(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Proceedings of the Linnean Society of New South Wales"

PROCEEDINGS 

fthe 



INNEAN 
OCIETY 



NEW SOUTH WALES 




NATURAL HISTORY IN ALL ITS BRANCHES 






THE LINNEAN SOCIETY OF 

NEW SOUTH WALES 

ISSN 0370-047X 




Founded 1874 
Incorporated 1884 

The Society exists to promote the cultivation and study 
of the science of natural history in all its branches. 
The Society awards research grants each year in the 
fields of Life Sciences (the Joyce Vickery fund) and 
Earth Sciences (the Betty Mayne fund), offers annually 
a Linnean Macleay Fellowship for research, contributes 
to the stipend of the Linnean Macleay Lecturer in 
Microbiology at the University of Sydney, and 
publishes the Proceedings. It holds field excursion and 
scientific meetings, including the biennial Sir William 
Macleay Memorial Lecture delivered by a person 
eminent in some branch of natural science. 



Membership enquiries should be addressed in the first instance to the Secretary. Candidates for elec- 
tion to the Society must be recommended by two members. The present annual subscription is 
$A56.00. 

The current subscription rate to the Proceedings is set at $A80.00 per volume. In recent years a 
volume consists of a single annual issue. 

Back issues of all but a few volumes and parts of the Proceedings are available for purchase. Prices 
are listed on our home page and can also be obtained from the Secretary. 

OFFICERS AND COUNCIL 2005/2006 

President: M.L. Augee 

Vice-presidents: K.L. Wilson, A. Ritchie, J.R Barkas, I.G. Percival 

Treasurer: I.G. Percival 

Secretary: J-C. Herremans 

Council: A.E.J. Andrews, M.L. Augee, J.R Barkas, M.R. Gray, J-C. Herremans, M.A. Humphrey, 

D. Keith, R.J. King, H.A. Martin, RM. Martin, E. May, M.S. Moulds, D.R. Murray, 

P.J. Myerscough, I.G. Percival, A. Ritchie, S. Rose, and K.L. Wilson 

Editor: M.L. Augee 

Assistant Editor: Elizabeth May 

Linnean Macleay Lecturer in Microbiology: PR. Reeves 

Auditors: Phil Williams Carbonara 



The postal address of the Society is: RO. Box 82, Kingsford NSW 2032, Australia 

Telephone: (International) 61 2 9662 6196; (Aust) 02 9662 6196 

E-mail: linnsoc@acay.com.au 

Home page: www.acay.com.au/~linnsoc/welcome.html 

© Linnean Society of New South Wales 

Cover motif: Vegetation profile of swamp and woodland association in a study site within 
Gibraltar Range National Park (Virgona et al. pages 39-47, this volume). 



PROCEEDINGS 

of the 

LINNEAN 
SOCIETY 



of 

NEW SOUTH WALES 



For information about the Linnean Society of New South Wales, its pubhcations and 

activities, see the Society's homepage 

www.acay.com.au/~Hnnsoc/welcome.htm 



VOLUME 127 

February 2006 



EDITORIAL 



This volume contains a special section composed of papers dealing with the biology 
and ecology of the Gibraltar Range National Park. The Gibraltar Range is in the NE comer of the 
state of New South Wales (29°3r 152° 10'), between the towns of Glen Innes and Grafton, on the 
eastern edge of the Great Dividing Range. Details are given in the introductory paper, pages 1-4. 

The Gibraltar papers have been sub-edited by Peter J. Clarke and Peter J. Myerscough, 
and the Linnean Society of NSW appreciates the work they have put into preparing the section. 

2007 is the three hundredth anniversary of the birth of Carl von Linne (1707-1778). He is of course 
more commonly known these days by the Latinised version of his name under which he published - Carolus 
Linnaeus. There will no doubt be various events across the globe to mark this important aimiversary in 
the history of natural science. The Linnean Society of NSW is plarming a special symposium, and arising 
from that, a special issue of this journal. Both will be concerned not only with the history of Linnaeus 
and Liimean Taxonomy, but also with the recent advances in systematics that have built upon the ground- 
breaking work of Liimaeus. That issue of the journal will also be open to papers dealing either with Lirmaeus 
himself or systematics at any level whether they have been part of the symposium or not, so start now! 



M.L. Augee 
Editor 



11 



Introduction to the Biology and Ecology of Gibraltar Range 
National Park and Adjacent areas: Patterns, Processes and 

Prospects 

Peter J. Clarke' and Peter J. Myerscough^ 

'Botany, School of Environmental Sciences and Natural Resources Management, University of New England, 

Armidale, 2351 (pclarke@une.edu.au) 
^School of Biological Science, The University of Sydney, Sydney, 



Clarke, P.J. and Myerscough, P.J. (2006). Introduction to the biology and ecology of Gibraltar Range National 
Park and adjacent areas: patterns, processes and prospects. Proceeding of the Linnean Society of New 
South Wales 111, 1-3. 



Papers on the biology and ecology of Gibraltar 
Range National Park were sought to reflect the 
increased research focus on the area over the past 
decade. The 12 papers, published here, come from a 
variety of natural history disciplines. This collection 
of papers reflects the start that has been made, and, 
hopefully, will stimulate further biological and 
ecological investigation of Gibraltar Range National 
Park. 

Gibraltar Range National Park was first dedicated 
in the 1960s following the construction of the Gwydir 
Highway connecting Glen Innes and Grafton in 
northern NSW. Prior to this the area had been used 
for grazing, prospecting, forestry and had been 
surveyed for the potential use of hydroelectricity. 
However, it remained little explored in terms of its 
biology and ecology until the 1960s and 70s when 
John B. Williams began to collate species lists and 
describe the broad patterns of vegetation (Williams 
1970, 1976). On his first exploration in 1958 he 
noted the similarity of the vegetation to that of the 
Sydney Region but also noted that many of the plant 
genera have species that are endemic to the granite 
flora (pers. comm.). This observation is still being 
examined today and is exemplified in the paper by 
Jones and Bruhl describing a new species oi Acacia. 
John Williams was also acutely aware of the influence 
of geology and soils on vegetation and the role of 
these differences in producing diverse habitats. 
These themes are explored by Williams and Clarke 
in their description of the vegetation, and by Vemes 
et al. and Mahony in their accounts of the mammals 
and amphibians respectively. Whilst we now have 
a good understanding of vascular plant distribution 
and abundance there are many gaps in knowledge of 



the more cryptic vertebrate fauna and invertebrates. 
Surprisingly, the more easily studied avifauna has not 
been well documented at Gibraltar Range despite the 
wealth of opportunities for behavioural and ecological 
studies in diverse habitats. 

The biological processes that influence the 
distribution and abundance of community dominants 
at Gibraltar Range National Park are being better 
understood through quantitative surveys, comparative 
biology and experimental manipulations. In particular, 
the influence of fire regimes on the sclerophyll and 
rainforest flora has been advanced by the papers in 
this volume by Campbell and Clarke, Croft et al., 
Knox and Clarke, and Williams and Clarke. At finer 
scales Virgona et al. have elucidated the proximal 
factors governing the distribution of Banksia species, 
which are a keystone resource in heaths and adjacent 
forests. Furthermore, Vaughton and Ramsey have 
experimentally examined the reproductive biology of 
one such Banksia species to explore the evolution of 
plant mating systems. Whilst all banksias set seed at 
Gibraltar Range National Park some other members 
of the Proteaceae family appear to be sterile as 
documented by Caddy and Gross in their population 
study of a rare species of Grevillea. 

Future prospects for the biota of Gibraltar Range 
National Park are seemingly assured through the 
management of the conservation reserve by NSW 
National Parks and Wildlife Service. However, 
the paper by Goldingay and Newell highlights that 
recreational use of protected areas may impact 
the quality of habitats for wildlife through the 
apparently innocuous disturbance of rocks. The 
complex task of fire management is also highlighted 
in the study of Knox and Clarke, who conclude that 



INTRODUCTION TO GIBRALTAR ROPERS 



short fire fi-equencies can reduce the resprouting 
ability of common shrubs. In short, it is clear that 
enticing prospects for future research and adaptive 
management are many in Gibraltar Range National 
Park. 



DEDICATION 

John B. Williams (12/2/1932 to 31/7/2005) 

This collection of papers is dedicated to John B. 
Williams who was instrumental in describing the flora 
of Gibraltar Range National Park and that of the New 
England Region more generally. John Williams will be 
remembered for his wealth of knowledge about plants 
and his intuitive guides and keys to various Australian 
plant groups. John lectured in Taxonomy and Ecology 
at the University of New England for nearly 40 years 
and after 'retirement' remained actively involved in 
teaching and research. His passion for botany, natural 
history and conservation was conveyed to a wide range 
of people through his lectures, public talks, activities 
in conservation, numerous checklists, ecological 
notes and published books. His interests in heaths, 
sclerophyll forests, and rainforests have inspired 
many to pursue the description and explanation of 
their ecological patterns and processes. This legacy is 
reflected in many of the papers published on research 
done in Gibraltar Range National Park. 



REFERENCES 

Williams, J.B. (1970). A preliminary list of the seed 
plants of the Gibraltar Range National Park. 
Unpublished Notes, University of New England, 
Department of Botany. 

Williams, J. B.(1976). Notes on the vegetation of 
Gibraltar Range National Park. Unpublished 
Notes, University of New England, Department 
of Botany (reproduced below) 



APPENDIX 

Reproduced from Williams (1976) 

Gibraltar Range National Park consists in its upper 
section of an undulating granite plateau, while the 
lower section is steeply dissected, and has a variety 
of underlying rock types. 

The plateau section is about 1000 to 1250 m in 
altitude and its natural features are dominated by the 
underlying pink granite (leuco-adamellite) - a very 



coarse-grained and siliceous rock. This weathers to 
form shallow, gritty soils with some extreme nutrient 
deficiencies (especially in phosphate), and the upper 
slopes and hilltops have extensive bare rock outcrops, 
and some spectacular tor-fields (groups of very large 
granite boulders on ridgetops). In the rock crevices and 
between the tors are patches of low heath and scrub 
vegetation with several unusual flowering shrubs. The 
slopes and gullies of the plateau landscape carry LOW 
OPEN-FOREST with stringybarks and peppermints, 
and a very large nvimber of shrub species (see separate 
list). 

Several eucalypts are found in these low forests, 
in varying associations. Four of them are very common 
and widespread; these are Youman's Stringybark {E. 
youmanii); Privet-leaved Stringybark {E. ligustrind). 
New England Blackbutt {E. andrewsii) and Coast 
Blackbutt {E. pilularis). Others with local occurrences 
are Needle-leaved Stringybark {E. planchoniand). 
Narrow-leaved Peppermint {E. radiatd) and Round- 
leaved Gum {E. deanei). The remaining eucalypts of 
the granite areas favour special habitats where they 
are often locally dominant. So we may find smooth- 
barked Mountain Ash {E. oreades) as a fringe of 
white-trunked trees around the base of some of the 
high tor-fields. Among the rocky outcrops there are 
patches of Mallee {E. approximans) in some areas, 
and stunted trees of the Red Mahogany {E. notabilis). 
Along watercourses in shallow valleys narrow bands 
of Mountain Gum {E. dalrympleand) and Peppermint 
{E. acaciiformis) may occur. In a few deeper gully 
areas with better, sandy soils and some shelter from 
wind, patches of TALL OPEN-FOREST are found, 
with Gum-topped Peppermint {E. campanulatd). 
Messmate {E. obliqud) and Diehard Stringybark {E. 
cameronii) as the dominants. Such patches are found 
on the Mulligan's Hut Track. 

In several of the shallow valleys of the plateau the 
forest cuts out abruptly, giving way to extensive open 
peat swamps with a natural treeless SEDGELAND 
(moorland) of sedges and rushes, other herbs and 
low shrubs. This plant community again is dependent 
on the special way in which the pink granite has 
weathered, to form swampy valleys with an acid, 
peaty soil in this high-rainfall area. 

The main plants in these wetlands are coarse 
tough-leaved herbs, including restiads such as Restio 
andLepyrodia, and large, tufted sedges such as Button- 
Grass {Gymnoschoenus), Spike-sedge (Schoenus) 
and Razor-sedge (Lepidosperma). [Beware of Razor- 
sedges, the flat, narrow, leaves and stems have sharp 
edges which can cause deep cuts.] Along the sluggish 
watercourses in the swamps are several small shrubs 
which flower well in late spring and summer. These 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.J. CLARKE AND P.J. MYERSCOUGH 



include myrtles such as Leptospermum, Baeckea 
and Callistemon and epacrids such as Epacris 
microphylla. Christmas-bells {Blandfordia) are a 
feature of the swamps in summer. Three small insect- 
trapping herbs with red, sticky leaves may be seen in 
parts of the swamps. These are the Sundews, Drosera 
spathulata, D. auriculata and the larger, showy D. 
binata with long, forked leaves. 

The lower section of the park and some areas 
near the edge of the plateau have steep slopes, high 
rainfall and different rock types giving richer, deep 
soils. Here are found some TALL OPEN-FORESTS 
with very fine, large specimens of Blue Gum {Euc. 
saligna). Tallow Wood {E. microcorys), Silver-topped 
Stringybark (E. laevopinea). Gum-topped Peppermint 
{E. campanulatd) and Brush Box {Tristania conferta). 
Some of these trees are over 160 ft high. The 
understorey in these forests contains wattles, treefems, 
some "rainforest" shrubs and vines, and some tall 
flowering shrubs such as Nightshade {Solarium 
cinereum), Mint-Bush (Prostanthera), Correa and 
Tall Everlasting {Helichrysum rufescens). 

Near the bottom of the range, the rainfall is much 
lower, and OPEN-FPREST of a drier sort occurs, with 
trees such as Ironbark, White Mahogany, Bloodwood 
and Broad-leaved Apple. 

In sheltered gullies and on some east-facing 
slopes, the open-forests give way to stands of 
rainforest, of which two forms are found in the 
Park. SUBTROPICAL RAINFOREST, with palms, 
strangling figs. Red Cedar, Yellow Carabeen, 
Rosewood, Stinging Tree and many large vines 
occurs on the scarp, and at mid and low altitudes 
generally. Fine stands may be seen in Cedar Valley, 
and on the steep descent along the highway below 
the tick-gate. WARM-TEMPERATE RAINFOREST 
with Coachwood, Sassafi"as, Crabapple, Corkwood, 
Prickly Ash, Laurels, and many ferns, is found above 
1000 metres, sometimes right on the plateau surface 
(e.g. a little north of the Washpool Road tumoff). 
Large epiphytes such as Birds-nest Fern (Asplenium 
nidus), Elkhoms {Platy cerium), Dictymia, and many 
orchids are common and conspicuous high up in the 
trees, especially in the Subtropical rainforests. 



Proc. Linn. Soc. N.S.W., 127, 2006 



Acacia beadleana (Fabaceae: Mimosoideae), a New, Rare, 
Localised Species from Gibraltar Range National Park, New 

South Wales 

Rodney H. Jones''^ and Jeremy J. Bruhl' 

'Botany, Centre for Ecology, Evolution and Systematics, The University of New England, Armidale, NSW 

2351 (jbruhl@une.edu.au), - current address: Department of Primary Industries, Primary Industries Research 

Victoria, Knoxfield, Private Bag 15, Femtree Gully Delivery Centre, Victoria 3156. 



Jones, R.H. and Bruhl, J. J. (2006). Acacia beadleana (Fabaceae: Mimosoideae), a new, rare, localised 
species from Gibraltar Range National Park, New South Wales. Proceedings of the Linnean Society of 
New South Wales 111, 5-10. 

A new, rare species of phyllodinous Acacia from granitic areas of the Gibraltar Range in northern New 
South Wales is described on the basis of phenetic analysis. Comparison of A. beadleana with other 
morphologically similar species, and notes on its biology and ecology are presented. Conservation status 
for .4. beadleana is proposed. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: Acacia, rarity, resprouting shrub, taxonomy. 



INTRODUCTION 

During a separate study (Quirin et al. 1995), 
two specimens of an Acacia housed in the N.C.W. 
Beadle Herbarium (NE) that had been determined 
variously as Acacia mppii Maiden & Betche, A. 
torringtonensis Tindale and A. brunioides A.Cunn. 
ex G.Don were recognised as not belonging to any 
of these species. Although clearly belonging to A. 
subgen. Phyllodineae sect. Phyllodineae, clarification 
of the identity of these specimens could not be 
achieved using currently published descriptions at the 
time (Pedley 1983; Morrison and Davies 1991), and 
it was therefore tentatively assigned the phrase name 
Acacia sp. nov. (Gibraltar Range). Information from 
morphology and published descriptions, supported 
by advice from Acacia specialists (B. Maslin pers. 
comm.; L. Pedley pers. comm.), suggested that these 
specimens and others collected from the original 
population within Gibraltar Range National Park had 
affinities with A. brunioides, A. conferta A.Curm. ex 
Benth., A. gordonii (Tindale) Pedley, A. mppii, A. 
tindaleae Pedley, and A. torringtonensis. 

Subsequent investigation of the taxonomy 
of these species revealed conflicting classifications. 
A multivariate analysis (Jones 1997; Jones and Bruhl 



in prep.) was undertaken to test and set species limits 
of Acacia sp. nov. (Gibraltar Range) and the others 
of the study group above. Our plan was to publish 
the description of this new species together with the 
supporting analysis (Jones and Bruhl in prep.), but 
given that this new species is endemic to Gibraltar 
Range National Park we accepted the^ invitation to 
formally describe it in this special issue that celebrates 
the biodiversity of the region. 



MATERIALS AND METHODS 

Herbarium specimens from BRI, CANB, NE 
and NSW were examined, but only NE was found to 
have specimens of Acacia sp. nov. (Gibraltar Range). 
Fieldtripswere undertaken in Gibraltar RangeNational 
Park to expand the sample of morphological features, 
permit observation of the habit and habitat of the 
species, determine the extent of the known populations 
and search for new populations. Terminology for 
indumentum features follows Hewson (1988), and for 
other features Radford et al. (1974). Herbarium codes 
follow the current online version of Holmgren et al. 
( 1 990) Ihttp://207. 1 56.243 .8/emu/ih/index.php]. 

The number of flowers per head is a useful 
character for distinguishing ^coc/a sp. nov. (Gibraltar 



ACACIA BEADLEANA, Al^EW Al^D RARE SPECIES 



Range) from morphologically similar species. Precise 
counts are necessary as estimates can easily lead 
to spurious counts. Flower number per head was, 
therefore, checked either by marking individual 
flowers with a pen to avoid counting flowers more 
than once, or by removal of all flowers from a head 
and counting the number in a Petri dish viewed under 
a dissecting microscope. 



TAXONOMY 

Acacia beadleana R.H. Jones & J.J.Bruhl, sp. nov. 
Ad A. gordonii (Tindale) Pedley similaris, a qua 
phyllodiis in sectione transversali oblongis, trichomis 
ad marginem abaxialem phyllodii limitatis, petalis 
piliferis, et floribus per capitulo numerosioris, 
differt. 

Typus: New South Wales: Northern Tablelands: 
Gibraltar Range National Park, Gwydir Highway 
[precise locality withheld for conservation purposes], 
J J. Bruhl 1584, 28 Jan. 1996 (holo.: NSW; iso.: BRI, 
CANE, HO, K, MEL, MO, NE, PERTH, PRE). Figs 
1-2. 

Description: Single to multi-stemmed, lignotuberous, 
erect to spreading evergreen shrub, 0.4—2.5 m high. 
Stems woody, terete, roughened by phyllode scars. 
Branchlets terete with persistent, densely pilose 
indumentum; trichomes simple, hyaline appearing 
silver to white, anfrorse to refrorse. Stipules 
subpersistent, narrowly triangular to triangular, 
0.4-1 mm long, hairy. Pulvinus 0.5-1 mm long, 
sparsely hairy or sometimes glabrous. Phyllodes 
alternate and spiralled, crowded along the branchlets; 
narrowly elliptic, elliptic, linear to broadly linear, 
narrowly oblong, or narrowly oblanceolate 5-12.7 
mm long, 0.6-1.4 mm wide, sfraight or recurved, 
often irregularly furrowed when dried; cross-section 
narrowly oblong to oblong; sparsely pilose; the 
hairs mostly restricted to abaxial margin, divergent, 
sometimes curved, antrorse to subappressed, hyaline 
and appearing silver to white; base cuneate; apex acute 
to short-acuminate and mucronate, mucro straight to 
oblique or hooked; two main veins (separating at 
proximal end of phyllode; one more or less central 
and the other closer to the abaxial edge) observed 
in cleared and stained phyllodes, nerves obscure in 
dried material; extrafloral nectary usually only one 
present, occasionally on the pulvinus or more often 
less than 2 mm distal to the pulvinus; stomata flush 
with phyllode surface, sometimes slightly raised. 



Inflorescence solitary, axillary; peduncles densely 
pilose, 5.8-15.5 mm long, proximally ebracteate; 
flower heads globular, bright golden-yellow, 32^6 
flowered, 7-10 mm diameter when dried; bracteoles 
hairy; sepals, more than two thirds united from the 
base, hairy; petals sparsely hairy. Pods oblong; 20- 
60 mm long, 7-10.4 mm wide, glabrous, pruinose 
and purplish red when young, maturing to very 
dark brown outside and mid-tan inside, coriaceous, 
sfraight. Seeds of fransverse orientation in pod; obloid 
or ovoid, 3.8-5 mm long, 2.5-3.5 mm wide; black 
to very dark brown; areole usually open, sometimes 
closed; aril extending to more than half the length of 
seed. 

Selected specimens examined: New South Wales: 
Northern Tablelands: Gibraltar Range National 
Park: Anvil Rock Track [precise locality withheld 
for conservation purposes]: J.J. Bruhl 1759, J.B. 
Williams & R.H. Jones (BRI, CANB, DNA, L, NE, 
NSW, P, UPS, WAIK), T Tame, 4992 (NE, NSW); 
Dandahra Crags Track [precise locality withheld for 
conservation purposes]: J.J. Bruhl 1757, J.B. WiUiams 
& R.H. Jones (AD, BRI, CANB, CHR, MEL, NE, 
NSW, NY), J.J. Bruhl 1758a, J.B. Williams & R.H. 
Jones (BOL, CANB, EIU, MO, NE, SI, TENN); 
Gwydir Highway [precise locality withheld for 
conservation purposes]: J.J. Bruhl 1508, F.C. Quinn 
& J.B. Williams (BRI, CANB, NE, NSW). 

Similar species: Acacia beadleana is most similar in 
habit, phyllode morphology, inflorescence structure 
and flower colour to A. gordonii, a species that 
grows on sandstone and is resfricted to the lower 
Blue Mountains (Bilpin, Faulconbridge) and the 
Sydney Hills (Glenorie), more than 450 km south 
of the Gibraltar Range. Apart from its geographical 
separation, A. beadleana is most readily distinguished 
from A. gordonii by the distribution of phyllode and 
sepal indumentum and the number of flowers per head 
(Table 1). The most similar, proximal species to A. 
beadleana is A. brunioides A.Cunn. ex G.Don subsp. 
brunioides. The latter is also native to Gibraltar Range 
National Park, but populations are separated by c. 6 
km. These two species are readily morphologically 
distinguishable (Table 1) as are the broadly similar 
but more distantly located taxa A. brunioides subsp. 
granitica and^. conferta (Jones 1997; Maslin 2001). 

Figure 1 (right). Isotype oi Acacia beadleana R.H. 
Jones & J.J.Bruhl, J.J. Bruhl 1548 (NE). Pre- 
cise locality withheld for conservation purposes. 



Proc. Linn. Soc. N.S.W., 127, 2006 



R.H. JONES AND J.J. BRUHL 




I 







N.C.VV. Bcadic Ucrhsrium (Nt;) 
liHivcrsily f>f New iins'^"*^ 

ISOTYPE 



"" J.J^'Al 2?J</'N«2ooS" 



N.C.W. Beadle Herbarium (NE) 

The University of New England 

Armidalo NSW 2351 Australia 

N:fit".:,iiion o( ctlaflgo of ooletminatior^ ivoulcl be appfecinted by NB 

NE 85360 

FabacGaR subfam. Mimosoideae ^ 

Acacia sfi (Gibraltar Range) 

Australia. New South Wales: Norlhero Tabl eland s: 
Gibraltar Range National ParkF 




derate rocky slope, mid-slopo ! , ^-^eietal 

sandy loam on granite between boijioe.-i. jnu m rock crevices. 
Patchy Eucalyptus williamsiana layered open woodland with 
Leplospermum Irinerviuni CaWtiis munticola, AHocasuahna 
rigida. Acacia sp, nov,, A. baeuerlenii, Boronia anethifolia, 
Leucopogon neo-anghca, Mitbelia speciosa, Calytrix 
tetragona. Isopogon petiolans. Lepidosperma gunnii 
L v^scidum, Caustis flexuosa, Schoenus turbinates, 
Conospermiim hurgessioruni, Trachymene incisa. 

Common at site, localised (c. 120 plants seen). Shrubs 
to 2 X 2 m. Flowers golden yellow. 



Coll: J.J. Bruhl 1548 

Dei,; 



28 Jan. 1996 
Rep(s) to: BRI, CANS, HO, K, MEL, MO, NSW, PERTH. PRE 



Proc. Linn. Soc. N.S.W., 127, 2006 



7 



ACACIA BEADLEANA, Al^EW AND RARE SPECIES 




Figure 2. Acacia beadleana. A = densely pilose branchlet; stipules pilose; phyllodes mucro- 
nate, pilose along the abaxial margin; B = globular inflorescence at anthesis; flower buds 
hairy; C = fruits showing transversely oriented cavities that indicate the in situ orientation 
of the seeds; D = black seeds with fleshy/oily funicle forming an elaiosome. Scale bars A-B = 
1 mm; D = 1 mm; C = 10 mm. A, B = L.M. Copeland 3892 (NE); C, D = J.J. Bruhl 1548 (NE). 



Etymology: The specific epithet honours Professor 
Noel C.W. Beadle (1914-1998), foundation Professor 
of Botany at The University of New England, noted 
ecologist and taxonomist. 



Ecology: Plants of Acacia beadleana grow in 
skeletal to deep sandy soils on granite in layered 
eucalypt woodland and heath. The type locality is 
heterogeneous in topography and aspect due to the 



8 



Proc. Linn. Soc. N.S.W., 127, 2006 



R.H. JONES AND J.J. BRUHL 



Table 1. Distinguishing morpliological features of Acacia beadleana,A. brunioides and /I. gordonii 

* Additional observations provided by P. Kodela (NSW) 



Cliaracter/Taxon 


A. beadleana 


A. brunioides subsp. 
brunioides 


A. gordonii 


Branchlet hair density 


Dense 


Absent, isolated or 
sparse 


Dense 


Phyllode base 


Cuneate 


Obtuse 


Cuneate to obtuse 


Phyllode indumentum 


Abaxial margin only 


Absent 


Over whole phyllode 


Pulvinus indumentum 


Usually present 


Absent 


Present 


Petal indumentum 


Present 


Absent 


Absent 


Sepal indumentum 


Present 


Sparse or absent 


Sparse or absent 


Flowers per head 


32^6 


21-26 


(12-)21-25(-34)* 


Flower colour 


Bright golden yellow 


Pale creamy yellow 


Bright golden yellow 



outcropping granite. Consequently the vegetation is 
also heterogeneous: patchy Eucalyptus williamsiana 
layered open woodland and heath with Leptospermum 
trinervium, Allocasuarina rigida, Callitris monticola. 
Acacia beadleana,A. baeuerlenii, Boronia anethifolia, 
Mirbelia speciosa, Leucopogon neo-anglica, Calytrix 
tetragona, Isopogon petiolaris, Lepidosperma gunnii, 
L. viscidum, Caustis flexuosa, Schoenus turbinatus, 
Conospermum burgessiorum and Trachymene incisa. 
Another population occurs on the lower slope of a 
broad, shallow valley on deeper soils in a eucalypt- 
layered woodland close to a swamp. 

Biology: Most plants appear to be single-stemmed, 
while some are clearly multistemmed. A lignotuber 
at about ground level is often apparent. We have 
observed plants resprouting after most main branches 
had died due either to senescence or drought. Plants 
on granite outcrops were also observed to resprout 
within months of the major fire of 2002 in GRNP 
(P.J. Clarke pers. comm.) and such fired, resprouting 
individuals were observed (by JJB) to be growing 
well in June 2005. 

Plants, especially those in the 'Gwydir 
Highway' population, appear generally to be 
parasitised by a scale or related hemipteran and 
consequently laden with sooty mould, especially 
along the stem. 

Flowering and fruiting phenology: Plants of Acacia 
beadleana have been observed to flower in all 



seasons of the year. Examination of herbarium 
material indicates that the main flush of buds occurs 
around November, and these buds are well developed 
by December- January. Flowering peaks in January- 
February. Abundant, young, immature finit is evident 
by July-August. While some mature fruit is probably 
held on the plants for months after seed drop, the 
collection with the most mature fruit containing seed 
in situ was the type collection of late January. 

Distribution and conservation status: 
Evidence from our study (Jones 1997; Jones and 
Bruhl in prep.) indicates that Acacia beadleana is 
rare and geographically restricted. It is only known 
from three discrete populations within Gibraltar 
Range National Park. Each population is composed 
of c. 100 plants. One population is bisected by the 
Gwydir Highway, so roadside maintenance and any 
plan to widen or alter the road or extend the verge 
in that vicinity is likely to impact the population and 
should be actively discouraged. Three populations 
with a total of fewer than 1000 plants occur within the 
National Park, therefore a ROTAP code (Briggs and 
Leigh 1996) of 2VCit is suggested for ^. beadleana. 
The population biology of A. beadleana merits close 
study. We predict that most likely range extensions 
are in the more inaccessible escarpment areas of 
Gibraltar Range National Park. 



Proc. Linn. Soc. N.S.W., 127, 2006 



ACACIA BEADLEANA, A^EW A^D RARE SPECIES 



ACKNOWLEDGMENTS 



Financial support to RHJ from the Noel C.W. 
Beadle Scholarship in Botany and the Keith and Dorothy 
Mackay Scholarship (Honours) is gratefully acknowledged. 
Thanks go to Carolne Gross, Frances Quinn, Warren 
and Gloria Sheather, and John Williams (all UNE) for 
field assistance, and staff and students of Botany, UNE, 
for support and advice to RHJ as an honours student. 
Thanks go to Bruce Maslin (CALM, WA) for personal 
communication and access to unpublished material; Terry 
Tame for discussions and specimens; Les Pedley (BRJ) 
and Peter Clarke (UNE) and Philip Kodela (NSW) for 
personal communications; directors of herbaria BRJ, 
CANB (including CBG) and NSW for loan material; and 
access to the N.C.W Beadle Herbarium and facilities is 
acknowledged. We thank National Parks and Wildlife 
Service of NSW for permits to collect, and access to the 
park. Thanks also to Alex George (as Australian Botanical 
Liaison Officer) for Latin diagnosis and comments; Ian R.H. 
Telford (NE) for advice during the project and comments 
on the manuscript; and Lachlan Copeland (UNE) and the 
two referees for comments on the manuscript. 



Radford, A.E., Dickison, W.C, Massey, J.R. and Bell, 

C.R. (1974). Vascular Plant Systematics. (New 
York, Harper & Row). 



REFERENCES 

Briggs, J.D. and Leigh, J. (1996). Rare or threatened 

Australian plants (Melbourne, CSIRO). 
Hewson, H.J. (1988). Plant Indumentum: A handbook of 

terminology. Volume 9 (Canberra, Australian 

Government Publishing Service). 
Hohngren, RK., Holmgren, N.H. and Bamett, L.C. (1990). 

Index Herbariorum. Part I: The Herbaria of the 

World, 8th Edition (Bronx, New York Botanical 

Garden). 
Jones, R.H. (1997). Systematic studies in Acacia subgenus 

Phyllodineae (Fabaceae: Mimosoideae). 

Honours thesis. The University of New England, 

Armidale. 
Jones, R.H. and Bruhl, J.J. (in prep.). Species limits within 

Acacia subgenus Phyllodineae (Fabaceae: 

Mimosoideae): A case for analytical assessment. 
Maslin, B. ed. (2001). WATTLE: Acacias of Australia 

(Perth, ABRS and CALM). 
Morrison, D.A. and Davies, S.J. (1991). Acacia. In Flora 

of New South Wales, vol. 2, 1st edn (Ed. G.J. 

Harden) pp. 327-392. (Sydney, UNSW Press). 
Pedley, L. (1983). Mimosaceae. In Flora of South-eastern 

Queensland (Eds TD. Stanley and E.M. Ross) 

pp. 332-386. (Brisbane, Queensland Department 

of Primary Industries). 
Quinn, F.C., Williams, J.B., Gross, C.L. and Bruhl, J.J. 

( 1 995). Report on rare and threatened plants of 

north-eastern New South Wales, p. 292. Report 

prepared for National Parks and Wildlife Service 

of New South Wales and Australian Nature 

Conservation Agency. 



10 



Proc. Linn. Soc. N.S.W., 127, 2006 



Population Structure and Fecundity in the Putative Sterile 
Shrub, Grevillea rhizomatosa Olde & Marriott (Proteaceae) 

H.A.R. Caddy & C.L. Gross 

Ecosystem Management, School of Environmental Sciences and Natural Resources Mangement, The 
University of New England, Armidale, NSW, 2351 (cgross@une.edu.au). 

Caddy, H.A.R. and Gross, C.L. (2006). Population structure and fecundity in the putative sterile shrub, 
Grevillea rhizomatosa Olde & Marriott (Proteaceae). Proceedings of the Linnean Society of New South 
Walesni,n-U. 

Grevillea rhizomatosa Olde & Marriott (Proteaceae) is a threatened species of shrub known only from 12 
populations within a 7 x 8 km area within Gibraltar Range and Washpool National Parks, northern New 
South Wales, Australia. Prior to this study it was believed that the species only reproduced from rhizomatous 
suckers as seed and fruit were never detected in the wild. A concern for the reproductive and evolutionary 
potential of the species in the event of a catastrophic disturbance was the basis for an investigation into the 
reproductive ecology of G. rhizomatosa. Such an event occurred in October 2002 with an intense wildfire 
affecting most of the populations. Five populations were studied in detail for demography and fecundity 
prior to this fire and two populations were resurveyed in August 2005. In 2000, 916 individual stems were 
recorded across these populations and only small to large shrubs were found; no seedlings were recorded. 
Post-fire response was documented in two populations where plants were found to be resprouting and 
suckering fi^om underground stems. In the pre-fire surveys of 2000 and 2001 flowering occurred in all 
populations, but since the fire of October 2002 flowering has only occurred in unbumt habitats. Flowers on 
shrubs in two of the five populations failed to produce fruit, but low fruit-set (7-13% of flowers) occurred in 
three populations. Seeds collected from two populations (n = 14) were tested for viability using tetrazolium 
chloride and were 100% viable. Ramets were detected in all populations and resprouting from underground 
stems was observed after wildfire. This is the first record of viable seed in this species and fertile populations 
require specific management to prevent loss of fertile plants. Loss of fertile plants could occur if repeated 
burning selects for vegetative reproduction and sterile plants. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: clonality, fire response, Grevillea rhizomatosa, population structure, rarity. 



INTRODUCTION 

Grevillea is one of the most successfully 
dispersed groups vi'ithin the Proteaceae. An estimated 
357 species occur variously in temperate, arid and 
tropical ecosystems throughout Australia with 
species also found in New Caledonia (3 endemic 
species), New Guinea (3 species, 1 endemic) and 
Sulawesi (1 endemic species) (Makinson 2000). 
About 14% of species have the capacity to reproduce 
asexually through vegetative reproduction, although 
the majority of these species combine both asexual 
and sexual reproduction (Makinson 2000; Makinson, 
unpub. data). An exception to this may be Grevillea 
rhizomatosa, which is described, by Olde and Marriot 
(1994) as sterile and an obligate clonal species with 
ramets produced from stem suckering. Grevillea 
rhizomatosa is restricted to Washpool and Gibraltar 



Range National Parks and is listed as a vulnerable 
species at both a State and Federal level {Threatened 
Species Conservation Act 1995 (NSW), Environment 
Protection and Biodiversity Conservation Act 1999 
(Commonwealth)) . 

The occurrence of clonality in rare and 
threatened plants can complicate the conservation of 
such species (e.g. Sydes and Peakall 1998) because 
ramet reproduction may cause population sizes 
to be overestimated (Ellstrand and Roose 1987). 
Moreover much of the genetic variation may exist 
among populations rather than within, thereby 
requiring all populations to be actively conserved. 
The clonal syndrome may be disadvantageous to 
species if low to nil genetic diversity is combined 
with sterility (e.g. Lomatia tasmanica. Lynch et al. 
1998). This can make such species highly susceptible 
to extirpation, as all individuals in the population are 



FECUNDITY IN A THREATENED SPECIES OF GREVILLEA 



likely to respond in a uniform fashion to a stochastic 
event (e.g. disease, fire). In addition, the chances of 
extirpation are exacerbated if populations are small 
and fragmented but simultaneously impacted upon by 
major disturbance events such as wildfire. 

The absence or poor seed production in 
populations can have many causes that may include 
ecological deficiencies (e.g. pollinator and/or 
pollen limitation; finait predation, e.g. Hampe 2005; 
Vesprini and Galetto 2000) nutrient shortages (e.g. 
Drenovsky and Richards 2005) and innate sterility 
mechanisms (Pandit and Babu 2003). As a first stage 
to understanding fecundity in Grevillea rhizomatosa, 
we investigated the distribution of populations 
with special reference to their fertility and post-fire 
response. 



as seedlings (< 10 cm height, with no obvious 
rhizomatous connections) or as a combined class 
called juvenile/adult. A plant was scored as an 
individual if growing as a single stem or as a multi- 
stem plant (on the proviso that the multi-stems 
were grouped within a 5 cm basal diameter). Plants 
with more than 5 cm between them were classed as 
separate individuals although it is possible that they 
exist as ramets. In each population a maximum of 1 00 
individuals was examined and in populations of less 
than 100 individuals, all were measured. 

2005 

Plant response to the October 2002 fire was 
scored in Wash and Dand in August 2005. In each 
population 25-50 individuals were measured for 
height and autonomy (seedling or resprouter/sucker). 



METHODS 

Species distribution 

Specimen data for Grevillea rhizomatosa 
from the New England Herbarium and flora survey 
data fi-om the NSW-NPWS were collated and mapped 
resulting in 22 potential populations. Seventeen of 
these locations were relocated in the field and visited 
during August and November 2000. This preliminary 
work showed that the species is found in at least 12 
populations in an area 8 km x 7 km in northern NSW 
(Fig. 1). 

Population and habitat description 

Five populations spanning the range of 
Grevillea rhizomatosa were chosen for fiirther study 
(Fig. 1). Sites were selected based on the parameters 
of population size (at least 25 individuals) and range 
so that reproductive outputs could be compared 
across the extremes of the species' distribution. One 
study site occurred in Washpool National Park and 
the remainder in Gibraltar Range National Park. 
The selected populations were Washpool National 
Park (Wash), Mulligan's Hut (MHut), Dandahra 
Trail (Dand), M'^Climonts Swamp (Swamp), and 
Murrumbooee Cascades (Cascade) (Fig. 1). Fieldwork 
was conducted between August and November 2000 
at Cascade, MHut and Dand, with the majority of 
work undertaken (in all populations) between March 
and September 2001 with follow up work in June to 
August 2005. 

Demography 

2000-2001 

Plants do not flower every year and ramets 
occur in all populations so plants were scored 



Fecundity 

To quantify the extent of sexual reproduction 
in populations, flowers on each of 10 plants were 
tagged in Cas, Dand, MHut and Wash over two 
flowering seasons (1999-2000 and 2000-2001, Dand 
and Cascade; and 2000-2001 MHut, Wash). Flowers 
open to all pollinators were tagged and then monitored 
for finait-set over the 2000 and 2001 flowering season 
(see Table 1 for sample sizes). Bags were placed 
over developing fi-uits to reduce fi-uit-loss. Data 
were pooled across seasons. The Swamp population 
was not used for fecundity experiments because of 
time restrictions, but plants in this population were 
extensively searched for fruit production over the two 
flowering seasons. 

Seed viability 

Seeds were encountered only infrequently 
during fieldwork (see below). Seven seeds were 
obtained from each of three individuals at Cas and 
Wash. We checked the viability of seeds using a 48- 
hour soaking solution of a 1.0% solution of 2,3,5- 
triphenyl-tetrazolium chloride (Scott and Gross 
2004). Seven of these seeds were killed by boiling 
and used as a control. Seed were dissected and scored 
as viable if they stained bright pink. Non- viable seeds 
do not stain (Lakon 1949). 



RESULTS 

Population and habitat description 

Grevillea rhizomatosa was only found 
growing on low-nutrient lithosols derived fi-om 
Dandahra Granite complex within Gibraltar Range and 
Washpool National Parks. 



12 



Proc. Linn. Soc. N.S.W., 127, 2006 



H.A.R. CADDY AND C.L. GROSS 




Figure 1. Distribution of Grevillea rhizomatosa plants in Gibraltar Range 
and Washpool National Parks. 



Washpool National Park (Wash) 

The northern most population of Grevillea 
rhizomatosa is located in the Washpool National Park 
(29° 28' 03"S, 152° 18' 18"E) at 870 m ASL. The 
topography is mid slope with a north-eastern aspect. 
The soil substrate is a deep to shallow sandy loam 
derived from leucogranite granite. The vegetation is 
tall open forest dominated by Eucalyptus campanulata 
and E. cameronii. Associated species include Banksia 
integrifolia subsp. monticola, Pultenaea sp. B, 
Acacia nova-anglica. Grevillea rhizomatosa grows 
in linear strips along the North and South sides of 



Moogem Road; at least 200 individuals grow south 
of the road and at least 25 scattered individuals occur 
to the north between the road and the Dandahra 
Gully. Fire records at 1 July 2002 show this area to 
the north of Moogem Road had not been burnt since 
1968, whereas south of Moogem Road was burnt in 
1988. The population on the southern side of the road 
was also extensively burnt in October 2002 and dense 
resprouting was observed in July 2005. 
Mulligan's Hut Camping Area TMHut) 

MHut is located 200 m north east 
of Mulligan's Hut along the world heritage 



Proc. Linn. Soc. N.S.W., 127, 2006 



13 



FECUNDITY IN A THREATENED SPECIES OF GREVILLEA 



Table 1. Population locations, demography and fertility for Grevillea rhizoma- 
tosa from Washpool and Gibraltar Range National Parks. 



Site 


Latitude 

longitude 

altitude 

29° 28' 03"S 


Number of 
plants 
(% seedlings) 


% fruit production 
(number of flowers 
treated over 10 plants) 


% seed viability 
(n = number of 
seed treated) 


Wash 


152° 18' 18"E 
870mASL 

29°31'00"S 


c. 225 (0) 


10.19(206) 


100% (7) 


MHut 


152°2r39"E 
910mASL 

29°31'42"S 


c. 250 (0) 


0(62) 




Dand 


152°20'30"E 
980mASL 


c. 250 (0) 


7.08 (367) 


' 


Swamp 


29°31'58"S 

152°20'27"E 

960mASL 

29° 32' 37"S 


165 (0) 


not quantified but none 
observed from 2000-2005 


- 


Cas 


152°2r29"E 
830mASL 


41(0) 


13.35 (337) 


100% (7) 



walking track (Table 1). Two hundred and fifty 
individuals of Grevillea rhizomatosa were found in 
this area. The topography is mid-slope with a south- 
western aspect. Shallow to skeletal sandy granitic 
soils occur at the site, with most plants growing 
between granite boulders. The vegetation is an open 
woodland with a dense shrub layer; the dominant 
tree species associated with this community include 
Eucalyptus olida, E. pyrocarpa, and E. planchoniana. 
Dominant understorey species include Leptospermum 
trinervium, Pultenaea sp. B, Persoonia nifa, Banskia 
spinulosa. Ground cover species include Platysace 
ericoides, Caustis flexuosa, Bossiaea scortechinii, 
Xanthorrhoea johnsonii, and Lomandra longifolia. 
The NPWS database indicates the area was burnt in 
1964 and possibly in 1988. The population was not 
burnt in the October 2002 fires. 

Dandahra Trail (Dand) 

Grevillea rhizomatosa grows on both sides 
of the Dandahra Trail into Mulligan's Hut (Table 1). 
Most plants (200 stems) grow south of the Dandahra 
trail, with only some 50 individuals growing to the 



north. Shallow sandy granitic soils occur at the site. 
Some plants grow between granite boulders. Dominant 
tree species include E. olida and E. cameronii. 
Common shrubs are Pultenaea sp. B, Persoonia rufa, 
and Acacia obtusifolia. Groundcover species include 
Platysace ericoides, Caustis flexuosa, and Bossiaea 
scortechinii. NPWS fire history for the area shows 
fire in 1964 and 1988. The area was intensively burnt 
in October 2002. 

M ^Climonts Swamp (Swamp) 

The Swamp population is located 
approximately 500 m down-slope fi-om Dand (Table 
1). A population of 165 individuals occurs in linear 
strips adjacent to the road. Soil substrate, vegetation, 
and fire history are similar to those described for 
Dand. 
MvuTumbooee Cascades (Cascade) 

Cascade is the southernmost G. rhizomatosa 
population detected in this study (Table 1). A small 
population of 41 individuals occurs on north and 
south ridges dissected by a drainage line. Soils are 



14 



Proc. Linn. Soc. N.S.W., 127, 2006 



H.A.R. CADDY AND C.L. GROSS 



shallow to skeletal and of granitic derivation, as 
described previously. Eucalyptus radiata subsp. 
sejuncta is present along the creek, with E. olida and 
E. cameronii on the ridges. Dominant shrubs include 
Leptospermum trinervium, Dillwynia phylicoides, 
and Hakea laevipes subsp. graniticola. The area was 
burnt in 1964 and 1988. The October 2002 fires burnt 
the northern half of this population. 

Demography 

2000-2001 

No seedlings were detected during the study. 
All plants were greater than 10 cm in height and most 
(c. 80%) appeared to be connected to nearby plants, 
as evidenced by plants growing in lines from larger 
plants and as confirmed from occasional excavations 
(Figure 2a-c). At MHut plants are large (0.5-1.20 m 
tall X c. 0.5-1.40 m wide) and many are connected 
underground by their stems. Large granite boulders 
partition this population into well-defined clumps. 
Flowering occurred in all populations in all years 
although not all plants flowered every year. 



2005 

No seedlings were found in the fire- 
recovering communities of Wash and Dand. The 
mean plant height of Grevillea rhizomatosa in the 
burnt habitat at Wash was 48.05+ 3.81 cm (n = 43), 
which was considerably shorter than the few plants 
that escaped the fire on the northern side of the road 
(mean height = 108.89 ± 23.99 cm, n=7). The unbumt 
plants flowered in 2004 and 2005, whereas the burnt 
plants did not. At Dand the recovering population had 
a mean height of 47.64 + 2.58 cm (n=51) in August 
2005. In 2004 and 2005 flowering was only detected 
on unbumt individuals at Wash and in the unbumt 
population of MHut. 

Fecundity 

Fmits were only detected in Wash, Dand 
and Cas (Table 1, Figure 3a, 2b). Flowers have two 
ovules, but fmits mainly contained one seed. Fmit 
was recorded on each of the 10 survey plants in 
Wash and Cas and on eight of the 10 survey plants in 




Figure 2. (a) Subterranean reprouting from a plant in Wash August 2005, (b). rhizomatous connections 
between small plants at Wash August 2005, (c) rhizomatous growth in G. rhizomatosa (scale bar = 100 
mm). 



Proc. Linn. Soc. N.S.W., 127, 2006 



15 



FECUNDITY IN A THREATENED SPECIES OF GREVILLEA 





MlffliJJJltJIlJJJIllJ 





B 




tHQt'K-2 



^ ^|HusrL£-ss~| 



'i : ' u^'"'* 'ii*^''^ 



iMMi: II IM 




Figure 3. (a) & (b) fruits of G. rhizomatosa (scale bars = 10 mm), viable (c) and inviable (d) seed of 
Grevillea rhizomatosa from Wash and Cas populations. Scale bar = 10mm. 



Dand. Although not quantified, there were, however, 
many plants that did not produce fiiiit in the fertile 
populations. No seed was produced fi-om tagged 
flowers at MHut and no fiaiit were ever found in any 
season at Swamp during cursory observations. 

Seed Viability 

Seed collected from Wash and Cas (n = 14) 
and treated chemically with tetrazolium were 100% 
viable (Figvires 3c, 3d) and controls were unviable (n 

= 7). 

DISCUSSION 

This is the first time that seed has been 
found on individuals of Grevillea rhizomatosa. 
Prior to our work the species was thought to be 
sterile and obligately clonal (Olde and Marriott 
1994; Makinson 2000). Within Gibraltar Range and 
Washpool National Parks all five study populations of 
Grevillea rhizomatosa contained clonal individuals 
with all plants in two populations failing to produce 
fi:iiit on any flowering plant. Seedlings were never 



encountered, even after fire had burnt populations 
containing fiaiit-bearing plants. These three fertile 
populations (Wash, Dand and Cascades) are widely 
separated and thus valuable for the conservation of 
the species. Natural fi-uit-set was low (< 0.14 fruit 
to flower ratio) but higher than that foimd in other 
species of Grevillea (e.g. 0.015-0.096 fruit to flower 
ratio at maturation, Hermanutz et al. 1998). 

Not all individuals flowered every year and 
after the hot fires of October 2002 flowers have not 
been initiated on recovering individuals in Wash, 
Dand, Swamp or Cascade as of August 2005. Instead 
the species in these populations has recolonised areas 
by resprouting fi-om stem bases (Fig. 2a) and from 
the advent of new suckers (Fig. 2b and 2c). It may 
be that the release from flowering allows resources 
to be redirected for vegetative reproductions and that 
clonality is selectively favoured in this irregular flower 
producer. This has major ramifications for the genetic 
stucture of populations such that near neighbours 
are likely to be genetically identical, which in turn 
may promote inbreeding when flowering occurs in 
populations. 



16 



Proc. Linn. Soc. N.S.W., 127, 2006 



H.A.R. CADDY AND C.L. GROSS 



Where clonality coexists with sexual forms 
it may provide populations with a flexible response 
to variable habitat or resource abundance and allow 
the transfer of resources among ramets. In habitats 
where large resource-reserves are required to initiate 
new growth, clonality may provide a more secure 
investment than seed-set alone. Clonal plants with 
pronounced vegetative reproduction can have lower 
rates of local extinction in nutrient-poor ecosystems 
than plants without pronoimced vegetative 
reproduction (Fischer and Stocklin 1997). Indeed 
the correlation that clonal plants are often found on 
nutrient-poor soils (see Fischer and van Kleunen 
2002) may, in part, explain why Australia, with 
nutrient-poor soils, seems to have so many threatened 
species that are clonal (Gross, unpub. data). 

The regenerative capacity of clonal growth 
also affords ramets increased longevity. Tyson et al. 
(1998) for example, foimd a clonal mallee eucalypt 
to be at least 900 years old, much older than the usual 
age of single stemmed eucalypts. Moreover, Smith 
et al. (2003) estimate from radial growth rates in 
Eucalyptus curtisii that some clones may be between 
4000 and 9000 years old. The population at MHut 
is comprised of at least 250 large, sterile shrubs 
that are nestled among granite boulders. Within this 
population the lateral spread of plants is restricted 
by boulders encircling clumps, suggesting that plant 
clumps may not be of recent origin. 

Management of Grevillea rhizomatosa 
should especially focus on the fertile populations of 
Wash, Dand and Cas. Of concern is the promotion 
of suckering in post-fire habitats, where plants can 
form thickets. If this is combined with sterility then 
seedling establishment of fertile individuals may 
be disadvantaged. Our work has shovm that plants 
do not flower in the first three seasons post-fire and 
thus further observations are required to determine 
the optimal fire interval. In addition, the reasons 
for an absence of fiiiit-set in some individuals of 
Grevillea rhizomatosa and the genetic composition 
of populations are important components to unravel 
for the conservation of the species (e.g. Grevillea 
infecunda, Kimpton, James and Drinnan 2002). 
Work is underway in these areas and will be reported 
elsewhere. 



ACKNOWLEDGEMENTS 

The 2000 and 200 1 work was undertaken by HAR Caddy as 
part of an Honours dissertation. The project was supported 
by fiinds to C.L. Gross from NSW National Parks and 
Wildlife Service, Glen Innes District and by the University 



of New England. Peter Croft is thanked for suggesting 
the project, arranging funding and for providing access 
to flora and tire records. Many thanks to Anna Coventry, 
Bruce Tailor and David Mackay for field assistance. The 
Director of the New England Herbarium is thanked for 
access to specimens and records. Bob Makinson (RBG- 
Sydney) is thanked for once again generously sharing 
his knowledge of Grevillea and for allowing us access 
to his database of rhizomatous species in Grevillea. 
This work was conducted under permit number NZ143. 



REFERENCES 

Drenovsky R.E. and Richards J.H. (2005). Nitrogen 
addition increases fecundity in the desert shrub 
Sarcobatus vermiculatus. Oecologia 143, 349-356. 

Ellstrand, N.C. and Roose, M.L. (1987). Patterns of 

genotypic diversity in clonal plant species. American 
Journal of Botany 74,123-131. 

Fischer, M. and Stocklin, J. (1997). Local extinctions of 
plants in remnants of extensively used calcareous 
grasslands 1950—85. Conservation Biology 11 (3), 
727-737. 

Fischer, M. and van Kleunen, M. (2002). On the evolution 
of clonal plant life histories. Evolutionary Ecology 
15, 565-582. 

Hampe, A. (2005). Fecundity limits in Frangula alnus 
(Rhamnaceae) relict populations at the species' 
southern range margin. Oecologia 143, 377-386. 

Hermanutz, L., Innes, D., Denham, A. and Whelan, R. 
(1998). Very low fhait:flower ratios in Grevillea 
(Proteaceae) are independent of breeding system. 
Australian Journal of Botany 46, 465-478. 

Kimpton, S.K., James, E.A. and Drinnan, A.N. (2002). 
Reproductive biology and genetic marker diversity 
in Grevillea infecunda (Proteaceae), a rare plant with 
no known seed production. Australian Systematic 
Botany 15, 485-492. 

Lakon, G. (1949). The topographical tetrazolium method 
determining the germination capacity of seeds. Plant 
Physiology 24, 389-394. 

Lynch, J.J., Barnes, R.W., Cambecedes, J. and 

Vaillancourt, R.E. (1998). Genetic evidence that 
Lomatia tasmanica (Proteaceae) is an ancient clone. 
Australian Journal of Botany 46, 25-33. 

Makinson, R.O. (2000). Grevillea. Flora of Australia 
17A, 1-460. 

Olde, P.M. and Mariott, N.R. (1994). A taxonomic revision 
of Grevillea arenaria and Grevillea cbtusifolia 
(Proteaceae: Grevilleoideae). TelopeaS, 711-733. 

Pandit, M.K. and Babu, C.R. (2003). The effects of loss of 
sex in clonal populations of an endangered perennial 
Coptis teeta (Ranunculaceae). Botanical Journal of 
the Linnean Society 143, 47-54. 

Scott, B. and Gross, C.L. (2000). Recovery directions 
for monoecious and endangered Bertya ingramii 
using autecology and comparisons with common B. 



Proc. Linn. Soc. N.S.W., 127, 2006 



17 



FECUNDITY IN A THREATENED SPECIES OF GREVILLEA 



rosmarinifolia (Euphorbiaceae). Biodiversity and 

Conservation. 13, 885-899. 
Smith, J.A. and Gross, C.L. (2002). The poUination 

ecology of Grevillea beadleana McGilhvray 

(Proteaceae), an endangered shrub from Northern 

NSW. Annals of Botany 89, 97-108. 
Sydes, M.A. and Peakall, R. (1998). Extensive clonality 

in the endangered shrub Haloragodendron lucasii 

(Haloragaceae) revealed by allozymes and RAPDSs. 

Molecular Ecology 7, 87-93. 
Tyson, M., Vaillancourt, R.E. and Reid, J.B. (1998). 

Determination of clone size and age in a malice 

eucalypt using RAPDs. Australian Journal of Botany 

46,161-172. 
Vesprini J.L. and Galetto L. (2000) The reproductive 

biology ofJaborosa integrifolia (Solanaceae): Why 

its fruits are so rare? Plant Systematics and Evolution 

225, 15-28. 



18 Proc. Linn. Soc. N.S.W., 127, 2006 



Selfed Seed Set and Inbreeding Depression in Obligate Seeding 

Populations of Banksia marginata 

Glenda Vaughton and Mike Ramsey 

Botany, School of Environmental Sciences and Natural Resources Management, University of New England, 

Armidale NSW 2351 (gvaughto@une.edu.au). 



Vaughton, G. and Ramsey, M. (2006). Selfed seed set and inbreeding depression in obligate seeding 
populations of Banksia marginata. Proceedings of the Linnean Society of New South Wales 127, 19-25. 

Self-compatible species can often produce seeds when pollinators are scarce or unreliable, but any 
advantage may be lessened if selfed progeny are less fit than outcrossed progeny due to inbreeding 
depression. We use hand self-pollinations to determine whether Banksia marginata is self-compatible and 
examine the relative fitness of seeds derived fi-om self- and open-pollination at several early life-cycle 
stages to gauge the likely impact of inbreeding depression. Substantial numbers of fruits and seeds were 
produced following selfing, indicating that plants are self-compatible. However, differences between self- 
and open-pollinated inflorescences indicated that relative self-fertility was less than one. Compared with 
open-pollinated seeds, selfed seeds were smaller and produced smaller seedlings that were less likely to 
survive. Percent germination of self- and open-pollinated seeds was similar Cumulative fitness estimated 
over several life-cycle stages, including seed production, indicated that selfed progeny were on average 
only 62% as fit as open-pollinated progeny. These differences in relative fitness indicate that despite 
self-compatibility, populations have experienced a history of outcrossing. Banksia marginata plants at 
Gibraltar Range National Park are killed by fire, and self-compatibility may be associated with this trait. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: plant breeding system, pollination, Proteaceae, self-compatibility, self-fertility. 



INTRODUCTION 

Reproductive assurance is thought to be a power- 
ful selective factor influencing the evolution of self- 
compatibility in plant populations. Self-compatible 
species do not require pollen from other plants in 
order to set seeds and can have an advantage when 
pollinators are scarce or uiureliable (Lloyd 1979, 
1992; Barrett 2003). A disadvantage of self-com- 
patibility, however, is that selfed progeny may be 
less fit than outcrossed progeny due to inbreeding 
depression (Charlesworth and Charlesworth 1987). 
One common cause of inbreeding depression is the 
expression of deleterious recessive alleles made ho- 
mozygous following selfing. Genetic load and the 
severity of inbreeding depression are expected to 
evolve with the mating system. Species with a his- 
tory of selfing often have low inbreeding depression 
because deleterious alleles have been purged from 
the gene pool. By contrast, genetic load is main- 
tained in species that are primarily outcrossing, and 
inbreeding depression can be severe such that the 



benefits of self-compatibility are substantially re- 
duced or even negated (Lande et al. 1994;Husband 
and Schemske 1996; Byers and Waller 1999). 
In their review of the breeding and mating sys- 
tems of the Australian Proteaceae, Goldingay and 
Carthew (1998) concluded that most Banksia spe- 
cies showed only low levels of self-compatibility 
and were highly outcrossing. Two exceptions were 
B. brownii, which is self- compatible and maintains 
a mixed mating system with selfing and outcross- 
ing (Sampson et al. 1994, Day et al. 1997), and B. 
spinulosa var. neoanglica, which is self-compatible 
but highly outcrossing (Vaughton 1988; Vaughton 
and Carthew 1993). Since this review, self-compat- 
ibility has been reported in other species of Banksia 
including, B. ericifolia var. macrantha (Hackett and 
Goldingay 2001), B. baxteri, B. media and B. nutans 
(Wooller and WooUer 2001, 2002, 2003). Self-com- 
patibility has also been demonstrated in B. ilicifolia, 
although firiit and seed set following selfing were 
much lower than following outcrossing (Heliyanto et 
al. 2005). In two of these species the relative fitness 



SELF-COMPATIBILITY IN B ANKSIA MARGINATA 



of selfed progeny was also examined. In B. baxteri, 
more selfed seeds aborted, but seed germination and 
seedling survival did not differ following self-pol- 
lination compared with natural pollination (Wooller 
and Wooller 2004). In B. ilicifolia, fewer selfed 
seeds germinated than crossed seeds, and survival 
of selfed seedlings was less when exposed to attack 
by a fungal pathogen (Heliyanto et al. 2005). Taken 
together, these results suggest that self-compatibility 
may be more common in Banksia than previously 
thought, and that in such species the relative fitness 
of selfed progeny warrants further investigation. 
Here we use hand self-pollinations to determine 
whether B. marginata plants occurring at Gibraltar 
Range National Park (GRNP) are self-compatible. 
We compare selfed seed set to that occurring naturally 
in populations and examine variation in the effect of 
pollination among years and sites. Finally, we assess 
the relative fitness of seeds derived from self- and 
open-pollination at several early life-cycle stages to 
gauge the likely impact of inbreeding depression. 



MATERIALS and METHODS 

Study species and sites 

Banksia marginata Cav. is widely distributed 
in south-eastern Australia and exhibits considerable 
variation in both its morphology and life history 
throughout its range (George 1998). At GRNP, B. 
marginata is killed by fire and relies on seeds for 
subsequent regeneration (i.e. plants are obligate 
seeders, Vaughton and Ramsey 1998; Virgona et 
al. 2006). Plants occur in sedge-heath in areas of 
impeded drainage on flats and hillsides (Virgona et 
al. 2006). Flowering occurs in late autumn and winter 
and plants produce multiple inflorescences with an 
average of 784 flowers (SE = 52.4, n = 20). Flowers 
open acropetally on inflorescences over 3-4 weeks 
(G. Vaughton unpublished data). Inflorescences are 
pollinated by nectarivorous honeyeaters, insects, 
including introduced honeybees, and probably 
mammals (see methods). Follicles are strongly 
serotinous and have up to two seeds (Vaughton 
and Ramsey 1998). Field studies were conducted 
at two sites within GRNP: Surveyors Creek (SC: 

29°32° S, 152° 18° E, 1044 m a.s.l) and Waratah 

Trig (WT: 29°29° S, 152° 19° E, 1050 m a.s.l.). 

Self-compatibility 

To assess self-compatibility, inflorescences on 
plants were either bagged and hand self-pollinated 
or left open to receive natural pollination by pollen 



vectors. Either one or two inflorescences at a similar 
stage of development on each plant were randomly 
assigned to the two treatments. Selfed inflorescences 
were covered with nylon mesh bags with apertures 
of < 1 mm in diameter just prior to flower opening. 
Every 4-6 days during flowering, bags were removed 
to self-pollinate flowers and then replaced. We 
removed pollen from newly opened flowers using 
pieces of soft cloth attached to small wooden sticks 
and self-pollinated flowers that had opened a few 
days previously. When flowering was complete, 
bags were removed firom inflorescences. Open 
inflorescences were marked with flagging tape but 
were otherwise left untouched. Cross-pollinations 
were not performed because we were unable to 
visit study sites sufficiently often to remove self 
pollen and thereby avoid possible autonomous self- 
pollination of flowers. The number of inflorescences 
developing follicles was scored about 10 months after 
flowering when follicle development was discernible. 
Inflorescences with follicles (hereafter cones) were 
harvested and the numbers of follicles on each were 
counted. Follicles were opened using a blowtorch and 
the number of filled seeds per cone was determined. 

To assess variation in the natural levels of seed 
set and the effects of selfing among years and sites, 
experiments were conducted in three consecutive 
years at SC (1997, 1998 and 1999) and at SC and 
WT in 1999. Sample sizes ranged between 12 and 
30 plants per year per site but were reduced for final 
analyses because mammals broke into some bags 
and cockatoos destroyed some cones before they 
could be harvested. For the plants with two cones per 
treatment, the mean value was used in the analyses. 
Different plants were used in each of the three years 
at SC. All plants had surplus inflorescences that were 
not used in the experiment. Pollination treatments 
were conducted on the same plant to control 
for plant genotype when assessing seed fitness. 

To examine the effect of pollination on the number 
of inflorescences that produced follicles, we used a 
logit model with a binomial error term and a logit link 
fimction. The response variable was the number of 
inflorescences with follicles. Explanatory variables 
were pollination treatment and year or site. Numbers 
of follicles and seeds per cone were compared 
between treatments with two-way ANOVAs with 
pollination treatment as a fixed factor and year or site 
as random factors. The interaction between the main 
factors was examined in preliminary analyses and, if 
not significant (P > 0.20), was omitted firom the final 
model to increase the degrees of freedom for testing 
the main effects. When the interaction was significant. 



20 



Proc. Linn. Soc. N.S.W., 127, 2006 



G. VAUGHTON AND M. RAMSEY 



Table 1. The percentage of Banksia marginata inflorescences producing follicles (i.e. cones) and 
the mean (± SE) numbers of follicles and seeds per cone following either experimental self- 
or natural open-pollination. The number of plants in each treatment is given in parentheses. 



Trait 



Year 



Site 



Self-pollinated 



Open-pollinated 



Inflorescences with follicles (%) 



Number of follicles per cone 



Number of seeds per cone 



1997 


SC 


94 (16) 


92 (25) 


1998 


SC 


94 (35) 


94 (36) 


1999 


SC 


83 (57) 


93 (45) 


1999 


WT 


85 (20) 


89 (44) 


1997 


SC 


22.7 ±2.4 (15) 


33.3 ±1.2 (15) 


1998 


SC 


23.0 ±1.9 (20) 


24.0 ±1.4 (20) 


1999 


SC 


22.6 ±2.8 (12) 


22.6 ±2.8 (12) 


1999 


WT 


19.6 ±2.4 (12) 


21.0± 1.9(12) 


1997 


SC 


32.1 ±3.4 (15) 


55.9 ±1.9 (15) 


1998 


SC 


33.6 ± 2.5 (20) 


46.1 ±3.7 (20) 


1999 


SC 


26.4 ±5.0 (12) 


33.3 ±4.8 (12) 


1999 


WT 


31.0 ±5.3 (12) 


35.0 ±4.3 (12) 



differences between the pollination treatments 
were examined separately for each year or site. 

Progeny fitness 

Seeds produced by self-pollinated inflorescences 
were self-fertilised, whereas seeds produced by open 
inflorescences may have been either self- or cross- 
fertilised. Progeny fitness was examined using a 
subset of 16 plants at SC in 1998. Seed mass was 
examined by weighing 20 seeds individually from 
selfed and open inflorescences on each plant to the 
nearest 0. 1 mg. Individual seeds were placed on the 
soil surface of tubes (282 cm^) containing a 1:1:1 
mixture of sand, loam and peat. Tubes were placed 
on a bench in a laboratory with natural light at about 
20 C and kept moist. Seeds were inspected every 
day and the number that germinated was scored. 
About four weeks after sowing when most seedlings 
had produced expanded cotyledons, tubes were 
relocated to the glasshouse and arranged randomly 
on benches. Plants were regularly watered and were 
fertilised once after 10 weeks with 30 ml of half- 
strength 'Aquasol'. Plants were inspected weekly 
and mortality recorded. After 12 weeks seedlings 
were harvested and plant mass (roots + shoots) 
was determined after drying at 80 C for 3 days. 

The effects of pollination on seed germination 
and seedling survival were assessed with logit models 
with a binomial error term and a logit link function. 



The response variable was either the number of 
germinated seeds or the number of surviving 
seedlings. Pollination treatment and maternal plant 
were explanatory variables. Differences in seed and 
seedling mass between the treatments were assessed 
using two-way ANOVAs, with pollination treatment 
as a fixed factor and maternal plant as a random factor. 
To satisfy the assumptions of ANOVA, seedling 
mass was transformed using natural logarithms. 
For each trait, we used individual maternal 
plants to calculate relative fitness as: Rf = Wg/wo, 
where Wg and Wq, are the mean performances of 
selfed and open progeny, respectively (Charlesworth 
and Charlesworth 1987). Cumulative relative 
fitness was calculated for each maternal plant as the 
product of relative fitness values for the number of 
seeds per cone, percent seed germination, percent 
seedling survival and seedling mass. These traits 
were chosen because they are related to overall 
fitness and are probably independent of each other. 



RESULTS 

Self-compatibility 

Over three years at SC at least 83% of 
inflorescences in both pollination treatments produced 
follicles (Table 1). The number of inflorescences 
producing follicles was not dependent on pollination 



Proc. Linn. Soc. N.S.W., 127, 2006 



21 



SELF-COMPATIBILITY IN BANKSIA MARGINATA 



Table 2. Effects of self- and open-pollination on progeny fitness in Banksia marginata. Sixteen 
maternal plants were examined at SC in 1998. Sample sizes are given parentheses. Cumula- 
tive relative fitness was calculated as the product of the relative fitness of individual traits ex- 
cept seed mass. Relative fitness estimates were calculated as the mean of the 16 maternal plants. 



Trait 



Self-pollinated 



Open-pollinated 



Relative fitness 



Number of seeds per cone 
Seed mass (mg) 
Seed germination (%) 
Seedling survival (%) 
Seedling mass (mg) 



34.7 ±2.9 (16) 
6.24 ±0.10 (320) 
97.2 (320) 
79.1 (311) 
156.2 ±2.9 (246) 



49.1 ±4.2 (16) 
7.32 ± 0.07 (320) 
97.8 (320) 
84.0(313) 
174.7 ±3.3 (263) 



0.73 ± 0.05 
0.85 ± 0.03 
0.99 ± 0.01 
0.95 ± 0.04 
0.90 ±0.03 



Cumulative relative fitness 




0.62 ± 0.05 



treatment (G = 2.96, df = 2, P = 0.227), year (G = 1 .44, 
df = 1, P = 0.231), or their interaction (G = 1.31, df = 
2, P = 0.518). Differences in the numbers of follicles 
and seeds per cone between the treatments varied 
among years as indicated by significant treatment x 
year interactions and were compared for each year 

separately (Table 1 ; treatment x year: follicles, F^ „„ = 

4.28, P= 0.017; seeds F^ gg = 2.71, P = 0.074). Se'lfed 
cones produced significantly fewer follicles than 
open cones in 1997; differences in other years were 

not significant (1997, Fj ^g = 15.58, P < 0.001; 1998, 

F, 33 = 0.22, P = 0.638; 1999, F, ^^= 0.10, P = 0.758). 
In' addition, selfed cones produced significantly 
fewer seeds than open cones in 1997 and 1998, but 

not 1999 (1997, F, ,, = 37.20, P < 0.001; 1998, F 



1,28 



1,38 



= 7.80, P = 0.008; 1999; ¥^^^= 0.97, P = 0.336). 

In 1999 at SC and WT,' 85-93% of selfed and 

open inflorescences produced follicles (Table 1). 

The number of inflorescences with follicles was not 



dependent on pollination treatment (G = 0.004, df 
= 1, P = 0.944), site (G = 2.34, df = 1, P = 0.126), 
or their interaction (G = 0.53, df = 1, P = 0.468). 
For the numbers of follicles and seeds per cone, 
treatment x site interactions were not significant 
and were removed from the final models (both, 

Fj^^< 0.26, P > 0.614). Numbers of follicles and 
seeds per cone did not differ between treatments 

or sites (Table 1, all F^^^^ 1.24, P > 0.272). 

Progeny fitness 

Seed and seedling mass were significantly less 
following self-pollination than open pollination 
(Tables 2, 3). For seed mass, the significant treatment 
x plant interaction indicated that selfing negatively 
affected seed mass to a greater extent in some plants 
than others. The treatment x plant interaction was 
marginally significant for seedling mass and was 
probably related to variation in seed mass. Variation 



Table 3. Results of two-way ANOVAs (F) for seed and seedling mass, and analyses of de- 
viance (G) for seed germination and seedling survival. The effects of self- and open-pol- 
lination, maternal plant and their interaction on progeny fitness were examined in Bank- 
sia marginata. Data are presented in Table 2. jP < 008; ** p < 0.01; *** P < 0.001. 



Trait 




Treatment 

df F or G 



Plant 
df 



ForG 



Interaction 

df F or G 



Seed mass 


1,15 


23.31 


15, 608 


29.76 


15, 608 


5.98 


Seedling mass 


1,15 


** 
13.00 


15,477 


*** 
8.52 


15,477 


1.58^ 


Seed germination 


1 


0.26 


15 


11.29 


15 


16.72 


Seedling survival 


1 


3.27' 


15 


*** 
56.83 


15 


21.77 



22 



Proc. Linn. Soc. N.S.W., 127, 2006 



G. VAUGHTON AND M. RAMSEY 



occurring among maternal plants was significant 
in both analyses (Table 3). Seed germination was 
independent of pollination treatment, but there was a 
marginally significant trend for lower survival of selfed 
progeny compared with open progeny (Tables 2, 3). 
The treatment x plant interaction was not significant 
for either trait. Seedling survival, but not seed 
germination, differed among maternal plants (Table 3). 
Relative fitness of selfed versus open progeny for 
the 16 plants varied from 0.73 for the number of seeds 
per cone to 0.99 for seed germination (Table 2). Mean 
cumulative relative fitness estimated fi^om the number 
of seeds per cone, seed germination, seedling survival 
and seedling mass was 0.62, indicating that on average 
selfed progeny were only 62% as fit as open progeny. 



DISCUSSION 

Substantial numbers of fiiiits and seeds were 
produced following experimental self-pollination, 
indicating that Banksia marginata plants at GRNP 
are self-compatible. Studies of other banksias have 
showTi that species fall into one of two groups with 
respect to self-compatibility; those that produce 
few or no seeds following selfing and those that 
produce moderate to large numbers of selfed seeds. 
The results of this study indicate that B. marginata 
should be included in the second group. Other species 
in this group include B. spinulosa var. neoanglica 
(Vaughton 1988), B. brownii (Sampson et al. 1994), 
B. ericifolia var. macrantha (Hackett and Goldingay 
2001) and B. baxteri (Wooller and Wooller 2001). 
Except for B. spinulosa var. neoanglica, which is 
able to resprout after fire, B. marginata and other 
Banksia species capable of producing high numbers 
of selfed seeds are killed by fire. The association 
between self-compatibility and obligate seeding has 
been noted in other studies of Banksia (Sampson et 
al. 1994; Wooller and Wooller 2001, 2002). Self- 
compatibility helps to buffer the effects of pollinator 
scarcity on seed set, and depending on pollinator 
availability, plants can produce a mixture of selfed 
and outcrossed seeds, resulting in mixed mating. 

Despite the substantial production of selfed seeds 
in B. marginata, differences between selfed and open 
inflorescences indicate that plants are not completely 
self-fertile and that some outcrossing occurs under 
natural conditions. As is common in banksias (Copland 
and Whelan 1989; Vaughton 1991), fruit and seed set 
of open inflorescences varied among years and sites, 
potentially reflecting the availability of pollinators 
and other factors. At SC in 1997, when follicle and 
seed production following open pollination were the 



highest, and hence the least likely to be limited by 
the availability of cross pollen, open inflorescences 
produced on average 56 seeds compared with only 
33 seeds by self-pollinated inflorescences. If all 56 
seeds on open inflorescences were outcrossed, then 
the maximum relative self-fertility can be estimated 
by dividing self seed set by open seed set, and would 
be 0.59. If, however, some of the seeds produced by 
the open inflorescences were selfed, then maximum 
crossed seed set is probably greater than 56 seeds. 
This would provide a lower estimate of self-fertility. 
Nevertheless, seed set in 1997 must have been close 
to the maximum because spatial constraints on cones 
would have limited the production of more follicles. 
Further, follicles can only produce two seeds, and on 
average 1.7 seeds per cone were produced on open 
inflorescences, indicating that our estimate of 0.59 
is probably close to the actual relative self-fertility. 

The minimum outcrossing rate at SC in 1997 can 
be estimated ifwe assume that levels of self-fertilisation 
on selfed and open inflorescences are similar. Thus, 
if 33 of the 56 seeds on open inflorescences were 
self-fertilised, then the remaining 23 seeds would be 
cross-fertilised, providing an estimated outcrossing 
rate of 0.41 (i.e. 23/56). The outcrossing rate may 
have been less in years and sites when selfed and 
open inflorescences produced similar numbers of 
seeds. Studies of outcrossing rates in Banksia species 
using genetic markers have generally indicated 
high outcrossing rates, even for species that exhibit 
substantial self-fertility. This has been attributed to 
inbreeding depression and selective abortion of selfed 
progeny (Vaughton and Carthew 1993; Carthew et 
al. 1996). An exception is B. brownii that appears to 
maintain lower outcrossing rates than other Banksia 
species (Sampson et al. 1994; Day et al. 1997). 

Selfed progeny were less fit than those resulting 
fi^om open-pollination, indicating that inbreeding 
depression occurs in B. marginata. Compared with 
open-pollinated seeds, selfed seeds were smaller 
and produced smaller seedlings that were less 
likely to survive. Maternal effects could not have 
been responsible for these differences because we 
specifically controlled for maternal genotype in our 
experimental design. Seed mass has been found 
to be a predictor of seedling size and survival in 
other plant species (Paz and Martinez-Ramos 2003; 
Khan 2004). In B. marginata, the N and P content 
of seeds increases linearly with increasing seed 
mass, rendering seedlings less dependent on external 
supplies of these nutrients in their natural habitat 
(Vaughton and Ramsey 1998). In many banksias, 
seedling establishment occurs after fire and seedlings 
may utilise N and P reserves in seeds to complement 



Proc. Linn. Soc. N.S.W., 127, 2006 



23 



SELF-COMPATIBILITY IN BANKSIA MARGINATA 



the high levels of other nutrients that are present in the 
immediate post-fire environment (Stock et al. 1990). 

Cumulative fitness estimated over several life- 
cycle stages, including the number of seeds per cone, 
indicated that on average selfed progeny vi^ere only 
62% as fit as open-pollinated progeny. Assuming 
all open-pollinated progeny were outcrossed, this 
equates to moderate inbreeding depression of 
0.38. If open inflorescences produced a mixture 
of selfed and outcrossed progeny, then inbreeding 
depression would be higher. Fitness differences 
between the pollination treatments also may have 
been underestimated in this study because only early 
life-cycle stages were examined, and plants were 
grown under benign conditions in the glasshouse 
(Ramsey and Vaughton 1998). The observed 
cumulative fitness estimate indicates that despite 
self-compatibility, these B. marginata populations 
have likely experienced a history of outcrossing. 
High levels of early-acting inbreeding depression are 
common in species with substantial outcrossing, and 
reflect a lack of opportunities for purging deleterious 
recessive alleles (Husband and Schemske 1996). 

Further study of the breeding and mating systems 
of 5. marginata is clearly warranted to determine the 
relative benefits of self-compatibility in providing 
reproductive assurance and the fitness costs associated 
with self-pollination. In particular, hand cross- 
pollinations in combination with self-pollinations 
would confirm our estimates of relative self-fertility 
and allow a more accurate estimate of inbreeding 
depression. Studies using genetic markers would also 
be valuable in determining realised outcrossing rates 
in populations and the effects of selfing on population 
genetic structure. The importance of pollinators for 
seed set also needs to be determined because banksias 
have an unusual pollen presentation mechanism, 
which in some species facilitates autonomous 
self-pollination and seed set in the absence of 
pollinators (Vaughton 1988). Finally, B. marginata 
exhibits considerable variation over its geographic 
range and both obligate seeding and resprouting 
populations occur (George 1998). Studies of the 
breeding capabilities of plants over the geographic 
range, and especially in resprouting populations, 
may provide insight into the factors favouring 
the evolution of self-compatibility in this species. 



ACKNOWLEDGEMENTS 

We thank Peter Clarke and the referees for comments on 
the manuscript and Stuart Cairns for statistical advice. 
Financial support was provided by a UNE research grant. 



REFERENCES 

Barrett, S.C.H. (2003). Mating strategies in flowering 

plants: the outcrossing-selfing paradigm and beyond. 

Philosophical Transactions of the Royal Society, 

London B.2>5%99\-\mA. 
Byers, D.L. and Waller, D.M. (1999). Do plant populations 

purge their genetic load? Effects of population size 

and mating history on inbreeding depression. Annual 

Review of Ecology and Systematics 30, 479-513. 
Carthew, S.M., Whelan, R.J. and Ayre, D.J. (1996). 

Experimental confirmation of preferential outcrossing 

in Banksia. International Journal of Plant Sciences 

157, 615-620. 
Charlesworth, D. and Charlesworth, B. (1987). Inbreeding 

depression and its evolutionary consequences Annual 

Review of Ecology and Systematics 18, 237-268. 
Copland, B.J. and Whelan, R.J. (1989). Seasonal variation 

in flowering intensity and pollination limitation 

of fiuit-set in four co-occurring Banksia species. 

Journal of Ecology 11, 507-523. 
Day, D.A., Collins, B.G. and Rees, R.G. (1997). 

Reproductive biology of the rare and endangered 

Banksia brownii Baxter ex R.Br (Proteaceae). 

Australian Journal of Ecology 11, 307-315. 
George, A.S. (1998). Proteus in Australia. An overview 

of the current state of taxonomy of Australian 

Proteaceae. Australian Systematic Botany 11, 257- 

266. 
Goldingay, R.L. and Carthew, S.M. (1998). Breeding and 

mating systems of Australian Proteaceae. Australian 

Journal of Botany 46, 421-437. 
Hackett, D.J. and Goldingay, R.L. (2001). Pollination of 

Banksia spp. by non-flying mammals in north-eastern 

New South Wales. Australian Journal of Botany 49, 

637-644. 
Heliyanto, B., Veneklaas, E.J., Lambers, H. and Krauss, 

S.L. (2005). Preferential outcrossing in Banksia 

ilicifolia (Proteaceae). Australian Journal of Botany 

52, 195-199. 
Husband, B.C. and Schemske, D.W. (1996). Evolution of 

magnitude and timing of inbreeding depression in 

plants. Evolution 50, 54-70. 
Khan, M.L. (2004). Effects of seed mass on seedling 

success in Artocarpus heterophyllus L., a tropical 

tree species of north-east India. Acta Oecologica 25, 

103-110. 
Lande, R., Schemske, D.W. and Schultz, S.T. (1994). 

High inbreeding depression, selective interference 

among loci, and the threshold selfing rate for purging 

recessive lethal mutations. Evolution 48, 965-978. 
Lloyd, D.G. (1979). Some reproductive factors affecting 

the selection of self-fertilization in plants. American 

Naturalist 113, 67-79. 
Lloyd, D.G. (1992). Self- and cross-fertilization in plants. 

II. The selection of self-fertilization. International 

Journal of Plant Sciences 15, 370-380. 



24 



Proc. Linn. Soc. N.S.W., 127, 2006 



G. VAUGHTON AND M. RAMSEY 



Paz, H. and Martinez-Ramos, M. (2003). Seed mass 
and seedling performance within eight species of 
Psychotria (Rubiaceae). Ecology 84, 439-450. 

Ramsey, M. and Vaughton, G. (1998). Effect of 
environment on the magnitude of inbreeding 
depression in a partially self-fertile perennial herb 
{Blandfordia grandiflora, Liliaceae). International 
Journal of Plant Sciences 159, 98-104. 

Sampson, J.F., CoUins, B.G. and Coates, D.J. (1994). 
Mixed mating in Banksia brownii Baxter ex R.Br. 
(Proteaceae). Australian Journal of Botany 42, 103- 
111. 

Stock, W.D., Pate, J.S. and Delfs, J. (1990). Influence of 
seed size and quality on seedling development 
under low nutrient conditions in five Australian and 
South African members of the Proteaceae. Journal of 
Ecology 78, 1005-1020. 

Vaughton, G. (1988). Pollination and seed set of Banksia 
spinulosa: Evidence of autogamy. Australian Journal 
of Botany 36, 633-642. 

Vaughton, G. (1991). Variation among years in pollen and 
nutrient limitation of fruit set in Banksia spinulosa 
(Proteaceae). Journal of Ecology 79, 389-400. 

Vaughton, G. and Carthew, S.M. (1993). Evidence for 
selective abortion in Banksia spinulosa (Proteaceae). 
Biological Journal of the Linnean Society 50, 35-46. 

Vaughton, G. and Ramsey, M. (1998). Sources and 
consequences of seed mass variation in Banksia 
marginata (Proteaceae). Journal of Ecology 86, 563- 
573. 

Virgona, S., Vaughton, G. and Ramsey, M. (2006) Habitat 
segregation of Banksia shrubs at Gibraltar Range 
National Park. Proceedings of the Linnean Society of 
New South Wales 111, 39-47. 

Wooller, S.J. and Wooller, R.D. (2001). Seed set in two 
sympatric banksias, Banksia attenuata and B. baxteri. 
Australian Journal of Botany 49, 597-602. 

Wooller, S.J. and Wooller, R.D. (2002). Mixed mating 
in Banksia media. Australian Journal of Botany 50, 
627-631. 

Wooller, S.J. and Wooller, R.D. (2003). The role of non- 
flying animals in the pollination of Banksia nutans. 
Australian Journal of Botany 51, 503-507. 

Wooller, S.J. and Wooller, R.D. (2004). Seed viability in 
relation to pollinator availability in Banksia baxteri. 
Australian Journal of Botany 52, 195-199. 



Proc. Linn. Soc. N.S.W., 127, 2006 25 



26 



Fire History and Soil Gradients Generate Floristic Patterns 
in Montane Sedgelands and Wet Heatlis of Gibraltar Range 

National Park 

Paul Richard Williams'-^ and Peter John Clarke^ 

'Botany, School of Environmental Sciences and Natural Resources Management, University of New Engiand, 

Arm iDAiE, NSW, 2351 (pcL!aacE@uNE.EDU.Au). 

^School OF Tropical Biology, James Cook University, and Queensiand Parks and Wiidufe Service, PO Box 

5597, Townsviue, Queenslwd 4810, Austraua (paulwiijuam s@epa.qid.gov.au). 



WiiiiAMS, p. AND Clarke P.J. (2006). Fire history and soilgradients generate Aoristic patterns in montane 
SEDGELWDS AND WET HEATHS OF GIBRALTAR Range NationalPark. Proceedings of the Linnean Society of 
New South Wales 111, 27-38. 

High rainfall escarpment areas along the Great Dividing Range provide habitats for sedgeland and wet 
heath vegetation in areas with impeded drainage. There are few studies of the processes that influence the 
floristic composition of montane sedgelands and heaths in relation to fires that sweep these landscapes. 
Gibraltar Range National Park contains extensive areas of sedge-heaths that remain mostly fi-ee fi'om 
anthropogenic disturbance. These areas have a well-known fire history which provides an opportunity to 
test whether: 1) plant resources are related to time-since-fire; 2) floristic composition is more strongly 
related to physiographic factors than time-since-fire, and 3) floristic composition of vegetation is related 
to fire fi-equency. Physiographic position strongly influenced the vegetation's structure and floristic 
composition, with taller heaths confined to better-drained edges whereas sedgelands were more common 
in poorly drained slopes regardless of fire regime. In turn, these patterns were related to soil conductivity 
reflecting the fertility status of the soils. Upper slope heaths were more species rich than those lower in 
the landscape where soil conductivity was higher Time-since-fire strongly influenced heath structure and 
species richness declined in the heaths with canopy closure at some sites. Floristic composition across the 
physiographic gradient was more divergent soon after fire and became more similar 1 5 years after fire. Fire 
fi'equency had no significant effect on shrub species richness, but frequent fires decreased the abundance of 
some woody species. Inter-fire intervals of less than seven years may reduce the abundance of some shrub 
species. Both the history of fire and ease of access make the sedgelands and wet heaths of Gibraltar Jlange 
an ideal location to assess the long-term effects of fire regimes in montane sedge-heaths. 

Manuscript received IMay 2005, accepted for publication 7 December 2005. 

KEYWORDS: Fire ecology, heaths, resource gradients, species richness, time-since-fire. 



INTRODUCTION 

Montane plateaux along the Great Dividing Range 
have high rainfall and low evaporation creating ideal 
conditions for sedgelands and wet heath communities 
where drainage is impeded (Keith 2004). Beadle 
(1981) described the mosaic of sedgelands and wet 
heaths as "sedge-heaths", but more recently Keith 
(2004) has circumscribed them as "Montane Bogs and 
Fens", reserving the term "Montane Heaths" to those 
heathlands with well-drained soils on rocky sites. The 
more poorly drained sedgelands are dominated by 
species of the monocotyledon families Cyperaceae, 



Juncaceae and Restionaceae whilst adjacent wet 
heaths are dominated by shrubs, especially of the 
families Ericaceae, Fabaceae and Myrtaceae (Keith 
2004). 

The earliest studies of sedge-heaths identified 
that sedgelands dominated the wettest areas, while 
shrubs were more common in better-drained positions 
(Pidgeon 1938). Early descriptions also considered 
sedge-heaths as a sere in succession leading to more 
complex vegetation (Pidgeon 1938; Davis 1941; 
Jackson 1968). With this focus, Millington (1954) was 
the first to describe the sedge-heaths of the Northern 
Tablelands of NSW describing the cyclic formation 
of Sphagnum hummocks and hollows following 



FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS 



the European tradition. More recently, Whinam and 
Chilcott (2002) surveyed the floristic composition 
and environmental relationships of Sphagnum bogs 
in eastern Australia but did not sample those areas 
where Sphagnum was absent. 

