cover, such as scrub, as it is essentially hollow Forests of the study area, though a population
dependent. still survives in the Warrumbungle National
Park. The Western Grey Kangaroo Macropus
The Dingo Canis familiaris dingo was
fuliginosus is regarded as common, though
recorded from observations made of two
only an occasional visitor to the north-west
sandy-coloured pups in Goonoo State Forest in
slopes of New South Wales, as is the Red
scrubby open woodland.
Kangaroo (Johnson et al. 1989). Another
macropod, the Rufous Bettong Aepyprymnus
6) Weight ranges
rufescens is known from throughout the region
The species most commonly detected. from historic and database records, though
including the macropods, the larger arboreal not positively identified for over 50 years
marsupials and the Echidna, all weigh 1 kg from the study area. It is known to prefer
or more. The Sugar Glider and the Yellow- a grassy understorey habitat (Johnson 1980;
footed Antechinus are the only two species Pople 1989).
214 Australian Zoologist 31(1)
Table 2. Native mammal species not detected during surveys (1993-95),
Species Habitat' Weight (kg) Records4 Status
Phascogob fnpoatafa mature forest .2 Singleton, 1969 (AM) vulnerable
Cobar, 198? (NA)
Pilliga, 19?? (SFR)
Sminthopris mocrourn shrublands, shrubby .02 Many records from northern vulnerable
woodland, scrub part of study area (AM)
Sminthogsis crmicnudnla grassland, shrubland, Throughout study area rare
open woodland, scrub ,015 (Marlow, 1958)
Sminthopsis sp. no". (Lim) shrubby woodland ? Pilliga NR, 1990 (Lim 1992) rare
Anterhinomys laniger grassland, shrubland, .03 Western edge of study area (AM) vulnerable
scrub
Aepyprymnw mfexcm grassy understorey, 1.6 Many records throughout study vulnerable
woodland edge area, north of the Castlereagh
R. (A.M., Rolls 1981)
Pctrogal~p enicillota rocky complex 7 Warrumbumgles (NA), vulnerable
Coolah, 1887 (AM),
Pilliga, 1992 (NA, Rolls 1981),
Mt Kaputar, 1975-81 (NA),
(Lunney d al. 1997b)
Macropus Jiliginorur grassland1 woodland 50 many recent records occasional
Petaurur australis mature forest .7 Mt Kaputar (Marlow 1958; NA) vulnerable
Petauroidcr volonr mature forest 1.5 Mt Kaputar (Marlow 1958) rare
Rottus uillorissimus grassland, shrubland .15 Merriwa. 1980 (AM), vulnerable
Coonamble, 1943 (AM)
Rnttur tunneyi grasslands, shrubby .075 Duck Creek IS?? (Marlow 1958) endangered
understorey
'Sources: Flannery 1990; Fax 1982; Marlow 1958; Pople 1989; Strahan 1995; Troughton 1941.
'Sources are given in brackts. Abbreviations: A.M.: Australian Museum records; N.A.: New South Wales National
Parks and Wildlife Service Atlas: SFR: State Forest Records.
The Brush-tailed Phascogale Phascogale wooded habitat and has not received much
tapoatafa has been recorded from the Pilliga scientific investigation. Most of the species
(State Forest records) and from an unidentified in Table 2 can be regarded as specialized in
State Forest near Cobar (NPWS Atlas). The their habitat requirements and sensitive to
large gliders have only been identified from disturbance, except for the Long-haired Rat,
the Mt KaputarINandewar area. The Kultarr which tends towards generalism, particularly
Antechinomys laniger once occurred throughout during irruptive periods (Redhead 1995).
the Liverpool Plains as did the Long-haired
Rat Rattus uillosissimus. The Kultarr and the b) Weight range
Pale Field Rat Rattus tunneyi are both thought
With the exception of the Western Grey
to be extinct in the study area (with most
Kangaroo and the Brush-tailed Rock Wallaby,
recent records over 100 years ago), while the
all species in this group are about one
most recently recorded possible irruption of
kilogram or less. There is a strong similarity
the Long-haired Rat in the study area was in
between weight classes of this group and those
1980 (Australian Museum record).
rarely detected during the survevs, most being
of
smali (<I kg). tie 14 sm'ai~t errestri;
a) Habitat preferences mammals ever recorded from the study area,
These unrecorded species are mostly reliant four are extinct, one is secure and the other
upon tall woodlandlforest tree resources, nine are rare, little known or declining
scrub, and shrubby and grassy understoreys. species. Of these only the Pilliga Mouse
Tall woodlandlforest was a habitat encountered Pseudomys pilligaensis, the Long-haired Rat, the
extensively in the, deep sands of Pilliga Stripe-faced Dunnart and the Kultarr are
West State Forest and adjacent areas which listed as threatened species in New South
were, and still are, prime timber-producing Wales.
compartments (Forestry Commission 1993).