Contemporary studies have considered geology, 
soil depth and soil moisture as interrelated factors 
controlling floristic patterns within sedge-heath 
communities (Burrough et al. 1977; Buchanan 1980; 
Brown and Podger 1982; Pickard and Jacobs 1983; 
Bowman et al. 1986; Myerscough and Carolin 1986). 
The importance of water level in determining the 
distribution of dominant montane sedge-heaths species 
was shown by Tremont (1991), who evaluated the 
effects of hydrological changes resulting from a dam 
built across a wet heath in Cathedral Rock National 
Park. Resource-driven processes were highlighted in 
the detailed studies of Keith and Myerscough (1993) 
who found species richness was inversely related to 
soil resources, consistent with resource competition 
models that predict greatest species diversity with 
lowest levels of resources (Tilman 1982). 

Fire is a regular event in sedge-heath 
communities, due to the dense graminoid biomass 
and the fine elevated fuels presented in the leaves of 
the sclerophyllous shrubs. Both obligate seeder (fire- 
killed) and resprouting species co-exist within wet 
heaths, although resprouting shrub species are more 
numerous than those killed by fire (Clarke and Knox 
2002). Plant species richness is usually highest in the 
initial post-fire community, due to the recruitment 
of short-lived species (e.g. Specht et al. 1958), with 
an inverse relationship between shrub canopy cover 
and understorey species richness (Specht and Specht 
1989; Keith and Bradstock 1994). Frequent fires in 
sedge-heath communities also have the potential to 
alter floristic composition if the life cycles of plants 
are not completed between fire intervals (Keith et al. 
2002). 

There are few studies of the processes that 
mediate the floristic composition of the montane 
sedge-heaths in northern NSW, unlike their coastal 
and southern counterparts (see Keith 2004). Gibraltar 
Range National Park contains extensive areas of 
montane sedge-heaths that remain mostly fi-ee fi-om 
anthropogenic disturbance. These bogs and heaths 
also have a well-known fire history which provides 
an opportunity to test whether: 1) plant resources 
(soil and light) are related to time-since-fire; 2) 
floristic composition is more strongly related to 
physiographic factors than time-since-fire, and 3) 
floristic composition is related to fire firequency. 



MATERIALS AND METHODS 

Study sites 

Study sites were located in Gibraltar Range 
National Park in February 1995 by choosing replicate 
sedge-heaths that were widely spaced with different 
fire histories. Fire histories of different sections of the 
park were determined through consultation with park 
staff and their fire history records. Fire firequency over 
the last 30 years was found to be similar for many 
sedge-heaths of the park, differing only in whether 
they had been burnt since 1980 (i.e. differing in the 
time since last fire). 

Sedge-heaths occur as distinct swampy low- 
lying islands surrounded by eucalypt forest. Six 
sedge-heaths were selected for this survey based 
on certainty and differences in known fire history 
and ease of access. All six sedge-heaths were burnt 
in wildfires of both 1964 and 1980. Two remained 
unbumt since 1980 and were considered in this study 
as long unbumt (i.e. 15 years since fire). Two sedge- 
heaths were burnt in a plarmed bum in 1994 and 
were considered regenerating communities, having 
been bumt only half a year prior to this study. The 
remaining two sedge-heaths had an intermediate age 
since last fire. One was bumt in a wildfire in 1989, the 
other in a plaimed bum in 1990. Therefore of the six 
sedge-heaths surveyed, two were last bumt 15 years, 
two 5-6 years, and two were bumt half a year prior 
to the survey (Williams 1995). Following the 1995 
survey all study sites were bumt by a landscape-scale 
wildfire seven years later in November 2002 and a 
subset of the original sites were re-sampled. 

Sampling design 

Preliminary inspections of the sites suggested 
that floristic patterns were likely to vary with the soil 
moisture gradient from the drier outer edge to the 
drainage channels flowing through the centre of each 
sedge-heath, as documented in similar communities 
in southem Australia (e.g. Buchanan 1980; Keith and 
Myerscough 1993). Therefore a stratified sampling 
design was used, where each sedge-heath was divided 
into three habitats: drier outer edge, mid-slope and 
drainage charmel. To survey spatial variation, three 
plots were placed in each of the three habitats in each 
of the top, central and lower sections of sedge-heath. 
Therefore 27 plots (3 habitats x 3 plots x 3 sections) 
were surveyed in each of the six sedge-heaths (2 areas 
X 3 time-since-fire), providing a total of 162 plots. In 
addition 36 plots (2 habitats x 3 plots x 2 areas x 3 
fire frequencies) were re-sampled in 2003 for woody 
species. In this sampling, the drier outer edge and 



28 



Proc. Linn. Soc. N.S.W., 127, 2006 



p. WILLIAMS AND P.J. CLARKE 



drainage channel were surveyed in each of two sedge- 
heaths for each fire frequency. 

The quantitative nested quadrat method 
(Morrison et al. 1995) was used to document species 
abundance at each plot. This method uses concentric 
sub-quadrats of increasing size, which were 1, 4, 9, 
16 and 25m^ in this study. An abundance score out 
of five was given to each of the species at each plot, 
derived fi-om the number of sub-quadrats it was 
present within. Plant nomenclature follows Harden 
(1990-93) with later modifications adopted by the 
National Herbarium, Sydney, and voucher specimens 
of uncommon species were lodged in the NCW 
Beadle Herbarium (NE) Herbarium. Fire responses 
(obligate seeder or resprouter) were documented 
for species within the recently burnt sites. Electrical 
conductivity and soil pH measurements were taken 
using electronic meters and a 1:5 ratio of soil to 
distilled water. Electrical conductivity is positively 
correlated with soil ionic concentrations and hence 
is a crude index of soil fertility. A light reading was 
taken at the soil surface at each plot and calculated 
as a percentage of a reading taken above the canopy. 
Aspect, degree of slope and canopy height were also 
recorded at each plot. 

Analyses 

The species composition and abundance data 
for each plot were correlated with environmental 
variables using a canonical correspondence analysis 
(CCA) through the CANOCO program. The CCA 
is calculated in two stages. Firstly the similarity of 
the 162 plots, based on species composition and 
abundance, is calculated to display the relative 
ordering of sites (i.e. ordination). The ordination is 
undertaken using a correspondence analysis (CA), 
which is a modal response model, which assumes 
species reach a maximum abundance at a point 
along an environmental gradient. The second step 
in the CCA is a multiple regression technique that 
evaluates the link between environmental variables 
at each plot and the initial ordination of plots based 
on species abundance. In addition, the 36 plots (2 
habitats x 3 plots x 2 areas x 3 fire frequencies) that 
were re-sampled in 2003 for woody species only were 
analysed using CCA with fire frequency and habitat 
as environmental variables. 

The relationships between environmental 
variables of canopy height, light, pH and soil 
conductivity, and habitat and time-since-fire were 
examined using a general linear model (GLM) with 
habitat (3 levels) and time-since-fire (3 levels) as 
orthogonal factors. This orthogonal design was also 
applied in GLM analyses for the richness response 



variables of total species, resprouters, obligate 
seeders, woody plants, graminoids, grasses, ferns 
and forbs. In addition analyses of covariance were 
performed with conductivity as a covariate. A fire- 
frequency orthogonal GLM analysis was also applied 
to species richness data collected in 2003 with fire 
frequency (3 levels) and habitat (2 levels). A Poisson 
error structure with a log link function was applied 
for species richness data, a binomial error structure 
with a logistic link function for species presence/ 
absence data and an identity link function was applied 
to normally distributed data. 



RESULTS 

Effects of time-since-fire and habitat 

Eighty-nine taxa were recorded from the 162 
plots sampled in 1995 (see Appendix 1). Shrubs were 
the most common growth form (41 spp.) followed 
by graminoids (21 spp.), forbs (13 spp.), grasses 
and trees (5 spp. each) and ferns (4 spp.). Among all 
growth forms, 19 species were killed by fire and 70 
were recorded as resprouting. Of those species killed 
by fire, only Banksia marginata had canopy-held seed 
banks. 

Ordination of sample sites in two dimensions 
showed distinct clustering of sites in relation to time- 
since-fire and physiography (Fig. 1). The strongest 
effects were time-since-fire with 15 years at the top 
of the ordination and the more recently burnt sites 
at the base, whilst the drier edge site to the wetter 
chaimel sites are distributed left to right (Fig. 1). This 
floristic gradient is initially wide in the short time- 
since fire sites but converges with longer time-since 
fire (Fig. 1). Both light and soil resources were related 
strongly to physiographic position and time-since-fire 
(Fig. 1) and when examined using univariate analyses 
they show the effect of canopy closure on light levels 
(Table 1, Fig. 2). Univariate analyses also show a 
strong resource gradient with soil conductivity being 
higher along the charmels with corresponding lower 
pH (Fig. 2). Both conductivity and pH were, however, 
not consistently related to time-since-fire (Table 1, 
Fig. 2). 

There was significant negative correlation 
between conductivity and species richness (r = 
- 0.53, P < 0.001) (Fig. 3). The relationship between 
species richness, fire response and growth form 
groups were further examined using GLM which 
showed inconsistent patterns of time-since-fire and 
habitats with a significant interaction term (Table 
2). The drier outer edge plots contained a greater 



Proc. Linn. Soc. N.S.W., 127, 2006 



29 



FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS 



Canopy 



Time-Since-Fire 




Light 



■1.0 



Figure 1. Biplot from the canonical correspondence analy- 
sis. Symbols represent plots, arrows represent environmen- 
tal gradients. Top cluster = 15 year old plots, middle cluster 
5 year old, bottom cluster 0.5 year old plots. Circles = edge 
plots, squares = mid slopes, triangles = drainage channel plots. 



number of species compared with the other plots 
(Fig. 4a). Species richness decUned with time-since- 
fire in these outer edge plots and reached a peak 
some five years later on the slopes (Fig. 4a). The 
four most abundant species in drier edge plots of 
recently burnt sedge-heaths were Ptilothrix deusta, 
Amphipogon strictus, Leptospermum arachnoides 
and Lepidosperma limicola. In areas unbumt for 15 
years, Ptilothrix deusta, Leptospermum arachnoides 



and Lepidosperma limicola remained 
the most abundant, but the grass 
Amphipogon strictus was replaced by 
the obligate seeding twiner, Cassytha 
glabella. The recently burnt edge plots 
contained a total of 62 species whilst 
48 species were documented in plots 
unbumt for 15 years. In these edge 
plots the mean number of obligate 
seeding species and resprouting 
species decreased over time, as did 
richness of herbaceous species (Fig. 

4). 

The mid slope and channel plots in 
recently biimt sedge-heaths contained 
a total of 40 species. The most abundant 
species in the wetter mid slope and 
channel plots of recently burnt sedge- 
heaths were Lepidosperma limicola, 
Baeckea omissa, Gymnoschoenus 
sphaerocephalus and Drosera 
binata. Lepidosperma limicola, 
Baeckea omissa and Gymnoschoenus 
sphaerocephalus were also the most 
abundant species in the sedge-heaths 
unbumt for 15 years. By this stage 
the herb, Drosera binata, was much 
less abundant and was replaced by 
Epacris obtusifolia, an obligate 
seeder subshrub. In the chaimel plots 
resprouter richness decreased through 
time whilst obligate seeder richness increased (Fig. 
4b,c). On the bog slopes species richness appeared to 
peak six years after fire then decline mainly due to the 
decrease in grass and sedge species (Fig. 4a,e). 



1.0 



Effects of fire frequency on shrub species 

No significant trend in species composition with 
fire frequency was detected for shmb species in the CCA 
analyses, which are shown. Similarly, no effect of fire 

Table 1. Summary results for two factor general linear models for time-since-fire and habitat for 
environmental variables. All models have a Poisson error structures with a log-link function and 
have scale estimated using Pearson Chi-squared. 




Factor 



df 



Canopy height 



% Light 



pH 



Conductivity 



F ratio 



F ratio 



F ratio 



F ratio 



Time-Since-Fire 


2 


141.7 


*** 


244.9 


*** 


2.9 


* 


2.6 


ns 


Habitat 


2 


30.0 


*** 


27.0 


*** 


8.3 


*** 


65.9 


*** 


TSF X Habitat 


4 


2.3 


ns 


2.5 


ns 


3.7 


*** 


3.3 


* 


Residual 


153 



















30 





Proc. Linn. Soc. N.S.W., 127, 2006 



p. WILLIAMS AND PJ. CLARKE 




Channel Mid slope Outer edge 



Channel Mid slope Outer edge 




25-1 d) 



20 



15 



to 



10 



0-" 




Channel Mid slope Outer edge 



Channel Mid slope Outer edge 



Figure 2. Mean (+se) for a) canopy height, b) % full sunlight, c) pH, and d) 
conductivity for each of three physiographic positions in the sedge-heaths and 
among three time-since-fire locations. 



35 



^ 30 i 

CM 



in 






U 






25 - 



20 - 



^151 

Ol 

10 H 



# 



^p 



Channel 
Mid slope 
Outer edge 








— I — I — I — I — I — I — I — I — I — I — I — I — I — I — 1 — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I — I 

5 10 15 20 25 30 35 



Conductivity 



Figure 3. Relationship between conductivity (dS/m), species richness and topograph- 
ic position across the sedge-heaths at Gibraltar Range. Lowness Une fitted. 



Proc. Linn. Soc. N.S.W., 127, 2006 



31 



FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS 



X 









o 



* 
* 
* 



to 



00 






0\ 



H 












3 






fO 




Tl 


^ 




P 


5' 




o 


o 




O 


a> 




•-t 


^ 












>^' 






ft 






K) 








►n 


zn 


ON 


^ 


^ 




tu 


a> 






o 
en 

o' 

3" 


* 
* 


"0 


3 
ft 








M 

^ 




2.^ 




o' 


O i£ 


* 
* 


■^3 


en S. 


* 








►n 


o 


^^ 


S 


ct; 


b\ 


^ s^ 




o' 


o' era 

3- S. 

3 ft 
55 ft 


3 


^3 


w ft 

D. 
ft 
"-t 


to 


►-1 


g 




B5 


o 


Lo 


o' 


o 

Q. 

3. 
o 

3" 


3 


"0 


3 
ft 

w 


U) 


"1 








II 

ft 3 

S 2. 


* 


^ 

^ 


a. 


(O 
4i- 


a 


Q 


Lu 








o' 


3. 
o 

3^ 


* 




3 


* 
* 




ft 

CO 
C/3 


U) 


"-t 


Tl 


ON 




O 




o' 


3. 
o 


* 


^ 


ft 



3 


H 


O 


&9 


a 


2 


en 


i>> 


BT 


• 


&> 


CZl 


-<l 


e 


rt 


3 


S9 


3 


Wl* 


^ 


o 
s 


3 


•1 




•s 


w 


o 


S" 


w 


■! 


c 


^ 


« 


O 


s 


5* 

9^ 


W! 


o 


^ 


"1 




S 


» 


fD 


CfQ 


n 

£, 


N^* 


s 


S 


rt 


?r 


» 


•^ 


•t 


s 
s 


3 

o 


(-f- 


o. 


o 


!L 


s 


M 


&» 


s* 


s 


"1 


a 


F^ 




N^> 


cr 


3 


< 


? 


rt 


I/) 




KM* 


»9 


S 
n 


5r 


Sh 


rs 


"t 




n 


3 


s 


p 


a. 


^t- 




ffi 


ST 


a 


» 


e 


a 


«> 


f^ 


5* 


&9 


tra 


i?> 




o 


&9 
"1 


S' 


W 


s 


o 


n 


9 




n 


o 


Br 


a 


^* 




2 


CTQ 


A 


i-S 


e 


o 


&> 


B 


•t 


"a 


fB 


vi 


a 


• 


• 


3^ 



frequency on species richness was detected (Table 
3). However, six common resprouting species had 
significantly different abundances across sites with 
different fire frequencies (Table 3, Fig. 5). Of these, 
Leptospermum gregarium, Hibbertia rufa, Boronia 
polygalifolia and Grevillea acanthifolia had lower 
abundances in sites with the highest fire frequency 
(Fig. 5). 

DISCUSSION 

Distinct floristic patterns occur in the sedge-heaths 
of Gibraltar Range representing both physiographic 
and fire-regime effects. Firstly, floristic composition 
varies along gradients in soil moisture, which are 
linked with increased electrical conductivity and 
nutrient accumulation along the drainage lines. 
These drainage-driven patterns are similar to those 
described by Keith and Myerscough (1993) at Darkes 
Forest on the southern Sydney plateau of NSW. 
Despite these structural similarities, major floristic 
differences separate central and southern NSW from 
northern regions (Keith 1995; Keith 2004), but more 
detailed surveys of the sedge-heaths in the Northern 
Tablelands and comparative analyses are required. 
Initial comparative analyses of life-history attributes 
suggest similar growth form composition and fire 
response syndromes to other east coast heaths (Keith 
et al. 2002). 

Species richness values were generally higher 
toward the outer edge of the heaths and lower on 
the slopes and drainage channel corresponding to 
patterns at Darkes Forest. This inverse relationship 
between species richness and electrical conductivity 
(positively correlated with soil fertility) was similar 
to that found in other heaths (Keith and Myerscough 
1993; Myerscough et al. 1996), suggesting a 
widespread resource-competition effect in heaths 
with resource gradients. However, the overall number 
of species encountered was much smaller than the 
high species richness found in coastal heaths (Keith 
and Myerscough 1993). 

Habitat segregation of serotinous shrub species 
along gradients of moisture and soil fertility has been 
explored in manipulative experiments by Williams 
and Clarke (1997) who suggest that a combination of 
seedling establishment and seedling survival in relation 
to moisture gradients segregates species within these 
sedge-heaths. Patterns of seedling establishment are 
initiated by fire and the effect of time-since-fire was 
prominent in our analyses. Following the passage of 
fire, the sedge-heath canopy is opened up and ground 
level insolation peaks, but as plants grow taller, 



32 



Proc. Linn. Soc. N.S.W., 127, 2006 



p. WILLIAMS AND PJ. CLARKE 



28 - 



24 - 



^20 
in 

£12 

a. 
w 

° 8 
d 



E 

ID 

(SI 

-^ 

M 

o 
at 

a. 
w 

o 

6 



-■ 



14 n 



12 



10 



6 - 



4 - 



a) Total 




Channel Mid slope Outer edge 
c) Obligate seeders 




14 



12 - 



CM 10 

E 



in 
'3 

Of 

a. 

tA 

*•- 
o 



6 - 



4 - 



2 - 



o-' 



Channel Mid slope Outer edge 
e) Grasses and sedges 



Channel Mid slope Outer edge 



b) Resprouters 





Channel Mid slope Outer edge 
d) Woody species 



14 -I 



12 - 



<N 10 

c 

in 

^ 8 
£ 6 

Q. 

'S 4i 



2 - 



-■ 



14 



12 - 



<gl0 
in 

(SI 

^ 8 
.1 

o ^ 

S. 6 
v\ 

o 4 

d 




Channel Mid slope Outer edge 
f) Forbs and ferns 



miilln 



Channel Mid slope Outer edge 



Figure 4. Mean (+se) for species richness a) total, b) resprouters, c) obligate seeders, d) woody species, 
e) grasses and sedges, f) forbs and ferns for each of three physiographic positions and among three 
time-since-fire locations. 



Proc. Linn. Soc. N.S.W., 127, 2006 



33 



FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS 



7 A O 

APo 

P O L^ 

o — 
o 



CD 
03 



c 

ft 

3 
o 



o 



00 



0^ 






p 



to 





"n 




p 




O 








o 




^ 




a. 




^ 


^►u 


t> 




a ^ 




■^ ^ 




Si ^ 




n o 


^ 


s: s 






^►o 


i^ 1 ^ 






^ ,:5 s-^ 




a. 00 a 


^ ( 


ksia 
inal 




» a 


?< 






. s 




bber 
ipari 




a =r. 


'TS 


a 


?^^ 


a 




g Ci 
s ^ 




S- S 






■13 


Sr a 
a 


^K. 


S 








^" S3 




d a 




r^ ??■ 




a "> 




.S c> 




^ 


"13 


a 


?^^ 


o 








£; a 




^B. 




S: ^ 


■td 


a 



© 


S"H 


■t 


&9 ta 


S-^ o* 


n 


n "- 


"t 


1. « 


o 


« w 


i-ti 


B8 o- 


"J 


. Sum 
inomi 


< 


» 3 


» 


« S 


e 
s 




a 


v> <» 


» 


^ e 


s 




n 


<^ ^ 




rt- «» 




B O 




^ -< 




-^ 




I.S 




«- "1 




ST as 








to IT 




■ri* 




©5 




era JS 




M. &» 




<» ^ 




rt- 




n 3 




» o 




sBr 




^2. 




K 




^^ 




? 




a » 




o 3 




3 « 




2 « 




-1^ 




i£ C 




T3 rt 




n B 




2. « 




rt^ 




■* » 




as 




» a 




?^ 




6S B" 




' K 




05 a* 




"5 s." 




2 ^ 








ib 








mod 
sted 




5'^ 



ft 
en 



C 
D 

o 

X 



N) 



to 






►tI 



ft 

c 

3 
o 



o 



to 

o 



<1 

p 

bo 



00 

io 






to 

o 



bo 



CT\ 



p 



« 
* 



o 
to 





■n 




p 




o 




<-f 




o 




■^ 




D. 




i-ti 


*n 






3. on 




O >T3 




g^S 


'xj 


i s- 




to 


?< 


^ 


to 






??• 




re> 




a 




o 




2 


"0 


5' 




Co 




a 




t- 


^M 


05 ^ 




^ '^ 




^ 




Oq ^ 




a ^ 




>! (t 


■TJ 






S 


?^ 


33 


to 


c> 




&■ 




fii 




>! 




^s. 








a' 


"Ta 


i 


^M 


s 




?;■ ^ 




s ^ 




^ ^ 




>S- ~s 




X K- 


^ 


a- 


?<^ 


^ 




^ to 




oronia 
•galifo 


^3 


a' 



«L 


^ 


M 


&9 


sr 


2 


» 


fT 


< 




» 


• 




B 


a 


3 



3 


3 


5' 


>s 




•^ 


ffi 


"s 


-( 


A 


•1 


w 





B_ 


•t 






f^ 


c« 


M 


^^ 




2 


s* 

^ 


rt- 


^^ 


B 

V9 


s 


sr 


•1 

3 


a 


Si 


n 






o" 


w 


(Q. 


o' 


^ 


"< 


^^ 










n 


Nri" 




S 


S> 


W 


S 


S»fi 


B 


B 


S 


n> 


« 


s 


a 


« 


v; 


B 


Si 


5?S 



■1 


a 


09 


BT 


•tJ 


Si 


ffi 


a* 


2^ 


ff 


S* 


Si 


M 


f-K 


a 

&9 


> 


i-t- 




F 


3 









e. 



34 



Proc. Linn. Soc. N.S.W., 127, 2006 



p. WILLIAMS AND P.J. CLARKE 



0) 

o 

c 

(0 

c 

3 
< 






CO 

s 
1 

CO 


•S 

1 

C 

2 

CO 


CO 

i 
a 


CO 

i 
t 


1 

CD 

0} 

-Q 
■Q 

2 


1 

CD 
O) 

& 
1 

1 




CO 

5 

c 

CD 


CD 
CO 

.CO 

i 

CO 

1 

CD 


1 

1 
1 

CO 


^ 

^ 


'5 

CD 


CO 

i 

o 

CQ 


■i 

CO 


CD 

C 

i 

1 


1 

CD 

1 
& 

-J 


0) 
OQ 


B 














CD 






UJ 














OQ 







Figure 5. Total abundance scores (frequency score) for the ten most common woody plants recorded in 
sedge-heaths in 2003, eight months after a fire among areas that been burnt 2, 3, and 5 times since 1964. 



ground layer insolation subsequently decreases. 
Neither soil pH nor conductivity showed consistent 
trends with time-since-fire although it is likely that 
post-fire soil nutrients peaked immediately after fire. 
Hence it is thought that competition for light is the 
main driver for differences in floristic composition 
with time-since-fire (Specht and Specht 1989; Keith 
and Bradstock 1994) or alternatively the differences 
simply reflect species' life spans. Decreases in woody 
species richness with time-since fire are prominent in 
the better-drained, outer-edge heaths. Hence we think 
that competition rather than variation in the life span 
of plants is the causal factor. 

There were no major decreases in species 
richness in the channel or slope plots, which may 
reflect the slower growth dynamics of montane 
sedge-heaths compared with coastal systems. 
Overall, variation in floristic composition along the 
drainage gradient was greatest immediately after fire. 



and then became less variable at 15 years time-since- 
fire. This may reflect the lack of strong competitive 
exclusion in the drainage channel heaths, possibly 
due to their narrow and patchy distribution. We 
think the alternative explanation of the lack of short- 
lived species immediately after fire unlikely because 
short-lived species were common along creek banks. 
Unfortunately, studies of long-unbumt sedge-heaths 
were halted in 2003 when all long-unbumt sedge- 
heaths were burnt in wildfires. 

Fire frequency appears to have much less 
influence on composition than time-since-fire, 
although only shrub data were sampled. When shrub 
species abundances were examined individually 
several dominant species had reduced abundances 
under frequent fire regimes. This is consistent with 
patterns in the adjacent dry sclerophyll forests 
(Knox and Clarke in this volume) where higher fire 
frequencies reduced plant performance. We would 



Proc. Linn. Soc. N.S.W., 127, 2006 



35 



FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS 



predict, however, that if the intervals between fires 
were less than eight years then the dominance and 
composition would change. 



ACKNOWLEDGEMENTS 

The Director of the NSW National Parks and Wildlife 
Service is thanked for allowing us to do this work in Gibraltar 
Range National Park under permit No. 1601. The Service 
staff at Glen Innes are thanked for their encouragement and 
help in providing accommodation and access to the site. 
The students of The Ecology of Australian Vegetation in 
2003 helped collect the fire irequency data. Kirsten Knox is 
thanked for her thoughtful input and David Keith is thanked 
for comments that improved the manuscript. This study 
was aided from funding from a University of New England 
Beadle Scholarship to one of us (PRW). 



REFERENCES 

Beadle, N.C.W. (1981). 'The Vegetation of Australia. ' 
(Cambridge University Press, Cambridge). 

Bowman, D.M.J.S., Maclean, A.R. and Crowden, R.K. 
(1986). Vegetation-soil relations in the lowlands of 
southwest Tasmania. Australian Journal of Ecology 
11, 141-153. 

Brown, M.J. and Podger, F.D. (1982). Floristics and fire 
regimes of a vegetation sequence from sedgeland- 
heath to rainforest at Bathurst Harbour, Tasmania. 
Australian Journal of Botany 30, 659-676. 

Buchanan, R.A. (1980). The Lambert Peninsula Ku- 
ring-gai Chase National Park. Physiography and 
distribution of podzols, shrublands and swamps, 
with details of the swamp vegetation and sediments. 
Proceedings of the Linnean Society of New South 
Wales 104, 73-94. 

Burrough, RA., Brown, L. and Morris, B.C. (1977). 

Variations in vegetation and soil patterns across the 
Hawkesbury Sandstone plateau from Barren Grounds 
to Fitzroy Falls, New South Wales. Australian 
Journal of Ecology 2, 137-159. 

Clarke, P.J. and Knox, K.J.E. (2002). Post-fire response 
of shrubs in the tablelands of eastern Australia: 
do existing models explain habitat differences? 
Australian Journal of Botany 50, 53-62. 

Davis, C. (1941). Plant ecology of the Bulh District 
part 2: plant communities of the plateau and scarp. 
Proceedings of the Linnean Society of New South 
Wales 66, 1-10. 

Harden, G. J. (Ed.) (1990-3) 'Flora of New South Wales.' 
Vol. 1-4. (New South Wales University Press, 
Sydney). 

Jackson, W.D. (1968). Fire, air, water and earth - an 
elemental ecology of Tasmania. Proceedings of the 
Ecological Society of Australia 3, 9-16. 

Keith, D.A. (1995). How similar are geographically 

separated stands of the same vegetation formation? 



A moorland example from Tasmania and mainland 

Australia. Proceedings of the Linnean Society of New 

South Wales 115, 61-75. 
Keith, D.A. (2004). 'Ocean shores to desert dunes: the 

native vegetation of New South Wales and the ACT'. 

(NSW Department of Environment and Conservation, 

Sydney). 
Keith, D.A. and Bradstock, R.A. (1994). Fire and 

competition in Australian heath: a conceptual model 

and field investigations. Journal of Vegetation 

Science 5, 347-354. 
Keith, D. A. and Myerscough, P. J. (1993). Floristics 

and soil relations of upland swamp vegetation near 

Sydney. Australian Journal of Ecology 18, 325-344. 
Keith, D.A., McCaw, W.L. and Whelan, R.J. (2002). Fire 

regimes in Australian heathlands and their effects 

on plants and animals. Pp. 199-237. In: Flammable 

Australia: The fire Regimes and Biodiversity of a 

Continent. Edited by Bradstock, R.A. Williams, 

J.E. and Gill, M.A. Cambridge University Press, 

Cambridge. 
Millington, R.J. (1954). Sphagnum bogs of the New 

England Plateau, New South Wales. Journal of 

Ecology 42, 328-324. 
Morrison, D.A., Le Brocque, A.F and Clarke, PJ. (1995). 

An assessment of some improved techniques for 

estimating the abundance (frequency) of sedentary 

organisms. Vegetatio 120, 121-135. 
Myerscough, P.J. and Carolin, R.C. (1986). The vegetation 

of the Eurunderee sand mass, headlands and previous 

islands in the Myall Lakes area, New South Wales. 

Cunninghamia 1, 399-466. 
Myerscough, RJ., Clarke, RJ. and Skelton, N.J. (1996). 

Plant coexistence in coastal heaths: habitat 

segregation in the post-fire environment. Australian 

Journal of Ecology 21, 47-54. 
Pickard, J. and Jacobs, S.W.L. (1983). Vegetation patterns 

on the Sassafras Plateau. In: 'Aspects of Australian 

Sandstone Landscapes.' (Eds R. W. Young and 

G. C. Nanson.) Pp.92 (Department of Geography, 

Wollongong University, Wollongong). 
Pidgeon, I M. (1938). Plant succession on the Hawkesbury 

Sandstone. Proceedings of the Linnean Society of 

New South Wales 63, 1-26. 
Specht, R.L., Rayson, R and Jackman, M.E. (1958). Dark 

Island heath (Ninety-mile Plain, South Australia) IV. 

Pyric succession: changes to composition, coverage, 

dry weight and mineral nutrient status. Australian 

Journal of Botany 6, 59-88. 
Specht, R.L. and Specht, A. (1989). Species richness of 

sclerophyll (heathy) plant communities in Australia - 

the influence of overstorey cover. Australian Journal 

of Botany 2,1, 337-50. 
Tilman, D. (1982). Resource Competition and Community 

Structure. Princeton University Press, Princeton. 
Tremont, R. (1991). Swamp Wetlands of the High Country 

of South-east Australia. Master of Letters Thesis, 

Botany Department, University of New England, 

Armidale. 
Whinam, J. and Chilcott, N. (2002). Floristic description 

and environmental relationships oi Sphagnum 



36 



Proc. Linn. Soc. N.S.W., 127, 2006 



p. WILLIAMS AND P.J. CLARKE 



communities in NSW and the ACT and their 
conservation management. Cunninghamia 7, 463- 
500. 
Williams, P.R. (1995). Floristic Patterns Within and 
Between Sedge-Heath of Gibraltar Range National 
Park, New South Wales. BSc Honours Thesis, 



Department of Botany, University of New England, 
Armidale, New South Wales. 
Wilhams PR. and Clarke, PJ. (1997). Habitat segregation 
by serotinous shrubs in heaths: Post-fire emergence 
and seedling survival. Australian Journal of Botany 
45,31-39. 



Appendix 1. Species recorded in sample sites of the sedge-heaths in 
Gibraltar Range National Park, their growth form and fire response. 
R = resprouting, S = obligate seeding. * exotic 



Species name 



Grovyth form 



Sprouting 



Allocasuarina littoralis 


Tree 


R 


Amphipogon strictus 


Grass 


R 


Aotus subglauca var. subglauca 


Shrub 


R 


Aristida ramosa 


Grass 


R 


A ustrostipa pubescens 


Grass 


R 


*Axonopus affinis 


Grass 


R 


Baeckea omissa 


Shrub 


R 


Baloskionfimbriatus 


Graminoid 


R 


Baloskion stenocoleus 


Graminoid 


R 


Banksia spinulosa 


Shrub 


R 


Banksia marginata 


Shrub 


S 


Baumea rubiginosa 


Graminoid 


R 


Blandfordia grandiflora 


Graminoid 


R 


Boronia microphylla 


Shrub 


R 


Boronia polygalifolia 


Sub-shrub 


S 


Bossiaea scortechinii 


Shrub 


R 


Brachyloma daphnoides ssp. glabrum 


Shrub 


R 


Caesia parviflora 


Graminoid 


R 


Callistemon pallidus 


Shrub 


R 


Callistemon pityoides 


Shrub 


R 


Cassytha glabella 


Forb 


S 


Caustis flexuosa 


Graminoid 


R 


Conospermum taxifolium 


Shrub 


R 


Cryptostylis subulata 


Graminoid 


R 


Dampiera striata 


Forb 


R 


Dianella caerulea 


Graminoid 


R 


Dillwynia phylicoides 


Shrub 


R 


Drosera binata 


Forb 


R 


Drosera spatulata 


Forb 


R 


Empodisma minus 


Graminoid 


R 


Entolasia striata 


Grass 


R 


Epacris microphylla var. microphylla 


Shrub 


R 


Epacris obtusifolia 


Shrub 


S 


Eucalyptus acaciiformis 


Tree 


R 


Eucalyptus campanulata 


Tree 


R 


Eucalyptus ligustrina 


Tree 


R 


Eucalyptus williamsiana 


Tree 


R 


Euphrasia collina ssp. paludosa 


Forb 


R 


Gleichenia dicarpa 


Fern 


R 


Gompholobium sp. "B" 


Shrub 


R 



Proc. Linn. Soc. N.S.W., 127, 2006 



37 



FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS 



Gonocarpus micranthus 


Forb 


S 


Gonocarpus teucrioides 


Forb 


R 


Goodenia bellidifolia 


Forb 


S 


Goodenia hederacea 


Forb 


s 


Grevillea acanthifolia ssp. stenomera 


Shrub 


R 


Grevillea acerata 


Shrub 


R 


Gymnoschoenus sphaerocephalus 


Graminoid 


R 


Hakea laevipes ssp. graniticola 


Shrub 


R 


Hakea microcarpa 


Shrub 


R 


Hibbertia rufa 


Shrub 


R 


Hibbertia riparia 


Shrub 


R 


Hovea heterophylla 


Sub-shrub 


R 


Hybanthus monopetalus 


Forb 


S 


Hypericum japonicum 


Forb 


S 


Isopogon petiolaris 


Shrub 


R 


Kunzea bracteolata 


Shrub 


S 


Lepidosperma limicola 


Graminoid 


R 


Lepidosperma tortuosum 


Graminoid 


R 


Leptospermum arachnoides 


Shrub 


R 


Leptospermum brevipes 


Shrub 


R 


Leptospermum gregarium 


Shrub 


R 


Leptospermum novae-angliae 


Shrub 


R 


Lepyrodia anarthria 


Graminoid 


R 


Lepyrodia scariosa 


Graminoid 


R 


Lindsaea linearis 


Fern 


R 


Logania pusilla 


Sub-shrub 


R 


Lomandra elongata 


Graminoid 


R 


Lomandra longifolia 


Graminoid 


R 


Lycopodium sp. 


Fern 


S 


Melichrus procumbens 


Shrub 


R 


Mirbelia rubiifolia 


Shrub 


R 


Monotoca scoparia 


Shrub 


R 


Patersonia sericea 


Graminoid 


R 


Persoonia rufa 


Shrub 


S 


Petrophila canescens 


Shrub 


R 


Phyllota phylicoides 


Shrub 


R 


Pimelea linifolia ssp collina 


Shrub 


S 


Platysace ericoides 


Shrub 


S 


Ptilothrix deusta 


Graminoid 


R 


Pultenaea polifolia 


Shrub 


S 


Pultenaea pycnocephala 


Shrub 


s 


Rhytidosporum diosmoides 


Sub-shrub 


s 


Schizaea bifida 


Fern 


R 


Schoenus sp. 


Graminoid 


R 


Sphaerolobium vimineum 


Shrub 


R 


Sprengelia incarnata 


Shrub 


S 


Thysanotus tuberosus 


Graminoid 


R 


Trachymene incisa ssp. incisa 


Forb 


R 


Xanthorrhoeajohnsonii 


Graminoid 


R 


Xvris nperculnta 


Graminoid 


R m^ 



38 




Proc. Liim. Soc. N.S.W., 127, 2006 



Habitat Segregation of Banksia Shrubs at Gibraltar Range 

National Park 

Shanti Virgona, Glenda Vaughton and Mike Ramsey 

Botany, School of Environmental Sciences and Natural Resources Management, University of New England, 

Aimidale NSW 2351 (gvaughto@une.edu.au) 



Virgona, S., Vaughton, G. and Ramsey, M. (2006). Habitat segregation of Banksia shrubs at Gibraltar 
Range National Park. Proceedings of the Linnean Society of New South Wales 127, 39-47. 

Events during seedling recruitment affect species' distributions, causing habitat segregation of congeneric 
species within the same area. We documented the segregation of Banksia marginata and B. spinulosa 
var neoanglica in adjacent swamp and woodland habitats at two sites by surveying adult and seedling 
distributions. We also examined seed banks and seed characters as factors contributing to segregation. 
Habitat segregation was pronounced, with 92% of B. marginata adults located in swamps and 98% of B. 
spinulosa adults located in woodlands. After fire, 84% of 5. marginata seedlings were in swamps, but 10 
months later this increased to 93%, indicating that although seeds dispersed to and germinated in adjacent 
woodlands, most seedlings failed to establish. Seedlings of B. spinulosa were confined to woodlands, 
indicating that seeds did not disperse into swamps or that, if they did, seeds failed to germinate or seedlings 
suffered early mortality. Canopy seed banks of both species were large (> 280 seeds/plant) and seeds 
of both species possess membranous wings, allowing dispersal between habitats. Overall, neither limited 
numbers of seeds nor limited seed dispersal are likely to cause habitat segregation. Instead, processes 
occurring during early seedling growth are probably more influential. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: Banksia marginata, Banksia spinulosa, fire, niche, Proteaceae, regeneration. 



INTRODUCTION 

Seedling recruitment is a critical stage in 
the demography of plant populations. Seeds and 
seedlings typically experience high mortality rates, 
and events during recruitment potentially affect 
the distribution of species in the landscape (Harper 
1977; Grime 1979; Silvertown and Chariesworth 
2001). When seeds are dispersed, they encounter a 
variety of abiotic and biotic conditions that affect the 
seed germination, seedling emergence, survival or 
growth stages of the life cycle. Interactions between 
these factors exert powerful effects on the spatial 
patterns of recruitment, allowing regeneration to 
occur in some microhabitats but not others (Lament 
et al. 1989; Mustart and Cowling 1993; Schutz et al. 
2002; Castro et al. 2004). Grubb (1977) coined the 
term "regeneration niche" to distinguish between the 
habitat conditions required for seedling recruitment 
as opposed to adult survival and reproduction. The 
regeneration niche is generally considered to be 
more complex than the niche experienced by adult 



plants, providing substantial opportunities to explain 
the distribution patterns of different species (Grubb 
1977). 

Habitat segregation of congeneric species in 
distinct habitats within the same general geographic 
area is a common feature of coastal plains and 
tableland areas of southern Australia (Siddiqi et al. 
1972; Bowman et al. 1986; Keith and Myerscough 
1993; Myerscough et al. 1995; Clarke 2002), and 
has been well documented in Banksia. On the Swan 
coastal sand plain near Perth, Banksia littoralis is 
restricted to swamp margins, whereas B. menziesii 
and B. attenuata occur more widely in drier woodland 
areas (Groom et al. 2001). Further north of Perth on 
the Eneabba sandplain, B. hookeriana, B. prionotes 
and B. attenuata are segregated along topographic 
gradients in the dune-swale system (Lamont et al. 
1989; Groeneveld et al. 2002). In NSW, on the coastal 
sand plains in the Myall Lakes area, B. oblongifolia 
and B. aemula are segregated into either wet heath 
or dry heath occurring on the slopes and ridges, 
respectively (Myerscough et al. 1996). Finally, on 
the north coast of NSW, B. ericifolia is most common 



HABITAT SEGREGATION OF BANKSIA SHRUBS 



in wet heath, while its congener B. aemula occurs in 
headland heath, dry heath and moist heath (Benwell 
1998). Although such patterns have been attributed 
to physiological intolerance and competition at later 
life-cycle stages, increasing evidence points to the 
importance of processes occurring during recruitment 
in mediating segregation (Myerscough et al. 1996; 
Clarke et al. 1996; Williams and Clarke 1997). Using 
field experiments Myerscough et al. (1996) and 
Clarke et al. (1996) demonstrated that segregation 
of 5. aemula and B. oblongifolia at Myall Lakes was 
related to processes operating during seed dispersal 
when safe sites for seeds are required, and during 
establishment when resources are critical for early 
seedling growth. 

At Gibraltar Range National Park (GRNP), 
sedge-heath communities occur in areas of impeded 
drainage on hillsides and on flats that are surrounded 
by a matrix of sclerophyll woodland. Segregation of 
congeneric species into either sedge-heath or woodland 
habitats is common (Sheringham and Hunter 2002). 
Here we quantified the distribution of both adult and 
seedling populations of Banksia marginata and B. 
spinulosa var. neoanglica to determine whether the 
two species are segregated and whether processes 
operating during recruitment mediate this pattern. 
We found pronounced segregation that is established 
during recruitment, and to explain the observed 
distribution pattern we examined the seed production 
and seed dispersal stages of the life cycle. Specifically, 
we examined seed bank size and seed attributes to 
determine whether both species produce viable seeds 
and whether seeds have the potential to disperse 
between habitats. 



MATERIALS AND METHODS 

Study species 

Banksia marginata Cav. is widely distributed 
along the coast and ranges of south-eastern Australia. 
At GRNP, the species is at the northern limit of its 
range (Harden 2002). In this area, plants are killed by 
fire and populations rely solely on seeds for recovery 
(i.e. are obligate seeders, Vaughton and Ramsey 1998; 
Benson and McDougall 2000). Adult plants are single- 
stemmed and grow to 2 m in height. The inflorescences 
are 5-10 cm long and bear straight styles. Flowers are 
self-compatible and two seeds are usually formed 
per follicle (Vaughton and Ramsey 1998, 2006). At 
this site follicles are strongly serotinous and open 
only after exposure to high temperatures during 
fires. Seeds have a membranous wing, allowing wind 



dispersal. Most seedling recruitment occurs in the 
first 12 months after fire. 

Banksia spinulosa Sm. var. neoanglica A.S. 
George (5. cunninghamii Sieber ex Rchb. subsp. 
A; Harden 2002; hereafter B. spinulosa) is found 
in northern NSW and southern Queensland. Plants 
have an underground lignotuber and are able to 
resprout after fire (i.e. are resprouting). Adult plants 
are multistemmed and grow to 2 m in height. The 
inflorescences are 6-15 cm long and bear hooked 
styles. Flowers are self-compatible, but most seeds 
are outcrossed (Vaughton and Carthew 1993). The 
follicles are strongly serotinous and have a single 
winged seed (Vaughton and Ramsey 2001). Seedling 
recruitment occurs after fire. 

Study sites 

Two sites were chosen at GRNP that were 
burned by bushfires during November 2002: Waratah 

Trig (WT: 29^29' S, 152^19' E, 1050 m a.s.l.) and 

Surveyors Creek (SC: 29o32' S, 152018' E, 1044 m 
a.s.l.). These sites occur on separate drainage channels 
and are approximately 5 km apart. Both sites are 
floristically and structurally similar, comprising two 
associated vegetation types, sclerophyll woodland 
and sedge-heath. The woodlands are dominated by 
Eucalyptus olida with a diverse shrub understorey. 
The sedge heaths are dominated by Lepidosperma 
limicola with emergent shrub species occurring along 
the swamp margins (Sheringham and Hunter 2002). 

Distribution of adults and seedlings 

For both species, the distributions of adult plants 
were surveyed 3 months after the fire. Burnt adults 
retained their cones, and were readily identified. 
Seedling distributions were surveyed twice, at 1 8 and 
28 months after the fire. Seedlings of the two species 
were identified by differences in cotyledon and leaf 
traits. 

We used stratified belt-transects for the adult 
and seedling surveys. Different transects were used 
for each survey, ensuring samples were independent 
of each other. Transects sparmed 40 m of the swamp 
habitat and 60 m of the woodland habitat, and were 
haphazardly placed. This stratification ensured that we 
surveyed the distributions of both B. marginata and 
B. spinulosa. Transects excluded the wetter regions of 
the swamp where banksias rarely occur (Sheringham 
and Hunter 2002). For adult and seedling surveys at 
each site, four, 5 m wide transects and six, 1 m wide 
transects were used, respectively. Within transects, all 
banksias were counted and recorded as being in either 



40 



Proc. Linn. Soc. N.S.W., 127, 2006 



S. VIRGONA, G. VAUGHTON AND M. RAMSEY 




0-20 



21-40 ' 41-60 ' 61-80 
Transect locations (m) 

Swamp ►"^ Woodland 



81-100 



Figure 1. Vegetation profile of tlie swamp and woodland 
association and the placement of stratified belt-transects 
used to examine tlie distribution of B. marginata and B. 
spinulosa at the SC and WT sites. The illustration is rep- 
resentative of the vegetation and is not drawn to scale. 



Similarly, the number of seeds per plant was 
estimated as the product of cones per plant and 
seeds per cone. Plant density was assessed using 
five haphazardly placed quadrats in each habi- 
tat at each site. In swamp and woodland habi- 
tats, 5 m X 5 m and 10 m x 10 m quadrats were 
used, respectively, to ensure both species were 
adequately represented. To assess the number 
of cones per plant, we counted cones with fol- 
licles on 50 plants of each species per site. For 
the number of viable seeds, we collected five 
cones fi-om each of 20 plants of each species 
per site prior to the fire in August 2002. Seeds 
were extracted in April 2003 by heating cones 
for 30 min in an oven set at 1 10 °C. Seeds were 
deemed viable if they were intact and plump. 



the swamp or woodland habitat (Fig. 1). 

To assess habitat segregation of 5. marginata and 
B. spinulosa, we calculated an analysis of deviance 
using a logit model with a binomial error term and 
a logit link fiinction. The three explanatory variables 
were site (SC, WT), plant age (adults, seedlings at 
1 8 and 24 months after the fire) and habitat (swamp, 
woodland). The response variable was the number of 
plants of one species expressed as a proportion of the 
sum of the number of plants of both Banksia species. 
With this approach, the explanatory variables and 
their interactions are interpreted as their effect on the 
response variable in a similar fashion to ANOVA. In 
preliminary analyses, the three-way interaction (site 
x habitat x age) and the site x age interaction were 
not significant and were omitted fi-om the final model 
{P > 0.50). We also calculated the percentage of the 
total deviance explained by each term. Because the 
age x habitat interaction was significant in the final 
model, we also examined how the distributions of 
adults and seedlings differed with respect to habitat 
for each species at each site using 3x2 contingency 
tables (G-tests). A significant G-test indicates that the 
distribution of plants of the different ages depends on 
habitat. We used a sequential Bonferroni correction to 
account for multiple tests. 

Seed bank 

2 
We estimated the seed bank (seeds/m ) of 

each species as: (plants/m ) x (mean number of 

cones/plant) x (mean number of viable seeds/cone). 



Seed dispersal 

To assess the dispersal potential of seeds, 
seed mass:seed area ratios (i.e. wing loading) 
were calculated; larger wing loadings imply 
shorter potential dispersal distances. Seeds 
were weighed and measured with the seed and wing 
intact, using three seeds fi-om each of 10 plants per 
species fi-om each site. Seeds were weighed to the 
nearest 0. 1 mg, and seed area was determined using a 
leaf area meter (A T Area Meter, Delta - T Devices). 

Seed bank and dispersal analyses 

Data were analysed with two-way ANOVAs, 
with species as a fixed factor and site as a random 
factor. When the site x species interaction was not 
significant {P > 0.2), it was pooled with the error term, 
resulting in a more powerful test of the differences 
between species (Quinn and Keough 2002). When 
the site x species interaction was significant, we 
calculated tests of simple main effects comparing 
the species at each site. To meet assumptions of 
ANOVA, plant density, the numbers of cones, seed 
mass and wing loadings were logio transformed. 
Other variables did not require transformation. 



RESULTS 

Distribution of adults and seedlings 

Of the 3610 plants found on transects, 15.5% 
and 61.7% were B. marginata adults and seedlings, 
respectively, and 3.2% and 19.6% were B. spinulosa 
adults and seedlings, respectively. The final logit 
model included the three main effects, and the age 
X habitat and the site x habitat interactions (Table 
1). The significant age x habitat interaction indicates 



Proc. Linn. Soc. N.S.W., 127, 2006 



41 



HABITAT SEGREGATION OF BANKSIA SHRUBS 



Table 1. Analysis of deviance (\ogit model) examining the dis- 
tribution of Banksia marginata and B. spinulosa adults and 
seedUngs (18 and 28 months after the fire) in swamp and 
woodland habitats at the SC and WT sites (n = 3610 plants). 
Habitat explained 97.7% of the total deviance. The site x 
habitat x age and the site x age interactions were not signifi- 
cant (P > 0.50), and they were omitted from the final model. 



Source 



df 



A Deviance 



Site 

Habitat 

Age 

Age X habitat 

Site X habitat 



1 



32.71 



2505 



11.33 



12.34 



2.97 



< o.oooT^ 

< 0.0001 

0.003 

0.002 



0.085 



that the relative frequencies of adults and seedlings 
varied between the two habitats (Table 1, Fig. 2). 
Although the main effects of site, age and habitat 
were significant, habitat singulariy explained 97.7% 
of the deviance in the model (Table 1), indicating 
that the two species exhibited pronounced habitat 
segregation. Overall, most B. marginata plants of all 
ages were found in the swamps (> 84%) and most 
B. spinulosa plants were found in the woodland (> 
95%). 

Fori?, marginata, adult and seedling distributions 
differed significantly at both sites (Table 2, Fig. 
2). Seedlings were distributed more widely than 
adults 18 months after the fire, but by 28 months, 
the distributions were similar. At 18 months, about 
84%) and 16%) of seedlings were found in the 
swamps and woodlands, respectively, but 10 
months later, about 93% and 1% of seedlings 
were in the two habitats, respectively. For B. 
spinulosa, adult and seedling distributions at 
WT did not differ significantly (Table 2, Fig. 
2). At SC, however, adults were distributed 
more widely than seedlings (Table 2, Fig. 
2). About 5% of adults were located in the 
swamp, whereas > 99%) seedlings were 
confined to the woodland. 

Seed bank 

2 

For plant density (plants/m ), the site x 
species interaction was not significant (Fi,i6 
= 0.48, P > 0.450), and it was pooled with 
the error term for the final analysis. Sites 
did not differ (F, j^ = 0.07, P > 0.750), but 
B. marginata plants were about 10-fold more 



numerous than B. spinulosa plants 
(Fii7= 173.67, P < 0.001; Table 3). 
For the number of cones per plant and 
viable seeds per cone, the site x species 
interaction was significant (cones: 
Fi,i96 = 16.85, P< 0.001; seeds: Fii9g= 
2 1 . 1 5, P < 0.00 1), and we examined the 
simple main effects of species at each 
site. The number of cones produced 
by B. marginata was significantly 
greater than B. spinulosa at both sites 
(WT: Fj 196= 103.49, P < 0.001; SC: 
Fi 196 =' 19.08, P < 0.001; Table 3). 
For the number of seeds per cone, B. 
marginata produced more seeds than 
B. spinulosa at WT (Fi ^^ = 24.25, P 
< 0.001), but at SC seed production 
of both species was similar (Fj ^g = 
2.49, P = 0. 109; Table 3). Overall, B. 
marginata plants produced about 9- 
fold and 1.4-fold more seeds than did B. spinulosa 
plants at WT and SC, respectively. Similarly, the seed 
bank (seeds/m ) of B. marginata was 182-fold and 
1 8-fold greater than that of B. spinulosa at WT and 
SC, respectively, the large differences resulting from 
differences in plant density. 

Seed dispersal 

For seed mass and seed area, the site x species 
interactions were significant (seed mass : Fiiig=11.13, 
P< 0.001; seed area: Fi 11^= 15.06, P < 6.001), and 
we examined simple main effects of species at each 
site. Seeds ofB. spinulosa weighed at least 22% more 
than B. marginata seeds at both sites (WT: Fi ug = 



Table 2. Results of 3 x 2 contingency analyses examining 
the effects of habitat on the distribution of plants of differ- 
ent ages. We compared the distributions of adults (A) ver- 
sus seedlings surveyed 18 and 28 months after the fire (SI 
and S2, respectively) between swamps and woodlands for 
Banksia marginata and B. spinulosa at the SC and WT sites. 
G- and P-values are presented. G-values are significant 
following Bonferroni correction at P = 0.0125. All df = 2. 



Population 




B. marginata 


B. spinulosa 


m 


m 


Avs Si vs 


S2 


Avs Si vs S2 


SC ■ 


H 


1 17.54 




10.07 U 




p 


0.0002 




0.0065 " 


WT 


G 


23.27 




0.15 




P 


< 0.0001 




0.926 



42 





Proc. Linn. Soc. N.S.W., 127, 2006 



S. VIRGONA, G. VAUGHTON AND M. RAMSEY 



1 
0.8-1 

I 0.4 
0.2-1 







SC adults 



fh 



ffi 



^ 



1 
0.8-1 
0.6 
0.4- 
0.2-^ 







WT adults 



m 



I I i ~ 






d 
o 

o 



In 
0.8 

0.6 

0.4 -j 
0.2 




SC seedlings 18 months 



Q 



r — • — T' — ' — T- 



In 



0.8 



0.6- 



0.4 



0.2 







WT seedlings 18 months 



T^ — r — — V — ' — T- 



1-, 

0.8 

O 

a, 

0.2-1 







SC seedlings 28 months 



^ A' s^ uS 



f .-f .J' .J' .# 



Intervals (m) 



^^ 



1-, 



0.8- 



0.6 



0.4 



0.2 







WT seedlings 28 months 

i- 



a 



jn 



J' ^ ^ <^ ^^ 



Intervals (m) 



*" ^ 



Figure 2. Distribution of Banksia marginata (open) and B. spinulosa (shaded) adults and seed- 
lings at the SC and WT sites. Adults were surveyed 3 months after the fire, and seedlings were sur- 
veyed 18 and 28 months after the fire. Data are mean proportions of plants (± SE) at 20 m in- 
tervals based on four transects for adults and six transects for seedUngs. The 0-20 m and 21-40 
m intervals were in swamp habitats and the other three intervals were in woodland habitats. 



Proc. Linn. Soc. N.S.W., 127, 2006 



43 



HABITAT SEGREGATION OF BANKSIA SHRUBS 

Table 3. Seed bank characters for Banksia marginata and B. spinulosa at the SC and WT sites. Data 
are means ± SE. 



Characters 


^1 


t 


WT 


m 




B. marginata 


B. spinulosa 


B. marginata 


B. spinulosa 


2 
Plants/m 


0.9 ±0.2 


0.1 ±0.1 


1.0 ±0.3 


0.1 ±0.1 


Cones/plant 


17.2 ±1.8 


9.2 ±1.2 


45.8 ±4.5 


9.4 ± 0.9 


Seeds/cone 


24.4 ±3.8 


33.3 ±3.8 


57.7 ±3.5 


30.0 ±4.7 


Seeds/plant 


420 ± 72 


306 ±45 


2642 ± 220 


282 ± 47 


2 
Total seeds/m 


386 ±53 


21±2 


2736 ±271 


15±2 



15.74, P< 0.001; SC: Fiiig= 75.42, P< 0.001; Table 
4). By contrast, B. marginata seeds were at least 11% 
greater in area than B. spinulosa seeds at both sites 
(WT: Fi 1,6 = 51.47, P < 0.001; SC: Fj ,i6 = 2.84, P < 
0.01; Table 4). 

For wing loading, the site x species interaction 
was not significant (Fj ],g = 1.51, P > 0.20), and it 
was pooled with the error term for the final analyses. 
No differences in wing loadings were found between 
sites (F, 117 = 2.04, P = 0.156). Wing loadings of 5. 
spinulosa seeds were significantly greater than the 
wing loadings of B. marginata seeds, indicating that 
B. spinulosa should disperse shorter distances than B. 
marginata (Fj n^ = 201.06, P < 0.001; Table 4). 



DISCUSSION 

Our results showed that B. marginata and B. 
spinulosa were segregated into different habitats, 
and that this pattern was established during seedling 
recruitment. Adult B. marginata plants were 



concentrated in the swamp margins, whereas B. 
spinulosa plants were located in the woodland. In 
B. marginata, seedlings were more widely dispersed 
than adults 18 months after fire, but by 28 months 
the distribution of seedlings had contracted so that 
it did not differ from that of adults. In B. spinulosa, 
virtually all seedlings were confined to the woodland 
habitats. Our results are consistent with other studies 
showing the importance of the regeneration niche in 
determining patterns of segregation of congeneric 
species (Lamont et al. 1989; Mustart and Cowling 
1993; Myerscough et al. 1996; Clarke et al. 1996; 
Williams and Clarke 1997; Schutz et al. 2002). 

In B. marginata, processes operating during 
seedling establishment appear to mediate habitat 
segregation. The wider distribution of 5. marginata 
seedlings than adults 18 months but not 28 months 
after fire, indicates that seeds dispersed into the 
woodland and germinated, but seedlings failed to 
establish. Other studies have shown that recruitment 
of Banksia seedlings is strongly influenced by abiotic 
conditions, especially drought (Lamont et al. 1989; 



Table 4. Seed characters relevant to dispersal for Banksia marginata and B. spinulosa at the 
SC and WT sites. Data are means ± SE (n = 30 seeds). 



Site 


Species 


Seed mass 
("ig) 


Seed area 
(cm ) 


Wing loading 
(mg/cm ) 


SC 


B. marginata 


8.44 ±0.31 


0.72 ± 0.04 


11.95 ±0.49 




B. spinulosa 


12.85 ± 0.44 


0.65 ± 0.07 


19.56 ±1.04 


WT 


B. marginata 


8.85 ±0.22 


0.85 ±0.03 


10.64 ±0.33 




B. spinulosa 


10.81 ±0.32 


0.57 ± 0.04 


18.51 ±1.07 




44 



Proc. Lirni. Soc. N.S.W., 127, 2006 



S. VIRGONA, G. VAUGHTON AND M. RAMSEY 



Myerscough et al. 1996; Lamont and Groom 1998). 
At GRNP, soils in both the swamp and woodland 
habitats are of granitic origin. However, the swamp 
soils are fine textured and poorly drained compared 
with the more well-drained soils of the woodland 
(Virgona 2004). Segregation therefore may be driven 
by adaptation to the particular abiotic conditions 
in the swamp and intolerance to conditions in the 
woodland. Specifically, compared with B. spinulosa, 
B. marginata seedlings may be less tolerant of 
fluctuating levels of soil moisture and thus the overall 
drier soils in the woodland. Manipulative field and 
glasshouse experiments are now needed to investigate 
this possibility. 

As expected of an obligate seeder that relies solely 
on seeds for recruitment (Lamont and Groom 1998; 
Lamont and Wiens 2003), B. marginata maintained a 
large canopy seed bank. Compared with B. spinulosa, 
adult5. marginata plants had 2-4 times more cones per 
plant, and at WT, more viable seeds per cone. Banksia 
marginata adults also occurred at higher densities 
than did B. spinulosa, resulting in seed densities that 
were 18-182 times greater in their preferred habitat. 
High seed densities provide maximum opportunities 
for seedling recruitment and also allow colonisation 
of new sites, as evidenced by the occurrence of B. 
marginata seedlings in woodland habitat 1 8 months 
after fire. Populations of obligate seeding plants, 
however, are susceptible to either short or very long 
fire intervals that can decrease the amount of stored 
seeds for recruitment and hence threaten population 
persistence (Morrison et al. 1995; Enright et al. 
1996). The effect of fire on demographic processes 
may therefore interact with other abiotic and biotic 
factors influencing segregation in fire-prone heaths 
and woodlands such as those occurring at GRNP. 

In B. spinulosa, > 99% of seedlings were located 
in the woodland 18 months after the fire. The lack 
of recruitment into swamps indicates that either seed 
availability was limited, seeds did not disperse into 
the swamps or seeds dispersed, but seedlings failed 
to establish. Although at present we are unable to 
distinguish between these possibilities, we suspect the 
latter. First, recruitment by B. spinulosa in swamps 
was unlikely to be limited by seed availability. 
Compared with other resprouting Banksia species, 
adult B. spinulosa plants maintained a large store of 
seeds in their canopy prior to the fire (~ 300 seeds for 
B. spinulosa vs < 16 seeds, n = 7 species, Lamont and 
Groom 1998). Further, seedling recruitment occurred 
in woodland but not adjacent to swamp habitats, 
indicating that seed availability was adequate. Second, 
the inability of seeds to disperse fi^om the woodland 
is unlikely to account for the absence of seedlings in 



the swamps. Although B. spinulosa seeds had larger 
wing loadings (greater mass but smaller area) than 
B. marginata seeds, this may not overly affect their 
dispersal potential. Using wind-tunnel and seed- 
release experiments, Hammill et al. (1998) reported 
that Banksia seeds weighing between 9 mg and 70 mg 
dispersed similar distances. They found that seeds of 
Banksia species with similar characteristics to those 
of 5. spinulosa and B. marginata were most abundant 
within 2-3 m of parent plants, but were readily 
dispersed 9-12 m and occasionally up to 40 m from 
parents. Assuming that B. spinulosa seeds behave as 
in Hammill 's study, the absence of seedlings in the 
swamps is unlikely to be due to the inability of seeds 
to disperse the short distance from the woodland to 
the swamp margins. 

At SC, about 5% of B. spinulosa adults were 
located in the swamp, but no seedlings were found 
in this habitat. These adult plants had canopy seed 
banks and would have released their seeds directly 
into the swamps. The lack of seedling establishment 
in the swamps may be due to the inability of seeds 
to find safe sites. At Myall Lakes, B. aemula, a dry 
heath species, failed to establish in nearby wet heath, 
even though experiments showed that seedlings were 
able to grow in this habitat. The lack of establishment 
in wet heath under natural conditions was attributed 
to insufficient soil disturbance, which reduced safe 
sites for seeds (Myerscough et al. 1996; Clarke et al. 
1996). Little or no recruitment in the swamps would 
be expected if B. spinulosa seeds experience similar 
problems in finding safe sites. 

If seeds of B. spinulosa germinated in the 
swamps, but seedlings died shortly afterwards, then 
they would not have been present when we surveyed 
the swamps 18 months after fire. Early mortality 
could have been mediated by abiotic conditions in the 
swamps. Compared with B. marginata, seedlings of 
B. spinulosa may lack the physiological capacity to 
cope with the poorly draining and often waterlogged 
swamp soils. Further, early seedling mortality could 
result from competitive exclusion. Seedlings of B. 
spinulosa grow more slowly than seedlings of B. 
marginata (Virgona 2004), and they may be unable 
to compete with the rapidly resprouting sedges 
in the swamps. Despite these potential obstacles, 
establishment of B. spinulosa in the swamps must 
occur occasionally as evidenced at SC by the presence 
of the small number of adult plants in this habitat. 
Given the ability of B. spinulosa plants to persist by 
resprouting, seedlings only need to be recruited rarely 
to maintain current plant densities. 



Proc. Linn. Soc. N.S.W., 127, 2006 



45 



HABITAT SEGREGATION OF BANKSIA SHRUBS 



ACKNOWLEDGEMENTS 

We thank M. Campbell and P. Clarke for comments on the 
manuscript and thank S. Cairns for providing statistical 
advice. Financial support was provided by UNE and a 
Botany NCW Beadle Scholarship to S.V. 



REFERENCES 

Benson, D. and McDougall, L. (2000). Ecology of Sydney 
plant species 7b: Dicotyledon families Proteaceae to 
Rubiaceae. Cunninghamia 6, 1017-1198. 

Benwell, A.S. (1998). Post-fire seedling recruitment in 
coastal heathland in relation to regeneration strategy 
and habitat. Australian Journal of Botany 46, 75-101. 

Bowman, D.M.J.S., Maclean, A.R. and Crowden, R. K. 
(1986). Vegetation-soil relations in the lowlands of 
southwest Tasmania. Australian Journal of Ecology 
11, 141-153. 

Castro, J., Zamora, R., Hodar, J.A. and Gomez, J.M. 
(2004). Seedling establishment of a boreal tree 
species (Pinus sylvestris) at its southernmost 
distribution limit: consequences of being in a 
marginal Mediterranean habitat. Journal of Ecology 
92, 266-277. 

Clarke, P.J. (2002). Habitat insularity and fire response 
traits: evidence from a sclerophyll archipelago. 
Oeco/ogza 132, 582-591. 

Clarke, RJ., Myerscough, RJ. and Skelton, N.J. (1996). 
Plant coexistence in coastal heaths: Between- and 
within- habitat effects of competition, disturbance 
and predation in the post-fire environment. Australian 
Journal of Ecology 21, 55-63. 

Enright, N.J., Lamont, B.B. and Marsula, R. (1996). 

Canopy seed bank dynamics and optimum fire regime 
for the highly serotinous shrub, Banksia hookeriana. 
Journal of Ecology 84, 9-17. 

Grime, J.P. (1979). 'Plant strategies and vegetation 
processes'. (Wiley and Son: Chichester). 

Groeneveld, J., Enright, N.J., Lamont, B.B. and Wissel, C. 
(2002). A spatial model of coexistence among three 
Banksia species along a topographic gradient in fire- 
prone shrublands. Journal of Ecology 90, 162-11 A. 

Groom, P.K., Froend, R.H., Mattiske, E.M. and Gumer, 
R.P. (2001). Long-term changes in vigour and 
distribution of Banksia and Melaleuca overstorey 
species in the Swan Coastal Plain. Journal of the 
Royal Society of Western Australia 84, 63-69. 

Grubb, P. (1977). The maintenance of species richness 
in plant communities: The importance of the 
regeneration niche. Biological Review 52, 107-145. 

Hammill, K., Bradstock, R. and Allaway, W. (1998). Post- 
fire seed dispersal and species re-establishment in 
Proteaceous heath. Australian Journal of Botany 46, 
407-419. 

Harden, G.J. (2002). 'Flora of New South Wales'. Vol II. 



(University of New South Wales Press, Sydney). 
Harper, J.L. (1977). 'Population biology of plants'. 

(Academic Press: London). 
Keith, D.A. and Myerscough, RJ. (1993). Floristics and 

soil relations of upland swamp vegetation near 

Sydney. Australian Journal of Ecology 18, 325-344. 
Lamont, B.B., Enright, N.J. and Bergl, S.M. (1989). 

Coexistence and competitive exclusion of Banksia 

hookeriana in the presence of congeneric seedlings 

along a topographic gradient. Oikos 56, 39-42. 
Lamont, B.B. and Groom, RK. (1998). Seed and seedling 

biology of the woody-firuited Proteaceae. Australian 

Journal of Botany 46, 387-406. 
Lamont, B.B. and Wiens, D. (2003). Are seed set and 

speciation rates always low among species that 

resprout after fire, and why? Evolutionary Ecology 

17, 277-292. 
Morrison, D.A., Cary, G.J., Pengelly, S.M., Ross, D.G., 

Mullins, B.J., Thomas, C.R. and Anderson, T.S. 

(1995). Effects of fire fi-equency on plant species 

composition of sandstone communities in the 

Sydney region: Inter-fire interval and time-since-fire. 

Australian Journal of Ecology 20, 239-247. 
Mustart, RJ. and Cowling, R.M. (1993). The role of 

regeneration stages in the distribution of edaphically 

restricted Fynbos Proteaceae. Ecology 74, 1490-1499. 
Myerscough, RJ., Clarke, RJ. and Skelton, N.J. (1995). 

Plant coexistence in coastal heaths: Floristic patterns 

and species attributes. Australian Journal of Ecology 

20, 482-493. 
Myerscough, RJ., Clarke, RJ. and Skelton, N.J. (1996). 

Plant coexistence in coastal heaths: Habitat 

segregation in the post-fire environment. Australian 

Journal of Ecology 21, 47-54. 
Quirm, G.R. and Keough, M.J. (2002). 'Experimental 

design and data analysis for biologists'. (Cambridge 

University Press, Cambridge). 
Schutz, W, Milberg, R and Lamont, B.B. (2002). 

Germination requirements and seedling responses 

to water availability and soil type in four eucalypt 

species. Acta Oecologica 23, 23-30. 
Sheringham, P. and Hunter, J.T (2002). 'Vegetation and 

floristics of Gibraltar Range National Park'. (NSW 

National Parks and Wildlife Service, Glen Innes). 
Siddiqi, M.Y., Carolin, R.C. and Anderson, D.J. (1972). 

Studies in the ecology of coastal heath in New South 

Wales. I. Vegetation structure. Proceedings of the 

Linnean Society of New South Wales 97, 21 1-224. 
Sivertown, J. and Charlesworth, D. (2001). 'Introduction 

to plant population biology'. 4th edition, (Blackwell 

Science, Oxford). 
Vaughton, G. and Carthew, S.M. (1993). Evidence for 

selective abortion in Banksia spinulosa (Proteaceae). 

Biological Journal of the Linnean Society 50, 35-46. 
Vaughton, G. and Ramsey, M. (1998). Sources and 

consequences of seed mass variation in Banksia 

marginata (Proteaceae). Journal of Ecology 86, 563- 

573. 
Vaughton, G. and Ramsey, M. (2001). Relationship 

between seed mass, seed nutrients, and seedling 



46 



Proc. Linn. Soc. N.S.W., 127, 2006 



S. VIRGONA, G. VAUGHTON AND M. RAMSEY 



growth in Banksia cunninghamii (Proteaceae). 
InternationalJournal of Plant Sciences 162, 599-606. 

Vaughton, G. and Ramsey, M. (2006). Selfed seed set and 
inbreeding depression in obligate seeding populations 
oi Banksia marginata. Proceedings of the Linnean 
Society of New South Wales 111, 19-26. 

Virgona, S.P. (2004). Habitat segregation of Banksia 
marginata and B. spinulosa. BSc Honours thesis. 
Botany, University of New England, Armidale. 

Williams, P.R. and Clarke, RJ. (1997). Habitat segregation 
by serotinous shrubs in heaths: Post-fire emergence 
and seedling survival. Australian Journal of Botany 
45,31-39. 



Proc. Linn. Soc. N.S.W., 127, 2006 47 



48 



Response of Resprouting Shrubs to Repeated Fires in the Dry 
Sclerophyll Forest of Gibraltar Range National Park 

KiRSTEN J. E. Knox''^ and Peter J. Clarke^ 

'Botany, School of Environmental Sciences and Natural Resources Management, University of New England, 

Armidale, 2351 (pclarkel@une.edu.au), 
^current address: Department for Environment and Heritage, PO Box 822, Clare, South Australia 5453. 

Knox, K.J.E. and Clarke, RJ. (2006). Response of resprouting shrubs to repeated fires in the dry 
sclerophyll forest of Gibraltar Range National Park. Proceedings of the Linnean Society of New South 
Wales 121, 49-56. 

Fire regimes affect survival and reproduction of shrub species in fire-prone vegetation such as occurs in 
Gibraltar Range National Park. The influence of fire regimes on resprouting shrubs is known for a range 
of species in coastal regions of Australia but is poorly known in montane sclerophyll communities. The 
fire responses of three Proteaceae shrubs {Banksia spinulosa, Hakea laevipes, Petrophile canescens) and 
a grasstree {Xanthorrhoea johnsonii) were measured after the wildfire of 2002 to determine whether: 1) 
storage organ size was related to post-fire growth and flowering response, 2) fire fi-equency influences post- 
fire mortality and if survival was related to the size of plant; 3) fire fi-equency influences the resprouting 
ability of plants, and 4) fire frequency affects pyrogenic flowering in the post-fire environment. We found 
the size of storage organs was positively related to post-fire sprouting in the three shrubs and to flowering 
in the grasstree. However, high fire fi-equency only affected the survival of Banksia spinulosa and decreased 
flowering in Xanthorrhoea johnsonii. Survival in all species ranged between 83 and 99% and it appears 
that the intervals between fires (7-22 years) had been sufficient for most adult plants to regain the ability to 
resprout. The ability of juvenile plants to develop the ability to resprout needs to be tested on seedlings that 
established after recent fires. 

Manuscript received IMay 2005, accepted for publication 7 December 2005. 

KEYWORDS: Fire frequency, fire regime, persistence, pyrogenic flowering, resource allocation 



INTRODUCTION 

The fire response of species is often simplified 
into resprouters and obligate seeders, but in reality 
a continuum fi-om 0-100% mortality of individuals 
within a population exists amongst species 
(Bellingham and Sparrow 2000; Vesk and Westoby 
2004; Clarke et al. 2005). Characteristics of a 
particular fire, distribution of size-classes and the 
physiological and anatomical features of a species will 
affect the percentage mortality in a population after 
fire. Shrub species capable of resprouting generally 
resprout from subterranean buds (lignotubers and 
roots suckers), but also occasionally from epicormic 
buds on aerial stems. Grasstrees on the other hand 
resprout via apical buds. The ability of an individual 
to resprout following fire depends on having adequate 
dormant buds and carbohydrate storage to facilitate 
resprouting (Bell 2001; Knox and Clarke 2005). 
Variation in mortality has been observed for different 
size-classes within populations (Morrison 1995; 



Bond and Van Wilgen 1996). Some species have been 
found to have greater resprouting potential in larger 
size-classes (e.g. Morrison 1995); in contrast, some 
species have been found to have greater resprouting 
potential in smaller size-classes (e.g. Burrows 1985). 
Frequent fires with short inter-fire intervals may 
result in the exhaustion of buds or carbohydrates 
stored in the lignotuber, resulting in the mortality of 
resprouters (Zammit 1988; Bowen and Pate 1993). 
The intensity of a particular fire can influence what 
proportion of a population survives. Some obligate 
seeders may survive a low-intensity fire if 100% 
leaf scorch does not occur (Gill 1981; Bond and van 
Wilgen 1996). On the other hand a very high-intensity 
fire may result in the death of a large number of 
individuals within a population that usually resprouts 
following fire. The minimum fire-tolerant stem size 
of resprouters often increases with fire intensity for 
some species (Bradstock and Myerscough 1988; 
Morrison 1995; Morrison and Renwick 2000). 



RESPONSE OF RESPROUTING SHRUBS TO REPEATED FIRES 



Resprouters that recruit seedlings into populations 
following fire generally have seed stored in the soil or 
in the canopy in woody fruits. Hence, understanding 
the post-fire growth and reproductive response 
of resprouting shrubs is critical in determining 
appropriate fire regimes in landscapes dominated 
by resprouting species. Resprouting shrubs typically 
have greater growth and reproductive vigour in the 
year following a fire (e.g. Auld 1987; Bowen and Pate 
2004). In many resprouting shrubs flowering occurs 
predominantly, or exclusively, following fire, e.g. 
Telopea speciosissima (Pyke 1983), Lomatia silaifolia 
(Denham and Whelan 2000), Xanthorrhoea preissii 
(Lamont et al. 2000), and Stirlingia latifolia (Bowen 
and Pate 2004). Little is known about the factors that 
influence reproductive output of pyrogenic flowering 
plants, although season of fire is known to strongly 
influence flowering in some Western Australian 
species (Lamont et al. 2000; Bowen and Pate 2004). 
One important component of fire regime that is likely 
to influence post-fire flowering is the frequency of 
bums as this may influence the starch storage capacity 
of plants. 

Little quantitative work has been conducted to 
determine the effects of frequent fires on post-fire 
performance and mortality of shrubs that resprout 
following fire. The fire regime in Gibraltar Range 
National Park provided an opportunity to examine 
these questions because records date back to the 
1960s and the number and extent of subsequent 
fires have spatially explicit records. Gibraltar Range 
National Park also has widespread and abundant 
populations of resprouting shrubs occurring within 
physiographically similar landscapes. In late 2002 an 
intensive crown fire burnt most of the dry sclerophyll 
forest in the National Park. This event afforded an 
opportunity to study the post-fire response of species, 
which have experienced different fire firequencies. 

Evidence of an effect of fire fi-equency would 
show that more frequently burnt sites had more dead 
plants and surviving plants with reduced grow1:h and 
reproduction. If, however, these sites had smaller 
plants, then these may appear to showreduced survival, 
growth and reproduction purely for allometric reasons. 
Hence we asked whether: 1) storage organ size was 
related to post-fire growth and flowering response, 
2) fire fi:equency influences the resprouting ability 
of plants 3) fire firequency and/or size of the storage 
organ influences post-fire mortality 4) fire frequency 
affects pyrogenic flowering (flower or inflorescence 
production) in the post-fire environment. 



METHODS 

Fire regime maps of Gibraltar Range National 
Park were examined and dry sclerophyll forest areas 
that had been burnt twice, four and five times since 
1 964 were identified. All sites were burnt in November 
2002 by an intense wildfire that removed most of the 
tree leaf canopy but did not totally incinerate the 
fruits of the target species. The minimum interval 
between fires was approximately seven years and the 
maximum 22 years. Fire records showed that all fires 
burnt in spring/summer, suggesting that all fires were 
of high intensity. All observations were at the same 
time since the last fire (8 months). In areas of each of 
the fire frequency regimes two patches were chosen 
that were at least 1 km apart. In each patch three 500 m 
transects were established and the post-fire response 
of three species of Proteaceae shrubs with canopy- 
held seed banks {Banksia spinulosa, Hakea laevipes, 
and Petrophile canescens) were measured. These 
species were selected because they are ubiquitous, 
easy to identify when dead, and they only recruit after 
fire, hence their minimum age can be estimated. For 
each species the number of shoots resprouted fi^om the 
lignotuber, the length of the longest shoot resprouted 
and the basal girth of the lignotuber were measured for 
the first 20 (approx.) individuals encountered in each 
transect. Dead plants were also recorded and their 
basal girth measured. Individuals were identified by 
their 'skeletal' remains and their canopy-held woody 
finits. In addition, the post-fire flowering of the 
grasstree Xanthorrhoea johnsonii was also recorded 
along each transect. Vox Xanthorrhoea basal girth and 
height of the caudex were measured and the presence 
and length of the inflorescence were recorded for the 
first 20 individuals encountered along each transect. 

Data for this study were mainly collected by 
undergraduate students. All students collected the 
equivalent amount of data fi-om each of the fire 
frequency areas. This was important so that the 
patterns in post-fire resprouting and flowering could 
be attributed to the different fire frequencies, and not to 
variation in sampling among different students. In each 
of the three species of shrubs, to test the relationship 
between storage organ size and post-fire response, 
basal girths were regressed against the number of 
shoots resprouted, height of shoots resprouted and 
size of inflorescence as independent variables. We 
then used analysis of covariance (ANCOVA) to test 
if number of shoots or stem height were reduced by 
fire firequency, with lignotuber size as the covariate. 
We also used ANCOVA in Xanthorrhoea to test if the 
inflorescence length was reduced by fire firequency, 
with the caudex size as a covariate. Plots of residuals 



50 



Proc. Linn. Soc. N.S.W., 127, 2006 



K.J.E. KNOX AND P.J. CLARKE 



established that no transformations of raw data were 
necessary. Homogeneity of slopes was determined 
by testing the interaction between the covariate and 
the main factors. We next tested the hypothesis that 
storage organ size and/or fire frequency affects post- 
fire survival by logistic regression using likelihood 
ratio tests. In these analyses the response variable is 
the number of plants alive or dead. In Xanthorrhoea, 
we also tested the hypothesis that caudex volume 
and/or fire fi-equency affects post-fire flowering by 
logistic regression using likelihood ratio tests. In this 



analysis the response variable was the number of 
plants flowering or not flowering. 

RESULTS 

Of the four species sampled, only the grasstree 
Xanthorrhoea johnsonii was observed to flower in 
the immediate post-fire period (August 2003), whilst 
the other species began to flower in the following 
year (August 2004). All resprouting Proteaceae shrub 
species had a positive and significant {P < 0.05) 



120 - 
100 - 

80 - 

60 - 

40 

20 1 



a) Banksia 




r^"" — I — """I — """T — "— 1 — ■— 1 — "-T — ■— 1 — ""T — •—I 

10 20 30 40 50 60 70 80 90 
Girth (cm) 




-I 1—1 1— I 1— I r— 1 1—1 1— I 1— 1 1— I 

10 20 30 40 50 60 70 80 90 
Girth (cm) 



o 
o 



c 
o 



O) 

X 



b) Hakea 




1 — "—I — f— I — >— 1 — "—I — <—\ — ■— I 
20 30 40 50 60 70 80 90 
Girth (cm) 



i 60 1 
o 50 - 


e 

• 


c40- 
o 

% 30 - 
o 

t 20- 

|io- 

1 oJ 


e 


iiijiifmnpu o ''^ o ' " 


1 ■ 1 • 1 ■ 1 ■ 1 




10 20 30 40 50 60 70 80 90 




Girth (cm) 



c) Petrophile 




30 40 
Girth 

Figure 1. Regression of shoot height and number with lignotuber basal girth with 18 months after fire 
for a) Banksia spinulosa, r^ = 0.77, 0.59; b) Hakea laevipes, r^ = 0.77, 0.70; and c) Petrophile canescens 
r^ = 0.76, 0.77, across all fire frequencies of fire at Gibraltar Range National Park. 



Proc. Linn. Soc. N.S.W., 127, 2006 



51 



RESPONSE OF RESPROUTING SHRUBS TO REPEATED FIRES 



Table 1. Summary results for analysis of covariance for height (Ht) and numbers 
of shoots (Shoot) resprouted for Banksia spinulosa, Hakea laevipes, Petrophile ca- 
nescens and length of inJBorescence (Infl.) in Xanthorrhoea johnsonii. The size co- 
variate for the three shrubs was basal girth and for the grasstree it was the cau- 
dex volume. *** indicates P<0.05, ** indicates P<0.01, and * indicates P<0,001 



£ 
o 

b 



Factors 



Fire frequency 
Size (covariate) 
Fire fq. x size 



B. spinulosa H. laevipes P. canescens X. johnsonii 

Ht Shoot Ht Shoot Ht Shoot Infl. 



NS NS 






NS NS 



NS NS 



NS 



24 - 
20 - 
16 - 
12 - 
8 - 
4 - 

-'■ 



a) Banksia 





NS NS NS * 



NS 



NS 



NS 



£ 






24 - 

20 - 

16 - 

12 - 

8 - 

4 - 

- 



b) Hakea 

n 2 Fires 
H 4 Fires 
H 5 Fires 



1 



1 







Dead 



Alive 



Dead 



Alive 



£ 
o 

J= 
b 



24 

20 1 

16 

12 - 
8 - 
4 - 




c) Petrophile 




Dead 



T , T 



Alive 



£ 
o 

Of 

£ 
> 

X 
Of 

■o 

O 



30000 1 



25000 



20000 



1 5000 - 



10000 - 



5000 - 







d) Xanthorrhoea 

D 2 Fires 
n 4 Fires 
H 5 Fires 



Not flowering 



J 



1 



Flowering 



Figure 2. Mean (+ se) basal girth of lignotubers in each of three fire frequencies in Gibraltar Range National 
Parkfora)5fl«^s/as/7/«H/os«,b)jyaA^ea/aevipes,andc)Pefro/?^//ecflnesce«s.Meanvolumeofthecaudex(+se) 
for%««?Aorr^o^flyo^«so«Hforeachfire frequency wheresmallerplantsflowerin sites withlessfrequentfires. 



52 



Proc. Linn. Soc. N.S.W., 127, 2006 



K.J.E. KNOX AND P.J. CLARKE 



Table 2. Number of plants recorded in each fire frequency category and the results of logistic 
regression for fire frequency and size of storage organ from likeUhood ratio tests. *** indicates 
P<0.05, ** indicates P<0.01, and * indicates P<0.001 



Species 



Logistic response 
variable 



Fire frequency 



Chi- Chi-squared 

squared Fire Basal girth/ 

frequency volume 



B. spinulosa 


Dead 


10 


21 


24 




Alive 


96 


99 


108 


H. laevipes 


Dead 


4 


5 


3 




Alive 


116 


115 


116 


P. canescens 


Dead 


9 


2 


4 




Alive 


120 


121 


117 


X. johnsonii 


Not flowering 


103 


93 


110 




Flowering 


17 


7 


8 



10.7* 



0.5 NS 



0.8 NS 



8.4* 



16.6*** 



8.7=* 



7 9** 



67.3*** 



relationship between basal girth of the lignotuber and 
post-fire response of shoots (numbers and height) 8 
months after fire (Fig. labc). In addition, the volume 
of the caudex in Xanthorrhoea johnsonii was also 
positively related to the length of the inflorescence (r^ 
= 0.71). 

Fire frequency did not reduce the height and 
number of shoots resprouting when basal girth was 
used as a covariate (Table 1). Fire fi-equency, however, 
appeared to increase the height of Hakea laevipes 
which is not consistent with the hypothesis that fire 
fi'equency would reduce height. In shrub species, size 
and number of shoots resprouting were significantly 
related to basal girth. Hence, the apparent reduction 
in size and number of resprouted shoots in Banksia 
simply reflects the decreased size of lignotubers 
with fire frequency (Fig. 2). Fire firequency did 
not affect the length of the inflorescence in the 
grasstree Xanthorrhoea johnsonii and the length of 
the inflorescence was not significantly related to the 
caudex volume (Table 1). 

Next we ask if fire fi-equency and/or size of 
storage organ affect the survival of species. Only 
two of 358 Xanthorrhoea johnsonii plants were 
killed by fire, hence it was not possible to examine 
the relationship between survival and caudex size. 
Mortality was sufficiently high in the shrub species 
to examine the effects of fire fi-equency and size 
on post-fire survival using logistic regression. All 



species had an increased likelihood of mortality as 
lignotuber size decreased (Table 3, Fig. 2.). However, 
fire firequency only influenced mortality in Banksia 
spinulosa where increased fire frequency increased 
the likelihood of mortality (Table 2). 

Finally, we examined whether fire fi-equency 
and/or size of the caudex influenced flowering in 
Xanthorrhoea johnsonii. Both size of caudex and 
fire frequency influenced whether plants flowered or 
not with increased proportions of plants flowering 
when the caudex was large (Fig. 2) and increased 
proportions of plants flowering when fire frequency 
was low (Table 2). In addition only larger plants 
tended to flower in populations with a high fire 
fi-equency (Fig. 2). 

DISCUSSION 

In this study we examined the influence of storage 
organ size and fire fi-equency on post-fire mortality, 
resprouting vigour and flowering. Generally, we 
found that larger storage organ size was related to: (i) 
greater post-fire survival, (ii) more resprouting shoots 
(iii) faster grov/ing resprouting shoots, and (iv) the 
presence, and size of inflorescences iox Xanthorrhoea 
johnsonii. Fire fi-equency influenced the post-fire 
survival of Banksia spinulosa and also influenced the 
presence of inflorescences of Xanthorrhoea johnsonii 
following fire (Table 2). 



Proc. Linn. Soc. N.S.W., 127, 2006 



53 



RESPONSE OF RESPROUTEMG SHRUBS TO REPEATED FIRES 



Table 3. Summary of the results on the effect of fire frequency and size of storage organ on post-fire 
survival and resprouting of four species with canopy-held seed banks. 



Species 



Are storage organ size 
and post-fire growth 
response correlated? 



Does increased fire 

fi-equency affect 

storage organ? 



Does increased 

storage organ size 

affect post-fire 

survival? 



Does increased fire 

firequency affect 

post-fire survival? 



Banksia spinulosa 
Hakea laevipes 
Petrophile canescens 



Xanthorrhoea 
johnsonii 



Yes +ive 
Yes +ive 
Yes +ive 



Yes -ive 
No evidence 
No evidence 



Does storage organ Does fire frequency 
size affect presence of affect size of 

inflorescence? inflorescence? 



Yes +ive 
Yes +ive 
Yes + ive 

Does increased 

storage organ size 

affect post-fire 

survival? 



Yes -ive 
No evidence 
No evidence 

Does fire frequency 

affect presence of 

inflorescence? 



Yes +ive 



No evidence 



No evidence 



Yes -ive 



Post-fire survival for the species studied 
ranged from 99% for Xanthorrhoea johnsonii to 
82% for Banksia spinulosa, although we may have 
underestimated mortality if small dead plants were 
overlooked. Y or Xanthorrhoea johnsonii, individuals 
were found to survive fire irrespective of storage 
organ size. However, for the three shrubs, individuals 
with small storage organs were more likely to be 
killed by fire than those that had larger storage organs. 
This pattern is consistent with the idea that many 
species develop greater fire-tolerance as the storage 
organ increases in size and age (Morrison 1995; 
Keith 1996). This increased tolerance is likely to be a 
result of a larger dormant bud bank and more stored 
carbohydrate in less frequently burnt plants (Knox 
and Clarke 2005). We do, however, acknowledge 
that the sampling technique used might have 
underestimated the mortality of some populations 
if the fire intensity was great enough to incinerate 
fhiits. This could have resulted in an under-sampling 
of dead individuals, as retained woody fruits were 
used to help identify individuals. There could have 
also been an underestimation of the mortality of pre- 
reproductive individuals, as there would not have 
been fi:iiits present to identify the species. Clearly 
if this occurred then the percentage mortality of the 
populations would have been underestimated. While 
this may have influenced the recorded percentage 
mortality of the population, we feel that it would have 
little influence on our general findings. 

When examining the relationship between the 
size of the storage organ and resprouting vigour 
we found that larger storage organ size was related 
to more resprouting shoots and faster growing 



resprouting shoots. Although this pattern was not 
imexpected, we had hoped to be able to determine 
whether carbohydrate storage or ntxmber of dormant 
buds was the limiting factor when it came to the 
ability of individuals to resprout, but we did not 
detect any trends. Rather, it would appear that 
larger storage organs have more dormant buds 
available for resprouting and greater carbohydrate 
stores. Whether this greater resprouting vigour for 
individuals with larger storage organs translates to 
a greater reproductive output remains to be tested at 
this site, but other studies have shown a relationship 
between storage organ size and reproductive output 
(Auld 1987; Bowen and Pate 2004). Furthermore, 
there was clear evidence supporting this idea in the 
grasstree {Xanthorrhoea johnsonii) where the length 
of the inflorescence was positively correlated with 
the volume of the caudex. We also found that plants 
that lacked or had a short caudex did not flower in 
the first year following fire. This result contrasts with 
the findings of Lamont et al. (2000) who foimd that 
for a Western Australian grasstree, plant size was 
not positively related to the proportion of plants 
flowering. 

Fire frequency did not affect the post-fire survival 
of Hakea laevipes or Petrophile canescens, but high 
fire fi-equency increased the mortality of Banksia 
spinulosa. Individuals of Banksia spinulosa in higher 
fire frequency sites were more likely to be killed by 
fire, were generally smaller in size when compared to 
less fi-equently burnt sites. Interestingly, we found no 
evidence that this increased mortality was a result of 
a depletion of the bud bank, as the number of shoots 
per plant did not differ among sites with different fire 



54 



Proc. Linn. Soc. N.S.W., 127, 2006 



K.J.E. KNOX AND P.J. CLARKE 



frequencies. Similarly, we found no evidence that 
the mortality was directly related to a depletion of 
carbohydrate reserves, as the length of the longest 
resprout did not differ among sites with different fire 
frequencies. Rather, it appears that the cohort that 
recruited since the previous fire (1990) had not had 
an opportunity to reach fire tolerance and it was these 
individuals that contributed to the higher mortality 
in the more frequently burnt site. A synthesis of 
post-fire survival of juvenile resprouting species by 
Keith (1996) suggests that some Proteaceous shrubs 
{Banksia oblongifolia, Telopea speciosissima) develop 
the ability to resprout at around five years, but others 
{Banksia serrata, Isopogon anemonifolius) may take 
more than 10 years to develop a strong resprouting 
ability. Whether juvenile plants that recruit after fire 
are able to develop persistence in less than 10 years 
needs to be tested on seedlings that established after 
recent fires. 

Xanthorrhoea johnsonii individuals were less 
likely to flower in the higher fire frequency site, 
but the length of the inflorescence was not affected 
by fire frequency. This is surprising given the large 
investment of resources in post-fire flush flowering 
in Xanthorrhoea johnsonii. The reduced flowering 
in the high fire frequency sites may be a result of 
previous fires depleting carbohydrate reserves. Knox 
and Morrison (2005) found a similar pattern for some 
resprouting shrubs where individuals in high fire 
frequency sites had lower reproductive output than in 
less frequently burnt sites. Interestingly, Taylor et al. 
(1998) found individuals of Xanthorrhoea fulva were 
more likely to flower in areas with high fire frequency 
than areas with less frequent fires. While this appears 
to contradict our findings, the intervals in that study 
were much shorter than those in the current study, and 
hence it is difficult to draw comparisons. 

Previous studies that have examined the effects 
of fire frequency in dry sclerophyll vegetation have 
often found resprouters to decline in abundance under 
very short inter-fire intervals (e.g. Gary and Morrison 
1995). In the current study, the shortest interval 
between fires was seven years and at this fire frequency 
two of the four species examined were adversely 
affected in the higher fire frequency sites. This is an 
important finding because the current Guidelines for 
Ecologically Sustainable Fire Management in NSW 
(Kenny et al. 2003) indicate that a lower minimum 
threshold between fires for dry sclerophyll shrub 
forest is seven years, and the results from this current 
study indicate that such an interval may be too short 
for these particular forests. 



ACKNOWLEDGEMENTS 

We thank the NSW National Parks and Wildlife Service 
(now Department of Environment and Conservation) for 
allowing us to measure the post-fire responses of plants in 
Gibraltar Range National Park and for logistic support for 
the research. The 2003 class of the Ecology of Australian 
Vegetation is thanked for their diligent collection of the data 
that have enabled this paper to be produced. Ian Simpson 
provided excellent field assistance for the staff and students, 
and the University of New England supported the research 
by extending the budget of the student excursion. 



REFERENCES 

Auld, T.D. (1987). Post-fire demography in the resprouting 
shrub Angophora hispida (Sm.) Blaxell: flowering, 
seed production, dispersal, seedling establishment 
and survival. Proceedings office Linnean Society of 
New South Wales 109, 259-269. 

Bell, D.T (2001). Ecological response syndromes in 
the flora of southwestern Western Australia: fire 
resprouters vs reseeders. Botanical Review 67, 417- 
440. 

Bellingham, P.J. and Sparrow, A.D. (2000). Resprouting 
as a life history strategy in woody plant communities. 
0/7co5 89, 409-416. 

Bond, W.J. and Van Wilgen, B.W (1996). Fire and Plants. 
Chapman & Hall, London. 

Bowen, B.J. and Pate, J.S. (1993). The significance of root 
starch in post-fire shoot recovery of the resprouter 
Stirlingia latifolia R. Br (Proteaceae). Annals of 
Botany 12, 7-16. 

Bowen, B.J. and Pate, J.S. (2004). Effect of season of 
bum on shoot recovery and post-fire flowering 
performance in the resprouter Stirlingia latifolia R. 
Br (Proteaceae). Austral Ecology 29, 145-155. 

Bradstock, R.A. and Myerscough, RJ. (1988). The 

survival and population response to frequent fires of 
two woody resprouters Banksia serrata and Isopogon 
anemonifolius. Australian Journal of Botany 36, 415- 
431. 

Burrows, N.D. (1985). Reducing the abundance of 
Banksia grandis in the Jarrah forest by the use of 
controlled fire. Australian Forestry 48, 63-70. 

Gary, G. J. and Morrison, D.A. (1995). Effects of fire 

frequency on plant species composition of sandstone 
communities in the Sydney region: combinations of 
inter-fire intervals. Australian Journal of Ecology 20, 
418-426. 

Clarke, RJ. Knox, K.J.E., Wills, K.E. and Campbell, M.L. 
(2005). Landscape patterns of woody plant response 
to crown fire: disturbance and productivity influence 
sprouting ability. Journal of Ecology 93, 543-555. 

Denham, A.J. and Whelan, R.J. (2000). Reproductive 
ecology and breeding system of Lomatia silaifolia 
(Proteaceae) following fire. Australian Journal of 
Botany 'X^, 261-269. 



Proc. Linn. Soc. N.S.W., 127, 2006 



55 



RESPONSE OF RESPROUTING SHRUBS TO REPEATED FIRES 



Gill, A.M. (1981). Adaptive responses of Australian 

vascular plant species to fire. In Fire in the Australian 
Biota. Gill, A. M., Groves, R. H. and Noble, I. R. 
(Eds). Australian Academy of Science, ACT, pp. 243- 
272. 

Keith, D.A (1996). Fire-driven extinction of plant 
populations: a synthesis of theory and review of 
evidence from Australian vegetation. Proceedings of 
the Linnean Society of New South Wales. 116, 37-78. 

Kenny, B., Sutherland, E., Tasker, E. and Bradstock, R. 
(2003). Guidelines for Ecologically Sustainable Fire 
Management, NSW Government, Sydney. 

Knox, K.J.E. and Clarke, RJ. (2005). Nutrient availability 
induces contrasting allocation and starch formation in 
resprouting and obligate seeding shrubs. Functional 
Ecology 19, 690-698 

Knox, K.J.E. and Morrison, D.A. (2005). Effects of 
inter-fire intervals on the reproductive output of 
resprouters and obligate seeders in the Proteaceae. 
Austral Ecology 30, 407-413 

Lamont, B.B., Swanborough RW. and Ward, D. (2000). 
Plant size and season of bum affect flowering and 
Suiting of the grasstree Xanthorrhoea preissii. 
Austral Ecology 25, 268-272. 

Morrison, D.A. (1995). Some effects of low-intensity fires 
on populations of co-occurring small trees in the 
Sydney region. Proceedings of the Linnean Society of 
New South Wales 115, 109-119. 

Morrison, D.A. and Renwick, J.A. (2000). Effects of 
variation in fire intensity on regeneration of co- 
occurring species of small trees in the Sydney region. 
Australian Journal of Botany 48, 71-79. 

Pyke, G.H. (1983). Relationships between time since last 
fire and flowering in Telopea speciosissima R. Br. and 
Lambertia formosa Sm. Australian Journal of Botany 
31, 293-296. 

Taylor, J.E., Monamy, V. and Fox, B.J. (1998). Flowering 
of Xanthorrhoea fulva: the effect of fire and clipping. 
Australian Journal of Botany 46, 241 - 251 

Vesk, P.A. and Westoby, M. (2004). Sprouting ability 
across diverse disturbances and vegetation types 
worldwide. Journal of Ecology 92, 310-320. 

Zammit, C. (1988). Dynamics of resprouting in the 

lignotuberous shrub Banksia oblongifolia. Australian 
Journal of Ecology 13, 311-320. 



56 Proc. Linn. Soc. N.S.W., 127, 2006 



Fire Responses in Four Rare Plant Species at Gibraltar Range 
National Park, Northern Tablelands, NSW 

^Peter Croft, ^Damien Hofmeyer and ^John T. Hunter 

' Department of Environment and Conservation (NSW), Parks and Wildlife Division, Glen Innes Area, 68 

Church St, Glen Innes, NSW 2370 
^Department of Environment and Conservation (NSW), Parks and Wildlife Division, Richmond River Area, 

Colonial Arcade 75 Main St, Alstonville, NSW 2477 
^School of Human and Environmental Studies, University of New England, Armidale, NSW 2351 



Croft, P., Hofrneyer, D. and Hunter, J.T. (2006). Fire response of four rare plant species at Gibraltar Range 
National Park, Northern Tablelands, NSW. Proceedings of the Linnean Society of New South Wales 127, 
57-62. 

Fire responses are reported in four rare species at Gibraltar Range National Park following hazard- 
reduction buming. Acacia barringtonensis Tindale, Grevillea rhizomatosa P.M.Olde & N.R.Marriot, 
Persoonia nifa L.A.S.Johnson & P.H.Weston and Telopea aspera M.D. Crisp & P. H. Weston were the 
species investigated. In each species, individuals were tagged prior to a hazard reduction fire and their 
fates followed for 34 months. In Acacia barringtonensis, individuals survived fire and resprouted fi'om 
buds at the base of stems and fi-om rhizomes but the resprouts were heavily browsed by insects and Swamp 
Wallabies {WaUabia bicoJor Desmarest). In Grevillea rhizomatosa, individuals survived and resprouted 
fi'om imderground rhizomes and no seedlings were found after fires. After fire in Persoonia rufa, all 
scorched plants died but seedling recruitment occurred fi-om a soil-stored seed. In Telopea aspera, most 
burnt individuals resprouted from basal shoots and survived despite heavy post-fire grazing pressure. 
Increasing fire frequencies by hazard-reduction buming may threaten the survival of all four species, and 
it is suggested that other methods of reducing fuel be used to manage fire in this area of Gibraltar Range 
National Park. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: fire ecology, fire response, obligate seeding, rare species, resprouting. 



INTRODUCTION 

Hazard-reduction buming for wildfire suppression 
is thought to have recently increased in fire-prone 
vegetation around the world (Moritz et al. 2004; DEC 
2004). This is particularly so on reserved lands in 
NSW where the NSW Department of Environment 
and Conservation annual report (DEC 2004) noted 
that there were twice the number of hazard-reduction 
bums in NSW National Parks in 2003-04 than in 2002- 
03. Hazard-reduction buming may be initiated when 
fiiel begins to accimiulate beyond specified thresholds 
(Gill et al. 1987; Morrison et al. 1996; Morrison and 
Renwick 2000; Femandes and Botelho 2003). On 
the Northern Tablelands and similar areas, fuel may 
be kept at or below these thresholds by buming as 
often as every three to five years (Raison et al. 1986; 
Smith et al. 1992). This is a higher frequency than 



recommended for perpetuation of many species and 
vegetation communities in the region (Clarke and 
Fulloon 1997). 

Fire can endanger the viability of species, 
especially if fire frequency is too high (Benson 
1985; Bradstock et al. 1995; Keith 1996; Morrison 
and Renwick 2000). Keith (1996) identified high fire 
frequency as a mechanism for plant population decline 
and extinction through depletion of buds or starch 
reserves in standing plants and also as a mechanism 
for depleting soil-stored seed banks before they can 
be replenished. Hence 'High Frequency Fire' has been 
listed as a Key Threatening Process in the Threatened 
Species Conservation Act (TSC) 1995. 

Some shrubs are killed by fire and, after fire, 
rely on seedling germination and a sufficient interfire 
period to survive and reproduce (obligate seeders), 
whilst others resprout after fire (Benson 1985; Gill 
& Bradstock 1992; Morrison & Renwick, 2000). 



FIRE RESPONSES IN FOUR RARE PLANT SPECIES 



Glenlnnes 




s_ 



Reserve Boundary 

Road 

^ Town 

i Study Location 



Figure 1. Location of Mulligan's Hut study area within Gi 
braltar Range National Park. 



However, fire responses in many species of vascular 
plants in Gibraltar Range National Park are unknown 
(Clarke and Fulloon 1997; Williams and Clarke 1997; 
Hunter 1995; Hunter 1998; Hunter 2003; Clarke and 
Knox 2002; Knox and Clarke 2004). 

Within Gibraltar Range National Park several 
areas have been subjected to hazard-reduction burning. 
One of them, Mulligan's Hut, contains populations 
of four rare shrub species, Acacia barringtonensis 
Tindale, Grevillea rhizomatosa P.M.Olde & 
N.R.Marriot, Persoonia rufa L.A.S.Johnson & 
P.H.Weston and Telopea aspera Crisp & P.H.Weston. 
As fire responses of plants in each of these species 
were poorly known, fates of plants burned in a 
hazard-reduction fire in 1999 were recorded together 
with any recruitment of seedlings of these species 
after the fire. 



The mean annual rainfall at Mulligan's 
Hut is 2100 mm. The study area has a 
mean annual temperature of 1 3°C on the 
plateau with a mean aimual maximum 
of 28°C and mean annual minimum of 
0°C. The warmest months of the year are 
November to March. The rock types are 
primarily granitic and the topography 
is generally undulating with extensive 
areas of exposed rock sheeting and 
boulder fields. 

The study area comprised four 
hectares of open forest immediately 
north of the Mulligan's Hut camping 
area in Gibrahar Range National Park 
where a small hazard-reduction bum 
was scheduled in 1999. Mulligan's Hut 
camping area is towards the centre of the 
Gibraltar Plateau, at an altitude of 900 
m. This bum was planned to help protect 
visitors and facilities in the camping area 
firom wildfire in the park. 

The open forest community at this 
locality is AormaaXeAhy Eucalyptus olida 
L.A.S.Johnson & K.D.Hill, Eucalyptus 
ligustrina DC. and Eucalyptus 
cameronii Blakely & McKie. The shrub 
layer is dominated by Leptospermum trinervium 
(Sm.) Joy Thomps., Dillwynia phylicoides A.Cunn., 
Hakea laevipes subsp. graniticola Haegi, Petrophile 
canescens A.Cunn. ex R.Br, and Daviesia umbellata 
Sm. The ground layer consists of: Caustis flexuosa 
R.Br., Platysace ericoides (Sieber ex Spreng.) 
C.Norman, Bossiaea neo-anglica F.Muell., Goodenia 
rotundifolia R.Br, and Entolasia stricta (R.Br.) 
Hughes. 

Prior to the hazard-reduction bum in 1999, 
NSW National Parks & Wildlife Service fire records 
indicated two large wildfires had bumt the Mulligan's 
Hut area in 1964 and 1988 with fire-history maps 
indicating the study site was bumt. Another fire 
occurred after the project was initiated in 2002 and 
the subject populations were bumt during back- 
buming operations. 



METHODS 

Study Area 

Gibraltar Range National Park is located 90 
km west of Grafton and 65 km east of Glen Innes in 
north-eastemNew South Wales (29°329 S 152° 189 E) 
(Fig. 1). The Gibraltar Range straddles the Northem 
Tablelands and North Coast Botanical Subdivisions. 



Target species 

Acacia barringtonensis is an erect shmb endemic 
to high altitude areas of northem New South Wales 
with a Rare or Threatened Australian Plant (RoTAP) 
code of 3RCa (Briggs and Leigh 1996). This shrub 
grows along swamp margins and creek edges in dry 
sclerophyll forests and woodlands reaching a height 
of 7 m. Flowering occurs primarily from September 
to early November (Tindale 1975). 



58 



Proc. Linn. Soc. N.S.W., 127, 2006 



p. CROFT, D. HOFMEYER AND T. HUNTER 



Table 1. Percent survival of tagged plants in four species in relation to amount of leaf scorched during 
a planned hazard-reduction burn in the Mulligan's Hut area of Gibraltar Range in 1999. 



Species 


% Leaf 
Scorch 


1 
month 


3 
months 


5 
months 


7 
months 


34 
months 


Number of 
plants 


Acacia 
barringtonensis 


0-50 


0% 


100% 


100% 


100% 


0% 


2 




51-75 


0% 


100% 


100% 


100% 


0% 


1 




76-100 


0% 


50% 


57% 


54% 


7% 


28 


Grevillea 
rhizomatosa 


0-50 


0% 


100% 


100% 


100% 


100% 


2 




51-75 


0% 


33% 


67% 


67% 


67% 


3 




76-100 


0% 


20% 


39% 


57% 


52% 


49 


Persoonia rufa 


0-50 


0% 


50% 


50% 


0% 


0% 


2 




56-75 


0% 


100% 


100% 


0% 


0% 


1 




76-100 


0% 


12% 


6% 


0% 


0% 


17 


Telopea aspera 


100% 


12% 


94% 


94% 


94% 


94% 


17 



Grevillea rhizomatosa is known from scattered 
populations within Gibraltar Range and adjacent 
areas of Washpool National Park (Sheringham and 
Hunter 2002) and is listed on Schedule 2 (Vulnerable) 
on the TSC Act. It is a shrub 0.3-1 m tall and is known 
to sucker from roots and grows in sclerophyll forests 
on sandy soils near creeks. The species flowers 
sporadically throughout the year (see also Caddy and 
Gross this volume). 

Persoonia rufa is endemic to the Gibraltar Range. 
It is a spreading shrub with a RoTAP code of 2RC 
(Briggs and Leigh 1996). The plant commonly grows 
to 1-2.5 m tall and is found in dry open forests on 
granitic soils (Sheringham and Hunter 2002; Weston 
and Johnson 1991). Flowering is primarily between 
December and February. 

Telopea aspera is a multistemmed shrub that 
grows to 3 m tall and has a RoTAP code of 2RCa 
(Briggs and Leigh 1996). It is largely restricted to 
Gibraltar Range and is known from dry sclerophyll 
forests on granitic soils. Flowering occurs between 
October and November (Sheringham and Hunter 
2002; Crisp and Weston 1993). The flowering 
response after fire has not been studied in Telopea 
aspera. The closely related Telopea speciosissima is 
a pyrogenic flowerer and recruits two years after fire 
(Pyke 1987; Bradstock 1995). 

Fire response traits 

Before the fire individuals of all four species 
were tagged with stainless steel straps with individual 



identification codes. Each individual was marked on a 
map of the study area, to aid relocation after the fire. 
Plant attributes measured included: basal diameter, 
height, number of stems, location of regrowth, 
flowering stage, condition and number of seedlings 
nearby. All the individuals of Telopea aspera within 
the study area were tagged (17 plants), the populations 
of the other three species were sub-sampled: 
Grevillea rhizomatosa (160 individuals). Acacia 
barringtonensis (62 individuals) and Persoonia rufa 
(60 individuals). 

The intensity of the fire was gauged by using 
flame height markers, photographs and scorch height 
post bum. Measurements were taken of all tagged 
plants before the experimental bum and at one, three, 
five, seven and 34 months after it. 



RESULTS 

The Grevillea rhizomatosa population was 
burned by a low-intensity fire (average flame height 
0.75 m) that affected 54 tagged plants; 26 were bumt 
and 28 were scorched by radiant heat. Grevillea 
rhizomatosa responded to fire by increasing the 
average number of stems per plant from 1.19 prior 
to burning to 1.78 after being bumt. Although all 
tagged plants unaffected by fire survived, only 55% 
of fire-affected individuals were alive at the end of 
the monitoring period (Table 1). Forty -five (83%) of 
all surviving fire-affected individuals recovered by 



Proc. Linn. Soc. N.S.W., 127, 2006 



59 



FIRE RESPONSES IN FOUR RARE PLANT SPECIES 



multiple rhizomes at a distance of up to 30 cm (12 
cm average) from the parent plant, and the remaining 
nine plants recovered by coppicing. 

A low-intensity fire (average flame height 1 .07 m) 
burned the Acacia barringtonensis population where 
20 plants were burned and a fiarther 11 scorched. 
All tagged Acacia barringtonensis unaffected by 
fire survived until the end of the monitoring period. 
Although 17 fire-affected plants recovered from 
basal stem buds by the 5-month post fire, only two 
fire-affected individuals survived until the end of 
monitoring. Most of these recovering individuals were 
heavily browsed by both insects and Wallabia bicolor 
(Swamp Wallaby). No seedlings were recorded in 
the first seven months within the vicinity of affected 
A. barringtonensis. However, at 34 months, 22 
putative 'seedlings' between 15-50 cm in height were 
observed. 

Twenty Persoonia rufa plants were burned with 
a low-intensity fire (average flame height 0.75 m). 
All unbumt tagged Persoonia rufa individuals (40) 
survived until the end of the monitoring period, but no 
individuals survived that were burnt (Table 1). Three 
scorched plants continued to survive for 5 months 
after the fire. Within the study area 30 seedlings were 
counted five months post-bum. All seedlings were of 
a uniform height (5 cm) and survived through until 
the last recording period. 

Flame heights exceeding five metres were 
recorded in the areas where Telopea aspera was 
tagged. This resulted in 100% of the tagged plants 
being burnt at moderate intensity. Ninety-four 
percent of tagged Telopea aspera plants survived the 
moderate intensity bum. The sole means of survival 
was by resprouting from the base/lignotuber. As a 
consequence of hazard-reduction buming, mean 
number of stems per Telopea aspera plant increased 
from 4.5 to 9.1. At least two individuals were noted 
resprouting after the first month, with all remaining 
surviving plants resprouting by the second month 
and surviving to the final sample date (Table 1). 
Heavy browsing of resprouting parts by insects and 
Swamp Wallabies {Wallabia bicolor) was observed 
on recovering plants after the fire. No plants had 
flowered within the post-fire monitoring period. 



DISCUSSION 

Species responses 

Whilst many individual plants died as a result 
of the hazard-reduction fire, all species persisted in 
the study area by different fire response syndromes. 



The immediate post-fire response of surviving 
Acacia barringtonensis was basal resprouting. These 
resprouting stems were heavily grazed, which may 
be a cause of the decline of this species towards the 
end of the monitoring period. Only at the last survey, 
at 34 months, were putative seedlings noted in the 
vicinity of dead individuals. Subsequently, these 
'seedlings' were revisited three years after the last 
monitoring (May 2005) and were found to be shoots 
from roots that extended back to 'dead' individuals 
that were tagged. Thus it would seem that, though 
delayed by nearly three years, this species' response 
to the hazard-reduction bum was resprouting. The 
delay may have been in part due to increased grazing 
pressure that inunediately followed this fire. Increased 
grazing pressure is not a necessary consequence of 
fire. Indeed, following an extensive wild fire, it may 
be less than before the fire. However, small burnt 
areas within large unbumt surrounding areas, such 
as may arise from some hazard-reduction fires, may 
be particularly attractive to browsing and grazing 
animals and experience much more pressure from 
them than surrounding unbumt areas. 

No seedlings of Grevillea rhizomatosa were found 
during the monitoring period and populations appear 
to be maintained by resprouting from underground 
rhizomes. Keith (1996) noted that resprouters might 
be killed if stored starch reserves are exhausted by 
repeated fires. Though numbers were too low for 
statistical comparisons, more of the smaller plants 
in terms of stem diameter and height survived, 
potentially indicating an age effect ability to recover 
post-fire (see paper by BCnox and Clarke this volume). 
The smaller Grevillea rhizomatosa plants in this 
study may have depleted smaller quantities of starch 
reserves. Following the second bum three years later 
about 65% of the original survivors of the hazard- 
reduction bum did not recover from the second fire. 

Standing populations of Persoonia rufa 
individuals were the most susceptible to extirpation 
by the end of the monitoring period from low- 
intensity bums. Although some temporary recovery 
occurred (due to coppicing after minor scorching), 
the species persisted in the fire-affected area mainly 
by germination from a soil-stored seed bank; hence 
this species should be classed as an obligate seeder. 

Telopea aspera was the most resilient in terms of 
recovery of those individuals present before hazard- 
reduction buming. Almost all of the tagged plants 
survived to the end of the monitoring period, despite 
the imposition of increased (and heavy) herbivory on 
newly forming foliage. This species responded to fire 
by resprouting from basal/lignotuberous buds, with 



60 



Proc. Linn. Soc. N.S.W., 127, 2006 



p. CROFT, D. HOFMEYER AND T. HUNTER 



fire increasing the number of stems per plant post- 
bum. Of the four species monitored, Telopea aspera 
was the first to show signs of recovery after fire but no 
seedHng recruitment was observed. 

Implications 

Landscape burning at short intervals can have 
major effects on plant populations (Bradstock 
1995; Bell 2001) and may drive the decline of plant 
populations (Keith, 1996). Consequently, fi-equent 
fire can have a significant effect on the composition 
of flora and fauna (Clark 1988; Andrew et al. 2000; 
York 2000; Moritz et al. 2004). The current study 
has identified varied fire responses of plants in four 
rare species within Gibraltar Range National Park. 
Although populations of all species persisted after a 
hazard-reduction bum, most were reduced in numbers 
and at least two were affected by an increase in post- 
fire herbivory. This herbivory may have delayed the 
regeneration of Acacia barringtonensis, and may 
have detrimental long-term effects on Telopea aspera 
in terms of depleting starch reserves. As no plant in 
any of the four species flowered 34 months after the 
fire, fire intervals of greater than three years will be 
needed to maintain their populations. 

Hazard-reduction bums may have minimal 
effects on the number of wildfire events particularly 
in fire-prone vegetation (Tumer et al. 2003; Moritz et 
al. 2004) and this is likely to be the case here at the 
Mulligan's Hut site (e.g. the 2002 fire that effected 
the Mulligan's Hut area). Though hazard-reduction 
bums are plarmed, wildfire events are not, thus by 
increasing the amount of fire in the landscape without 
being able to predict wildfires there is an increased 
risk of population decline and extinction. Whilst 
hazard-reduction buming can reduce the intensity of 
a wildfire for several years (Raison et al. 1986) it may 
not prevent the area from re-buming during a wildfire, 
especially in severe fire weather (Tumer et al. 2003; 
Moritz et al. 2004). Additionally, hazard reduced 
ground is often chosen as an area from which back 
bums are plarmed during wildfires control operations 
because of lower fuel levels. If hazard-reduction 
buming is undertaken at the maximum frequency 
without considering unplanned fires then critical 
thresholds of fire frequency for long-term survival 
of populations can be exceeded as has occurred at 
Mulligan's Hut. To ensure the persistence of our focal 
species a fire interval that allows seedlings to mature 
and a seed bank to accumulate is required. Whilst 
these demographic factors are yet to be quantified it 
is suggested that the minimum interval between fires 
to ensure the persistence of the focal species will be 
more than ten years. 



ACKNOWLEDGMENTS 

Dr Kathryn Taffs is thanked for reviewing preliminary 
drafts. The staff of Department of Environment and 
Conservation (NSW) is also thanked for providing assistance 
in carrying out the hazard-reduction bum in 1 999. The input 
from the editors and referees was appreciated. 



REFERENCES 

Andrew, N., Rodgerson, L. and York, A. (2000). Frequent 

fuel-reduction buming: the role of logs and associated 

leaf litter in the conservation of ant biodiversity. 

Austral Ecology 25, 99-107. 
Bell, D.T. (2001). Ecological response syndromes in 

the flora of southwestern Western Australia: Fire 

resprouters versus reseeders. The Botanical Review 

67,417-440. 
Benson, D.H. (1985). Maturation periods for fiire sensitive 

shrub species in Hawkesbury sandstone vegetation. 

Cunninghamia 1, 339-349. 
Bradstock, R.A. (1995). Demography of Woody Plants in 

Relation to Fire: Telopea speciosissima. Proceedings 

of the Linnean Society of New South Wales 115, 25- 

33. 
Bradstock, R.A., Keith,D. and Auld, T.D. (1995). Fire 

and conservation: imperatives and constraints on 

managing for diversity. In 'Conserving biodiversity: 

threats and solutions' (Eds R.A. Bradstock, T.D. 

Auld, D.A. Keith, R.T. Kingsford, D. Lunney and 

D.R Sivertsen) pp. 323-334 (Surrey Beatty and Sons, 

Chipping Norton). 
Briggs J.D. and Leigh J.H. (1996). 'Rare or Threatened 

Australian Plants'. (CSIRO publishing, 

Collingwood). 
Clark, S.S. (1988). Effects of hazard-reduction buming 

on populations of understorey plant species on 

Hawkesbury sandstone. Australian Journal of 

Ecology 13, 473-484. 
Clarke, RJ. and Fulloon, L. (1997). 'Fire and rare 

plants: Torrington State Recreation Area. Botany 

Department'. (University of New England, 

Armidale). 
Clarke, P.J. and Knox, J.E. (2002). Post fire response 

of shrubs in the tablelands of Eastern Australia: 

do existing models explain habitat differences? 

Australian Journal of Botany 50, 53-62. 
Crisp, M.D. and Weston, PH. (1993). Geographic and 

ontonogenic variation in morphology of Australian 

Waratahs {Telopea: Proteaceae). Systematic Biology 

42, 49-76. 
Department of Environment and Conservation (2004). 

Annual Report. Department of Environment and 

Conservation (NSW). Sydney. 
Femandes, RM. and Botelho, H.S. (2003). A review 

of prescribed buming effectiveness in fire hazard 



Proc. Linn. Soc. N.S.W., 127, 2006 



61 



FIRE RESPONSES IN FOUR RARE PLANT SPECIES 



reduction. InternationalJournal of Wildland Fire 12, 
117-128. 

Gill, A.M. and Bradstock, R.A. (1992). A national register 
for the fire responses of plant species. Cunninghamia 
2, 653-660. 

Gill, A.M., Christian, K.R., Moore, P.H.R. and Forrester, 
R.I. (1987). Bushfire incidence, fire hazard and fiiel 
reduction burning. Australian Journal of Ecology 12, 
299-306. 

Hunter, J.T (1995). Some observations on the fire 

responses of two rare species in the Girraween and 
Bald Rock National Parks. Queensland Naturalist 35, 
5-6. 

Hunter, J.T. (1998). Notes on the occurrence ofMonotaxis 
macrophylla Benth. (Euphorbiaceae), with particular 
reference to New South Wales. Queensland 
Naturalist 36, 21-24. 

Hunter, J.T. (2003). Persistence on inselbergs: the role 
of obligate seeders and resprouters. Journal of 
Biogeography 18, 497-510. 

Keith, D. (1996). Fire-driven extinction of plant 

populations: a synthesis of theory and review of 
evidence from Australian vegetation. Proceedings of 
the Linnean Society of New South Wales 116, 37-78. 

Knox, K.J.E. and Clarke, P.J. (2004) Fire response 

syndromes of shrubs in grassy woodlands in the New 
England Tableland Bioregion. Cunninghamia 8, 348- 
353. 

Moritz, M.A., Keeley, J.E., Johnson, E.A. and Schaffher, 
A.A. (2004). Testing the basic assumption of 
shrubland fire management: how important is fiiel 
age? Frontiers in Ecology and the Environment 2, 
67-72. 

Morrison, D.A., Buckney, R.T., Renwick, B.J. and Cary, 
G.J. (1996). Conservation conflicts over burning bush 
in south eastern Australia. Biological Conservation 
76, 167-175. 

Morrison, D.A. and Renwick J.A. (2000). Effects of 
variation in fire intensity on regeneration of co- 
occurring species of small trees in the Sydney region. 
Australian Journal of Botany 48, 71-79. 

Pyke, G.H. (1987). Pollination biology of Telopea 
speciosissima. In ' Waratahs - their Biology, 
Cultivation and Conservation' (Ed. J.A. Armstrong) 
pp. xx-xx Australian National Botanic Gardens. 
Occasional Publication No. 9. 

Raison, R.J., Woods, RV. and Khanna, RK. (1986). 

Decomposition and accumulation of litter after fire 
in sub-alpine eucalypt forests. Australian Journal of 
Ecology 11, 9-19. 

Sheringham, P. and Hunter, J.T. (2002). Vegetation and 
floristics of Gibraltar Range National Park. (A report 
to the NSW National Parks & Wildlife Service). 

Smith, A.R, Moore, D.M. and Andrews, S.R (1992). 
Proposed Forestry Operations in the Glen Irmes 
Management Area, Fauna Impact Statement. Prepared 
for the Forestry Commission of New South Wales. 

Tindale, M.D. (1975). Notes on Australian taxa of Acacia 
No. 4. Telopea 1, 68-83. 



Turner, M.G., Romme, W.H. and Tinker, D.B. (2003). 
Surprises and lessons fi-om the 1988 Yellowstone 
fires. Frontiers in Ecology and the Environment 1, 
351-358. 

Weston, P.H. and Johnson, L.A.S. (1991). Taxonomic 
changes in Persoonia (Proteaceae) in New South 
Wales. Telopea 4, 269-306. 

Williams, P.R and Clarke PJ (1997). Habitat segregation 
by serotinous shrubs in heaths: post fire emergence 
and seedling survival. Australian Journal of Botany 
45,31-39. 

York, A. (2000). Long-term effects of frequent low- 
intensity burning on ant communities in coastal 
Blackbutt forests of south-eastern Australia. 
Austral Ecology 25, 83-98. 



62 



Proc. Linn. Soc. N.S.W., 127, 2006 



Response of Montane Wet Sclerophyll Forest Understorey 
Species to Fire: Evidence from High and Low Intensity Fires 

Monica L. Campbell and Peter J. Clarke 

Botany, School of Environmental Sciences and Natural Resources Management, University of New England, 

Armidale NSW 2351 (jpclarkel@une.edu.au). 



Campbell, M.L. and Clarke, P.J. (2006). Response of montane wet sclerophyll forest understorey species 
to tire: evidence from high and low intensity fires. Proceedings of the Linnean Society of New South 
Wales 111, 63-73. 

On the New England Tablelands wet sclerophyll forests typically form the ecotone between rainforest and 
dry sclerophyll forest. Currently there are few data on the response of wet sclerophyll plant species to fire. 
We compared the fire-response traits of woody understorey and sub-canopy species in wet sclerophyll forest 
after high and low intensity fires. The majority of species (>80%) resprouted after fire and the prevalence 
of resprouting did not differ with fire intensity. Obligate seeders were rare in these communities (<10% 
of species), and similar numbers of rainforest and sclerophyllous species were killed by fire. Resprouting 
from basal stems and root suckering were the most common mechanisms of vegetative regeneration; 
however, these traits may have arisen more in response to canopy disturbance than fire regime. We found 
that most rainforest taxa resprouted but lacked post-fire seedling recruitment, whereas most resprouting 
sclerophyllous taxa recruited from seed after fire. This dichotomy in seedling recruitment could reflect 
the productivity and disturbance gradients across the ecotone. We propose that gap-phase recruitment is 
favoured towards the rainforest margin and fire-related recruitment is more prevalent at the eucalypt forest 
edge. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: functional groups, obligate seeding, rainforest, resprouting, seedling recruitment, wildfire. 

wet sclerophyll forest have been used for hardwood 

INTRODUCTION timber production and cattle grazing. These forests 

have consequently been exposed to a regime of 
Wet sclerophyll forests typically form an interface frequent, low-intensity fires that are often associated 
between two broad vegetation types, rainforest and with grazing (stimulation of green pick). A high fire- 
dry sclerophyll forest. These tall eucalypt forests have frequency is thought to reduce diversity of woody 
been described as a stage in long-term succession as species in the forest understorey by removing 
their structure and composition approaches that of mesophyllous taxa and promoting growth of fire- 
rainforest in the prolonged absence of disturbance. tolerant grasses and forbs (Binns 1991; Henderson 
Hence, changes in community properties have been and Keith 2002). Currently there are few data on the 
closely linked to the passage of fire and length of demographic links between disturbance frequency 
fire-fi-ee periods (Ashton and Attiwill 1994). Wet and changes in understorey composition in the wet 
sclerophyll forests are not highly flammable most of the sclerophyll ecotone. 

time, and as time-since-fire increases the probability xhe classification of plant species into fianctional 

of a subsequent fire is reduced as mesophyllous taxa groups based on fire-response fraits can be usefiil in 

become more dominant in the standing vegetation preliminary modelling of how vegetation will change 

(Unwinl989;Adam 1992; Harrington and Sandersen with more frequent or less fi-equent fires (Whelan 

1994). Generally, one intense crown fire every 100- 1995; Bond and van Wilgen 1996). Baseline data on 

200 years is thought to be sufficient to maintain the response of plant populations to crown fire are 

the sclerophyllous component and restrict the more being sought to confrast the effects of seed-based 

mesophyllous taxa of the wet sclerophyll ecotone recruitment (persistence ofpopulations) with those of 

(Gilbert 1959; Chesterfield et al. 1991). resprouting (persistence of individuals) (Clarke and 

On the New England Tablelands many areas of Knox 2002; Pausas et al. 2004; Vesk and Westoby 



RESPONSES OF UNDERSTOREY SPECIES TO FIRE 



2004). However, generalisations about vegetation 
change with fire-frequency may be complicated 
by variable responses of species to fire of different 
intensities (e.g. Ashton and Martin 1996; Morrison 
and Renwick 2000). In addition, shade-tolerant 
rainforest species are not expected to have fire-driven 
recruitment, and their regeneration syndromes are 
more likely to be linked to small-scale disturbances 
such as tree fall that create light gaps. 

In spring 2002, areas of wet sclerophyll forest in 
Washpool National Park were burnt by fire following 
the severe drought that affected most of eastern 
Australia. We took advantage of this one in fifty 
year event to record the response of wet sclerophyll 
understorey species to crown fire. We focused on the 
shrub and sub-canopy taxa, as the dominant overstorey 
eucalypts all resprout after fire and their dynamics 
have been documented elsewhere (e.g. Ashton and 
Attiwill 1994; Florence 1996). To test the generality 
of responses to fire we compared data from the crown 
fire at Washpool National Park with a lower intensity 
bum at Mummel Gulf National Park the previous 
year. We addressed the following questions: 1) Do 
fire-response fraits vary with fire intensity? 2) Do fire- 
response traits vary between sympatric rainforest and 
sclerophyllous species? and 3) Are there correlations 
between environmental variables, fire-response traits 
and other life-history traits? 



METHODS 

Study areas 

The New England wet sclerophyll forests 
are restricted to areas of high rainfall along the 
eastern edge of the escarpment. These forest are 
characterised by a eucalypt-dominated overstorey, 
typically exceeding 30 m in height, and a well- 
developed, layered understorey of mesomorphic and 
sclerophyllous growth forms (Specht 1970; Ashton 
and Attiwill 1994). The area selected for study at 
Washpool National Park (hereafter WPNP) was in 
the recent western additions that were acquired by 
the New South Wales National Parks and Wildlife 
Service (NPWS) in 1998 and had not been burnt for 
at least 50 years. The fires that occurred in November 
2002 were high-intensity fires that burnt all vegetation 
strata in the sclerophyll forests and the understorey 
of the warm temperate rainforest. Mummel Gulf 
National Park (hereafter MGNP) lies approximately 
60 km east of Walcha and was also gazetted as 
National Park in the late 1990s. The fire at MGNP 
was a back-bum, initiated by the NPWS to contain a 



grass fire from a neighbouring property in October of 
2001. The fire was of low to moderate intensity and 
resulted in the complete burning of the understorey 
and ground layer, but with minimal canopy scorching. 
All study sites occurred at altitudes greater than 900 
m on metasediment-derived soils. Prior to the fires, 
areas of wet sclerophyll forests within each park 
were surveyed for full floristic composition using 20 
X 20 m quadrats (21 vegetation survey sites in total: 
WPNP, 12 sites; MGNP, 9 sites). Vegetation in each 
quadrat was described in terms of the growth form, 
height and cover of dominant species in the ground 
storey, understorey and canopy strata. All species 
were recorded in each quadrat and their abundance 
estimated using Braun-Blanquet cover-abundance 
scale. For each survey quadrat geographical 
position, altitude, slope, local soil characteristics, 
physiography, evidence of fire and other disturbances 
were recorded. 

Responses of adult plants to fire 

Responses of woody plant species to fire were 
recorded at four sites within burnt forest at WPNP 
and MGNP. At each of these sites two transects (20 
X 2 m) were placed along each of three topographical 
positions - ridge, slope and gully. Topographical 
position accounted for differences in vegetation 
composition prior to fire and differences in fire 
intensity. Post-fire responses of adult plants were 
recorded along each transect in four categories: 
1) killed by fire, 2) resprouting via root suckers, 3) 
resprouting via basal stem buds, and 4) resprouting via 
stem buds. The presence of post-fire seedling recraits 
was also recorded. In addition to these observations, 
post-fire responses of other woody species outside of 
transects (16 species) were recorded to gain a more 
comprehensive overview of fire-response traits of wet 
sclerophyll forest understorey species. 

Analyses of fire response traits 

Plant species were allocated to one of five fire- 
response syndromes based on categories defined by 
Gill and Bradstock (1992). Note that because of the 
low frequency of occurrence, species killed by fire 
were classified into one group regardless of seed 
bank type. An additional group was also formed with 
the combination of regeneration by root suckering 
(Category IV) and resprouting by basal stem buds 
(Category V). Fire-response traits of woody species 
were compared between sites of different fire 
intensities, i.e. high-intensity fire (WPNP) and low- 
intensity fire (MGNP), and between topographic 
positions (ridge, slope and gully) within WPNP and 
MGNP. The relative frequency of fire-response traits 



64 



Proc. Linn. Soc. N.S.W., 127, 2006 



M.L. CAMPBELL AND P.J. CLARKE 



Table 1, Summary table for life-history attributes 
of 61 woody taxa occurring in wet sclerophyll for- 
est in Washpool and Mummel Gulf National Parks on 
the New England Tablelands. Note eucalypts not in- 
cluded. Figures are the No. of species in each category. 



Attribute 



WPNP 



MGNP 



Growth form 
Shrub (<3 m) 
SmallTree (3-10 m) 
Tree (>10m) 

Leaf type 

Sclerophyllous 

Coriaceous 

Mesophyllous 

Dispersal syndrome 
Vertebrate 
Invertebrate 
Wind 
Passive 

Seed bank type 
Soil 
Canopy 
Dispersed 

Species richness 



17 


19 


13 


9 


12 


6 



15 


15 


6 


7 


21 


12 



21 


21 


9 


4 


4 


5 


8 


4 



31 
3 
8 

42 



Analyses of persistence syndromes and foliage 
types 

To test for differences in persistence syndromes 
between rainforest and sclerophyll forest taxa growing 
in the same habitat, species were divided into three 



26 
3 
5 

34 



was compared with a G-test for independence (Sokal 
and Rolf 1981). Fire-response traits of adult plants 
after crown and understorey fires were reclassified 
into one of the four plant persistence syndromes based 
on the hierarchical persistence scheme of Pausas et 
al. (2004). The relative fi-equencies of persistence 
syndromes after crown and understorey fire were 
compared with a G-test for independence (Sokal and 
Rolf 1981). 



foliage types; sclerophyllous, coriaceous 
and mesophyllous. These categories were 
chosen as preliminary representatives 
of the different life-history syndromes 
of plant species in the wet sclerophyll 
ecotone. Rainforest taxa generally fell 
into the mesophyllous and coriaceous leaf 
types, while most sclerophyll forest taxa 
were classified as sclerophyllous. Note that 
because of low frequencies of occurrence, 
taxa with coriaceous leaves were grouped 
with the mesophyllous taxa for the analysis. 
The frequencies of persistence syndromes 
(as described above) were tested between 
foliage groups with a G-test for independence 
(Sokal and Rolf 1981). 



Life-history traits and environmental 
variables 

To examine the relationship between 
plant traits and enviroimiental variables, 
constrained ordinations were derived from 
the full floristic data set. Patterns of life- 
history traits and envirormiental correlates 
were tested with Canonical Correspondence 
Analysis (CCA) using CANOCO™ v4.5 (ter 
Braak and Smilauer 1 992). A life-history trait 
data set was constructed using observations 
in the field and from other sources. Growth 
form, leaf type, dispersal syndromes and 
seed bank classes are shown in Table 1. 
Plant persistence syndromes were also 
included in the analysis. An environmental 
matrix was constructed using the variables 
recorded in each vegetation survey 
quadrat and climatic parameters extracted 
from BIOCLIM™ (Busby 1991). For the 
purposes of analysis, Australian Map Grid 
Eastings and Australian Map Grid Northings 
were included as covariables. The significance of 
envirormiental variables was determined using 499 
Monte Carlo Permutations and the forward selection 
option in CANOCO™ v4.5 (ter Braak and Smilauer 



1992). 



RESULTS 



Fire-response traits, persistence syndromes and 
fire intensity 

Data on the fire-response of 49 woody taxa 
were collected from transects at WPNP and MGNP. 
Resprouting was the dominant response to fire (> 
80% of taxa. Table 2) and there was no difference in 



Proc. Linn. Soc. N.S.W., 127, 2006 



65 



RESPONSES OF UNDERSTOREY SPECIES TO FIRE 



Table 2. Summary of contingency-table analysis for fire-response traits of woody understorey species 
following high intensity and low intensity fires in National Parks on the New England Tablelands. Fig- 
ures are the No. of species in each category. 





High intensity fire (WPNP) 
Ridge Slope Gully Total 


Low intensity Fire (MGNP) 
Ridge Slope Gully Total 


Fire Response (II- VI) ^^W 


















II. Killed, soil-stored seed bank 


5 


5 


4 


6 


4 


1 


1 


2 


IV. Resprouts via root suckers 


3 


4 


4 


3 


1 


2 


4 


3 


V. Resprouts via basal stem 


13 


14 


12 


16 


10 


14 


13 


18 


VI. Resprouts via stem bud bank 


1 


1 


1 


1 


1 


1 


1 


1 


Both IV & V 


5 


4 


5 


8 


3 


4 


4 


3 


Fire Response 


















Killed (> 70% killed) 


5 


5 


5 


6 


3 


1 


1 


2 


Resprout (< 30% killed) 
Variable (30 - 70% killed) 


21 

1 


23 



20 

1 


27 

1 


15 
1 


20 
1 


21 
1 


22 
3 



the frequency of fire-response traits between crown 
fire (WPNP) and understorey fire (MGNP) (G^ = 
2.68, P > 0.05). The fi^equency of resprouting traits 
was also consistent across topographic gradients 
within parks (WPNP, G^ = 0.87, P > 0.05; MGNP G^ 
= 4.19, P > 0.05). Resprouting firom basal stems was 
the most common fire-response, followed by species 
regenerating from both basal stems and root-suckers 
and then those regenerating from root-suckers alone 
(Table 2). Species killed by fire were less common in 

Table 3. Summary table of contingency analysis for persistence syn- 
dromes of woody understorey species after crown and understorey fires 
and between leaf-type classes. Persistence syndromes defined by Pau- 
sas et al. (2004), presence or absence of seedlings refers to post-fire re- 
cruitment only. Species killed by fire and lacking post-fire recruitment 
rely on the dispersal of propagules into a burnt area for recruitment. 





Low intensity 


High intensity Sclerophyllous 


Mesophyllous/ 




(WPNP) 


(MGNP) 


..iv»'^-i 


. Coriaceous 


Resprouters 








1 


+ seedlings 


9 


10 


12 


1 


- seedlings 


21 


15 


2 


29 


Killed 










+ seedlings 


6 


2 


3 


4 


- seedlings 


1 


2 


1 


2 



the landscape, although a greater number of species 
killed by fire were recorded at the high-intensity fire 
sites (WPNP, 17%) than at the low-intensity fire sites 
(MGNP, 7%) (Table 2). 

Additional observations recorded outside 
transects were included in the persistence syndrome 
data set, with persistence traits of 54 woody species 
included in the analysis. Resprouting was the most 
common persistence syndrome, although resprouting 
without post-fire seedling recruitment was more 

frequent than resprouting with 
post-fire seedling recruitment 
(Table 3). Species killed by 
fire were low in frequency 
at both sites. There was no 
significant difference in the 
relative frequencies of any 
of the persistence syndromes 
between high-intensity fire 
sites and low-intensity fire 
sites (G2= 2.84, P> 0.05). 

Persistence syndromes and 
leaf types 

Sclerophyllous (dry forest 
taxa) species had a higher 
frequency of resprouting 
species with post-fire 
seedling recruitment than 
mesophyllous and coriaceous 



66 



Proc. Linn. Soc. N.S.W., 127, 2006 



M.L. CAMPBELL AND PJ. CLARKE 




Figure 1. Biplot diagram of CCA ordination for environmental variables and 
life history traits of 61 woody taxa in wet sclerophyll forest. D= life history traits; 
solid arrows = significant environmental variables (499 Monte Carlo Permu- 
tations, P< 0.05). RS-, RS+, K-, K+ = persistence syndromes defined by Pau- 
sas et al. (2004); passive, wind, vertebrate, invertebrate = dispersal syndromes; 
Spp Rich = species richness; AMMI = annual mean moisture index; MICV = 
moisture index coefficient of variation. Plot axes are 1 by 1 units of ordination. 



taxa {&= 22.66, P < 0.0001) (Table 3). Conversely, 
rainforest taxa (mesophyllous and coriaceous plants) 
had significantly greater frequency of species that 
resprouted and lacked post-fire seedling recruits 
(Table 3). There was no difference in the frequency 
of species killed by fire with or without post-fire 
seedling recruitment between leaf type groups (Table 
3). 

Plant traits and environmental variables 

Shrubs were the most common woody growth 
forms at both sites, but trees and small trees were 



more common at WPNP (Table 1). There were 
more sclerophyllous and mesophyllous taxa than 
coriaceous at both sites and most species had 
vertebrate-dispersed propagules (Table 1). Soil-stored 
seed banks were the predominant type with less than 
20% of species having a canopy-held seed banks or 
relying on dispersal of seed for regeneration (Table 
1). More detailed information on the life-history traits 
of individual species is given in Appendix 1 . 

Life-history traits of woody taxa in the wet 
sclerophyll understorey were significantly correlated 
with canopy cover, slope and understorey cover 



Proc. Linn. Soc. N.S.W., 127, 2006 



67 



RESPONSES OF UNDERSTOP^Y SPECIES TO FIRE 



(Fig. 1). Mesophyllous and tree attributes correlated 
with increasing canopy cover, as did resprouting 
species that lack post-fire seedling recruitment (Fig. 
1). Species killed by fire and with post-fire seedling 
recruitment irom in situ seed banks were negatively 
associated with increasing canopy cover, but positively 
correlated with increasing ground cover. Resprouting 
species with post-fire seedling recruitment were 
weakly associated with increasing understorey cover, 
but correlated closely with moisture predictability 
and temperature (Fig. 1). Species with canopy-stored 
seed banks and wind-dispersed seeds were positively 
correlated with landscape slope (Fig. 1). Species with 
passive dispersal, those lacking in situ seed banks, 
and those killed by fire with seeds dispersed into the 
post-fire environment, were all negatively correlated 
with increasing understorey cover (Fig. 1). 



DISCUSSION 

Do fire response traits vary with fire intensity? 

We found that the majority (>80%) of woody 
understorey species in montane wet forests resprouted 
after fire irrespective of fire intensity. Whilst fire 
intensity was not replicated across sites our general 
observations in wet forests in the region support this 
finding. This contrasts with observations that high- 
intensity fires limited vegetative regeneration in 
Victorian wet sclerophyll forests, where rootstocks 
and bud banks did not survive high temperatures. 
Conversely, after lower- intensity fire, the predominant 
mechanism of regeneration was vegetative with a 
high density of root-suckers and resprouting adults in 
the post-fire environment (Ashton and Martin 1996). 
In our study, less than 10% of species demonstrated 
a variable response to fire, indicating that the 
dichotomous classification of species into obligate 
seeders (< 30% resprout) and resprouters (> 70% 
resprout) is a useful generalisation for these systems. 

The proportion of resprouting species recorded 
here is high in comparison to wet sclerophyll 
forests in Victoria (30%) and southwest Western 
Australia (24%), but comparable to data for coastal 
wet sclerophyll forest in northern New South Wales 
(60%) (Ashton 1981). Similarly high proportions of 
resprouting species have been reported for other highly 
competitive systems such as wet heaths and grassy 
woodlands on the New England Tablelands (Clarke 
and Knox 2002; Clarke et al. 2005). Resprouting 
may be favoured in productive habitats as vegetative 
recruits and regeneration are competitively superior to 
seedlings (Clarke et al. 2005). This may also explain 



the low frequency of obligate seeding species and 
post-fire recruitment in these communities. We also 
recorded high levels of root-suckering, and roughly 
one quarter of the species were capable of resprouting 
from the basal stem and roots after fire. Resprouting 
from basal stem tissue suggests that moderate fire- 
frequencies have been a selective force (Bellingham 
and Sparrow 2000), but resprouting may also confer 
an advantage by enabling individuals to survive and 
regenerate after mechanical damage inflicted by tree- 
and limb-fall (Ashton 2000; Paciorek et al. 2000; 
Kanno et al. 2001). Similarly, root-suckering is an 
effective means of invading unoccupied space after 
disturbance events such as tree-fall (Stocker 1981; 
Kammesheidt 1999; Bellingham and Sparrow 2000). 

Do fire-response traits vary between rainforest 
and sclerophyllous species? 

There was a clear dichotomy between rainforest 
(mesophyllous and coriaceous) and sclerophyllous 
taxa in relation to post-fire seedling recruitment. The 
majority of sclerophyllous taxa that resprouted also 
had post-fire seedling recruitment, whereas most 
rainforest taxa resprouted but lacked post-fire seedling 
recruitment. This may be explained by resource 
gradients across the wet sclerophyll ecotone affecting 
species composition. At the rainforest interface the 
quantity and quality of light reaching the forest floor 
is much lower than at the eucalypt forest edge and 
these conditions are generally unfavourable for the 
recruitment of shade-intolerant taxa (Turton and Duff 
1992). The prevalence of mesophyllous species at the 
rainforest interface reduces the probability of fire and 
species with gap-phase regeneration dominate the 
community (Unwin 1989; Adam 1992; Harrington 
and Sandersen 1994). The general absence of post- 
fire seedling recruitment in rainforest taxa is likely 
to reflect that recruitment syndromes in rainforest are 
linked to canopy disturbance rather than ^re perse. In 
contrast, post-fire seedling recruitment was common 
in sclerophyllous taxa: these species respond to 
broad-scale disturbance in order to regenerate and 
gap-phase recruitment is rare (Melick 1990). 

The dichotomy in seedling recruitment 
syndromes was reflected in the CCA, with species 
that lack post-fire recruitment closely associated 
with increasing canopy cover. In contrast, species 
with post-fire seedling recruitment, regardless of the 
adult plant response to fire, were positively associated 
with more open habitats as indicated by increasing 
ground and understorey cover. At the landscape scale, 
productivity gradients have been linked to ratios 
of obligate seeders to resprouters across habitats 
(e.g. Clarke et al. 2005). However, within the wet 



68 



Proc. Linn. Soc. N.S.W., 127, 2006 



M.L. CAMPBELL AND P.J. CLARKE 



sclerophyll ecotone, the productivity gradient appears 
to be driving the prevalence of post-fire seedhng 
recruitment more than adult fire responses. 

Paradoxically, most rainforest species vigorously 
resprouted and similar numbers of rainforest and 
sclerophyll species were killed by fire. Other 
studies have reported rainforest species coppicing 
or resprouting after fire (Stocker 1981; Chesterfield 
et al. 1991; WiUiams 2000). Hence the notion of a 
split 'fire-intolerant' vs. 'fire-tolerant' flora does 
not appear to be explained simply by differences in 
resprouting ability. Recently, however, Fensham et 
al. (2003) demonstrated that recurrent fires caused 
increased mortality in tree species from monsoon 
rainforest compared to surrounding savannah, 
suggesting fundamental differences in sprouting 
ability. Remaining unresolved is the question of 
whether quintessential sclerophyllous species are 
more 'fire-tolerant' than their mesophytic cousins in 
the same genus or family, and what the mechanisms 
for this tolerance are. 



ACKNOWLEDGEMENTS 

We thank Richard Willis and Shanti Virgona for 
assistance in the field. Financial support was provided to 
MLC by an Australian Post-graduate Award (Industry), 
New South Wales National Parks and Wildlife Service, 
University of New England and NCW Beadle scholarship. 
Lachlan Copeland kindly assisted with the identification 
and nomenclature of plant taxa. 



REFERENCES 

Adam, P. (1992). 'Australian rainforests'. (Clarendon 

Press, Oxford). 
Ashton, D.H. (1981). Fire in tall open-forests (wet 

sclerophyll forests). In 'Fire and the Australian biota' 

(Eds A.M. Gill, R.H. Groves and I.R. Noble) pp. 339- 

366. (Australian Academy of Science, Canberra). 
Ashton, D.H. (2000). The Big Ash forest, Wallaby Creek, 

Victoria - changes during one lifetime. Australian 

Journal of Botany 4S, 1-26. 
Ashton, D.H. and Attiwill, RM. (1994). Tall open-forests. 

In 'AustraUan Vegetation' (Ed. R.H. Groves) pp. 157- 

196. (Cambridge University Press, Cambridge). 
Ashton, D.H. and Martin, D.G. (1996). Regeneration in a 

pole-stage forest of Eucalyptus regnans subjected to 

different fire intensities. Australian Journal of Botany 

44, 393-410. 
Bellingham, P.J. and Sparrow, A.D. (2000). Resprouting 

as a life history strategy in woody plant communities. 

Oiito5 89, 409-416. 
Binns, D.L. (1991). Vegetation dynamics oi Eucalyptus 



microcorys — E. saligna wet sclerophyll forest in 
response to logging. MSc Thesis, University of New 
England, Armidale. 

Bond, W.J. and van Wilgen, B.J.R (1996). 'Fire and 
plants'. (Chapman and Hall, London). 

Busby, J.R. (1991). Bioclim - a bioclimate analysis and 
prediction system. Plant Protection Quarterly 6, 8-9. 

Chesterfield, E.A., Taylor, S.J. and Molnar, CD. (1991). 
Recovery after wildfire: warm temperate rainforest 
at Jones Creek, East Gippsland, Victoria. Australian 
Forestry SA, 157-173. 

Clarke, PJ. and Knox, K.J.E. (2002). Post-fire response 
of shrubs in the tablelands of eastern Australia: 
do existing models explain habitat differences? 
Australian Journal of Botany 50, 53-62. 

Clarke, RJ., Knox, K.J.E., Wills, K.E. and Campbell, M. 
(2005). Landscape patterns of woody plant response 
to crown fire: disturbance and productivity influence 
sprouting ability. Journal of Ecology 93, 544-555. 

Fensham, R. J., Fairfax, R.J., Butler, D.W. and Bowman, 
D.J.M.S. (2003). Effects of fire and drought in a 
tropical eucalypt savanna colonized by rain forests. 
Journal of Biogeography 30, 1405-1414. 

Florence, R.G. (1996). 'Ecology and silviculture of 

eucalypt forests'. (CSIRO Publishing, Collingwood). 

Gilbert, J.M (1959). Forest succession in the Florentine 
Valley, Tasmania. Proceedings of the Royal Society of 
Tasmania 93, 129-151. 

Gill, A.M. and Bradstock, R.A. (1992). A nafional register 
for the fire response of plant species. Cunninghamia 
2, 653-660. 

Harden, G. J.E. (1990). 'Flora of New South Wales Vol. 
1 '. (NSW University Press, Kensington). 

Harden, G.J.E. (1991). 'Flora of New South Wales Vol. 2'. 
(NSW University Press, Kensington). 

Harden, G.J.E. (1992). 'Flora of New South Wales Vol. 3'. 
(NSW University Press, Kensington). , 

Harden, G.J.E. (1993). 'Flora of New South Wales Vol. 4'. 
(NSW University Press, Kensington). 

Harrington, G.N. and Sandersen, K.D. (1994). Recent 
contraction of wet sclerophyll forest in the wet 
tropics of Queensland due to invasion by rainforest. 
Pacific Conservation Biology 1, 319-327. 

Henderson, M.K. and Keith, D.A. (2002). Correlation 
of burning and grazing indicators with composition 
of woody understorey flora of dells in a temperate 
eucalypt forest. Austral Ecology 27, 121-131. 

Kammesheidt, L. (1999). Forest recovery by root 

suckers and above-ground sprouts after slash-and- 
bum agriculture, fire and logging in Paraguay and 
Venezuela. Journal of Tropical Ecology 15, 143-157. 

Kanno, H., Hara, M., Hirabuki, Y, Takehara, A. and Seiwa, 
K. (2001). Population djoiamics of four understorey 
shrub species during a 7-yr period in primary beech 
forest. Journal of Vegetation Science 12, 391-400. 

Melick, D.R. (1990). Ecology of rainforest and sclerophyll 
communities in the Mitchell River National Park, 
Gippsland, Victoria. Proceedings of the Royal Society 
of Victoria 102, 7 1-S7. 



Proc. Linn. Soc. N.S.W., 127, 2006 



69 



RESPONSES OF UNDERSTOREY SPECIES TO FIRE 



Morrison, D.A. and Renwick, J.A. (2000). Effects of 

variation in fire intensity on regeneration of co- 
occurring species of small trees in the Sydney region. 

Australian Journal of Botany 48, 71-79. 
Paciorek, C.J., Condit, R., Hubbell, S.R and Foster, R.B. 

(2000). The demographics of resprouting in tree and 

shrub species of a moist tropical forest. Journal of 

Ecology 88, 165-111. 
Pausas, J.G., Bradstock, R.A., Keith, D.A. and Keeley, 

J.E. (2004). Plant functional traits in relation to fire in 

crown-fire ecosystems. Ecology 85, 1085-1100. 
Sokal, R.R and Rolf, F.J. (1981). 'Biometry'. (Freeman 

and Company, San Francisco). 
Specht, R.L. (1970). Vegetation. In 'The Australian 

environment' (Ed. G.W. Leeper) pp. 44-67. (CSIRO B .- S 

? ^ I 



K 



^ H 5 



and Melbourne University Press, Melbourne). 

Stocker, G.C. (1981). Regeneration of a north Queensland E "^ ^ «« 

rainforest following felling and burning. Biotropica <S, ^ Si ^ 

13, 86-92. '^ 1 "S 3 

ter Braak, C.J.F. and Smilauer, R (1992). 'CANOCO "S, J s "I 

version 4.5 reference manual & users guide to 2 ^ "^ S 

CANOCO for Windows'. (Microcomputer Power, "3 ^ '.^ | 

Ithaca, NY). ^ s § Z 

Turton, S.M. and Duff, G.A. (1992). Light environments I — S S 

and floristic composition across open forest-rainforest « S *« "S 

boundary in northeastern Queensland. Australian 5 S S" 5 

Journal of Ecology \1,A\5A23. o 2 « ^ 

Unwin, G.L. (1989). Structure and composition of ^ ■« y « 

the abrupt rainforest boundary in the Herberton vo § -3 "O 

Highland, North Queensland. Australian Journal of *© "o •■" ? JJ^ 

Botany 37, 413-42%. ^ g, *^ g g; 

Vesk,P.AandWestoby,M. (2004). Global patterns of 1 * a ^^ 

sprouting ability: can all plant species be divided into ' ^ 



a «\ 



Sprouters and non-sprouters? Journal of Ecology 92, ^- « ^ - 

310-320. 1 «. r« ^ 




.a « f^ (« ov 



Whelan,R.J. (1995). 'Ecology of fire'. (Cambridge ^ _ _ - 

(^N flj f* fM ' • 

University Press, Cambridge). "j a. ^ ■* iT 

Williams, P.R. (2000). Fire-stimulated rainforest seedling i-h '^ s' * on 

recruitment and vegetative regeneration in a densely .S o '^^ S ^ 

grassedwetsclerophyll forest of north-eastern "g ^ « « § 

Australia. Australian Journal of Botany 48, 651-658. O. -S ^ cc "2 

3'^ ^ * 5 

-< s ^ * PC 



70 Proc. Linn. Soc. N.S.W., 127, 2006 



M.L. CAMPBELL AND P.J. CLARKE 



■i •= + 



IX 



^ 












■a 








c 












u 








CB 


^^ 


^^ 


_^ 


^^ 


^^ 


S2 


^^ 


, 


^^ 


CQ 




















o 


o 


o 


o 


o 


o 


o 


o 


o 




t>0 


(Zl 


C/3 


c« 


Ol 


.2 


c/o 


(Zl 


CO 



(U 

o 






O 



o 

o 



u 

a. 

C/5 



E 



z z 

A. O^ 

3= ^ 



Oh 


dn 




&, 




CL, 




eu 




z 


z 




Z 




z 




Z 




eu 


Pu 




CL| 




a. 




cu 




^ 


^ 




^ 




^ 




5^ 




oT 


oT 


a. 


oT 


CL. 


cC 


0- 


ClT 


a, 


Z 


Z 


Z 


Z 


z 


Z 


z 


Z 


Z 


o 


a 


a 


o 


a 


o 


0. 


a 


o 


2 


2 


2 


s 


s 


s 


5: 


S 


s 



Oh 

z 

Oh 



o. 

Z 

o 



CU Oh 

z z 

Oh Oh 



« 

Oh 

z 

Oh 



o- 

Z 

a 



o 





0. 

Z 










1 


1 






Oh 










Oh 


CL. 






^ 










^ 


5: 




CU 


Oh' 


o- 


Oh 


o- 


Oh 


Oh' 


Oh' 


Oh 


Z 


z 


z 


z 


z 


z 


Z 


z 


z 


Oh 


C3 


Oh 


o 


o. 


Oh 


a 


a 


CU 


^ 


s 


^ 


s 


^ 


5^ 


s 


s 


^ 



> 
> 



> 
> 



S S > > > K 



> 
> 



> 



> = 



> 
> 



> 
> 



>> 


>> 


■a 
1) 




























O 
g 


c3 




'S 


!Z1 


■3 


'0 
1/5 


'0 
!/3 


'S 
t/3 


'5 

tzi 


'0 


'0 


'0 
t/3 


.1 


'S 

IZl 


'S 

»2 



u 


u 





















cd 


CO 


cd 


rt 


"5 


tH 


tH 


Uh 


Uh 


iM 


XI 


.C 


J2 


XI 


XI 


u 


u 


U 








■c 


■e 


■c 


■c 


■c 


u 


u 


u 





u 


> 


> 


> 


> 


> 


c 


c 


a 


c 


c 



u 


U 











3 






u 


u 





u 







U 









u 


<L> 


CS 


13 


es 


5^ 


T3 


-a 


rt 


U 


<u 


« 


« 


^ 


"cd 


td 


U 


"^ 


u 





rt 


"S 


■a 


h. 


Ui 


w 


_> 


;. 


> 


> 


a 


b< 


c. 


b^ 


Lh 


> 


tH 


> 


> 


C^ 


u- 


t: 


X3 


X> 


X5 


M 


C 


c 


X) 






X5 


XI 


J2 


Xi 


X> 


"rr> 


XI 


*« 


'w 


X) 


XS 


X3 








U 


C/5 


§ 


^ 


1) 


O] 


05 


U 








u 











u 


u 


■c 


■c 


■c 


«^ 


C 


JS 


ca 


■c 


■c 


t; 


t 


t 


CO 


■c 


C8 


CB 


t 


■d 


t 


J!> 


JU 


ju 


CU 






^D 


CU 


CU 


JU 


Ji> 


JU 


^w 


.<!> 


Oh 


^w 


Oh 


CU 


JU 


^ 


^w 


> 


> 


> 








> 






> 


> 


> 


> 


> 




> 






> 


> 


> 



>. 


>. 


>^ 


>> 


>. 




>-, 


J5 




>. 


>. 




j3 


>1 


D 


x: 


x: 


x: 


x: 


J3 


x: 


J= 


x: 





Lh 


§■ 






s 


a. 




Ih 


0, 


s 


1 


§■ 

Ui 


0, 





D. 
g 


8 

CO 

■g 














tzi 




on 


s 




00 


S 


s 




00 





2 


s 


S 



x; 

Q. 
O 



00 



x: 
a, 
o 



O 



3 

8 


3 

§ 


x: 


3 
g 


3 

g 




>> 

x: 





CJ 


&■ 





u 


S< 


g. 


CO 


CO 





.2 


.2 












w 






Cfl 


t/3 


'C 


'C 


w 


'C 


*c 


a> 







u 



U 


s 




u 




u 


2 


2 



x: 

B. 
O 

CO .-H .— 



-- -- r^ r^ 



XI 

5 



X3 

2 

x: 
00 



XI 

00 



X) 

2 



H 



cd A A A cQ A r^ 

E E E S E E 

C/D M C/l 05 t/3 (Zl 



— <u 2 

=^ fi: ^ 

















■| 





•1 
1 

c« 
















iS 
































.^ 




X) 
















S 












CO 






C/3 


CO 

1 


'5 
■2 


X 


1 


* 

.2 

1 


E 

CO 
CO 

i 


a 
to 
00 

OS 
CO 




x> 

c« 

E 


.2 

1 
u 
fc 




3 
.2 
1 
■fe 


* 


CO 

3 

E 



2 

'S 


CO 
CA 

_o 

■3 
u 

1 


.2 
1 

2 


> 

"o 

E 

3 


CO 

•a 
tc 
*c 

T3 
CO 

§■ 


CO 

'S 

8 

> 

cS 

CO 




1 
00 

CO 


CO 

'5b 

•c 

cS 


-a 
■| 

CO 

■«> 


t/5 

1 

CO 


* 

CO 
00 

& 
CO 

ao 
u 

E 


CO 

eg 

1 


2 
"3 

u 

CO 


1 

E 


J 


3 

CO 




1 

.2 


E 
.2 


E 

CO 


]3 

.2 


(U 

X3 



•a 


■s 

1 


2 

CO 

2 

to 


.2 

C3 


CO 


2 
'> 

_3 
"0 

T3 

CO 


CO 

E 


"a 


E 


c 
c 


CO 

E 

CXI 

1 




1^ 






CO 

8 

0. 


1 


CO 

1 

Xl 

g 


CO 

2 

CO 




CO 

c 

■a 



2 


x: 

&> 




3 

e- 

CO 

8 

CO 


2 

•§ 
.2 
•5 

c 


.2 

n. 

3 


< 


< 


< 


< 


< 


< 


<; 


<: 


CQ 


CQ 


DQ 








U 





U 








Q 


5 


Q 


Q 


m 


PL) 


W 



u 





lU 


<D 







s 


CO 


CO 


CO 




u 


u 


u 


(U 


CJ 


t) 


tj 








CO 


CO 


CO 


CO 


CO 


x> 


X) 


X) 


X5 


X) 


CO 


CO 


CO 


CO 


CO 


tu 


Pu 


pu 


tu 


CU 



u 

CJ 

CO 



u 

CO 
D. 
PJ 



■c 



3 





CO 





<u 















CO 


CO 













CO 


<u 


u 


<U 











A 







y 


rt 


CO 


CO 


CO 


OJ 


■Q 


CO 


.2 


(U 





a> 


<D 





Lh 



x: 


1 


'S 



2 


2 


2 


2 


2 


0. 


c 


e 


3 


^ 


3 


3 


3 


3 


3 


CO 


3 


CO 


CO 


CU 


CQ 


U 


U 


hJ 


Bi 


hJ 


J 



•a 

c 



X) 

PU 



pj 



i:i 





3 


0. 


CO 


3 


J 


PU 



Proc. Linn. Soc. N.S.W., 127, 2006 



71 



RESPONSES OF UNDERSTOREY SPECIES TO FIRE 












o 







a a a 



a 



o 



o 



&< C^ a, 

Z Z g 

a o ph 

S S ^ 



o 



o 




« £ 



o 

Oh 



> 






> 






> 




> 


=a 


> 


> 


'^ 


> 


> 


•^ 


> 


^ 


> 






> 






t 




> 



> 









ea 
03 

-a 



o 

C/3 



o 
c 

U 



o 
t/l 



o 

00 



o 

C/2 



ft 


c 


o 


o 


c 


c 


S 


5 


U 


U 



o 



O 
1/3 



■a 


■T3 


T3 


(U 


U 


<L) 


C/3 


CA 


M 


Vh 


;-i 


tH 


(U 


O 


1) 


ft 


ft 


ft 



o 

00 



T3 


-o 


U 


<u 


M 


c« 


t-i 


l-i 


1) 


<u 


ft 


ft 



o 


o 


o 


o 


o 


o 


o 


o 


00 


00 


00 


00 


C/5 


on 


1/3 


(/5 



Q Q 



•a 
o 



o 

ft 



s 




a> 


a 


a> 


U 






o 


a> 








a> 






a> 


u 


u 


u 






u 


ed 






ca 








> 
















> 










> 


1 






-a 


2 




2 


B 


T3 


1 


2 


•T3 


T3 


•a 


■§ 


-a 


1 


B 


i 


1 


B 


u 


c 


J2 


^ 


x> 


C 




^ 


^ 


C 


C 


B 


X) 


a 




XI 


X3 


X) 


X) 


S 

s 


M 

W 
i^ 


X3 




'$ 






■c 


S 


■^ 


^ 


^ 


^ 


^ 


^ 


^ 


'$ 




^ 


s 


^ 


^ 


g 




^ 


> 


^ 


;^ 




a. 


;^ 


;^ 








^ 




Ph 


:^ 


;^ 


:^ 


:^ 


(X 


Oh 


;^ 



X! 

u 

> 






>1 

ft 
O 



ft 
o 



o 

00 



x: 
ft 
o 

u 



ft 

O 

(A 

o 



>. 


>-, 


>> 


>, 


x: 


3 


3 


g 


3 


x: 


x: 


JS 


JS 


O 


O 


o 


o 


r 


& 


ft 
o 


ft 
o 






.2 


0^ 


jj 


_B 


<D 


a> 




•c 


•c 


c 


■c 


oo 


00 


00 


00 


o 
U 


o 

U 


o 

U 


o 

U 






■g. 


?-. 


■g. 


>-. 


?-. 


>. 


>> 


j3 


>. 


J3 


>. 


x: 


x: 


J= 


x: 


x: 


x: 


J3 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


ft 


o 
u 


1 


o 


2 


a 


2 


1 


o 

s 


2 


o 


1 


s 


00 


s 


00 


V3 


73 


u 

00 


s 


00 


S 


OO 



o 

I 

2 
a 



"3 "3 
OO VI 



a 
B 

00 



F^-r .^^ -J-^ 

^11 



X3 

2 



a 

B 

!/3 



x: 
oo 



XS 
C/l 



I 



ca 

E 



^ 



s 



ft 

00 



.2 


ca 


a 


s 


.1 


.a 


g 


CO 




ca 


ca 


ca 


cd 


■c 

u 


^ 


2 




ca 


o 


o 


"3 


u 


>. 


ap 


o 


M 


■o 




o 


ca 


u 


^ 


a 


ffi 


ffi 


C3 



a> 


O 


1 


u 

C 
_2 


C 


c 


o 


o 


M 


CiO 


O 


o 


ft 


ft 


O 


o 


o 


o 


3 


3 


(U 


u 


l-J 


J 



.2 



ca 

E 
o 



" 2 

c 

o 

s 



o 

X5 
ft 
O 



ft 

o 
o 



o 



a 

> 

o 



.2 



o 
ft 
ft 
o 
ca 



3 

x: 



ca 
E 
O 






•c 
O 



o 



§ 

o 
o 

CB 

'S 
O 
O 

CA 
b« 
O 



CO 

ca 

■c 



s E s a 



o 
o 

CA 

Oh 



c 

O 
O 

CA 

Ui 
<L> 
BLi 



o 
o 

CA 

U 

PLi 



E 






I 

O 
u 

a 
ca 

"u 

E 



6 

S 
o 

ft 

CA 

o 



X3 

_o 
"o 

-O 

a 

Oh 



i 



ca 

X) 

ca 





ca 




«3 


ca 


o 


(U 




<U 


u 


s 


ca 

E 


u 


1^ 
■a 


■a 
■C 


<1> 

2 


1 




ft 


o 

ca 

ft 


ft 


tin 


w 


W 



ca 


d^ 


(U 


ca 


CJ 


u 




u 


"t^ 


ca 


>> 


^ 


S 


Oi 



u 


U 


Q 


u 


« 


« 






o 

X3 
ft 
3 

W 



o -s o o o o 

;-l ^ )-• (-1 u ^ 



o 
ft 



o 



ca 

X) 



72 



Proc. Linn. Soc. N.S.W., 127, 2006 



M.L. CAMPBELL AND P.J. CLAIIKE 



o 
S 




PL, 



a o o 






> > > 



> 

■a > 
> 



> H 1=1 



o o 
CO oo 






o o o o o o 
c/3 en 00 (/I M OQ 



o o o o 

CO CO CO C/3 



o o 

CO CO 






(1> 4=> -g 



■a -H -s « 



.£3 


•u 


x> 


43 


.o 


.r> 


J2 


XI 


-D 


-D 


XI 


X 


X 


U 


i 


o 


lU 


u 


d) 


u 


u 


<L> 


<U 


(U 


(U 


u 


-c 


■c 


■c 


■c 


■c 


■c 


■c 


■c 


■c 


-c 


•c 


•c 


^ 


> 
a 


;^ 


^ 


^ 


:^ 


^ 


:^ 


:^ 


1^ 


;^ 


^ 


:^ 



X 


X 


u 


u 


■c 


•c 


u 


1) 


> 


> 


e 


c 



ex 
o 



o 



x; 
a. 
o 

u 

2 



3 


3 














ta 


ca 


■c 


•c 


o 


o 


U 


U 



>, 


>, 


>, 


>> 


>. 


>-, 


X 


X 


X 


X 


X 


X 


o. 


EX 


a. 


o. 


o. 


o. 


o 


O 


o 


o 


o 


o 


M 


w 


w 


M 


M 


M 


u 


1) 


o 


1) 


o 


(U 


s 


s 


s 


s 


s 


2 



>. 


>. 


">> 


*>» 


X 


X 


JZ3 


>• 


1 


1 




& 

S 




00 


s 


S 









u 


1> 










P 


u 


^ 


•§ 


f^ 


^ 


£ 






















CO 


CO 


C8 

6 

CO 


CS 

00 



B 
on 



!: I S 

"rt ^ 'rt 



-§ 



3 
X 



"o 

Oh 



CQ 



S3 

o 



•5 ■S >> 



s 
ca 

C 

'% 
o 

X 
ca 

c 
ca 



> 

ca 
<u 

c 
ca 



3 
-O 



-3 g. 



ca 

.g 
ca 



f2 



S. 



§ 



& e- 



o 

X 



a 
u 
u 



o 
ca 

•c 



^ 



ca 
■g 



N N 



ca 


gl 


K 




o 




ca 


o 




c3 


03 


X 




ca 


'^ 


fc 





o 


u 




u 






fll 


<U 










s 


S 




ca 


C^ 


Q> 






u 


u 


a> 


u 


U 


CB 


2 




ca 


s 


w 


ca 


CJ 


o 


U 


2 


2 


H 


ca 


o 


u 


ca 


c« 


u 


c 


e 


o 


1 


ca 


o 


h 


C; 


ca 


'^ 


'^ 


-^ 


a 


.2 


a 


^ 


1 




^ 

S 


s 


U 


00 









o 


u 






CS 


ca 






u 


u 






^ 


i^ 


52 


K 


•c 


13 


ca 


ca 

8 


^ 


CJ 

ca 


3 


S 


o. 


o. 


3 


3 


u 


W 


»:; 


Oi 



Proc. Linn. Soc. N.S.W., 127, 2006 



73 



74 



A Preliminary Assessment of Disturbance to 
Rock Outcrops in Gibraltar Range National Park 

Ross L. GOLDINGAY AND DaVID A. NeWELL 

School of Environmental Science and Management, Southern Cross University, Lismore, NSW, 2480 

(rgolding@scu.edu.au) 



Goldingay, R.L. and Newell, D.A. (2006). A preliminary assessment of disturbance to rock outcrops in 
Gibraltar Range National Park. Proceedings of the Linnean Society of New South Wales 111, 75-81. 

The signilicance of habitat disturbance within protected areas remains poorly understood. This study 
assessed habitat disturbance to granite rock outcrops within a protected area in north-east New South 
Wales. Survey sites were classed as near (<350 m) or far (>500 m) from roads and walking fracks. Habitat 
disturbance was dependent on site category, occurring at 8 of 10 near sites compared to 1 of 12 far sites. 
Disturbance mostly consisted of the construction of rock cairns that may deplete the availability of loose 
rocks at a site. Reptiles were frequently found sheltering under loose rocks, attesting to the valuable 
microhabitat that this type of substrate provides. Further research is required to understand the significance 
of this disturbance and the extent of dependence by the local reptile fauna on this substrate. Our data 
provide a baseline against which ftiture surveys can be compared. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 
KEYWORDS: habitat disturbance, rock-dwelling reptiles, rock outcrops. 



INTRODUCTION 

A common assumption in developed countries 
is that species and their habitats contained within 
protected areas will be adequately conserved (e.g. 
Primack 1998; Brooks et al. 2004; Higgins et al. 2004; 
Molnar et al. 2004). Indeed, much effort and many 
resources have been put into expanding protected 
area networks to extend such protection, particularly 
in New South Wales (Davey et al. 2002; Pressey et al. 
2002; Newell and Goldingay 2004). Many protected 
areas are managed specifically for recreational use. 
Where this occurs management is often focused 
on minimizing the impacts of users in areas where 
recreational activities are concentrated (e.g. NPWS 
2000a,b). However, recognition is emerging that 
protected area users may diminish the quality of some 
wildlife habitats over broad areas (Goldingay 1998; 
Newell and Goldingay in press). 

One important case study that implicates 
protected area users in the widespread degradation 
of wildlife habitat is that of the broad-headed snake 
(Hoplocephalus bungaroides). This endangered 
species has a geographic range completely restricted 
to the Sydney basin (Swan 1990; Cogger 1992), 



where it shelters within sandstone rock outcrops 
during the cooler months of the year (Webb and Shine 
1998). It is known from a number of protected areas 
and its conservation appears dependent on how well 
these areas are managed (Cogger et al. 1993). Several 
studies have demonstrated that disturbance to rock 
outcrops is widespread and continuing to threaten this 
snake (Schlesinger and Shine 1994; Goldingay 1998; 
Shine et al. 1 998; Goldingay and Newell 2000; Webb 
et al. 2002). Until recently, collection of sandstone 
bush-rock for landscaping from protected areas was 
viewed as the primary cause of the decline of this 
species (Hersey 1980; Shine and Fitzgerald 1989; 
Mahony 1 997; Shine et al. 1 998). It is now recognized 
that much of this disturbance can be attributed to 
protected area users, of which there appear to be three 
types involved: hikers, reptile poachers and vandals 
(Goldingay and Newell 2000; Newell and Goldingay 
in press). 

Whilst concern about rock habitat degradation 
in Australia has been driven by its impact on the 
broad-headed snake, this is not the only species that 
is affected (see Schlesinger and Shine 1994). For 
example, Newell and Goldingay (in press) detected 
a fiirther 19 reptile species under loose rocks (eight 



DISTURBANCE TO ROCK OUTCROPS 




1b 




assessment of rock habitat disturbance 
in Gibraltar Range National Park. This 
protected area occurs in north-eastern New 
South Wales and is characterized by many 
areas of distinctive granite rock formation 
that have associated rock outcrops. 



METHODS 

Study Area 

Gibraltar Range National 
Park (Gibraltar Range NP) is located 
approximately 100 km west of Grafton. 
It has an area of approximately 25,000 ha 
and is boimded to the north by Washpool 
National Park, which is 67,000 ha (NPWS 
2003). These parks are included in the 
World Heritage area known as the Central 
Eastern Rainforest Reserves of Australia 
(DEC 2005). 

Gibraltar Range NP contains broad 
areas of rainforest, heathland, open forest 
and woodland. Rainforest is common along 
the eastern and northern sides of the Park 
while open forest occurs through much of 
the remainder of the Park. Heathland areas 
are restricted in area and are associated 
with drainage lines that traverse the Park. 
Granite rock outcrops are widespread 
through the Park (Fig. 1). A wildfire burnt 
through much of the Park in December 
2002. 



Figure 1. (a). Granite outcrops near the Waratah 
(b). Anvil roclc showing associated outcrop. 

snakes, three geckos, seven skinks and one dragon) 
during a regional sxirvey for the broad-headed snake. 
Rock crevices are commonly used as retreat sites 
by many species of reptile and fi-og, some of which 
may be dependent on such habitat during periods of 
the year and are likely to be affected by rock habitat 
degradation. Furthermore, there is no reason to expect 
that this kind of habitat degradation will be limited 
to sandstone substrates. Therefore, there is a need 
to conduct studies at many locations to assess how 
ubiquitous rock habitat disturbance may be. Indeed, 
Goode et al. (2005) have identified that destruction of 
rock habitats is widespread in parts of the USA and 
was associated with a decreased abundance of rock- 
dwelling reptiles. 

The aim of this study was to provide a preliminary 



Trig. Survey Sites 

Areas of rock habitat suitable for 
survey were identified fi^om topographic 
maps and from ground truthing. Only areas 
with a north through west aspect were included in 
the survey because these aspects are more highly 
preferred by reptiles that rely on sheltering under 
loose rocks (Webb and Shine 1998; Pringle et al. 
2003). If outcrops with these aspects were affected by 
habitat disturbance then others would be also. Sites 
were purposefully selected to fall into one of two 
categories: either near (<350 m) or far (>500 m) fi-om 
a road or walking track. Sites had to be at least 250 m 
apart to be considered as individual sites. Sites were 
selected in the vicinity of the Waratah Trig (located 
either side of the boundary between Gibraltar Range 
NP and Washpool NP) and the Anvil Rock (Gibraltar 
Range NP) walking tracks (see Table 1 for location 
details). 

We selected rock platforms that contained at 



76 



Proc. Linn. Soc. N.S.W., 127, 2006 



R.L. GOLDINGAY AND D.A. NEWELL 




consisted of bare rock (Goldingay 1998; 
Newell and Goldingay in press). Such 
rocks were classed as "good" rocks for 
reptile use and counted. Any sheltering 
reptiles were identified. Most transects 
were surveyed by two people. If no 
evidence of disturbance was obtained on 
a transect, then a search for disturbance 
was also conducted of areas within a 50 
m radius of the transect. This was simply 
a recognition that disturbance may be 
patchy and that transects may be too 
short to adequately sample an area. Each 
site was surveyed on one occasion in 
March 2005. 



RESULTS 




Figure 2. (a). A rock cairn, showing stencils left when rocks 
have been moved from their original position, (b). An older 
rock cairn. 



least 10 loose rocks along a 50 x 20 m transect. This 
provided a reasonable minimum number of rocks 
from which to determine whether any disturbance 
had occvirred. Only rocks >10 cm in length were 
included in the assessment. Once a site was selected, 
all loose rocks along the transect were counted and 
inspected for evidence of disturbance (e.g. rock 
cairns, rock camp fires, rocks flipped over). Rocks 
were lifted to determine their suitability to provide 
habitat for reptiles. This was ascertained by scoring 
whether rocks sat neatly on the platform, whether 
they formed a narrow crevice with the platform and 
whether at least 50% of the underlying substrate 



Of 22 sites chosen for survey, 1 
occurred near and 12 occurred far from 
roads and tracks. Eight of the near sites 
showed some evidence of disturbance 
compared to one of the far sites (Table 
1). For one near site, no disturbance 
was found on the transect but a rock 
cairn was observed within 50 m of the 
transect. This distribution of disturbance 
across sites shows that disturbance was 
highly dependent on site category (G = 
12.88, P=0.001). Disturbance consisted 
of rock cairns (Fig. 2), fireplaces (Fig. 3) 
and less commonly a broken or flipped 
over rock. The one instaiice of rock 
disturbance at a far site was a single 
rock (ca 35 x 42 cm in size) that had 
been flipped over to reveal a stencil fi^om 
where it rested originally (Fig. 4). There 
were no other rocks around this site that 
showed evidence of disturbance. 

There was a significant difference 
(t = 2.50, P = 0.021) in the total number of rocks 
counted along near (27.7 ±3.2) versus far (38.6 ± 2.9) 
transects. When only good rocks is considered, there 
was no significant difference (t = 1.16, P = 0.26) in 
the number of rocks counted along near (5.0 ± 1.0) 
versus far (7.3 ± 1.2) transects. 