Extinct species
Today much of this habitat is highly
structurally altered. The KaputarINandewar All species listed in Table 3 are known to
system still contains significant stands of taller have once occurred in the study area but are
Australian Zoorogisf 31(1) 215
Tabb 3. Extinct mammals of the study area.
Species Habitat preferences' Weight (kg) Recordsa
Das~nuui urrrinvr grasslandlwoodland edge, heath 1.3 Singleton, 1895 (AM)
Dorpm geofioii semi-arid habitats, inc. forests 1.3 Liverpool Plains 18??
(Marlow 19581
Pernrnclcs bougainuille shrubland, scrub, rocky complex .22 Liverpool Plains, 1863 (NA)
Macrotk lngotir grassland, shrubland, grassy woodland 2.5 Gilgandra, 1899 (AM)
Bettongin bmmr grassland, shrubland 1 Macquarie R. l8??
(Marlow 1958)
~ - -
Betton@ p~nicilhlo( tmplco?) scrub, woodlands 1.3 Liverpool Plain. 1863 (NA)
Lngorchcrtcr bponks grassland, grassy woodland 4 Liverpool Plain, 1863 (NA)
Onychogaba frninata shrubby, grassy woodland 6 Manilla, 1840 (NA),
Pilliga (Rolls 1981)
Aeudmys aurholk grassland ,065 Manilla, 1840 (NA)
Priudomys gouldii grassland .05 Liverpool Plain IS??
(Marlow 1958)
Conilunu ol6ipa mature forest, woodland .2 Macquarie River'? 1846 (NA)
'Sources: Flannery 1990; Fox 1982; Marlow 1958; Pople 1989; Strahan 1995; Troughton 1941.
PSources are given in brackts. Abbreviations: AM: Australian Museum records; NA: New South Wales National Parks
and Wildlife Service Atlas.
currently listed (in the Threatened Species area. The degree of habitat specialization for
Conservation Act 1995) as species probably these animals can only be speculative.
extinct (not positively identified in the last
50 years). The identity of the "Brush-tailed b) Weight mnge
Bettong" fromqhe CypressIIronbark Belt is
All species regarded as extinct in the study
uncertain at present, as it may have been
area at present fall into a weight range
either Bettongia penicillata or B. tropica (C.
between 50 gm and 6 kg, with seven of those
Dickman, pers. comm.).
I1 species weighing between 1 and 6 kg,
Historical fauna records from this region the other four below 1 kg. So in contrast to
of Australia are few and restricted to reports the other groups, extinct mammals were
from early European pioneers, some species mostly medium-sized (1-7 kg) along with four
being recorded only once or twice from small rodents and a small bandicoot.
imprecise locations. It is impossible to say
precisely when, or even if, any of these species Significant records from the study
became extinct in the study area.
a) Quolk
a) Habitat preferences Hair collected from a hair tube at Quegohla
State Forest (Site 7) was identified to the
Habitat categories assigned to the extinct
genus Dasyurus, probably the Tiger Quoll
fauna here are based on old and new sources
D. maculatus (B. Triggs, pers. comm.). This is
(see Table 3). The main difference between the
the first positive identification of this genus
habitat preferences of this group compared
from the Pilliga. The quo11 was detected in
with the others is that out of twelve species
December, well outside the known breeding
which may have been present in the study
season for this species (Mansergh 1983a,b).
area, nine (including both the Eastern and
This points to the possibility that there is a
Western Quolls Dasyurus viverrinus and D.
relict population of quolls in the Pilliga, at
geofiii) are known to have been dependent
least in the northern section, whose status
upon grassy groundcover for forage, cover
urgently needs assessing.
andln.r- n- -e- s~ t~ in0r r-eso-urc~es.~ T he onlv tree-
dependent species regarded as extinct is Three species of Dasyurus (D. muculatus, D.