Due to the time of year when surveys were 
conducted (autumn), only a small number of reptiles 
was observed sheltering under loose rocks. Eulampms 
tenuis was the most common species, being detected at 
1 1 of the sites (4 near, 7 far). Mcphee's skink {Egernia 
mcpheei) and White's skink {Egernia whitii) were 
observed at two sites. Cunningham's skink (Egernia 



Proc. Linn. Soc. N.S.W., 127, 2006 



77 



DISTURBANCE TO ROCK OUTCROPS 



Table 1. Survey site details and reptiles detected under rocks. AMG = Australian Map Grid references 
(Eastings, Northings). Near sites were located <350 m from a walking track or road, while far sites 
were located >500 m from these. Rocks are the number of rocks along a 50 x 20 m transect. Good is the 
number of rocks with traits most suitable for use by reptiles. Reptiles: Et = Eulamprus tenuis', Em = 
Egernia mcpheei; Ew = Egernia whitii', Ec = Egernia cunninghami; Bp = Bassiana platynota. 



Site 


AMG reference 


Distance 
(m) 


Rocks 


Good 


Reptiles 


Types of Disturbance 


1 


0433091 6736800 


Near (50) 


29 


5 


- 


Cairn (of 3 rocks), broken 
rock, displaced rock 


2 


0433091 6736839 


Near (20) 


42 


'3 


- 


Cairn (of 12 rocks) 


3 


0432567 6736726 


Near (200) 


36 


5 


2Et 


Cairn outside transect only 


4 


0433539 6737182 


Near (150) 


27 


3 


3 Em 


None 


5 


0433761 6737538 


Near (50) 


24 


4 


lEt 


Broken rock 


6 


0434263 6737888 


Near (20) 


12 


2 


- 


3 cairns (13, 15,16 rocks) 


7 


0433590 6730564 


Near (100) 


41 


6 


6Et 


Fire place, rock seat, 2 cairns 
(3, 3 rocks) 


8 


0433556 6730491 


Near (300) 


31 


13 


1 Et, 2 Ew 


None 


9 


0430090 6732380 


Near (50) 


16 


3 


- 


Fireplace, broken rock 


10 


0429603 6732331 


Near (200) 


19 


6 


Ew 


Cairn 


11 


0432309 6736252 


Far 


41 


3 


IBp 


None 


12 


0432232 6736539 


Far 


29 


5 


3 Em 


None 


13 


0432268 6736821 


Far 


43 


11 


- 


None 


14 


0432140 6737080 


Far 


46 


1 


lEt 


None 


15 


0432161 6737590 


Far 


41 


10 


- 


None 


16 


0432542 6737299 


Far 


26 


5 


lEt 


1 flipped rock 


17 


0432776 6737429 


Far 


27 


7 


- 


None 


18 


0433024 6737329 


Far 


30 


3 


1 Et, Ec 


None 


19 


0433834 6730303 


Far 


40 


10 


2Et 


None 


20 


0433924 6730049 


Far 


31 


7 


lEt 


None 


21 


0434210 6730149 


Far 


51 


12 


3Et 


None 


22 


0434288 6730664 


Far 


58 


14 


2Et 


None 



cunninghami) was seen in a number of rock crevices 
at various sites but was recorded under loose rocks at 
only one site. There was no difference (t = 0.39, P = 
0.35) in the mean number of lizards per site (near: 1.6 
± 0.6; far: 1.3 ± 0.3) across site categories. 



DISCUSSION 

This study has provided some important 
insights that will extend our understanding of habitat 
disturbance within protected areas. We detected 



78 



Proc. Linn. Soc. N.S.W., 127, 2006 



R.L. GOLDINGAYAND D.A. NEWELL 




Figure 3. Granite rocks used to form a bush campfire. 




Figure 4. A rock that has been flipped over at the far site. Two 
coins (20 cent, one dollar) are present near the stencil for scale 



evidence of rock habitat disturbance at many sites 
and this showed a highly significant association with 
whether sites were near or far fi-om tracks or roads. 
This finding is consistent with what we have observed 
in rock habitats around Sydney (Goldingay 1998; 
Goldingay and Newell 2000; Newell and Goldingay 
in press). There is clearly an influence of distance fi-om 
access points on the likelihood that disturbance will 
occur. This provides Park managers with a clear insight 
for managing rock habitats. That is, areas within 500 
m of existing tracks are likely to be associated with 



disturbance, and development of 
new walking tracks will attract 
habitat disturbance. 

In the present study, most 
disturbance consisted of rock 
cairns and fireplaces that had been 
constructed by hikers. In contrast, 
most of the rock disturbance 
observed in Sydney was caused by 
vandals and reptile poachers, and 
led to severe habitat degradation 
(e.g. rocks were often smashed). 
We found almost no evidence of 
rock disturbance consistent with 
searching for reptiles. The one 
observation of a rock that was 
overturned was quite isolated, 
unlike that in protected areas 
near Sydney where several rocks 
in an area show evidence of 
such disturbance (Goldingay and 
Newell unpubl. data). Therefore, 
we conclude that the overturning 
of this one rock was likely caused 
by a hiker rather than by someone 
searching for reptiles. 

This study provides a 
usefiil baseline for a protected 
area in which reptile poaching is 
currently of low significance. It 
is unknown whether this is due to 
the Park's relative isolation, away 
from a large city, or because it lacks 
an endangered species that might 
be targeted by reptile poachers. 
However, follow-up surveys 
in another 5-years time would 
be a worthwhile management 
consideration to ensure that rock 
habitat disturbance remains at 
a low level. Surveys of similar 
habitat in other Parks in north- 
east NSW should be conducted to 
establish a baseline of data for many fiarther areas. It 
is likely that rock habitat disturbance is widespread, 
though possibly different in intensity to that seen in 
the Sydney basin (see Shine et al. 1998; Newell and 
Goldingay in press). 

Rock cairns were quite common, occurring at 
6 of 10 near sites. Indeed, the walking track to the 
Waratah Trig was marked by >30 small rock cairns 
for most of the distance (see Fig. 5). It is not clear 
whether any of these were recent but it highlights 
an issue that the impacts of this activity are not well 



Proc. Linn. Soc. N.S.W., 127, 2006 



79 



DISTURBANCE TO ROCK OUTCROPS 




Figure 5. Rock cairns marking the 
track on the way to the Waratah 
Trig. 

understood. The plan of management for Gibraltar 
Range and Washpool NPs notes that among several 
objectives, "National Parks are managed to provide 
for sustainable visitor use and enjoyment that is 
compatible with conservation of natural and cultural 
values" (DEC 2005). It is unlikely that the habitat 
disturbance identified in this study is compatible with 
the conservation of natural values based on studies of 
rock habitat disturbance in the Sydney basin (Shine 
et al. 1998; Goldingay and Newell 2000) and the 
thermal requirements of rock-dwelling reptiles (e.g. 
Webb and Shine 1998). Providing education to the 
general public may be needed to mitigate habitat 
impacts. The on-going need for this could be assessed 
by photographic monitoring of a number of near sites 
over several years to assess whether rock cairns and 
rock campfires are continuing to be constructed. Such 
an assessment was used successfully by Goldingay 
and Newell (2000) in Royal National Park in Sydney 
to monitor rock disturbance. This would be consistent 
with the identified need for research into visitor-use 
impacts in these Parks. 

We found significantly fewer rocks on near 
transects compared to far transects. This is consistent 
with the greater fi-equency of disturbance on the near 
transects. This may have little consequence for rock- 
dwelling reptiles because the number of rocks suitable 
for use by reptiles did not differ across site categories. 
The number of reptiles across sites was not different. 
However, the time of the survey was not optimal for 
assessing the number of reptiles that use loose rocks 
and it is likely that some species are more sensitive 
to disturbance than others. Surveys conducted during 



late winter would be more appropriate (see Newell 
and Goldingay in press). It would be worthwhile for a 
detailed study to be conducted so that species that are 
highly dependent on the loose rocks in rock outcrops 
can be identified and their management needs better 
imderstood. 

This study highlights that disturbance to loose 
rock habitats is not confined to areas around Sydney. 
We could generalize fi-om this study that such habitat 
disturbance is a widespread phenomenon regardless 
of where that rock habitat occurs. Goode et al. (2005) 
have revealed that it occurs in many rocky habitats in 
arid areas of the USA. Understanding the ecological 
significance of such habitat disturbance will depend 
on understanding the number of species that are 
dependent on rocky habitats. 



ack:nowledgements 

This paper was improved by the comments of two 
referees. 



REFERENCES 

Brooks, T., da Fonseca, G.A.B. and Rodrigues, A.S.L. 
(2004). Species, data and conservation planning. 
Conservation Biology 18, 1682-8. 

Cogger, H.G. (1992). 'Reptiles and Amphibians of 
Australia'. (Reed Books: Sydney). 

Cogger, H.G., Cameron, E.E., Sadlier, R.A. and Eggler, 
R (1993). 'The Action Plan for Australian Reptiles'. 
(Australian Nature Conservation Agency, Endangered 
Species Program Project Number 124). 

Davey, S. M., Hoare, J.R.L. and Rumba, K.E. (2002). 
Science and its role in Australian regional forest 
agreements. International Forestry Review 4, 39-55. 

DEC (2005). 'Gibraltar Range Group of Parks 

(incorporating Barool, Capoompeta, Gibraltar 
Range, Nymboida and Washpool National Parks and 
Nymboida and Washpool State Conservation Areas), 
Plan of Management'. (Department of Environment 
and Conservation, NSW). 

Goldingay, R.L. (1998). Between a rock and a hard place: 
conserving the broad-headed snake in Australia's 
oldest National Park. Proceedings of the Linnean 
Society of New South Wales 120, 1-10. 

Goldingay, R.L. and Newell, D.A. (2000). Experimental 
rock outcrops reveal continuing habitat disturbance 
for an endangered Australian snake. Conservation 
Biology U, 1908-1912. 

Goode, M.J., Horrace, W.C, Sredl, M.J. and Howland, 
J.M. (2005). Habitat destruction by collectors 
associated with decreased abundance of rock- 
dwelling lizards. Biological Conservation 125, 47-54. 



80 



Proc. Linn. Soc. N.S.W., 127, 2006 



R.L. GOLDINGAYAND D.A. NEWELL 



Hersey, F. (1980). Broad-headed snake Hoplocephalus 
bungaroides. In 'Endangered Animals of New South 
Wales'. (Ed. C. Haigh) pp. 38-40 (National Parks and 
Wildlife Service, Sydney). 

Higgins, J.V., Ricketts, T.H., Parrish, J.D., Dinerstein, E., 
Powell, G., Palminteri, S., Hoekstra, J.M., Morrison, 
J., Tomasek, A. and Adams, J. (2004). Beyond Noah: 
saving species is not enough. Conservation Biology 
18, 1672-3. 

Mahony, S. (1997). Efficacy of the "threatening processes" 
provisions in the Threatened Species Conservation 
Act 1995 (NSW): bush-rock removal and the 
endangered broad-headed snake. Environmental and 
Planning Law Journal 14, 3-16. 

Molnar, J., Marvier, M. and Kareiva, P. (2004). The sum 
is greater than the parts. Conservation Biology 18, 
1670-1. 

Newell, D.A. & Goldingay, R.L. (2004). Conserving 

reptiles and frogs in the forests of New South Wales. 
In 'Conservation of Australia's Forest Fauna'. 2"'' 
edition (Ed. by D. Lurmey) pp. 270-96 (Royal 
Zoological Society of NSW, Sydney). 

Newell, D.A. and Goldingay, R.L. (in press). Distribution 
and habitat assessment of the broad-headed snake 
{Hoplocephalus bungaroides). Australian Zoologist. 

NPWS. (2000a). 'Royal National Park, Heathcote National 
Park and Garawarra State Recreation Area, Plan of 
Management'. (NSW National Parks and Wildlife 
Service:, Hurstville). 

NPWS (2000b). 'Kosciuszko National Park, Plan of 
Management'. (NSW National Parks and Wildlife 
Service, Hurstville). 

NPWS (2003). 'Visitor Guide: Gibraltar Range and 

Washpool National Parks'. (NSW National Parks and 
Wildlife Service, Hurstville). 

Pressey, R.L., Whish, G.L., Barrett, T.W and Watts, 
M.E. (2002). Effectiveness of protected areas in 
north-eastern New South Wales: recent trends in six 
measures. Biological Conservation 106, 57-69. 

Primack, R. B. 1998. 'Essentials of Conservation 
Biology'. (Sinauer Associates:, Sunderland, 
Massachusetts). 

Pringle, R.M., Webb, J.K. and Shine, R. (2003). Canopy 
structure, microclimate, and habitat selection by 
a nocturnal snake, Hoplocephalus bungaroides. 
Ecology 84, 2668-79. 

Schlesinger, C.A. and Shine, R. (1994). Choosing a rock: 
perspectives of a bush-rock collector and a saxicolous 
lizard. Biological Conservation 67, 49-56. 

Shine, R. and Fitzgerald, M. (1989). Conservation 
and reproduction of an endangered species: the 
broad-headed snake, Hoplocephalus bungaroides 
(Elapidae). Australian Zoologist 25, 65-67. 

Shine, R., J. Webb, M. Fitzgerald, and Sumner, J. (1998). 
The impact of bush-rock removal on an endangered 
snake species, Hoplocephalus bungaroides 
(Serpentes: Elapidae). Wildlife Research 25, 285-95. 

Swan, G. (1990). 'A Field Guide to the Snakes and Lizards 
of New South Wales'. (Three Sisters Productions, 
Winmalee). 



Webb, J. K. and Shine, R. (1998). Using thermal ecology 
to predict retreat-site selection by an endangered 
snake species {Hoplocephalus bungaroides: 
Serpentes, Elapidae). Biological Conservation 86, 
233-42. 

Webb, J. K., Brook, B.W. and Shine, R. (2002). Collectors 
endanger Australia's most threatened snake, Broad- 
headed snake Hoplocephalus bungaroides. Biological 
Conservation 81, 21-33. 



Proc. Linn. Soc. N.S.W., 127, 2006 



81 



82 



Amphibians of the Gibraltar Range 

Michael Mahony 
School of Environmental and Life Sciences, Newcastle University, University Drive, Callaghan NSW 2308 



Mahony, M. (2006). Amphibians of Gibraltar Range. Proceedings of the Linnean Society of New South 
Wales. m,S3-9l. 

The Gibraltar Range supports a relatively high diversity of amphibians and thirty frog species, with equal 
numbers of tree frogs (Hylidae) and ground frogs (Myobatrachidae) having been recorded there. It is 
postulated that the geological history of the Great Dividing Range and the rugged landforms on its eastern 
edge, known as the Great Escarpment, provides the underlying explanation for the amphibian diversity 
present. Among the amphibians four major biogeography groups are recognized based on distribution 
and association with major vegetation communities. The largest group consists of 15 species that have 
wide distributions within and beyond the range and occur in several vegetation communities, and only one 
member is categorized as threatened. The second group consists of 12 species and is associated with wet 
forest habitats of the escarpment and coastal belt, with four threatened species. The third group is restricted 
to rainforest habitats and consists of three species of ground frog, two of which are threatened. The final 
group is associated with the drier open forests and grasslands of the tablelands and western slopes and 
consists of four species, three of which are threatened. No frog is endemic to the range, although one ground 
frog, Philoria pughi, is found only in the range and the nearby New England Range and Timbarra Plateau. 
This species and Assa darlingtoni, another ground frog, are closely associated with the warm temperate 
rainforest that is restricted to the higher altitudes on the Gibraltar Range, and their distribution is considered 
to be relictual. Their broader distribution is within isolated montane rainforest that occurs on the higher 
peaks of the Great Escarpment and coastal ranges. Among the frogs of the Gibraltar Range, 11 of the 30 
species are categorized as threatened, eight of which are associated with sfream habitats. This is despite 
the large areas of undisturbed natural habitat present on the range. In contrast species associated with pond 
habitats are less represented in this group. 

Manuscript received 4 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: amphibians, Gibraltar Range, Great Dividing Range, Great Escarpment, Hylidae, mesic 
forests, Myobafrachidae, rainforests. 



INTRODUCTION 

An investigation of the biogeography of the 
amphibian fauna of the Gibraltar Range in northeast 
New South Wales was undertaken to shed light on their 
origins, relationships and the implications of these for 
conservation management. The study of biogeography 
is fundamentally concerned with the documentation 
and interpretation of the distribution of flora and fauna 
and their interrelationship. Uncovering origins and 
dispersal routes of organisms largely depends upon 
the degree of resolution of distributional data and 
robust phylogenetic reconstructions of evolutionary 
relationships (Tyler et al. 1974). 

An understanding of the composition and ancestry 
of the amphibian faima of the Gibraltar Range is 
underpiimed by interpretations of the geological 
history of the landforms of the range and its climate. 



The Gibraltar Range occurs on the eastern boimdary 
of the Great Dividing Range. The Great Dividing 
Range is the dominant landform feature of the east 
coast of Australia, and running along its eastern edge 
is the Great Escarpment. Oilier (1982) postulated 
that the escarpment originated by scarp retreat from 
a new continental edge of eastern Australia about 80 
million years ago. From a biogeographic perspective 
this results in two principle axes; the first is in the 
north-south direction of the Great Divide and the 
associated Great Escarpment that extends in the order 
of hundreds of kilometres, the second is in the east- 
west direction from coast to tablelands that extends in 
the tens of kilometres. 

The geologic history of the Gibraltar Range and its 
landforms are significant factors in our understanding 
of the composition and ancestry of the amphibian 
fauna. The higher mountains on the Great Escarpment 



AMPHIBIANS OF THE GIBRALTAR RANGE 



experience a climate of moderate temperatures and 
high rainfall and support mesic forest vegetation 
communities at mid to high altitudes. These forests 
contain ancestral elements of wet forest communities 
that were once more widespread, particularly along the 
Great Divide, and have contracted as the Australian 
climate has dried (Nix 1991). 

In the north-south axis the higher mountains 
along the Great Divide provide refiiges for the flora 
and fauna adapted to these mesic habitats and provide 
a view of their evolutionary history. The frog species 
found in the mesic forest habitats are postulated to 
reflect a long evolutionary relationship between 
the flora and fauna. In the north-south axis the 
Gibraltar Range is one of many ranges that form the 
relatively continuous Great Divide. While it may be 
relatively continuous as a major landform feature it 
has considerable variation in altitude and ruggedness 
along its considerable length. The Gibraltar Range 
is one of the higher ranges along the length of the 
Great Divide with its highest peaks being above 1400 
metres in altitude, and along with a rugged topography 
and complex underlying geology (Barnes et al. 1995) 
result in a complex mosaic distribution of rainforests 
and wet sclerophyll forests. 

In the east-west axis the Gibraltar Range stands at 
the junction of two major geomorphic provinces, the 
tablelands to the west and the coastal belt to the east. 
In this axis the formation of the Great Escarpment and 
the mountain ranges, river valleys and coastal plains 
associated with it, provide a diverse topography from 
low to high altitude. On its western side the Gibraltar 
Range has upland areas of low relief with gently 
flowing streams and tableland swamps. On the east 
is a steep escarpment with rapidly flowing sfreams 
and deep gorges, and to the northeast is an area of 
moderate to high relief with rapid flowing streams. 
To the south and east the scarp is clearly defined by 
the Marm River and its smaller tributaries, while to 
the northeast the range is almost cut off by the Rocky 
(Timbarra) River to form an isolated plateau. This 
river runs in a northerly direction along the line of 
Demon Fault that separates the Gibraltar Range from 
the tablelands to the west. 

While the Gondwanan origin and relationship of 
Australia's two major frog families, the Hylidae (tree 
frogs) and Myobatrachidae (ground frogs) is well 
accepted, the geographic context of the evolution 
and diversification of the Australian frog fauna 
remains a matter of considerable debate (Tyler 1979; 
Roberts 1998). Two major features of their evolution 
can be investigated by studies of the fauna of the 
Great Divide. The first is evidence of the ancestral 
composition of the amphibian fauna of the mesic 



forests that have been associated with the Great 
Divide for tens of millions of years, and the second 
is the extent of diversification that has occurred as the 
Great Divide has been eroded away and as climate 
has changed. 

The objective of this paper is to provide an 
overview of the diversity of frogs in the major 
vegetation communities of the Gibraltar Range along 
with an interpretation of the composition and ancestry 
of the amphibian fauna. Where appropriate, details of 
habitat use and conservation status will be discussed 
along with the implications for management. 



MATERIALS AND METHODS 

To compile a list of the amphibian species 
of the Gibraltar Range a number of sources were 
consulted. A primary species list was assembled by 
consulting the records of the Australian Museum, 
Queensland Museum, Victorian Museum and South 
Australian Museum. To these were added the records 
in the Wildlife Atlas of New South Wales (NSW 
DEC, NPWS, accessed April 2005). A selected 
literature search was conducted that included large 
and comprehensive surveys such as the North East 
Forest Biodiversity Survey (NSW NPWS 1994) and 
Fauna Surveys for Forestry Enviroimiental Impact 
Statements (Smith et al. 1994; State Forest NSW, 
1995). Lastly, records from targeted surveys for 
taxonomic studies and from a long-term monitoring 
site in Washpool National Park were included 
(Knowles et al. 2004; Donnellan et al. 2002, 2004; 
Mahony unpubl. data). In addition, information on 
the vegetation communities and habitats occupied by 
each species was collated. 

Based on distribution records and association 
with major vegetation communities frog species were 
assigned to one of four categories, 1) widespread 
occurrence across the region in all major vegetation 
communities, 2) eucalypt-dominated forest 
communities of the escarpment and coast belt, 3) 
rainforest specialists, and 4) woodlands, dry forests 
and grasslands of the tablelands. Within these 
categories the frogs were subdivided on the basis of 
primary breeding habitat. 

Conservation status of species was based on 
listings in the New South Wales Threatened Species 
Conservation Act 1995 (NSW TSC Act 1995) 
supported by a recent assessment of Australian 
amphibians that applied the International Union 
for the Conservation of Nature (lUCN) categories 
(Global Amphibian Assessment 2004). 



84 



Proc. Linn. Soc. N.S.W., 127, 2006 



M. MAHONY 



RESULTS 

A total of 30 frog species has been recorded from 
the Gibraltar Range and a ftirther four are considered 
likely to occur there (Table 1). Equal numbers of tree 
frogs (Hylidae) and ground frogs (Myobatrachidae) 
are found. All hylids present are members of the 
genus Litoria, while there are eight genera of 
myobafrachids. Despite the relatively high species 
diversity many species are represented by a small 
number of location records, and for several species 
by a single location record. As a result the regional 
distribution, abundance and habitat associations 
of many species are incomplete. Fauna surveys 
conducted for the North East Forest Biodiversity 
Study (NSW NPWS 1994) and Forestry EISs (SF 
NSW 1995) provide the most detailed picture of the 
distribution and abundance of species. Discoveries 
made during surveys in the last two decades indicate 
that significant disfributional records, and even new 
species, may be found there (Dormellan et al. 2002; 
Knowles et al. 2004). 

Frogs with a widespread distribution 

Division of the frog fauna into major distribution 
patterns and broad vegetation community associations 
reveals that the largest group numerically is species 
that have a widespread distribution and that occur 
in several vegetation communities. Fifteen species 
are placed in this group, seven tree frogs and eight 
ground frogs (Table 1). Most of these frogs have 
extensive distributions in south-eastern Ausfralia 
(see distribution maps in Cogger 2002 and Robinson 
2002). This is not to say that they are necessarily 
habitat generalists. Subdivision of these species by 
preferred breeding habitat shows that the majority, 
14 of the 15, make use of ponds or swamps, four 
use both ponds and streams and could be considered 
to be generalists in respect of breeding habitat, and 
only one is a stream specialist. Two species that use 
ponds show a preference for ephemeral ponds and a 
third {Crinia signifera) makes use of a wide range 
of water bodies from small ephemeral pools to large 
swamps, indeed the only habitat it is not found in is 
fast flowing streams. This species occurs in disturbed 
sites and therefore is common where human activity 
opens or modifies habitats. 

Only one species in this group is categorized as 
threatened. The New England Tableland population of 
Adelotus brevis is listed as an "endangered population" 
under the NSW TSC Act. No recent records of this 
species were found at high altitudes in the Gibraltar 
Range, but several populations are known from lower 



altitude in Washpool National Park (NP) and Ewingar 
State Forest (SF). 

Frogs of forest communities of tlie escarpment 
and coastal belt 

The second largest group is associated with 
eucalypt-dominated forest vegetation communities 
of the escarpment and coastal belt. Twelve species 
are placed in this group, eight tree frogs and four 
ground frogs (Table 1). As might be expected due 
to the rugged topography of the escarpment the 
majority of these species are associated with stream 
habitats. Six species, three tree frogs and three 
ground frogs, breed in streams and have tadpoles 
adapted to sfream habitats. Among these, three 
species, Litoria subglandulosa, L. piperata and 
Mixophyes balbus(Fig. lb), are resfricted to higher 
altitudes on the escarpment. Vegetation community 
is not the major factor determining their distribution; 
they occur in sfreams within heath, dry forest, wet 
forest and rainforest communities. One stream frog, 
Mixophyes iteratus (Fig. Ic), is found only at low 
to moderate altitudes and is always associated with 
rainforest or wet forest habitats. The remaining two 
species {L. barringtonensis and M. fasciolatus) occur 
across the range of altitudes but always in wet forest 
communities. 

Five species in this group breed in ponds and 
swamps, and two of these often breed in ephemeral 
situations (Table 1). For three of the species included 
in this group (Z,. brevipalmata, L. revelata and L. 
tyleri) there are no confirmed records in the Gibraltar 
Range. They are included here because they occur in 
wet forest habitats to the north, south and east of the 
Gibraltar Range and it is considered possible that they 
occur in the range. If these species do occur they are 
not abimdant because they have not been detected in 
systematic surveys (NSW NPWS 1994; Smith et al. 
1994; State Forests NSW 1995) or targeted searches 
(Mahony unpubl.). Litoria brevipalmata is often 
difficult to detect in field surveys because adults are 
active at breeding sites on only one or two evenings 
of the year. Records of Z,. revelata may be absent for 
a different reason. This species was overlooked in 
the past and until recently it was not distinguished 
from Litoria verreauxi, a close relative. Field guides 
do not indicate that L. revelata is found south of the 
Border Ranges region, which are approximately 120 
kilometres to the north-east of the Gibraltar Ranges, 
yet recent field studies (Price 2004) indicate that it 
occurs in a series of apparently isolated populations 
along the escarpment and coastal ranges as far south 
as the Sydney Basin. Targeted searches for this species 
have been conducted in the Washpool National Park 



Proc. Linn. Soc. N.S.W., 127, 2006 



85 



AMPHIBIANS OF THE GIBRALTAR RANGE 



^^^^^9 


^^^^^^^^^^_ 












Association with major vegetation community 




lypical breeding 


location 






Pond 










Conser- 


Stream 


and/or 


Ephemeral 


Terrestrial 


Species scientific and common name 


vation 
status 


breeding 
(lentic) 


swamp 

breeding 

Gotic) 


pool 
breeding 


eggs and 

embryonic 

stage. 


Widespread occurring 


J^H^^HHIF 












in many vegetation 


^^^^^^^B 












communities from 


^^^^^V 












rainforest to grassland 














Litoria caerulea 


Green Tree Frog 






X 






L. dentata 


Bleating Tree Frog 






X 


X 




L. fallax 


Dwarf Tree Frog 






X 






L. latopalmata 


Broad-palmed Frog 






X 






L. peronii 


Peron's Tree Frog 






X 






L. verreauxi 


Whistling Tree Frog 






X 






L. wilcoxi 


Rocky River Frog 




X 








Adelotus brevis 


Tusked Frog 


EP 


X 


X 






Crinia signifera 


Eastern Froglet 




X 


X 


X 




Limnodynastes dumerillii 


Banjo Frog 




X 


X 






L. ornatus 


Ornate Burrowing Frog 




X 


X 


X 




L. peronii 


Striped Marsh Frog 






X 






L. tasmaniensis 


Spotted Grass Frog 






X 


X 




Uperoleia fusca 


Dusky Toadlet 






X 






U. laevigata 


Smooth Toadlet 




^^B 


X 




^■B 


Escarpment and coastal 














belt. Wet forests excluding 














rainforest specialists 














Litoria barringtonensis 


Barrington Tree Frog 




X 








L. brevipalmata* 


Green-thighed Frog 


V 




X 


X 




L. chloris 


Red-eyed Frog 






X 


X 




L. gracilenta 


Dainty Tree Frog 






X 






L. piperata 


Peppered Frog 


E 


X 








L. revelata* 


Revealed Frog 






X 






L. subglandulosa 


Glandular Tree Frog 


V 


X 








L. tyleri* 


Tyler's Tree Frog 






X 






Mixophyes balbus 


Stuttering Frog 


E 


X 








M. fasciolatus 


Great Barred Frog 




X 








M. iteratus 


Giant Barred River Frog 


E 


X 








Pseudophryne coriacae 


Red-backed Toadlet 










X 


Restricted to high altitude. 














Rainforest specialists 














Assa darlingtoni 


Hip-pocket Frog 


V 








X 


Lechriodus fletecheri 


Sandpaper Frog 








X 




Philoria pughi 


Mountain Mist Frog 


V 








X 


Tablelands and western 














species. Woodlands, dry 














forest and grasslands 














Litoria booroolongensis 


Booroolong Frog 


E 


X 


X 






L. castanea* 


Yellow-spotted Bell Frog 


E 


X 








Crinia parinsignifera 


Beeping Froglet 






X 






Pseudophryne bibroni 


Brown Toadlet 


EP 




X 




X 


Total 






13 


21 


6 


4 



86 





Proc. Linn. Soc. N.S.W., 127, 2006 



M. MAHONY 



without success. The final species, L. tyleri, is readily 
distinguished and the lack of records may indicate 
that it does not occur in the Gibraltar Range. 

Five species in this group are classed as 
threatened, and each of these breed in stream habitats. 
In contrast, no pond-breeding species in this grouping 
is threatened. One of the two stream-breeding species 
that is not threatened, Mixophyes fasciolatus, breeds 
in both ponds and streams. 

Frogs that are found only in rainforest habitats 

The group with the narrowest distribution is the 
rainforest specialists, with only three species, all of 
which are ground frogs. The absence of tree fi-ogs 
from this group is not unexpected, given that there is 
no tree fi"og that is restricted to rainforest vegetation 
communities of the Great Escarpment in NSW and 
south-east Queensland. It is not until the rainforests 
of far north Queensland that we encounter tree fi^ogs 
that are restricted to rainforest habitats. 

Two of the ground frogs, Assa darlingtoni (Fig. 
la) and Philoria pughi (Fig. Id) reflect refugial 
distributions. They occur only at higher altitudes in 
warm temperate rainforest or deeper gullies with 
subtropical rainforest. In the Gibraltar Range the 
distribution of Assa is limited to a relatively small 
area of high altitude warm temperate rainforest 
(above 1000 m) and Philoria pughi has a slightly 
wider distribution in warm temperate and subtropical 
rainforest from mid to high altitudes (800 to 1000 m). 
These vegetation communities are relicts of former 
more widespread vegetation communities. They attest 
to a past when the climate was wetter and milder and 
when their distribution was more continuous along 
the great escarpment. The last member of this group, 
Lechriodus fletcheri, is found in rainforest from 
low to high altitude and thus its distribution is more 
extensive. 



Table 1. LEFT 

Major habitat association, breeding location and 
conservation status of the frogs of the Gibraltar 
Range. Conservation status is based on lUCN cat- 
egories (Stuart et al. 2004). For a small number 
of species there are no records for the Gibral- 
tar Range; they are included because popula- 
tions are known in forested habitats to the north, 
south and east and it is likely that they occur in 
the Gibraltar Range. They are identified by an 
asterisk. An ephemeral water body is defined as 
a non-perennial; it can be a pool that lasts for a 
matter of days or weeks or up to several months. 



Frogs of the woodlands, dry forests and grasslands 
of the tablelands 

Another relatively small group are the frogs that 
are associated with the vegetation communities of 
the tablelands and western slopes. Four species, two 
tree frogs (Litoria booroolongensis and L. castanea) 
and two ground frogs {Crinia parinsignifera and 
Pseudophyrne bibroni) are placed in this group 
(Table 1). The group may be even smaller because 
there is no direct evidence that the two tree frogs 
{Litoria booroolongensis and L. castanea) occur in 
the Gibraltar Range. They are included here because 
of proximity of records on the tablelands and the 
presence of suitable habitat in the range. Both species 
have disappeared from the New England Tableland 
(Hines et. al. 1999; Mahony 1999) and it may be 
that we will never know whether they occurred on 
the Gibraltar Range. Litoria boorolongensis had an 
extensive distribution on the New England Tableland 
and on the western slopes south to the Australian Alps, 
and suitable habitat in the Gibraltar Range occurs 
along the upper reaches of the Mann River and Rocky 
(Timbarra) River. Litoria castanea had a far narrower 
distribution that was centred on tableland habitats. Its 
preferred habitat was tableland swamps and lagoons 
and the upper altitudes in the southern areas of the 
Gibraltar Range contain significant tableland swamps 
in undisturbed condition. 

Of the two ground frogs in this group, one, P. 
bibroni, has also disappeared from the tablelands 
(Mahony unpubl. data). There are no records of this 
species from the Gibraltar Range, but once again 
its was formerly widespread across the tablelands 
(Heatwole et al. 1995). The remaining species in this 
group, C. parinsignifera, is common and widespread 
being found in ponds and swamps in open vegetation 
communities, and is often associated with disturbed 
areas. The limit of the distribution of the two tree 
fi"Ogs (L. booroolongensis and L. castanea) is at the 
upper or western edge of the escarpment on the other 
hand the two ground frogs are also distributed to the 
east on the coastal plain, but they are not found in the 
wet forest habitats of the escarpment. Pseudophyrne 
bibroni is replaced by a congener P. coriacae in the 
wet forests of the escarpment, and C. parinsignifera 
shows a preference for open habitats. 

Each member of this group has a distinct 
breeding biology and behaviour and there is no 
apparent link between these features and those that 
have disappeared. Litoria booroolongensis breeds in 
flowing streams, L. castanea in swamps and pools, 
sometimes in large still pools on streams, and P. 
bibroni lays its eggs in terrestrial sites near swamps 
and pools. 



Proc. Linn. Soc. N.S.W., 127, 2006 



87 



AMPHIBIANS OF THE GIBRALTAR RANGE 



a 


^\ ■"■■■ „ 


- . - V' 


^ 


*^*'- 


(■ ' ";~ 


i.-fS^^-^i^': 


i .-^ _ 








^%^ 






■< 


. . .11 


h^, " 




^^^ 








Figure 1. a) Adult male Assa darlingtoni surrounded by hatching embryos prior to their entering into 
his lateral pouches where they will undergo the tadpole stage of their life cycle. This terrestrial frog 
is found only in warm temperate rainforests at high altitude in the Gibraltar Range, b) Adult male 
Mixophyes balbus, an endangered stream-breeding species that occurs in high altitude streams of the 
Gibraltar Range, c) A pair of Mixophyes iteratus in embrace prior to egg deposition. This vulnerable 
species is found in stream habitats at low altitude in the Gibraltar Range, d) A male Philoria pughi 
within its terrestrial nest chamber that has been exposed by lifting away a covering of leaves. A clutch 
of embryos in early stages of development and still within their egg capsules can be seen beneath 
the male. After the embryos hatch, the tadpoles remain in the nest and leave after metamorphosis. 



No frog species is endemic to the Gibraltar 
Range. Philoria pughi has the narrowest distribution, 
it is known only from the Gibraltar Range, and the 
New England Range and Timbarra Plateau to the 
north. Two others, Assa darlingtoni and Lechriodus 
fletcheri, occupy refiigial mesic forest habitats, and 
their populations in the Gibraltar Range are isolated 
from other restricted populations along the Great 
Escarpment. 



DISCUSSION 

The high diversity of amphibians found in the 
Gibraltar Range can be explained by a combination of 
factors; the antiquity of the Great Dividing Range, the 
abrupt change in altitude and the rugged landscape 
of the Great Escarpment, and the consequent climate 
differences. The range stands at the junction of two 



ancient geomorphic regions, the tablelands to the 
west and the coastal plain to the east, and provides 
habitats for species that have evolved in these 
regions. These differences are reflected in the aquatic 
habitats that are present, from tableland swamps with 
slow flowing streams to fast flowing streams on the 
escarpment. The rugged topography of the region, its 
altitudinal range and climate result in the presence of 
several major vegetation communities. 

All of the frogs found on the Gibraltar Range 
belong to two families that have a long evolutionary 
relationship with the Ausfralian continent; the tree 
frogs of the family Hylidae and the ground frogs 
of the family Myobafrachidae. These families are 
recognized as being of Gondwana origin (Tyler 
1979); they are old endemics. Molecular genetic 
evidence indicates that the ancesfral tree and ground 
frogs were already well differentiated at the time 
Australia separated from Antarctica some 52 million 



88 



Proc. Linn. Soc. N.S.W., 127, 2006 



M. MAHONY 



years ago (Daugherty and Maxson 1982; Hutchinson 
and Maxson 1988). Apart from the introduced cane 
toad (family Bufonidae) Australia has members of 
two other families of frog, the Michrohylidae and 
Ranidae. Members of these families are considered 
to have arrived in Australia in more recent geological 
time, when the Australian continent came into closer 
contact with south-east Asia (Tyler 1979), and their 
members are found only in rainforest habitats in north 
Queensland and the Northern Territory. 

Several genera and species groups that have 
a long association with the mesic forest habitats of 
the Great Divide and escarpment can be identified 
in the Gibraltar Range. Two examples are briefly 
considered, one from each of the major families, to 
illustrate this point. The five species of Mixophyes 
are found only in wet forest habitats along the 
Great Divide and escarpment from east Gippsland 
in Victoria to the Atherton Tablelands in far north 
Queensland, with a fiirther species found in montane 
rainforest in Papua New Guinea (Dormellan et al. 
1990). Phylogenetic studies (Heyer and Leim 1996; 
Kluge and Farris 1976) place this genus in a basal 
position among the myobafrachids and their current 
distribution and habitat preferences sfrongly suggest 
they have had a long association with the wet forests 
of the Great Divide and escarpment. Among the free 
frogs members of the Litoria citropa species group 
(Tyler and Davies 1978) are closely associated with 
the wet forests of the Great Divide and escarpment 
from southern Victoria to mid east Queensland 
(Donnellan et al. 1999; Mahony et al. 2000). 

A detailed account of the frogs of the New 
England Tablelands region, an area about nine 
times larger in extent than the Gibraltar Range, was 
presented by Heatwole et al. (1995). The Gibraltar 
Range is adjacent to the north-east of this region 
and the western portion of the range was included in 
their investigation. They reported 46 species in the 
New England region and concluded that the largest 
numbers were associated with moist habitats that are 
disfributed along the east coast and onto the Great 
Dividing Range. They did not have extensive data 
from the Gibraltar Range region and inspection of 
their data reveals that most of their records were 
from along the Gwydir Highway, which cuts east- 
west across the range, and a small number of sites in 
the Gibraltar Range National Park. Nonetheless, the 
current study provides sfrong support for their major 
conclusion. The 30 species present in the Gibraltar 
Range account for 65% of the total number they 
reported for the larger region. It is evident that the 
mesic habitats of the Great Escarpment and coastal 
belt provide a diversity of habitats and this is reflected 



in the number of amphibians present. 

The significance of the geomorphic processes 
that have shaped the Great Escarpment in relation to 
the evolution of its terrestrial fauna is evident in the 
Gibraltar Range. With respect to the north-south axis 
the Gibraltar Range is an isolated area of uplands. 
Scarp retreat created firstly steep gorges and then 
wider valleys, as these valleys widened and their 
headwaters refreated further west the higher altitude 
ranges of the Great Divide and their fauna and flora 
were isolated (Oilier 1982). It is postulated that 
dispersal was limited where large valleys with drier 
vegetation communities dissected the ranges. For 
example, in the Gibraltar Range isolated populations 
of a small number of rainforest frogs are found at 
higher altitudes {Assa darlingtoni, Philoriapughi, and 
Lechriodusfletcheri) in mesic rainforest communities. 
In addition to the isolation resulting from landscape 
barriers are the barriers that were created as climate 
changed. In the past the climate was warmer and 
wetter and the mesic vegetation more widespread on 
the Great Divide (Nix 1 99 1 ), providing an opportunity 
for species adapted to the mesic forest habitats to 
disperse. From the perspective of the amphibian 
fauna the period or extent of isolation of the Gibraltar 
Range has not been extensive because only one frog, 
Philoria pughi can be described as endemic to the 
Gibraltar and the nearby New England Ranges. 

From the perspective of the evolution of its 
amphibian fauna it is perhaps more appropriate to 
view the Gibraltar Range as part of a larger unit of the 
eastern escarpment of the New England Tableland, 
which extends from the Macleay River incursion 
in the south to the Clarence River incursion in the 
north. Two species associated with the fast-flowing 
streams of the upper escarpment, L. piperata 
and L. subglandulosa, are found only within this 
region. Litoria daviesae, a sibling species of L. 
subglandulosa, occurs to the south of the Macleay 
River catchment, and L. pearsoniana, a sibling of L. 
piperata, occurs in the mesic forests on the northern 
side of the Clarence catchment. Among the ground 
frogs, M balbus reaches the extent of its disfribution 
at the northern incursion of the Clarence River, and 
to the north a sibling species, M. fleayi, occurs in 
mesic forest habitats. A similar pattern occurs within 
Philoria, to the north of the incursion of the Clarence 
River P. pughi is replaced by P. kundagungan, and to 
the northeast by P. loveridgei and P. richmondensis 
(Rnowles et al. 2004). This genus more than any other 
is indicative of the isolation of mesic forest habitats in 
north-eastern New South Wales in the past 15 million 
years (Knowles et al. 2004). 



Proc. Linn. Soc. N.S.W., 127, 2006 



89 



AMPHIBIANS OF THE GIBRALTAR RANGE 



Despite the protection of large portions of 
the Gibraltar Range in conservation reserves a 
considerable number of the frogs found there are 
classified as threatened. Nine of the 30 species are 
categorized as either endangered or vulnerable. In 
the case of those species found in isolated rainforest 
remnants the categorization is related to small 
population size and limited distribution, and the 
potential factors threatening their short-term survival 
are associated with habitat loss, changes in hydrology 
and pollution. In the long-term their evolutionary 
potential may be impacted by climate change. A 
similar explanation is not possible for those threatened 
species that are found in vegetation communities that 
are more widespread or those not limited to specific 
vegetation communities. 

Most threatened are frogs that breed in streams 
and are associated with stream habitats, they include 
L. piperata, L. subglandulosa, M. balbus and M 
iteratus. There is extensive habitat for these species 
in the Gibraltar Range and in the wider region. It is 
difficult to argue that declines in abundance and the 
disappearance of their populations are due primarily 
to habitat loss or degradation. Undoubtedly, habitat 
modification, particularly on the tablelands where 
there is a long history of agricultural activity may 
have impacted on species such as L. booroolongensis, 
but this explanation is not tenable across the wider 
distributions of these species. It is most likely that 
the cause of declines is due to the impact of an 
invasive pathogenic fungus that causes the disease 
chytridiomycosis in fi-ogs (Berger et al. 1998, 2004). 
High altitude stream fi-ogs are known to be most 
susceptible to this disease (Berger et al. 2004) and the 
threat to their long-term persistence remains in the 
balance. 

One species of conservation significance, the 
peppered fi-og (Z,. piperata), deserves more detailed 
consideration. This fi"og was described in 1985 from 
a small number of high altitude locations distributed 
on the edge of the Great Escarpment in the New 
England region, extending from the Oxley River 
Gorge (Gara River) in the south to several sites on the 
headwaters of the Clarence River in the north (Mann, 
Oban, Henry and Sara Rivers; Tyler and Davies 
1985). All specimens, with the exception of two 
collected at the Gara River in 1952, were collected 
in the early 1970s. Several specimens were obtained 
fi-om Diehard Creek, which drains south-west from 
the Gibraltar Range to the Mann River. Conservation 
assessments of the peppered fi-og have been fi-aught 
with difficulty. No specimens of this or other members 
of its species group {Litoria citropa species group, 
Tyler and Davies 1978; Donnellan et al. 1999) were 



detected during intensive searches conducted in the 
1990s at any of the locations named in the species' 
description (NSW NPWS 1994; SF NSW 1995). 
Searches were extended to likely habitats within the 
region and small "peppered" tree fi-ogs were foimd at 
Rockadooie and Seven Mile Creeks in the catchment 
of the Rocky (Timbarra) River in the north-west 
region of the Gibraltar Range, and at Cooraldooral 
Creek, a catchment of the Mann River, in the south- 
west region in Gibraltar Range. Other populations 
were detected on the Timbarra Plateau (Nelsons 
Creek) to the north of the Gibraltar Range. 

Genetic comparisons of the "peppered" frogs 
fi-om each of these sites with a larger collection 
of specimens of members of the Litoria citropa 
species group fi-om across the Great Escarpment 
and coastal belt placed these specimens within the 
species recognized as the Barrington Tree Frog 
{Litoria barringtonesis; Donnellan et al. 1999). Such 
a result would normally lead to a questioning of the 
taxonomic status of the Peppered Frog. However, 
because no specimens could be collected from any of 
the historical sites listed in the species' description, 
and suitable genetic material could not be extracted 
fi-om the fixed museum specimens to be included in 
appropriate genetic comparisons, the question remains 
open. Furthermore, the type series of L. piperata, 
which consists of over 70 specimens, has been closely 
examined, and there is general agreement among 
herpetologists that L. piperata is distinctly different 
from L. barringtonensis . 

The Peppered Frog is listed as endangered and a 
Recovery Plan has been prepared (NS WNPWS 200 1 ). 
If we accept the position that it is morphologically 
distinct, then there is no evidence of an extant 
population and the species should be considered 
as presumed extinct. Whatever the situation, the 
Gibraltar Range provides important high altitude 
plateau and escarpment streams considered to be the 
habitat of this frog. 



ACKNOWLEDGEMENTS 

I am most grateful for the assistance of numerous colleagues 
during fieldwork, in particular Steve Donnellan, Ross 
Knowles, Andrew Stauber, Karen Thumm and Stephen 
Mahony. Long-term monitoring of stream frogs was 
supported by a grant from Earthwatch and many volunteers 
assisted with the fieldwork. 



90 



Proc. Linn. Soc. N.S.W., 127, 2006 



M. MAHONY 



REFERENCES 

Anstis, M. (2002). 'Tadpoles of south-eastern Australia: a 
guide with keys'. (Reed New Holland, Sydney). 

Berger, L., Speare, R., Daszak, P., Green, D.E., 

Cunningham, A.A., Goggin, C. L., Slocombe, R., 
Ragan, M. A., Hyatt, A.D., McDonald, K. R., Hines, 
H. B., Lips, K.R., Marrantelli G. and Parkes, H. 
(1998). Chytridiomycosis causes amphibian mortality 
associated with population declines in the rainforest 
of Australia and Central America. Proceedings of the 
National Academy of Science USA 95, 9031-9036. 

Berger, L., Speare, R., Hines, H.B., Marantelli, G., Hyatt, 
A.D., McDonald, K.R., Sherratt, L.F., Olsen, V, 
Clarke, J.M., Gillespie, G., Mahony, M.J., Sheppard, 
N., Williams, C. and Tyler, M.J. (2004). Effect of 
season and temperature on mortality in amphibians 
due to chytridiomycosis. Australian Veterinary 
Journal 82(7), 434-439. 

Cogger, H.G. (2000). 'Reptiles and Amphibians of 
Australia'. (6* ed. Reed, Sydney). 

Daugherty C.H. and Maxson L.R. (1982). A biochemical 
assessment of the evolution of Myobatrachine frogs. 
Herpetological 3Si3), 341-348. 

Donnellan, S.C, Mahony, M.J. and Davies, M. (1990). A 
new species of Mixophyes (Anura: Leptodactylidae) 
and first record of the genus in New Guinea. 
Herpetologica 46(3), 266-274. 

Doimellan, S.C, McGuigan, K., Knowles, R., Mahony, 
M.J. and Moritz, C. (1999). Genetic evidence for 
species boundaries in fi-ogs of the Litoria citropa 
species group (Anura: Hylidae). Australian Journal 
of Zoology 47, 275-293. 

Heatwole, H., de Bavay, J., Webber, R and Webb, G. 
(1995). Faunal survey of New England. IV. The 
Frogs. Memoirs of the Queensland Museum 38(1), 
229-249. 

Heyer, W.R. and Leim, D.S. (1996). Analysis of 
intergeneric relationships of the Australian frog 
family Myobatrachidae. Smithsonian Contributions 
to Zoo/ogy No. 233, 1-29. 

Knowles, R., Mahony, M.J., Armstrong, J. and Doimellan, 
S. (2004). Systematics of sphagnum fi-ogs of the 
genus Philoria (Anura: Myobatrachidae) in eastern 
Australia. Records of the Australian Museum 56, 57- 
74. 

Mahony, M. J. and Knowles, R. (1994). 'Ataxonomic 
review of selected frogs of north-east NSW forests'. 
In North East Forests Biodiversity Report No. 3g. 
New South Wales National Parks and Wildlife 
Service. 

Mahony, M.J., Knowles, R., Foster, R. and Doimellan. S. 
(2000). Systematics of the Litoria citropa (Anura: 
Hylidae) complex in northern New South Wales 
and southern Queensland. Records of the Australian 
Museum. 53, 37-48. 

McGuigan, K., McDonald, K.M., Parris, K. and Moritz, C. 
(1998). Mitochondrial DNA diversity and historical 
biogeography of a wet forest-restricted fi-og {Litoria 



pearsoniana) from mid-east Australia. Molecular 
Ecology 7,175-1^6. 

New South Wales National Parks and Wildlife Service. 
(1994). Fauna of north-east NSW forests. North East 
Forests Biodiversity Report No. 3. New South Wales 
National Parks and Wildlife Service. 

New South Wales National Parks and Wildlife Service. 
(2001). Yellow-spotted Bell Frog (Litoria castanea) 
and Peppered Tree Frog (Litoria piperata) Recovery 
Plan. (NPWS Hurstville, NSW). 

Nix, H.A. (1991). Biogeography: patterns and processes. 
In: 'Rainforest Animals. Atlas of Vertebrates Endemic 
to Australia's Wet Tropics (Kowari 1)' (Eds H.A. Nix 
and M.A. Switzer) pp. 10-39. (Australian National 
Parks and Wildlife Service, Canberra). 

Oilier, CD. (1982). The Great Escarpment of eastern 
Australia: tectonic and geomorphic significance. 
Journal of the Geological Society of Australian 29, 
13-23. 

Price, L. (2004) Honours Thesis, University of Newcastle, 
NSW Australia. 

Quilty, P.A. (1994). The background: 144 million years of 
Australian palaeoclimatic and palaeogeography. In 
'History of The Australian Vegetation: Cretaceous 
to Recent' (Ed. R.S. Hill) pp. 14-43. (Cambridge 
University Press, Cambridge). 

Roberts, J.D. and Watson G.F. (1993). Biogeography and 
phylogeny of the Anura. In 'The Fauna of Australia 
Vol. 2a Amphibia and Reptilia'. (Eds C.J. Glasby, 
G.J.B Ross and PL. Beesley) pp. 35-40. (Australian 
Government Printing Service, Canberra). 

Robinson, M. (1993). 'Afield guide to Frogs of Australia. 
From Port Augusta to Eraser Island including 
Tasmania'. (Reed, Sydney). 

Smith, A.R, Andrews, S.P and Moore, D.M. (1994). 
Terrestrial fauna of the Grafton and Casino State 
Forest Management Areas. Description and 
Assessment of Forestry Impacts. Grafton-Casino 
Management Area EIS Supporting document Number 
1. State Forest of New South Wales, Northern 
Regions, Coffs Harbour. 

State Forest of New South Wales. (1995). Proposed 
forestry Operations in the Tenterfield Management 
Area. Vol. D. Environmental Impact Statement. 
Fauna Impact Statement. 

Tyler M.J. (1979). Herpetofaunal relationships of South 
America with Australia. In 'The South American 
herpetofauna: Its origin, evolution, and dispersal'. 
(Ed. W.E. Duellman) pp. 73-106. University of 
Kansas Museum Natural History Monogrograh 7, 1- 
485. 

Tyler, M.J. and Davies, M. (1978). Species groups within 
the Australo-Papuan Hylid genus Litoria Tschudi. 
Australian Journal of Zoology, Supplementary Series 
63,1-47. 

Tyler, M.J. and Davies, M. (1985). Anew species of 
Litoria (Anura: Hylidae) from New South Wales, 

Australia. Copeia 1985(1), 11 45- 11 49. 



Proc. Linn. Soc. N.S.W., 127, 2006 



91 



92 



Species Richness and Habitat Associations of Non-flying 
Mammals in Gibraltar Range National Park 

Karl Vernes, Stuart Green, Alison Howes and Linda Dunn 

School of Environmental Sciences and Natural Resources Management, Ecosystem Management, The 

University of New England, Armidale, NSW 235 1 . 



Vernes, K., Green, S., Howes, A. and Dunn, L. (2006). Species richness and habitat associations on non- 
flying mammals in Gibraltar Range National Park. Proceedings of the Linnean Society of New South 
Wales 127, 93-105. 

We surveyed mammals in Gibraltar Range National Park using a range of census methods between May 
2003 and September 2005. Our primary survey techniques included 5780 trap nights and more than 40 
km of walked spotlighting transects, and our observations, coupled with previously collected datasets, 
revealed the occurrence of 28 native species and six introduced species of non-flying mammal. To examine 
the importance of habitat heterogeneity in influencing this high manootnal species richness, we surveyed 
mammals across a steep vegetation gradient from swamp, through two eucalypt forest types, to rainforest. 
The mammal community responded strongly to this gradient, with different suites of species favouring 
different parts of the gradient. We also attempted to describe the entire mammal community in one of these 
forest types, wet eucalypt forest, because we suspected it to be one of the more species-rich habitats in the 
park. The mammal community in this forest type was assessed on two 2.6-ha grids using Elliot and cage 
trapping (plus incidental observations), and comprised at least 12 species of non-flying native mammal. 
Brown antechinus (Antechinus stuartii), bush rats {Rattus fuscipes), and fawn-footed melomys {Melomys 
cervinipes) were the most abundant groimd-dwelling mammals in this community. 

Manuscript received 1 May 2005, accepted for publication 7 December 2005. 

KEYWORDS: Bioregion, ecotone, habitat heterogeneity, mammal species richness. New England, non- 
volant. 



INTRODUCTION some records is difficult to confirm. 

One of us (Vernes) has begun a long-term study 

Despite Gibraltar Range National Park being one on the manunals of Gibraltar Range, particularly 

of the oldest parks in New South Wales (reserved in those species that consume and disperse the spores 

1963; NSW NPWS 2005), relatively little pubUshed of hypogeous ectomycorrhizal fungi, otherwise 

information exists on the mammals in the park. Surveys known as 'truffles'. As a first step in understanding 

by Osboume and Marsala (1982) and Pulsford (1982) the mammal community structure and dynamics in 

are summarised by Clancy (1999) who provided a list the region, we have surveyed mammals at a range of 

of 25 native and six introduced species for the park, sites in Gibraltar Range, and present these data here, 

but few other details about habitat associations or in addition to providing a simple species list for the 

relative abundances. More recently, records of flora park, we have also attempted to summarise the broad 

and fauna gathered over many decades by a number habitat preferences of mammals that are present, and 

of government agencies in New South Wales have to show how habitat heterogeneity in the park leads to 

become available on a single web-accessible database the structuring of distinct mammal communities. 
(BioNet Database, 2005). Although this database 
has great utility in the generation of species lists for 

any given region, it provides no quantitative data on MATERIALS AND METHODS 

faunal abundances, and because contributions to the 

database can be made by any interested individual Study area 

(through submissions to the NSW Department of We undertook broad, observational surveys 

Environment and ConservationAtlas), the veracity of throughout Gibrahar Range National Park in 



MAMMALS IN GIBRALTAR RANGE NATIONAL PARK 



northeast New South Wales, but focused our trapping 
and spotUghting in the north-eastern section of the 
park and an adjacent area in the southern part of 
Washpool National Park (Fig. 1). This region of 
these adjacent parks includes the wetter forest types 
to be found in the area (including rainforest) and we 
expected mammal species richness to be highest here. 
The north-eastern region of the park is on the extreme 
eastern edge of the New England Tableland bioregion, 
and straddles the interface between the Tableland and 
the Great Escarpment, a part of the Great Dividing 
Range characterised by rugged topography and 
dramatic changes in elevation. The study region is 
characterised by high ridges and plateaus, with a 
mean elevation of 1000 m, although altitude in the 
park ranges from 200 m to 1175 m (NSW NPWS 
2004). The regional topography and relatively high 
altitude contributes to a high local rainfall of around 
2000 mm annually at the highest elevation around 
the Great Escarpment (NSW NPWS 2004), although 
rainfall decreases rapidly westward away from the 
scarp to be around 1100 mm annually in the drier 
parts of the park (NSW NPWS 2004). Winters are 
usually dry and cold, with average winter daytime 
temperatures of 13°C (NSW NPWS 2004). Most 
rains occur in the months of November to April, with 
average daytime temperatures in summer of around 
25°C (NSW NPWS 2004). 

A diversity of vegetation types is present in the 
park, and they occur in a tortuous mosaic that reflects 
combinations of soil type, a complex underlying 
geology, local rainfall and fire history (NSW NPWS 
2004). Over distances of a few hundred metres 
vegetation can grade from open sedge swamps and 
wooded heaths to tall wet forest and rainforest, and the 
ecotones between these habitats are often sharp. The 
dominant vegetation type can broadly be described 
as eucalypt woodland with a heath-dominated 
understorey; although considerable tracts of open 
sedge swamp, tall open eucalypt forest and rainforest 
are present in the landscape. The importance of 
the more mesic habitats in Gibraltar Range and the 
adjoining Washpool National Park was recognised by 
their listing as part of the Central Eastern Rainforest 
Reserves of Australia (CERRA) World Heritage Area. 
Sheringham and Hunter (2002) provide a detailed 
description of vegetation in these parks. 

The study consisted of three elements. The 
first comprised a survey of mammal species present 
within Gibraltar Range National Park identified 
through observation during spotlighting and other 
visual searches, from their scats and diggings, and 
by examination of road kills. The second element 
of the study was concerned with understanding the 



changes to the small mammal community over a 
continuous ecological gradient spanning a range of 
locally common vegetation types found in the north- 
eastern part of the park. The third element focused 
on the small mammal community in one of these 
vegetation types, wet open eucalypt forest, in order 
to understand more fially the structure of the small 
mammal conmiunity present. This element of the 
study is ongoing, and here we present the first year of 
data. 

Mammal survey of Gibraltar Range 

We conducted spotlighting surveys along ten 
transects ranging in length from 500 m to 1500 m 
in various regions of the park (see Fig. la) between 
May 2003 and September 2005. These transects 
were traversed on foot with 1-3 operators using 30 
W spotlights, beginning at least one hour after dusk. 
Each transect was traversed between 1 and 3 times 
during the study, and all observations included exact 
locations of mammals and dates and times, recorded 
using a handheld GPS (Garmin GPS72). Whenever 
we encountered other signs of mammals in the park 
(scats, calls, diggings etc), or when mammals were 
seen at any time during the study, we also recorded 
the exact location of the observation, date, and time 
of day using a GPS. We also trapped grovmd-dwelling 
mammals in selected areas of the park (see following 
sections, and Fig. lb). To augment our species list, 
we also drew upon data gathered by government 
agencies in Gibraltar Range National Park, and 
lodged with the BioNet Database (2005). This 
database is a compilation of all records from NSW 
State Forests, the NSW Department of Environment 
and Conservation, and the Australian Museum. 

The mammal community along the swamp-to- 
rainforest gradient 

We chose a site where vegetation associations 
graded from open sedge swamp, and graded into dry 
open eucalypt woodland with a heath understorey, then 
into wet open eucalypt forest with a fern understorey, 
and finally into rainforest (Fig. lb). The ecotonal 
boundary between each habitat was sharp, being 
no greater than 25 m wide. We sampled mammals 
across the habitat gradient using four trapping and 
spotlighting transects (T1-T4; Fig. lb) arranged 
so that each transect traversed each habitat and the 
intervening ecotones. A trapping cluster of nine Elliot 
traps arranged in a 3x3 grid with 20-m spacings 
was positioned along each transect in each habitat 
as well as on the ecotone between habitats (Fig. lb) 
for a total of seven clusters (63 traps) per transect. 
The distance between each cluster was variable. 



94 



Proc. Linn. Soc. N.S.W., 127, 2006 



K. VERNES, S. GREEN, A. HOWES AND L. DUNN 



Wadipool MP 






Wa^pool NP 



1000 m 




Gibraltar 
Range MP 



Uharrad 



^ . . Pa>^ bounty 

1^ laiTMCt^ 









m 



Gibraltar 
ftange MP 





isnans- 



1Sf2ff2Qr 




□ Dyapen Sxifypltjf vnodbndwlii 

■ 'blapaq Sxslypkis timn'.vAn 
tab. jndeis(x»y 



[^is aiJIxraps 



Mharrad 



% 



T2 OodbUlanHcl 
2JS-iBlapp^giti 




Figure 1. (a) Map of study area showing major roads and tracks, and location of spotlighting transects, 
(b) Detailed map of main study area (outlined by the box enclosed by a dotted line in Fig. la) showing 
major vegetation types, trapping grids, gradient transects. Inset shows the detail of transects that tra- 
versed the swamp-dry open woodland-wet open forest-rainforest gradient, and the intervening ecotones. 



Proc. Linn. Soc. N.S.W., 127, 2006 



95 



MAMMALS IN GIBRALTAR RANGE NATIONAL PARK 



and depended upon the width of the habitat type; 
however, all trapping clusters were at least 100 m 
apart. Each transect was trapped for 3 nights per field 
trip, with field trips undertaken in November 2003 
and February, March and April 2004. There were 756 
trap nights per field trip, totalling 3024 trap nights for 
this part of the study. Each time a small mammal was 
captured during the study, we identified it to species, 
and collected data on sex, reproductive condition 
and body weight. A numbering system using an ear 
punch was employed to identify individuals over the 
duration of the study. Scat samples from all captured 
mammals were also collected for analysis of diet (to 
be reported elsewhere). 

During November 2003, and February, March 
and April 2004, we also spotlighted each of the four 
transects that traversed the swamp-to-rainforest 
gradient (T1-T4; Fig. 1) twice, on different nights. 
Spotlighting began one hour after dusk, using two 
observers, each using a handheld 30 W spotlight. 
When an animal was sighted, we noted our own 
position with a GPS, estimated the distance and 
recorded the compass bearing to the sighted animal, 
thereby allowing us to later determine where the 
animal was in relation to habitat type. 

Mammal community of the wet open-forest 

Two 160x160 m (2.6-ha) trapping grids in wet 
open eucalypt forest were sampled for small mammals 
in April, June, August and September 2004. Each of 
these grids (Gl and G2; Fig. lb) had a 9x9 grid of 
Type A Elliot traps spaced 20 m apart, with a 5x5 grid 
of larger cage traps spaced 40 m apart superimposed 
upon it. These grids were trapped for 2 - 4 nights 
per sampling period, yielding a total of 2106 Elliot 
trap nights and 650 cage trap nights. We selected this 
forest type based upon our previous survey work that 
identified this habitat as supporting a high diversity 
of mammals. We conservatively estimated relative 
density of trapped mammals on these grids as mean 
minimum numbers of animals known-to-be-alive 
(KTBA), although fixture work at this site aims to 
calculate more robust estimates of population size 
and density for all trappable mammals. 



RESULTS 

Mammals detected in the study region 

We detected 1 1 mammal species across our ten 
spotlighting transects (Table 1). Amongst the seven 
arboreal species seen, the greater glider {Petauroides 
volans) was the most common, being detected at a rate 



of up to 9 animals per km of transect (Table 1). The 
common ringtail possum {Pseudocheirus peregrinus) 
was also regularly encountered (up to 9 animals per km; 
Table 1 ). The mountain brushtail possum {Trichosurus 
caninus) was often seen on transects that traversed 
rainforest, and we also recorded the presence of the 
common brushtail possum {Trichosurus vulpecula) 
in eucalypt forest, but this species appears to be 
considerably less common than T. caninus. We made 
three observations of koalas {Phascolarctos cinereus) 
in the wetter tall open forest along Washpool Way and 
Cedar Track (Fig. la). These records are all within 
Washpool National Park, but one of them was 200 
m west of the Gibraltar Range park boimdary (near 
that end of Cedar Track), and we have included it 
in our species list because the Sydney blue gum {E. 
salignd) habitat it was seen in continues east into the 
park, and we suspect the koala population does too. 
We also recorded three macropods on these spotlight 
surveys, the swamp wallaby {Wallabia bicolor), 
the red-necked pademelon {Thylogale thetis), and 
the parma wallaby {Macropus parmd). The latter is 
listed as vulnerable in NSW, but appears to be locally 
common in the Mulligan's Hut area, where most of our 
sightings were made. Additionally, parma wallabies 
have been sighted at the Coachwood Picnic Area 
and along the Anvil Rock track by park staff" (Kate 
Harrison, pers. comm.), and we also saw one during 
a vehicle spotlighting transect along the Raspberry 
Lookout road, near the western boundary of the park. 
Additional species not detected by spotlight were 
encountered during our mammal trapping (see Table 
2), and these data will be discussed in the following 
sections. 

We made incidental observations of other 
mammals in the region (see Table 2), some of 
which were not detected at any of our trapping 
and spotlighting sites. A macropod that we did not 
detect during spotlighting, the red-necked wallaby 
{Macropus rufogriseus), was regularly seen by 
us during daylight in the eucalypt woodlands and 
forests, and appears to be common and widespread. 
Furthermore, although we encountered swamp 
wallabies only once while spotlighting, evidence of 
them in the form of scats was ubiquitous throughout 
the study area, with the exception of rainforest. Dingo 
{Canis lupus dingo) scats are common along all roads 
and tracks in the park, and we recorded a road-kill 
dingo on the Gwydir Highway near the junction of 
the North West Fire Trail. We saw spotted-tail quoll 
{Dasyurus maculatus) scats in the wet forest areas 
too, but less commonly. Additionally, the northern 
brown bandicoot {Isoodon macrourus), rufous bettong 
{Aepyprymnus rufescens), brush-tailed rock wallaby 



96 



Proc. Linn. Soc. N.S.W., 127, 2006 



K. VEPINES, S. GREEN, A. HOWES AND L. DUNN 






(U 



1) 

.6 

d 

Z 



•a 

3 



c 
o 

■c 

o 

u 

T3 



O 



(D 


^'— ^ 




^ 


o3 


c^ 


z 


E^ 




u< 




C3 




T3 




u 




u 



o 53 

S s 

Ei5 I 



o 



O 



ON 

o 



00 

o 






o 



O 






en 



d 



o 
o 
o 



d 



d 



d 



o6 






d 



d 



00 



n-1 



d 



o 



t^ 




■* 


NO 


T-H 




d 


d 




p 


o\ 


ON 




'-' 


CN 


fNJ 


fN 


O 


^^ 


^^ 


OS 


00 


vd 


r-^ 



o 
o 
o 



o o 
o o 
00 m 



o 
o 



o 
o 



o 

o 

NO 



o 
o 
in 



o o 
o o 



— ■ ^ CN| 



00 




c 


•a 




c 


•3 


(L> 


2 


c« 


60 


•H 


-+-»^ 


a 




1 



o 



o 

'3 

43 



c 



-a 



T3 

o 






B t 



T3 

§ 



C 



DO 

C 



13 



^ = ^ =1 60 == ^ 



•a 
o 






to 



CO 

o 

g 



■s p: ^ S 



o 
u 

C 

O 






*- o 



0) 
SO 



a 

33 

c 
u 

O 




C3 
O 

c 
ex 

o 



0) 

Q 



60 



o 
(U 

C 
(U 

a 
O 



o 

u 

o 

Pi 



-a 
o 
o 



c 
u 
ex 
O 



3 
o 



c 

60 



■a 

c 

O 

o 



C3 

o 

3 



C 



-^ > -^ 






-a 



o 
o 

_>. 
o 

3 

C 
u 
ex 
O 



CO 



J3 
ca 

T3 

C 
ca 

Q 



61 

c 



^ C8 

.2 

"■C BJD 

« .- 

IX ^ 

*- s 

e o 

« ? 

« -S 

e u 

o 

■a J3 

« © 

« C 

Clh o 

— ^ 

S U 

o ;. 

•■C o 

es « 

(1) O) 

GiX) .A 

s s 



s: 



J5 



a. II II 






6n 



2 o 



WD S S 

-5 ^ II 

-^& 



es w 
±i 4> 

2 S 
•-9 i 

w S 

5 -^ 
a S 
o ^ 

" D. 

u 
*> "B 

= H 
« « 

OX) o 

•mm 

■£, « 

UD ^ 
•♦J 

o e 
S^ 



a 
s 



fc« CX ^ 
1^ 

Si ^ 

g § II 

*S -5; H 

"s II •- 

(« PQ S 

a 

s 
s 



ce 

-a 



§1 



S ^ 

£ 0H 



o R 

V .2 ct 
'3 « S 

aZ 

o 
II 



^ 2 

a ^ 

•i: ^ 

^^ II 

« "^ ■< 

u 

e 

B Oh 

c8 s 5 

=« -2 S 

o « 2 



S 



« a; 



s 



£ 



I 

2 II 



13 II 

PES ^ 



2h 






Proc. Linn. Soc. N.S.W., 127, 2006 



97 



MAMMALS IN GIBRALTAR RANGE NATIONAL PARK 



Table 2. Species of mammal detected in the study area at Gibraltar Range and Washpool Na- 
tional Parks, and their habitat associations. In each case, the method we primarily used to de- 
tect the presence of these species is listed. Species that were not detected during this study, but 
that are recorded as being present according to the BioNet Database (2005) are also recorded. 



Common and Scientific Name 



Detection 
Method 




PROTOTHERIA (Monotremes) 
Family Tachyglossidae 

Short-beaked Echidna, Tachyglossus aculeatus 

Family Ornithorhynchidae 

Platypus, Ornithorhynchus anatinus 

METATHERIA (Marsupials) 

Family Dasyuridae 

Spotted-tailed QuoU, Dasyurus maculatus^ 

Brown Antechinus, Antechinus stuartii 

Family Peramelidae 

Northern Brown Bandicoot, Isoodon macrourus 

Long-nosed Bandicoot, Perameles nasuta 

Family Burramyidae 

Eastern Pygmy-possum, Cercartetus nanus^ 

Family Acrobatidae 

Feathertail Glider, Acrobates pygmaeus 

Family Petauridae 

Sugar Glider, Petaurus breviceps 

Yellow-bellied Glider, Petaurus australis^ 

Greater Glider, Petauroides volans 

Family Pseudocheiridae 

Common Ringtail Possum, Pseudocheirus peregrinus 

Family Phalangeridae 

Mountain Brushtail Possum, Trichosurus caninus 

Common Brushtail Possum, Trichosurus vulpecula 

Family Phascolarctidae 

Koala, Phascolarctos cinereus^ 

Family Potoroidae 

Rufous Bettong, Aepyprymnus rufescens^ 

Long-nosed Potoroo, Potorous tridactylus^ 

Family Macropodidae 

Parma Wallaby, Macropus parma^ 

Red-necked Wallaby, Macropus rufogriseus 

Brush-tailed Rock Wallaby, Petrogale penicillata^ 

Red-legged Pademelon, Thylogale stigmatica^ 

Red-necked Pademelon, Thylogale thetis 

Swamp Wallaby, Wallabia bicolor 




Diggings, seen 
Seen* 



Scat 
Trap 

BioNet 
Trap, spotlight, scat 

Trap 

Spotlighting 

Spotlight, calls 

BioNet 

Spotlight 

Spotlight 

Trap, spotlight, scats 
Spotlight 

Spotlight 

BioNet 
Trap 

Spotlight, stag-watch 
Scat, seen, road-kill 

BioNet 

BioNet, road-kiir 

Spotlight, heard, scats 

Scat, seen, spotlight 



98 



Proc. Linn. Soc. N.S.W., 127, 2006 



K. VERNES, S. GREEN, A. HOWES AND L. DUNN 



TABLE 2 CONTINUED 

Common and Scientific name 
EUTHERIA ('Placental' Mammals) 

Family Muridae 

New Holland Mouse, Pseudomys novaehollandiae 

Fawn-footed Melomys, Melomys cervinipes 

House Mouse, Mus musculus^ 

Bush Rat, Rattus fuscipes 

Swamp Rat, Rattus lutreolus 

Black rat, Rattus rattus^ 

Family Canidae 

Dingo, Canis lupus dingo 

European fox, Vulpes vulpes^ 

Family Felidae 

Feral Cat, Felis catus^ 

Family Leporidae 

European rabbit, Oryctolagus cuniculus^ 

Family Suidae 

Feral Pig, Sus scrofa^ 



Detection method 



Trap 
Trap 
Trap 
Trap 
Trap 
Clancy (1999) 

Scat, road-kill 
BioNet 

Seen 

Seen 

Diggings 



IListed as 'Vuhierable' in NSW 

2Listed as 'Endangered' in NSW 

3Introduced species 

* K. Harrison (Park Ranger), personal communication 

'^ R. Goldingay (Southern Cross University), personal 

communication 



{Petrogale penicillata), red-legged pademelon {T. 
stigmatica), feathertail glider {Acrobates pygmaeus), 
and yellow-bellied glider {Petaurus australis) have 
been recorded within our study area by others (BioNet 
Database, 2005). 

Amongst introduced species, cats {Felis catus) 
and foxes {Vulpes vulpes) have been recorded in the 
park (BioNet Database 2005) and we have seen a 
rabbit {Oryctolagus cuniculus) in the Mulligan's Hut 
area. Feral pigs {Sus scrofa) have not been previously 
reported from the park, but we have noted diggings 
characteristic of pigs on the edges of swamps along 
Mulligan's Drive, but their presence needs to be 
verified with a sighting. 

In all, ova work in Gibraltar Range National 
Park in 2003 and 2004, coupled with data gathered 
from the BioNet Database, indicates the presence of 
28 native and six infroduced species of non-flying 
mammal (Table 2). 

The mammal community along the swamp-to- 
rainforest gradient 

Seven species of small mammal were detected 



in our fraps across the habitat gradient (Sites Tl- 
T4): four species of native rodent (bush rat Rattus 
fuscipes, swamp rat R. lutreolus, fawn-footed 
melomys Melomys cervinipes, and New Holland 
mouse Pseudomys novaehollandiae), the infroduced 
house mouse {Mus musculus), the browTi antechinus 
{Antechinus stuartii) and the eastern pygmy possum 
{Cercartetus nanus). Spotlighting yielded a fiarther 
four species: the greater glider, common ringtail 
possum, sugar glider {Petaurus breviceps), and 
swamp wallaby. 

The small mammal community changed 
markedly across the habitat gradient spaiming 
swamp to rainforest (Fig. 2), despite this representing 
a distance of only about 700 m. Amongst small 
frappable mammals, several patterns in distribution 
emerged. R. fuscipes and M. cervinipes changed 
significantly in abundance (KTBA) between habitats 
(P = 0.004 and P = 0.00 1 respectively; Kruskal-Wallis 
Nonparametric ANOVA), with abundance increasing 
from the dry eucalypt woodland and the wet eucalypt 
forest, peaking on the open foresf rainforest ecotone, 
before declining inside the rainforest (Fig. 2a). P. 
novaehollandiae and M. musculus abundances were 
greatest on the ecotone between swamp and open 
woodland, declining either side of this region (Fig. 
2b), significantly for P. novaehollandiae {P = 0.001; 
Kruskal-Wallis Nonparametric ANOVA with Duim's 
Multiple Comparison Test), but the few captures 
of M musculus precluded statistical comparisons. 



Proc. Linn. Soc. N.S.W., 127, 2006 



99 



MAMMALS IN GIBRALTAR RANGE NATIONAL PARK 



XI 10.0 

o 



a. 

< 
PQ 

c 



7.5 



5.0- 



2.5- 



0.0 



R.fuscipes 
M. cervinipes 









— ^ ^ 1 , 1 , , — 

Swamp Ecotone Woodland Ecotone Wet forest Ecotone Rainforest 



T3 
O 

CI, 

;-( 

<U 

< 

c 



5.0-1 
4.0- 
3.0- 
2.0- 
1.0- 



S 0.0- 



j_ \ 



I 

I 






-^ 



-■ — R. lutreolm 

-♦- - - P. novaehollandiae 

■ ▲ — M. musculus 






Swamp Ecotone Woodland Ecotone Wet forest Ecotone Rainforest 



T3 
O 

a. 

lU 

< 

PQ 

H 

S3 
0) 



5.0-1 
4.0- 
3.0- 
2.0- 
1.0- 



0.0- 



A. Stuart a 
C. nanus 



-4^ 



^ 



.4i 



Swamp Ecotone Woodland Ecotone Wet forest Ecotone Rainforest 



c 


2.5 


o 




• — < ^-^ 






2.0 


*-> iys 




<o c 




T3 cd 

^ fa 


1.5 


oix 




^"^ 


1.0 


2.^ 




§§ 


0.5 


o 




s 


0.0 



— p. volans 
■- P.peregrinus 



i 



\ 



-r 



Swamp Ecotone Woodland Ecotone Wet forest Ecotone Rainforest 

Habitat type 

Figure 2. Changes in the mammal community across the gradient from swamp to rainforest, for the 
species detected in traps (Figs a-c) and by spotlighting (Fig. 2d). 



100 



Proc. Linn. Soc. N.S.W., 127, 2006 



K. VEPUSIES, S. GREEN, A. HOWES AND L. DUNN 



Likewise, our few captures ofR. lutreolus suggest that 
this species may also be associated with the swamp/ 
open forest ecotone (Fig. 2b). C. nanus was captured 
in small numbers from rainforest to dry open eucalypt 
forest (Fig. 2c), but again, too sparsely to allow 
statistical analysis. A. stuartii was captured across the 
gradient from dry open woodland to rainforest, and 
no significant differences in abundance were detected 
(Fig. 2c; P > 0.3, Kruskal-Wallis Nonparametric 
ANOVA). 

The two arboreal species that we sighted regularly 
{P. volans and P. peregrinus) during spotlighting 
along gradient transects (Tl - T4) also showed 
distinct habitat association. P. volans was seen in 
all eucalypt-dominated habitats and ecotones, but 
was significantly more common in the wet eucalypt 
forest and the wet forest/dry woodland ecotone 
(Fig. 2d; P = 0.004, Kruskal-Wallis Nonparametric 
ANOVA), whereas P peregrinus was found from 
dry eucalypt woodland to rainforest. Although this 
species appeared to reach its greatest abundance 
on the wet forest/rainforest ecotone (Fig. 2d), high 
variance in the rainforest ecotone may have masked 
any differences in detection rate across the gradient 
(P = 0.15, Kruskal-Wallis Nonparametric ANOVA; 
Fig. 2d). Three other arboreal mammals {P. breviceps, 
T. caninus and T. vulpecula) were each seen only once 
during this part of the study (all near the wet forest/ 
rainforest ecotone), so we were unable to determine 
their local habitat associations across the gradient. 

The small mammal community in wet open-forest 

Elliot trapping in the wet open eucalypt forest 
yielded a sub-set of the small mammals detected 
in sites T1-T4, although cage traps captured some 
species not detected at those sites. Our Elliot traps 
mostly captured Rattus fuscipes (148 captures; 43 
individuals), Antechinus stuartii (107 captures; 21 
individuals) and Melomys cervinipes (68 captures; 
24 individuals), with relatively fewer captures of 
R. lutreolus (7 captures; 2 individuals). For these 
four species combined, the capture success of small 
mammals on our two grids was about 16.7% (Table 
2). During any one field trip, the minimum numbers 
of individual animals known-to-be-alive (KTBA) on 
each 2.6-ha grid averaged about 10 i?. fuscipes, 8 A. 
stuartii, 5 M cervinipes and <1 R. lutreolus. Cage 
traps captured Trichosurus caninus (21 captures; 6 
individuals) on both grids, long-nosed bandicoots 
(Perameles nasuta; 4 captures; 3 individuals) on Gl 
and a single capture of a long-nosed potoroo {Potorous 
tridactylus) on G2. Incidental observations made of 
other mammals on these grids included sugar gliders, 
common ringtail possums, greater gliders, common 
brushtail possums, and swamp wallabies. 



DISCUSSION 

Mammal richness in Gibraltar Range National 
Park 

In a survey of mammals in a 2400-km^ area in 
the upper Richmond and Clarence River catchment 
in north-eastern NSW, Calaby (1966) recorded the 
presence of 35 non-flying native mammals, noting 
that, at the time, it represented one of the richest 
Australian mammal faunas that had been reported 
for a comparable area. Bamett et al. (1976) surveyed 
the mammal fauna in a 1 1 8-km^ area at Clouds Creek 
on the eastern edge of the New England Bioregion, 
recording 27 non-flying native mammals, and 
again, this area was heralded for its high species 
richness. The Clouds Creek area was very similar in 
geographic context to our own, and serves as a usefial 
benchmark for our study at Gibraltar Range National 
Park (area = 253 km^) where we recorded 28 native 
and six introduced species. Based upon information 
in the BioNet Database (2005), this species list would 
include at least 36 native mammals if the adjacent 
Washpool National Park had been included in our 
survey, making these parks, and those adjacent to 
them, of great importance in the protection of the 
regional mammal biodiversity of north-eastern New 
South Wales. Recently, Jarman and Vemes (in press) 
summarised the mammals of the New England 
Bioregion, concluding that there were 43 species of 
non-flying native mammal still present there. Based 
upon the data we have gathered, Gibraltar Range 
National Park accommodates 65% of the bioregional 
non-flying mammal fauna, and together with Washpool 
National Park, these reserves accommodate 83% of 
the bioregional non-flying mammal fauna. 

Macropods (kangaroos, wallabies and 
rat-kangaroos in the families Potoroidae and 
Macropodidae) are one of the most species rich 
groups of mammal that we recorded in Gibraltar 
Range National Park (eight species), and again, this 
richness is comparable to other studies in the region. 
Calaby (1966) recorded 11 species of macropod 
in the Upper Richmond and Clarence catchment, 
Bamett et al. (1976) recorded nine species at Clouds 
Creek, and Jarman et al. (1987) recorded ten species 
at Wallaby Creek, which is located in the northern 
headwaters of the Clarence River within the region 
where Calaby (1966) worked. Jarman and Vemes (in 
press) noted that 12 species of macropod persist in the 
New England Bioregion. Interestingly, two of these 
species, the eastem grey kangaroo (M giganteus) and 
common wallaroo (M robustus) appear to be absent 
from Gibraltar Range National Park, despite their 
being the most common macropods across the largely 
modified landscape of the New England Tableland. 



Proc. Linn. Soc. N.S.W., 127, 2006 



101 



MAMMALS IN GIBIIALTAR RANGE NATIONAL PARK 



Density of vegetation at ground level is typically high 
in most habitats in the park, which would favour the 
smaller wallabies and restrict the movement of the 
larger species. 

As with previous studies in north-eastern New 
South Wales, the macropod diversity we recorded can 
be attributed to the great diversity of habitat types 
present at Gibraltar Range National Park, within 
a relatively small area. For example, we recorded 
pademelons (Thylogale spp.) in rainforest, and based 
on other research in north-eastern New South Wales 
(Calaby 1 966; Bamett et al. 1 976; Jarman and Phillips 
1989) we suspect that T. thetis is more likely to 
occur around the wet sclerophyll/rainforest ecotone, 
whereas T. stigmatica is likely to occur deeper within 
the rainforest. P. tridactylus was detected in wet 
forest with a dense understorey, and M. rufogriseus 
was detected primarily in the dry open forest. W. 
bicolor is probably the most widespread macropod 
in the park, and we detected its presence in all non- 
rainforest habitats. 

M. parma inhabits wet eucalypt forest and 
rainforest margins throughout its distribution, but 
at Gibraltar Range, it also occurs in drier eucalypt 
woodland with a heath understorey (Maynes 1977). 
The presence of M. parma in the dry forest habitat 
is unusual for this species; in a survey of M. parma 
throughout New South Wales, Maynes (1977) noted 
that the area along Mulligan's Drive was the only 
dry sclerophyll forest site in their range where he 
recorded M. parma as being resident. He attributed 
this occurrence to the availability of dense shrubby 
cover in the forest understorey for shelter that was in 
close proximity to open grassy areas around swamps 
where the wallabies could feed. 

Although P. penicillata has apparently been 
sighted in the steep, rocky escarpment region at the 
eastern edge of the park (BioNet Database, 2005), 
this record appears to unsubstantiated (Clancy 1999) 
and needs to be verified, as do the few sightings in the 
BioNet Database (2005) fox A. rufescens of which at 
least one may have been a misidentification (Clancy 
1999). Both species occur in the adjacent Washpool 
National Park (BioNet Database 2005). Another 
three species of macropod (eastern grey kangaroo 
M. giganteus, common wallaroo M robustus, and 
whiptail wallaby M parry i) also occur in the adjacent 
Washpool National Park. Thus, the only macropod 
species that occurs in New England (see Jarman 
and Vemes in press), but does not occur locally in 
the Gibraltar Range/Washpool region, is the black- 
striped wallaby (M dorsal is). 

Another species-rich group in the park was 
the possums and gliders (see Table 2). Of the eight 



species reported to be present in the park, we recorded 
seven, with the most common and widespread of 
these being P. volans and P. peregrinus. Although we 
only recorded the small, cryptic feathertail glider {A. 
pygmaeus) once, it is almost certainly widespread and 
common in the park, despite only a single record of 
this species in the BioNet Database (2005). However, 
we could not verify the presence of the yellow-bellied 
glider {P. australis), of which one sighting has been 
recorded in the park near its northern boundary with 
Washpool National Park (BioNet Database 2005). 

Threatened species in the park 

Nine threatened species of mammal are listed as 
occurring in Gibraltar Range National Park (Table 
2). In particular, the park is reputed to have a large 
population of D. maculatus (NSW NPWS 2005), 
and together with Washpool and Barool National 
Parks, contains a significant percentage of the state 
population of M parma (NSW NPWS 2005). Other 
macropods of conservation interest in the park include 
T. stigmatica and P. tridactylus, and, if records are 
substantiated, A. rufescens and P. penicillata. 

Mammal community dynamics 

The diversity of habitats within a relatively 
small area is one of the factors that contribute to the 
high species richness that we recorded in Gibraltar 
Range National Park. We tested this by trapping 
and spotlighting mammals across a steep gradient 
in vegetation fi-om swamp to rainforest, and found 
that despite the short distance (-700 m) there were 
significant and consistent changes in the structure 
of the mammal commvmity. One suite of species {R. 
fuscipes, M. cervinipes and P. peregrinus) appeared 
to have wide habitat tolerances but reached their 
highest abundances at the ecotone between eucalypt 
forest and rainforest, whereas another suite of species 
{P. novaehollandiae, M. musculus and R. lutreolus) 
favoured the ecotone between swamp and the dry, 
heath-dominated eucalypt woodland. Although 
we had fewer captures of eastern pygmy possums 
{Cercartetus nanus), our data point towards this 
species favouring the intermediate vegetation types 
along the gradient (wet and dry eucalypt forest and 
woodland), particularly the ecotone between the two. 
These are the floristically more diverse habitats along 
our habitat gradient in terms of flowering heath plants 
such as banksias {Banks ia spp.) and bottlebrushes 
(Callistemon spp.) (Howes 2004), and they are 
therefore likely to support the highest numbers of 
this primarily nectar-feeding marsupial (Ward 1990). 
A. stuartii occurred across much of the gradient and 
appeared to be the only habitat generalist that we 



102 



Proc. Linn. Soc. N.S.W., 127, 2006 



K. VERNES, S. GPIEEN, A. HOWES AND L. DUNN 



detected. P. volam was widespread within the open 
forest habitat across the entire gradient, but reached 
highest densities in the wet eucalypt forest, an 
observation that is consistent with other studies (e.g. 
see Bennett et al. 1991). Although too few sightings 
were made of brushtail possums (Trichosurus spp.) 
during this part of the study, previous work on T. 
caninus indicated that it is a rainforest/wet forest 
specialist (How 1972). We saw this species during our 
various spotlighting surveys throughout the park only 
in the rainforest and its wet eucalypt forest ecotone, 
whereas T. vulpecula is a species of more open forest 
(How 1972) and we saw it in low numbers in the wet 
open eucalypt forest. 

Williams and Marsh (1998) studied ground- 
dwelling mammals across a rainforest/open-forest 
ecotone in north Queensland, and our observations 
from Gibraltar Range are consistent with their work, 
despite some differences in the way individual 
species responded. They noted significant changes 
to the mammal community across their vegetation 
gradient, with some species being more generalist in 
habitat preference (e.g. R.fuscipes and M cervinipes), 
whereas others were strictly associated with rainforest 
(e.g. A. stuartii) or open forest (e.g. R. lutreolus). 

On our intensively sampled grids in wet open 
eucalypt forest, R. fuscipes and A. stuartii were the 
most dominant species in terms of animals known- 
to-be-alive (KTBA), followed by M cervinipes. 
By comparison, R. lutreolus was considerably less 
common. Population sizes of other species were more 
difficult to discern, mainly because these animals 
are harder to trap using conventional techniques. 
As a continuation of this study, we will trial a range 
of methods for the capture of some of the larger 
mammals, including bandicoots, potoroos, possums 
and wallabies. 

Summary and conclusions 

The data we gathered on habitat associations 
of mammals from trapping grids and transects, and 
spotlighting transects throughout this study, as well 
as other direct and indirect observations of mammals 
within the park, yielded a total of 28 species of 
non-flying native mammals. The most species-rich 
habitats in the park appear to be the wet eucalypt 
forests and the dry open eucalypt woodland with 
a heath understorey (Fig. 3). Importantly though, 
rainforest, swamps and rocky outcrops accommodate 
species not found in these dominant habitat types, 
and the overall habitat complexity at Gibraltar Range 
serves to generate its high species richness. Although 
the richness of mammals in Gibraltar Range is high 
by regional standards, we feel that some records 



of mammals in the park that were not gathered by 
us require further validation (e.g. P. penicillata, A. 
rufescens, P. australis), and we plan to target these 
species as part of our future work. Furthermore, there 
are species in the national parks adjacent to Gibraltar 
Range that have not been recorded in the park (such 
as the brush-tailed Phascogale Phascogale tapoatafa 
and the common dunnart Sminthopsis murina), despite 
suitable habitat probably being present. Thus, our 
continuing work will also aim to provide a definitive 
and comprehensive list of mammal species in time. 



ACKNOWLEDGEMENTS 

We thank the Department of Environment and 
Conservation, particularly park ranger Kate Harrison, 
for allowing us to undertake this work in Gibraltar 
Range National Park, and for providing such hospitable 
accommodation for us while we were in the field. Thanks 
also to Tani Cooper for reading an earlier version of the 
manuscript and suggesting valuable improvements. We 
are grateful also to The University of New England for 
providing the funds through their URG scheme that made 
this research possible. 

REFERENCES 

Bamett, J.L., How, R.A. and Humphreys, W.F. (1976). 
Mammals of Clouds Creek, north-eastern New 
South Wales, and their distribution in pine and native 
forests. Australian Zoologist 19, 23-34. 

Bennett, A.F., Lumsden, L.F., Alexander, J.S.A., Duncan, 
P.E., Johnson, P.G., Robertson, P. and Silveira, C.E. 
(1991). Habitat use by arboreal mammals along an 
environmental gradient in north-eastern Victoria. 
Wildlife Research 18, 125-146. 

BioNet Database (2005). URL: www.bionet.nsw.gov.au. 
Data accessed on August 2, 2005. 

Calaby, J.H. (1966). Mammals of the upper Richmond and 
Clarence Rivers, New South Wales. CSIRO Wildlife 
Research Paper 10, 1-55. 

Clancy, G. (1999). Report on the fauna of national parks 
and nature reserves in the Glen Innes District. 
Unpublished report to the Glen Innes District of 
NPWS. 

How, R.A. (1972). The ecology and management of 
Trichosurus species (Marsupialia) in NSW PhD 
Thesis, Department of Zoology, The University of 
New England. 

Howes, A. (2005). Structure of the mammal community 
across a swamp-woodland-rainforest ecotone in 
northern NSW. BSc Honours Thesis, The University 
of New England, Armidale, NSW. 

Jarman, P.J., Johnson, C.N., Southwell, C.J. and Stuart- 
Dick, R., 1987, Macropod studies at Wallaby 
Creek. I. The area and animals. Australian Wildlife 
Research 14, 1-14. 



Proc. Linn. Soc. N.S.W., 127, 2006 



103 



MAMMALS IN GIBRALTAR RANGE NATIONAL PARK 



s 


hS 


•Tl 


Si 




5< 


a* 


A 


hd 


&i 


U> 


^ 


R, 


!ZI 


T 




rs 


•^ 


B 


ST 


o 


O 




ffi 


a 


Si 


s 


^^ 


ii^ 


hH 


sr 


n 


S 

s 


a 


S 


ffi 


Si 


"O 




«^ 


•1 




Si 


S 




S! 






rs 


S 




ST 


Si 




Si 


o 
s 




Si 


o 




ST 


sr 




t 


re 




a 


a* 




H 


o 

Si 




sr 


a 




re 

-*• 


5 




n 


» 






B 






3 




a 


Si 




o 

D 


n 

o 



n 


s 


a 


s 


re 


*^ 


v 


re 






n 




f^ 


Si 


re 


[-1 


a 


a 


a 


sr 


^^ 


m 


ar 






■s 


Ml 


-f- 


as 


re 


(IQ 


orq 


B 


re 


3 


Si 


sr 





Si 


s 


a- 


Si 


re 


VI 


re 





S 


r> 




■■• 


S 


Si 


K 


f^ 


re 



s 


Si 


w 


8 


Si 


a 


«■ 


5 





o 


w 


fr 


1-1 


» 


Si 


re 


<i*- 


a 


Si 


5; 


» 


p^ 


Si 


S" 


s 


-a 


^^ 


re 


2 


3 


Si 

a 






9 





Si 


s 






B> I <!t> ^ 

i ll I 

*" F 



I 



Ji 



Co 



tJ r^ tf i- U "n tr 



e B a ti ena 

I I II 






P r*: 3* .ts Tl TB 






I 



l.l§" 






104 



Proc. Linn. Soc. N.S.W., 127, 2006 



K. VERNES, S. GREEN, A. HOWES AND L. DUNN 



Jarman, P. J. and Phillips, CM., 1989, Diets in a 
community of macropod species In Grigg, G., 
Jarman, P. and Hume, I., (Eds) Kangaroos, wallabies 
and rat-kangaroos, Surrey Beatty and Sons Pty 
Limited, Australia. Pp. 143-149. 

Jarman, P. and Vemes, K. (in press). The Wildlife of New 
England In "High lean country foil of old stories": 
Envirorunent, Peoples and Traditions in New 
England. (Eds J. Ryan, A. Atkinson, I. Davidson, and 
A. Piper) (Heritage Futures Research Centre, The 
University of New England). 

Maynes, G.M. (1977). Distribution and aspects of the 
biology of the parma wallaby, Macwpus parma, in 
New South Wales. Australian Journal of Wildlife 
Research 4, 109-125. 

NSW NPWS (2004). Gibraltar Range National Park 

- Climate. URL: http://wvifw.nationalparks.nsw.gov. 
au/ Data accessed on October 20, 2005. 

NSW NPWS (2005). Gibraltar Range Group of Parks 
(Incorporating Barool, Capoompeta, Gibraltar 
Range, Nymboida and Washpool National Parks and 
Nymboida and Washpool State Conservation Areas) 
Plan of Management. Department of Environment 
and Conservation (NSW). 

Osborne, W. S. and Masala, V. 1982. Vertebrate Faunal 
Studies in the Washpool - Gibraltar Range Region. 
Unpublished report prepared for the Washpool Faunal 
Study Management Committee, Total Environment 
Centre, Sydney. 

Pulsford, I. 1982. Mammals of the Gibraltar Range. 
Unpublished report to NPWS. 

Sheringham, RS. and Hunter J.T (2002). Vegetation and 
floristics of Gibraltar Range National Park. NSW 
National Parks and Wildlife Service, Glen Innes. 

Ward, S.J. (1990). Life history of the eastern pygmy 
possum, Cercartetus nanus (Burramyidae: 
Marsupialia), in south-eastern Australia. Australian 
Journal of Zoology 28, 287-304. 



Proc. Linn. Soc. N.S.W., 127, 2006 105 



106 



The Discovery and Early Natural History of the Eastern Pygmy- 
Possum, Cercartetus nanus (Geoffroy and Desmarest, 1817) 

Jamie Mark Harris 

School of Environmental Science and Management, 
Southern Cross University, Lismore NSW 2480 (jharril l@scu.edu.au) 



HARRIS, J.M. (2006). The discovery and early natural history of the eastern pygmy-possum, Cercartetus 
nanus. Proceedings of the Linnean Society of New South Wales 127, 107-124. 

Early accounts of the eastern pygmy-possum, Cercartetus nanus (Marsupialia: Burramyidae), are 
reviewed and the history of its discovery is reported. Fran9ois Peron discovered the species when on a short 
stay on Maria Island in 1802. Various names have been conferred upon it, but C nanus is now accepted. 
The early natural history literature on C nanus has some very interesting and highly relevant accounts of 
morphology, distribution, behaviour, habitat and diet. Some discrepancies and misinterpretations in the 
early literature are identified, and several interesting 1 9* Century illustrations of C. nanus are reproduced. 
This study documents the significance of the primary soiu^ce material pertaining to this small elusive 
marsupial. 

Manuscript received 4 May 2005, accepted for publication 21 September 2005. 

KEYWORDS: Burramyidae, Cercartetus nanus, discovery, natural history, nomenclature 



INTRODUCTION 

The eastern pygmy-possum, Cercartetus nanus, 
is broadly distributed in Tasmania and along the 
eastern seaboard of mainland Australia from south- 
eastern Queensland, through coastal New South 
Wales and Victoria, and into south-eastern South 
Australia (Strahan 1995). Currently there are two 
recognised subspecies: C. nanus nanus in Tasmania; 
and C. n. unicolor on the mainland (Wakefield 
1963; McKay 1988). It is a small (~24g) and agile 
tree-dwelling marsupial that feeds chiefly on nectar, 
pollen and invertebrates within a range of habitats 
including heathland, woodland, sclerophyll forest and 
rainforest. Modem studies have documented some 
aspects of the population biology of this species and 
it is understood that it depends on the presence of a 
diverse range of flowering plants (particularly Banksia 
in certain areas), and that seasonal food availability 
influences both the timing and duration of breeding 
(Turner 1984, 1985; Ward 1990; Turner and Ward 
1995; Bladon et al. 2002). During winter, C. nanus is 
able to store up fat in its body and tail, and can exhibit 
torpor (Geiser 1993; Turner and Ward 1995; Bladon 
et al. 2002). Pygmy-possums have a prehensile tail, 
which resembles that of a ringtail possum, and also 
syndactylous hind feet and an opposable clawless 
hallux (Turner and McKay 1989). 



Cercartetus nanus shares the family Burramyidae 
with four other extant species: the long-tailed pygmy- 
possum, C. caudatus, little pygmy-possum, C. 
lepidus, western pygmy-possum, C. concinnus and 
mountain pygmy-possum, Burramys parvus (Strahan 
1995). This paper investigates the discovery and early 
accounts of the natural history of C. nanus, which was 
the first of the burramyids to be formally described 
by Europeans (Desmarest, 1817). Subsequently, 
C. concinnus (Gould, 1845) was recognised, then 
C. caudatus (Milne-Edwards, 1877), C. lepidus 
(Thomas, 1888) and 5. /?arvM5 Broom, 1896. 



MATERIALS AND METHODS 

The work of Thomas (1888) is instructive 
for early accounts of Cercartetus spp., and in this 
regard 36 references for C. nanus (and its synonyms) 
were provided from literature published from 1817 
to 1875. The Kinetica and Firstsearch databases 
were used to identify libraries within Australia and 
overseas that held the relevant early natural history 
titles from which copies of the relevant articles 
were obtained. I also supplemented these papers 
by searching for mention of the species in the early 
volumes (<1970) of the Australian Zoologist and the 
Victorian Naturalist (Harris 2005). The literature was 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 



examined and reviewed for information on discovery, 
taxonomy, morphology, distribution, abundance, diet, 
habitat and behaviour. 



HISTORICAL RECORDS 

Discovery 

The first specimen of C. nanus known to Europeans 
was collected by Fran9ois Peron, a naturalist aboard 
Nicolas Baudin's voyage to the south seas on the ships 
Le Geographe and Le Naturaliste (1800-1804). His 
discoveries and observations whilst in Australia have 
long interested historians (Triebel 1948; Faivre 1953; 
Cornell 1965; Plomley 1983; Wallace 1984; Homer 
1987; Plomley et al. 1990; Hunt 1999; Anderson 
2001). He is credited with the collection of about 
100,000 zoological specimens, 2500 of which were 
new to science, including C. nanus. Whilst on a short 
stay on Maria Island, off eastern Tasmania between 
19 and 27 February 1802, Peron traded with the 
Aboriginal inhabitants (the Tyreddeme people; Ryan 
1981) for a single small marsupial. Peron (1809:233) 
wrote (in translation) 'In the class of mammiferous 
animals, I only saw one kind of Dasyurus, which was 
scarcely as large as a mouse. I obtained one that was 
alive, in exchange for a few trifles, fi-om a savage who 
was just going to kill and eat it'. In an unpublished 
manuscript (now held in the Le Havre Museum in 
France) Peron also wrote that the animal 'was given 
to me by the natives; it was still alive; I believe it to 
be a new species and have described it as Didelphis 
muroides because of its resemblance to the D. mus 
of Lirmaeus' (Observations zoologiques by Fran9ois 
Peron, on Maria Island, unpublished manuscript 
# 18043:31). The specimen collected by Peron (a 
juvenile male) was transported back to France, and 
is now held in the Museum National d'Historie 
Naturelle in Paris as the holotype (Julien-Laferriere 
.1994). Cercartetus nanus still presumably inhabits 
Maria Island, as there is a relatively recent record 
from 1969, when two young animals were found in a 
dead tree being cut for firewood (Animals and Plants 
Protection Board 1969). 

Plomley et al. (1990) erroneously stated that the 
single small marsupial collected on Maria Island by 
Peron was the type specimen for Antechinus minimus. 
This was probably based on a similar mistake made 
by Waterhouse (1846) which was highlighted by 
Wakefield and Wameke (1963). Waterhouse (1846) 
interpreted Peron's statement of finding a 'Dasyurus' 
as meaning that the dasyurid A. minimus was also 
collected from Maria Island, when evidently C. nanus 
was the only mammal species collected (Desmarest 



1817, 1820; Cuvier 1826; Lesson 1827, 1838, 
1842; Temminck 1827; Fischer 1829; Schinz 1844; 
Iredale and Troughton 1934; Tate 1945; Wakefield 
and Wameke 1963). The type specimens for both C. 
nanus and A. minimus were collected by Peron, but 
the latter is considered to have come from Waterhouse 
Island, which lies close to the north-eastem coast of 
Tasmania (Wakefield and Wameke 1963; Rounsevell 
1989). 

Taxonomy and nomenclature 

Upon the retum of the Baudin expedition to 
France in 1804, several of the great French zoologists 
of the period, including Anselme-Gaetan Desmarest 
and Etieime Geoffroy Saint-Hilaire worked rapidly 
describing and classifying the specimens collected 
by Peron. In the encyclopedic Nouveau Dictionairie 
d'Histoire Naturelle, Desmarest (1817) described 
the small marsupial collected from Maria Island as 
Phalangista nana Geoff. (=Geoffroy). However, 
subsequently there has been uncertainty as to 
whom the specific name nana ('dwarf') should 
be attributed to, with some authors allocating it to 
Geoffroy (e.g. Cuvier 1827; Temminck 1827; Lesson 
1828, 1830 1838; Fischer 1829; Gray 1841; Schinz 
1844; Waterhouse 1846; Gunn 1852) and others 
to Desmarest (e.g. Giebel 1859; Lydekker 1896; 
Lucas 1897; Le Souef and Burrell 1926; Iredale and 
Troughton 1934; Wakefield 1963). McKay (1988) 
stated that it must be dated from Desmarest [and hence 
not Geoffroy] as 'Geoffroy's (1803) manuscript was 
never published'. However, Julien-Laferriere (1994) 
stated that the species is not mentioned in Geoffroy's 
(1803) Catalogue des Mammiferes, contrary to what 
McKay (1988) allows to be assumed. Furthermore, 
the specimen did not arrive in France until 1804. 
Although Geoffroy did not write on the species, 
Beaufort (1966) believed that Desmarest's allocation 
of the name to his colleague was intentional (also see 
Desmarest 1820), and accordingly he proposed that it 
should officially be attributed to both as Cercartetus 
nanus (Geoffroy and Desmarest, 1817). In this, I have 
followed Beaufort (1966). 

In a new edition of Nouveau Dictionairie 
d'Histoire Naturelle, published in 1818, a description 
of P. nana equivalent to Desmarest (1817) was 
also published. This is significant because the 
1818 edition is sometimes incorrectly referred to 
as the first description for the subject species (e.g. 
by fredale and Troughton 1934; Marlow 1962; 
Wakefield 1963; Green 1974; McKay 1988; Tumer 
and McKay 1989; Flannery 1994; Menkhorst 1995; 
Tumer and Ward 1995). Following Desmarest 
(1817), brief descriptions of the species appeared in 



108 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 




'%-i^-: >-^.__ ...-<r„ 




Figure 1. This illustration above of two Phalangista gliriformis {=Cercartetus nanus) appeared in an article 
by Thomas Bell published in the Transactions oftheLinnean Society of London in 1829. The animals appear 
to be quite large due to the disproportionally small tree trunk and branches upon which they are standing. 



Proc. Linn. Soc. N.S.W., 127, 2006 



109 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 





\f'r,ff '' 



a/ 




^ ^ ^ 




, ^mm^ 







Figure 2. Pouch and extremities of Phalangista gliriformis {=Cercartetus nanus) by Bell (1829). 
a. Pouch and teats, shortly after the period of suckling; b. Pouch and teats of the unimpregnat- 
ed animal; c. Prehensile extremity of the tail; d. Fore-foot, upper part; e. Fore-foot, under part; 
f. Hind-foot, upper part; g. Hind-foot, under part; h. Curl of the tail, observed during sleep. 



110 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HAI^RIS 



subsequent zoological publications printed in French 
(Desmarest 1820; Cuvier 1826; Lesson 1827; 1828; 
1830; 1838; Temminck 1827; Fischer 1829), English 
(Cuvier 1827) and German (Schinz 1844) and were 
either taken from the original reference or from the 
specimen which formed the subject of it. 

On 4 November 1828, Thomas Bell read before 
the Lirmean Society of London a description of a 
supposed new species of Phalangista, which he 
named P. gliriformis (Bell 1829). The species name 
was derived from the latin word 'glires' meaning 
'dormouse'. His address was based on close 
examination of two live females which were 'received 
from New Holland' (Australia), but from what part 
was not stated. Bell (1829) detailed a great deal of 
carefiil observation, but he failed to persuasively 
distinguish P. gliriformis from P. nana. According to 
the description, the distinction was proposed because 
of slight differences in the colouring, and principally 
because fur was absent from the ears. Bell's 
coirfidence in the distinction relied on the phrase 'les 
oreilles sont arrondies et couvertes de polls' from 
Temminck's (1827) description of P. nana, which 
quoted Desmarest (1817), and translates as 'the ears 
are round and covered with hair'. Later, Waterhouse 
( 1 84 1 ) stated that ' Temminck should have said that the 
ears are covered with very minute hairs, for so small 
are they that to the naked eye they appear naked' (see 
also Wagner 1 843). The holotype of P. nana contained 
in the Paris Museum, and also the type specimen 
for P. gliriformis, were re-examined by Waterhouse 
(1841) and no specific differences were perceived 
by him (see also Waterhouse 1846; Wagner 1855). 
Despite this taxonomic oversight. Bell's observations 
on living specimens resulted in some very interesting 
notations on the habits of the species and he also 
provided some remarkable illusfrations (reproduced 
as Figs 1 and 2). However, one inaccuracy in Fig. 1 
(lower animal) is the inclusion of a claw on the hallux. 
It should be highlighted that a very similar illustration 
to Fig. 1 appeared in Cobbold (1868), but the hind 
feet were also drawn incorrectly (see reproduction of 
this image and comments in Strahan 1981). 

There is some conftision in the literature regarding 
a statement made by Burmeister (1837) which 
translates as 'a specific genus (Cercaertus Glog.) is 
formed by the common brush tailed Ph. vulpina\ 
It has occasionally been presumed that Cercaertus 
was a mis-spelling or synonym of Cercartetus (e.g 
Simpson 1945; Marlow 1958; Hickman and Hickman 
1960; Sharman 1961; Bartholomew and Hudson 
1962; Grzimek 1975). In fact, the name Cercaertus 
was used in reference to Phalangista vulpina, which 
is an absolute synonym for Trichosurus vulpecula, the 



common brush-tail possum. According to Wakefield 
(1963), the reference was drawn from an unpublished 
manuscript written by Constantin Gloger, but when 
the work was published in May 1841, the name 
Cercaertus was not mentioned. Instead, Gloger ( 1 84 1 ) 
proposed the quite different name PsilogrammUrus 
for P. vulpina, and used Cercartetus for P. nana. 
Cercartetus makes some reference to the tail (from 
the Greek kerkos) but the significance is obscure 
(Strahan 1981). It is not known whether Burmesiter 
(1837) incorrectly cited Gloger (unpublished) or if 
substantial changes were made to the work prior to 
publication. Perhaps due to the conftision, the name 
Cercartetus was at that time basically disregarded for 
P. nana. However, it is clear that the name Cercaertus 
is a junior synonym of Trichosurus and not of 
Cercartetus (Iredale and Troughton 1934; Wakefield 
1963; McKay 1988). 

In a report dated 10 July 1841, and published in 
November of that year, Dr I.E. Gray of the British 
Museum set out a review of locality data on Australian 
mammals wherein he proposed the genus Dromicia 
for P. nana because 'the dentition and the peculiar 
form and character of the tail of this species, at once 
point out that it should constitute a distinct genus from 
the other Phalangers, from which it differs in many of 
its habits' (Gray 1841). This was later accepted by Dr 
G.R. Waterhouse of the British Museum (Waterhouse 
1846), and subsequently the name D. nana was 
widely applied, although the synonym 'Phalangista 
nana' persisted in a small number of articles (e.g. 
Gunn 1852; Gulliver 1875). Cobbold (1868) reported 
that Professor Richard Owen, of the Royal College 
of Surgeons London, disagreed with Gray (1841) 
on the justification of Dromicia. Owen stated that 
'modifications of the teeth are unaccompanied by any 
change of general structure or of habit, whilst those 
teeth which most influence the diet are constant' 
and also that 'these differences of dentition are 
unimportant, and afford no grounds for subgeneric 
distinctions'. However, in this case at least, Owen's 
view did not gather support. 

The species was not found on the Australian 
mainland until Gerrard Rrefft of the Australian 
Museum made a report of a Dromicia found near St. 
Leonards, North Shore, Sydney, New South Wales. 
Krefft (1863) believed it represented a new species 
and described it as D. unicolor, which was a reference 
to its uniform mouse-colour. However, M.R. 
Oldfield Thomas of the British Museum doubted the 
significance of the find, and believed that Krefft's 
Dromicia was probably a D. nana from Tasmania 
which had escaped from captivity (Thomas 1888). 
He argued that apart from Krefft's specimen, the 



Proc. Linn. Soc. N.S.W., 127, 2006 



111 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 



species had never been recorded from the mainland, 
also adding the questionable statement that it is 'to 
be found in the collection of almost every dealer in 
live animals'. Thomas (1888) also remarked that he 
had inspected drawings of the premolars of the D. 
nana held in the Paris Museum, and compared these 
with Bell's D. gliriformis. He concluded they were 
synonymous, which supported Waterhouse's (1841) 
earlier view, although Thomas did not mention 
Waterhouse in relation to this. 

In 1925, Frederic Wood Jones, Professor of 
Anatomy at the University of Adelaide, communicated 
some observations in the Transactions of the Royal 
Society of South Australia on what he believed was a 
new species of Dromicia (Wood Jones 1925). An adult 
male, collected at Millicent in south-eastern South 
Australia, was described as the type of Dromicia 
britta. Certain measurements were provided which 
suggested that his specimen was considerably smaller 
than Krefft's D. unicolor and the average specimens 
of D. nana. For this reason, and also because his 
specimen had a greyer colouration, and shorter tail 
than D. nana. Wood Jones (1925) believed that it 
should be given species status. It is worth noting that 
measurements for two D. nana individuals were also 
presented by Wood- Jones (1925), but it is, apparent 
that these statistics are in error since they represent data 
from more than two animals (see Thomas 1888). This 
inaccuracy may or may not have influenced Iredale 
and Troughton (1934) to reject the proposed specific 
distinction, but britta was nevertheless recognised by 
them at the subspecific level (see below). 

The genus name Dromicia Gray had been applied 
for close to a century when Iredale and Troughton 
* (1934) noted that Cercartetus Gloger antedated 
Dromicia by several months. They advanced the 
name Cercartetus nanus to supersede D. nana, which 
included a change in the ending of the specific name 
from nana to nanus to accord with the gender of the 
new genus (Strahan 1981). Iredale and Troughton 
(1934) then somewhat arbitrarily accepted three sub- 
species: (1) C. nanus nanus for Tasmania, with/! nana 
and P. gliriformis as synonyms; (2) C. nanus britta 
for south-eastern South Australia with D. britta as a 
synonym; and (3) C. nanus unicolor for New South 
Wales and Victoria with D. unicolor as a synonym. 

From the type of C. nanus held in the Paris 
Museum, G.H.H Tate of the American Museum of 
Natural History, had the skull extracted and cleaned 
for study in 1937 (Tate 1945). He examined the 
dentition of this and other specimens in London and 
sought to determine whether the type of gliriformis 
was from mainland Australia (as implied by several 
authors subsequent to Bell 1829, e.g. Gould 1863; 



Forbes-Leith and Lucas 1884) or from Tasmania 
(as accepted by Iredale and Troughton 1934). He 
compared the teeth of nanus (Desmarest 1817), 
gliriformis (Bell 1829), unicolor (Kreffl 1863) and 
britta (Wood Jones 1925), but could not resolve 
the matter with the specimens available to him. 
Nonetheless, he suggested that the subspecies should 
be C. nanus nanus for Tasmania; C. nanus gliriformis 
(=unicolor) for New South Wales and Victoria, and 
C. nanus britta for South Ausfralia, which was at 
variance from Iredale and Troughton (1934). Tate's 
(1945) proposal was not adopted because he failed 
to demonstrate unequivocally that gliriformis was 
from the mainland. However, Iredale and Troughton 
(1934) had not proved that Bell's specimens were 
Tasmanian. 

The next important contribution on the taxonomy 
of C. nanus was a review by Norman Wakefield of 
Monash University, who discussed the distribution, 
habitat and taxonomy of this species and the pygmy- 
possums more broadly (Wakefield 1963). He revised 
the taxonomy insofar as reducing the number of 
subspecies advanced by Iredale and Troughton 
(1934) from three to two, because he believed that 
on the mainland there was only one subspecies, 
which was reasonably uniform and continuous in 
distribution from South Australia through Victoria 
and into New South Wales (see also Le Souef and 
Burrell 1918). That is, Wakefield (1963) accepted C. 
n. unicolor as the mainland subspecies, and made C. 
n. britta an equivalent synonym, while also accepting 
C. n. nanus as the Tasmanian subspecies. However, 
in a subsequent note, Wakefield (1970) questioned 
his own sub-specific assignment, stating that the four 
cranial specimens available to him from Tasmania 
were 'insufficient to demonstrate difference from or 
affinity with' mainland populations. Despite this, the 
arrangement of Wakefield (1963) has been in place 
for more than 40 years (McKay 1988; Turner and 
Ward 1995; van Weenen 2002), and this is despite the 
absence of any review, testing or elaboration upon 
which to substantiate this hypothesis. 

Confiision is even greater in vernacular 
nomenclature. Names included dwarf phalanger 
(Desmarest 1817; Cuvier 1926; 1827), minute 
phalanger (Waterhouse 1838), dwarf cuscus (Gloger 
1841), pigmy phalanger (Waterhouse 1841), Bell's 
Dromicia {Gray 1843; Gerrard 1862), opossum mouse 
(Gunn 1852; Bonwick 1858; Lord and Scott 1924; 
Tate 1945), dusky Dromicia, pygmy opossum (Kreffi 
1864), thick-tailed Dromicia (Krefft 1868; 1871; 
Le Souef 1907), mouse-like phalanger (Cobbold 
1868), common dormouse-phalanger (Thomas 1888; 
Lydekker 1896), dormouse phalanger (Waterhouse 



112 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 



1846; Lucas 1890; Le Souef and Burrell 1926; 
Marlow 1958), common dormouse-opossum (Ogilby 
1892); dormouse possum (Brazenor 1950), pigmy 
opossum (Le Souef and Burrell 1918), pigmy possum 
(Iredale and Troughton 1934; Wakefield 1963) and 
eastern pigmy possum (Ride 1970). A standard name 
finally eventuated when a committee of the Australian 
Mammal Society recommended 'eastern pygmy- 
possum' in 1980 (Strahan 1980). 

Dentition and Morphology 

Desmarest (1817) stated that the teeth, as far as 
it was possible to observe them on this little animal, 
appeared to be arranged like those of phalangers. 
Similarly, Bell (1829) stated that the incisors did 
resemble other species of the genus Phalangista, 
but complained of the difficulty of examining the 
minuscule teeth on living subjects. Owen (1845) 
pointed out that the species 'has only three true molars 
on each side of the jaw', and also that 'the last and 
penultimate premolars on the lower jaw are. shaped 
like canines'. Subsequently, Krefft (1863, 1864) was 
able to provide the following dental formula: 



13-3/1-1 C 1-1/1-1 
Total = 36 



P 3-3/3-3 M 3-3/3-3 



The basic phalangerid dentition is three 

premolars and four molars in each row (Tate 

1945), although Cercartetus is unusual in having 

only three molars in each row, and C. nanus has a 

diagnostic P^ which is large and double-rooted (see 

also Smith 1971; TumbuU and Schram 1973; Green 

and Rainbird 1983; Menkhorst and Knight 2001). 

In terms of morphology, Desmarest (1817) made 

a description fi-om a spirit specimen and briefly noted 

it as the size of a mouse, and with a brown circle 

around the eyes, and imprecisely described the ears 

as short, rounded and 'covered with hair'. As already 

mentioned, it should have been stated that the ears 

appear nearly naked. A more articulate description 

was provided by Bell (1829) who stated that: 

'the general form of this 

animal resembles that of the 

common dormouse; but it is larger, 

broader and more depressed. The 

head is broad across the ears, fi-om 

whence it tapers to the nose, which 

is somewhat pointed. The nostrils 

are narrow, and of a semicircular 

form: the upper jaw, which is 

elongated, overhangs the imder, 

and almost entirely conceals it. 

The lips are scantily covered with 



soft short hair, of a whitish colour, 
and are fiimished with four rows of 
long black vibrissae, the posterior 
ones tipped with light brown. The 
eyes are very large, remarkably 
prominent, and of a jet-black 
colour: the ears of considerable 
size, erect, totally destitute of hair, 
and of a uniform mouse-colour'. 

In terms of colouration, the fiir was first described 
as grey lightly frosted with a reddish tinge and white 
underneath (Desmarest 1817) and more simply 
as upper parts grey, but white underneath (Cuvier 
1826; Lesson 1827; Schinz 1844; Krefft 1871). In 
characteristic detail. Bell (1829) stated that his living 
examples were: 

'covered with a very soft and 
thick fur; the hairs which compose 
it being of a gray colour tipped with 
reddish brown, give the general hue 
of rufous-gray. The under parts are 
more sparingly covered with fur of 
a pale yellowish-gray colour, the 
yellow predominating at the sides, 
and especially at the throat. The 
general colour of the face is also 
yellowish, the upper and back part 
of the head assuming the rufous- 
gray colour of the back'. 

Bell (1829) also noted a blackish ring around 
the eye, and remarked on 'a darkish ring partially 
surrounding the ears at the anterior part, interrupted by 
a distinct white spot behind each (ear)'. Krefft (1863) 
described the fiir as 'a uniform mouse-colour lighter 
on the sides and beneath, with a blackish patch in fi-ont 
of the eye'. Gould (1863) stated that 'considerable 
diversity of colour exists in different individuals; in 
some the upper surface is nearly uniform grey, while 
in others a fine tawny or rufous tint pervades the same 
parts; and examples are constantly met with exhibiting 
every variety of intermediate shade'. Wakefield 
(1963) pointed out that the Tasmanian members of the 
species (C. n. nanus) 'have a warm brown infiision 
in the general body colour and are yellowish on the 
sides and underneath', while the mainland form (C.«. 
unicolor) 'is less brown and less yellow' (see also Le 
Souef and Burrell 1918). 

Early naturalists noted that C. nanus have several 
features in common with other possums, such as the 
prehensile tail and feet specially adapted for climbing. 
They also noticed the incrassated base of the tail, 
and considered this to be a unique and characteristic 



Proc. Linn. Soc. N.S.W., 127, 2006 



113 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 




PLATE XVII. 



"X? 



COMMON DOfi.MOUSE-PHAl.ANGER 



Figure 3: This illustration of the Common Dormouse Phalanger {=Cercartetus nanus) appeared in 
Lydekker's Handbook to the Marsupialia and Monotremata in 1896. 



attribute of this species (Bell 1829; Lesson 1830; 
Gray 1841; Waterhouse 1846; Le Souef and Burrell 
1926). Lydekker (1896) noted the tail as 'rather long 
with the basal inch thickened', but the incrassation 
was not evident in the illustration he provided which 
was originally published in Waterhouse (1841) (Fig. 
3). Le Souef and Burrell (1926) explained that 'when 
captured in summer the tail is not usually incrassated, 
and the animal is slender and mouse-like; but as 
winter approaches it becomes bulkier, the base of 
the tail becomes very swollen, and the appearance 
of the animal is very much changed' (see also Le 
Souef and Burrell 1918). An assessment of the 
female reproductive organs by Bell (1829) revealed 
four teats, and many subsequent naturalists concurred 
with this observation (Lesson 1830; Wagner 1843; 
Giebel 1859; Thomas 1888; Ogilby 1892; Le Souef 
and Burrell 1926; Troughton 1943; Wakefield 1963). 
However, in more recent times Wakefield (1970) 
reported an individual with five nipples, and Turner 
(1981) found that there are actually six teats, four 
developed and two rudimentary. 



Bell (1829) noted that two toes on each of 
the hind feet were 'united together' (Fig. 1). This 
morphological feature (syndactyly) is an adaptation 
for fur cleansing (Ride 1978) and for an arboreal 
lifestyle (Hall 1987). Krefft (1863) noticed that the 
tongue is 'fiimished with a slight brush at the tip', 
and he interpreted this as an adaptation for nectar- 
feeding. Thomas (1888) noticed that there were five 
large pads on each of the palms and soles. There are 
various other minor descriptions of morphological 
features outlined in the early literature, but I have 
only covered those of most significance. 

Distribution and abundance 

In the early years of European settlement of 
Australia it was presumed that the species was 
peculiar to Maria Island and mainland Tasmania 
(Cuvier 1827; Waterhouse 1838; Gray 1841; 1842; 
Gunn 1852; Gould 1845; Waterhouse 1846; Gervais 
1955; Giebel 1859; Cobbold 1868). It is now clear 
that the species also has a broad distribution in the 
coastal regions of south-eastern mainland Australia 



114 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 



(Turner and Ward 1995). In the early years however, 
the specimens which reached the British Natural 
History Museum were mainly Tasmanian (Gray 
1843; Gerrard 1862; Thomas 1888; Wakefield 1963) 
which probably led Gould (1863) to postulate that 
the species was 'abundant ...in Van Dieman's Land 
(=Tasmania), particularly the northern parts of the 
island'. Lord and Scott (1924) also suggested that it 
was more common in northern Tasmania. However, 
by the early 1960s it was considered that the species 
was rare in this State because of 'marked changes 
in vegetation' brought about by periodic forest fires 
(Wakefield 1963). Important early literature records 
for Tasmania include Hobart, Waratah, Launceston, 
Westbury district, and Fury Gorge near Cradle 
Mountain, Cloudy Bay, Mount Wellington (see 
Wakefield 1963), and also Maria, Bruny, Flinders, 
King and Cape Barren Islands (Le Souef 1929; 
Hickman and Hickman 1960; Wakefield 1963; Green 
1969; Green and McGarvie 1971; Whinray 1971; 
Hope 1973). More recent Tasmanian records and a 
comprehensive distribution map are provided by 
Munks et al. (2004). 

While C. nanus was apparently not found on the 
mainland prior to 1 854 (Seebeck 1995), the main credit 
for its discovery on the continent should go to Krefft 
(1863), who collected a specimen at St. Leonards, a 
suburb of Sydney, NSW. However, it is acknowledged 
that Bonwick (1858) had earlier noted that 'opossum 
mice' occurred at Warmambool, Victoria, but no 
specimen was collected. The first collected specimen 
fi^om Victoria appears to have come fi-om Western 
Port in 1880 (Wakefield 1963), and subsequently 
Forbes-Leith and Lucas (1884) accepted the species 
as a component of the Victorian mammalian fauna. 
Other very early Victorian records include specimens 
collected fi-om Gerabrook and Muckleford in 1886, 
and Mordialloc in 1887 (Wakefield 1963). Thomas 
(1888) was evidently unaware of these Victorian 
records when he dismissed Krefft's( 1863) observation 
of the mainland occurrence of the species. 

In 1896, Dr Robert Broom recorded that he 
found a large number of teeth and upper jaws of C. 
nanus in a sub-fossil bone breccia deposit near the 
Wombeyan Caves (Broom 1896). In the same year, 
Professor Baldwin Spencer of the University of 
Melbourne provided details of several specimens 
secured in southern Victoria (Spencer 1896). 
Surprisingly however, its natural occurrence on the 
mainland was still disputed. Waite (1904) provided 
details of a specimen collected at Jindabyne, NSW, 
but was reluctant nonetheless, to declare that the 
species definitely occurred naturally on the continent. 
Hall (1904) finally put the controversy to rest, and 



responded to Waite (1904) with a convincing list 
of reliable mainland records. Further relatively 
early (<1970) locality records for Victoria include 
Heathcote, Blacks Spur, Sale, Avoca, Buanger, 
Portland, Erica, Wilson's Promontory, Mount Lock, 
Tamboon Inlet, Mallacoota, Whitlands, Nowa Nowa, 
Snake Valley, Rushworth Forest, Cape Conran, 
Grenville, Yackandandah and Mount Drummond 
(Harris 2005). A comprehensive review of more recent 
Victorian records is given by Harris and Goldingay 
(2005). 

Early C. nanus records fi-om NSW include those 
fi-om St. Leonards in 1863 and Jindabyne in 1903, 
Fitzroy Falls in 1914, La Perouse prior to 1918, 
Royal National Park in 1925 and Bowral in 1939 
(Le Souef and Burrell 1918; Wakefield 1963). Krefift 
(1864) stated that 'the range of this species probably 
does not extend beyond the east coast districts' but 
qualified this by noting that because it is diminutive 
and nocturnal 'it will be a difficult task to obtain 
many examples, and so define its geographical 
distribution with certainty'. As further information 
became available, Marlow (1958) was able to state 
that its range in NSW was 'between the Hastings 
River and Sydney' and extended west only to the 
Blue Mountains. Subsequently, Wakefield (1963) 
remarked that Newcastle was the northern limit of its 
range. However, a recent review of the distribution 
of C. nanus in NSW (Bowen and Goldingay 2000) 
indicates that its range in NSW extends to Grafton, 
Maclean and Tweed Heads and on the far north NSW 
coast, although most records are from the south coast 
and on the eastern side of the Great Dividing Range. 
A few scattered western records have been identified 
for Pilliga, Coonabarabran, Dubbo, Parkes and 
Molong. The scarcity of recent records in Bowen and 
Goldingay (2000) has led to its current recognition as 
a 'Vulnerable' species in NSW. 

South Australia (SA) and Queensland form 
the western and northern limit, respectively, of the 
distribution of C. nanus. There are only a small 
number of records from each of these States. Wood 
Jones (1925) reported that the first SA specimen was 
discovered at Millicent, and this specimen is now 
held in the collection of the British Natural History 
Museum (Wakefield 1963). Only three specimens 
from this State were acquired by the South Australian 
Museum prior to 1997, and its status was considered 
rare. These records are confined to the far south- 
east of SA. An intensive survey of this region which 
targeted C. nanus in 1997 produced a fiirther 27 
records, and subsequently the status of C. nanus in 
SA was changed to 'Vulnerable' (under Schedule 8 
of the National Parks and Wildlife Act 1972) (van 



Proc. Linn. Soc. N.S.W., 127, 2006 



115 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 



Weenen 2002; Carthew 2004). In Queensland, the 
species was first discovered by Molly O'Reilly in 
Lamington National Park in 1936 (O'Reilly 1941). 
Further examples were later found in the same general 
vicinity (Fleay 1966; Wakefield 1970), but as far as is 
known, the range of C. nanus extends only marginally 
into Queensland, where it is at present paradoxically 
rated as 'Common' (Eyre 2004; Harris et al. in prep). 

Diet and habitat 

Bell's (1829) captive C. nanus (housed in 
London) fed 'on nuts and other similar food'. 
Captive animals are known to accept a range of foods 
including bread, cake, seed, honey, milk, cream, 
biscuits, lollies, finits and insects (Lord and Scott 
1924; Le Souef and Burrell 1926; Troughton 1931; 
Rocking 1939; Conway 1939; Hickman and Hickman 
1960). In the wild, the first feeding observation was 
made by Kreffl (1863) who saw C. nanus 'feeding 
on the blossoms of the Banksiae\ He later wrote 
that 'they live principally on honey and soft insects' 
(Krefft 1 867). Gould (1 863) stated that they feed upon 
the tender buds and spikes of flowers, which Ogilby 
(1892) and Lucas and LeSouef (1909) interpreted 
as meaning that C. nanus was phytophagous. This 
possum is now generally regarded as omnivorous 
(McKay 1988; Menkhorst 1995; Menkhorst and 
Knight 2001), but not herbivorous, and microscopic 
analysis of faeces supports the contention that a range 
of dietary items (particularly pollen and insects) are 
consumed (Huang et al. 1986; Dickman and Happold 
1988; Tulloch 2004). 

As early as 1863 it was recognised that 'of all 
trees it prefers banksias' (Gould 1 863), an observation 
which is supported by modem ecological studies 
(Turner 1985; Ward 1990). Bowen and Goldingay 
(2000) and Harris and Goldingay (2005) also note its 
penchant for Banksia habitat. Early naturalists reported 
that 'they inhabited open wooded country', usually 
among banksias as well as eucalypts, angophora, 
grevilleas, melaleucas and other small flowering 
shrubs (Le Souef and Burrell 1926; Chaffer 1930a,b). 
While it has been recorded from both wet and dry 
sclerophyll forests (Marlow 1958; Green 1973; Harris 
and Goldingay 2005), it has been suggested that dry 
forests are preferred over wet forests (Wakefield 
1963). However, there are both historic and more 
recent evidence that wet forests/rainforest is probably 
favoured habitat on the edges of its range in Tasmania 
(Green 1973; Munks et al. 2004) and in Queensland 
(O'Reilly 1941; Bowen and Goldingay 2000; Harris 
et al. in prep). 

A little information is available fi'om the literature 
about the nesting requirements of C. nanus. Le Souef 



and Burrell (1918) found nests of this species in 
hollow limbs of Eucalyptus squamosa, E. piperata 
and E. haemastoma. Later, these zoologists remarked 
that 'they live in any convenient nook or cranny in a 
tree, but usually in a hollow limb protected fi'om the 
weather, making their nest at an angle. The nest is 
composed of soft bark, which the animals sometimes 
have to travel a considerable distance to procure' 
(Le Souef and Burrell 1926). They also detailed an 
observation that in one case 'it was a quarter of a mile 
(~400m) to the nearest tree on which bark similar to 
that in the nest [of C. nanus] was found'. Nesting 
observations are scant, but those published include the 
discovery of C. nanus nesting in the decaying stumps 
of grass trees Xanthorrhoea spp. (Green 1969), and 
also in deserted bird and bat nests (Chaffer 1930a,b; 
Schulz 2000). Lord and Scott (1924) commented 
that 'Searching for the retreats of these animals is 
a tedious task', and that most sightings are 'fi^om 
bushmen who come across them when felling and 
cutting up trees in the bush'. They also added that 
their habits 'naturally make them difficult to obtain, 
and it is more by accident than design that specimens 
are secured'. 

Behaviour 

Bell ( 1 829) was in possession of living examples, 
and this furnished him with the opportunity to closely 
observe the habits of the species while in confinement. 
He observed that: 

'in their habits they are 
extremely like the dormouse, 
feeding on nuts and other similar 
food, which they hold in their fore 
paws, using them as hands [see 
also Fig. 1]. They are nocturnal, 
remaining asleep during the 
whole of the day, or, if disturbed, 
not easily roused to a state of 
activity; and coming forth late in 
the evening, and then assuming 
their natural rapid and vivacious 
habits. They run about a small tree 
which is placed in their cage, using 
their paws to hold by the branches, 
and assisting themselves by their 
prehensile tail, which is always 
held in readiness to support them, 
especially when in a descending 
attitude. Sometimes the tail is 
thrown in a reversed direction, 
turned over the back; and at other 
times, when the weather is cold, it is 
rolled closely up towards the under 



116 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 



part, and coiled almost between the 
thighs. When eating they sit up on 
their hind quarters, holding the food 
in their fore paws, which, with the 
face, are the only parts apparently 
standing out from the ball of fur, of 
which the body seems at that time 
to be composed. They are perfectly 
harmless and tame, permitting any 
one to hold and caress them without 
ever attempting to bite, but do not 
evince the least attachment either 
to persons about them or even to 
each other'. 

Bell's observations were wrongly attributed to 
John Gould by Waterhouse (1846). However, when 
Gould published his meticulous work Mammals 
of Australia in 1863, he made some very original 
remarks, an extract of which follows: 

'I am sufficiently acquainted 
with the habits and economy of 
the Dromicia gliriformis to state 
that it is a strictly nocturnal animal, 
and that of all trees it prefers 
the Banksias, whose numerous 
blossoms supply it with a never- 
ceasing store of food, both of 
insects and sweets; if I mistake not, 
it also feeds upon the tender buds 
and spikes of the flowers. During 
the day it generally slumbers 
coiled up in some hollow branch 
or fissure in the trees, whence if 
its retreat be discovered it is easily 
taken by the hand; this state of 
inactivity is totally changed at 
night, when it runs over the smaller 
branches and leaps from flower to 
flower with the utmost ease and 
agility. This disposition is just as 
strongly displayed by it when kept 
in confinement; being so drowsy 
during the daytime as to admit of 
its being handled without evincing 
the least anxiety to escape, while 
the contrary is the case as soon 
as night approaches. I have also 
observed that during the months of 
winter it is less active than in the 
summer; undergoing in fact a kind 
of hibernation, somewhat similar, 
but not to the same extent, as the 
Dormouse'. 



Gould provided an illustration of a pair of C. 
nanus (Fig. 4), which at that time were 'alive in 
the possession of Her Most Gracious Majesty at 
Windsor Castle', having been brought to England by 
the Very Reverend the Archdeacon Marriott, and set 
before Queen Victoria (1837-1901) as a gift. Archer 
(1982) later commented that 'anyone who has seen 
one of these utterly charming creatures struggling to 
wake itself up after a deep sleep in torpor will fially 
understand why the Queen insisted that these little 
colonials had to stay with her inside Windsor Castle'. 
Many others have also made complimentary portrayals 
of this little animal, such that it has been described as 
'interesting', 'elegant', 'graceful', 'beautiful', 'cute', 
'harmless', 'tame', and an 'endless source of interest 
and amusement' (Bell 1 829; Lesson 1 830; Waterhouse 
1846; Bonwick 1858; Krefft 1863; Lydekker 1896; 
Lord and Scott 1924; Le Souef and Burrell 1926; 
Flannery 1994). They obviously fared well in Royal 
confinement, evidenced by their corpulence (Fig. 3), 
and Gould (1863) noted that these captive animals were 
'inclined to obesity'. The tendency for individuals to 
over-eat and become fat has also been referred to by 
other authors (Waterhouse 1846; Thomas 1888; Le 
Souef and Burrefl 1918; Conway 1939; Baines 1962; 
Bartholomew and Hudson 1962). 

Early naturalists were quick to liken the 
species to the English dormouse (Bell 1829; Schinz 
1844; Waterhouse 1846; Gervais 1855; Krefift 
1871; Thomas 1888). Befl (1829) explained that the 
superficial resemblance is: 

'shown in their nocturnal 
activity, the nature of their food, their 
manner of taking it, their attitudes 
and motions, no less than in many 
circumstances connected with their 
external form and characters; as, the 
general form of the body, the nature 
of the fur, the character of the feet, 
the prominence and remarkable size 
of the eyes, &c. There is, however, 
one very important peculiarity of the 
dormouse, which has not as yet been 
observed to appertain to our animal, 
and that is its hybernation'. 

However, Befl (1829) was certainly mistaken 
in asserting that C. nanus does not imdergo torpor, 
which is a significant aspect of its behaviour (see 
also Waterhouse 1846; Gould 1863; Le Souef 1907; 
Lord and Scott 1924; Hickman and Hickman 1960; 
Bartholomew and Hudson 1962; Geiser 1993). An 
amazing story was told by A.H.E Mattingley of a 



Proc. Linn. Soc. N.S.W., 127, 2006 



117 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 




Figure 4. This charming illustration of a pair of Dromicia gUriformis (=Cercartetus nanus) ap- 
peared in Gould's (1863) Mammals of Australia. These animals that were at that time in the 
possession of the Queen of England, at Windsor Castle, and subject to the excesses of roy- 
al life, became quite obese. The signature shows that it was drawn by Gould and H.C. Richter. 



118 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 



dormant one found while felling a dead tree in the 
Goulbum valley, Victoria (in Le Souef and Burrell 
1926; Troughton 1943). To try to rouse it he 'hung it 
to a twig by its prehensile tail, but it grasped the fur 
of its abdomen with its paws and remained hanging 
and dormant, its tail automatically suspending it'. It 
apparently stayed in this position 'for several hours 
without attempting to seek a different pose'. Le Souef 
and Burrell (1926) further remarked that C. nanus: 
'are the most harmless little 
creatures, quiet in disposition, 
rather slow in movement, and quite 
defenceless. They spend the day 
coiled up in their nests, coming out to 
feed at night. Then they become alert, 
running and jumping from limb to 
limb, making use of their prehensile 
tail, especially when descending from 
one branch to another'. 



CONCLUSION 

The history of European knowledge of C. nanus 
starts with its collection from Maria Island more 
than 200 years ago. The subsequent accounts of its 
biology and of its classification were made by some 
of the best-known professional zoologists of the 19* 
Century such as Desmarest, Gould, Krefft, Thomas 
and Waterhouse. However, important contributions 
on this possum were also made by lesser known 
researchers, naturalists and bushmen, including 
Bonwick, Hall, Le Souef, Mattingley and Waite. 
The early records and narratives are of historical 
importance and add appreciably to our knowledge of 
this species. 



ACKNOWLEDGEMENTS 

It is a pleasure to acknowledge the assistance of 
a number of people and organisations in preparing this 
historical note. The staff of the State Library of New 
South Wales helped with the decryption of abbreviated 
citations in Thomas (1888). Steven Smith contacted Mme 
Bonnemain of the Le Havre Museum in France and she 
forwarded a translation of the relevant parts from Peron's 
unpublished manuscript. Henri Jeanjean helped me draft a 
letter to Professor Michel Tranier of the Museum National 
d'Histoire Naturelle in Paris who checked on the C. nanus 
holotype held in the Museum's collection and sent a copy 
of the mammal catalogue. Several of the early natural 
history texts I required were not held or otherwise easily 
accessed within Australia, but the relevant pages were 
sent to me courtesy of the American Museum of Natural 



History, British Library, Library of Congress, Smithsonian 
Institution, University of Glasgow, University of Southern 
California Library and Wellcome Institute for the History of 
Medicine. I am indebted to Robyn Williams for translating 
French articles and Benjamin Teeuwsen for translating 
German. I also acknowledge the CSIRO Black Mountain 
Library for permission to reproduce Fig. 1, the Australian 
Museum Research Library for Fig. 2, Museum Victoria 
for Fig. 3 and the Queensland Museum Library for Fig. 
4. Finally, I would like to thank Ronald Strahan, Ross 
Goldingay and Mike Augee for helpful advice on earlier 
drafts of this report. 



REFERENCES 

Anderson, S. (2001). French anthropology in 

Australia, the first fieldwork: Fran9ois Peron's 
Maria Island - anthropological observations. 
Aboriginal History 25, 228-242. 

Animals and Plants Protection Board (1969). 'Maria 
Island'. (Animals and Plants Protection Board: 
Hobart). 

Archer, M. (1982). Possums. Pp. 32-35. In 

'Mammals in Australia'. (Australian Museum: 
Sydney). 

Baines, J. A. (1962). Fauna Survey Group - March 1, 
1962. Victorian Naturalist IS, 367-368. 

Bartholomew, G. A. and Hudson, J. W. (1962). 
Hibernation, aestivation, temperature 
regulation, evaporative water loss and heart 
rate of the pygmy possum, Cercaertus nanus. 
Physiological Zoology 35, 94-107. 

Beaufort, F. de (1966). Catalogue des types des 
mammiferes du Museum National d'Histoire 
Naturelle, Paris. VI Monotremata. VII 
Marsupialia. Bulletin du Museum National 
d'Histoire Naturelle 38, 509-553. 

Bell, T. (1829). Description of a new species of 
Phalangista. Transactions of the Linnean 
Society of London 16, 121-128. 

Bladon, R.V, Dickman, C.R. and Hume, I.D. 
(2002). Effects of habitat fragmentation 
on the demography, movements and social 
organisation of the eastern pygmy possum 
(Cercartetus nanus) in northern New South 
Wales. Wildlife Research 29, 105-116. 

Bocking, J.M. (1939). The story of Twinkle. 
Victorian Naturalist 56, 134-135. 

Bonwick, J. (1858). 'Western Victoria: Its 

Geography, Geology and Social Condition. 
The Narrative of an Educational Tour in 1857'. 
Republished in 1970, with an introduction 
and editorial commentary by C.E. Sayers. 
(Heinemann: Melbourne), p. 66. 



Proc. Linn. Soc. N.S.W., 127, 2006 



119 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 



Bowen, M. and Goldingay, R. (2000). Distribution 
and status of the eastern pygmy possum 
{Cercartetus nanus) in New South Wales. 
Australian Mammalogy 2\, 153-164. 

Brazenor, C.W. (1950). 'The mammals of Victoria'. 
(National Museum of Victoria: Melbourne). 

Broom, R. (1896). On a small fossil marsupial with 
large grooved premolars. Proceedings of the 
Linnean Society of New South Wales 10, 563- 
567. 

Burmeister, C.H.C. (1837). Phalangista. Handbuch 
der Naturgeschichte 2, 814. 

Carthew, S.M. (2004). Distribution and conservation 
status of possums and gliders in South Australia. 
Pp. 63-70 in 'The Biology of Australian 
Possums and Gliders' (Eds R.L Goldingay 
and S.M Jackson). (Surrey Beatty and Sons: 
Sydney). 

Chaffer, N. (1930a). The opossum mouse (Dromicia 
nana). Australian Zoologist 6, 109. 

Chaffer, N. (1930b). The opossum mouse. Victorian 
Naturalist 41 , 18-19. 

Cobbold, T.S. (1868). Mammalia. In Richardson, 
J., Dallas, W.S., Cobbold, T.S., Baird, W. and 
White, A. 'The Museum of natural history, 
being a popular account of the structure, habits, 
and classification of the various departments 
of the animal kingdom : quadrupeds, birds, 
reptiles, fishes, shells and insects, including the 
insects destructive to agriculture'. 2nd edition. 
Glasgow ; London ; Edinburgh : William 
MacKenzie. p. 212 and plate 30 (figure 94). 

Conway, K. (1939). The life of Bluey. Victorian 
Naturalist 56, 133-4. 

Cornell, C. (1965). 'Questions relating to Nicolas 
Baudin's Australian expedition, 1800-1804'. 
(Libraries Board of South Australia: Adelaide). 

Cuvier, F. (1826). Le phalanger nain. In 

'Dictionnaire des Sciences naturelles, dans 
lequel on traite methodiquement des differens 
Etres de la Nature' 39, 415. 

Cuvier, G.B. (1827). Synopsis of the species 
of the class Mammalia as arranged with 
reference to their organization with specific 
characters, synonyma etc etc. In Cuvier, B. and 
Griffith, E. 'The animal kingdom, arranged 
in conformity with its organization with 
additional descriptions of all the species hitherto 
named, and of many not before noticed'. (G.B. 
Whittaker: London). 5, 198. 

Desmarest, A.G. (1817). Nouveau Dictionairie 
d'Histoire Naturelle. Deterville, Tome 25, 477. 

Desmarest, A.G. (1820). Mammalogie ou description 
des especes de mammiferes. Encyclopedic 



Methodique Histoire Naturelle. Volume 1 . 
(Mme Veuve Agasse: Paris). 1, 268. 

Dickman, C.R. and Happold, D.C.D., (1988). The 
eastern pygmy-possum, Cercartetus nanus 
(Marsupialia: Burramyidae), in the Australian 
Capital Territory. Australian Mammalogy 11, 
77-79. 

Eyre, T.J. (2004). Distribution and conservation 
status of the possums and gliders of southern 
Queensland. In 'The Biology of Australian 
Possums and Gliders' (Eds R.L Goldingay and 
S.M Jackson), pp. 1-25. (Surrey Beatty and 
Sons: Sydney). 

Faivre, J.P (1953). 'L' expansion Fran9aise dans le 
Pacifique de 1800 a 1842'. (Nouvelles editions 
latines: Paris). 

Fischer, J.B. (1829). Synopsis mammalium. 
Stuttgardtiae : Sumtibus J.G. Cottae. p. 276. 

Fleay, D. (1966). David Fleay's Nature Notes. A 
'needle' out of Qld's haystack. The Courier 
Mail, Brisbane, 18 January 1966, p. 8. 

Flarmery, T. (1994). Eastern Pygmy-Possum, pp. 
54-55. In 'Possums of the world: a monograph 
of the Phalangroidea'. (GEO Productions: 
Chatswood). 

Forbes-Leith, TA. and Lucas, A.H. (1884). 

Catalogue of the fauna of Victoria. Vertebrates: 
Mammalia. Victorian Naturalist 1, 4-6. 

Geiser, F. (1993). Hibernation in the eastern pygmy 
possum, Cercartetus nanus (Marsupialia: 
Burramyidae). Austraian Journal of Zoology 41, 
67-75. 

Gerrard, E. (1862). Dromicia. In 'Catalogue of 
the bones of Mammalia in the collection of 
the British museum'. (Printed by order of the 
Trustees: London), p. 120. 

Gervais, P. (1855). Tribu des Phalangistins. In 
'Histoire naturelle des mammiferes avec 
r indication de leurs moeurs, et de leurs rapports 
avec les arts, le commerce et 1' agriculture'. (L. 
Curmer: Paris). 2, 275-276. 

Giebel, C. (1859). Die Saugethiere in zoologischer, 
anatomischer und palsontologischer Beziehung 
umfassend dargestellt von C.G. Giebel. Leipzig, 
A. Abel. p. 699-700. 

Gloger, C.W.L. (1841). Cercartetus. In 

Gemeinntitziges Hand-und Hilfsbuch der 
Naturgeschichte. Schulz and Co. p. 85. 

Gould, J. (1845). Dromicia concinna. Proceedings of 
the Zoological Society of London 13, 2. 

Gould, J. (1863). 'The mammals of Australia': 
incorporating the 3 original volumes with 
modem notes by Joan M. Dixon. (Macmillan 
1977: South Melbourne). 



120 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 



Gray, J.E. (1841). Contributions towards the 
geographical description of the Mammalia 
of Australia, with notes on some recently 
discovered species. Appendix in Grey, G. 
'Journal of two expeditions of discovery in 
northwest and Western Australia during the 
years 1837, 38 and 39'. (T. and W. Boone, 
London), p. 397-414. 

Gray, J.E. (1842). Catalogue of Australian 

Mammalia, with descriptions of several new 
species. Tasmanian Journal of Natural Science, 
Agriculture, Statistics 1, 382-385. 

Gray, J.E. (1843). List of the specimens of 
Mammalia in the collection of the British 
Museum. (Printed by order of the Trustees: 
London), p.85. 

Green, R.H. (1969). Birds of Flinders Island with 
references to other eastern Bass Strait Islands 
and aimotated lists of other vertebrate fauna. 
Records of the Queen Victoria Museum 34, 1- 
32. 

Green, R.H. (1973). 'The mammals of Tasmania'. 
(Foot and Playsted: Launceston). 

Green, R.H. (1974). Mammals. In 'Biogeography 
and ecology in Tasmania'. Ed by W.D. Dr W. 
Junk Publishers, The Hague, pp. 367-96. 

Green, R.H. and McGarvie, A.M. (1971). The birds 
of King Island, with references to other western 
Bass Strait Islands and aimotated lists of the 
vertebrate fauna. Records of the Queen Victoria 
Museum 40, 1-42. 

Green, R.H. and Rainbird, J.L. (1983). An illustrated 
key to the skulls of mammals in Tasmania. 
Laimceston. Queen Victoria Museum. 

Grzimek, B. (1975). Pygmy Possums p. 114. In 
'Grzimek's Animal Life Encyclopedia. Volume 
10. Mammals I. Van Nostrand Reinhold: 
Melbourne. 

Gulliver, G. (1875). Observations on the sizes and 
shapes of the red corpuscles of the blood of 
vertebrates, with drawings of them to a uniform 
scale, and extended and revised tables of 
measurements. Proceedings of the Zoological 
Society of London 1875, 474-495. 

Guim, R.C. (1852). A list of the mammals 

indigenous to Tasmania. Proceedings of the 
Royal Society of Tasmania 1851, 77-90. 

Hall, L.S. (1987). Syndactyly m marsupials. In 

'Possums and Opossums: studies in evolution'. 

(Ed. M. Archer) (Surrey Beatty and Sons, 

Chipping Norton). 1, 245-255. 
Hall, T.S. (1904). The genus Dromicia on the 

Australian mainland. Victorian Naturalist 20, 



176. 

Harris, J.M (2005). Aimotated records of the eastern 
pygmy-possum Cercartetus nanus from the 
Victorian Naturalist 1884-2004. Victorian 
Naturalist 122, 146-150. 

Harris, J.M and Goldingay, R.L. (2005). 

Distribution, habitat and conservation status of 
the eastern pygmy-possum Cercartetus nanus in 
Victoria. Australian Mammalogy 11, 185-210. 

Harris, J.M., Eyre, T.J., Goldingay, R.L. and 
Gynther, I.C. (in prep). Status of the eastern 
pygmy-possum Cercartetus nanus in 
Queensland. Memoirs of the Queensland 
Museum. 

Hickman, VV. and Hickman, J.L. (1960). Notes 
on the habits of the Tasmanian dormouse 
phalangers Cercaertus nanus (Desmarest) and 
Eudromicia lepida (Thomas). Proceedings of 
the Zoological Society of London 135, 365-374. 

Hope, J.H. (1973). Mammals of the Bass Strait 
islands. Proceedings of the Royal Society of 
Victoria S5, 163-95. 

Homer, F.B. (1987). 'The French reconnaissance: 
Baudm in Australia 1801-1803'. (Melbourne 
University Press: Carlton). 

Huang, C, Ward, S., and Lee, A.K. (1986). 

Comparison of the diets of the feathertail glider, 
Acrobates pygmaeus, and the eastern pygmy- 
possum, Cercartetus nanus (Marsupialia: 
Burramyidae) in sympatry. Australian 
Mammalogy 10, 47-50. 

Hunt, S. (1999). 'Terre Napoleon: Australia through 
French eyes, 1800-1804'. (Historic Houses 
Trust of New South Wales in association with 
Hordem House: Sydney). 

Iredale, T. and Troughton, E.L.G (1934). A check- 
list of the mammals recorded from Australia. 
Australian Museum Memoir 6, 1-122. 

JuHen-Laferriere, D. (1994). 'Catalogue des types 
de mammiferes du Museum National d'Histoire 
Naturelle. Order des Marsupiaux. Extrait de 
Mammalia'. Tome 58. 

Krefft, G. (1863). Description of a new species 
of the genus Dromicia discovered in the 
neighbourhood of Sydney. Proceedings of the 
Zoological Society of London 1863, 49-50. 

Krefft, G. (1864). Dromicia in 'Catalogue of the 
Mammalia in the collection of the Australian 
Museum'. (Government Printer: Sydney), pp. 
42-44. 

Krefft, G. (1867). Mammalia. In 'Australian 
vertebrata, recent and fossil, representing all 
the genera known up to the present time'. 
(Government Printer: Sydney), p. 8. 



Proc. Linn. Soc. N.S.W., 127, 2006 



121 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 



Kreffl, G. (1868). 'Notes on the fauna of Tasmania'. 

(F.White Printer: Sydney), pp. 4-5. 
Kreffi, G. (1871). 'The mammals of Australia, 

with a short account of all the species hitherto 

described'. (Government Printer, Sydney), p. 3. 
Le Souef, A.S. (1929). Notes on some mammals 

from Bass Strait Islands. Australian Zoologist 5, 

329-332. 
Le Souef, A.S. and Burrell, H. (1918). Notes on 

some of the smaller marsupials of the genera 

Phascogale, Sminthopsis, Acrobates and 

Dromicia. Australian Zoologist 1, 147-152. 
Le Souef, A.S. and Burrell, H. (1926). Dormouse- 

Phalangers. In 'The wild animals of 

Australiasia'. (George Harrap and Company: 

Sydney), pp. 244-248. 
Le Souef, W.H. (1907). 'Wild Life in Australia'. 

(Whitcombe and Tombs: Christchurch). p. 11. 
Lesson, R.P. (1827). Phalanger nain, Phalangista 

nana. In 'Manuel de Mammalogie, ou histoire 

naturelle des mammiferes'. (Roret: Paris), pp. 

218-219. 
Lesson, R.P. (1828). Phalanger nain, Phalangista 

nana. In 'Dictiormaire Classique d'Histoire 

Naturelle'. (Eds J.V. Audouin and J.B.G.M. 

Bory de Saint- Vincent) 13, 334. 
Lesson, R.P. (1830). Histoire naturelle generale et 

particuliere des mammiferes et des oiseaux 

decoverts depuis 1788 jusqu'a nos jours. 

Volume IV, Suite des mammiferes. (Baudouin 

Freres: Paris), pp. 466-471. 
Lesson, R.P (1838). 'Complement de Buffon, 

races humaines et mammiferes', 2"'' edition. (P. 

Pourrat Freres: Paris). 1, 447-448. 
Lesson, R.P. (1842). Famille Phalangistae. 

In 'Nouveau Tableau du Regne Animal. 

Mammiferes'. (A. Bertrand: Paris), p. 188. 
Lord, C.E. and Scott, H.H. (1924). 'A synopsis of 

the vertebrate animals of Tasmania'. (Oldham, 

Beddome & Meredith: Hobart). 
Lucas, A.H.S. (1890). Zoology. Vertebrata. 

In 'Handbook of Melbourne'. (Ed. WB. 

Spencer) pp. 61-62. (Spectator Publishing Co: 

Melbourne). 
Lucas, A.H.S. (1897). On some facts in the 

geographical distribution of the land and fresh- 
water vertebrates in Victoria. Proceedings of the 

Royal Society of Victoria 9, 34-53. 
Lucas, A.H.S. and Le Souef, W.H.D. (1909). 'The 

animals of Australia: Mammals, Reptiles and 

Amphibians'. (Whitcombe and Tombs Ltd: 

Melbourne). 
Lydekker, R. (1896). The Dormouse Phalangers 

Genus Dromicia. In 'A hand-book to the 



Marsupialia and Monotremata'. (Llyods Natural 
History: London), pp. 111-116. 

Marlow, B.J. (1958). A survey of the marsupials of 
New South Wales. CSIRO Wildlife Research 3, 
71-114. 

Marlow, B.J. (1962). 'Marsupials of Australia'. 
(Jacaranda Press: Brisbane). 

McKay, G.M. (1988). Burramyidae. In 'Zoological 
Catalogue of Australia 5. Mammalia'. (Eds J.L. 
Bannister, J.H. Calaby, L.J. Dawson, J.K. Ling, 
J.A. Mahoney, G.M. McKay, B.J Richardson, 
W.D.L. Ride and D. W. Walton) pp. 98-102. 
(Australian Government Publishing Service: 
Canberra). 

Menkhorst, P.W. (1995). Eastern Pygmy-possum. In 
'Mammals of Victoria distribution, ecology and 
conservation'. (Ed. P.W Menkhorst) pp. 101- 
102. (Oxford University Press: Melbourne). 

Menkhorst, PW and Knight, R (2001). 'Afield 
guide to the mammals of Australia'. (Oxford 
University Press: Melbourne). 

Milne-Edwards, C.R. (1877). Note sur quelques 

Mammiferes noveaux provenant de la Nouvelle- 
Guinee (^'Dromicia caudata"). Comptes Rendus 
Hebdomadaires des Seances, Academic des 
Sciences (Hebd. Seanc. Acad. Sci.) Paris. 55, 
1079-1081. 

Munks, S.A., Mooney, N., Pemberton, D. and Gales, 
R. (2004). An update on the distribution and 
status of possums and gliders in Tasmania, 
including off-shore islands. In 'The Biology 
of Australian Possums and Gliders' (Eds R.L 
Goldingay and S.M Jackson) pp. 111-129. 
(Surrey Beatty and Sons: Sydney). 

Ogilby, J.D. (1892). 'Dromicia'. In 'Catalogue of 
Australian mammals, with introductory notes 
on general mammalogy'. Australian Museum 
Catalogue 16, 35-36. 

O'Reilly, B. (1941). 'Green Mountains', p. 21. (WR. 
Smith and Patterson: Brisbane). 

Owen, R. (1845). 'Odontography; or a treatise on 
the comparative anatomy of the teeth; their 
physiological relations, mode of development, 
and microscopic structure, in the vertebrate 
animals'. Volume 1 (text) p. 383 and Volume 2 
(atlas) plate 100, figure 3 (skull). 

Peron, M.F. (1809). 'A voyage of discovery to the 
southern hemisphere, performed by order of 
the Emperor Napoleon during the years 1801, 
1802, 1803 and 1804'. (Translation published in 
1975 by Marsh Walsh Publishing: Melbourne), 
p. 233. 



122 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.M. HARRIS 



Plomley, N.J.B. (1983). 'The Baudin expedition 
and the Tasmanian Aborigines, 1802'. (Blubber 
Head Press: Hobart). 

Plomley, B., Cornell, C. and Banks, M. (1990). 
Franfois Peron's natural history of Maria 
Island, Tasmania. Records of the Queen Victoria 
Museum 99, 1-50. 

Ride, W.D.L. (1970). 'A guide to the native 
mammals of Australia'. (Oxford University 
Press: Melbourne). 

Ride, W.D.L. (1978). An historical introduction to 
studies on the evolution and phylogeny of the 
macropodidae. Australian Mammalogy 2, 1-14. 

Rounsevell, D.E. (1989). Managing offshore island 
reserves for nature conservation in Tasmania. 
In 'Australian and New Zealand islands: Nature 
conservation values and management' (Ed. 
A. Burbridge) pp. 157-161. Department of 
Conservation and Land Management, Perth. 

Ryan, L. (1981). 'The Aboriginal Tasmanians'. (St 
Lucia: Queensland). 

Schinz, H.R. (1844). Systematisches verzeichniss 
aller bis jetzt bekannten saugethiere; oder. 
Synopsis mammalium nach dem Cuvier'schen 
system. Solothum, lent und Gassmann. p. 530. 

Schulz, M. (2000). Roosts used by the golden- 
tipped bat Kerivoula papuensis (Chiroptera: 
Verpertilionidae). Journal of Zoology (London) 
250, 467-478. 

Seebeck, J.H. (1995). Terrestrial mammals in 

Victoria - a history of discovery. Proceedings of 
the Royal Society of Victoria 107, 11 -23 . 

Sharman, G.B. (1961). The mitotic chromosomes of 
marsupials and their bearing on taxonomy and 
phylogeny. Australian Journal of Zoology 9, 
38-60. 

Simpson, G.G. (1945). The principles of 

classification and a classification of mammals. 
Bulletin American Museum Naturalist History 
85, 1-350. 

Smith, M.J. (1971) Small fossil vertebrates fi-om 
Victoria Cave, Naracoorte, South Australia. I. 
Potoroinae (Macropodidae), Petauridae and 
Burramyidae (Marsupialia). Transactions of the 
Royal Society of South Australia 95, 185-198. 

Spencer, B. (1896). 'Report on the work of the 

Horn Scientific Expedition to Central Australia. 
Part 1 - Introduction, Narrative, Summary and 
Results'. Supplement to Zoological Report, 
Map. (Dulau and Co: London), p. 184. 

Strahan. R. (ed.) (1980). Recommended common 
names of Australian mammals. Australian 
Mammal Society Bulletin 6, 13-23. 



Strahan, R. (1981). 'A dictionary of Australian 

mammal names: Pronunciation, derivation, and 
significance of the names, with bibliographical 
notes'. (Angus and Robertson: Sydney). 

Strahan, R. (ed.) (1995). 'The mammals of 
Australia'. (Reed Books: Chatswood). 

Tate, G.H.H. (1945). Results of the Archbold 
Expeditions. No. 55. Notes on the squirrel- 
like and mouse-like possums (Marsupialia). 
American Museum Novitates 1305, 1-12. 

Tenmiinck, C.J. (1827). Phalangista nana. In 

'Monographies de Mammalogie ou description 
de quelques genres de mammiferes, don't les 
especes ont ete observees dans les differens 
musees de I'Europe'. (G. Dufour et E. 
D'Ocagne: Paris). 1: 9 

Thomas, O. (1888). Dromicia. In 'Catalogue of the 
Marsupialia and Monotremata in the collections 
of the British Museum' (Natural History), 
London. Pp. 140-148. 

Triebel, L.A. (1948). Peron in Tasmania. Papers and 
Proceedings of the Royal Society of Tasmania 
1947, 63-68. 

Troughton, E.L.G. (1931). Habits and food of some 
Australian mammals. Australian Zoologist 7, 
77-83. 

Troughton, E. (1943). Furred animals of Australia. 
Second Edition. (Angus and Robertson: 
Sydney) 

Tulloch, A. (2004). The importance of food and 
shelter for habitat use and conservation of the 
burramyids in Australia. In 'The Biology of 
Australian Possums and Gliders' (Eds R.L 
Goldingay and S.M Jackson) pp. 268-84. 
(Surrey Beatty and Sons: Sydney). 

Tumbull, W.D. and Schram, F.R. (1973). Broom 
Cave Cercartetus, with observations on pygmy 
possum dental morphology, variation, and 
taxonomy. Records of the Australian Museum 
28, 437-64. 

Turner, V. (1981). Aspects of the ecology of the 
eastern pygmy-possum Cercartetus nanus. 
Australian Mammal Society Bulletin 7, 62. 

Turner, V. (1984). Banksia pollen as a protein 
source in the diet of the Australian marsupials 
Cercartetus nanus and Tarsipes rostratus. Oikos 
43,53-61. 

Turner, V. (1985). The ecology of the eastern 
pygmy-possum, Cercartetus nanus, and its 
association with Banksia. PhD thesis, Monash 
University. 

Turner, V. and McKay, G.M. (1989). Burramyidae. 
In 'Fauna of Australia. Vol. IB Mammalia'. 
(Eds D.W. Walton and B.J. Richardson) pp. 



Proc. Linn. Soc. N.S.W., 127, 2006 



123 



EARLY NATURAL HISTORY OF CERCARTETUS NANUS 



652-664. (Australian Government Publishing 
Service: Canberra). 

Turner, V. and Ward, S.J. (1995). Eastern pygmy- 
possum Cercartetus nanus. In 'The Mammals 
of Australia' (ed. R. Strahan). pp. 217-18. (Reed 
Books: Chatswood). 

van Weenen, J. (2002). Distribution and status of 
the eastern pygmy possum Cercartetus nanus 
unicolor (Marsupialia: Biirramyidae) in South 
Australia. (Nature Conservation Society of 
South Australia: Adelaide). 

Wagner, J. A. (1843). 'Die Saugethiere, 
in Abbildungen nach der Natur, mit 
Beschreibungen. Fortgesetzt von A. Goldfiiss 
Supplementband von J.A. Wagner'. (Ed. J.C.D. 
von Schreber). Suppl. 3 hft 109-110. Erlangen : 
Voss. pp. 82-83. 

Wagner, J.A. (1855). ' Schreber 's die Saugethiere, 
in Addildungen nach der Natur, mit 
Beschreibungen. Fortgesetzt von A. Goldfiiss. 
Supplementband von J.A. Wagner'. Suppl. 5, 
276-278. 

Waite, E.R. (1904). The genus Dromicia in New 
South Wales. Records of the Australian Museum 
5, 134. 

Wakefield, N.A. (1963). The Australian pigmy 
possums. Victorian Naturalist 80, 99-1 16. 

Wakefield, N.A. (1970). Notes on Australian 
pigmy possums {Cercartetus, Phalangeridae, 
Marsupialia). Victorian Naturalist %1 , 11-18. 

Wakefield, N.A. and Wameke, R.M. (1963). Some 
revision in Antechinus (Marsupialia) - 1. 
Victorian Naturalist %{i, 194-219. 

Wallace, C. (1984). 'The lost Australia of Fran9ois 
Peron' (London). 

Ward, S.J. (1990). Life history of the eastern pygmy- 
possum, Cercartetus nanus (Burramyidae: 
Marsupialia), in south-eastern Australia. 
Australian Journal of Zoology 38, 287-304. 

Waterhouse, G.R (1838). Minute Phalanger. In 
'Catalogue of the Mammalia preserved in the 
Museum of the Zoological Society of London'. 
(Richard and John E. Taylor: London). (2nd 
edition) p. 68. 

Waterhouse, G.R. (1841). The natural history of 
Marsupialia or pouched animals. In 'The 
Naturalist's Library. Mammalia'. (Ed. W. 
Jardine). (WH. Lizars & H.G. Bohn: Edinburgh 
& London). 11, 279-282. 

Waterhouse, G.R. (1846). 'A natural history of 
the Mammalia. Vol. 1 . Containing the Order 
Marsupialia, or pouched animals'. (Bailliere: 
London), pp. 307-317. 



Whinray, J.S. (1971). The present distribution of 
some mammals in the Fumeaux Group, Bass 
Strait. Victorian Naturalist 88, 279-286. 

Wood Jones, F. (1925). A new South Australian 
dormouse opossum. Transactions of the Royal 
Society of South Australia 49, 96-98.. 



124 



Proc. Linn. Soc. N.S.W., 127, 2006 



Additions to Knowledge of the Early Pleistocene Wallaby, 
Baringa nelsonensis Flannery and Hann 1984 (Marsupialia: 

Macropodinae) 

K.J. PiPER^ AND N. Herrmann^ 

^School of Geosciences, PO Box 28E, Monash University, Clayton, Victoria 3800, Australia 
^Geological Museum, University of Copenhagen, 0stervoldgade 5-7, DK-1350, Copenhagen K, Denmark 

Piper, K.J. and Herrmann, N. (2006). Additions to Knowledge of the Early Pleistocene Wallaby, Baringa 
nelsonensis Flannery and Hann 1 984 (Marsupialia: Macropodinae). Proceedings of the Linnean Society of 
New South Wales 111, 125-131. 

Following the recovery of more specimens of the extinct wallaby, Baringa nelsonensis, from early 
Pleistocene deposits at Nelson Bay, near Portland, Victoria, dental elements that were previously unknown, 
or only tentatively associated with Baringa at the time of its establishment, are described here. Specimens 
from the early Pliocene Curramulka Local Fauna, Yorke Peninsula, South Australia, previously allied 
with Baringa, are re-examined, and it is concluded that they do not belong to this genus. Baringa is an 
intermediate browser-grazer, but the relatively enlarged I' and characteristic vertical wear facet on I, 
suggest an unusual feeding specialisation. 

Manuscript received 24 January 2005, accepted for publication 21 September 2005. 

KEYWORDS: Baringa, Curramulka, early Pleistocene, Macropodinae, Nelson Bay, Victoria, wallaby. 



INTRODUCTION 

Baringa nelsonensis is a small to medium-sized 
macropodine first described by Flannery and Harm 
(1984) from the early Pleistocene Nelson Bay Local 
Fauna (LF), Portland, Victoria (Hann 1983). It is 
the most abundant species in the fauna, accounting 
for approximately 30% of all specimens. Further 
collection and study oi Baringa material from Nelson 
Bay was reported by Herrmarm (2000), who described 
much of the new material, including teeth previously 
unknown at the time of the original description of 
Baringa. Collecting is still being carried out at Nelson 
Bay, and current research on the fauna by one of us 
(K. P.) has produced more dental specimens referable 
to B. nelsonensis. 

Upper incisors and premolars are often highly 
diagnostic of genera within the Macropodidae. 
This paper describes elements of the incisor and 
premolar dentition previously unknown for the 
genus. In addition, following the discovery of an 
upper deciduous premolar (dP-^) in association with 
undoubted Baringa upper molars, the single dP^ 
specimen (NMV PI 73573) referred to Baringa by 
Flarmery and Harm (1984) is no longer considered 
correctly assigned. Other features of the dentition are 
also discussed based on the more complete material 



now available. 

In addition to the Nelson Bay specimens, 
specimens tentatively aligned with Baringa (cf 
Baringa sp., cf. Baringa nelsonensis) have been 
reported from the Curramulka Local Fauna (Pledge 
1992). The affinities of these specimens have been 
re-examined in the light of the new, mof e complete 
topotypic material. 

All Baringa specimens described here are 
registered in the palaeontology collection of Museum 
Victoria (NMV P). A full list of specimens examined 
is given in the Appendix. The Curramulka Local 
Fauna specimens are registered in the palaeontology 
collection of the South Australian Museum (SAM P), 
a list of which is given in Pledge (1992). 

Classification within the Macropodidae follows 
Kear and Cooke (2001) and dental terminology 
follows Luckett (1993). All measurements are in 
millimetres. 

SYSTEMATICS 

Order: Diprotodontia Owen, 1866 

Family: Macropodidae Gray, 1821 

Subfamily: Macropodinae Gray, 1821 

Tribe Macropodini Flarmery, 1989 

Baringa nelsonensis Flannery and Harm, 1984 



AN EARLY PLEISTOCENE WALLABY 



Description 

Deciduous Premolars: 

Eight dP2S are known, three of which are 
certainly associated with Baringa nelsonensis 
molars (Fig Ij). They consist of a simple blade with 
a prominent anterior cuspid, posterior cuspid, and a 
single intermediate cuspule and associated ridgelet, 
all of which are approximately sub-equal in height. 
The anterior cuspid is occasionally slightly lower 
and is separated from the intermediate cuspule by a 
deep groove. The main blade terminates anteriorly 
in a small, low, rounded cuspule. A small, lower 
posterolingual cuspid is also present, separated 
from the posterior cuspid by a shallow groove. A 
second smaller cuspule is present posterior to the 
posterolingual cuspid in NMV P200410 (Figs la-c; 



Table 1). 

Nine complete dP^s are knovm, five of which are 
certainly associated with B. nelsonensis molars (Fig 
1 k). They are all morphologically similar, consisting of 
a main blade, a posterolingual cusp and a very poorly- 
developed, lingual cingulum. The blade consists of a 
well-defined anterior and posterior cusp with a single 
intermediate cuspule and ridgelet, which often appears 
to be merged with the posterior cusp. The posterior 
cusp is higher than the anterior cusp. The anterior 
cusp is separated from the intermediate cuspule by a 
groove. The weak lingual cingulum comprises a low, 
narrow bulge extending from the posterolingual cusp, 
and terminating at a small anterolingual tubercle. The 
posterolingual cusp is lower than the main crest and 
is separated from the lingual cingulum and posterior 




Figure 1. Baringa nelsonensis. (a) NMV P200449 right dPj labial view, (b) lingual view, (c) oc- 
clusal view, (d) NMV P200482 left dP^ labial view, (e) occlusal view. Scale bar = 2 mm, (f) 
NMV P216028 right I^ buccal view, (g) Ungual view (h) NMV P200702 left I^ labial view, (i) lin- 
gual view, (j) NMV P200410 left dPj, dPj, Mj 2 in dentary fragment with associated I^ occlu- 
sal view, (k) NMV P201155 left associated dP^, dP^, M*-^ occlusal view. Scale bar = 10 mm. 



126 



Proc. Linn. Soc. N.S.W., 127, 2006 



KJ. PIPER AND N. HERRMANN 



Table 1. Dimensions (mm) of Baringa nelsonensis deciduous premolars. L = length, AW = anterior 
width, PW = posterior width. 



Specimen 



_L. 



dP7 



:£M. 






.SM. 



NMVP200410 
NMVP200449 
NMVP200450 
NMV P200690 
NMVP201155 
NMV P2 15777 
NMVP215789 
NMV P2 15790 



5.4 
5.7 
5.9 

5.5 
5.5 
5.7 
5.3 



2.5 
2.4 
2.3 

2.0 
2.6 
2.5 
2.4 



2.7 
2.6 

2.5 
2.4 
2.8 
2.5 
2.8 



NMVP200444 


5.6 


3.0 


3.9 


NMV P200482 


5.8 


3.1 


4.1 


NMVP201155 


6.3 


3.0 


4.1 


NMV P2 15774 


5.7 


2.9 


4.0 


NMV P21 5777 (R) 


6.1 


3.2 


4.4 


NMV P21 5777 (L) 


6.0 


3.3 


4.4 


NMV P2 15966 


6.5 


3.3 


4.0 


NMV P2 16888 


5.8 


3.2 


- 



cusp by a deep groove. A weakly developed medial 
posterior fossette is present (Figs Id-e; Table 1). 

Upper incisors: 

Sixteen partial and complete I's are known. They 
are large relative to the size of I^ and the molar teeth, 
a condition similar to that seen in Protemnodon. 
They are arc-shaped, possessing a convex labial 
surface, which is twisted slightly medially to bring 
the anterior-most tips into contact. They are widest 
near the root (7.2 mm), tapering slightly towards the 
tip (6.3 mm unworn). A moderately thick enamel 
covers the labial surface, extending over the sharply 
curved anterior edge onto the anterolingual surface to 
form a wide band. The lingual surface is only thinly 
enamelled close to the tip, which is removed by wear. 
The labial surface is occasionally ornamented by fine 
grooves and ridgelets, which follow the curvature of 
the tooth. They are similar morphologically to the I' 
of Protemnodon, but are readily distinguished, being 
smaller, less robust, and are narrower buccolingually, 
therefore producing a much smaller occlusal wear 
facet (2.7 mm average width) (Figs If-g; Table 2). 

Unfortunately none of the I^s have been found in 



association with B. nelsonensis cheek teeth. They are 
here assigned to B. nelsonensis as they are relatively 
abundant in the assemblage, are too small to be 
referable to either Protemnodon brehus, P. roechus or 
P. sp. nov. present in the fauna, and too large to be 
referred to any other genus of macropodid identified 
so far in the Nelson Bay Local Fauna. 

At the time of its description, only one moderately 
worn V- (NMV P173591 originally identified as I^) 
was known for 5. nelsonensis, and was not associated 
with any other specimens (Flannery and Harm 1984). 
Eight unworn complete and partial I^s are now 
known, but still none are associated with other B. 
nelsonensis material. However, they are relatively 
abundant in the assemblage, and are not referable to 
any other genus in the Nelson Bay Local Fauna, so 
their assignment to B. nelsonensis is still followed 
here. Many of the specimens lack a lingual surface 
due to damage, but they are all similar in morphology 
to NMV PI 73 591, being narrow anteroposterorly, 
and in possessing a short labial groove very close to 
the posterior edge, which continues onto the occlusal 
surface. The occlusal edge is notched approximately 
halfway along its length. The prominent cuspule at 



Proc. Linn. Soc. N.S.W., 127, 2006 



127 



AN EARLY PLEISTOCENE WALLABY 



Table 2. Occlusal length (mm) of Baringa nelsonensis upper incisors. OL = occlusal length, e = esti- 
mated. 























CO 












X) 


















00 


(S 


m 


^ 


r- 


r- 


00 


(N 


lO 


lO 


1-^ 


(N 


VO 


o 


f— 1 


(N 


(N 


■* 




Tf 


(N 


(N 


(N 


(N 


fN 


(N 


m 


m 


^ 


0^ 


O 


o 


i-^ 


r- 


0\ 


O 


(N 




so 


O 


O 


O 


o 


O 


o 


o 


o 




>n 


I^ 


00 


oo 


00 


0^ 


(N 


(N 




m 


^ 


VO 


VO 


VO 


SO 


vo 


V£> 


VO 


VO 


m 


o 


W-) 


*o 


>ri 


>o 


SO 


vo 




t-~ 


,—1 




r-H 


t— ( 


1—1 


1—1 


•— < 


>— 1 


1— t 


t^ 


o 


1-H 


1-H 


.-H 


*-H 


^^ 


1-^ 






CN 


CN 


<N 


(N 


(N 


(N 


(N 


(N 


(N 




<N 


(N 


(N 


(N 


cs 


(N 


tN 




cu 


Oh 


Dh 


Oh 


IX 


0. 


Dh 


CU 


Oh 


Oh 


CU 


o. 


Oh 


Oh 


&, 


CL, 


CLi 


eu 




> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 


> 




s 


S 


S 


S 


S 


s 


s 


s 


s 


S 


S 


S 


S 


S 


S 


s 


S 


S 




z 


:z 


2 


Z 


;z 


z 


?: 


Z 


z 


2 


JZ 


2 


2 


2 


2 


2 


2 


2 


11 

OL 


6.2 


6.1 


6.4 


7.0 


6.9 


7.5 


6.1 


7.4 


7.6 


7.2 


















12 






















4.3 


5.0 


5.1 


4.8e 


5.3 


5.1 


5.4 


5.1 


OL 







































the posterior end of the occlusal crest present in NMV 
PI 73 591 is variably developed in the present sample 
(Figs Ih-i; Table 2). 