the White-footed Rabbit-rat Conilurus albipes. viuerinnw and D. geoffroii) have been described
Based on available sources the Brush-tailed from the north-west slopes area in historic
Bettong and the Western Barred Bandicoot times, making this the only region in Australia
Perameles bougainuille seem to have preferen- in which the three species had sympatric
tially selected shrubby understoreys, though ranges. The Tiger Quoll, the most frequently
habitat preferences were never noted in detail recorded species, has been recorded in the
for most of the extinct fauna in the study study area near Coolah in 1887 and 1973,
216 Austmlian Zoologist 31(1) June 1999
Dubbo in 1985, Gilgandra 1957, near Mt All individuals were flushed from shrubby
Kaputar in 1972, 1977 and 1982 (NPWS, understorey woodland, and moved rapidly in
Australian Museum databases) with only front of slow moving vehicles at night.
unconfirmed reports from the Pilliga prior
Three species of potoroid macropods have
to this study. Recent specimen-based identi-
been recorded from the region historically
fications of this species have also come
(Gould 1841; New South Wales NPWS database).
from Walgett and Carroll, New South Wales
These are the Brush-tailed Bettong, the
(T. Mazzer, NPWS, pers. comm. and The
Burrowing Bettong and the Rufous Bettong.
Northern Daily Leader, September, 1997).
Both the Brush-tailed and Burrowing Bettongs
Judging by the spread of recent and historical
were once common in the study area. John
records of the Tiger Quoll, it was once found
Gould reported Brush-tailed Bettongs on the
throughout the North-west Cypressllronbark
Liverpool Plains and the Namoi River in 1863
Belt.
(Marlow 1958) and the Burrowing Bettong
The known distribution of the Tiger Quoll once occurred on the western edge of the
is along the eastern escarpment and its study area (Robertson 1981; Seebeck et al.
eastern and western ranges, also occupying 1989). There are claims of sightings of Brush-
nearby coastlands and river flats, wherever tailed and one of a Burrowing Bettong in the
there is mature forest. It seems to depend Pilliga area during the 1980s (NPWS database),
on the oldest forest communities, with high though positive identifications are lacking.
levels of moisture retention, prey availability Information on their habitat and behaviour is
and suitable refugia (old trees and logs). The also limited (Stoddart 1966; Christensen 1980;
Quegobla forest is a much drier environment Johnson 1980; Robertson 1981).
than expected for the Tiger Quoll. This
Previous records of the Rufous Bettong in
forest is characterized by a structural and
the study area are old. They include one near
floristic complexity with relatively low levels
Coonabarabran in 1906, one from Narrabri in
of human disturbance, with some evidence of
1927 (Marlow 1958) and seven from the
old logging activity (50-100 years ago), old
Manilla area during the 1920s (Australian
bushfires, feral rabbits and goats, as well as
Museum database). Rolls (1981) records
the presence of an exotic carnivore (dog scats
several first-hand accounts of rat-kangaroos
were found along roads in this area). There
(probably Rufous Bettongs) over many years
were high numbers of mature Ironbarks,
in the Pilliga, with an observation of their
Red Gums and Box species, a very diverse
decline in more recent times.
understorey of White Cypress Pine, Bull Oak,
grasstrees Xanthorrhoea sp., and Acacia spp.
6) Wallabies
as well as patches of low shrubs. Moisture
and nutrient levels within this habitat were A Black-striped Wallaby Macropus dorsalis was
thought to be relatively high, due primarily to spotlighted as it crossed a road in Euligal
its good understorey and groundcover and State Forest, near Etoo Creek. This observa-
overall structural complexity. The area could tion was made in an IronbarWCypress
be classed as a mature, mixed eucalypt tall forest with a dominant, tall acacia understorey.
woodland with a developed shmbbylgrassy This habitat type is consistent with known
mosaic understorey. The extent of such habitat preferences for this species, which inhabits
within the Pilliga forests is uncertain, though scrubby or vine wet forest in north-east New
may be helped by forthcoming vegetation South Wales and Brigalow scrub in Queens-
survey information (State Forests of New South land (Kirkpatrick 1995). Though a social
Wales, in prep.). Data from this study showed species (Kirkpatrick 1995), the individual
that mixed eucalypt/cypress pine habitats are sighted in Euligal State Forest was probably a
often characterized by a diverse vertebrate solitary male, standing over one metre tall,
fauna (Date and Paul1 1996). with a prominent hip-stripe and a conspicuous
vertebral stripe, visible as it moved into
b) Rat Kangaroos the scrub in an unhurried way, with its head
held low and its tail held aloft (pers. obs.