Remarks 

The dP^s described above differ from NMV 
P173573, the isolated dP^ originally referred to 
Baringa by Flannery and Harm (1984) in the following 
details: they are shorter and broader; possess only 
one intermediate cuspule instead of two; lack a sharp 
lingual ridge on the anterior cusp; the posterior cusp is 
the highest, and is separated from the posterolingual 
cusp by a groove. NMV P173573 is very similar in 
both size and form to the P^ of Thylogale billardierii. 
This genus has since been recognised in the Nelson 
Bay Local Fauna, but was not known at the time of 
Flaimery and Harm's (1984) description. 

The identification of the posterior incisor 
as V- rather than I^ is based on the following 
observations. In all grazer and intermediate grazer- 
browser macropodines, P is relatively elongate 
anteroposteriorly and divided into two lobes by a 
labial groove, which is positioned approximately 



centrally or towards the posterior (Ride 1957). In 
contrast, V- is narrower and not divided into two 
lobes, with the short groove occurring very close to 
or on the posterior margin of the tooth. Flannery and 
Hann (1984) described NMV P173591 as an I^ based 
on its superficial similarity to V' of Onychogalea 
unguifera. But even in the latter species, where the 
posterior incisors are very reduced and narrow, I^ 
still possesses a labial groove, which is positioned 
approximately centrally. The unworn B. nelsonensis 
incisors described above are more consistent in 
morphology with that of I^. 

Examination of the much larger sample of B. 
nelsonensis material from Nelson Bay has shown 
there is little morphological variation within the 
species, and all other specimens are consistent with 
the holotype and referred specimens. 

Unfortunately none of the upper incisors have been 
found in association, either with each other or with 
other B. nelsonensis material. Due to the lack of more 
complete maxillae or premaxillae material, details of 
the palate and the shape of the incisor arcade are still 
unable to be described. 




Figure 2. Baringa nelsonensis. NMV P201156 right adult dentary, lateral view. Scale bar 
= 10 mm. 



128 



Proc. Linn. Soc. N.S.W., 127, 2006 



K.J. PIPER AND N. HERRMANN 



CURRAMULKA LOCAL FAUNA 'BARING A ' 
SPECIES 

Two species from the early Pliocene Curramulka 
Local Fauna, Yorke Peninsula, South Australia were 
tentatively allied with Baringa by Pledge (1992). 

The first species, cf. Baringa sp., is about 30% 
smaller than B. nelsonensis from Nelson Bay. It was 
referred to Baringa on the basis of similarities in the 
dentary shape, depth of the buccinator groove and 
morphology of P3 (Pledge 1992). Our re-examination 
of these specimens can confirm the possession of only 
the first of the features used by Flarmery and Harm 
(1984) to diagnose Baringa (i.e. a well-developed 
crest on the dentary just venfral to the ventral rim 
of the masseteric foramen). In cf Baringa sp. the 
anterior cingula on the lower molars are much shorter 
and broader than those of 5. nelsonensis, and the Ij is 
relatively narrower dorso-ventrally, and possesses a 
more horizontally-inclined wear facet. 

Compared to B. nelsonensis, which possesses 
only two intermediate cuspules, the P3 of cf Baringa 
sp differs in possessing three intermediate cuspules 
with more defined associated ridgelets. The lower 
deciduous premolar, dP2 also differs in lacking the 
small anterior and posterolingual cuspules present 
in the dPj of B. nelsonensis described in this paper. 
An unusual feature of this Curramulka species is the 
presence of a small shelf-like posterior cingulum 
or bulge on at least the M3 and M^ of some of the 
specimens (e.g. SAM P31337, SAM P29863). Pledge 
(1992) noted that the dentary was even in depth 
below the teeth in cf. Baringa sp., a feature he used to 
ally it to B. nelsonensis. However, the dentary of B. 
nelsonensis is deeper below M j than M4 (Flarmery and 
Harm 1984). The P^ of cf. Baringa sp. also possesses a 
lingual cingulum which is better developed, although 
only very slightly, than that seen in B. nelsonensis. 

Cf Baringa sp. appears instead to be closer to 
Thylogale, which is phenetically similar to Baringa 
(Flannery and Hann 1984), particularly in the 
morphology of the premolars and lower molars, 
but is smaller. In particular, the Curramulka Local 
Fauna specimens are most similar to extant T. 
stigmatica in the relative length of the premolars to 
the molar row, and to the extinct T. ignis from the 
Early Pliocene Hamilton Local Fauna (Flarmery et 
al. 1992) in the form of the premolars (i.e. presence 
of three intermediate cuspules on the main blade, a 
small cingulum around the base of the teeth, a low 
posterolingual cusp and the lack of a distinct lingual 
cingulum on P^). The similarity of the specimens to 
Thylogale was noted by Pledge (1992), however he 
considered them closer to Baringa based on features 



of the dentary. We believe these features differ 
significantly from those of 5. nelsonensis and suggest 
the specimens of cf Baringa sp. be referred to cf. 
Thylogale sp. pending a more thorough review of the 
Curramulka Local Fauna macropodids. 

The second species described by Pledge (1992), 
cf Baringa nelsonensis, is similar in size to the 
Nelson Bay specimens. However, as in cf Baringa 
sp., it resembles Baringa only in the possession of 
a well-developed crest on the rim of the masseteric 
foramen. Cf B. nelsonensis also differs from the type 
series of B. nelsonensis in having: a dentary that is 
even in depth below the teeth; the ascending ramus 
inclined slightly less vertically; a smaller Ij that is 
shallower dorsoventrally, and has a more horizontal 
wear facet; shorter and broader anterior cingula on 
the lower molars; a longer dP2 that lacks the small 
anterior cuspule; a larger P3 that is more rectangular 
in shape; a dP^ with a very well-developed lingual 
cingulum forming a shallow basin, and a well-defined 
intermediate cuspule and posterior fossette; a P^ with 
a better-developed lingual cingulum, a deeper groove 
separating the posterolingual cusp from the posterior 
cusp, a well-developed posterior fossette, and the 
three intermediate cuspules on the main blade sub- 
equal to, rather than lower than, the anterior and 
posterior cusps. In some respects the P^ is similar to 
that of Petrogale spp. and the dP^ is similar to that of 
Wallabia bicolor. Cf B. nelsonensis may represent an 
as yet unknown genus or species, but it is unlikely, for 
those reasons listed above, to be referable to a species 
of Baringa. 

DISCUSSION 

Flannery and Harm (1984) suggested that the 
lower incisors of Baringa nelsonensis might have 
been used to scrape off bark or lichens, or to sever 
hard plant stems. The enlarged crest on the rim of the 
masseteric foramen, and excavated jugal also noted 
by Flannery and Hann (1984), indicate the presence 
of an enlarged masseter muscle. This suggests that B. 
nelsonensis possessed an increased ability to move 
the dentaries anteriorly when compared to other 
macropodines (Sanson 1980; Flannery and Hann 
1984). Although the upper incisors have not been 
found in life position, their relative sizes and other 
general browsing features of the dentition suggest the 
I' probably extended below the occlusal line of I^'^ 
(Sanson 1989). The anterior movement of the dentaries 
would bring the lower incisors into occlusion with the 
large, robust I's, giving a possible mechanism for the 
production of the vertical wear facet observed on the 
lower incisors. 



Proc. Linn. Soc. N.S.W., 127, 2006 



129 



AN EARLY PLEISTOCENE WALLABY 



Observations on the stage of eruption and wear 
of molars associated with lower incisors supports 
Flannery and Harm's (1984) hypothesis that the 
majority of incisor wear occurs after the eruption of 
Pj, although some wear is seen to occur while dP2 is 
still part of the functional dentition (Herrmann 2000), 
indicating that the specialised feeding style described 
above is initiated early in the animal's life. 

Strong morphological similarities are observed 
between the I^s of B. nelsonensis and Pwtemnodon, 
as well as in the wear patterns observed on the 
lower incisors. The vertical wear pattern, one of 
the diagnostic characters of Baringa, has also been 
seen in some species of Protemnodon (Flannery and 
Hann 1984), and in the recently described Silvaroo 
bila (Dawson 2004). No I's are known for Silvaroo, 
however it is likely that they would also be relatively 
robust and enlarged relative to the cheek teeth, and 
that the feeding habits of Silvaroo may have been 
similar to that of Baringa. 

The most complete B. nelsonensis dentary from 
Nelson Bay, NMV P201156 (Fig 2), was found 
with I J attached in an apparent life position. This 
specimen therefore appears to have a relatively 
elongate diastema (94% the length of the cheek tooth 
row in B. nelsonensis compared to 75% in Thylogale 
billardierii), a feature usually associated with grazers 
(Ride 1959; Dawson and Flannery 1985). Baringa 
nelsonensis otherwise possesses dental features more 
indicative of browsing macropodids, i.e. narrow 
anterior cingula and weak midlinks on molars, 
moderately low-crowned molars, relatively large 
premolars, no evidence of molar progression, and 
only a very slightly curved lower tooth row, resulting 
in the eventual occlusion of both the anterior and 
posterior cheek teeth at the same time (Sanson 1980, 
1982, 1989). However, the lack of a lingual valley 
on the P^, and transverse striations on the molars 
indicating lateral movement of the lower jaw during 
mastication suggest that abrasive vegetation may also 
have been a part of its diet (Sanson 1980), possibly on 
a seasonal basis. 

If, as argued here, the Curramulka Local Fauna 
specimens are not referable to Baringa, the extension 
of the range of Baringa to the Early Pliocene by some 
workers (e.g. Tedford 1994) is no longer supported, 
returning its only named occurrence to the early 
Pleistocene. Interestingly, an un-named macropod 
from the Plio-Pleistocene Nullarbor Caves possesses 
upper incisors that bear a strong resemblance to those 
of Baringa (J. Long pers. comm.). This material is 
very well preserved and includes complete skulls 
and associated postcranial material. If this material 
is referable to Baringa or a new closely-related 



genus, it will add considerably to our knowledge 
of this extremely unusual macropod and its unique 
adaptations. 



ACKNOWLEDGEMENTS 

We would like to thank the Museum Victoria 
Palaeontology and Mammalogy departments, Monash 
University School of Geosciences and the University of 
Copenhagen for the provision of the facilities and access 
to the collections whilst conducting this research. Dr Jim 
McNamara of the South Australian Museum organised the 
loan of the Curramulka 'Baringa' specimens. Many thanks 
to Drs Tom Rich and Leah Schwartz, who read earlier drafts 
of the manuscript, and to the reviewers whose comments 
were very helpfiil. We are indebted to David Pickering of 
Museum Victoria, for his support, and untiring enthusiasm 
for collecting and preparing material from Nelson Bay. The 
renewed study of the Nelson Bay Local Fauna is funded by 
a Northcote Graduate Scholarship, Kings College London 
(awarded to K. Piper). 



REFERENCES 

Dawson, L. (2004). A new fossil genus of forest wallaby 
(Marsupialia, Macropodinae) and a review of 
Protemnodon from eastern Austtalia and New 
Guinea. Alcheringa 28, 275-290. 

Dawson, L. and Flannery, T. (1985). Taxonomic and 
phylogenetic status of living and fossil 
kangaroos and wallabies of the genus Macropus 
Shaw (Macropodidae: Marsupialia), with a new 
subgeneric name for larger wallabies. Australian 
Journal of Zoology 33, 473-498. 

Flannery, T.F. (1989). Phylogeny of the Macropodoidea; a 
study in convergence. In 'Kangaroos, wallabies 
and rat-kangaroos'. (Eds G. Grigg, P. Jarman 
and I. Hume) pp. 1-46. (Surrey Beatty and Sons: 
Sydney). 

Flannery, T., Rich, T.H., Tumbull, W.D. and Lundelius, 
E.L.Jr. (1992). The Macropodoidea 
(Marsupialia) of the Early Pliocene Hamilton 
Local Fauna, Victoria, Australia. Fieldiana: 
Geology 25, 1-37. 

Flannery, T.F and Hann, L. (1984). Anew macropodine 
genus and species (Marsupialia: Macropodidae) 
from the early Pleistocene of southwestern 
Victoria. Australian Mammalogy 7, 193-204. 

Gray, J.E. (1821). On the natural arrangement of 

vertebrose animals. London Medical Repository 
15,296-310. 

Hann, L.M. (1983). The vertebrate palaeontology and age 
of the Nelson Bay Formation, Portland Victoria. 
BSc (Honours) thesis, Monash University, 
Melbourne. 



130 



Proc. Liim. Soc. N.S.W., 127, 2006 



K.J. PIPER AND N. HERRMANN 



Herrmann, N.D. (2000). Dental analysis and 

palaeoecological assessment of Baringa 
nelsonensis (Marsupialia: Macropodidae: 
Macropodinae), an intermediate browsing/ 
grazing kangaroo from the Early Pleistocene 
Nelson Bay Formation, Victoria, Australia. MSc 
thesis, University of Copenhagen, Copenhagen. 

Kear, B.P. and Cooke, B.N. (2001). A review of 

macropodoid (Marsupialia) systematics with 
the inclusion of a new family. Memoirs of the 
Association of Australasian Palaeontologists 25, 
83-101. 

Luckett, W.P. (1993). An ontogenetic assessment of dental 
homologies in Therian mammals. In 'Mammal 
phylogeny'. (Eds F.S. Szalay, M.J. Novacek and 
M.C. McKenna.)pp. 182-204. (Springer-Verlag: 
New York). 

Owen, R. (1866). 'On the anatomy of vertebrates; volume 
2'. (Longmans, Green and Co.: London). 

Pledge, N.S. (1992). The Curramulka Local Fauna; a late 
Tertiary fossil assemblage from Yorke Peninsula, 
South Ausfralia. The Beagle, Records of the 
Northern Territory Museum of Arts and Sciences 
9, 115-142. 

Ride, W.D.L. (1959). Mastication and taxonomy in the 

macropodine skull. In 'Function and taxonomic 
importance'. (Ed. A.J. Cain.) pp. 33-59. 
(Systematics Association Publication No. 3: 
London). 

Ride, W.D.L. (1957). Protemnodon parma (Waterhouse) 
and the classification of related wallabies 
{Protemnodon, Thylogale and Setonix). 
Proceedings of the Zoological Society of London 
128, 327-347. 

Sanson, G.D. (1980). The morphology and occlusion 
of the molariform cheek teeth in some 
Macropodinae (Marsupialia: Macropodidae). 
Australian Journal of Zoology 28, 341-365. 

Sanson, G.D. (1982). Evolution and feeding in fossil and 
recent macropodoids. In 'The fossil vertebrate 
record of Ausfralasia'. (Eds P.V. Rich and E.M. 
Thompson.) pp. 489-506. (Monash University: 
Melbourne). 

Sanson, G.D. (1989). Morphological adaptations of teeth 
to diets and feeding in the Macropodoidea. In 
'Kangaroos, wallabies and rat-kangaroos'. (Eds 
G. Grigg, P. Jarman and I. Hume.) pp. 151-168. 
(Surrey Beatty and Sons: Sydney). 

Tedford, R.H. (1994). Succession of Pliocene through 
medial Pleistocene mammal faunas of 
southeastern Ausfralia. Records of the South 
Australian Museum 27, 79-93. 



APPENDIX 

Specimens of Baringa nelsonensis from Nelson 
Bay examined and discussed in tlie text 

NMV P173648 right I^; NMV P200659, partial right 
P; NMV P200685, partial left I'; NMV P2 16022, left 
and right I^; NMV P 216023, tip of right I^; NMV 
P216024, left I^; NMV P216026, right I'; NMV 
P216027, left and right I^; NMV P216028, right I'; 
NMV P216032, left andrightli;NMVP216035, worn 
left I'; NMV P216036, right Ji; NMV P216145a, right 
:•; NMV P216221; partial left P; NMV P173591, 
right I2; NMV P200702, left I^; NMV P2 15806, left 
I2; NMV P215810, left P; NMV P215811, right P; 
NMV P2 15992b left I^; NMV P2 16202, left I^; NMV 
P2 16224, left P; NMV P200444, left dP^ (associated 
with left dV\ M^-^); NMV P200482, left dP^; NMV 
P200490, anterior cusp of left dP^; NMV P2 15774, 
left dP-^ (in maxilla fragment also containing left dP^); 
NMV P2 15794, right dP^; NMV P2 15966, right dP^; 
NMV P216888, right dP^; NMV P201155, isolated 
left dP^ and right dP2 (associated with isolated left and 
right dP^, M'"^, left and right Ij, dP3, Mj 2, unerupted 
left and right M3, right P3); NMV P2 15777, isolated 
right and left dP^, and left dP2 (associated with 
isolated left and right I,, isolated right P3, left dPj, 
Mj 2 and unerupted P3 and M3 in partial dentary, right 
Mj 2 and unerupted M3 in partial dentary, isolated left 
P^ and left dP^, left M^'^ in maxilla fragment, right 
dP^-M^ in maxilla fragment); NMV P200410, left 
dP2 (in dentary fragment with dP3, Mj j, associated 
with Ij); NMV P200449, right dP2; NMV P200450, 
partial left dP2; NMV P200690, partial left dP2; NMV 
P2 15789, right dP2; NMV P2 15790, right dP2; NMV 
P201156, partial dentary containing Ij, P3, Mj^ 



Proc. Linn. Soc. N.S.W., 127, 2006 



131 



132 



Eastonian (Upper Ordovician) Graptolites from Michelago, 

near Canberra 

p. L. Williamson' and R.B. Rickards^ 

'School of Earth and Environmental Sciences, University of Wollongong, N.S.W. 2522, Australia 

(pennyw@uow.edu.au); and 
^Department of Earth Sciences, The University, Downing Street, Cambridge, CB2 3EQ, UK. 



Williamson, P.L. and Rickards, R.B. (2006). Eastonian (Upper Ordovician) Graptolites from Michelago, 
near Canberra. Proceedings of the Linnean Society of New South Wales 111, 133-156. 

A diverse Upper Ordovician (Eastonian) graptoloid fauna of some 20 taxa has been obtained from 'black 
shales' of the uppermost Foxlow Beds near Ryrie Hill south of Michelago. Eighteen of these are figured and 
described. The age indication is Eastonian 2 and 3, possibly about the caudatus/morrisi Biozone boundary 
in global graptolite terms. Some specimens exhibit a peculiar preservation, possibly of associated soft parts, 
though not necessarily graptolite soft parts. 

Manuscript received 23 June 2005, accepted for publication 7 December 2005. 

KEYWORDS: Eastonian, graptolites, Michelago, Ordovician, unusual preservations. 



INTRODUCTION 

Ordovician graptolites have been known for a 
number of years from the black shales towards the 
top of the Foxlow Beds. Richardson (1979) gives a 
detailed account of past work, beginning with the 
recognition of the Foxlow Beds by Oldershaw (1965) 
and records Ordovician graptolites fi-om two locations 
west of the Ryrie trigonometrical station (appendix 1, 
pp 182-183). Richardson and Sherwin (1975) discuss 
the Silurian outcrop slightly further west (Fig. 1). 
The present collections were made from the quarry 
east of the Ryrie trigonometrical station (Michelago 
1:100,000 map sheet 8726: 96953825, Fig. 1). The 
Ordovician forms recorded by Richardson (1979) 
are Orthograptus and Climacograptus species in 
open nomenclature, a possible amplexograptid, 
leptograptids and Dicellograptus forchammeri 
(sic) flexuosus. Richardson (1979, p. 30) notes that 
the graptolites represent "various zones within the 
Eastonian. Some possible late Gisbomian and early 
Bolindian fossils are also present." Our work entirely 
accords with this (Table 1 ), although we have recorded 
20 taxa with just one in open nomenclature: these are, 
for the most part, illustrated and described below. 

Our collections were made in 2002 by RL.W. In 
addition to the interestingly diverse fauna, embracing 
seven genera, some of the specimens exhibit a peculiar 
structure associated with parts of the rhabdosome of 



some biserials. It is possible that this represents soft 
tissue of some kind and it is illustrated and discussed 
in more detail below. 



PRESERVATION: GENERAL 

The graptolites are found in what was once black 
shale but which is mostly deeply weathered buff or 
whitish, soft mudstone: even traces of the original 
hemipelagic laminae are muted. The rock splits readily 
along the bedding planes and, except where stained 
with hematite, the graptolites are inconspicuous and 
in places very faint. A little of the original periderm 
is left in some instances but clay mineral replacement 
may also have occurred. In the few areas where 
the original black shale is preserved, in blobs and 
patches, the graptolites are dark silver-grey on a dark 
background, and are poorly preserved. The most 
serious drawback to this preservation is recognition of 
proximal end features, especially of proximal spines 
and early thecal growth. Even identification of early 
thecal apertures is often difficult. On the positive side 
there is no tectonic deformation: slabs with variously- 
orientated graptolites show no stretching, and there is 
no tectonic lineation on the bedding surfaces. Equally, 
there is no tectonic flattening parallel to the bedding, 
at least not to an extent that alters the dimensions of 
the graptolites. All the specimens are more or less 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



6040 



6039 :- 



6038 r^ 



6037 



6036 



6035 r^:: 



6034 




Figure 1. Location of the quarry at Ryrie Hill, south of Michelago and Canberra. Topography 
based on 1:100,000 Topographic Sheet 8726 Michelago 1974 Edition 1, generalised geology after 
Richardson & Barron 1977. 



134 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKAIODS 



V) 




9i 




e 




o 




N 




O 








.s 


^ 


^^ 


w 


a 


-a 


A 


i> 


o 


o 




u 


or 


<u 




h. 




x: 


e 


-^^ 


DC 


o 


o 


B 


4> 


o 


b> 


b 




«M 


>> 


>^ 


4> 




T> 


.fi 


& 


0£ 


«M 


V9 


o 


1/) 


*- 


u 


tu 


<1> 


I« 


«M 




S 


R 




"O 


•o 


f* 


a> 


cs 


OJ) 

e 




R 


R 


e 




R 






S 


R 


<s 


U 




•^^ 




Vi 




3 


s 


B 


< 


R 




a> 


1/1 


b< 


B 


0* 




£ 


R 
DXI 


^ 


« 


a; 






-o 


R 


at 


u 






o 


e 


a. 


•^ 






i« 


V 


s 


u 


01 


u 


E 


u 


4> 


• •V 


V 




« 

e 
s 


a 

R 

a 


«2 


'.a 


o 


^^rf 


ec 


PI 


« 








0) 


"O 


.fi 


a> 


u 


,o 






§ 




V 


!» 
«> 




"O 


C«a 


l> 


o 


o 


S»l 


T3 


4> 


4) 


Oil 


u 


B 


S 


R 


OX) 


U 


CC 


"r 


-4^ 

o 


.fi 


s 


o 






II 


U 




^ 


1-H 


yl 






tu 


u 




f\> 


.A 


-^^ 


m 


V) 


H 


< 



CO 

o 

c 

CO 



en 

a 

d 

"a 

B 
o 
o 



o 

£ 



St 

Si 
o 



en 

c 



c 
•5 

"o 



c 
o 

w 



c 
2 
'5 
o 

W 



C 

S 

'S 

o 
-«— * 



o 



O 

O 



(U 

CM 



H 



o 

X) 

CO 

5 



o 



§ 

• I— I 



to 
S 

S 

Co 



O 






o 

=0 

I 

bo 
o 

<3 



U 
O- 



Co 



u 

o- 



>1 
-C) 

=0 



03 

-Ki 

=0 



=0 



Co 

i 



bJD bo 
<3 Q 



O 



O 



<>3 

o 

o 



>1 

bo 
o 

o 



!>3 

a 



Co 



=1, 

bd bo 



o 






10 
CO 

S 

I 



CO 



CO 

O 






CO 

O 



5«. 



5< 



?1 
Co 

bo^ 



* 

S 

§■ 
I 



I 

bo 
o 



Co 
Co 

f 



to 

53 



* 



S^ 5^ 






!>3 



Co 
S 



I 

Co 
en 

I 
I 



Proc. Linn. Soc. N.S.W., 127, 2006 



135 



EASTONIAN GRAPTOLITES FROM MICHELAGO 





[*^v 



V ^ 



Figure 2a (left) Climacograptus caudatus Lapworth, AMF 114974, sketch of specimen showing 
possible soft tissue preservation, (full explanation in text); b (right) original of the same speci- 
men as in 2a for comparison with interpretation, scale bar 1 mm. 



diagenetically flattened and there is little pyritisation 
in the black shale, which is unusual. A number of 
specimens show what may be a thick layer of chloritic 
material entombing them, though this cannot be related 
to any tectonic strain shadow, at least in the outcrops 
we have dealt with. There is, however, a great deal of 
secondary hematite along veins, joints and as patches 
and circular blobs on the bedding planes: in places the 
whole rock is suffused with a pink colour as a result 
of hematite staining. Thus there are some difficulties 



attending the identification of these graptolites and 
this we have tried to reflect in the systematics section 
below. 



PROBLEM PRESERVATION: POSSIBLE SOFT 
TISSUE 

There is one interesting problem of a rather 
striking preservation (Fig. 2) concerning rings of 



136 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



hematite surrounding the graptolites. Of the hundreds 
of graptoHtes collected only nine or ten display this 
feature, some very strikingly. This preservation may 
represent soft tissue, not necessarily graptolitic, or it 
is more likely hematite staining, part of the overall 
process of preservation of the fauna. The phenomenon 
is two dimensional, restricted to bedding planes. 

The hematite rings, or "bubbles" affected only 
the biserial graptolites, Orthograptus calcaratus s.l. 
and Climograptus caudatus. Several of the rings 
have an internal structure (Fig. 2) and occasionally 
the rings appear on the bedding plane unconnected 
with a graptolite. Other red stainings have a patchy, 
blob-like, or ring-like arrangement and these are 
unequivocally secondary hematitic staining of the 
sediment. 

The most striking specimen is of C. caudatus 
(Fig. 2) which has two ring-like structures from about 
the 15* to the 30* thecal pair. The uppermost, larger 
circular body has about 40 radiating lines, looking like 
septae, each connected by short bars. However, under 
high magnification this is a more patchy structure 
- vesicular almost - than rigidly radiating (see 
interpretation in Fig. 2a). The lower, smaller circular 
area shows the same structures less well. Surrounding 
both circular areas, but not separating them, is a thick, 
red, hematitic band followed on the outside by an 
apparently vesicular layer. A small circular structure 
to one side of the graptolite shows the same features, 
and may be associated with graptolitic fi-agments or a 
smaller specimen of biserial graptolite. The specimen 
of C. caudatus is itself preserved in hematite and is 
possibly a scalariform view. It extends beyond thp 
uppermost circular structure for some fiirther 13 mm, 
beyond which is a slightly expanded virgula for 4 
mm before the end of the slab. The preservation of 
the graptolites as a whole is poor with little except the 
virgella and its tube-like structure visible. 

Other specimens do not show the above detail, but 
do show circular structures "attached" to specimens 
of O. calcaratus s.l. that are also stained/preserved in 
hematite. In some the colour of the ring-like structures 
is a yellow-orange, suggesting alteration to goethite. 
It is difficult to decide whether such structures are 
organic or not. If they eventually prove to be organic 
then they could be algal or coelenterate in nature. The 
only previously-known graptolite soft parts are those 
recorded by Kozlowski (1949) (eggs and embryos), 
Bulman and Rickards (1966) (embryos), Rickards 
and Stait (1984) (zooids), Crowther et al. (1987) 
(cellular tissue), and the work of Bjerreskov (1987) 
on pyritisation features possibly representing soft 
parts. Loydell et al. (2004) also describes soft tissue 
within the thecal tubes. 



AGE OF FAUNA 

Table 1 gives the global range of the identified 
graptolites. It is clear that the most probable horizon 
is Eastonian and perhaps Eastonian 2 and 3: that is, in 
global graptolite terms, the upper half of the clingani 
Biozone, (caudatus level), morrisi Biozone, and 
linearis Biozone. The most likely level, which we 
justify below, is the caudatus/morrisi boundary. Much 
of the 'wooliness' in this age attribution is due to the 
difficult preservation of the fauna and consequent 
difficulties of identification. We feel certain that 
some smaller species have been missed - there are, 
for example, small specimens we have provisionally 
labelled as Cryptograptus, and other climacograptids 
with proximal spines may occur. 

There are a few anomalies in our identifications 
but these do not affect the overall judgement on the 
age of the fauna. The most obvious anomaly is our 
identification of a few specimens oilClimacograptus 
uncinatus, normally considered a Bolindian species. 
We may have accidentally collected these from a 
loose block that came from a different part of the 
section; certainly the rock type is slightly different 
in its preservation. But if the identifications are 
correct it does suggest the presence of Bolindian 
strata nearby. There is a great deal of poor exposure 
in the region, as well as the quarry itself, from which 
we have not successfully collected. Orthograptus 
quadrimucronatus quadrimucronatus has been 
recorded fi-om the Bolindian in Australia (see 
VandenBerg and Cooper 1992) but is much more 
common, globally, as shown in Table 1 . 

Climacograptus caudatus Lapworth' has been 
identified with certainty and is recorded in Australia 
fi-om Gisbomian 2 to Eastonian 2, but not higher. 
Elsewhere it occurs in the caudatus and morrisi 
biozones, roughly equivalent to the upper part of 
Eastonian 1 and Eastonian 2. The similar species C. 
tubuliferus occurs in Eastonian 3 and 4 and ranges 
into the Bolindian in Australia, but elsewhere occurs a 
little earlier, in the morrisi Biozone, thus overlapping 
slightly with C. caudatus (see Williams 1982, p. 246). 
The occurrence of these two forms together strongly 
suggests an age for the fauna of around the caudatus/ 
morrisi boundary, that is Eastonian 2. 

There seems little in the remaining fauna that 
conflicts with the age attribution to Eastonian 2, apart 
from the examples mentioned above. The occurrence 
of Orthograptus amplexicaulis intermedius does 
reach the level of the clingani Biozone on previous 
records, but not the upper parts of that biozone: given 
the difficulties of distinguishing subspecies of O. 
amplexicaulis pending the radical revision needed for 



Proc. Linn. Soc. N.S.W., 127, 2006 



137 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



that species group, we cannot at present place much 
weight on the known range of O. a. intermedius. 

Some small specimens preserved in bedding 
planes covered in graptolite debris may be referable to 
the genus Cryptograptus. Whilst generally considered 
as ranging into the low clingani Biozone, in Australia 
the genus reaches Eastonian 3 (VandenBerg and 
Cooper 1992) roughly the equivalent of the linearis 
Biozone. There is also a considerable number of 
climacograptid specimens that we have been unable 
to identify with certainty. They may be either early 
growth stages of C. caudatus or ones in which the 
tubular grow1:h along the virgella has not occurred; or 
they may be referable to another species such as C. 
styloideus, which they generally resemble except in 
the absence of the distal nemal vane. 

Finally there is a problem, it seems to us, in 
distinguishing Climacograptus wilsoni from C. 
tubulifenis; we have opted for the latter because 
the proximal thecae in our material show no signs 
of spines. Therefore we consider the most likely 
stratigraphic level represented by this assemblage is 
either the caudatus Biozone or the morrisi Biozone, 
or some horizon close to the boundary of the two, and 
to be unequivocally Eastonian. 



SYSTEMATICS 

NOTE: FIGURES 3-9 ARE AT THE END OF THE 
TEXT 

Class Graptolithina Bronn 1849 

Order Graptoloidea Lapworth 1875 

Family Nemograptidae Lapworth (ex Hopkinson 

ms) 1873 

Genus Leptograptus Lapworth 1 873 

Type species (by original designation) Graptolithus 

flaccidus RaW 1865. 

Diagnosis Biramous, occasionally multiramous 

stipes, slender, flexed, often slightly reclined, with 

simple, long, low-angled thecae mostly without 

spines. 

Leptograptus flaccidus (Hall, 1 865) cf. macer Elles 

and Wood 1903 

Figures 3a-e 

cf 

1903 Leptograptus flaccidus var. macer var. nov.; 

Elles and Wood, pp. 110-111, pi. 15, figs2a-i. 
1934 Leptograptus flaccidus Hall var. macer Elles 

and Wood 1903; Ruedemann and Decker, p. 



306, pi. 40, figs 5-6. 
71963 Leptograptus cf L. flaccidus var. macer Elles 
and Wood 1903; Ross and Berry, p. 101, pi. 6, 

fig. 1. 
1982 Leptograptus flaccidus macer Elles and Wood 
1903; Williams, pp. 233, 236, figs 4a-e. 

Lectotype Only relatively recently proposed by 
Williams (1982 p. 233) BU1377 figures by Elles and 
Wood (1903, plate 15, fig. 2e). 
Material About ten specimens and numerous 
fragments, probably referable to this species. 
Diagnosis Rhabdosome with a proximal dorsoventral 
width of 0.25-0.35 mm more distally 0.60 mm; 
variously gently flexed, but usually gently declined 
proximally and reclined or reflexed distally; thecal 
spacing 6-9 in 10 mm proximally and 8-9 in 10 mm 
distally. 

Description The rhabdosome is variously flexed, in 
some specimens very gently declined or horizontal 
initially, becoming gently reclined or reflexed distally. 
A few specimens show greater curvature distally and 
the two stipes may have different curvature. It is 
uncertain how much of this variation is a result of 
diagenetic flattening. The sicula is often conspicuous 
but only a millimetre or so is preserved in the best 
specimens; it could be much longer because many of 
the siculae are clearly broken (e.g. Figs 3a, d, e). The 
apparent prothecal curvature seen in some specimens 
(e.g. Fig. 3b) does not seem to be real prothecal folding 
and may be a reflection of difficult preservation. 
The thecae are simple tubes with seemingly quite 
denticulate apertures in places perhaps reflecting a 
slight apertural expansion. The virgella is a short, 
conspicuous spine (see Figs 3b, c). 
Remarks These specimens closely resemble L. 
flaccidus macer as described by Elles and Wood 
(1903) and WilHams (1982) but have even lower 
thecal spacing proximally. Elles and Wood (1903) 
do give a thecal spacing of 6 in 10 mm, but this 
is for distal thecae; Williams (1982) also gives a 
reduction in the distal figure (9 in 10 proximally and 
8 in 10 distally). The reverse is true in the Michelago 
specimens. Otherwise this material is very close to 
previous descriptions. L. flaccidus cf macer differs 
fi-om L. eastonensis in having slightly more robust 
stipes, in having a lower thecal spacing (10-11 in 10 
mm given by Keble and Harris 1925, p. 514): Keble 
and Harris (1925) comment that L. flaccidus macer 
is the closest species to L. eastonensis. L. flaccidus 
subjectus has strongly reclined stipes in the proximal 
region, which later become reflexed: it otherwise 
resembles L. flaccidus flaccidus and is more robust 
than L. flaccidus arcuatus and L. capillaris which 



138 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



have greatly flexed stipes, unlike L. flaccidus cf. 
macer, whilst the former is more robust. L. flaccidus 
macilentus has more rigid stipes and is more robust 
also. The remaining leptograptids described by Elles 
and Wood (1903) namely L. validus, L. grandis, L. 
latus andZ. ascendens are all either more robust, have 
different thecal spacing, or both. However, it must 
be said that leptograptids of the L. flaccidus group 
do seem to us to show such variation as to suggest 
that some of the subspecies may be unrecognisable: 
further work on the group is necessary. Neither L. 
flaccidus macer nor L. flaccidus spinifer (see below) 
have been previously recorded from Australia. 
VandenBerg and Cooper (1992) list L. capillaris, L. 
eastonensis, and L. flaccidus arcuatus, which we have 
commented upon. Some slender dicellograptids have 
not dissimilar rhabdosomal proportions, but thecae 
are more complicated and the stipes ususally reclined 
not declined or deflexed. 

Leptograptus Iflaccidus spinifer Elles and Wood 

(1903) 

Figures 3f-i 

71903 Leptograptus flaccidus var. spinifer var. nov; 
Elles and Wood. 

Holotype BUI 03 7, figured by Elles and Wood 
(1903)platel4,fig. 2a. 
Material Three specimens, all figured herein. 
Description The sicula is well-preserved 1.50-2.00 
mm long and has a short nema and a conspicuous 
if short virgella (Fig. 3g). The virgella is deflected 
back across the sicular aperture. The origin and 
early growth of thl' and thP is not clear but both are 
prominently spined: either a subapertural spine or 
strong denticulation, probably the latter. Subsequent 
thecae have no spines, are low angled (5°- 10°) and 
have apparently simple apertures. Thecal overlap is 
low and thecal spacing 7-8 in 10 mm. Dorsoventral 
width proximally, excluding denticles, is about 0.30 
- 0.40 mm and distally may reach 0.60 mm with a 
thecal spacing there of 9 in 10 mm. 
Remarks There is a superficial resemblance to the 
proximal ends of some spinose dicellograptid species, 
but the thecae of this form are elongate, apparently 
simple, and low angled. Only the first two thecae are 
spinose/denticulate. This form has not previously been 
recorded from Australia: VandenBerg and Cooper 
(1992) record only the subspecies L. f arcuatus. The 
closest dicellograptid is probably D. forchhammeri 
but this does have slightly introverted thecae and a 
higher thecal spacing value (9-12 cf. 7-8 in 10 mm). 



Family Dicranograptidae Lapworth 1873 
Genus Dicellograptus Hopkinson 1871 

Type Species Subsequently designated Gurley 
(1896), Didymograpsus elegans Carruthers 1868. 
Diagnosis Rhabdosome of two reclined uniserial 
stipes, straight or curved usually symmetrically: 
thecae almost simple to strongly introverted, mostly 
with prothecal folds proximally; variously spined, 
especially proximal thecae. 

Dicellograptus morrisi Hopkinson 1871 
Figures 4a-c 

71867 Didymograpsus flaccidus Hall; Nicholson, pp. 

110-111, pi. 7, figs 1-3. 
1868 Didymograpsus elegans Carruthers; Carruthers 

(pars) pi. 5, figs 8b, c (non figs 8a, d = D. 

elegans sensu stricto). 
1871 Dicellograpsus morrisi sp. nov.; Hopkinson, p. 

5, pi. 1, figs 2a-h. 

1876 Dicellograptus morrisi Hopkinson; Lapworth, 
pi. 4, fig. 85. 

1877 Dicellograptus morrisi Hopkinson; Lapworth, 
pi. 7, fig. 6. 

1904 Dicellograptus morrisi Hopkinson; Elles & 

Wood, pp. 155-157, pi. 21, figs 6a-d, text-figs 

98a-e. 
1904 Dicellograptus pumilus Lapworth: Elles & 

Wood (pars), pi. 21, fig. 3c, {non p. 149, pi. 21, 

figs 3a, b, d-f = Z). pumilus sensu stricto/ 
1963 Dicellograptus morrisi Hopkinson; Skoglund, 

pp. 31-32, pi. 1, figs 1, 2. 
1970 Dicellograptus morrisi Hopkinson; Toghill, pp. 

17-18, pi. 7, figs 1-4, text-figs 4d-f 
1976 Dicellograptus morrisi Hopkinson; Erdtmann, 

pp. 92-93, pi. 5, figs L/2b, M/6a, pi. 11, fig. 

K/2b, pi. 12, fig. YUA. 
1982 Dicellograptus morrisi Hopkinson; Williams, 

pp. 238-239, figs 7e, f, 8a-c. 
71983 Dicellograptus morrisi Hopkinson; Williams 

& Bruton, pp. 169-170, figs lOD, 14A-E. 
2002 Dicellograptus morrisi Hopkinson, 1871; 

Rickards, p. 4, figs 4A-D. 

Type specimens Not yet designated. 

Material About thirty specimens including some 

fragmentary uniserial stipes probably belonging to 

this species. 

Diagnosis Stipes more than 80 mm long widening 

rapidly from 0.50-0.60 mm proximally to 1.2 mm 

distally. Axial angle from 30°-55°, axil itself slightly 

rounded. Thecae number 11-13 in 10 mm proximally, 

with curved supragenicular walls and sub-apertural 



Proc. Linn. Soc. N.S.W., 127, 2006 



139 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



spines at least for the first nine thecae in each stipe. 
More distally the thecal spacing figure falls to 9-11 in 
10 mm. The proximal dorsoventral width is 0.50 mm 
and more distally reaches 1.20-1.30 mm. 
Description The complete rhabdosome is very large 
with stipes in excess of 70 mm. Some of the stipes are 
almost straight but mostly they have a gentle, ventral, 
distal curvature with suggestions of an equally gently 
spiral growth. There are a few distal fi-agments that 
reach 1.40-1.50 mm but it is uncertain whether these 
are referable to D. morrisi or not: they may be very 
distal fragments of very large specimens, or they may 
represent a second species for which we have no 
proximal region. 

Remarks Our material supports the description of 
Skoglund (1963) and Rickards (2002) which gave 
up to eleven and eight spinose thecae respectively. In 
all other respects they closely agree with much of the 
previously described material. Dicellograptus morrisi 
has not previously been recorded from Australia. 

Dicellograptus cf. caduceus Lapworth 1876 
Figures 4d, 5a, b 

1876 Dicellograptus caduceus Lapworth; Lapworth, 

pp. 141-2, pi. 7, fig. 3. 
1904 Dicellograptus caduceus Lapworth; Elles and 

Wood, pp. 161-3, pi. 23, figs 4a-c, text-figs 102 

a-c. 



descriptions in that the stipes distally seem less than 
1 mm in our material. Some of the specimens figured 
by Elles and Wood (1903, pi. 23) are also less than 1 
mm and where they reach 1 mm may be tectonically 
widened. The loops seem more variable in the original 
material and the first loop smaller. 

What is surprising, in the descriptions of a 
similarly enrolled species, D. complexus, is that 
neither Davies (1929, pp 3-4) nor Williams (1983, pp. 
36-7) who revised the species, discuss D. caduceus at 
all. Yet the two forms are very similar in dimensions 
and appearance, although D. complexus is restricted 
to the anceps Biozone and D. caduceus to the morrisi 
Biozone (Eastonian 2) and, in Australia, Eastonian 3- 
4. The loops of D. complexus are smaller and tighter 
than in our specimens of Z). cf caduceus and Williams 
(1983, p. 36) implies, but does not state specifically, 
that D. complexus is distinguished fi"om D. caduceus 
in that the former has left-handed torsion of the 
stipes. That feature is uncertain in our material, but 
may be right-handed. Dicellograptus complexus has 
not been recorded firom Australia; D. caduceus has 
(see VandenBerg and Cooper, 1 992) but the species, 
considered globally is rarely recorded. It may be that 
fiirther research will recognise more variation and 
more species. 

Dicellograptus sp. 
Figures 5c, d 



Type specimens Not yet identified. 
Material About 60 specimens, some slabs with up to 
10 per slab. 

Diagnosis Spirally coiled stipes crossing at least 
twice, firom a proximal region slightly rectangular 
and lacking, as a rule, a preserved sicula. Proximal 
dorsoventral width 0.40-0.50 mm, distally up to 0.70- 
0.80 mm; proximal thecal spacing 12-14 in 10 mm, 
distally about 10-11 in 10 mm. 
Description The only specimen with a sicula 
preserved (Fig. 5b) shows a length of 1 .30 mm with 
no attached nema. The sicula is midway between the 
two stipes. Early thecal development has not been 
seen but thl' and thP have short spines as a rule, 
occasionally well-developed (Fig. 5a). Some later 
thecae may also have short and inconspicuous spines. 
The coiled stipes are conspicuous, the first crossing 
of stipes being at about 15-18 mm from the sicula, 
the maximum distance between the stipes, in the 
first loop, being a little under 10 mm. Two loops are 
common in this material, and possible three loops in 
some cases. Loops are similar in dimension whether 
the first or the third. 
Remarks This form seems to differ fi-om the original 



Material A single specimen, AMF 1 14903. 
Description A conspicuously robust rhabdosome 
proximally with very large proximal spines positioned 
at 90° to each other; the longer spine is 8.25 mm long. 
Both spines may be incomplete as seen, and at their 
bases are about 0.50 mm wide. Approximately 0.75 
mm of sicula is faintly visible but whether this is the 
real length is unclear. A virgella has not been identified. 
If the apex of the sicula is positioned as indicated the 
total length of the sicula could be around 1 mm if the 
apertural region is obscured in this specimen. There 
is a web of material spanning the two stipes that helps 
obscure the proximal region. The stipes diverge at 50° 
and initially have a dorsoventral width of 0.50 mm or a 
little more but reach 1 .0 mm after only eight thecae or 
so and have a thecal spacing of 13 in 10 mm. We have 
one distal dicellograptid fi-agment (Fig. 5d) possibly 
referable to this form, with a dorsoventral width of 2 
mm and a thecal spacing of 9-10 in 10 mm. 
Remarks Dicellograptus sp. is remarkably similar 
in overall appearance to D. ornatus Elles and Wood 
(1904) but is a much larger and more robust form 
with longer, broader spines. Whilst the proximal end 
template is very similar, as well as the thecal spacing, 



140 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



the sicula is smaller and the stipes rapidly become 
more robust. Early growth stages of Dicellograptus sp. 
would have to be at least twice the size as comparable 
stages ofD. ornatus. 

Family Diplograptidae Lapworth 1 873 
Genus Climacograptus Hall 1 865 

Type species Graptolithus bicomis Hall (1847) by 

original designation. 

Remarks Because of the general nature of the 

preservation of this material we have adopted a rather 

conservative classification of Climacograptus which 

contrasts slightly with that of VandenBerg and Cooper 

(1992). 

Climacograptus caudatus Lapworth 1876 
Figures 6a-d 

1876 Climacograptus caudatus sp. nov.; Lapworth, 
pi. 2, fig. 48. 

1 877 Climacograptus scalaris var. caudatus 
Lapworth; Lapworth, pi. 6, fig. 34. 

1906 Climacograptus caudatus Lapworth; EUes & 

Wood, pp. 202-203, pi. 27, figs 7a-e, text-figs 

133a-d. 
1 908 Climacograptus caudatus Lapworth; 

Ruedemann, pp. 438-439, pi. 28, figs 17-18, 

text-fig 405. 
71913 Climacograptus caudatus Lapworth; 

Hadding, pp. 49-50, pi. 3, figs 18-19, text-fig. 

19. 
71934 Climacograptus caudatus Lapworth; 

Ruedemann & Decker, p. 319, pi. 43, figs 1-la. 
1947 Climacograptus caudatus Lapworth; 

Ruedemann (pars), p. 424, pi. 72, figs 57-65 

{non pi. 71, figs 51-52). 
71955 Climacograptus caudatus Lapworth; Harris & 

Thomas, pp. 38-39, pi. 1, figs 4-6. 
1971 Climacograptus caudatus Lapworth; Strachan, 

p. 32. 
1981 Climacograptus? caudatus Lapworth 

1876; Williams, pp. 135-6, pi. 33, figs 1-6, 3 

unnumbered text-figs. 

1989 Ensigraptus caudatus (Lapworth 1876); Riva 
and Kettner, p. 89. 

1990 Climacograptus caudatus VandenBerg, fig. 1. 
1992 Ensigraptus caudatus (Lapworth); VandenBerg 

and Cooper, pp. 46, 48, 81, fig. 9A. 
2002 Ensigraptus caudatus (Lapworth); 
VandenBerg, p. 45, fig. 5.1.4/11. 

Type specimen According to Strachan (1971) the 
type specimen has not been traced. 



Material At least 50 specimens, possibly more (see 
under Remarks). 

Diagnosis Climacograptus lacking proximal thecal 
spines and with characteristic proximal growth of 
a virgellate or siculate structure (= parasicula of 
VandenBerg, 1990) and distal growth of a moderately 
robust, long virgula; rhabdosome with proximal 
dorsoventral width of 0.75-1.00 mm and a distal 
dorsoventral width of up to 2.5 mm; proximal thecal 
spacing 12-13 in 10 mm, distal thecal spacing 9- 
11 in 10 mm; distal thecae with outward-sloping 
supragenicular wall. 

Description Few certain early growth stages 
have been identified, almost certainly because of 
preservational difficulties, but Fig. 6c is of an early 
growth stage with virgula and virgella preserved, the 
latter with traces of the typical process that grows 
along the virgella. The nature of this process cannot 
be seen in this material. In this early growth stage thl ' 
seems unusually prominent as it does in the specimen 
illustrated as Fig. 6c; this feature has not been seen 
on any other specimens, all of which seem to have 
typical climacograptid thecae numbering 12-13 in 10 
mm. In a few specimens (Fig. 6a) the supragenicular 
wall seems inclined outwards slightly as it ofi;en does 
more commonly in distal thecae. Whether this is real 
or a preservational feature is uncertain, but it has been 
commented upon by other workers (see Remarks). The 
virgella grows very long, possibly up to 10 mm and in 
mature specimens the characteristic virgellate growth 
may reach 5 mm. The virgula is always long, up to 
15 mm, and fairly robust without being dramatically 
expended (Fig. 6d; and see following description). 
Remarks VandenBerg (2002) gives a^ range of 
Eastonian 1 and 2 for this species in Victorian strata, 
although VandenBerg and Cooper (1992) give 
Gisbomian 2 to Eastonian 2 as the Australian range; 
this latter is in accord with the global range (Table 
1). Williams (1981) was the first to draw attention 
to the apparently orthograptid/glyptograptid distal 
thecae of C. caudatus and hence questioned the 
generic attribution, although he abandoned this later 
(Williams 1994). It seems to us that many Ordovician 
climacograptids have gently outward-sloping 
supragenicular walls and this effects a contrast 
with the largely Silurian genus Normalograptus. 
Subsequently VandenBerg (2002) along with other 
workers recognised C caudatus as the type of 
Ensigraptus (Riva and Kettner 1989) on the grounds 
that the early development was slightly more primitive 
than otherwise similar climacograptids. We cannot 
comment on that from this material: it is possible that 
the slightly conspicuous appearance of thl' in two 
specimens, referred to above, reflects the tendency 



Proc. Linn. Soc. N.S.W., 127, 2006 



141 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



of that thecae to grow downwards and outwards as 
described by Riva and Kettner 1989. 

In addition to this considerable number of 
specimens attributed without doubt to C. caudatus 
we have a large number of climacograptids of similar 
dimensions yet lacking the pronounced virgella, the 
parasicula, or the robust virgula (Figs 7a, b). These 
could be specimens of C caudatus in which the robust 
virgella and virgula have not developed; or they could 
be referred to C. tubuliferus (see next description) 
in which the expanded virgular vane had not yet 
developed; or they could be part of a plexus marking 
a possible evolutionary transition from C. caudatus 
to C. tubuliferus (see Rickards et al. 2001). Similar 
forms to these, lacking a robust virgella or expanded 
virgules, may have been previously identified as C. 
pulchellus (Hadding 1915) (see Rickards et al. 2001 
p. 79, fig. lOB). It is a pity that the preservational 
state of the Michelago assemblage does not allow 
pursuance of these questions. 

Climacograptus tubuliferus Lapworth 1876 
Figures 6e-h 

1876 Climacograptus tubuliferus Lapworth; 
Lapworth, pi. 2, fig. 49. 

1877 Climacograptus scalaris var. tubuliferus 
Lapworth; Lapworth, pi. 6, fig. 33. 

1902 Climacograptus tubuliferus Lapworth; Hall, p. 

55,pl. 13,fig.5,pl. 14,fig.4. 
1906 Climacograptus tubuliferus Lapworth; Elles 

& Wood, pp. 203-204, pi. 27, figs 8a-d, text-figs 

134a-c. 
1947 Climacograptus tubuliferus Lapworth; 

Ruedemann, p. 440, pi. 75, figs 54-56. 
71948 Climacograptus styloideus Lapworth; 

Henningsmoen, p. 404. 
1955 Climacograptus tubuliferus Lapworth; Harris 

& Thomas, p. 40, pi. 1, figs 10-12. 
1960 Climacograptus tubuliferus Lapworth; Berry, 

p. 85, pi. 19, fig. 5. 
1963 Climacograptus tubuliferus Lapworth; Ross & 

Berry, p. 132, pi. 10, figs 1,2. 
71963 Climacograptus styloideus Elles & Wood; 

Skoglund, pp. 38-40, pi. 2, figs 1-4. pi. 3, fig. 3. 
1969 Climacograptus tubuliferus Lapworth; Moors, 

pp. 268-270, figs 3a-c. 
1977 Climacograptus tubuliferus Lapworth; Carter 

& Churkin, pp. 23-24, pi. 7, fig. 5. 

1982 Climacograptus tubuliferus Lapworth; 
Williams, pp. 245-246, figs lla-n. 

1983 Climacograptus tubuliferus Lapworth; Williams 
& Bruton, pp. 170-172, figs 12c-e, 15a-n. 

1983 Climacograptus tubuliferus Lapworth; Koren' 



and Sobolveskaya (pars), pp. 139-141, pi. 41, 

figs 1-3, {non pi. 40, figs 6-117). 
1987 Climacograptus tubuliferus Lapworth, 1876; 

Williams, p. 80, figs 4F, H, I, 6G, 70-Q. 
71988 Scalarigraptus tubuliferus (Lapworth); Riva, 

figs 2i, j (7=Normalograptus normalis). 
1989 Normalograptus tubuliferus (Lapworth); Riva 

(in Riva and Kettner), pp. 87-89, figs lOa-i, 11a- 

e. 

1991 Climacograptus tubuliferus (Lapworth, 1876); 
Williams, pp. 593-4, pi. 1, figs 2-4, 75, figs 8A- 
C. 

1992 Climacograptus tubuliferus Lapworth; 
VandenBerg and Cooper, p. 81. 

1992 Normalograptus tubuliferus tubuliferus; 
VandenBerg and Cooper, p. 50, fig lOA. 

Type Specimens Lapworth's original specimen has 
not yet been recognised (Strachan, 1971, p. 35). 
Material Around 40 specimens. 
Diagnosis Climacograptus lacking proximal thecal 
spines but with a characteristically expanded, 
?vane-like virgula, and a small virgella; thecae 
broadly climacograptid numbering 10-14 in 10 mm; 
rhabdosome proximally with dorsoventral width of 
0.70-0.75 mm rising distally to 2.50 mm. 
Description Some rhabdosomes have a length of 
13 cm but do not exceed 2.50 mm in dorsoventral 
width. The vane-like structure is up to 1 mm wide 
and extends distally, often as much as 20 mm, and 
even then may be incomplete. The proximal end 
usually has a small virgella but in some specimens 
it is more robust. It does not have a parasicula. The 
thecae are climacograptid throughout, except for a 
few specimens (Fig. 6f) where the supragenicular 
wall does appear to be outward leaning, though this 
could be a preservational feature. 
Remarks C tubuliferus ranges from Eastonian 2 to 
Bolindian 1 in Australia (VandenBerg and Cooper 
1992) but elsewhere has been recorded from the 
latest clingani level (Table 1). The variation referred 
to above has already been commented upon under 
"Remarks" in the preceding description. 

1 Climacograptus lanceolatus VandenBerg 1990 
Figure 7g 

71990 Climacograptus lanceolatus sp. nov.; 

VandenBerg, pp. 44-49, fig. 1, figs 7A-P, 8A-C. 

Remarks A single problematical specimen is 
undoubtedly a Climacograptus species with a 
maximum dorsoventral width of 2 mm and a thecal 
spacing of 8-10 in 10 mm, which agrees with the 



142 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



original dimensions given by VandenBerg (1990 
p.47). The proximal end has two spines and the shorter 
of the two directed ventrally may derive from thl'. 
The longer spine, in exactly the correct disposition 
for C. lanceolatus, is possibly the virgella, though 
this cannot be proved. We have no other specimens of 
spinose climacograptids in the collection displaying 
these features. C. lanceolatus is Eastonian 1 according 
to VandenBerg (1990). One of the referees suggested 
the possibility that this form was referable to 
Pseudoclimacograptus, but it should be noted that in 
some views and preservations climacograptid thecae 
can be slit-like without being pseudoclimacograptid. 

Climacograptus mohawkensis (Ruedemaim 1912) 
Figures 7h, i 

1906 Climacograptus minimus (Carruthers); Elles 

and Wood, p. 191, pi. 27, figs la-g, text-figs 

124a-d. 
non 1868 Diplograptus minimus sp. nov.; 

(Carruthers); p. 74, pi. 5, figs 12a, b. 
1912 Diplograptus (Mesograptus) mohawkensis sp. 

nov.; Ruedemann; pp. 80-2, pi. 2, figs 18, 19, 

text-figs 19, 20. 

1 947 Diplograptus (Mesograptus) mohawkensis 
Ruedemann; Ruedemaim, pp. 419-20, pi. 71, 
figs 24-6. 

1948 Climacograptus cf. minimus (Carruthers); 
Henningsmoen, pp. 404-5. 

1960 Climacograptus minimus (Carruthers); Berry, 

p. 80, pi. 19, fig. 2. 
71963 Climacograptus minimus (Carruthers); Ross 

and Berry, pp. 125-6, pi. 8, fig. 7. 
1964 Climacograptus minimus (Carruthers); Obut 

and Sobolevskaya, pp. 57-8, pi. 11, figs 8,9. 
1969 Climacograptus minimus (Carruthers); Riva, p. 

521, text-figs 3h-j. 
non 1969 Climacograptus minimus (Carruthers); 

Strachan, p. 191-2, pi. 4, fig. 3, text-figs 4a. 
1977 Climacograptus mohawkensis (Ruedemaim); 

Walters, pp. 937-8, pi. 2, figs f, h, i. 
1982 Climacograptus mohawkensis (Ruedemann); 

Williams, pp. 246-7, figs lOc-j. 
2002 Climacograptus mohawkensis (Ruedemaim 

1912); Rickards, pp. 8-9, fig. 3N. 

Holotype The specimen figured by Ruedemann, 

1947, pi. 71, fig. 24. 

Material About 20 specimens, all indifferently 

preserved with thecal preservation faint. 

Diagnosis Small Climacograptus lacking proximal 

thecal spines but with a short, sharp virgella; proximal 

thecal spacing 12-16 in 10 mm, distally 11-12 in 10 



mm; dorsoventral width proximally 0.65-0.90 mm 
and distally 1.75 mm. 

Description This is a small and inconspicuous 
species with details of thecae difficult to ascertain: the 
apertures appear to be slit-like and hence difficult not 
only to detect, but difficult to distinguish firom "pressed 
through" apertures in flattened specimens like these. 
Consequently the above thecal spacing figures must 
be considered approximate. The virgula is relatively 
long and robust, though not expanded: it is preserved 
in most specimens. The most distinguishing features 
are the parallel-sided nature of a slim rhabdosome 
and the slit-like apertures. 

1 Climacograptus uncinatus Keble and Harris 1934 
Figures 7c, d 

71934 Climacograptus uncinatus sp. nov.; Keble and 
Harris pp. 173-4, pi. 20, figs 5a-c. 

71972 Climacograptus uncinatus, Keble and Harris 
1934; Carter, pp. 48-9, pi. 1, figs 2-7, 10, text- 
figs 2J, L-0. 

Type specimen A type has never been designated. 
Material Only the two specimens figured. 
Remarlcs The thecal details of this form have never 
been ascertained and our material does not help 
much. Two of the specimens, if really referable to 
IC. uncinatus, appear to have almost orthograptid 
thecae, as does one of the Keble and Harris originals 
(1934, pi. 20, fig. 5a). The pair of proximal spines is 
clearest in scalariform views (Fig 7d. herein; Keble 
and Harris 1934, pi. 20, figs 5b, c). In our rnaterial the 
spines are 2.5 mm from the proximal end, but in the 
types they are only 1.5 mm Irom the proximal end. 
In this respect our specimens are closer to the Carter 
(1972) specimens from Idaho than the specimens 
firom Victoria. The Idaho specimens are from the 
linearis Biozone (approximately Eastonian 3) and the 
Victorian specimens firom Bolindian 1 . There is also 
the problem of the relationship, if any, of C7 uncinatus 
to O. quadrimucronatus spinigerus; whether the pair 
of spines in the latter species are thecal spines or 
divisions of the virgula is not known. The questions 
must be raised on to whether uncinatus group has a 
longer range than recorded previously in Australia 
(VandenBerg and Cooper 1992), and whether more 
species are involved than previously supposed. Such 
questions cannot be answered until better material is 
available. 

Genus Orthograptus Lapworth 1 873 
Type species Graptolithus quadrimucronatus Hall, 



Proc. Linn. Soc. N.S.W., 127, 2006 



143 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



1865, by original designation. 

Diagnosis Thecae straight or with slight sigmoidal 

curvature, thecal spines in one (type) group, proximal 

thecal spines common, and large basal spines not 

uncommon. 

Orthograptus quadrimucronatus (Hall 1865) 
Figures 7e, f 



1865 Graptolithus (Diplograptus) quadrimucronatus 
sp. nov.; Hall, J., p. 144, pi. 13, figs 1-10. 

1876 Diplograptus aculeatus Lapworth; Lapworth, 
pi. 2, fig. 44. 

1877 Diplograptus quadrimucronatus Hall; 
Lapworth, p. 133, pi. 6, fig. 20. 

1906 Diplograptus (Orthograptus) 
quadrimucronatus (Hall); Hall, T.S. p. 277, pi. 
34, figs 10, 11. 

1 907 Diplograptus (Orthograptus) 
quadrimucronatus (Hall); Elles and Wood, pp. 
223-4, pi. 28, figs la-d, text-figs 145a-f. 

1908 Glossograptus (Orthograptus) 
quadrimucronatus (Hall); Ruedemann pp. 385- 
92, text-fig. 336. 

1915 Diplograptus quadrimucronatus Hall; Hadding 
pp. 12-3, text-figs 3a-f. 

1947 Glossograptus quadrimucronatus (Hall); 
Ruedemann pp. 452-4, pi. 78, figs 1-5. 

1948 Diplograptus (Orthograptus) quadrimucronatus 
(Hall); Hermingsmoen, pp. 403-4. 

1955 Diplograptus (Orthograptus) 

quadrimucronatus (Hall); Harris and Thomas, p. 

37, pi. 2, figs 37. 
1970 Orthograptus quadrimucronatus (Hall); 

Toghillp. 23,pl. 13, figs 10, 11. 

1982 Orthograptus quadrimucronatus (Hall); 
Williams, pp. 247-248, figs 12a-12d. 

1983 Orthograptus quadrimucronatus (J. Hall); 
Koren' and Sobolevskaya, pp. 152-154, pi. 45, 
figs 1,2,58. 

1987 Orthograptus quadrimucronatus (Hall); 
Mitchell, text-figs 9a-d, 9f-h. 

1991 Orthograptus quadrimucronatus (J. Hall 
1865); Williams, p. 594-5, pi. 2, figs 1-4, figs 
9o-q. 

1 992 Orthograptus quadr. quadrimucronatus (J. 
Hall); VandenBerg and Cooper, p. 82, fig. 9k. 

Type specimen Not designated. Bolton (1960 p. 104) 
listed Geological Survey of Canada, Ottawa, GSC 
1898a, GSC 1898b and GSC 1898d, fi-om the Utica 
Shale east of Pointe Bleue, Lake St. John, Quebec as 
syntypes. 



Material Only two specimens, both figured. 
Diagnosis Wide rhabdosome with dorsoventral width 
in excess of 3 mm within 5 mm of the proximal end 
firom a proximal dorsoventral width of 1.50 mm; 
thecae denticulate and spinose with clear indications 
of more than one spine per theca; thecal spacing about 
14 in 10 mm. 

Remarlcs The thecal apertures appear to be not quite 
so intumed as in the O. calcaratus groups (see below); 
but the presence of spines along the rhabdosome is 
sufficient to distinguish O. quadrimucronatus fi"om 
the O. amplexicaulis group (see below). Specimens 
of O. quadrimucronatus are easily missed because 
biprofile views do not show the spines too well and 
in badly-preserved collections such forms could 
easily be grouped in with O. calcaratus sensu lato. 
The similar species O. whitfieldi is a much narrower 
species. 

Orthograptus calcaratus calcaratus (Lapworth 

1876) 

Figures 8a, b 

1 876 Diplograptus foliaceus Murchison v. 

calcaratus Lapworth; Lapworth pi. 1, fig. 30. 
1 907 Diplograptus (Orthograptus) calcaratus 

Lapworth; Elles and Wood, pp. 239-241, pi. 30, 

figs la-c, text-figs 159a-c. 
1960 Orthograptus calcaratus; Thomas; pp. 12, 19, 

pi. 10, fig. 132. 
1992 Orthograptus calcaratus calcaratus 

(Lapworth, 1876); VandenBerg and Cooper, p. 

82. 
200 1 Orthograptus calcaratus calcaratus 

Lapworth); Rickards et al. p. 82, figs IIH-J. 

Holotype Specimen figured by Elles and Wood 1907, 
pi. 30, fig. lb. 

Material Numerous specimens. 
Diagnosis Robust Orthograptus up to 35 mm long 
and a distal dorsoventral width of 3.20 mm; virgula 
robust; proximal end with three conspicuous spines: 
a virgella, a robust spine on thl' and a spine low on 
thP; thecal apertures very slightly everted proximally 
and more or less horizontal distally; thecal spacing 1 1 - 
14 in 10 mm proximally and 8-10 in 10 mm distally; 
development possibly pattern G of Mitchell (1987). 
Description The sicular aperture is usually visible 
(Fig. 8a) but it is difficult to ascertain which is thl' 
and which thP. If that theca to the right is thl' then 
the virgella is in a strange position, unless the two left; 
hand spines are antivirgellar spines and the virgella 
itself is small or missing. The second alternative 
seems most likely, for a short virgella is visible on 



144 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



the counterpart in the position marked on Fig. 8a by 
dashed lines. Most of the specimens show only three 
proximal spines, including the virgella, as do the other 
subspecies (see below). The thecal apertures are more 
nearly opposite than in many biserial graptolites, 
and in the proximal region they are normal to the 
thecal tube giving a very slightly everted appearance 
on flattening. More distally the apertures become 
horizontal or slightly introverted. 
Remarks Orthograptus calcaratus calcaratus is 
still a little-understood species both in terms of its 
development and in terms of its relationship to several 
described subspecies (see also Rickards et al., 200 Ij. 
Because of consequential identification difficulties the 
known ranges of the subspecies must be considered 
provisional. Considered globally the type subspecies 
seems to range from the clingani Biozone to low in 
the linearis Biozone, that is from Gisbomian 2 to 
Eastonian 3. 



c. priscus (see VandenBerg and Cooper 1992), and it 
occurs in Gisbomian 2 and Estonian 1 . Orthograptus 
calcaratus priscus is thought to be earlier, around the 
gracilis Biozone (approximately Gisbomian 1). We 
do wonder whether there is much difference between 
these two subspecies, and whether our forms, despite 
their very robust proximal end, might not be better 
identified as O. c.lacutus. 

Orthograptus calcaratus cf vulgatus (Lapworth 

1875) 

Figures 8c-e 

cf 1907 Diplograptus (Orthograptus) calcaratus 

var. vulgatus var. nov.; Elles and Wood, pp. 241- 

2, pi. 30, figs 5a-d, text-figs 160a-c. 
cf. 1 992 Orthograptus calcaratus vulgatus 

Lapworth; VandenBerg and Cooper, p. 82, fig. 

8M. 



Orthograptus calcaratus Ipriscus (Elles and Wood 

1907) 

Figure 8f 

71907 Diplograptus (Orthograptus) calcaratus var. 
Ipriscus var. nov.; Elles and Wood, pp. 244-5, 
pi. 30, figs 6a-c, text-fig. 164. 



Type specimen Not designated according to Strachan 
(1971). 

Material A small number of specimens, including 
possible fragments, about 10. 
Diagnosis Strikingly robust form of O. calcaratus, 
proximally with a dorsoventral width at thlVthP of 
1.50 mm (excluding spines) reaching 3.50 mm by 
the 10* thecal pair and widening distally to 4 mm; 
rhabdosomes several cm long; proximal thecal spines 
present; thecal spacing 12-7 in 10 mm. 
Description The proximal end is very robust with a 
"square" appearance and prominent but short spines. 
On Fig. 8f the interrogative marks an area that may 
be a firagment of an adjacent graptolite: even so the 
thecal spine on that side of the rhabdosome may be on 
the third theca. The proximal ends of other specimens 
are less clear still. The virgula is robust and the thecal 
apertures horizontal to gently introverted fi-om the 
start. 

Remarks The main distinguishing feature of this 
form from the almost equally robust O. c. acutus is 
that the proximal end of the latter is less "square" 
and less robust. Distally there is little difference 
between the two. Orthograptus calcaratus acutus 
has been recorded fi-om Australia before, unlike O. 



Type specimen Not yet designated according to 
Strachan (1971). 

Material Five specimens, all figured, including two 
early growth stages. 

Diagnosis Orthograptus calcaratus with virgella and 
two small but conspicuous proximal spines on thl' 
and at the base of thP; proximal end dorsoventral 
width is 1 .40 mm (excluding spines), distally reaching 
in excess of 2.5 mm; thecal spacing proximally 12-16 
in 10 mm, distally 10 in 10 mm. 
Description The virgella is short and spike-like and 
thl' can be seen growing down it a short distance 
before turning upwards and outward, making it 
sometimes rather conspicuous (Fig. 8c). One early 
growth stage (Fig. 8e) shows a sicula with a length 
of 2 mm. The spine on thl' is subapertural and the 
spine associated with thP is either at the base of thl ^ 
or is an antivirgella spine (?one of a pair). The thecae 
are typical of the species as a whole and are either 
slightly everted in appearance or slightly introverted. 
Remarks Orthograptus calcaratus vulgatus ranges 
firom Gisbomian 2 to Eastonian 2 (Table 1). Our 
specimens do not have definite distal parts so we are 
unable to confirm the distal robustness given by Elles 
and Wood for the original material. 

Orthograptus calcaratus aff. tenuicornis (Elles and 
Wood 1907) 
Figures 9b, c 

cf. 1 907 Diplograptus (Orthograptus) calcaratus 
var. tenuicornis, van nov.; Elles and Wood, pi. 
30, figs 4a-c, text-figs 163a,b. 



Proc. Linn. Soc. N.S.W., 127, 2006 



145 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



Type specimen Not yet designated according to 
Strachan(1971). 

Material Five specimens, all figured; some possible 
distal fragments. 

Diagnosis Orthograptus calcaratus with a small 
virgella but with two robust spines, one on thl' and 
one associated withthP; rhabdosomal dimensions as 
type subspecies; thecal spacing 8-10 in 10 mm. 
Description The rhabdosome proximally is possibly 
a little more slender than the type subspecies in the 
Michelago material, having a dorsoventral width at 
thlVthP of 0.75 -1.00 mm and a dorsoventral width 
of 2.10-2.20 mm after 10 mm. Thl' has a spine 
positioned mesially or sub-aperturally and this bends 
downwards after 1 mm to reach a length of up to 3.20 
mm. ThP has a similar spine associated with it, but, as 
in the type subspecies, its base is either in the siculate 
anti-virgellar position or is low down on the free 
ventral wall of the theca. When anti-virgellar spines 
occur in biserial graptolites they are usually as a pair, 
and this is suggested by one specimen AMF 1 14913, 
which certainly has two spines in this position. 
Remarks These forms fit the original Elles and Wood 
(1907) material quite well, except that the spine or 
spines associated with thP seem to be in a different 
position. The specimens figured by Elles and Wood 
(1907 text-fig. 163d, b) clearly have a sub-apertural 
or mesially-positioned spine on thP. Our forms more 
closely resemble the type subspecies, at least in this 
respect. Thomas (1960) recorded O. c. tenuicornis 
from Australia, but VandenBerg and Cooper (1992 
p. 82) considered it more likely to be referable to O. c. 
vulgatus and to O. quadrimucronatus; they regarded 
O. c. tenuicornis as very doubtfiil in Australian 
strata and specimens from Victoria they refer to O. 
thorsteinssoni. The Michelago specimens differ from 
O. thorsteinssoni in having a tiny virgella at similar 
growth stages and, indeed, does not grow a long and 
robust virgella. The general dimensions are similar 
but O. calcaratus aff. tenuicornis is more slender. 

Orthograptus amplexicaulis pauperatus (Elles and 
Wood 1907) 
Figures 9d, e 

1 907 Diplograptus (Orthograptus) truncatus var. 

pauperatus var. nov.; Elles & Wood, p. 237, pi. 

29, figs 5a-d. 
1915 Diplograptus truncatus Lapworth var. 

pauperatus Lapworth mscr.; Hadding, p. 15, pi. 

2, figs 8-11. 
1948 Diplograptus truncatus pauperatus Elles & 

Wood; Henningsmoen, p.403. 
1963 Orthograptus pauperatus Elles & Wood; 



Skoglund, pp. 45-46, pi. 1, fig. 11. 
1970 Orthograptus truncatus pauperatus Elles & 

Wood; Toghill, p. 24, pi. 16, figs 1,2. 
1976 Orthograptus amplexicaulis pauperatus Elles 

& Wood; Erdtmann, pp. 113-114, pi. 4, fig. 

M/4a, b. 

1982 Orthograptus? pauperatus Elles & Wood; 
Williams, p. 251, figs 14a, f, h. 

1983 Orthograptus pauperatus Elles & Wood, 1907; 
Williams and Bruton, p. 181-2, figs 2 IP, 22A-C, 
23E. 

Type species Not designated according to Sfrachan 
(1971) and Williams (1983). 
Material At least 50 specimens. 
Diagnosis Orthograptus amplexicaulis with relatively 
short rhabdosome, up to 30 mm long and with a 
maximum dorsoventral width of 2 mm; thecae simple 
tubes, numbering 10-14 in 10 mm; thl' and thP with 
short spines. 

Description The sicula is faintly visible in some 
specimens and may have a length of about 1.50 
mm. The thecal spacing is usually around 12 in 10 
mm proximally but can reach 14 in 10 mm in a few 
specimens. Distally the spacing is consistently 10 
in 10 mm. Thl' has a small mesial spine and thF a 
submesial spine (but one seemingly well clear of the 
sicular aperture so no confusion with anti-virgellar 
spines arises). Thecal apertures are normal to thecal 
length and thecal overlap approximately one half 
Remarks Orthograptus amplexicaulis is considered 
common in Australia (VandenBerg and Cooper 1992 
p. 82) but has usually been recorded as O. truncatus. 
The same authors cast doubt on previous records of 
the subspecies O. a. pauperatus, but the evidence 
from Michelago seems clear. The subspecies 
considered globally ranges from the middle of the 
clingani Biozone to the linearis Biozone, which is 
approximately Gisbomian 2 to Eastonian 3. 

Orthograptus amplexicaulis intermedius (Elles and 

Wood 1907) 

Figure 9f 

1907 Diplograptus (Orthograptus) truncatus var. 
intermedius var. nov.; Elles and Wood, p. 236, 
pi. 29, figs 4a-c, text-figs 156a, b. 

Type species Not yet designated according to 

Strachan(1971). 

Material One good specimen, figured, and a few 

doubtful fragments. 

Description The rhabdosome reaches a dorsovenfral 

width of 2.50 mm by the 11* thecal pair and 



146 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



thereafter increases very slightly to 2.70 mm. There 
are very long fragments of rhabdosome which may be 
referable to this subspecies but without proximal ends 
attached: these fragments have a dorsoventral width 
of 2.50 - 2.70 mm and a thecal spacing of 10-12 in 10 
mm. The specimen illustrated herein has a proximal 
thecal spacing of 14 in 10 mm and a more distal one 
of 11-12 in 10 mm. The most striking feature of the 
rhabdosome is the relatively high angle of thecal 
inclination (50°-60° distally). Thl' has a sub-apertural 
spine and thP a spine low on the free ventral wall, 
close to the sicula. 

Remarks Orthograptus truncatus intermedius was 
recorded from Ausfralia by Thomas (1960) but this 
was rejected by VandenBerg and Cooper (1992). So 
ours may be the first record of the form from NSW 
and Australia. 

Genus Glyptograptus Lapworth 1873 

Type species Diplograpsus tamariscus Nicholson 
(1868) by original designation. 
Diagnosis (emended Koren' and Rickards 1996) 
Proximal development of tamariscus (I) Pattern: 
thecae with sigmoidal curvature varying from gentle to 
sharp ('climacograptid'); supragenicular wall vertical 
in some, to, more commonly, sloping outwards; 
apertures generally everted but may be horizontal; 
may be septate, aseptate or partially septate; thecal 
and sicular spinosity uncommon; nemal vanes not 
uncommon; sicula usually less than 2 mm long. 

Glyptograptus Jov/e^/ Williams 1982 
Figure 9a 

1982 Glyptograptus daviesi sp. nov.; Williams pp. 
251-2,figsl4b-d. 

Holotype From the clingani Biozone, North Cliff 
trench, Dob's Linn, Southern Uplands, Scotland, 
figured Williams (1982) as 14c. 
Material A single definite specimen and a small 
number of other less well-preserved specimens. 
Description A diminutive Glyptograptus with sharp 
virgella and thread-like virgula and typically gently 
geniculate thecae numbering 15-16 in 10 mm. The 
free venfral wall of thl' is relatively short at 0.50 
mm compared with that of thP at 0.75 mm. The 
down-growing part of thl' is not visible. Thecal 
apertures are more or less normal to the thecal length. 
Overlap cannot be seen. The proximal dorsoventral 
width is 0.90 mm and by the seventh thecal pair the 
dorsoventral width is 1 .40 mm. 
Remarlcs The best specimen is identical to those 



recorded by Williams (1982) from Southern 
Scotland and is a first record for Australia. 



ACKNOWLEDGEMENTS 

PLW would like to thank Jennifer Zicker for help in the 
field and RBR thanks the Department of Earth Sciences 
at Cambridge and the Royal Society for support. 



REFERENCES 

Berry, W.B.N. 1960. Graptolite faunas of the marathan 

region, West Texas. University of Texas Publications, 

6005, 1-179. 
Bjerreskov, M. 1987. Discoveries on graptolites by X-Ray 

Studies. Acta Palaeontologica Polonica, 23, 463-471. 
Bolton, I.E. 1960. Catalogue of type invertebrate fossils 

of the Geological Survey of Canada, Volume 1. 

Geological Survey of Canada, Ottawa. 
Bronn, H.G. 1849. Index Palaeontologicus B, Enumerator 

palaeontologicus. Stuttgart, E. Schweizerbartsche, 

1-980. 
Bulman, O.M.B. and Rickards, R. B. 1966. A Revision of 

Wiman's Dendroid and Tuboid Graptolites. Bulletin 

of the Geological Institutions of the University of 

Uppsala, 43, 1-72. 
Carruthers, W. 1868. Graptolites: their structure and 

position. Intellectual Observer, 11, 283-292, 365-374. 
Carter, C. 1972. Ordovician (Upper Carodocian) 

Graptolites fi'om Idaho and Nevada. Journal of 

Paleontology 46, 43-49. 
Carter, C. and Churkin, M. 1977. Ordovician and Silurian 

graptolite succession in the Trail Creek area, 

central Idao - a graptolite zone reference section. 

Professional Papers of the U.S. Geological Survey, 

1020, 1-33. 
Crowther, RR., Rickards, R.B. and Urbanek, A. 1987. 

Graptoblast zooidal tissue and a review of graptolite 

soft parts. Geological Magazine, 124, 67-72. 
Davies, K.A. 1929. Notes on the graptolite faunas of the 

Upper Ordovician and Lower Silurian. Geological 

Magazine, 66, 1-27. 
Elles, G.L. and Wood, E. M. R. 1901-18. Monograph 

of British Graptolites. Palaeontographical Society 

(Monograph), i-clxxi, a-m, 1-536. 
Erdtmann, B.D. 1976. Die graptolithen fauna der Exploits 

Gruppe (Oberes Ordovizium, Caradoc) von Zentral 

Neufundland. Sonderabdruck aus Mitteil Geologisch- 

Paldontologischen Institut der Universitdt Hamburg, 

45, 65-140. 
Gurley, R.R. 1896. North American graptolites: new 

species and vertical ranges. Journal of Geology, 4, 

63-102. 
Hadding, A. 1913. Undre Dicellograptusskiffem i Skane. 

Lunds Universitet Arsskrift, new series, 9, 1-91. 



Proc. Linn. Soc. N.S.W., 127, 2006 



147 



EASTONIAN GRAPTOLITES FROM MICHELAGO 



Hadding, A. 1915. On Glossograptus, Cryptograptus och 
trenne dem narstaende graptolitstlakten. Geologiska 
Foreningens i Stockholm Forhandlingar, 37, 303-336. 

Hall, J. 1847. Descriptions of the organic remains of the 
lower division of the New York system. Paleontology 
of New York, 1, 1-338. Albany. 

Hall, J. 1865. Graptolites of the Quebec Group. Canadian 
Organic Remains. Geological Survey of Canada, 2, 
i-iv, 1-151. 

Hall, T.S. 1902. The graptolites of New South Wales in the 
collection of the Geological Survey. Records of the 
Geological Survey of New South Wales, 7, 49-59. 

Hall, T.S. 1906. Report on Graptolites. Records of the 
Geological Survey of Victoria, 1, 275. 

Harris, W. J. and Thomas, D.E. 1955. Victorian 

Graptolites, pt xiii. Graptolites from the Wellington 
River, part 1 . Mining and Geological Journal of 
Victoria, Department of Mines, 5, 35-45. 

Henningsmoen, G. 1948. The Tretaspis Series of the 
KuUatorp Cove. In Waem, B. Thorslund, R and 
Henningsmoen, G. Deep boring through Ordovician 
and Silurian strata at Kinnekulle, Vestergotland. 
Bulletin of the Geological Institute of the University 
of Uppsala, 32,374-432. 

Hopkinson, J. 1 87 1 . On Dicellograptus a new genus of 
Graptolite. Geological Magazine, 8, 20-26. 

Keble, R. A., and Harris, W.J. 1925. Graptolites from Mt. 
Eastern. Geological Survey of Victoria Records, 4, 
507-516. 

Keble, R. A., and Harris, W.J. 1934. Graptolites of 
Victoria, new species and additional records. 
National Museum of Victoria, Melbourne, Memoir, 8, 
166-183. 

Koren', T. N. and Rickards, R. B. 1996. Taxonomy and 
evolution of Llandovery biserial graptoloids from the 
Southern Urals, Western Kazakhstan. Special Papers 
in Palaeontology, 54, 1-103. 

Koren', T.N., and Sobolevskaya, R.F. 1983. Graptolites. 
In The Ordovician and Silurian Boundary in the 
Northeast of the USSR. (In Russian). Ed. Sokolov 
B.S. et al. Nauka Publishers, Leningrad, (St. 
Petersburg), 97-160. 

Kozlowski, R. 1949. Les graptolithes et quelques 
nouveaux groups d'animaux du Tremadoc de las 
Pologne. Palaeontologica Polonica, 3, 1-235. 

Lapworth, C. 1873. On an improved classification of the 
Rhabdopora. Geological Magazine, 10,500-504, 
555-60. 

Lapworth, C. 1875. Descriptions of graptolites of the 
Arenig and Llandeilo rocks of St. Davids. Quarterly 
Journal of the Geological Society of London, 31, 
631-672. 

Lapworth, C. 1876. On Scottish Monograptidae. 

Geological Magazine, 23, 308-321, 350-360, 499- 
507, 541-52. 

Lapworth, C. 1877. On the graptolites of County Down. 
Appendix (107-23) in Swanston, W. On the Silurian 
rocks of the County Down. Proceedings of the Belfast 
Naturalists Field Club 1876-1877, 107-147. 



Lloydell, D.K., Orr, P.J., and Keams, S. 2004. Preservation 

of soft tissues in Silurian Graptolites from Latvia. 

Palaeontology, 47, 503-513. 
Mitchell, C. E. 1987. Evolution and plylogenetic 

classification of the Diplograptacea. Palaeontology, 

30, 353-405. 
Moors, H.T. 1969. On the first occurrence of a 

Climacograptus bicornis with a modified basal 

assemblage, in Australia. Proceedings of the Linnean 

Society of New South Wales, 93, 227-231. 
Nicholson, H.A. 1867. Graptolites of the Moffat Shale. 

Geological Magazine, 4, 108-11. 
Nicholson, H.A. 1868. On the graptolites of the Coniston 

Flags; with notes on the British species of the genus 

graptolites. Quarterly Journal of the Geological 

Society, London, 24, 521-545. 
Obut, A.M. and Sobolevskaya, R. F. 1904. Graptolity 

Ordovika Taimyra. Moscow, Akademia Nauk, SSSR, 

1-92. 

Oldershaw, W 1965. Geological and geochemical 

survey of the Captains Flat area. New South Wales. 

Bureau of Mineral Resources, Australia. Report, 101, 

55pp. 
Richardson, S.J. 1979. Geology of the Michelago 1:100 

000 sheet 8726. 253pp. Geological Survey of New 

South Wales, Sydney. 
Richardson, S.J. and Barron, L. 1977. Michelago 1:100 

000 Geological Sheet 8726. Geological Survey of 

New South Wales, Sydney. 
Richardson, S.J. and Sherwin, L. 1975. Early Silurian 

Graptolites near Bredbo. Quarterly Notes of the 

Geological Survey of New South Wales, 21, 17-19. 
Rickards, R. B. 2002. The graptolitic age of the type 

Ashgill Series, (Ordovician), Cumbria, UK. 

Proceedings of the Yorkshire Geological Society, 54, 

1-16. 
Rickards, R.B., Sherwin, L. and Williamson, RL. 2001. 

Gisbomian (Caradoc) graptolites from New South 

Wales, Ausfralia: systematics, biostratigraphy and 

evolution. Geological Journal, 36, 59-86. 
Rickards, R.B. and Stait, B. 1984. Psigraptus, its 

classification, evolution and zooid. Alcheringa, 8, 

101-111. 
Riva, J. 1969. Middle and Upper Ordovician graptolite 

faunas of St. Lawrence lowlands of Quebec, and of 

Anticosti Island. American Association of Petroleum 

Geologists, Memoir, 12, 513-556. 
Riva, J. 1988. Graptolites at and below the Ordovician 

- Silurian boundary on Anticosi Island. Bulletin of the 

British Museum Natural History, 43, 221-237. 
Riva, J. and Kettner, K.B. 1989. Ordovician graptolites 

from the northern Sierra de Cobachi, Sonora, Mexico. 

Transactions of the Royal Society of Edinburgh, 80, 

71-90. 
Ross, R.J. and Berry, W.B.N. 1963. Ordovician graptolites 

of the Basin Ranges in California, Nevada, Utah and 

Idaho. United States Geological Survey Bulletin, 

1134, 1-77. 



148 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 



Ruedemann, R. 1908. Graptolites of New York, pt. II. 

Graptolites of the higher beds. Memoir New York 

State Museum and Science Service, 11, 1-228. 
Ruedemann, R. 1912. The Lower Siluric shales of the 

Mohawk Valley. New York State Museum Bulletin, 

162, 1-151. 
Ruedemann, R. 1947. Graptolites of North America. 

Geological Society of America, Memoir, 19, 1-652. 
Ruedemann, R., and Decker, C. 1934. The Graptolites 

of the Viola Limestone. Journal of Paleontology, 8, 

303-327. 
Skoglund, R. 1963. Uppermost Viruan and Lower Harjuan 

(Ordovician) stratigraphy of Vastergoland and 

Lower Harjuan graptolite faunas of central Sweden. 

Bulletins of the Geological Institute of the University 

of Uppsala, 42, 1-55. 
Strachan, I. 1969. A redescription of W. Carruther's type 

graptolites. Bulletin of The British Museum (Natural 

History), 17, 183-206. 
Strachan, I. 1971. A synoptic supplement to "A 

monograph of British Graptolites by Miss G. L. 

Elles and Miss E. M. R. Wood." Monographs of the 

Palaeontographical Society, London, 1-130. 
Thomas, D. E. 1960. The zonal distribution of Australian 

graptolites. Journal and Proceedings of the Royal 

Society of New South Wales, 94(1), 1-58. 
Toghill, R 1970. Highest Ordovician (Hartfell Shale) 

graptolite faunas from the Moffat area, south 

Scotland. Bulletin of the British Museum (Natural 

History), 19(1), 1-27. 
VandenBerg, A.M. 1990. The ancestry of Climacograptus 

spiniferus Ruedemann. Alcheringa, 14, 39-51. 
VandenBerg, A.H.M. 2002. The Victorian Ordovician 

Graptolite Succession. In VandenBerg, A.H.M. et al., 

(Eds). First International Palaeontological Congress, 

Post Congress Field Excursion Guide 2, 41-53. 
VandenBerg, A.M. and Cooper R.A. 1992. The Ordovician 

graptolite sequence of Australasia. Alcheringa, 16, 

33-85. 
Walters, M. 1977. Middle and Upper Ordovician 

graptolites from the St. Lawrence lowlands, Quebec, 

Canada. Canadian Journal of Earth Sciences, 14, 

932-952. 
Williams, S.H. 1981. Upper Ordovician and lowest 

Silurian graptolite biostratigraphy in southern 

Scotland. Ph D Thesis, University of Glasgow. 
Williams, S.H. 1982. Upper Ordovician graptolites from 

the top lower Hartfell Shale Formation (£>. clingani 

and P. linearis zones) near Moffat, southern Scotland. 

Transactions of the Royal Society of Edinburgh, 72, 

229-55. 
Williams, S.H. 1983. The late Ordovician graptolite 

fauna of the Anceps Bands at Dob's Liim, southern 

Scotland. Geologica et Palaeontologica, 16, 29-56. 
Williams, S.H. 1987. Upper Ordovician graptolites from 

the D. complanatus zone of the Moffat and Girvan 

districts and their significance for correlation. Scottish 

Journal of Geology, 23, 65-92. 



Williams, S.H. 1991. Stratigraphy and graptolites of the 
Upper Ordovician Point Leamington Formation, 
central Newfoundland. Canadian Journal of Earth 
Sciences, 2%, 5%\-6Q0. 

Williams, S.H. 1994. Revision and definition of the C. 
wilsoni graptolite zone (Middle Ordovician) of 
southern Scotland. Transactions of the Royal Society 
of Edinburgh, 85, 143-157. 