In Pilliga East State Forest, near Gilgai
1993). The Black-striped Wallaby may be
FR, a sighting of two small macropods, closely
easily confused with the Red-necked Wallaby,
resembling bettongs or rat-kangaroos was
however, the vertebral stripe is a distinctive
made during this study. Another observation
feature of the Black-striped Wallaby and is the
was made of one individual in the vicinity
basis of identification in this instance.
of Dandry Creek on private land near the
Pilliga Nature Reserve (pers. obs. 1994). The Black-striped Wallaby is distributed
Identification to species level was not possible from Townsvilles, Queensland, south along
due to the fleeting nature of the observations the coast, ranges and inland slopes and into
and could have been any species of bettong. northeastern New South Wales west of the
Australian Zoologist 31(1) 21 7
ranges (Kirkpatrick 1995). It is regarded as from a fox scat near Lanes Mill Flora Reserve
common in Queensland, even a pest (Calaby in the central Pilliga. The hair was identified
and Grigg 1989; Pople 1989), but has declined as belonging to the genus Notmnys, resembling,
over most of its range (Calaby and Grigg in cross-section, a hair from N. mitchellr (B.
1989), especially in New South Wales (Marlow Triggs, pers. comm.). However, there is some
1958; Johnson et al. 1989). Previous records conjecture as to the identity of this hair
in the study area include 11 near Manilla, sample at present, particularly whether it can
registered in 1910 and 1929 (Australian be distinguished from hair of a species of
Museum database), four near Narrabri in Pseudomys. There are no other records of this
1978, 1986 and 1991, one from the Ban Baa genus from the study area, apart from
area on the eastern side of the Pilliga, one an earlier report by Rolls (1981) of possible
near Tambar Springs in 1979 and some from Notomys footprints from somewhere in the
the Brigalow Nature Reserve in 1979 (NPWS central Pilliga area. Hopefully future small
database). Most of these records are from areas mammal surveys in the Pilliga will clarify this
now largely altered for stock and cropping record.
agriculture. The Brigalow Nature Reserve is Scats thought to belong to a native species
still thought to support a population of this of Rattus were discovered at a creekside
species. As far as we are aware, this Nature location in Kelvin State Forest, a mountaid
Reserve is the only protected patch of forest isolate north of Gunnedah. The scats,
Brigalow habitat in New South Wales, though uncovered from shallow burrows in and near
unprotected patches of mixed Brigalow also
a creek bank, appear to resemble those of
occur in the northern Pilliga and surrounding the Bush Rat R. fuscipes, (B. Triggs, pers.
region. Due to its size, the Pilliga may comm.), much larger than a Pseudomys yet not
represent one of the last important refuges for exhibiting features typical of introduced
the Black-striped Wallaby in New South Wales.
Black Rat R. rattus scats, having a brown
colour and thicker shape. The Bush Rat is
d) Gliders
not generally found more than 100 km from
Two unconfirmed observations were made the coast (Lunney 1995). The local vegetation
of a large glider (Family: Petauridae) during was a grassy Red Gumwhite Bodcypress
the surveys. One was made by spotlight in woodland surrounding a shallow, intermittent
Mehi State Forest, in Grey BoxJRed Gum tributary.
woodland near the Gwydir River. It was a Though tenuous, a positive identification
large, dark glider though the specific identity of this rat species needs to be the focus of
remains inconclusive because the individual future research. Two Rattus species are known
glided about 40 m from its tree position to historically from the north-west slopes of New
the ground and escaped (Clive Barker, pers. South Wales, however, neither has been
comm.). The other unconfirmed record comes observed recently. The Pale Field-rat R. tunneyi
from a hair tuft collected at a dry creek near is known only from black soil country west
Gwabegar, Pilliga West State Forest (Site 2). of the study area (Marlow 1958) and also
The hair filaments were considered too thick from north-east New South Wales (Redhead
to be those of the Squirrel Glider Petaurus 1995). The Long-haired Rat R. uillosissimus has
norfolcmis or the Sugar Glider F! breuiceps, it been recorded from the Liverpool Plains area
was considered more likely to be hair from a (see Table 2). Fluctuations in the distribu-
Yellow-bellied Glider Petaurus australis or a tion and abundance of the Long-haired Rat
Greater Glider Petauroides uolans (B. Triggs, seem to correspond to cycles of wet seasons
pers. comm.). The southern subspecies of the (Redhead 1995) and recent drought may
Yellow-bellied Glider F! a. australis has disjunct account for its apparent scarcity in the study
populations in coastal western Victoria and area. However, its presence may be overlooked
from the Nandewar Range (1991, NPWS Atlas or confused with introduced species.