Williams, S.H. and Bruton, D.L. 1983. The Caradoc- 
Ashgill boundary in the central Oslo Region and 
associated graptolite faunas. Norsk Geologisk 
Tidsskrift, 63,147-191. 



Proc. Linn. Soc. N.S.W., 127, 2006 



149 



EASTONIAN GRAPTOLITES FROM MICHELAGO 




Figure 3 a-e Leptograptus flaccidus cf. macer EUes and Wood, respectively AMF114895, 114938, 114939, 
114934, 114892; f-i Leptograptus ?flaccidus spinifer EUes and Wood, respectively AMF 114942, 114886, 
114887, 114941; scale bars 1 mm. 



150 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 




Figure 4 a-c Dicellograptus morrisi Hopkinson, respectively AMF 114925, 114947, 114910; d Dicello- 
graptus of. caduceus Lapworth, AMF 114922-3, specimens adjacent on same slab; scale bars 1 mm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



151 



EASTONIAN GRAPTOLITES FROM MICHELAGO 




Figure 5 a, b Dicellograptus cf. caduceus Lapworth, respectively AMF 114958, 114924; c,d Dicellograp- 
tus sp., respectively AMF 114903, 114891; scale bars 1 mm. 



152 



Proc. Linn. Soc. N.S.W, 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 




Figure 6 a-d Climacograptus caudatus Lapworth, respectively AMF 114904, 114906, 114949, 114905; 
e-h Climacograptus tubuliferus Lapworth, respectively AMF 114896, 114900, 114897, 114899; scale bars 
1 mm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



153 



EASTONIAN GRAPTOLITES FROM MICHELAGO 




a b 



Figure 7 a, b Climacograptus sp., respectively 114952, 114919; c,d Climacograptusl uncinatus Keble 
and Harris, respectively AMF 114950, 114889; e,f Orthograptus quadrimucronatus (J. Hall), respectively 
AMF 114916, 114917; g 1 Climacograptus lanceolatus VandenBerg; AMF 114959; h,i Climacograptus 
mohawkensis (Ruedemann), respectively AMF 114911, 114915; scale bars 1 mm. 



154 



Proc. Linn. Soc. N.S.W., 127, 2006 



P.L. WILLIAMSON AND R.B. RICKARDS 




Figure 8 a,b Orthograptus calcaratus calcaratus (Lapworth), AMF 114920 respectively proximal and 
distal parts of a long specimen; c-e Orthograptus calcaratus cf. vulgatus (Lapworth), respectively AMF 
114945, 114956, 114951; i Orthograptus calcaratus Ipriscus (EUes and Wood), AMF 114937; scale bars 1 
mm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



155 



EASTONIAN GRAPTOLITES FROM MICHELAGO 




Figure 9a Glyptograptus daviesi Williams, AMF 114946; b,c Orthograptus calcaratus aff. tenuicornis 
(EUes and Wood), respectively AMF 114933, 114888; d, e Orthograptus amplexicaulis pauperatus (EUes 
and Wood), respectively AMF 114901, 114898; f Orthograptus amplexicaulis intermedius (EUes and 
Wood), AMF 114893; scale bars 1 mm. 



156 



Proc. Linn. Soc. N.S.W., 127, 2006 



The Geomorphology and Hydrology of Saline Lakes of the 
Middle Paroo, Arid-zone Australia. 



Brian V. Timms 

School of Environmental and Life Sciences, University of Newcastle, Callahgan, NSW 2308, Australia. 

Email: brian.timms(S),newcastle. edu.au 



Timms, B.V. (2006). The geomorphology and hydrology of saline lakes of the middle Paroo, arid-zone 
Australia. Proceedings of the Linnean Society of New South Wales 127: 157-174. 

Sixteen subsaline (0.5 - 3 gL"') and saline lakes (> 3 gL') of the Paroo have been studied for periods of 
up to 18 years. Many were formed by drainage routes being blocked by dunes, some lie in dune swales, 
some lie at the edge of the Paroo floodplain where alluvial sediments are thinner, and Lake Wyara lies 
in a depression on a fault line. All developed further by deflation and owe their form to wind-induced 
currents and wave action shaping shorelines. Most saline lakes have lunette dunes on the eastern shore, and 
many larger ones have migrated westwards. Lakes of low salinity have sandy beaches and no, or poorly 
developed, lunettes. Lakes with N-S axes have the southeastern comer cut off by spits generated by currents 
induced by northwesterley winds. A few lakes are filling with sediment derived fi-om the overgrazing of 
catchments associated with European settlement. 

Larger lakes with inflowing streams fill in El Nino years, then dry over the next few years. Smaller lakes 
without surface inflows may fill a few times in wet years but dry quickly. Most lakes remain dry in La 
Nina years. Salinity regimes fluctuate widely and, while instantaneous faunal lists may be depauperate, 
cumulative species lists can be long. However, lakes which normally are fresh, but become saline in their 
final stage of drying, develop only a limited saline lake fauna. 

Manuscript received 27 July 2005, accepted for publication 7 December 2005 

KEYWORDS: biodiversity, El Nino, lake compartmentilisation, lake migration, lake origins, lake 
sedimentation, lunette dunes, saline lakes, spits. 

partly explored at the large scale (lake areas > 1 00 km^ 

INTRODUCTION ) on Lake Eyre (Kotwicki, 1986) and its predecessor 

Lake Dieri (de Vogel et al, 2004), on Lake Victoria in 

In most hot arid lands, geomorphic processes and southwestern New South Wales (Gill, 1 973 ; Lees and 

resultant landforms are dominated by wind action on Cook, 1991; Chen, 1992), and on lakes of Salinaland 

unconsolidated surfaces (Thomas, 1989). Therefore in Western Australia (Van de Graaf et al., 1977). The 

depressions and their lakes are likely to owe their SLEADS program on large salinaplayas in Australia's 

origin to aeolian processes, or at least have their arid and semi-arid inland (Chivas and Bowler, 1986), 

basins and shorelmes modified by wind. Furthermore, besides its main aim of interpreting past climates firom 

because drainage is often uncoordinated, most lakes lake sediments and lunettes, has confirmed the role of 

are closed hydrologically (Cole, 1968, 1983), so that wind in lake basin evolution. Besides these studies on 

saline waters abound. Lakes fill and dry intermittently large salinas, much can also be learnt fi-om comparative 

(Williams, 1984), either seasonally or episodically studies on smaller lakes (A< 50 km^ often < 5 km^ 

according to prevailmg climate. The extent of ) of a confined area where the hydrological pattern 

filling is influenced by the interaction between is known. The middle Paroo of northwestern New 

rainfall, evaporation, lake basin geomorphology and South Wales and southwestern Queensland has many 

hydrological character of the catchment. In total, the small saline lakes and a few freshwater lakes (Fig. 1) 

geomorphology and hydrology of arid-zone lakes, that salinise as they dry, and moreover hydrological 

particularly if saline, are likely to be distinctive. data covering many years (up to 18) are available for 

In the Australian context, these issues have been most. It is the aim this contribution to explore role 



SALINE LAKES OF THE MIDDLE PAROO 




SLAND 



BOURKE 



WILCANNIA 



Figure 1. Map of the Paroo catchment, southwestern Queensland and northwestern New South Wales. 
The location of most of the lakes mentioned in the text are shown. 



158 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



of hydrology and geomorphology in the Hmnology of 
the Paroo lakes, as well as the significance of wind 
action for determining lake basin process and form. 



METHODS 

Most of the middle Paroo study lakes are 
closed hydrologically, so that water levels fluctuate 
according to the balance between precipitation and 
evaporation, both on the lake basin and its catchment. 
Evenso, each lake generally has a distinct shoreline 
visible on an aerial photograph to which it has filled 
many times. This was designated the 'full' level and 
used as the lake outline on the accompanying maps. 
Occasionally, perhaps once in 20-100 years, a lake 
may fill to a greater depth, as Lake Wyara (Fig. 1) has 
done four times in the last 110 years (Timms, 1998a); 
such fillings are not accounted for geomorphogically 
in this study (i.e. shorelines, areas and depths refer to 
normal 'fiall' conditions, unless noted otherwise). 

Lakes (Fig. 1) were mapped when dry using a 
dumpy level, often fitted with laser technology. In 
small lakes a cart-wheel system of transects were 
used, with the dumpy in the deepest part of the lake 
and measurement lines radiating outwards at 25 
to 35° intervals and readings taken every 10-50 m, 
depending on lake size and change in elevations. If 
transect lines were longer than 250 m (e.g. Lower 
Bell Lake, Gidgee Lake, Lake Burkanoko), subsidiary 
lines were used commencing 100 - 250m from the 
central pivot point and radiating out at 15 to 25° 
angles, so that the shoreline was intercepted regularly 
at intervals of 25 -100 m, depending on lake size and 
lakebed irregularities. Some lines crossed each other 
and hence provided checks on elevations. In larger 
lakes (e.g. Lake Yumberarra, North Blue, Taylors 
Lake) cartwheels were used at each end and parallel 
transects in between with some lines crossing for 
checks on accuracy. This method enabled contours 
with an accuracy of ±1 cm or better to be drawn. 
Contour intervals of 10 to 50 cm were adopted, 
though occasionally intervals as low as 2.5 cm were 
employed. In some lakes (Lower Bell, Gidgee, 
Burkanoko and Barakee) it was easy to detect new red 
clayey sediments on older white gypseous surfaces, 
so it was possible to collect data on recent sediment 
depths at the same time as surface elevations were 
being recorded. 

Three lakes (Lakes Wyara, Numalla and 
Horseshoe) were too big to be mapped efficiently by 
these methods, so analyses are restricted to shoreline 
features. There were also problems mapping Mid 
Blue Lake (namely, cross correlation of transects), so 



a detailed map of this lake is not available. 

The lakes were visited at varying intervals 
between August 1987 and June 2005, more often 
in wet years (e.g. eight times in 1998) and rarely in 
lingering drought years (e.g. twice in 2004). On each 
visit, lake levels were noted and salinity (i.e. TDS) 
determined by gravimetry. Between visits, further 
information on water levels in most lakes was gained 
from local landowners. Rainfall data from Warroo 
Station (Fig. 1), in the northern part of the study area, 
was used as representative for the study area, though 
it varied monthly by up to 26% and yearly by 15% 
from figures for individual station properties with 
lakes included in this study. 

Although this paper is concerned mainly with 
geomorphology and hydrology, some biological 
data on salinising freshwater lakes were collected. 
Methods used were as described in Timms (1998a) 
and Timms and McDougall (2005). 



RESULTS 

Rainfall 

Yearly rainfall at Warroo Station fluctuated 
between 70.5 mm in 2002 to 685 mm in 2000 (Fig. 
2), both near records for Warroo (Rand M. Dunk, 
pers. com.), with 1998-2000 well above the 76- 
year average of 301 mm and 2001-2004 well below. 
Rainfall events > 100 mm in a few days, of the kind 
that fills lakes, occurred in December 1987 (108 mm), 
April 1988 (103 mm), May 1989 (122 mm), April 
1990 (265 mm), January 1995 (192 mm), January 

1998 (162 mm), June 1998 (163 mm),~ November 

1999 (119 mm). May 2000 (253 mm), and November 

2000 (217 mm). The exceptionally wet years had 
positive Southern Ossication Indexes (hereafter SOIs 
and based on monthly fluctuations in air pressure 
differences between Tahiti and Darwin - Bureau of 
Meteorology, website). Thus from May 1998 to April 

2001 all monthy SOIs were positive except two and 
for 2000 the average was 7.6), whereas during the dry 
years of 2001-2004 SOIs were almost continuously 
negative for 44 months and with a 2002 average of 
-6.1 (Bureau of Meteorology, website). Generally 
these rain events, and some outside the study area 
(the 'dry floods'), caused moderate to major flooding 
in the Paroo which contributed to the filling of two 
of the study lakes, Numalla and Wombah. The most 
recent inflows into Numalla and Wombah were in 
November 2000 (major) and January 2004 (minor). 



Proc. Linn. Soc. N.S.W., 127, 2006 



159 



SALINE LAKES OF THE MIDDLE PAROO 



800 
700 
600 



— 


v500 


(0 




>f> 




(0 


400 


Q^ 






300 


(0 




3 




C 


200 


C 




< 


100 









87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 



Figure 2. Variation in annual rainfall 1987 -2004 at Warroo Station, middle Paroo. Three year noioving 
average shown by solid line. 



Werewilka Ck. .\ n 



Benanga Ck 




Youlainge Ck. &"= -i"- 



/ / /beach 



ridges 



Kaponye Ck. 



Figure 3. Lake Wyara showing the main inflow- 
ing creeks, beach ridges and the depositional area 
(stippled) behind Pelican Island. After Timms 
(1998a). 



Lake Wyara 

Lake Wyara is the largest of the lakes studied, 
with an area of 3400 ha. It is D-shaped with the 
longest axis N-S of 8.5 km and width 4.5 km (Fig 
3). The eastern shoreline is evenly curved, with well 
developed beaches and spits at each end largely 
occluding the mouths of the two major inflowing 
creeks. There is an ancient lunette, 400-700m east 
of the average shoreline, which is hardly visible on 
the ground but noticeable on satellite images. The 
western shore is irregular but smoothed somewhat 
with offshore islands which are inundated when water 
levels are high but connected to the mainland at low 
levels (the large island to the southwest is connected 
at average 'lakefuU' stage while Pelican Island is 
isolated (Fig. 3). Details of beaches and islands are 
given in Timms (1998a). The catchments of Benanga 
and Youlainge Creeks to the middle west of the lake 
are severely eroded so that much clayey soft sediment 
has been deposited recently behind the islands (i.e. 
over the last few decades including during 1987-1996 
when the lake was visited regularly - Timms, 1998a). 
The deepest area is ca 750m north of the southern 
shore; depth fluctuates widely, often up to ca 2.6m, 
sometimes to ca 4m, and rarely to ca 6.9 m, at which 
level it overflows (see Timms, 1998a, for details). 

Lake Wyara fills from its own catchment (mainly 
from Werewilka Ck) and occasionally overflows via 
Kaponyee Creek to the Paroo River. It holds water 
most of the time (Fig. 4) but dries in moderate to 
major droughts and has overflowed just four times 



160 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



88 89 90 


91 


92 


93 


94 95 


96 97 98 99 00 

1 1 1 1 1 


01 02 


03 


04 


Wyara 3 




15 


350 


3 


34 


12 


41 


9 


13 


Numalla 2.5 










2.5 


3 


4 


Yumberarra 






131 


1 

<1 
,11 1 
,16 1 
,59 2 

7I 
3 88 

5 96 

24 12 

30 


4 13 1 
1 


<1 


4 114 




1 

2 
4 
18 
2 


Karatta 
North Blue 
Mid Blue 
Bulla 
Wombah 
Gidgee 5 


? 

? 

? 

7 
7 

32 

31 
301 




17 2 


<1 


31 


40 2 
30 193 2 
26 2 
19 17 3 
22 28 
1 164 47 
171 _209 


2 102 
11 25 158 
2 30 
13 19 182 

26 192 
44 350 
89 69 


Low Bell 8 


Horseshoe 
Bells Bore 29 



Figure 4. Comparsions 
(blanks) of Paroo lakes. 



of wet (black line) and dry periods 
Some salinities (TDS in gL') are given. 



in the last 1 1 8 years. Details are provided in Timms 
(1998a). Periods of being full and dry are strongly 
correlated with pulses of rainfall-drought explained 
by the SOI (r = 0.622, p> 0.001, n =34). Salinities 
vary greatly from almost fresh to crystallising brine. 

Lake Numalla 

Lake Numalla is the second largest lake (A = 
2900 ha, Timms, 1999, 2001a) of the middle Paroo 
(Figs 1 & 5). It lies near the edge of the Paroo 
floodplain along Boorara Ck and is connected to the 
main river by a distributary channel of Carwarra Ck. 



Boorara Ck 



Northwest 
Arm 



Public 
Beach 




Carwarra Ck 



Figure 5. Lake NumuUa showing beaches (A), 
spits (B), major occluded bays (C), and minor oc- 
cluded bays (D). 



Shorelines are sandy everywhere 
and usually gently shelving, but 
there are parallel beach ridges on 
the southern and eastern shores 
(marked A in Fig. 5), the inner 
beach inundated at higher water 
levels. Major spits occur at sudden 
changes in shore orientation 
(B on Fig. 5) and in two places 
these almost occlude two large 
backwaters, the Northwest Arm 
and a lakelet north of the Public 
Beach (C on Fig. 5). Smaller 
sandy spits partially cut off a few 
small bays and an incipient spit 
north of The Point is building out 
from the northeast, but has only 
partially occluded this comer of 
lake (D on Fig. 5). The lake is 6.5 
m deep when full; when levels are low, as in 2002-05, 
creek inlets are dry, the Northwest Arm drying first, 
followed by Carwarra Ck. There is no lunette dune 
associated with this lake. 

Lake Numalla held water throughout the study 
period (but dried in mid 2005) and besides receiving 
local runoff via the three northern arms, its main 
source of water comes fi-om Paroo 'freshes', which 
reach the lake via Carwarra Ck. Water in the lake is 
generally subsaline (0.5 - 3 gL"'), but at low water 
levels, salinity increases to hyposaline conditions 
(Fig 4) and finally becomes hypersaline (L. Fabbro 
in Hobson et al., 2005, recorded a conductivity of 
104,000 jLiS/cm in May 2005). Inflowing water is of 
very low salinity (<100 ^.S/cm) and mixes poorly 
with incumbent water because of the embayments in 
the lake, so salinity can vary spatially (see Timms, 
1997a). 

Lake Numalla supports abundant waterbird 
and turtle populations (Kingsford and Porter, 1994; 
Hobson et al., 2005), though the invertebrate fauna is 
neither rich nor abundant compared with other lakes 
in the area (Table 1 cf Hancock and Timms, 2002; 
Timms, 2001b; Timms and Boulton, 2001; Timms 
and McDougall, 2005). As the lake naturally salinised 
between 2002-2005, the invertebrate fauna became 
less diverse and dominated by salt-tolerant species 
together with some typical saline lake species (Table 

!)• 

Lake Yumberarra 

This lake is a triangular-shaped, 170 ha in 
area and 3.4 m deep when fiill (Fig. 6). It lies in a 
depression in Quaternary alluvium at the edge of the 
Paroo floodplain. It is fed by Paroo floodwater via 



Proc. Linn. Soc. N.S.W, 127, 2006 



161 



SALINE LAKES OF THE MIDDLE PAROO 



Table 1. Invertebrates in Lake Numalla. Code: xxx = often abundant; xx = common or present often; x 
= present occasionally; r = found sometimes in smaU numbers. 



Years 


1995-2001 


2002 


JuiOa & Feb04 NovOS &Nov04 


Conductivities (mS/cm) 


<3.5 


3.6-4.2 


6.4-9.6 13-17 


Number of lake visits 


n=60 


n=12 


n=6 n=6 



Species 

Boeckella tharticulata Tliomson 

Calamoecia Canberra Bayly 

Calamoecia lucasi Brady 

Apocyclops dengizus Lepeschkin 

Metacyclops sp. 

Mesocyclops cf woutersi Van de Velde 

Cletocamptus deitersi Richard 

Diaphanosoma unguiculatum Gurney 

Moina australiensis Sars 

Moina micrura Kurz 

Bosmina meridionalis Sars 

Daphnia carinata s. I. King 

Ceriodaphnia cornuta Sars 

chydorids (mainly Alona spp.) 

Heterocypris sp. 

Mytilocypris splendida (Chapman) 

Asplanchna sieboldi (Leydig) 

Brachionus calyciflorus Pallas 

Brachionus ibericus Ciros-Perez et al. 

Filinia australiensis Koste 

Filinia cfpejieri Hutchinson 

Hexarthra sp. 

Keratella sp. 

Macrobractiium australiense Holthuis 

Ctierax destructor Clark 

Cloeon sp. 

Tasmanocoenis tillyardi (Lestage) 

Xanttioagrion erytlironeurum Selys 

Diplacoides spp. 

Hemianax papuensis (Burmeister) 

IHemicordulia tau (Selys) 

Austrogomptius sp. 

Agraptocorixa eurynome Kirklady 

Agraptocorixa parvipunctata Hale 

Micronecta sp. 



XX 
X 

xxx 



XXX 

XX 
X 



XX 



xxx 



X 


X 






X 










X 


XX 




XX 


XX 






r 








r 








r 








r 


r 








r 


X 


XX 






X 


X 


X 


X 


X 


X 


X 


XXX 


X 


X 






XX 


XX 






X 





X 








X 








XX 


XX 






r 








X 








X 








X 








r 








r 




r 




r 




r 




X 








XX 


XX 


XX 


XX 


X 


X 




X 


xxx 


xxx 


xxx 


xxx 



162 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



Table 1 Continued: Invertebrates in Lake Numalla. Code: xxx = often abundant; xx = common or 
present often; x = present occasionally; r = found sometimes in small numbers. 



Years 


1995-2001 


2002 


JulOS & Feb04 Nov03 &Nov04 


Conductivities (mS/cm) 


<3.5 


3.6-4.2 


6.4-9.6 13-17 


Number of lal<e visits 


n=60 


n=12 


n=6 n=6 


Species 








Anisops calcaratus Hale 


XX 


XX 


X X 


Anisops gratus Hale 


XX 


XX 


XX XX 


Anisops thienemanni Lundbald 


X 


X 


X 


Ranatra dispar Montandon 


r 






Naucoris congrex Stal 


r 






Limnogonus sp. 


r 






Oecetis sp. 


r 






Triplectides australicus Banks 


r 






Allodessus bistrigatus (Clark) 


r 






Antiporus gilberti Clark 


r 




r 


Berosus munitipennis Blackburn 


r 






Berosus australiae Mulsant 


r 


r 


r 


Enochrus eyrensis (Blackburn) 


r 




r 


Hydaticus christi Nilsson 




r 


r 


Rhantus suturalis (W. MacLeay) 


r 






Sternopriscus multimaculatus (Sharp) 






r 


unident. tanypodine chironomid 


X 




X X 


unident. chironomini chironomid sp. a 


r 




XX X 


unident. chironomini chironomid sp. b 


X 


X 




Chironomus sp. 


X 


X 


X 


unident. ceratopogonind larva 


X 




X 


unident. tabanid larva 


r 






Arrenurus sp. 


X 




r 


Elyais sp. 


X 




r 


Corbiculina sp. 


X 


r 




Alathyria sp. 


r 







Carwarra Ck. and/or local runoff via Stinking Well 
Ck. When full, water exits via an outflow to Six Mile 
Creek to the Paroo and/or back along Carwarra Ck 
(see Timms, 1999 for details). A well developed 
spit of decreasing height southwards, cuts off the 
southeastern comer totally (at 0.5 m depth) to partially 
(at 2 m depth). No enhanced sedimentation in the 
main lake was detected. A lunette only 1.5 m higher 
than the fiill shoreline flanks the eastern shore. 



Lake Yumberarra had three filling-drying cycles 
during the 1 7 years of study. It usually fills fi-om Paroo 
floods, but can fill from local runoff, as it did in July 
1998 (see Timms and McDougall, 2005). The lake 
is usually fresh, but it naturally salinises as it dries. 
During such periods it gains some saline species, 
but some salt-tolerant fi-eshwater species persist (see 
Timms and McDougall, 2005). 



Proc. Linn. Soc. N.S.W., 127, 2006 



163 



SALINE LAKES OF THE MIDDLE PAROO 



3 m 




^from 
Carwarra 
Ck. 



Stinking 
Well Ck 
from 
L. Karatta 



Stinking 
Well Creek 



flood outflow 

to 

Six Mile Ck. 



Figure 6. Bathymetric map of Lake Yumberarra. Contour in- 
tervals 0.5m. Map based on Fig 1 in Timms & McDougall (2005). 
Key: beach ridges - long dashes; creek channels - short dashes. 



Lake Karatta 

Lake Karatta is hourglass-shaped, 
aligned N-S, 57 ha in area and near 1.2 
m deep when full (Fig. 7). The basin lies 
in Quaternary alluvium at the edge of 
the Paroo floodplain. At the constriction, 
marked by two long spits, it receives a 
deeply incised Stinking Well Ck., the 
channel turning to the south, shallowing 
and eventually dividing. A small charmel 
connects the two parts of the lake near the 
eastern shore (Fig. 7). The lake overflows 
to the northeast when it is >1.25 m above 
the deepest point in the southern basin. The 
lake basin contains much recent sediment, 
largely clays in the centre of the southern 
basin and loams and sands nearer the creek 
mouth. This recent sediment is 42 cm thick 
in the southern basin and > 1 m thick near 
the creek mouth (corer could not penetrate 
coarser bottom sediments). There is a 
broad, low lunette up to 1 m high abutting 
much of the eastern and southeastern 
shoreline. 



Lake Karatta generally fills fi-om 
local runoff via Stinking Well Creek, but 
occasionally Paroo floodwater reaches it 
via Lake Yumberarra (details in Timms, 
1999). During the wet years of 1998- 
2000 it remained full, but soon dried in 
the 2002 drought (Fig. 3). At other times 
it may partially fill and soon dry, as in 
1997 and 2004 (Fig. 3). Water is generally 
fi-esh, but in 1993 it was hyposaline. 

North Blue Lake 

North Blue Lake on Rockwell 
Station is elongate oval shaped, 205 ha in 
area and up to 2.3 m deep when fiill, but 
usually depths are < Im (Fig. 8). The long 
axis runs NNW-SSE. This lake is the first 
in a series (North Blue, Mid Blue, Bulla, 
and sometimes Lake Wombah) fed by 
Number 10 Creek, a major drainage line 
about 25 km long and partially blocked by 
dunes south of each lake. The indistinct 
shoreline varies firom ~2 to 3 m above 
the deepest point. The western shore is 
partly cliffed and the eastern shore has a 
gypseous lunette highest in the southeast. 
The eastern shore has well-defined 
beaches, decreasing in height from north 

To Lake 
Yumberarra 




N 



b 100 m 



Figure 7. Bathymetric map of Lake Karatta with position 
of creek channels and spot heights in these above lowest 
point in the lake. Contour intervals 25 cms. Key: creek 
channels - short dashes. 



164 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



Number 10 Ck. 



Cliffs 
to 3 m 




Overflow to 
Mid Blue Lake 



335 



lunette 



Figure 8. Bathymetric map of North Blue Lake with 
heights of the lunette dune on the eastern and southern 
shores and location of cUffs on the western shore. Contour 
intervals 25 cms. Key: beach ridges - dot and dashed lines. 



least 300m from the western cliffs where they 
are buried by 30-50 cm of grey mud. 

Mid Blue Lake contained water continuously 
from 1994 to early 2002 and again in mid 2004. 
Its mean salinity (4.1 gL') was similar to that 
in North Blue Lake, but the maximum salinity 
of 103 gL"'was much higher. Further data are 
given in Timms (in press a). 

Lake Bulla 

Lake Bulla is a complex lake, with a western 
basin connected to extensive waterways 
backed up inflowing creeks and with many 
gypseous lunettes on its northern, eastern and 
southern shores. It is 420 ha in area and up to 
4.8 m deep when full. Generally it is the final 
lake of the series on Number 10 Ck., as there 
is a dune system totally blocking the creek 
southwestwards. It receives water in the same 
pattern as the two lakes upstream (Fig.4), but 
has a greater salinity range (2 - 262 gL'), higher 
median salinity (9.8 gL'') and slightly shorter 
wet period. See Timms (in press a) for ftirther 
data. 



to south; one cuts off the southeast comer of the lake. 
Lake sediments are deep muds which, when dry, are 
readily moved in dust storms and partially redeposited 
in the lee of samphires {Arthrocnemum halocnemoides 
Nees) in the littoral zone, on the beaches and beyond. 
Other data are given in Timms (in press a). 

North Blue Lake held water for most of 1994- 
early 2002, but dried briefly three times. It also held 
some water in mid 2004 (Fig 4). Salinity varied from 
fresh to 31 gL"', with a median salinity of 4.2 gL'. 
Details are given in Timms (in press a). 

Mid Blue Lake 

The next lake downstream on Number 10 Creek 
is Mid Blue Lake which is also oval-shaped, but 
slightly bigger (at 210 ha) and considerably deeper 
when full (3.4 m). The bathymetric map (Fig. 9) is 
not as detailed as other maps, but together with the 
transect (Fig. 10), is sufficient to show relatively 
steeply shelving shores above the 1 m contour, a irmer 
lunette system ending both north and south in a beach 
system and a massive outer lunette system. The lake 
has largely retreated from the occluded parts in the 
southeastern and northeastern comers. Much of the 
western shoreline is cliffed soft sandstones cemented 
by carbonates; on the transect (Fig. 10) these rocks 
are exposed in the shore zone and beyond this to at 



Lake Wombah 

Lake Wombah is the largest of the Rockwell- 
Wombah system at 740 ha and 2.3 m deep. It is 
connected to the Paroo River and, like Lake Numalla, 
receives Paroo floodwater, but unlike Numalla, has 
limited beach and spit development. The western 
and northem shoreline is cliffed (up to 7.5 m high). 



cliffs to 4 m 
above high 
shoreline 



from North 
/ Blue Lake 




position of 
transect in 
Fig.10 



<csiSS> 



To Lake Bulla 



Figure 9. Incompete bathymetric map of Mid Blue 
Lake together with position of lunettes on eastern 
shore and cliffs on the western shore. Contours 
at 0, 0.5, 1 and 2.5m. Key: beach ridges - dot and 
dashed Hues. 



Proc. Linn. Soc. N.S.W., 127, 2006 



165 



SALINE LAKES OF THE MIDDLE PAROO 



outer 
lunette/ 



« 

z 




inner 
lunette 



'full' water level 



250 




500 750 

Distance across lake (m) 



1000 



1250 



Figure 10. Transect across Mid Blue Lake west to east through the deep- 
est portion 



while the eastern shoreline abuts subdued inner and 
outer lunettes. Because Wombah fills mainly fi-om 
the Paroo and not Number 10 Creek, it has different 
fluctuations in water levels than the Rockwell Lakes 
(Fig. 4), though salinity range (1-30 gL') and 
median salinity (4.9 gL') are similar. It dries more 
regularly than Lake Numalla, because it is less than a 
third its depth. Timms (in press a) presents more data 
on this lake. 

Gidgee Lake 

Gidgee Lake is an oval-shaped lake with a N-S 
major axis lying in a depression east of a dune system 
and connected by a channel to Bells Creek (Figs. 1 1 
& 12). In normal fillings it is 160 ha in area and ca 5 
cm deep, but in imusually large fillings (as in 1974 
and 1976, D. Leigo, pers. 
com.) it is larger in area and 
much deeper (to 1.5 m). The 
southeastern comer is cut off 
by a recurved spit; this spit 
and adjacent southern beach 
are each overlaid with a small 
lunette (Fig. 11 A). There 
is another clayey lunette 
adjacent to the old shoreline 
and beyond this, a large (5- 
8 m high) gypseous lunette 
(Fig. 12). The lake floor is 
of red clay up to 24 cm thick 
over gypseous mud. The clay 
is laminated, mainly near its 
base with the thick upper 
part believed to have been 
deposited in either of the 
big 1974 or 1976 fillings (D. 
Leigo, pers. com.). Recent 
sedimentation has moved 



the lake's deepest point to 
the south and halved the 
normal filling depth (Fig. 
UA&B). 

Generally, Gidgee Lake 
holds water for a few months 
then remains dry for many 
months, particularly during 
droughts (Fig. 4). The 
filling of 1998-2001 was 
much longer than usual and 
associated with the above 
average rainfall of 1998- 
2000. In that Bells Creek 
flows after most rain events 
>10mm, and these minor 
flows may reach Gidgee Lake, it is possible that there 
were even more minor inflows than indicated in Fig. 
4. Salinity ranges in Gidgee Lake firom 3-182 gL' 
but typically the lake is hyposaline. A filling-drying 
cycle in 1995 is documented in Timms (1997b). 

Lower Bell Lake 

At Lower Bell Lake, the 23 km long Bells 
Creek is blocked by a large dune advancing fi-om the 
northwest. The lake is wedge-shaped with the main 
axis SW-NE and the creek entering in a wide channel 
at the southeastern comer (Figs 12 & 13). When fiill, 
the lake is 1 85 ha in area and about 30 cm deep. There 
is a bar across the mouth of Bells Creek; this is part 
of a beach system extending across the southeast 



Bells Ck. 




former 
, shoreline 



Figure 11. A, Bathy metric map of Gidgee Lake with main contour inter- 
vals at 5 cm. B, map showing extent of recent sedimentation in Gidgee 
Lake. Main contour interval 5 cm. 



166 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



clifte 




Figure 12. Map showing streams and lakes in the vicinity of the 
terminus of Bells Creek. 



comer of the lake. The rudiments of another beach 

further into the lake and at lower elevation is marked 

by two low gypseous mounds and slight elevations 

in the lake floor as evidenced in the bathymetric 

map (Fig. 13 A). The lake basin extends further east, 

is marked by some minor beach/ 

dune systems near Bells Creek, and 

is bordered by a large gypseous 

lunette (Fig. 12). The lake is floored 

with gypseous muds, covered by 

laminated red clays up to 13 cm 

deep and alternating with layers of 

small gypsum crystals. The bar and 

associated beach is composed of at 

least Im of gypsum. There is a large 

(5-8 m high) gypeous lunette lying 

to the east of the lake 

Lower Bell fills less often than 

Gidgee Lake and tends to dry sooner 

after filling. It is dry for many 

months to years. Salinity regime is 

similar to, but slightly more saline 

than, that of Lake Gidgee (Fig. 4). 

Like Gidgee Lake, it filled well 

beyond its normal shores in 1974 

and 1976, so that it was possible 

to water ski on, and between, both 

lakes (D. Leigo. pers. com.). Events 

during a filling-drying cycle in 1995 
are given in Timms (1997b). 



Horseshoe Lake 

Horseshoe Lake (A = 746 ha) has a 
flat floor with slightly deeper parts at 
the southern end of each arm (Fig. 12), 
and a mound of sediment at the mouth 
of Bartons Creek partly occluding the 
southeastern portion. This mound is 
interpreted as an alluvial fan rather than 
a delta, as it has the profile and plan of a 
fan and is believed to form subaerially 
as the lake fills. Water depth is rarely 
> 30 cm. Besides a typical gypseous 
lunette on the eastern side and cliffs 
on the western shore, parts of the 
shoreline are backed by beach ridges. 
The most significant of these are in an 
area of the lake now abandoned in the 
northeastern comer (Figs 12 and 14), 
where there are three ridges increasing 
m average height landwards. There is 
no marked vertical differentiation in 
the bottom sediments, but those of the 
alluvial fan of Bartons Creek are more 
silty and give the appearance of recent 
deposition. Lake Horseshoe now never 
overflows, but two former pathways 
are evident to Lower Bell Lake (Fig. 12). Better 
evidence for a drainage change in this area is seen 
nearby at Palaeolake and Freshwater Lake (Fig. 12) - 
once Palaeolake with its older gypseous lunettes was 




Figure 13. A, Bathymetric map of Lower Bell Lake with main 
contour intervals of 5 cm. Mounds of gypsum shown dotted. B, 
map of Lower Bell Lake showing extent of recent sedimentation. 
Contour intervals at 5, 7.5 and 10 cm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



167 



SALINE LAKES OF THE MIDDLE PAROO 




beach ridges 



Distance from lalte shore (m) 

Fig 14, Transect through the northeastern corner of Horseshoe Lake from 
the lake shore to the gypseous lunette, showing three former beaches at in- 
creasing elevation above the present lake floor. 



the only ponding place in this catchment, whereas 
now, water ponds mainly in Freshwater Lake, with 
its younger irmer clayey lunette. Sometimes water 
flows on to Palaeolake, creating an unusual situation 
of water abutting an older gypeous lunette. 

Horseshoe Lake fills fi-om Bartons Creek and, 
like Lower Bell Lake, did not overflow during the 
study period. Filling is even more intermittent than 
for Lower Bell Lake, and prevailing salinities higher, 
so that meosaline - hypersaline conditions mostly 
prevail (Fig. 4). Salinities often increase along the 
axis of the lake from the inflow of freshwater to the 
southeastern comer to the blind southwestem comer, 
e.g. in July 2001, the gradient was 64 to 
182 gL-'. 

Bells Bore Salt Lake 

Bells Bore Salt Lake on Bloodwood Station 
is a small oval salina, orientated SW-NE (Fig 12 
& 1 5). When fiall, lake area is 24 ha and a potential 
depth of 50 cm, but during 1987-2004 maximum 
depths rarely exceeded 10 cm. There is a small 
island of gypseous sand and two lunette dunes on 
the east and southeastem shore. The irmer lunette 
is of clayey silt and the higher outer lunette is 
of gj^sum. With no inflowing creeks, lake water 
is mainly exposed groundwater together with 
overland flow from adjacent flats, so that filling 
events are limited (Fig. 4), and water does not 
persist for more than a few months, even during 
the wet years of 1998-2000. 



northeast for about 11 
km. The eastern shoreline 
is evenly curved and 
bordered by a lunette 
dune up to 330 cm 
above the lake floor and 
higher gypseous lunette 
fiarther eastwards. The 
lake floor is of gypseous 
mud covered with a red 
clayey layer up to 10 cm 
deep, but thinning away 
from the inlet (Fig 16B). 
There are also short, 
discontinuous alluvial 
fans up to 50 cm deep 
in the northwestern and 

southwestem comers of the lake. 

Lake Burkanoko had water on five occasions out 

of 19 visits during 1988-1994, with a salinity range of 

6-37 gL"' and median salinity of 22.6 gL' (Timms, 

1993, 1998b). 

Lake Barakee 

One of many small salinas on Barakee and 
adjacent Goonery Stations, Barakee Lake (Fig. 
17A) is a small oval salina (A = 90 ha) with a N- 
S axis, lying between western cliffs up to 5 m high 
in a transgressive dune and two lunettes to the east. 




420 cm 

Outer 
lunette 



100 m 



Lake Burkanoko 

Lake Burkanoko on Wangamaima Station is 
oval shaped with a N-S major axis and is 280 ha 
in area and ca 40 cm deep when fiall (Fig. 16A). 
It is the terminus for a creek flowing from the 



Figure 15. Bathymetric map of Bells Bore Salt Lake 
with contours at 10 cm intervals and position and spot 
heights above the lake bottom of two lunette dunes. 
Island of gypeous sand stipped. 



168 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



inner 
'^lunette 




330 



280 



the lake. Superficial examination 
this delta suggests it is composed of 
sands and gravels. There is a small 
lunette to the east (not shown on 
Fig. 18). 

During 1988-2004, Taylors Lake 
had water 1 8 times on 20 visits, with 
a salinity range of 0.7 - 9.1 gL"' and 
median salinity of 2.1 gL"' (Timms, 
1993, 1998b). Despite usually 
having water, the lake dried in late 
2002 and has not held water since 
(T. Nielson, pers. com.). 



Figure 16. A, Bathymetric map of Lake Burkanoko with contour in- 
tervals at 10 cm and location of a lunette dune on the eastern shore 
and cliffs on the western shore. B, map of Lake Burkanoko showing 
extent of recent sedimentation. 



The inner lunette of clay is much dissected and 
with a present day maximum height above the lake 
floor of ca 3 m, while the outer lunette of gypsum 
is much larger and higher, to ca 9 m. The lake has a 
'shoreline' 50 cm above the lowest point, but when it 
contains water, depth rarely exceeds 10 cm. There is a 
boomerang-shaped beach in southeast sector reaching 
23 cm above the lake floor. Superficial 
sediments are of recently deposited red 
silty clay up to 25 cm deep beneath the 
beach, and generally 15 cm deep in the 
centre of the lake and thinning to < 5 cm 
towards the margin (but deeper at the 
edges due to fans from the lake edge (Fig. 
17B)). 

Lake Barakee had water on eight 
occasions out of 20 visits during 1988- 
2004, with a salinity range of 23 - 2 1 8 gL" 
' and median salinity of 1 1 5 gL"' (Timms, 
1993, 1998b). 



Taylors Lake 

Taylors Lake on Ballycastle Station 
is a relatively deep (1.2 m) hj^sosaline 
lake in a hollow among dunes (Fig 18), 
probably made smaller by an advancing 
transgressive dune fi-om the northwest. 
The lake is orientated S W - NE and has an 
area of 62 ha. It receives a major stream 
(about 4 km long) which has built a multi- 
channelled delta on the southern shore of 



DISCUSSION 
Geomorphology 

Aeolian deflation is a major force 
in lake geomorphology in arid 
lands (Shaw & Thomas, 1989; 
Timms, 1992), and the Paroo 
is no exception. Some playas 
such as Bells Bore Salt Lake and 
Barakee Lake are simply hollows 
in the Quaternary sandscape deepened by wind. 
Timms (1993) lists fiirther examples in the Paroo 
and inspection of topographic maps suggests many 
other lakes were formed in this way. Blockage by 
dunes as they move transgressively across the land 
has formed many others, notably Lower Bell Lake 




Figure 17. A, bathymetric map of Lake Barrakee with con- 
tour intervals of 5 cm and location of lunette dunes on the 
eastern side and cliffs on the western shore. B, map of the 
extent of recent sedimentation in Lake Barakee. Note the 5 
cm depression contour, indicting recent deposition of sedi- 
ment is least within this contour. 



Proc. Linn. Soc. N.S.W., 127, 2006 



169 



SALINE LAKES OF THE MIDDLE PAROO 




Figure 18. Bathymetric map of Taylors Lake. Contour inter 
vals 20 cm. 



where a large transgressive dune from the northwest 
has blocked Bells Creek. Other examples include 
the lakes on Number 10 Creek on Rockwell Station 
- here the creek line has been totally occluded south 
of Lake Bulla and partial blockages south of Mid 
Blue Lake and North Blue Lake accounts for these 
lakes. Gidgee Lake, Lake Burkanoko and Taylors 
Lake are three further examples and Timms (1993) 
lists others. For some lakes, however, the initial 
formative process is not wind. Lake Wyara lies on a 
Tertiary fault (Timms, 1998a) and Lakes Yumberarra 
and Karatta are 'embankment lakes' (Timms, 1992) 
located at the edge of the greater Paroo floodplain, 
suggesting less deposition there well away from 
the main stream and associated ponding of riverine 
floodwater and also local runoff (Timms, 1999). In 
a slightly different version of this, water can also be 
ponded in a side valley by fluvial sediments; Lake 
Numalla and Wombah are examples of such blocked 
valley lakes (Timms, 1992). 

While there is no evidence of ancient megalakes 
in the Paroo (cf. the former Lake Dieri stage of Lake 
Eyre - DeVogel et al., 2004), some of the study 
lakes have shrunk since initial formation. Horseshoe 
Lake, Lower Bell Lake, Bells Bore Salt Lake, Lake 
Burkanoko and Barakee Lake now never reach their 
outer lunette dune (base 2-3 m above present lake 
floor), and Mid Blue Lake and Gidgee Lake do so only 
rarely. In both of these lakes the irmermost lunette is 
truncated, which is believed to have happened in the 
exceptionally high water levels during 1974 and/or 
1976. Horseshoe Lake has abandoned beaches with 
intervening lake floors up to 2m above present lake 
floor and stepped downwards towards the present lake 



floor (Fig. 14). This, and the high base of 
lunettes, points to lowering of lake floors 
by deflation, so that while lake areas have 
decreased, the potential volume of water 
held may not have. On the other hand, 
Barakee Lake now rarely fills beyond 
10-20 cm deep and a third beach/lunette 
precursor is forming at 15-25 cm above 
the deepest point, well inside the irmer and 
outer lunettes. . 

Lakes in arid lands tend to have 
regular outlines due to the smoothing 
influence of wind-induced currents 
(Hutchinson, 1957). The best examples are 
small playas in unconsolidated sediments, 
such as Lake Barakee and Bells Bore Salt 
Lake, which are almost perfectly oval- 
shaped. Both have an ellipiticity (E = (L- 
W)/L) of 0.5, within the range of playas in 
Western Ausfralia, but a little more than the 
0.33 average (Killigrew and Gilkes, 1974). 
The eastern shores of most other lakes are smoothed, 
the most striking example being Lake Wyara (Fig. 
3) probably because it is the largest lake so wave 
action and currents are strongest. With winds largely 
bidirectional (southeasteries and northwesterlies are 
sfrongest winds) (Bureau of Meteorology, website) 
and sandy shorelines, lake segmentation would 
be expected (Zenkovitch, 1959; Lees, 1989) and 
indeed Lake Karatta is divided into two lakelets and 
Lake Numalla has two major cut-off lakelets, many 
separated bays and an incipient cut-off southeastern 
portion. In other lakes, such as Yumberarra, North 
Blue, Lower Bell, and Gidgee (Figs. 6, 8, 13, 11) 
(listed in decreasing stage of development), the 
partially occluded southeastern part is well developed, 
with the major spit development always from the 
north. Significantly, these partial occlusions are found 
in lakes with a N-S axis which facilitates action by 
northwesterly winds to generate southerly-flowing 
currents on the southeastern shore. These occlusions 
increase habitat diversity, for in Lake Numalla, the 
segmented lakelets maybe of different salinity and 
hence invertebrate composition (Timms, 1997a) 
and in Lake Yumberarra the increased shoreline and 
shallow waters of the occluded bay increase bird 
habitat (Timms and McDougall, 2005). 

Like most intermittent lakes in southeastern 
Ausfralia, almost all of these Paroo lakes have 
lunette dunes on their eastern shores (Bowler, 1968, 
1983). Lake Numalla is the only lake without one; 
significantly it is mostly fresh and nearly permanent 
and hence lacks the proper environment for lunette 
development (Bowler, 1 976). The same enviroiunental 



170 



Proc. Linn. Soc. N.S.W., 127, 2006 



B.V. TIMMS 



factors apply, to a lesser degree, in Lakes Yumberarra, 
Karatta and Wombah, and not surprisingly their 
lunettes are weakly developed. The biggest lunettes 
are associated with intermittent salinas, such as 
Lakes Barakee, Lower Bell, Gidgee, Mid Blue and 
North Blue. In most lakes there are two or even three 
lunette dunes: an outer large gypseous dune some 
distance from the lake, then one or sometimes two 
smaller inner clay lunettes close to the present shore. 
The gypseous dunes were probably formed 40,000 
to 14,000 yBP (Pearson et al., 2004) and hence are 
contemporaneous with the lunette formation in 
southern Australia (Bowler, 1976). The inner clay 
lunettes must therefore be of younger age and some 
give the appearance of present activity (e.g. at Lakes 
Barakee and North Blue). The lunette on Lake Wyara 
is of quite different character (hardly visible on the 
ground, and no gypsum) and is possibly much older, 
as Lake Wyara may date back to the Tertiary (Timms, 
1998a). Finally, Freshwater Lake on Bloodwood 
Station (Fig. 12) has only an inner clay lunette and 
therefore is likely to be of Holocene origin, probably 
because of drainage change to Palaeolake which has 
only a gj^seous lunette (Pearson et al, 2004). 

Lakes with cliffs on the western and northern 
shores seem to have migrated a little (at least up to 
300 m) westwards. When fiiU, waves generated by 
southeast and southerly winds attack the cliffs and 
afterwards fresh debris can be found at their bases. 
Further evidence of cliff retreat is provided by sloping 
platforms below cliffs in southern Lake Wombah 
and by buried rock in the littoral zone adjacent to 
western cliffs in Lakes Burkanoko and Mid Blue. In 
the Paroo, cliffs occur only in medium-sized lakes; 
smaller lakes lack cliffs probably because fetch for 
wave production is insufficient, but cliff absence in 
the large Lake Wyara and Numalla must be due to 
other factors. Perhaps in the latter there are sufficient 
shore sediments (sandy beaches in Lake Numalla and 
offshore bars and gravelly beaches in Wyara (Timms, 
1998a, 1999) to protect the shore. On the other hand, 
large playas in Salinaland in Western Australia (Jutson, 
1934) and playas in South Australia (Madigan, 1944) 
have cliffs on their western shores and some of them, 
at least, lack protective shore sediments (author, 
unpublished data). Perhaps the explanation for the 
difference lies in the difference in filling regimes, with 
the Salinaland lakes filling only occasionally (Van de 
Graaf etal. 1977). Interestingly, Jutson (1935) claims 
the Salinaland lakes have migrated westwards, just 
like some, especially Mid Blue Lake, in the Paroo. 

Hydrology 

Most of the lakes of the middle Paroo are 



episodic, with only Lake Numalla almost permanent. 
This contrasts with saline lakes in southern Australia, 
where some are permanent (Tinmis, 1976; Williams, 
1 995), but most are seasonal (DeDeckker and Geddes, 
1980; Timms, in press b). In the Paroo, filling-drying 
regimes vary from highly intermittent in the shallow 
salinas with no inflowing streams, such as Bells Bore 
Salt Lake and Lake Barakee, to a pattern of holding 
water much of the time in closed lakes with major 
inflowing streams, like Lake Wyara. Lakes on lesser 
streams, such as those on Bartons and Bells Creeks 
(e.g. Gidgee Lake) and Number 10 Creek (e.g. Mid 
Blue Lake) have intermediate hydrological regimes. 
Those receiving water from the Paroo fill more reliably 
(e.g. Lake Yumberarra) or even almost permanently 
(Lake Numalla). Lakes connected to the Paroo tend 
to be fresh, largely because, when full, they are open 
hydrologically, but as they dry they become closed 
hydrologically and naturally salinise. The other lakes 
are closed permanently; the most intermittent ones 
tend to be the most saline (generally hypersaline) 
while those with inflowing creeks tend to spend much 
of their time when holding water in the hyposaline- 
mesosaline range, but overall, with a large salinity 
range as they progress from frill to dry. 

Eastern and northern Australia, including the 
inland, is affected by the El Nino/Southern Oscillation 
(ENSO) phenomenon (Bureau of Meteorology, 
website). This influences rainfall and river flow 
periodicity as shown for the fillings and drying of 
Lake Eyre (Kotwicki and Allan, 1998). In the Paroo, 
there is also a highly significant relationship between 
fiiU and dry periods over 118 years in Lake Wyara 
and the SOI. For the shorter period covered by this 
study, all lakes held water during the wet phase of 
1998-2000 when the SOI was positive and all dried, 
sooner or later during 2001 - 2004 when the index 
was negative. This relationship is not so intense 
during the previous wet period of 1988-1990 and 
drought of 1992 -1993, with most lakes filling at least 
intermittently in the wet years, and only the larger 
ones persisting during 1992 and into 1993 (Fig. 4). 

As a corollary to the wide fluctuations in 
salinity in most of these Paroo lakes, many salt lake 
invertebrates have wide salinity tolerances (Williams, 
1984; Timms, 1993). Furthermore, cumulative species 
lists for these lakes are unusually long (Timms, 
1998a, in press a) because the lakes pass through 
hyposaline, mesosaline and hypersaline stages and 
hence have components of all faunas (Timms and 
Boulton, 2001). On the other hand, freshwater lakes 
which rarely have saline phases, e.g. Lakes Numalla 
and Yumberarra, have a restricted salt lake faunal 
component, consisting mainly of readily dispersable/ 



Proc. Liim. Soc. N.S.W., 127, 2006 



171 



SALINE LAKES OF THE MIDDLE PAROO 



tolerant rotifers and cyclopoid copepods. 

Sedimentation 

Recent sedimentation in natural lakes in arid 
Australia has gone undocumented (Australian State of 
the Environment Advisory Council, 1996; Australian 
State of the Environment Committee, 2001), unlike 
that in reservoirs (e.g. Wasson and Galloway, 1986; 
Jones, 2003) and streams (e.g. Pickard, 1994). Either, 
there is none readily apparent, as in Lake Yumberarra, 
or lakes are too remote to know, or the problem too 
fragmented to be of interest (Timms, 2001c). Yet 
many of these Paroo lakes have suffered extensive 
sedimentation since European settlement, certainly 
during the wet years of 1974, 1976 and since. Lake 
Karatta, the terminus of a severely eroded stream 
channel, has a minimum of 42 cm of recent sediments 
(Fig. 7); Gidgee Lake, a side basin on Bells Creek, 
has up to 24 cm of clayey sediments very different 
to the gypseous sediments below (Fig. IIB); and 
Lakes Lower Bell (Fig. 13B), Burkanoko (Fig. 16B) 
and Barakee (Fig. 17B) have lesser amounts of recent 
clayey sediments. Alluvial fans and deltas are filling 
significant parts of Lake Wyara (Fig. 3), Taylors Lake 
(Fig. 1 8) and Horseshoe Lake, and most lakes have 
small fans at the entrance of every channel to the 
lake. These red, sticky clayey sediments originate 
from small catchments with severe erosion. In the 
lakes on Number 10 Creek, the recent sediments 
are fiiable muds which deflate during dry periods, 
so that there is little, if any, accumulation of recent 
sediments. Friable muds also floor Lakes Wyara, 
Numalla, Yumberarrra and in addition the BindegoUy 
Lakes near Thargomindah (M. Handley, pers. com.). 
In all these cases the inflowing stream is from a large 
catchment, in which isolated severe erosion of red 
clayey soils is masked by the less sticky grey clays 
transported by western rivers. 

The consequences of rapid recent sedimentation 
are largely unknown, apart from geomorphological 
modification of the affected lakes (e.g. the location 
of the deepest point in Lake Gidgee has changed). 
Certainly the affected lakes hold water for a shorter 
period after a major fill (in Lake Gidgee's case this can 
be as much as a 50% shorter period), but the influence 
of this on their ecology is xmknown. One known 
affect in Lake Karatta is for (the associated) greatly 
increased turbidity to devalue the lake as a waterbird 
feeding site (McDougall and Timms, 2001). Another 
problem is the predicted imminent connection of bird 
breeding islands to the lake shoreline in Lake Wyara 
and the consequent invasion of the islands by the 
predatory foxes and cats (Timms, 2001c). Beyond the 
lake shores, lunette building could be affected - the 



red clayey sediments seem not to readily deflate when 
dry, so that any contemporary lunette building in these 
lakes (e.g Lakes Gidgee, Lower Bell, Burkanoko, 
Barakee) is inhibited. On the other hand, lunette 
building could be enchanced in the lakes on Number 
10 Creek by its delivery of friable sediments. 



CONCLUSIONS 

The middle Paroo catchment of northwest New 
South Wales and southwest Queensland has numerous 
lakes, some of which are saline or become saline as 
they dry. Eleven lakes have been mapped and these 
plus five others have been studied for periods of up to 
1 8 years. Many lakes were formed by dunes or river 
sediments blocking drainage routes, some lie in dune 
swales, some lie at the edge of the Paroo floodplain 
where alluvial sediments are thinner, and Lake Wyara 
lies on a faultline. All developed fiirther by deflation 
and owe their form to wind-induced currents and 
wave action shaping shorelines. Eastern shorelines 
are of often evenly curved and western shorelines 
may be indented, or smooth. Typically, lakes are 
flat-floored and shallow (<2 m deep), but two have 
maximum depths of- 6.5 m. Most saline lakes have 
shrunk, leaving double, sometimes three or more, 
lunette dunes on the eastern shore, and many larger 
ones have migrated westwards due to wave action on 
cliffs on the western shore. Lakes of low salinity have 
sandy beaches and no, or poorly developed lunettes, 
but may be compartmentalised by spit growth across 
bays. Lakes with N-S axes have the southeastern 
comer cut off by spits generated by currents induced 
by northwesterley winds. A few lakes are filling with 
sediment derived from the overgrazing of catchments 
associated with European settlement. In small eroded 
catchments, sediments are sticky red clays which 
accumulate and are filling the lakes, but if the added 
sediments come from large, less eroded, catchments, 
they are friable and present deflation can keep 
pace with sedimentation so that such lakes are not 
infilling. 

Larger lakes with inflowing streams fill in El 
Niiio years, then dry over the next few years, i.e. are 
episodic. Smaller lakes without surface inflows may 
fill a few times in wet years but dry quickly. Most lakes 
remain dry in La Nina years, but those with major 
inflowing streams get occasional small inflows which 
evaporate within months. Salinity regimes fluctuate 
between subsaline (0.5-3 gL"') and euhypersaline > 
200 gL"' and, while instantaneous faunal lists may 
be depauperate, cumulative species lists can be long. 
However, lakes which normally are fresh, but become 



172 



Proc. Linn. Soc. N.S.W, 127, 2006 



B.V. TIMMS 



saline in their final stage of drying, develop only a 
limited saline lake fauna. 



ACKNOWLEDGEMENTS 

For ready access to lakes, I wish to thank the landholders of 
the Paroo, and for hospitality I thank the Bremner family of 
Muella Station, the Davis family of Rockwell Station and 
the staff of Currawinya National Park. For field assistance I 
am grateful to numerous students and friends including Alec 
Gaszik, John Vosper, and Sarah Wythes who survived two 
or more trips. For provision of rainfall data, I am indebted 
to the Neilsons of Ballycastle, the Leigos of Dangarvon, 
the Dunns of Warroo, the Davis' of Rockwell and the 
staff at CNP. For identifying rotifers and little copepods, 
I thank Rus Shiels. For drafting Figure 1 I thank Olivier 
Rey-Lescure, and for helpfiil comments on the manuscript 
I am gratefiil to Conjoint Professor Robert Loughran and 
Professor Wayne Erskine, all of Newcastle University. 



REFERENCES 

Australian State of the Environment Advisory Council 
(1996). Australia: State of the Environment 
2001, Independent Report to the Department of 
Environment, Sport and Tourism, CSIRO Publishing, 
Collingwood, Australia. 

Australian State of the Environment Committee 

(2001). Australia State of the Environment 2001, 
Independent Report to the Commonwealth Minister 
for the Environment and Heritage, CSIRO Publishing 
on behalf of the Department of Environment and 
Heritage, Canberra. 

Bowler, J.M. (1968). Australian landform example: 
lunette. Australian Geographer 10: 402-404. 

Bowler, J.M. (1976). Aridity in Australia: Age, origins 
and expressions in aeolian landforms and sediments. 
Earth Science Review 12: 279-310. 

Bowler, J.M. (1983). Lunettes as indices of hydrologic 
change: A review of Australian evidence. Proceedings 
of the Royal Society of Victoria 95: 147-168. 

Bureau of Meteorology website www.bom.gov.au/climate/ 
enso viewed 15 June 2005. 

Chen, X.Y. (1992). Lakes Menindee, Cawndilla and 
Victoria in western New South Wales: Their 
geomorphology, stratigraphy and shoreline erosion. 
Report by New South Wales National Parks and 
Wildlife Service for New South Wales Department of 
Water Resources, 190pp. 

Chivas, A.R. and Bowler, J.M. (1986). Introduction 
- The SLEADS project. Palaeogeography, 
Palaeoclimatology, Palaeoecology 54: 3-6. 

Cole, G.A. (1968). Desert limnology. In 'Desert Biology' 
(Ed. G.W Brown) pp 423-486. (Academic Press, 
New York). 



Cole, G.A. (1983). 'Textbook of Limnology'. (Waceland 

Press, Prospect Heights, Illinois). 
De Deckker, R and Geddes, M.C. (1980). Seasonal fauna 

of ephemeral saline lakes near the Coorong lagoon. 

South Australia. Australian Journal of Marine and 

Freshwater Research 31: 677-699. 
De Vogel, S.B., Magee J.W., Manley W.F. and Miller, 

G.H. (2004). AGIS-based reconstruction of late 

Quaternary paleohydrology: Lake Eyre, arid central 

Australia. Palaeogeography Palaeoclimatology. 

Palaeoecology 2M: 1-13. 
Gill, E.D. (1973). Geology and geomorphology of the 

Murray Region between Mildura and Renmark, 

Australia. Memoirs National Museum of Victoria 34: 

1-97. 
Hancock, M.A. and Timms. B.V. (2002). Ecology of four 

turbid clay pans during a filling-drying cycle in the 

Paroo, semi-arid Australia. Hydrobiologia 479: 95- 

107. 
Hobson, R., Peck, S. and Handley M. (2005). Freshwater 

Turtle Kill Lake Numulla Currawinya National 

Park, 2005. Queensland Parks and Wildlife Service, 

Environmental protection Agency, Queensland 

Government. Unpublished report. 
Hutchinson, G.E. (1957). 'A Treatise on Limnology' 

Volume I, ( John Wiley & Sons, New York). 
Jones, P.A. (2003). Examining the ability of the caesium- 

137 technique to quantify rates of soil distribution 

and sedimentation in arid western New South Wales, 

Australia. Unpublished Ph. D. Thesis, University of 

Newcastle, Australia. 
Jutson, J.N. (1934). The physiography (geomorphology) 

of Western Australia. ^QSiQxn Australian Geological 

Survey Bulletin Number 95. 
Killigrew, L.F. and Gilkes, R.J. (1974). Development of 

playa lakes in south Western Australia. Nature 247: 

454-455. 
Kingsford, R. T. and Porter, J.L. (1994). Waterbirds on 

an adjacent freshwater lake and salt lake in arid 

Austtalia. Biological Conservation 69: 219-228. 
Kotwicki, V. (1986). 'Floods of Lake Eyre'. (Engineering 

and Water Supply Department, Adelaide). 
Kotwicki, V. and Allan, R. (1998). La Nina de Australia 

- contemporary and palaeo-hydrology of Lake Eyre. 

Palaeogeography Palaeoclimatology Palaeoecology 

144:265-180. 
Lees, B.G. (1989). Lake segmentation and lunette 

formation. Zeitschrift fiir Geomorphologie 33: 475- 

484. 
Lees, B.G. and Cook, RG. (1991). A conceptual model 

of lake barrier and compound lunette formation. 

Palaeogeography Palaeoclimatology Palaeoecology 

84:271-284. 
Madigan, C.T. (1944). 'Central Ausfralia'. New Rev Ed. 

(Oxford University Press, Melbourne). 
McDougall, A. and Timms, B.V. (2001). The influence of 

turbid waters on waterbird numbers and diversity: 

A comparison of Lakes Yumberarra and Karatta, 

Currawinya National Park, south-west Queensland. 

Core//a: 25: 25-31. 



Proc. Linn. Soc. N.S.W., 127, 2006 



173 



SALINE LAKES OF THE MIDDLE PAROO 



Pearson, S, Gayler, L., Hartig, K and Timms, B. 
(2004). Ecosystem health in the Paroo: an arid 
frontier? In 'Proceedings of the Airs Waters Places 
Transdisciplinary Conference on Ecosystem Health 
in Australia'. (Ed. G. Albrecht). pp. 252-264. (School 
of Environmental and Life Sciences, University of 
Newcastle, Newcastle). 

Pickard, J. (1994). Post-European changes in creeks of 
semi-arid rangelands, "Polpah Station", New South 
Wales. In: 'Envirormiental Change in Drylands: 
Biogeographic and Geomorphologic Perspectives' 
(Eds. A.C. Millington & K. Pye). Pp 271-283. (John 
Wiley & Sons, Chichester). 

Shaw, P.A. and Thomas, D.S.G. (1989). Playas, pans 
and salt lakes In ' Arid Zone Geomorphology' (Ed. 
D.S.G. Thomas) pp 184-205. (Belhaven Press, 
London). 

Thomas, D.S.G. (Ed) (1989). 'Arid Zone 

Geomorphology'. ( Belhaven Press, London). 

Timms, B.V. (1976). A comparative study of the 

limnology of three maar lakes in western Victoria 
I. Physiography and physicochemical features. 
Australian Journal of Marine and Freshwater 
Research 27: 35-60. 

Timms, B.V. (1992). ' Lake Geomorphology' (Gleneagles 
Publishing, Adelaide). 

Timms, B.V. (1993). Saline lakes of the Paroo, inland New 
South Wales, Australia. Hydrobiologia 161: 269-289. 

Timms, B.V. (1997a). A Study of the Wetlands of 
Currawinya National Park. A report to the 
Queensland Department of Environment, 
Toowoomba. University of Newcastle. 

Timms, B.V. (1997b). A comparison between saline and 
freshwater wetlands on Bloodwood Station, the 
Paroo, Australia, with special reference to their use 
by waterbirds. International Journal of Salt Lake 
Research 5: 2S7-3U. 

Timms, B.V. (1998a). A study of Lake Wyara, an 
episodically filled saline lake in southwest 
Queensland. International Journal of Salt Lake 
Research 1: 113-132. 

Timms, B.V. (1998b). Further studies on the saline lakes of 
the eastern Paroo, inland New South Wales, Australia 
Hydrobiologia 381: 31-42. 

Timms, B.V. (1999). Local Runoff, Paroo Floods and 
Water Extraction Impacts on the Wetlands of 
Currawinya National Park. Pp51-66 In 'A free- 
flowing river: the ecology of the Paroo River' (Ed. R. 
T. Kingsford). Pp 51-66. (NSW National Parks and 
Wildlife Service, Sydney). 

Timms, B.V. (2001a). Large freshwater lakes in arid 
Australia: A review of their linmology and threats 
to their future. Lakes and Reservoirs: Research and 
Management 6: 183-196. 

Timms, B.V. (2001b). Limnology of the intermittent pools 
of Bells Creek, semi-arid Ausfralia, with special 
reference to community structure and succession of 
invertebrates. Proceedings of the Linnean Society of 
New South Wales 123: 193-213. 



Timms, B.V. (2001c). Wetlands of Currawinya National 
Park: Conservation and Management. In 'Research 
needs for managing a Changed Landscape in the 
Himgerford/Eulo Region - A Workshop held at 
Currawinya National Park 16th May 2001 ' (Eds. 
M. Page, C. Evenson, and A. Whittington), pp 9-12 
(University of Queensland, Gatton). 

Timms, B.V. in press a. The Rockwell- Wombah Lakes, 
Paroo, Eastern Ausfralia: a ten year window on five 
naturally salinised lakes. Hydrobiologia 

Timms, B.V. in press b. A study of salt lakes and springs 
of Eyre Peninsula, South Ausfralia. Hydrobiologia. 

Tinuns, B.V. and Boulton, A. (2001). Typology of arid- 
zone floodplain wetlands of the Paroo River, inland 
Ausfralia and the influence of water regime, turbidity, 
and salinity on their aquatic invertebrate assemblages. 
Archiv fur Hydrobiologie 153: 1-27. 

Timms, B.V. and McDougall, A. (2005). Changes in the 
waterbirds and other biota of Lake Yumberarra, an 
episodic arid zone wetland. Wetlands (Australia) 11: 
11-28. 

Van de Graaf, W.J.E., Crowe, R.W.A., Bunting J.A. and 
M.J. Jackson, M.J. (1977). Relict early Cainozoic 
drainages in arid Western Ausfralia. Zeitschrift fur 
Geomorphologic 1\: 379-400. 

Wasson, R.J. and Galloway, R.W. (1986). Sediment yield 
in the Barrier Range before and after European 
settlement. Rangeland Journal S: 79-90. 

Williams, W.D. (1984). Biotic adaptations in temporary 
lentic waters, with special reference to those in semi- 
arid and arid regions. Hydrobiologia 125: 85-110. 

Williams, W.D. (1995). Lake Corangamite, Ausfralia, 
a permanent saline lake: Conservation and 
management issues. Lakes and Reservoirs: Research 
and Management 1:54-64. 

Zenkovich, VP. (1959). On the genesis of cuspate spits 
along lagoon shores. Journal of Geology 67: 269- 
277. 



174 



Proc. Linn. Soc. N.S.W., 127, 2006 



Pseudoplasmopora (Cnidaria, Tabulata) in the 
Siluro-Devonian of Eastern Australia with comments on its 

global biogeography 



G.Z. FOLDVARY 



School of Geosciences, University of Sydney, NSW 2006 



Foldvary, G.Z. (2006). Pseudoplasmopora (Cnidaria, Tabulata) in the Siluro-Devonian of eastern 
Australia with comments on its global biogeography. Proceedings of the Linnean Society of New South 
Wales 111, \15-\%9. 

The tabulate coral Pseudoplasmopora is widely distributed in Eastern Australia, China, central and 
southeastern Asia, the Rhenish- Alpine region of central Europe, Gotland and eastern U.S.A. Occurrences 
of the genus in Australia are reviewed: Pseudoplasmopora follis, P. heliolitoides and P. gippslandica are 
reassessed, and Pseudoplasmopora sp. A and B are discussed in open nomenclature. During Late Silurian 
times Pseudoplasmopora was confined to Eurasia (predominantly Kazakhstan), eastern Gondwana (Tasman 
Fold Belt of eastern Australia), South China, Gotland and eastern Laurentia. Though disappearing from the 
latter two regions before the end of the Silurian, elsewhere during the Early Devonian Pseudoplasmopora 
underwent considerable biogeographic expansion, particularly within China and central Europe, whilst 
persisting in eastern Gondwana. The youngest species are of Eifelian age. This widespread record suggests 
that it may have potential in palaeobiogeographic analysis of the mid-Palaeozoic continental distribution. 

Manuscript received 16 February 2005, accepted for publication 7 December 2005. 

KEYWORDS: Biogeography, Devonian, Gondwana, heliolitine corals, Silurian, systematics. 



INTRODUCTION 

The Early Silurian to Middle Devonian heliolitine 
coral Pseudoplasmopora Bondarenko, 1963 is 
widely distributed v^^ithin the Tasman Fold Belt from 
Queensland to Victoria. In this paper, all Australian 
species attributed to this genus are reviewed. 
Species previously recognised in this region, 
though referred at the time of original description 
to other genera such as Plasmopora and Heliolites, 
include Pseudoplasmopora follis (Milne-Edwards 
and Haime, 1851), P. heliolitoides (Lindstrom, 
1899) and P. gippslandica (Chapman, 1914). These 
species are here redescribed, and two other forms 
- Pseudoplasmopora sp. A and B - are discussed in 
open nomenclature. 

Pseudoplasmopora is also known from central 
Asia (Kazakhstan) from where it was first distinguished 
by Bondarenko (1963), who additionally included 
in this genus some species from Australia, Gotland 
and eastern U.S.A. that had previously been assigned 
to Plasmopora. Subsequently Pseudoplasmopora 
has been identified in China and cenfral Europe 



(Rhenish - Alpine region). During Silurian times 
Pseudoplasmopora was confined to Eurasia, eastern 
Gondwana, South China, Baltica and eastern Laurentia 
(Figure 1). Although disappearing from the latter two 
areas by the close of the Silurian, during the Early 
Devonian it underwent considerable biogeographic 
expansion (Figure 2), prior to becoming extinct in the 
Middle Devonian (Eifelian). A review of all known 
occurrences suggests that Pseudoplasmopora may 
have potential in palaeobiogeographic analysis of 
mid-Palaeozoic continental distribution. The local 
species P. gippslandica in particular seems to be 
widespread, having been additionally recorded from 
Kazakhstan and central Europe. 



GLOBAL BIOGEOGRAPHIC DISTRIBUTION 

Bondarenko (1963) established 

Pseudoplasmopora on basis of two species, the 
type P. conspecta and P. arguta from central 
Kazakhstan. Interestingly he also assigned some 
Australian forms to this new genus, recognizing P. 
gippslandica (Chapman) from the northern slopes 
of the Tarbagatai Mountains near the mining town 



SILURO-DEVONIAN TABULATE CORALS 



of Karajal (Karadzhal) and the Nura Synclinorium 
in the Karaganda region (composed mainly of 
Silurian to Lower Devonian formations). Kovalevsky 
(1965) described fiirther new Late Silurian species 
of Pseudoplasmopora: - P. bella, P. karaespensis, 
P. subambigua and P. subdecipiens - from the Lake 
Balkhash area of Kazakhstan. Bondarenko (1967) 
reported on the distribution of these species in 
Kazakhstan, and subsequently Bondarenko (1975) 
described fiirther new species from central Asia, 
including P. dzhungaria, P. isenica, and P. septosa. 

Numerous species of Pseudoplasmopora 
have been described from China. Lin et al. (1988) 
showed that the genus was widespread there, with 
the recognition of P. aseptata (Regnell, 1941) and 
P. microsa Wang, 1981 from the eastern Tien-Shan 
Mountains in strata now known to be of Lochkovian 
(Early Devonian) age, and P. shiqianensis Yang, 1978 
from probable Late Silurian rocks of South China. 
More recently, several Early Devonian species from 
the Jilin province of North China were described 
including/! turpanens is Deng, 1997,/! aseptata minor 
Deng, 1997 and P. yaokengensis Deng, 2000. Deng 
(in Deng and Zheng 2000) compared P. yaokengensis 
with P. regularis (Dun) [= P. gippslandica herein] 
noting that the latter has larger corallites that are 
usually separated by two rows of tubuli. 

Amongst the youngest species referable to 
the genus are those known from cenfral Europe, in 
Early to Middle Devonian strata. A new undescribed 
species is present in the late Emsian to early Eifelian 
of the Rhenish Schiefergebirge, and Ghassan (1971) 
documented the occurrence of P. gippslandica from 
Eifelian-age rocks of the Camic Alps in Austria. 

Species originally placed in Plasmopora, such 
as Pfollis (Milne-Edwards and Haime, 1851) and P. 
heliolitoides (Lindstrom, 1899) from the Late Silurian 
of eastern U.S.A. and Gotland, respectively, appear 
to have received little systematic attention since their 
initial descriptions, apart from their reassignment to 
Pseudoplasmopora. 

Australian heliolitines now referred to 
Pseudoplasmopora were first described by Chapman 
(1914), Dun (1927) and Jones and Hill (1940). Hill 
et al. (1969) documented several informally assigned 
species with annotated illustration, in the same year 
that P. sp. cf P. gippslandica was described by Jell and 
Hill (1969). Foldvary (2000) illustrated P. sp. from 
central New South Wales. Usefiil biostratigraphic data 
on Silurian species of Pseudoplasmopora in eastern 
Australia was presented by Munson et al. (2000), 
based on a then-unpublished compilation by Pickett 
(subsequently made available via Internet access 
in 2002). Published species of Pseudoplasmopora 



from Ausfralia include P. follis (Milne-Edwards and 
Haime, 1851), P heliolitoides (Lindstrom, 1899), P. 
gippslandica (Chapman, 1914), and two informally- 
designated species illustrated by Hill et al. (1969). 

Kaljo and Klaamaim (1973) and Pickett 
(1975) briefly mentioned the distribution of 
Pseudoplasmopora in relation to Silurian coral 
biogeography (though surprisingly the genus was 
indicated to be endemic to central Asia, despite 
Bondarenko 's earlier identification of the Australian 
species P. gippslandica in Kazakhstan, and 
recognition of Queensland occurrences by Hill et al. 
1969). Since then much new information has come 
to light regarding mid-Palaeozoic palaeogeography, 
that when combined with increased knowledge of the 
world-wide disfribution of Pseudoplasmopora - here 
plotted on two terrane maps for the Late Silurian 
and Early Devonian respectively (Figures 1 and 2) 
- allows a more complete picture of biogeographic 
relationships between regions where this genus 
occurs. The distribution of Pseudoplasmopora reveals 
its restriction predominantly to the terranes of central 
Asia, China, and the Tasman Fold Belt of eastern 
Ausfralia, with an additional group of occurrences 
in Baltica (Gotland) and eastern Laurentia. It might 
be expected to also be found in terranes forming 
southeast Asia, providing a link between eastern 
Australia and China, but no records are currently 
known of Pseudoplasmopora from this region. 

During the Early Silurian to Early Devonian 
interval, Pseudoplasmopora was distributed between 
30° N and 30° S palaeolatitudes, encompassing (1) 
eastern Gondwana (eastern Australia: Hill, 1978, 
1981; Munson et al. 2000), (2) terranes and continental 
blocks in eastern and central Asia (Kazakhstan, North 
and South China: Bondarenko, 1963, 1975; Lin et 
al., 1988), (3) Baltica and (4) eastern Laurentia, 
confined to Silurian beds. As shown on the terrane 
reconstruction maps of Cocks and Torsvik (2002), by 
the Early Devonian Gondwana had shifted clockwise 
south-eastwards by 90° (Figure 2). Concurrently 
the blocks of North and South China and South- 
East Asian terranes became more separated from 
Gondwana, though remaining near the equator. Such 
dispersal brought about changes in the distribution of 
Pseudoplasmopora, partly retreating (from Laurentia 
and Baltica), and elsewhere expanding in space and 
time into Central Europe where it survived into the 
early mid-Devonian. The Australian part of Gondwana 
remained below 30° S, which explains the continued 
presence of Pseudoplasmopora in eastern Australian 
localities (Figure 2; Cocks and Fortey, 1990). 



176 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLDVARY 




Figure 1. Distribution of Late Silurian Pseudoplasmopora occurrences (indicated by • and terrane 
numbers) throughout the world, based on the terrane reconstruction map of Cocks and Torsvik (2002); 
Lambert Azimuthal Projection centred 30° Long,, -40° Lat. 
Symbols for the Australian species are: 

m=P. follis, A =P. heliolitoides, + = P. gippslandica, = P. sp. A, ^ = P. sp. B. 
Microcontinents and terranes shown thus: 

(1) Queensland, (2) New South Wales, and (3) Victoria of Australia, (4) Annamia, (5) Sibumasu, (6) North 
China, (7) South China, (8) Japan, (9) Taurides and (10) Pontides of Turkey, (11) Hellenic Terrane, (12) 
Perunica, (13) Armorica, (14) Iberia, (15) Baltica (Gotland and eastern Europe), (16) Siberia, (17) Taimyr 
and the Kara Block, (18) Tarim, (19) Sanand and (20) Alborz of Iran, (21) Afghan Terrane, (22) South Tibet, 
(23) Qintang (Qiangtang, Qantang), (24) Tien Shan Mtns., (25) Mongolia (inner part), (26) Altai Mtns. 
and the Tuva Terrane, (27) Kazakhstan, (28) Uzbekistan, (29) Eastern Laurentia (Michigan, Tennessee). 



Proc. Linn. Soc. N.S.W., 127, 2006 



177 



SILURO-DEVONIAN TABULATE CORALS 




Figure 2. Distribution of Early Devonian Pseudoplasmopora occurrences (indicated by • and terrane 
numbers) throughout the world, after the terrane reconstruction map of Cocks and Torsvik (2002); 
Lambert Azimuthal Projection centred 40° Long., -40° Lat. 

Symbols for the Australian species are: ■ =P.follis, A =P. heliolitoides, + = /I gippslandica. 
Microcontinents and terranes shown thus: 

(1) Queensland, (2) New South Wales, and (3) Victoria of Australia, (4) Annamia, (5) Sibumasu (Shan- 
Thai), (6) North China, (7) South China), (8) Taurides and (9) Pontides of Turkey, (10) Hellenic Ter- 
rane (including the Carpathian Basin and Dinarids), (11) Perunica (Bohemia), (12) Armorica, (13) 
Iberia, (14) Rhenish-Alpine area of Central Europe and Podolia (15) Siberia (Platform) and Kuzetsk 
Basin, (16) Taimyr, (17) Tarim, (18) Sanand and (19) Alborz Terranes of Iran, (20) Afghan Terrane, (21) 
South Tibet, (22) Qintang (Qiangtang, Qantang), (23) Tien Shan Mtns., (24) Mongolia, (25) Altai Mtn. 
Range, (26) Kazakhstan with Tarbagatai Mtn. further south-south east, (27) Uzbekistan, (28) Laurentia. 



178 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLDVARY 




[m] GWambOfH Group mat»mofphic» E] Vofcarfcs LEGEND 

Ej lotfutivt £3 . Harv«y Group 

'" Vtrra Yam Cr««k Group 

Janila Limastone Memtmr 
ipn Invartalth Sandstona 
^gl OaalMro Sandstone 
^ra Myamlay Sandstone 

Trundle Group 

t'S'-^ TroHi Formation 
P"^ Connemarra Formation 

Oarfiwong Group 
T~:a Yarrabandal Formation 
^^ Cookayi Plains Formalton 



Figure 3. Simplified locality and schematic geological map of the Trundle - Condobolin area of central 
New South Wales, showing the occurrence of the more important fossil localities. Based on the Narromine 
1:250,000 Geological Map (Sherwin, 1996) and the Forbes 1:250,000 Geological Map (Duggan et al. 1999). 
Fossil localities are indicated by Roman numerals I to XXX, less important ones by Arabic numerals. 



SYSTEMATIC PALAEONTOLOGY 

NOTE: TABLE 1 AND FIGURES 4-8 ARE AT THE 
END OF THE PAPER 

All new specimens are housed in the Australian 
Museum, Sydney; catalogue numbers prefixed by the 
acronym AMF refer to specimens, those prefixed AM 
to thin sections. Listed and illustrated specimens or 
thin-sections fi-om the New South Wales Geological 
Survey are prefixed by MMF, those fi-om the University 
of Queensland are designated UQF. Stratigraphical 
and locality details for the Central West area of 
New South Wales are given in Foldvary (2000) and 
shown herein in Figure 3. The classification follows 
Hill (1981) with updated zoological nomenclature as 
recommended by the ICZN (4* edn. 2000). 



Table 1 provides a concise summary of the 
principal distinguishing features of those species 
described below. 

Suborder Heliolitina Freeh, 1 897 

Superfamily Heliolitoidea Lindstrom, 1876 
Family Pseudoplasmoporidae Bondarenko, 1963 

Genus Pseudoplasmopora Bondarenko, 1963 

Type species 

Pseudoplasmopora conspecta Bondarenko, 1963. 
Late Silurian (Ludlow) age, firom the top of the 
Isen Suite, southern border of the Karaganda Basin, 
Akbastau, Central Kazakhstan [Note that Hill (1981) 
assigns an Early Devonian age to the type horizon]. 



Proc. Linn. Soc. N.S.W., 127, 2006 



179 



SILURO-DEVONIAN TABULATE CORALS 



Diagnosis 

Pseudoplasmoporidae with tabularium surrounded by 
an aureole of mostly 12 tubuli of varying diameter, 
with coenenchyme composed of tubuli of almost the 
same diameter. Tabularia and tubuli walls thin and 
smooth, diaphragms in the tubuli are horizontal and 
complete, rarely oblique or incomplete. Septa, when 
present, appear as septal spines, but they are often 
absent (after Hill, 1981, p. 609). 

Pseudoplasmopora follis (Milne-Edwards and 

Haime, 1851) 

(Figures4,A-F;8, E-F) 

Synonymy 

Plasmopora follis Milne-Edwards and Haime, 1851, 

p. 223, pi. 16, figs. 3, 3a. 
Plasmopora follis Lindstrom, 1899, p. 82, pi. 7, figs. 

19-20. 
Pseudoplasmopora sp. nov. Hill et al. 1969, pi. Ill, 

fig. 6. 

Diagnosis 

Pseudoplasmopora with dense tabularial spacing; 
average tabularium diameter of 1 .0 mm, surrounded 
by aureole formed by 12 regularly polygonal tubuli 
of smaller diameter; tabulae in the tabularia and 
diaphragms in the tubuli are closely spaced. 

Description 

Tabularia spaced between 0.7 and 2 mm apart 
(measured between centres) and number 25 - 30 per 
cm^ within the coenenchyme. Diameter of tabularia 
0.9 - 1 . 1 mm, each contain 10-15 tabulae in 5 mm. 12 
polygonal tubuli form the aureole to each tabularium; 
tubuli in coenenchymal tissue are also polygonal, their 
diameter is 0.1 - 0.2 mm; diaphragms within tubuli 
number 15 - 16 in 5 mm. Septa form node-shaped 
swellings, sometimes with blunt rounded spines. 

Remarks 

In Lindstrom 's type material of P. follis the 
diameters of tabularia and coenenchymal tubuli are 
intermediate between those of Bondarenko's (1963) 
two original species, P. conspecta and P. arguta. The 
type species P. conspecta has tabularial diameters 
of 0.7 - 0.8 mm, those for P. arguta 1.0 - 1.1 mm. 
Pseudoplasmopora dzhungaria Bondarenko, 1975 
has tabularial diameters of 0.8 - 0.9 mm, and the 
diameter of the tubuli is 0.1 - 0.15 mm; tabularia are 
surrounded by 12 (occasionally 13) tubuli: values 
which are comparable with P. follis. Of Chinese 
species, Pseudoplasmopora aseptata (Regnell, 1941) 



has tabularial diameters of 0.8 - 1.2 mm with 2 to 
4 tubuli (4 to 6-sided) interposed between tabularia, 
whereas P. shiqianensis Yang, 1978 has tabularia 
0.75 - 0.85 mm in diameter, spaced 0.4 - 0.8 mm 
apart (measured between centres); P. microsa Wang, 
1981 has tabularial diameters of 0.5 - 0.7 mm (Lin 
et al. 1988). Tabularia of the latter two species are 
considerably smaller in diameter than those of P. 
follis, while those of P. aseptata are practically 
identical in size. 

In Australia Pseudoplasmopora follis occurs 
mainly in the Late Silurian Bowspring Limestone 
and Hume Limestone Members of the Silverdale 
Formation (Gorstian and Ludfordian) at Hattons 
Comer, south of Yass, New South Wales (Munson 
et al. 2000). Additional unconfirmed, undescribed 
occurrences recorded by Munson et al. (2000) 
include: (1) limestone lenses of the Mirrabooka 
Formation west of Orange, (2) Borenore Limestone, 
west of Orange, (3) Jenolan Caves Limestone 
at Jenolan Caves near Oberon, and (4) Quidong 
Limestone at Delegate, near the Victorian border of 
New South Wales. Further specimens, listed below, 
are known fi-om Siluro-Devonian strata in the Trundle 

- Condobolin district of central western New South 
Wales. A species from Queensland, uimamed at the 
time of its illustration by Hill et al. (1969) but thought 
by them to represent a new species, is here assigned to 
P. follis on basis of characters including the presence 
of uniformly polygonal tubuli in both the aureole and 
coenenchyme (UQF60059, here shown on Figure 8). 

Material 

AMF 105567 (cf Figures 4 and 8), a complete colony, 
12x12x6 cm in size, and AMF 105568 are fi-om Loc. 
XXX; AMF116146, AMF116148, AMF116149 and 
AMF116150arefi-omLoc.XX,5kmSWofLoc.XXX. 
These localities are about 40 km NNE of Condobolin, 
New South Wales, near "Meloola" Homestead 
and "Moorefield" Station respectively (Figure 4, A 

- D). Both localities are in the Meloola Volcanics 
of Cookeys Plains Formation, Derriwong Group, of 
Pridolian age. MMF31447 is from a locality 3.5 km 
west of "Moorefield" Station, also in the Meloola 
Volcanics, of Pridoli age (Pickett and McClatchie, 
1991 - listed by them as Pseudoplasmopora sp. and 
not previously figured). 

Pseudoplasmopora heliolitoides (Lindstrom, 1899) 
(Figure 5, A -F; 6, A -F) 

Synonymy 

Plasmopora heliolitoides Lindstrom, 1899, p. 86, pi. 
7, figs. 32-33. 



180 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLD VARY 



Heliolites distans Dun, 1927, p. 258, pi. XIX, figs. 

3-6. 
Heliolites distans var. humewoodensis Dun, 1927, p. 

261, pi. XX, figs. 3,4. 
Heliolites distans var. intermedia Dun, 1927, p. 261, 

pi. XX, figs. 5, 6. 
Heliolites distans var. minuta Dun, 1927, p. 262, pi. 

XXI, figs. 1-4. 
Plasmopora heliolitoides Jones and Hill, 1940, pi. 

IX, figs. 4 and 5; pi. X, figs. 1-4. 
Pseudoplasmopora sp. Foldvary, 2000, p. 91, fig. 8, 

3-4. 

Diagnosis 

Pseudoplasmopora with aureole of tabularium 
composed of tubuli of irregular shape and varying size. 
Septa absent or appear in form of blunt swellings. 

Description 

Tabularia spaced 1.5-5 mm apart, characteristically 
5-7 tabularia per cm^; each tabularia 1.0 - 1.75 mm 
in diameter, with 12-16 tabulae in 5 mm. Aureole 
formed by 12 polyhedral tubuli, each 0.25 - 0.30 mm 
in diameter. Tabularia walls 0.05 mm thick, more than 
twice the thickness of the tubuli walls (0.02 mm). 
Diaphragms in tubuli are spaced 15 in 5 mm. Septal 
spines, when present, dilated at the base. 

Remarks 

The main distinctions between Pseudoplasmopora 
heliolitoides and P. follis from the Trundle - 
Condobolin area are differences in tabularial spacing 
(measured fi-om centre to centre of adjacent tabularia), 
and their density within the coenenchyme. The average 
tabularial spacing in P. heliolitoides is 0.8 mm, closer 
than in P. follis. Pseudoplasmopora heliolitoides 
has only 5 to 7 tabularia per cm^, whereas P. follis is 
crowded with tabularia. Their diameter is distinctly 
larger (1.0 - 1.3 mm) in P. heliolitoides compared 
to P. follis (0.9 - 1.1 mm). Parameter ranges for P. 
heliolitoides given by Jones and Hill (1940) are 1.0 
- 1.75 mm for tabularia diameters, 1.5 - 5.0 mm for 
tabularial spacing, and 12-16 tabulae in 5 mm. 

Additional Material 

Pseudoplasmopora heliolitoides occurs in limestone 
lenses of latest Silurian (Pridoli) age {eosteinhornensis 
Zone) in the Meloola Volcanics, Cookeys Plains 
Formation, Derriwong Group, about 40 km north- 
north-east of Condobolin (Munson et al, 2000; 
Foldvary, 2000). AMF69668 (Figure 5, A-D), from 
which a number of transverse and longitudinal thin 
sections have been prepared (AM 13547, AM13635- 
AM13637, AM13772-AM13774) is fi-om Loc. XX, 



33 km north-north-east of Condobolin, situated east 
of the road, and 0.5-1 km east of 'Moorefield' Station 
(Foldvary, 2000). AMF78962 from the Yass area 
(Figure 5, E-F) comes from beds of slightly older 
(Ludlow) age. 