records). The Greater Glider was also known
from the KaputarJNandewar area (Marlow Predation by feral species
1958). A follow-up survey to the Pilliga West
Feral predator (fox and dog) scats were
site failed to detect any evidence of either
collected for hair analysis (B. Triggs) to
glider (R. Kavanagh and G. King, State Forests
investigate their diets particularly with relation
of New South Wales, pers. comm.) and neither
to mammal prey (Table 4). Only one cat scat
has been identified previously from the Pilliga
was located and it contained insect material.
forests.
Dog and fox scats were usually found along
roads and tracks and around dams, but fox
e) Natiue Rodents
scats were also detected away from roads in
Possibly the most unexpected discovery of a variety of situations, but usually found on
the surveys was that of a hair sample taken or near logs and other landmarks. It is not
218 Australian Zoofogisf 31(1)
Table 4: Feral carnivore prey items collected from fox and not form a large part of the diet of these
dog scats (from Date and Paull 1996). carnivores (1 item out of 125).
Species Fox Dog Total Foxes took a wider range of native mammals
Bos taums 5 9 14 (10 spp.), including five small species. This
Capra hincus 1 6 7 evidence, along with the fox's willingness
Ovis aries 4 7 11 to hunt widely away from roads, suggest
Oryctolagus cuniculus 5 8 13
the fox is a more effective hunter of small
Felis cattus 2 - 2
native mammals than the dog. How much
Rattus rattw 1 3 4
Large Macropus spp. 3 14 17 these species rely on scavenging is unclear.
Small Macropus/Mallabia 6 11 17 In this study the preferred prey species of
Trichosurus sp. 14 9 23 the fox was the Brushtail Possum, with the
Phascolarctos sp. - 1 1
Swamp Wallaby, Yellow-footed Antechinus,
Tachyglossus sp. - 5 5
Petaurus sp. 1 - 1 cattle and rabbits also important. Interestingly,
Antechinus jlauipes 5 - 5 two fox scats were found to contain cat Felis
Cercartetus nanus 1 1 2 cattus hair, suggesting possible predation.
Notomys sp. 1 - 1
Sminthopsis murina 1 - 1 Overall, the most favoured food items of the
-
Acrobates bvpmaeus 1 1 feral carnivores were the large native mammals,
Totals 51 items 74 items 125 items with predationlscavenging of cattle, sheep and
15 species 1 1 species 17 species rabbits also important. Small mammals seem
to be more important for the fox. Of a
possible six species of Macropus from the study
area, only two, the Eastern Grey Kangaroo and
known whether the dog scats come from feral
the Wallaroo, were positively identified from
or domesticated dogs or Dingos which were
hair samples (B. Triggs, pers. comm.).
only positively identified in Goonoo State
Forest.
DISCUSSION
Dogs took fewer native mammal species than
foxes (5 spp), with a greater reliance on stock, For the purposes of indicating the decline
feral herbivores and large macropods. The in native non-flying mammal. species in the
most favoured items for dogs were large study area, rare species detected during the
macropods and the Swamp Wallaby, with the survey and those species not detected but
Echidna, cattle, sheep, goats and rabbits also not considered extinct in the study area were
favoured. Only dogs were found to prey on lumped together as "rare or unsure status"
the Echidna. Other interesting species found species. Figures 2 and 3 compare the relative
in dog scats were the Eastern F'ygmy-possum number of common, declining and extinct
Cercartetus nanus and the Koala. Despite their non-flying native mammal species according to
willingness to come to the ground, Koalas did resourcelhabitat preferences (grass-, shrub-,
Common Rare or Extinct species
unsure
status
I I
Grass Shrublscrub Mature treenog
I
Wildlife habitat resources
Figure 2. Relative decline in wildlife habitat resources on the western slopes of New
South Wales.
June 1999 Australian Zoologist 31 (1) 219