Pseudoplasmopora gippslandica (Chapman, 1914) 
(Figure 7, A-D) 

Synonymy 

Heliolites interstincta Linne, var. gippslandica, var. 

nov. Chapman, 1914, PI. LX, figs. 35-36. 
Heliolites regularis Dun, 1927, p. 256, pl. XVIII, 

figs. 2, 3. 
Plasmopora gippslandica (Chapman, 1914), Jones 

and Hill, 1940, p. 206, pl. X, fig. 5, pl. XI, fig. 1. 
Pseudoplasmopora gippslandica (Chapman, 1914), 

Bondarenko, 1963, p. 1863. 
Pseudoplasmopora sp. cf gippslandica Jell and Hill, 

1969,p. 23,p. 9,fig.l0a,b. 
Pseudoplasmopora gippslandica (Chapman, 1914), 

Ghassan, 1971, p. 593, pis. 1-2. 

Diagnosis 

Characterized by elongated tubuli in the tabularium, 
continuous walls and the absence of septa. Aureole 
consists of 12 tubuli, usually with two rows of tubuli 
between tabularia. 

Description 

Diameter of tabularia 1.25 - 1.75 mm. Tabulae 
strongly concave, 20 - 25 in 5 mm. Within tubuli are 
20 - 35 diaphragms in 5 mm. Septa absent. 

Remarks 

Pseudoplasmopora gippslandica (Chapman, 1914) 
occurs in eastern Australia in a number of localities 
firom Queensland to Victoria. Although documented 
fi-om Hattons Comer, Yass area, in Late Silurian strata 
(Dun, 1927), other Silurian occurrences are mentioned 
only in unpublished works listed in Munson et al. 
(2000). Otherwise this species is mainly restricted 
to the Devonian. New South Wales occurrences are 
mostly fi-om Lower Devonian beds, and in Victoria 
it is known fi-om Lower Devonian limestones at 
Cave Hill, Lilydale and Waratah Bay. A comparable 
species, P. sp. cf P. gippslandica was described by 
Jell and Hill (1969) from beds of Eifelian age from 
Ukalunda near Bowen in Queensland. Bondarenko 
(1963) noted that P. gippslandica differed fi-om his 
type species P. conspecta only by the coenenchymal 
tubuli having thickened walls, considered to be a 
Devonian trait (Hill, 1967). 
Ghassan (1971) illustrated P. gippslandica? firom the 



Proc. Linn. Soc. N.S.W., 127, 2006 



181 



SILURO-DEVONIAN TABULATE CORALS 



Middle Devonian (Eifelian?) of the Camic Alps in 
Austria. The 12 tubuli forming the aureole are slightly 
larger than the coenenchymal tubuli, and there are two 
to three tubuli between the tabularia. This description 
conforms to P. gippslandica. 

Material 

Specimens shown in Figure 7: AMF5512 (AM66) 
near Rockhampton, Queensland, and AMF6936 
(AM271), Nundle Road, near Tamworth, New South 
Wales. 

Pseudoplasmopora sp. A 

(Figure 8, A -B) 

Synonymy 

Pseudoplasmopora sp. cf. P. heliolitoides; Hill et al. 
(1969),pl. III,fig. 7. 

Description 

Pseudoplasmopora with tabularia 1.0 - 1.2 mm in 
diameter, spaced 12-15 per cm^, and having 14-16 
tabulae in 5 mm; tabularial walls thin, maximum 0.05 
mm. The 12 irregularly polyhedral tubuli forming 
the aureole are clearly differentiated from tubuli 
of the coenenchymal tissue which are 0.3 - 0.6 
mm in diameter, with 8-10 diaphragms in 5 mm. 
In transverse section many tubuli, both tabularial 
and coenenchymal, appear to have small bud-like 
structures internally. Septa when present are blunt. 

Remarks 

With tabularial diameters of 1.0 - 1.2 mm 
Pseudoplasmopora sp. A is comparable with P. 
follis but is readily distinguished from that species 
in displaying tubuli in the aureole that differ from 
those in the coenenchyme in both size and shape. 
It differs from P. heliolitoides in having denser 
tabularial spacing. Pseudoplasmopora gippslandica 
is a distinctly different species with a greater range 
(0.7 - 2.0 mm) for tabularial diameter and wider 
spacing between tabularia. Presence of small bud-like 
structures inside the tubuli is unknown in other forms 
of Pseudoplasmopora from eastern Australia, and 
appears to be a distinguishing feature oiP. sp. A. 

The only confirmed occurrence of P. sp. A is in the 
basal horizon of the Upper Jack Limestone Member, 
Graveyard Creek Formation (Late Silurian) of the 
Broken River area, Queensland (Hill et al. 1969). 

Material 

UQF52829 from Loc. B76F of Jell and Hill, 1969; 
UQF58203 (Hill et al. 1969, pi. Ill, fig. 7) here re- 
illustrated on Figure 8, A-B; and Fl 1587 (Geological 



Survey of Queensland collection). 

Pseudoplasmopora sp. B 

(Figure 8, C - D) 

Synonymy 

Pseudoplasmopora sp. Hill et al. (1969), p. 6, pi. Ill, 
fig. 8. 

Description 

Pseudoplasmopora characterised by densely packed 
(50 - 60 per cm^) very small tabularia (diameter 0.4 
to 0.5 mm); generally 25 tabulae in 5 mm within 
tabularia. 12 tubuli (occasionally 13) forming aureoles 
around tabularia; coenenchymal tubuli are smaller 
and regularly polyhedral. 

Remarks 

The unusually small tabularial diameter sets 
Pseudoplasmopora sp. B apart from the other 
Australian forms of Pseudoplasmopora, though 
the tabularial spacing is the same as in P. follis. 
The density of tabularia (50 - 60 per cm^) is very 
much greater than that of P. heliolitoides (5 - 7) and 
considerably exceeds that in P. sp. A (16 - 18), or in 
P follis {25-2,01). 

Pseudoplasmopora sp. B was first documented 
from Queensland in limestone lens horizons of the 
Jack Formation, Graveyard Creek Group, extending 
into the Upper Ludlow (Hill et. al. 1969). It has been 
reported (but not described) from various localities in 
the Silurian of N.S.W., such as the Narragal Limestone 
and the Catombal Park and Wylinga Formations near 
Wellington (Munson et al. 2000). 

Material 

UQF60060 (Hill et al. 1969, pi. Ill, fig. 8) here re- 
illustrated on Figure 8, C - D. 



ACKNOWLEDGEMENTS 

For use of the facilities of the School of Geosciences, 
University of Sydney, I am much obliged to the Head of 
School, Dr. Geoff Clarke. The author thanks the Director 
of the Australian Museum, and the Manager of the Fossil 
Collections (Mr. Robert Jones), for their kindness in 
incorporating the author's fossil material; Dr Yongyi Zhen, 
also from the Australian Museum, was of great assistance 
in translating Chinese literature. I am grateful to Dr Barry 
Webby for providing some of this literature from his 
extensive library. The author is much indebted to Dr. Ian 
Percival (Geological Survey of NSW) for his perceptive 
comments on early versions of the manuscript; also thanks 
to Dr. John Pickett fi-om the same organization for his 



182 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLD VARY 



generous help, particularly in provision of his OZCORALS 
database. I sincerely acknowledge the two referees for 
their very thorough reviews, which have greatly improved 
the final version. Dr. Alex Cook and Kristen Spring of 
the Queensland Museum kindly made available some 
of Dorothy Hills' UQF thin-sections for re-examination 
and re-illustration. I thank Dr. Carmen Gaina (School of 
Geosciences, University of Sydney) for preparing two 
palaeogeographic base maps using Dr. Trond Torsvik's 
computer poles (via pers. comm.), and Mr. Peter McNiece 
(University of Sydney Library) for his assistance in tracking 
down obscure publications. 



REFERENCES 

Bondarenko, O.B. (1963). Revision of the genus 

Plasmopora. International Geology Review 6 (10), 
1858-1867. 

Bondarenko, O.B. (1967). K istorii razvitya geliolitoidey 
V Kazakhstane. Moskovi Universitet Vestnik, Sen 4, 
Geologii, 22 (3), 39-50. 

Bondarenko, O.B. (1975). Podklass Heliolitoidea. In 
'Kharakteristika fauny silura i devona Tsentalnogo 
Kazakhstana'. Menner, V.V. (Ed.). Materiali 
Geologii. Tsentrala Kazakhstana, 12, 48-61, pis. 4- 
10. 

Chapman, F. (1914). Newer Silurian fossils of eastern 
Victoria Pt.3. Victoria Geological Survey, Records 
Vol. 3, Pt. 3,301-316. 

Cocks, L.R.M. and Fortey, R.A. (1990). Biogeography 
of Ordovician and Silurian faunas. In 'Palaeozoic 
Palaeogeography and Biogeography' (Eds. W.S. 
McKerrow and C.R. Scotese), pp. 97-104. 

Cocks, L.R.M. and Torsvik, T.H. (2002). Earth geography 
from 500 to 400 million years ago: a faunal and 
palaeomagnetic review. Journal of the Geological 
Society, London, 159, 631-644. 

Deng, Z.-Q. (1997). Silurian and Devonian corals 
fi'om the Tarim Basin and adjacent areas. Acta 
Palaeontologica Sinica 36 (Supplement), 116-135. 

Deng, Z.-Q. and Zheng, C.-Z. (2000). Tabulatomorphic 
corals fi-om the Erkhtaopou Formation of Jilin 
Province. Acta Palaeontologica Sinica, 39 (2), 222- 
225. 

Duggan, M.B. et al. (1999). Forbes 1:250,000 geological 
Sheet SI 55-7. 2"'' edition. Australian Geological 
Survey Organisation, Canberra and Geological 
Survey of New South Wales, Sydney. 

Dun, W.S. (1927). Descriptions of Heliolitidae from the 
Upper Silurian, Yass, New South Wales. Records of 
the Australian Museum 15, 255-268; Pis. XVIII-XXI. 

Foldvary, G.Z. (2000). Siluro-Devonian invertebrate 
faunas from the Bogan Gate - Trundle - Mineral 
Hill area of central New South Wales. Records of 
the Western Australian Museum Supplement No. 58, 
81-102. 

Fortey, R.A. and Cocks, L.R. (2003). Palaeontological 



evidence bearing on global Ordovician-Silurian 
continental reconstructions. Earth-Science Reviews, 
61, 245-307. 

Freeh, F. (1897). Refarat, J. Wentzel: Zur Kenntniss 
der Zoantharia tabulata. Neues Jahrbuch der 
Mineralogie, Geologic und Palaeontologie, 1897, 
Part 2, 212-214. 

Ghassan, K.M. (1971). Korallen aus dem Unterdevon der 
Kamischen Alpen. Verhandlungen der Geologischen 
Bundesanstalt Wien, 576-607. 

Hill, D. (1967). The Sequence and Distribution of 

Ludlovian, Lower Devonian, and Couvinian Coral 
Faunas in the Union of Soviet Socialist Republics. 
Palaeontology, 10 (4), 660-693. 

Hill, D. (1978). Bibliography and Index of Austrahan 
Palaeozoic corals. Papers, Department of Geology, 
University of Queensland % (4), 1-38. 

Hill, D. (1981). Part F, Coelenterata, Supplement 1, 
Rugosa and Tabulata 'Treatise on Invertebrate 
Paleontology' (Ed. R.C. Moore) Vols 1-2. The 
Geological Society of America, Inc. and the 
University of Kansas, Boulder, Colorado, and 
Lawrence, Kansas. 

Hill, D., Playford, G. and Woods, J.T. (1969). Ordovician 
and Silurian Fossils of Queensland. Queensland 
Palaeontological Society: 1-18. 

Jell, J.S. and Hill, D. (1969). Devonian corals fi-om the 
Ukalunda district, north Queensland. Geological 
Survey of Queensland, Publication 340, 
Palaeontological Papers, 16, 1-27. 

Jones, O.A. and Hill, D. (1940). The Heliolitidae of 
Australia, with a discussion of the morphology and 
systematic position of the family. Proceedings of the 
Royal Society of Queensland, 51 (12), 183-215. 

Kaljo, D. and Klaamann, E. (1973). Ordovician 
and Silurian corals. In A. Hallam Ed. 'Atlas of 
Palaeobiogeography'. Elsevier, Amsterdam. 

Kovalevsky, O.P. (1965). Tabulyaty i geliolitodei 
Karaesinskogo gomzonta. V. Kn.: Stratigrafiya 
nizhnepaleozoishikh i siliuiyskikh otlozheniy 
tsentral'noro Kazakhstana. (Tabulates and heliolitids 
of the Karaespink horizon. In; Stratigraphy of Lower 
Palaeozoic and Silurian fossil remains of Central 
Kazakhstan). Moskow, Nedra. 

Lin, B., Tchi, Y., Jin, C, Li, Y. and Yan, Y. (1988). 

Tabulatomorphic Corals. Monograph of Palaeozoic 
Cora/5 1,328-336; 2, 171-351. 

Lindstrom, G. (1876). On the affinities of the Anthozoa 
Tabulata. Annals and Magazine of Natural History 
series 4, 18, 1-17. 

Lindstrom, G. (1899). Remarks on the Heliolitidae. 
Handlingar Kongliga Svenska Vetenskaps- 
Akademiens, XXXH, No.l, 1-140. 

Milne-Edwards, H. and Haime, J. (1849). Memoire sur 
les polypiers appurtenant aux groupes naturels des 
Zoanthaires perfores et des Zoanthaires tabules. 
Academic Science Paris, Comptes Rendus, 29, 257- 
263. 



Proc. Linn. Soc. N.S.W., 127, 2006 



183 



SILURO-DEVONIAN TABULATE CORALS 



Milne-Edwards, H. and Haime, J. (1851). Monographic 

des polypiers fossils des terrains paleozoiques. 

Museum Histoire Naturales, Paris, Archives 5, 1-502, 

pi. 1-20. 
Munson, T.J., Pickett, J.W. and Strusz, D.L. (2000). 

Biostratigraphic review of the Silurian tabulate corals 

and chaetetids of Australia. Historical Biology, 15, 

41-60. 
Pickett, J.W. (1975). Continental reconstructions and 

the distribution of coral faunas during the Silurian. 

Journal and Proceedings, Royal Society of New 

South Wales, 108, 147-156. 
Pickett, J.W. (2002). OzCorals database. Version 2. http:// 

www.es.mq.edu.au/mucep/aap/downloads/ozcorals2. 

htm 
Pickett, J.W. and McClatchie L. (1991). Age and relations 

of stratigraphic units in the Murda Syncline area. 



Geological Survey of New South Wales, Quarterly 

Notes 85, 9-32. 
Regnell, G. (1941). On the Siluro-Devonian fauna of 

Choltagh, Eastern Tien-shan. Palaeontologia Sinica, 

17, Part I: Anthozoa, 1-63. Nanking, Geological 

Survey of China. 
Sherwin, L. (1996). Narromine 1:250,000 Geological 

Sheet SI/55-3: Explanatory Notes, 1-104. Geological 

Survey of New South Wales, Sydney. 
Wang, H.C. (1981). Tabulate and heholitid corals. 

Palaeontological Atlas of Northwest China Sinkiang 

Autonomous Region, 39-72. Geological Publishing 

House, Beijing (in Chinese). 
Yang, S. et al (1978). Tabulata. In: Palaeontological 

Atlas of Southwest China, Guizhou volume. Part 

I, Cambrian to Devonian, 161-250. Geological 

Pubhshing House, Beijing (in Chinese). 



Table! 

Comparison of parameters distinguishing species of Pseudoplasmopora 

discussed in text 



Diameter No. oftubuli Tabularial Spacing of Spacing of 

of surrounding spacing tabulae diapliragms Septa 
tabularium tiie aureole per cm^ in 5 mm in 5 mm 



P.follis 0.9-1.1 12 25-30 10-15 

P. heliolitoides 1.0-1.75 12 5-7 12-16 

P. gippslandica 1.2-1.5 12 4-5 10-15 

P. sp. A 1.0-1.1 12 12-15 14-16 

P. sp. B 0.4-0.5 12 50-60 20-25 



15-16 absent 
15 lump 

15-20 absent 
8-10 blunt 

16-18 spines 



184 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLDVARY 




Figure 4. Pseudoplasmopora follis A, B. Longitudinal sections of AMF105567 (AM14105) from 
Loc. XXX, 1.5 km ENE of "Meloola" Homestead, about 40 km NNE of CondoboUn, NSW. 
C. Transverse section of AMF116146 (AM14098) from Loc. XX, east of "Moorefield" Sta- 
tion, 40 km north of Condobolin, NSW. D. Transverse section of AMF116148 (AM13783) also 
from Loc, XX. E. Transverse section and F. Longitudinal section of MMF31447 (Geological Sur- 
vey of N.S.W.), 3.5 km W of 'Moorefield' Station, N of Condobolin, N.S.W. Scale bar = 1 cm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



185 



SILURO-DEVONIAN TABULATE CORALS 




Figure 5. Pseudoplasmopora heliolitoides. A, B. Transverse sections of AMF69668 (AM13547) from Loc. 
XX). Other transverse sections (unillustrated) are: AM13635, AM13772 and AM13773. C, D. Longitu- 
dinal sections of AMF69668. Otiier longitudinal sections (unillustrated) are: AM13636, AM13637 and 
AM13774. E. Transverse section and F. Longitudinal section of AMF78962 (AM257), probably from Hume 
Limestone scree at mouth of Booroo Ponds Creek, Hattons Corner, Yass River, N.S.W. Scale bar = 1 cm. 



186 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLD VARY 





A fe^gF-^^?a!5t^>«3a9^^3^ 



W V 'u, i. 



B 





Figure 6. Pseudoplasmopora heliolitoides. A. Longitudinal section and B. Transverse section of AMF5556 
(AM76), syntype of Heliolites distans var. intermedia Dun, 1927. C. Longitudinal section and D. Transverse 
section of AMF5173 (AM56), lectotype ofH. distans chosen by Jones and Hill, 1940. E. Longitudinal section 
and F. Transverse section of AMF4082 [AM 140 (AM65)], paralectotype ofH. distans. Scale bar = 1 cm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



187 



SILURO-DEVONIAN TABULATE CORALS 




'•1' •i'SI .'in.' . •,«■ •' '■ 





Figure 7. Pseudoplasmopora gippslandica. A. Transverse section and B. Longitudinal sec- 
tion of AMF5512 (AM66), near Rockhampton, Queensland, figured by Jones and Hill 
(1940). C. Transverse section and D. Longitudinal section of AMF6936 (AM271), Nundle 
Road, near Tamworth, New South Wales, figured by Jones and Hill (1940). Scale bar = 1 cm. 



188 



Proc. Linn. Soc. N.S.W., 127, 2006 



G.Z. FOLDVARY 






Figure 8. Pseudoplasmopora sp. A. A. Transverse section and B. Longitudinal section of UQF58203, 
from Graveyard Creek Formation, Silurian; figured also by Hill et al. (1969), pi. Ill, fig. 7, 
Pseudoplasmopora sp. B. C. Transverse section and D. Longitudinal section of UQF60060, from Grave- 
yard Creek Formation, Silurian; figured also by Hill et al. (1969), pi. HI, fig. 8. 

Pseudoplasmopora follis. E. Transverse section and F. Oblique section of UQF60059, from Perry Creek 
Formation, Silurian; figured also by Hill et al. (1969), pi. HI, fig. 6. 
Scale bar = 1 cm. 



Proc. Linn. Soc. N.S.W., 127, 2006 



189 



190 



Vegetation Responses to Pinus radiata (D. Don) Invasion: A 
Multivariate Analysis Using Principal Response Curves 

Andrew C. Baker*, Grant C. Hose and Brad R. Murray 

Institute for Water and Environmental Resource Management, Department of Environmental Sciences, 
University of Technology Sydney, Broadway, NSW 2007, Australia 
* Corresponding author Email: Andrew.C.Baker@student.uts.edu.au 



Baker, A.C., Hose, G.C. and Murray, B.R. (2006). Vegetation responses to Pinus radiata (D. Don) 
invasion: a multivariate analysis using principal response curves. Proceedings of the Linnean Society of 
New South Wales 121, 191-197. 

Radiata pine (Pinus radiata D. Don) has been introduced to many new regions outside its native range as a 
plantation species. Plantations are frequently located adjacent to native vegetation. This proximity allows 
not only pine wildings, but also large amounts of non-native leaf litter, to enter the surrounding natural 
vegetation. Our aim in the present study was to assess the composition of plant communities in vegetation 
surrounding plantations in relation to proximity to pine plantations. Using multivariate Principal Response 
Curves (PRC) analysis, we show significant differences in the composition of native vegetation between 
transects adjacent to and not adjacent to pine plantations. Species-level analysis identified a suite of native 
species that were frequently found in fransects adjacent to pine plantations, and a second suite of native 
species that were reduced in abundance in fransects next to pine plantations. This second group of species 
should be the focus of future conservation work, since they appear to be sensitive to disturbance wrought by 
pine plantations. We show that the ability of PRC analysis to reveal both community-level and species-level 
responses to disturbance brought about by exotic species can lead to the generation of testable hypotheses 
bridging species and community ecology. 

Manuscript received 4 May 2005, accepted for publication 16 December 2005. 

KEYWORDS: Invasion, pines, Pinus radiata, PRC, principal response curves, remnant vegetation. 

INTRODUCTION invasive and can readily escape and establish in native 

vegetation, beyond the boundaries of the plantation 

The timber of Pinus radiata (D. Don) is (Richardson and Higgins 1998). For instance, prior 

valuable because of its numerous applications in to removal of the Joimama Pine Plantation (southern 

manufacturing and construction industries (Sutton NSW, Australia), an estimated 16 000 Pinus spp. 

1999). Consequently, extensive areas now support individuals were growing in 24 000 ha of adjacent 

plantations globally, making P. radiata the most native vegetation in Kosciusko National Park (Leaver 

commonly grown conifer in the world (Lavery 1983). 

and Mead 1998). In Australia, pine plantations are The establishment of P. radiata within 

commonly bordered by native vegetation (Williams remnant native vegetation has been linked to the 

and Wardle 2005), and are frequently established displacement of native plant species (Richardson et 

amongst native vegetation and/or contain areas of al. 1994, Richardson and Higgins 1998, Holmes et 

remnantvegetationwithintheplantation(Lindenmayer al. 2000, Morgan et al. 2000). When pines become 

et al. 2002, Lemckert et al. 2005). Of particular well established, as in plantations, the richness of 

concern is the occurrence of plantations in close bryophytes, vertebrates and invertebrates is reduced 

proximity to areas set aside for conservation purposes. relative to undisturbed native forest (Bonham et al. 

For example, to the west of Sydney, pine plantations 2002, Lindenmayer and Hobbs 2004, Parris and 

border Blue Mountains National Park, Kanangra- Lindenmayer 2004, Pharo et al. 2004). In this study, 

Boyd N.P., and Jenolan Caves Karst Conservation we compare patterns of plant distribution between 

Reserve. The close proximity of plantations to native patches of remnant vegetation that are adjacent to, and 

vegetation is problematic because P. radiata is highly not adjacent to pine plantations. Those areas adjacent 



VEGETATION RESPONSES TO PINUS RADIATA 



to pine plantations are not only subject to invasion 
from pine wildlings, but also receive a large amount 
of pine litter (needles, pollen cones) from the adjacent 
plantation (Baker 2004). This material may smother 
ground-covering plants, and may alter soil chemistry 
and frirther facilitate change in the composition of 
vegetation communities. Our hypotheses are; 1) that 
there will be a significant difference in plant species 
richness and abundance between areas adjacent to 
and not adjacent to P. radiata plantations, and 2) that 
these differences will decrease with distance from the 
plantation as the number of wildings and pine litter 
also decreases. 

To test these hypotheses, we use the multivariate 
method of Principal Response Curves (PRC). The 
PRC technique has been used widely in ecotoxicology 
(e.g. Cuppen et al. 2000, Hose et al. 2003, Belanger et 
al. 2004) and is gaining some recognition as a tool for 
biomonitoring studies in aquatic ecology (e.g. Leonard 
et al. 1999, Pardal et al. 2004). PRC analysis has only 
very recently been used in vegetation ecology (e.g. 
Pakeman et al. 2003, Heegaard and Vandvik 2004, 
Vandvik et al. 2005). The utility of PRC analysis is 
not yet widely recognised for ecological studies. Our 
investigation of the impact of pine plantations on the 
composition of remnant vegetation provided us with 
an ideal opportunity to demonstrate the applicability 
of the PRC technique for ecological studies. 



METHODS 

Study Location and Design 

This study was conducted in the Jenolan Karst 
Conservation Reserve (33° 49'S, 150° 02'E), in 
southeast Australia, from April to August 2004 
(Baker 2004). The native vegetation of the study 
area is Eucalyptus spp. dominated woodland, with 
an understorey dominated by herbs, grasses and the 
occasional larger shrub species (e.g. Acacia longifolia 
and Exocarpus strictus). 

A narrow trail (~5 m wide) separates the Jenolan 
Karst Conservation Reserve from adjoining areas of 
mature P. radiata plantation and remnant forest. This 
allowed us to place sampling transects in woodland 
areas adjacent to P. radiata plantations (nominally 
'disturbed' areas) and in woodland areas adjacent to 
renmant vegetation (nominally 'undisturbed' areas). 
Nine replicate transects were randomly placed in 
disturbed and nine in undisturbed areas. The areas 
were all similar in altitude, aspect, slope, fire history, 
topography and soil type, but differed in being 
either adjacent to or not adjacent to pine plantations. 
Consequently, we have confidence in attributing 



any observed differences in vegetation composition 
between sites to the disturbance, having minimised 
confounding inter-site differences. Transects were 
laid parallel to each other and perpendicular to the 
trail, extending 50 m into the reserve. Sampling was 
conducted at 10 m intervals along the fransect line (i.e. 
6 samples per transect), using a 2 m x 5 m quadrat. 
Our study focused on herbs and shrubs as these are 
most likely to be affected by pine invasions, hence 
we ignored the canopy species {Eucalyptus spp.) as 
we considered them to have been established prior to 
the pine plantation and thus minimally affected. All 
vascular plants within the quadrat were identified and 
the percent canopy cover of each species within the 
quadrat was recorded as a measure of abundance. 

Data Analysis 

The technique of Principal Response Curves is 
a novel multivariate statistical method. PRC analysis 
was developed for the analysis of multi-species 
data from experiments designed with replicated 
controls and treatments and, specifically, repeated 
temporal sampling (van den Brink and Ter Braak 
1999). The PRC method focuses on differences 
between treatment and controls at each sampling 
time. It provides simplified ordination plots in which 
temporal gradients are presented along a horizontal, 
unidirectional axis. Here, we show that PRC analysis 
is equally applicable to transect studies with repeated 
sampling over spatial, rather than temporal gradients. 
In the following description, a simplified overview 
of the PRC technique is provided, and readers are 
referred to van den Brink and Ter Braak (1999) for 
full details of the method. 

The PRC method is based on Redundancy 
Analysis (RDA) but is extended to adjust for changes 
in the controls over time (van den Brink and Ter 
Braak 1999). RDA can be considered a constrained 
form of Principal Components Analysis, meaning 
that patterns in the biological data are limited to that 
which can be explained by explanatory variables. 
Because in RDA the explanatory variables are fixed 
a priori, the total variance can be partitioned into 
explained and residual variances. 

A major problem with traditional ordination plots 
is that they may be congested, and differences among 
treatments and confrols, and temporal frajectories 
can be confiising, particularly when ordination plots 
contain repeated sampling over time or space (van den 
Brink and Ter Braak 1999). To avoid these problems, 
PRC uses explanatory variables that distinguish the 
controls from the treatment, and individual sampling 
events (times or in our case distances). Explanatory 
variables identifying the controls are deleted from 



192 



Proc. Linn. Soc. N.S.W., 127, 2006 



A.C. BAKER, G.C. HOSE AND B.R. MURRAY 



the analysis to ensure that the treatment effects are 
expressed as a deviation from the control, at each 
distance (van den Brink and Ter Braak 1999). 

We used CANOCO version 4 (Ter Braak and 
Smilauer 1998) to carry out the PRC analysis. Like 
its precursor RDA, PRC requires a linear response 
model (van den Brink and Ter Braak 1999). To 
test the suitability of a linear model, Detrended 
Correspondence Analysis was used to determine the 
gradient length (the beta diversity, or extent of species 
turnover) of the first axis. The gradient length of the 
first axis was 3.08. Gradients >4 suggest a unimodal 
model is needed because data are heterogeneous 
and many species deviate from the assumed linear 
response model. Gradients < 3 are better suited to 
analyses with linear models such as PCA or RDA 
(Leps and Smilauer, 2003). 

PRC also requires the use of the Euclidean 
Distance for sample (dis)similarities (van den Brink 
and Ter Braak 1999). Euclidean distance weights 
shared absences and shared presences of species 
equally in its assessment of similarity among 
samples, thus distances among samples are driven 
by differences in the abundance of taxa irrespective 
of whether those species are present or absent in 
either sample. This is desirable for this study because 
differences among samples are not greatly influenced 
by chance recordings of uncommon species. 

Plant abundance data were log(10x+l) 
transformed prior to analysis and all other 
recommended settings were used (van den Brink and 
Ter Braak 1 999). Because plants multiply or die, count 
data are naturally modelled by proportional changes, 
i.e. by a multiplicative model. We used a logarithmic 
fransformation, so as to turn the multiplicative model 
for the counts into a linear model (van den Brink and 
Ter Braak 1999) and to down- weight the dominant 
species in the vegetation assemblages. Multiplication 
by the constant (10) avoided false discrepancy 
between zero and low abundance values (van den 
Brink etal. 1995). 

In a PRC diagram, the horizontal axis represents 
the distance along transects of the experiment and the 
vertical axis represents the freatment effect (Canonical 
Coefficient C^,) expressed as deviations from the 
control. The accompanying species weights allow 
an interpretation of effects at the species level. In the 
present study, taxa with negative species weights are 
expected to increase in abundance in the disturbed 
areas relative to the undisturbed areas, and taxa with 
positive species weights are expected to decrease 
in abundance in the disturbed areas relative to the 
undisturbed areas. Taxa with near zero weights either 
show no response or a response that is unrelated to 



the PRC (van den Brink and Ter Braak 1999). 

The significance of the treatment regime was 
tested using Monte Carlo tests and permuting whole 
transects among disturbed and undisturbed areas. 
Further Monte Carlo tests were performed to test 
the significance of differences at each distance along 
transects. This was achieved by conducting Monte 
Carlo tests using only those data for the distance of 
interest (van den Brink et al. 1996). For the Monte 
Carlo tests, a binary coded explanatory variable 
was used to distinguish transects in disturbed and 
undisturbed areas in the analysis. The Monte Carlo 
permutation tests are based on an F-type statistic, and 
the significance level (a) was 0.05. 



RESULTS 

Our PRC analysis detected significant differences 
(p = 0.005) in the composition of plant communities 
between disturbed and undisturbed areas. Differences 
among treatments accounted for 10.5% of all 
variance, while differences among sampling distances 
accounted for 6.1% of all variance. The remainder 
was attributed to variability among replicates. The 
response pattern in the first PRC axis was significant 
(p = 0.03), and this axis captured a much greater 
proportion of the total variance explained by the 
treatment regime than the second axis, which was not 
significant (p = 0.205). For this reason, only the first 
axis of the PRC analysis is presented. 

At the frail (distance = m) and closest to the 
plantations, the disturbed and undistjirbed areas 
differed greatly, but became more similar with 
increasing distance along transects into the remnant 
vegetation (Fig. 1). This pattern was consistent with 
the results of Monte Carlo tests, which detected 
significant (p<0.05) freatment effects at and 10 m 
but not at greater distances into the remnant vegetation 
(p>0.05,Fig. 1). 

The difference among disturbed and undisturbed 
areas was most sfrongly driven by differences in 
the abundance of P. radiata, which had a sfrongly 
negative species weight (Fig. 1). Native species 
with large negative species weights were Lomandra 
longifolia, Leucopogon lanceolatus, Poranthera 
microphylla, and Cassinia aculeata, suggesting an 
affinity of these species to the disturbed areas. Native 
species with strongly positive species weights, such as 
Persoonia acuminata, Monotoca scoparia, Clematis 
aristata and Stellaria pungens were more abundant 
in the undisturbed sites, demonstrating the sensitivity 
of these species to the disturbance associated with the 
penefration of pine litter. 



Proc. Linn. Soc. N.S.W., 127, 2006 



193 



VEGETATION RESPONSES TO PINUS RADIATA 



Undisturbed areas not adjacent to plantations 



O — O Disturbed areas adjacent to plantations 



Dt3eanoe(m) 



0.0 



-fl-l 



S -0-2 



10 



20 



30 



40 



50 



fNJ.10 



- — O 



fH).07 



p=0.19 



-0.3 



1 



p=0.19 



-0.4 



-OS 



^}.6 




p*0 01 



p=O.0l 



1 I 

\ 
- 

.1 ; 

.1 

I 
I 

-2 



-8 



P^rsoonia scuminsta 
hkmatoca scoparia 
Qematis aristata 
Steflafia pungens 
Viola betofMCifolia 
Asperala scopoma 
Dtchondra mpem 
Se/wciospp. 
L»ucopt2gim jitnlpsmus 
Undet. Grass 2 
fta/teftesp. 1 
Hypochomris rmMsia 
Lomatia tnyricoid^ 
Vsronica calycina 
Poa siebere/ra 
Plants^ o'etNfe 

DmnelSa sp 2 
BilSardiera scanderts 
Ummndra Mformis 
Cassifia aculeata 
Pofanthera microphyHa 
Leucopogon lanceomtus 
LomaiKira Song/folia 



Pinus mcHeia 



Figure 1. Principal response curve with species weights for vegetation assemblage data from veg- 
etation transects in areas adjacent to and not adjacent to pine plantations. Species with weights 
between 0.5 and -0.5 have been omitted for clarity. Probability (p) values indicate the outcomes of 
Monte Carlo tests per sampling distance. 



DISCUSSION 

The significant differences in the distribution of 
plant species located at disturbed and undisturbed 
areas in this study are highly consistent with 
previous research in the southern hemisphere that has 
documented the displacement of native plant species 
following invasion by P. radiata (Richardson et al. 
1994, Richardson and Higgins 1998, Holmes et al. 
2000, Morgan et al. 2000). We have also shown that 
differences in the composition of plant communities 
between disturbed and undisturbed areas decrease 
with distance fi-om plantations, such that there is 
no significant difference in plant species richness 
and abundance at 20 m and beyond. The increasing 
similarity of plant communities with increased 
distance fi-om plantations is consistent with work 
showing an exponential decline in the mass of pine 
litter with increasing distance from plantations (Baker 
2004). 

Relative changes in the composition of native 
plant communities between disturbed and undisturbed 



areas are clearly evident and easily interpretable in the 
PRC, therein highlighting a significant advantage of 
this approach over traditional ordination techniques. 
The method also has the distinct advantage over 
other approaches in that the ordination allows a 
simplified interpretation of species-level patterns in 
the data. Thus, this method is likely to detect subtle 
changes that may occur in only a few species in the 
assemblage (Pardal et al. 2004). The PRC method is 
currently limited to using Euclidean distance as the 
(dis)similarity measure. The trade off is the ability 
for PRC to include a species-level analysis. Other 
analyses that permit a broader range of indices (e.g. 
similarity analysis, MDS, ANOSIM) have a limited 
ability to display effects on particular species (although 
species patterns can be shown through supplementary 
analyses such as SIMPER). An advantage of PRC 
analysis is that it integrates sample ordinations and 
species patterns in a single analysis (van den Brink 
and TerBraak 1999). 

Accompanying the increase in pine abundance, 
the vegetation of disturbed sites contained a greater 
abundance of Lomandra longifolia, Leucopogon 



194 



Proc. Linn. Soc. N.S.W., 127, 2006 



A.C. BAKER, G.C. HOSE AND B.R. MURRAY 



lanceolatus, Poranthera microphylla, and Cassinia 
aculeata than at undisturbed sites. Several of these 
species are able to colonise disturbed habitats (B.R. 
Murray pers. obs.), however, there are very few 
herbarium records that describe whether or not these 
species are characteristic of disturbed areas. The 
exception is Cassinia aculeata, which is known to be 
a fast growing pioneer species that regenerates from 
seed following a disturbance (CSU herbarium 2005) 
and frequently inhabits disturbed areas (Fairly and 
Moore 2002). In this study, the occurrence of these 
species in the sites adjacent to the pine plantations 
suggests an ecological disturbance in those areas, as 
a result of the close proximity of pine plantations to 
native vegetation. In contrast, the vegetation at the 
undisturbed areas contained a greater abundance of 
Persoonia acuminata, Monotoca scoparia. Clematis 
aristata and Stellaria pungens. The species in this 
group are all small native shrubs, herbs and climbers 
(Harden 1990-1993, Fairly and Moore 2002). 
Their relatively lower abundance in disturbed areas 
suggests they are sensitive to the disturbance caused 
by the close proximity of pine plantations to native 
vegetation. 

All the species discussed above are common 
in woodland communities across the study region 
(Fairly and Moore 2002, PlantNET 2005), and cannot 
be distinguished into two groups based on regional 
abundance (i.e. none of these species are uncommon 
on a broader scale). The species are all perennial 
except for Poranthera microphylla (which is a 
small annual herb (Fairly and Moore 2002), are of 
similar growth form (herb or small shrub), but vary in 
maximum height (PlantNET 2005). Future research 
should include manipulative experiments to contrast 
the growth of these potentially sensitive species 
between areas with and without pines and pine litter. 

Clearly, our findings are correlative and further 
experimental work is required to link pine plantations 
with changes in plant species richness and abundance. 
However, our results do suggest that there is an edge 
effect associated with pine plantations. Such an edge 
effect may be caused by factors including altered 
microclimate surrounding pine plantations or the 
presence of pine litter that can penetrate remnant 
vegetation up to 50 m from adjoining plantations 
(Baker 2004). Pine litter may smother herbs or small 
shrubs, possibly explaining the reduced abvmdance of 
some such species in the disturbed sites. Plantations 
where pine litter dominates the forest floor have 
altered litter decomposition and nutrient cycling rates 
compared to native vegetation (Scholes and Nowicki 
1998). Similar patterns may occur in native forests 
and woodlands where pine material also dominates 



the litter, which may also explain the patterns in 
vegetation composition we observed. Indeed, Burdon 
and Chilvers (1994) report that a discontinuous carpet 
of pine needles and shading from individual pines 
growing in native vegetation results in a changed 
environment, and ultimately changes to plant 
communities. 

The close proximity of pine plantations to native 
vegetation appears to have a significant impact on 
composition of plant communities. It is our prediction 
that the patterns we observed in the vegetation 
assemblages are the result of the introduction 
of pine litter and altered ecosystem functioning. 
Consequently, we expect similar edge effects to occur 
wherever remnant vegetation abuts plantations and 
pine litter is exchanged. 

Our novel use of PRC analysis has identified 
significant effects at the community level, as well as 
particular species that may be tolerant or sensitive to 
disturbance brought about by the close proximity of 
native vegetation to pine plantations. It is the focus 
of our future research to better understand both the 
intrinsic factors (e.g. life-history traits) and extrinsic 
factors (e.g. seedling growth under soils exposed to 
pine litter leachate) that lead to this dichotomy. 



ACKNOWLEDGMENTS 

We thank the Department of Envirormiental Sciences 
(UTS), the Jenolan Caves Reserve Trust, and the 
Lirmean Society of NSW for financial and logistical 
support. Pete Mitchell kindly commented on a draft 
of the manuscript. 



REFERENCES 

Baker, A.C. (2004). Ecosystem responses to Pinus radiata 
invasion. Unpublished Honours thesis. University of 
Technology, Sydney. 

Belanger, S.E., Lee, D.M., Bowling, J.W. and LeBlanc, 
E.M. (2004). Responses of periphyton and 
invertebrates to a tetradecyl-pentadecyl sulfate 
mixture in stream mesocosms. Environmental 
Toxicology and Chemistry 23, 2202-2213. 

Bonham, K.J., Mesibov, R. and Bashford, R. (2002). 
Diversity and abundance of some ground- 
dwelling invertebrates in plantation vs. native 
forests in Tasmania, Australia. Forest Ecology and 
Management 158, 237-247. 

Burdon, J.J. and Chilvers, G.A. (1994). Demographic 
changes and the development of competition in a 
native Australian eucalypt forest invaded by exotic 
pines. Oecologia 97, 419-423. 



Proc. Linn. Soc. N.S.W., 127, 2006 



195 



VEGETATION RESPONSES TO PINUS RADIATA 



Charles Sturt University Herbarium (2005). http://www. 
csu.edu.au/herbarium/ viewed 18/4/2005 

Cuppen, J.G.M., Crum, S.J.H., van den Heuvel, H.H., 
Smidt, R.A. and van den Brink P.J. (2002). Effects 
of a mixture of two insecticides in freshwater 
microcosms: I. Fate of chlorpyrifos and lindane and 
responses of macroinvertebrates. Ecotoxicology 11, 
165-180. 

Facelli, J.M. and Pickett, S.T.A. (1991). Plant litter: its 
dynamics and effects on plant community structure. 
The Botanical Review 57, 1-32. 

Fairly, A. and Moore, P. (2002). 'Native plants of the 
Sydney district, an identification guide'. (Kangaroo 
Press: Sydney). 

Harden, G.J. (1990-1993). 'Flora of New South Wales'. 
(New South Wales University Press: Sydney). 

Heegaard, E. and Vandvik, V. (2004). Climate change 
affects the outcome of competitive interactions - an 
application of principal response curves. Oecologia 
139, 459-466. 

Holmes, P.M., Richardson, D.M., van Wilgen, B.W. and 
Gelderblom, C. (2000). Recovery of South Afiican 
fynbos vegetation following alien woody plant 
clearing and fire: implications for restoration. Austral 
Ecology 15, 62,1-6^9. 

Hose, G.C., Lim, R.P, Hyne, R.V. and Pablo, F. (2003). 
Short-term exposure to aqueous endosulfan affects 
macroinvertebrate assemblages. Ecotoxicology and 
Environmental Safety 56, 282-294. 

Lavery, P.B. and Mead, D.J. (1998). Pinus radiata takes on 
the world. In 'Ecology and Biogeography of Pinus' 
(ed. D.M. Richardson) pp. 432-449. (Cambridge 
University Press: Cambridge). 

Leaver, B.H. (1983). 'Harvesting and Rehabilitation of 
Jounama Pine Plantation, Kosciusko National Park. 
Environmental Impact Statement'. (National Parks 
and Wildlife Service of New South Wales: Sydney). 

Lemckert, F., Brassil, T. and Towerton, A. (2005). Native 
vegetation corridors in exotic pine plantations 
provide long term habitats for fi^ogs. Ecological 
Management and Restoration 6, 132-134. 

Leonard, A.W., Hyne, R.V., Lim R.P. and Chapman, J.C. 
(1999). Effect of endosulfan runoff fi^om cotton 
fields on macroinvertebrates in the Namoi River. 
Ecotoxicology and Environmental Safety 42, 125- 
134. 

Leps, J. and Smilauer, P. (2003). 'Multivariate analysis 
of ecological data using CANOCO'. (Cambridge 
University Press: Cambridge). 

Lindenmayer, D.B. and Hobbs, R.J. (2004). Fauna 
conservation in Australian plantation forests - a 
review. Biological Conservation 119, 151-168. 

Lindenmayer, D.B., Curmingham, R.B., Dormelly, C.F., 
Nix, H. and Lindenmayer, B.D. (2002). Effects of 
forest fragmentation on bird assemblages in a novel 
landscape context. Ecological Monographs 71, 1-18. 

Morgan, V.C, Hoffmann, J.H., Donnelly, D., van Wilgen, 
B.W. and Zimmermann, H.G. (2000). Biological 
Control of Alien Invasive Pine Trees {Pinus 



species) in South Africa. In 'Proceedings of the X 
international Symposium on Biological Control of 
weeds 4-14 July 1999' (Ed. N.E. Spencer) pp. 941- 
953. (Montana State University: Montana, USA). 

Pakeman, R.J., Hulme, P.D., Torvell, L. and Fisher, J.M. 
(2003). Rehabilitation of degraded dry heather 
[Calluna vulgaris (L.) Hull] moorland by controlled 
sheep grazing. Biological Conservation 114, 389-400. 

Pardal, M.A., Cardoso, P.G., Sousa, J.P., Marques, J.C. 
and Raffaelli, D. (2004). Assessing envirorunental 
quality: a novel approach. Marine Ecology-Progress 
Series 161, 1-8. 

Parris, K.M. and Lindenmayer, D.B. (2004). Evidence 
that creation of Pinus radiata plantations in south- 
eastern Australia has reduced habitat for frogs. Acta 
Oecologica 15, 93-101. 

Pharo, E.M., Lindenmayer, D.B. and Taws, N. (2004). The 
effects of large-scale fragmentation on bryophytes 
in temperate forests. Journal of Applied Ecology 41, 
910-921. 

PlantNET (2005). New South Wales Flora Online, http:// 
plantnet.rbgsyd.nsw.gov.au/. Viewed 18/4/2005. 

Richardson, D.M, Williams, PA. and Hobbs, R.J. 

(1994). Pine invasions in the Southern Hemisphere: 
determinants of spread and invadability. Journal of 
Biogeography 21, 51 1-527. 

Richardson, D.M. and Higgins, S.I. (1998). Pines as 

invaders in the southern hemisphere. In 'Ecology and 
Biogeography of Pinus' (ed. D.M. Richardson) pp. 
432-449. (Cambridge University Press: Cambridge). 

Scholes, M.C. and Nowicki, TE. (1998). Effects of pines 
on soil properties and processes. In 'Ecology and 
Biogeography of Pinus' (Ed. D.M. Richardson) pp. 
432-449. (Cambridge University Press: Cambridge). 

Semmartin, M., Aguiar, M.R., Distel, R.A., Moretto, 
A.S. and Ghersa CM. (2004). Litter quality and the 
nutrient cycling affected by grazing-induced species 
replacements along a precipitation gradient. Oikos 
107, 148-160. 

Sutton, W.R.J. (1999). The need for planted forests and the 
example of radiata pine. New Forests 17, 95-109. 

Ter Braak, C.J.F. and Smilauer, P (1998). 'CANOCO 
reference manual and user's guide to Canoco for 
windows: Software for Canonical ordination (Version 
4).' (Microcomputer Power: New York). 

van den Brink, PJ. and Ter Braak, C.J.F. (1999). Principal 
response curves: analysis of time-dependant 
multivariate responses of a biological community to 
stress. Environmental Toxicology and Chemistry 18, 
138-148. 

van den Brink, P.J., van Donk, E., Glystra, R., Crum, 
S.J.H. and Brock, T.C.M. (1995). Effects of chronic 
low concentrations of the pesticides chlorpyrifos 
and atrazine in indoor freshwater microcosms. 
Chemosphere 31, 3181-3200. 

van den Brink, P.J., van Wijngaarden, R.P.A., Lucassen, 
W.G.H., Brock, T.C.M and 

Leeuwangh, P. (1996). Effects of insecticide Dursban 
4E (active ingredient chlorpyrifos) in outdoor 



196 



Proc. Linn. Soc. N.S.W., 127, 2006 



A.C. BAKER, G.C. HOSE AND B.R. MURRAY 



experimental ditches: II Invertebrate community 
responses and recovery. Environmental Toxicology 
and Chemistry 15, 1143-1153. 

Vandvik, V., Heegaard, E., Maren, I.E., Aarrestad, P.A. 
(2005). Managing heterogeneity: the importance 
of grazing and environmental variation on post-fire 
succession in heathlands. Journal of Applied Ecology 
42, 139-149. 

Williams, M.C. and Wardle, G.M. (2005). The invasion 
of two native eucalypt forests by Pinus radiata in 
the Blue Mountains, New South Wales, Australia. 

Biological Conservation 125, 55-64. 



Proc. Linn. Soc. N.S.W, 127, 2006 197 



198 



Silurian Linguliformean Brachiopods and Conodonts from the 
Cobra Formation, Southeastern New South Wales, Australia 

James L. Valentine, Damian J. Cole and Andrew J. Simpson 

Centre for Ecostratigraphy and Palaeobiology, Department of Earth and Planetary Sciences, Macquarie 

University, NSW 2109, Australia 



Valentine, J.L., Cole, D.J. and Simpson, A.J. (2006). Silurian linguliformean brachiopods and conodonts 
from the Cobra Formation, southeastern New South Wales, Australia. Proceedings of the Linnean 
Society of New South Wales 111, 199-234. 

Silurian linguliformean brachiopods and conodonts are documented and described from the type section 
through the Cobra Formation (Taralga Group) in Murruin Creek, near Taralga. The linguliformean brachiopod 
fauna includes linguloids (six taxa), discinoids (three taxa), acrofretoids (four taxa) and a siphonotretoid. 
These are the first Late Silurian linguliformean brachiopods to be documented from eastern Ausfralia. 
New taxa include Acrotretella dizeugosa sp. nov., upon which is based the first detailed description of the 
ontogeny of Acrotretella Ireland, 1961. Eleven multi-element conodont taxa are recognised, including the 
temporally significant taxon, Kockelella maenniki Serpagli and Corradini, 1998. Based on these conodont 
data, and other faunal elements, the Cobra Formation in Murruin Creek appears to range from mid-Wenlock? 
to mid-Ludlow (early to vcaA-siluricus Zone) in age. 

Manuscript received 23 June 2005, accepted for publication 7 December 2005. 

KEYWORDS: Brachiopods, Cobra Formation, Conodonts, Linguliformea, Ludlow, New South Wales, 
Silurian. 



INTRODUCTION 

The Cobra Formation (Taralga Group) crops out 
in a thin, north-south trending belt east of Taralga 
in southeastern New South Wales (Fig. 1). Despite 
extensive studies of a number of sections through the 
Cobra Formation (eg. Jongsma 1968; Roots 1969; 
Scheibner 1973; Morritt 1979; Powell and Fergusson 
1979a; Pickett 1982; Matthews 1985), no detailed 
accounts or systematic descriptions of the numerous 
fossil groups from these sections have been published. 
The present investigation focuses on linguliformean 
brachiopods and conodonts recovered from the Cobra 
Formation in Murruin Creek, approximately 20 km 
north of Wombeyan Caves (Fig. 2). 

The only report of linguliformean brachiopods 
from the Taralga Group is restricted to a single 
occurrence of Schizotreta sp. from the base of the 
Cobra Formation in Murruin Creek (Sherwin 1970). 
Silurian linguliformean brachiopods from eastern 
Ausfralia are generally poorly known, with the only 
well-documented fauna being from the Early Silurian 



(Llandovery-Wenlock) Boree Creek Formation of 
central- western New South Wales (Dean- Jones 1979; 
Valentine and Brock 2003; Valentine et al. 2003). 
These are the first Late Silurian linguliformean 
brachiopods to be documented and described from 
eastern Australia. 

Previous accounts of conodonts from the 
Taralga Group are restricted to the Wombeyan 
Limestone (Sherwin 1969; revised by Pickett 1982), a 
biohermal unit interpreted as Late Silurian in age, and 
stratigraphically equivalent to the base of the Cobra 
Formation in Murruin Creek (Nay lor 1937; Jongsma 
1968; Scheibner 1973). Based on a single Pa element 
assigned to ' Spathognathodus' (= Pandorinella) 
exigua (Philip, 1966), Sherwin (1969) suggested that 
the Wombeyan Limestone was Early Devonian in age. 
However, in his biostratigraphic review of Australian 
Silurian conodonts, Simpson (1995a:339) stated that 
this element could be a morphotype of Ozarkodina 
confluens (Branson and Mehl, 1933), a late Silurian 
species. No conodonts have previously been reported 
from the Cobra Formation. 



SILURIAN BRACHIOPODS AND CONODONTS 



GEOLOGY AND STRATIGRAPHY 

The Early Silurian (mid-Wenlock?) to Early 
Devonian Taralga Group, cropping out east of Taralga, 
is an upward-shallowing, deepwater sequence 
deposited along the eastern limb of the Cookbundoon 
Synclinorium, on the western edge of the Capertee 
High in the Hill End Trough (Scheibner 1973; Powell 
and Fergusson 1979b; Matthews 1985) (Fig. 1). The 
Cobra Formation forms the basal unit of the Taralga 
Group and consists of -670 m of interbedded fine- 
grained micrites, siltstones and limestones (Pickett 
1982; Matthews 1985). Based on the fine detrital 
nature of the Cobra Formation, the orientation of 
fossil corals, and the occurrence of the calcareous 
alga, Pseudochaetetes Haug, 1883 in association with 
the tabulate coral, Entelophyllum sp., from the base 
of the Cobra Formation in Little Wombeyan Creek 
(Fig. 1), Pickett (1985) concluded that the Cobra 
Formation was of turbiditic origin. Disarticulated 
rhynchonelliformean brachiopods from the same 
horizons are all deposited concave side down, 
suggesting post-mortem transportation via traction 
currents (Matthews 1985). 

The Cobra Formation overlies the low grade 
metamorphic shales and greywackes of the Late 
Ordovician to Early Silurian Burra Burra Creek 
Formation (uppermost unit of the Triangle Group) 
(Figs 1, 2). The contact between the two units is 
widely stated to faulted, or a high angle unconformity, 
and a significant time break has been implied to exist 
between them (Jongsma 1968; Roots 1969; Scheibner 
1973; Talent etal. 1975; Powell etal. 1976; Powell and 
Fergusson 1979a, b; Pickett 1982). In contrast, both 
Morritt (1979) and Matthews (1985) have argued that 
this contact is paraconformable (though sometimes 
faulted) as in Murruin, Kerrawary, Guineacor and 
Cowhom creeks, or gradational over about 15 m as in 
Little Wombeyan Creek (Fig. 1). 

No evidence of a high angle unconformity 
between the Burra Burra Creek and Cobra formations 
was observed in Murruin Creek. The contact is 
marked by a prominent, 14 m thick conglomeritic 
horizon, whose upper boundary marks the start of 
the MU section (Figs 2, 3). Matthews (1985) argued 
that this conglomeritic horizon only occurs where 
faulting (parallel and/or subparallel to bedding) exists 
between the Burra Burra Creek and Cobra formations. 
The fault, and associated conglomerate, can occur 
within either formation, or as in Murruin Creek, at the 
contact between the two. Where faulting is absent, as 
in Little Wombeyan Creek, the conglomeritic horizon 
is also absent. Therefore, this horizon would appear 



to have originated through post-lithification tectonic 
activity (Matthews 1985). 

The first 468 m of the MU section through the 
Cobra Formation consists of well-bedded, grey- 
black shales (4-25 cm thick) interbedded with pale 
coloured, nodular limestone bands (1-6 cm thick) 
and dark-grey limestone beds (up to 1.8 m thick) 
(Fig. 3). However, continuously exposed horizons 
are restricted to 126-171 m and 431-468 m above 
the base of the MU section (Fig. 3). Between these 
intervals, only sporadic outcrops of grey-black shales 
and nodular limestones, identical to those occurring 
in the interval 126-171 m above the base of the MU 
section, were observed. 

The only linguliformean brachiopod recovered 
fi"om this part of the MU section was a single 
dorsal valve of Orbiculoidea sp. from sample MU 
21 (174.6 m above the base of the section) (Table 
1). Conodonts fi^om this part of the MU section are 
all predominantly long ranging forms and include 
Panderodus unicostatus (Branson and Mehl, 
1933), Panderodus recurvatus (Rhodes, 1953) and 
Dapsilodus obliquicostatus (Branson and Mehl, 
1933) (Table 2). This fauna is broadly suggestive of a 
Wenlock to Pridoli age. 

Jongsma (1968), Roots (1969) and Scheibner 
( 1 973) all recorded Batocara mitchelli (Foerste, 1888) 
within the first 175 m of their respective sections 
through the Cobra Formation in Murruin Creek. This 
species ranges fi^om the mid-Wenlock to mid-Ludlow 
in Australia (Pickett et al. 2000). Corals identified by 
Sherwin (1969) and Pickett (1985) from the base of 
the Cobra Formation in Little Wombeyan Creek (Fig. 
1) belong to the Hatton's Comer coral assemblage 
(Strusz and Munson 1997; Munson et al. 2000) and 
suggests a late Wenlock to Ludlow age. Therefore, 
the base of the Cobra Formation would appear to be, 
at most, mid-Wenlock in age. 

Continuously cropping out horizons occur for 
the last 64.2 m of the MU section, beginning 605 m 
above its base (Fig. 3). This part of the MU section 
consists of well-bedded, dark-grey limestone horizons 
(up to 20 cm thick) interbedded with thicker intervals 
of soft, light brown mudstones between 605-623.1 
m above the base of the MU section. Several faults 
also occur in this part of the Cobra Formation (Fig. 
2) — one at 623. 1 m above the base of the MU section, 
where massive black limestones replace the mudstone 
horizons. These limestone horizons continue through 
to the top of the Cobra Formation (Fig. 3). This part 
of the MU section has undergone folding as part of 
the latest Devonian to early Carboniferous regional 
deformation event that affected the Hill End Trough 
(Powell etal. 1976). 



200 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



34°00'S 



34''00'S 




20 km 



Late Devonian to 
Early Carboniferous 

Early to IVIiddle 
Devonian 

? Early Silurian to 
Early Devonian 

? Late Ordovician 
to ? Early Silurian 



^1 > 

> V 



Sydney Basin 

Granite Plutons 

Lambie Group 

Bindoolc Porphyry 
Complex and equivalents 

Taralga Group 
Triangle Group 

Anticline 
Synclina 



Figure 1. Generalised regional geological map of the Taralga area showing where the Taralga Group 
crops out (modified after Powell and Fergusson 1979a). Study area in Murruin Creek is indicated by 
boxed area and enlarged in Fig. 2. 



Proc. Linn. Soc. N.S.W., 127, 2006 



201 



SILURIAN BRACHIOPODS AND CONODONTS 




Figure 2. Detailed geological map of study area in Murruin Creek, showing location of MU section. 
Note that the MU section runs from right to left. 



The change in Hthology to massive black 
limestones coincides with a dramatic increase in the 
number of linguliformean brachiopods and conodont 
elements recovered. The linguliformean brachiopod 
fauna is dominated by acrotretoids, particularly 
Opsiconidion ephemerus (Mergl, 1982) (Table 1). 
This species ranges from the upper Ludlow of the 
Kopanina Formation to the Pridoli Pozary Formation 
of the Czech Republic and broadly agrees with the 
Ludlow age determination for the upper part of the 
MU section based on conodont data (see below). In 
fact, the Murruin Creek linguliformean brachiopod 
assemblage is similar to that described by Mergl 
(2001) from the deepwater Ludlow Kopanina 
Formation in the Barrandian of the Czech Republic. 
The Kopanina fauna, also dominated by acrotretoids, 



Figure 3. (OPPOSITE) Stratigraphic column of 
the MU section showing lithology and all sam- 
pled horizons. Numbers on the left of each col- 
umn represent metres above the base of the MU 
section and those on the right, sample numbers. 
Detail of lithology and sampling for the 47 m of 
section from sample MU 1 to MU 21 is enlarged 
to the left. Note that due to scale, only those nodu- 
lar limestone horizons sampled in this interval 
are included. Detail of lithology and sampling 
for the 2.51 m of section from sample MU 25 to 
MU 30, and for the 0.9 m of section from sam- 
ple MU 32 to MU 38, are enlarged to the right. 



202 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



644 m 



170 m- 



160 m 



150m-: 



140 m- 



130m-i 




Flaggy sandstones and shales 

Massive black limestones 

Calclte bands 

No outcrop 

Dark - grey to black limestone beds 

Light coloured nodular limestone 
bands 

Grey - black shales and mudstones 

Grey - black shales interbedded with 
light coloured nodular limestone bands 

Sporadically cropping out horizons of 
grey - black shales interbedded with 
light coloured nodular limestone bands 





i_ 






Burra Burra 

Creek 
Formation 




o o o 
o o o o 


Conglomerate 

Black, thinly bedded shales 


$^ 






27 


Sample numbers 



Proc. Linn. Soc. N.S.W., 127, 2006 



203 



SILURIAN BRACHIOPODS AND CONODONTS 



Metres above base of MU section 




od 






0^ 




C/3 

e2 


Sample Numbers 
Brachiopod Taxa 


21 


31 


32 


34 


35 


36 


pseudolingulid gen. et sp. indet. 1 


w 










1 


1 


2 


dv 






1 




2 


2 


5 


Kosagittellal sp. 


w 






1 




1 




2 


Rowellellal sp. 


w? 












1 


1 


dv? 












2 


2 


frag 










1 


19 


20 


Paterula sp. 


w 












1 


1 


dv 












2 


2 


CO 










1 




1 


linguloid gen. et sp. indet. 1 


frag 












8 


8 


linguloid gen. et sp. indet. 2 


dv 










1 




1 


Orbiculoidea sp. 


w 








2 






2 


dv 


1 


2 


1 


2 


2 


6 


14 


frag 




22 


1 


6 


4 


7 


40 


Schizotreta sp. 


w 




2 








1 


3 


dv 




3 


4 




3 


6 


16 


frag 




4 


2 




12 


21 


39 


discinoid gen. et sp. indet. 1 


dv 










1 


1 


2 


Artiotreta sp. cf. A. longisepta 


dv 




29 


9 


1 


5 


19 


63 


CO 




1 


1 






1 


3 


Acrotretella dizeugosa sp. no v. 


w 




25 


6 




3 


2 


36 


dv 




23 


7 




5 


4 


39 


Opsiconidion sp. cf. O. ephemerus 


w 




4 


8 




8 


14 


34 


dv 




19 


24 




16 


20 


79 


Opsiconidion sp. 


dv 




8 


2 




3 


3 


16 


siphonotretid gen. et sp. indet. 1 


dv 








2 




1 


3 



Table 1. Distribution and abundance of each linguliformean brachiopod species recovered 
from productive samples along the MU section through the Cobra Formation in Murruin 
Creek. Abbreviations: w = ventral valve(s); dv = dorsal valve(s); frag = fragment(s); co = 
conjoined specimen(s). 



204 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



--(N — S — r-i — — rN^2«>2<» — — — L) 

tN '— •— ' ^ ^T 



SIBJOX 


fN 


r- 


m 


fN 


oo 


.— 1 


«N 


OS 
in 


Ov 


00 




'"' 


**** 


(N 


<N 


fN 


00 


r<-i 


r<-| 


OS 
fN 


•"* 


m 


^^ 


'-^ 


m 


so 


o 


fN 




00 


— ' 


^- 


tN 


0^179 


00 




- 






— 




>o 


r- 


















OS 


r^ 


OS 


o 










- 


so 


W"l 


r<-i 


o 


- 








cet'Q 


r- 






— 








t-~ 


o 


m 


- 


(N 












o 




so 














O 


OS 


OS 


OS 


>n 








V£f9 




<N 








— 






oo 


rn 




00 










fN 




r- 


OS 


so 












r<-i 


m 


- 


r^ 


'S- 






OS 


6Sf9 


•o 
m 


























































tN 










93^9 






o 


<N 


(Nl 


o 


- 


o 




— 


■* 


o 

(N 


- 




(N 


(N 




tN 


fN 


c^ 


fN 

OS 


- 


m 


— 


— 


fN 


so 


oo 


OO 


m 


oo 


— 


- 


^ 


t^IW 










































- 




























eiw 
























































fN 












- 


9829 


m 




































- 
















- 






tN 










9£S9 


















CN 




r<-i 


■* 




- 








v-> 


fN 


m 


— 


























fN 


t^EW 
























































fN 














9HI 


<S 


































fN 
































tN 


t7>Z,I 


o 
















m 


<N 
















SO 


■^ 




fN 


















tN 










6691 


ro 


































- 


































6191 


(N 








































fN 




























6'6t'l 


— 




































— 
































izn 


O 






































— 






























eot'i 


<y. 


































-> 


fN 
































i££\ 


00 




































^^ 




- 




























eiei 


^o 


































— 






- 




























t'oei 


r<\ 


































fN 


































r62i 


(N 


































fN 




- 






























9Z,Zl 


- 






















































- 














c 
o 

1 

O 

u 
o 

XI 

CO 

s 


1 ; 

a. / 

•1 

1 
1 . 

§ 


to 
C/5 


X5 

C/2 


on 


•a 

C/5 


2 


H 


(/5 


2 


1/2 


a 
on 


X> 

00 


00 


2 


XI 


t/5 


2 


CO 

00 


X 


o 

00 


2 


S. 


X 

1/3 


CO 

l>0 


X 

on 


2 


a 


X 

a. 


V3 


X 

1/3 


oo 


a. 


on 




1 
1 

05 


1 

1 

5 
o 

} 


.=2 
§ 

8 
1 


! 

a 
'o 

£; 

1 
s 

1 


3 
1 

1 

1 


8 
1 

1 
S 

1 


.1 
to 

a 
■« 

1 
1 

a 


1>1 

1 

-Si 

1 

a 

O 

<*^ 
o 

d. 

1 

a 
O 


C3 

i 
i 
i 

1 


:5 

s: 
s: 

-Si 

1 


W5 

C 

1 
u 

u 

•o 

c 
op 

C/3 
Crt 

CO 

C 



oc 

s 
o 
u 

JS 
■*.* 

e 
_o 

u 



01 

JS 

c 



O) 

p « 

S e 

2 s 

*■ .S 

« s 

a o 

c U 

o 






C 
a 
•o 

s 
s 

« 

e 

S 
o 



'u 



fN 

e2 



Proc. Linn. Soc. N.S.W., 127, 2006 



205 



SILURIAN BRACHIOPODS AND CONODONTS 



Abbreviation 


Explanation 


L 


valve length 


W 


valve width 


H 


valve height 


WI 


width of pseudointerarea 


LI 


maximum length of pseudointerarea 


Fa 


length of pedicle foramen 


F'w 


width of pedicle foramen 


Fp 


point of origin of pedicle foramen 


LS 


length of dorsal valve median septum 


BS 


point of origin of dorsal valve median septum 


MHS 


maximum height of dorsal valve median septum 


OSP 


point of origin of surmounting plate 


LSP 


length of sui-iiiounting plate 


WSP 


width of surmounting plate 


LP 


length of larval shell 


WP 


width of larval shell 


HP 


height of larval shell 


N 


number of measurements 


MEAN 


average value 


SD 


standard deviation 


MAX 


maximum value 


MIN 


minimum value 



Table 3. Abbreviations used for measurements (in |am) of linguli 
formean brachiopods. Abbreviations based on those of Popov and 
Holmer (1994:35, fig. 39). Where applicable, all measurements are 
made from the posterior margin. 



includes numerous small discinids and rarer 
occurrences of larger discinids, obolids, linguloids 
and a siphonotretoid (Mergl 2001). 

The majority of the conodonts recovered from 
this part of the MU section were the same long ranging 
forms occurring in the lower part of the MU section 
(Table 2). The lenticular and triangular elements of 
Belodella anomalis Cooper, 1974 recovered from 
Murruin Creek (Table 2) are all broad-based (Fig. 
4a-g). Simpson (1995b: 310) and Jeppsson (1989) 
noted a general morphological trend in this taxon 
of broad-based elements in the Ludlow (eg. Cooper 
1974:pl. 1, figs 1-10; Simpson et al. 1993:fig. 4G- 
I; Simpson 1995b:pl. 16, fig. 15) and narrower- 
based elements in the Pridoli (eg. Jeppsson 1989: 
pi. 1, fig. 15; Simpson 1995b:pl. 16, fig. 21). Farrell 
(2004), however, documented a relatively broad- 
based population of elements from the Camelford 
Limestone and interpreted this sequence as Pridoli in 
age. Sample MU 34 (642.6 m above the base of the 
section) also yielded a single Pa and M element of 
the temporally significant taxon, Kockelella maenniki 
Serpagli and Corradini, 1998 (Table 2). This species 
is restricted to the early to rmdi-siluricus Zone (mid- 
Ludlow) of Europe and North America (Corradini 
and Serpagli 1999; Serpagli and Corradini 1999). In 
addition, a pygidium, possibly of B. mitchelli, was 
recovered from sample MU 28 (624.6 m above the 
base of the section) (Fig. 3). This is also in general 



agreement with the Ludlow age 
determination for this part of the 
MU section. 

Conformably overlying 
the Cobra Formation in 
Cowhom, Kerrawary, Guineacor 
and Little Wombeyan creeks 
(Fig. 1), are thinly bedded (<1 
m thick), deep-water, turbiditic 
arenites, lutites and siltstones of 
the Argyle Formation (Scheibner 
1973; Pickett 1982; Matthews 
1985). In Murruin Creek, the 
Cobra Formation is conformably 
overlain by the Whipbird Creek 
Formation (Fig. 2), a turbiditic 
sequence of interbedded 
sandstones and shales that may 
represent a distal facies of the 
Argyle Formation. Matthews 
(1985) reported rare boundary 
faulting between the Cobra 
and Argyle formations in Little 
Wombeyan Creek and similar 
faulting occurs in the upper 
part of the Cobra Formation in 
Murruin Creek (Fig. 2). The contact between the 
Cobra and Whipbird Creek formations lies -670 m 
above the base of the MU section (Fig. 3), compared 
to only 550 m reported by Scheibner (1973), Pickett 
(1982) and Matthews (1985). Given the folding and 
faulting occurring in this part of the Cobra Formation, 
the possibility of repeated horizons in the MU section 
carmot be dismissed. 



SYSTEMATIC PALAEONTOLOGY 

Discussion 

All type, paratype and figured materials are 
lodged in the palaeontological collections of the 
Australian Museum, Sydney (AM F). 

Phylum Brachiopoda Dumeril, 1 806 

Measurements 

Measurements (in jiim) of linguliformean 
brachiopods are based on those of Popov and 
Holmer (1994:35, fig. 39). Abbreviations used for 
the measurement of all taxa are listed in Table 3. 
Note that the width of some incomplete specimens 
was determined by measuring half the width 
and multiplying by two, assuming a bilaterally 
symmetrical organism. 



206 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



Order Lingulida Waagen, 1885 
Superfamily Linguloidea Menke, 1 828 
Family Pseudolingulidae Holmer, 1991 

pseudolingulid gen. et sp. indet. 1 
Fig. 4a-f 

Figured material 

AM F328314 (Fig. 4a-c): ventral valve; AM 
F128315 (Fig. 4d): dorsal valve; AM F128316 (Fig. 
4e, f): dorsal valve, sample MU 36. All from sample 
MU 35 unless otherwise mentioned (Table 1). 

Discussion 

The ventral valve pseudointerarea has a well- 
developed pedicle notch and small, subtriangular 
propareas (Fig. 4b, c). The posterior margin of the 
dorsal valve is thickened and has an undivided, 
anacline pseudointerarea (Fig. 4d). The larval shell is 
smooth and the post larval shell ornament consists of 
fine concentric filae (five per 10 |j.m) (Fig. 4a, e, f). 

'Lingula' lewisii Sowerby, 1839, from the 
lower Ludlow Aymestry Limestone of Wales, was 
questionably referred to the pseudolingulids by Holmer 
(1991) based on similarities in vascular impressions 
and muscle scars with PseudoUngula quadrata (von 
Eichwald, 1829). 'Lingula' lewisii differs by being 
more rectangular with sharper cardinal angles and 
is larger (average length 11.5 mm) (Chems 1979; 
Bassett 1986). IWadiglossa perlonga (Barrande, 
1879) from the Ludlow Kopanina Formation of 
the Czech Republic, is distinguished by its acutely 
pointed beak and post-larval shell ornament of low, 
poorly developed concentric growth lines (Mergl 
2001). 

Family Obolidae King, 1846 

Subfamily Obolinae King, 1 846 

Kosagittella Mergl, 2001 

Type species 

Kosagittella clara Mergl, 2001. 

Kosagittella'? sp. 
Fig. 4m-o 

Figured material 

AM F 128321 (Fig. 4m-o): ventral valve, sample 
MU 32 (Table 1). 

Discussion 

The ventral valve has a thickened posterior wall 
and a weakly apsacline to orthocline pseudointerarea, 
medially divided by a parallel sided pedicle groove 



that continues forward of the pseudointerarea a short 
distance as a shallow groove (Fig. 4n). The subcircular 
larval shell is smooth and located marginally. The 
post-larval shell ornament consists of widely spaced 
concentric lamellae that are best developed on the 
lateral slopes (Fig. 4m). These characteristics recall 
Kosagittella, and in particular, Kosagittella pinguis 
Mergl, 200 1 fi-om the Lochkovian Lochkov Formation 
of the Czech Republic. However, the ventral valve 
pseudointerarea of the Murruin Creek specimens 
differ fi"om Kosagittella in lacking laterally inclined 
propareas (Fig. 4o). 

Family Zhantellidae Koneva, 1986 
Rowellella Wright, 1963 

Type species 

Rowellella minuta Wright, 1963. 

Rowellellal sp. 
Fig. 4p-r 

Figured material 

AM F128322 (Fig. 4p): dorsal? valve; AM 
F 128323 (Fig. 4q, r): ventral? valve. Both firom 
sample MU 36 (Table 1). 

Discussion 

Although incomplete, these specimens appear 
to be elongately subrectangular to subtriangular 
in outline (Fig. 4p). The post-larval shell ornament 
consists of distinct concentric lamellae (six to seven 
per 100 |im) separated by flat interspaces bearing filae 
that are initially discontinues laterally, but become 
concentric during later growth stages (Fig. 4q, r). The 
post-larval shell microomament of Rowellella cf. R. 
lamellosa Popov, 1976 (in Nazarov and Popov 1976) 
fi^om Middle Ordovician strata in Sweden (Holmer 
1989) consists of similar sets of discontinuous 
filae, but these are developed over the entire shell. 
Rowellella distincta Bednarczyk and Biemat, 1978 
from the lower Arenig of the Holy Cross Mountains 
in Poland (Bednarczyk and Biemat 1978) and the 
Arenig Klabava Formation of the Czech Republic 
(Mergl 1995, 2002), also has a similar post-larval 
shell microomament, but has more prominent and 
widely spaced concentric lamellae. The post-larval 
shell microomament of Rowellella sp. firom the Early 
Ordovician Bjorkasholmen Limestone of Sweden 
and Norway (Popov and Holmer 1994) also consists 
of discontinuous sets of concentric filae, but these are 
only developed anteriorly. 

The dorsal? valve interior of the Murruin 
Creek specimens has an elongate muscle field that 



Proc. Linn. Soc. N.S.W., 127, 2006 



207 



SILURIAN BRACHIOPODS AND CONODONTS 




Figure 4. a-f. Pseudolingulid gen. et sp. indet. 1 all from sample MU 35 unless otherwise mentioned, a- 
c. Ventral valve AM F328314; a, exterior; b, interior; c, detail of pseudointerarea. d. Dorsal valve AM 
F128315, interior, e, f. Dorsal valve AM F128316, sample MU 36; e, exterior; f, detail of larval shell, g, h. 
Paterula sp. both from sample MU 36 g. Ventral valve AM F128317; interior, h. Dorsal valve AM F128318; 
exterior, i-k. Linguloid gen. et sp. indet. 2. Dorsal valve AM F128319, sample MU 35; i, exterior; j, interior; 
k, detail of pseudointerarea. 1. Linguloid gen. et sp. indet. 1. Fragment of post-larval shell AM F128320, 
sample MU 36; exterior, m-o. Kosagittella? sp. Ventral valve AM F128321, sample MU 32; m, exterior; 
n, interior; o, anterior view. p-r. Rowellellal sp. both from sample MU 36 p. Dorsal? valve AM F128322; 
interior (scale bar equals 1000 |j^m). q, r. Ventral? valve AM F128323; q, exterior; r, detail of post-larval 
shell microornament (scale bar equals 10 jam). Unless otherwise mentioned all scale bars equal 100 \x\a. 



208 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



expands slightly in width anteriorly, and is divided 
by a low, broad median ridge (Fig. 4p). This is 
similar to the dorsal valve interior of RoweUellal 
parvicapera Valentine, Brock and MoUoy, 2003 from 
the Llandovery-Wenlock Boree Creek Formation 
near Orange in central-western New South Wales. 
The Murruin Creek specimens, however, lack the 
microomment of irregularly arranged wrinkles 
possessed by Rl parvicapera on the interspaces 
between the concentric ridges on the post-larval shell 
(Valentine et al. 2003). 

Family Patemlidae Cooper, 1956 
Paterula Barrande, 1 879 

Type species 

Paterula bohemica Banande, 1879. 

Paterula sp. 
Fig. 4g, h 



similar to the post-larval shell ornament of Lingulops 
austrinus Valentine, Brock and Molloy, 2003 from 
the Llandovery-Wenlock Boree Creek Formation 
near Orange in central-western New South Wales and 
Lingulops barrandei Mergl, 1999b from the Ludlow 
Kopanina Formation of the Czech Republic (Mergl 
1999b, 2001). In comparison, the concentric ridges 
of the Boree Creek material are spaced at intervals of 
30 i^m and the concentric filae of the Czech material 
are confined to the concentric ridges. No evidence of 
a muscle supporting platform or limbus, diagnostic 
features of Lingulops Hall, 1872 (Holmer and Popov 
2000), were observed. 

linguloid gen. et sp. indet. 2 
Fig. 4i-k 

Figured material 

AM F 1283 19 (Fig. 4i-k): dorsal valve, sample 
MU 35 (Table 1). 



Figured material 

AM F128317 (Fig. 4g): ventral valve; AM 
F128318 (Fig. 4h): dorsal valve. Both from sample 
MU 36 (Table 1). 

Discussion 

The suboval outline, poorly impressed muscle 
scars, small pedicle notch (Fig. 4g) and dorsibiconvex 
profile of the Murruin Creek specimens are similar to 
P. argus from the Llandovery Zelkovice and Wenlock 
Motol formations of the Czech Republic (Mergl 
1 999a). The Murruin Creek specimens differ in having 
a wider limbus that creates a distinctly flattened rim 
externally, particularly along the posterior margin 
of the dorsal valve (Fig. 4h). Internally, the venfral 
valve differs by possessing a prominent, raised, 
subperipheral rim along the posterior margin (Fig. 
4g). 

linguloid gen. et sp. indet. 1 
Fig. 41 

Figured material 

AM F128320 (Fig. 41): post-larval shell frag- 
ment, sample MU 36 (Table 1). 

Discussion 

Known only from only a few post-larval shell 
fragments, these specimens have an ornament of 
low, broadly rounded concentric ridges spaced 
at regular intervals of 250 ^m. The ridges, and 
concave interspaces, bear closely spaced, rounded 
concentric filae (six per 100 (im) (Fig. 41). This is 



Discussion 

The well-developed limbus, ?elongate outline 
and lack of post-larval shell pitting (Fig. 4i, j) 
suggest affinities with the Elliptoglossinae. Unlike 
both Elliptoglossa Cooper, 1956 and Lingulops, the 
Murruin Creek material has a well-developed, broadly 
depressed and anacline dorsal valve pseudointerarea 
(Fig. 4k) and can be further differentiated from 
Lingulops by lacking a muscle supporting platform 
(Fig. 4j). 

Superfamily Discinoidea Gray, 1 840 
Family Discinidae Gray, 1 840 
Orbiculoidea d'Orhigay, 1847 

Type species 

Orbicula forbesii Davidson, 1848. 

Orbiculoidea sp. 
Fig. 5a-g 

Figured material 

AM F 128324 (Fig. 5a): ventral valve fragment, 
sample MU 34; AM F128325 (Fig. 5b, c): dorsal 
valve, sample MU 35; AM F 128326 (Fig. 5d, e): 
dorsal valve, sample MU 32; AM F128327 (Fig. 5f, 
g): dorsal valve, sample MU 31 (Table 1). 

Discussion 

Juvenile specimens are subrounded with a 
straight to weakly convex posterior margin and 
evenly convex lateral and anterior margins (Fig. 5b). 
In lateral profile they are weakly convex (Fig. 5c). 



Proc. Linn. Soc. N.S.W., 127, 2006 



209 



SILURIAN BRACHIOPODS AND CONODONTS 




Figure 5. a-g. Orbiculoidea sp. all from sample MU 35 unless otherwise mentioned, a. Fragment of ventral 
valve posterior slope AM F128324, sample MU 34; exterior; b, c. Dorsal valve AM F128325; b, exterior; 
c, lateral view, d, e. Dorsal valve AM F128326; d, exterior; e, lateral view; f, g. Dorsal valve AM F128327, 
sample MU 32; f, exterior; g, lateral view. h-n. Schizotreta sp. all from sample MU 36. h, i. Dorsal valve AM 
F128328; h, exterior; i, detail of larval shell, j. Ventral valve AM F128329; interior, k-n Ventral valve AM 
F128330; k, exterior; 1, detail of larval shell; m, detail of pedicle track and foramen; n, interior, o, p. Disci- 
nid gen. et sp. indet. 1. o. Dorsal valve AM F128331, sample MU 36; interior, p. Dorsal valve AM F128332, 
sample MU 35; exterior. q-yt.Artiotreta longisepta Valentine, Brock and Molloy, 2003. q-t. Dorsal valve AM 
F128333, sample MU 32 ; q, interior; r, exterior; and s, lateral views; t, interior in lateral view. u-w. Conjoined 
valves AM F128334, sample MU 36; u, plan; v, anterior; and w, posterior views. All scale bars equal 100 jam. 



210 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



Mature dorsal valves are more elongate, with longer, 
more gently curved, lateral margins (Fig. 5d, f) and 
are weakly convex to low subconical in lateral profile 
(Fig. 5e, g). The ventral valves have a long, narrow, 
parallel-sided pedicle track covered for most of its 
length by a concave listrum (Fig. 5a). The post-larval 
shell ornament consists of well-developed concentric 
ridges arising through insertion on the lateral slopes 
(Fig. 5b, d, f). 

Numerous Silurian discinids have been assigned 
to Orbiculoidea (eg. Biemat 1 984; Bassett 1986; Mergl 
1996, 2001). These are generally distinguishable from 
the Murruin Creek specimens by their more circular 
dorsal valves, greater convexity, and centrally located 
apices. Orbiculoidea sp. C from the Pridoli Pozary 
and Lochkovian Lochkov formations of the Czech 
Republic (Mergl, 2001) is similar to the Murruin 
Creek taxon. Both species have low, subconical 
dorsal valves with a subcentral apex and an ornament 
of well-developed concentric ridges arising through 
insertion on the lateral slopes. Orbiculoidea sp. C 
differs in having a subcircular dorsal valve outline 
and by being wider and less elongate (Mergl 2001). 

Schizotreta Kutorga, 1 848 

Type species 

Orbicula elliptica Kutorga, 1846. 

Schizotreta sp. 
Fig. 5h-n 

Figured material 

AM F128328 (Fig. 5h, i): dorsal valve; AM 
F128329 (Fig. 5j): ventral valve; AM F128330 (Fig. 
5k-n): ventral valve. All fi-om sample MU 36 (Table 
1)- 

Discussion 

Both valves of this species fi"om Murruin Creek 
are subcircular with a weakly flattened posterior 
margin and have a post-larval shell ornament of low, 
continuous, concentric lamellae (two to four per 100 
fim) that become more prominent toward the valve 
margins (Fig. 5h, k). The large larval shell, located 
submarginally in the ventral valve (averaging 438 
fxm long; 500 jam wide) and marginally in the dorsal 
valve (averaging 354 ^m long; 399 ^im wide), bear 
fine growth filae on their anterior and anterolateral 
slopes (Fig. 5h, k, 1). The ventral valve has a short, 
elliptical pedicle track covered for most of its length 
by a concave listrum. The foramen, preserved in only 
one specimen, is quadrate and has a raised rim (Fig. 
5k, m). The pedicle track continues internally as a 



posteriorly directed pedicle tube that is flattened along 
the valve floor and ends just prior to the posterior 
margin (Fig. 5n). 

Schizotreta elliptica from the Early Ordovician 
of the Leningrad district in Russia, differs from the 
Murruin Creek species by its elongately oval dorsal 
valve, submarginally located larval shell, shorter, 
more strongly elliptical pedicle track and elliptical 
foramen. Valentine et al. (2003) described two species 
of Schizotreta fi^om the Llandovery- Wenlock Boree 
Creek Formation near Orange in central-western 
New South Wales. Schizotreta corrugaticis Valentine, 
Brock and Molloy, 2003 has a flatter dorsal valve with 
a submarginally located larval shell and an ornament 
of well-developed concentric ridges arising through 
insertion on the lateral slopes (Valentine et al. 2003). 
Schizotreta cristatus Valentine, Brock and Molloy, 
2003 is distinguished by its elongately subrectangular 
dorsal valve outline and more widely spaced continuous 
concentric lamellae. Internally, the ventral valve has 
a low, broad, crescentic-shaped ridge bounding the 
anterior margin of the muscle field (Valentine et al. 
2003). Schizotreta rarissima (Barrande, 1879) from 
the Wenlock Motol Formation of the Czech Republic, 
has a narrow, elongately oval dorsal valve (Mergl 
2001). Biemat (1984) assigned a single dorsal? valve 
fi-agment of a discinid, from the Wenlock Podlasie 
Depression of Poland, to Schizotreta which possesses 
well-developed concentric ridges arising through 
insertion on the lateral slopes. The ventral valve of 
Schizotreta sp. fi-om the early Llandovery of Wales 
(Temple 1987), is flatter than the Murruin Creek 
taxon and has a shorter, posteriorly widening, pedicle 
track. The concentric lamellae of the Welsh taxon are 
also more widely spaced (six per mm) and separated 
by concave interspaces (Temple 1987). 

discinid gen. et sp. indet. 1 
Fig. 5o, p 

Figured material 

AM F128331 (Fig. 5o): dorsal valve, 
sample MU 36; AM F128332 (Fig. 5p): dorsal valve, 
sample MU 35 (Table 1). 

Discussion 

This taxon from Murruin Creek has a large 
(averaging 475 fim long; 500 |im wide), smooth, 
marginally located dorsal valve larval shell and a 
post-larval shell ornament of weakly developed, 
continuous concentric lamellae (Fig. 5p). However, 
unlike other discinids, the Murruin Creek taxon has a 
transversely elliptical dorsal valve outline (Fig. 5o, 

P)- 



Proc. Liim. Soc. N.S.W., 127, 2006 



211 



SILURIAN BRACHIOPODS AND CONODONTS 



Order AcrotretidaKuhn, 1949 

Superfamily Acrotretoidea Schuchert, 1893 

Family Scaphelasmatidae Rowell, 1965 

Artiotreta Ireland, 1961 

Type species 

Artiotreta parva Ireland, 1961. 

Artiotreta longisepta Valentine, Brock and MoUoy, 

2003 
Figs 5q-w, 6a-d 

Synonymy 

2003 Artiotreta longisepta sp. nov. Valentine, 
Brock and Molloy, p. 314; pi. 2, figs 9-18. 

Description 

See Valentine et al. (2003:3 14). 

Figured material 

AM F128333 (Fig. 5q-t): dorsal valve, sample 
MU 32; AM F 128334 (Figs 5u-w, 6a): conjoined 
valves; AM F 128335 (Fig. 6b-d): dorsal valve. All 
firom sample MU 36 unless otherwise mentioned 
(Table 1). 

Discussion 

The Murruin Creek material is conspecific with 
A. longisepta from the Llandovery- Wenlock Boree 



Creek Formation near Orange in central-western New 
South Wales (Valentine et al. 2003). Some specimens 
from Boree Creek have a median septum with a 
slightly thickened posterior margin (see Valentine et 
al. 2003 :pl. 2, fig. 16) and concentric lamellae that 
tend to be weaker and more irregularly developed 
(compare Figs 5q, r, 6b with Valentine et al. 2003 :pl. 
2, figs 11,13). These minor differences are considered 
insufficient to exclude conspecificity. 

Artiotreta krizi Mergl, 2001 from the Llandovery 
Zelkovice and Wenlock Motol formations of 
the Czech Republic, has a similar dorsal valve 
outline to A. longisepta, although its anterior margin 
tends to be more rounded. The median septum of ^4. 
krizi is also shorter, arising around valve midlength 
(Mergl 2001:33, pi. 28, fig. 3). Artiotreta krizi attains 
a larger maximum size than ^. longisepta (up to 1 100 
\xm wide), but most of the material illustrated by 
Mergl (2001:pl. 28: figs 1, 3, 9, 10) has comparable 
dimensions (Table 4). 

Artiotreta parva from the Wenlock Chimney 
Hill Limestone (Ireland 1961), Bainbridge Formation 
(Satterfield and Thompson 1969) and Clarita 
Formation (Chatterton and Whitehead 1987) of the 
USA, is distinguished by its rounder dorsal valve 
outline, shorter median septum arising around valve 
midlength and finer growth lamellae. Artiotreta 
longisepta is also larger (averaging 538 fxm long; 708 
^m wide; Table 4) than A. parva (averaging 400 |j.m 



Table 4. Artiotreta longisepta Valentine, Broclc and Molloy, 2003, dorsal valve 
dimensions (in |am) and ratios. 

Artiotreta longisepta Valentine, Brock and Molloy, 2003, 
dorsal valve dimensions (|-im) and ratios 

L W LI Wl LS MHS BS LP WP 

N 17 43 17 15 18 42 16 17 18 

MEAN 533.8 694.8 40.1 255.8 475.0 164.0 245.3 150.7 175.7 

SD 106.32 140.85 10.14 46.74 89.22 63.03 32.56 20.48 26.94 

MIN 375 400 18.75 150 325 25 212.5 125 150 

MAX 675 975 50 312.5 600 287.5 325 175 225 

LAV LFWI WIAV LS/L BS/L LPAVP LP/L WPAV 

N 17 15 13 15 12 17 12 14 

MEAN 83.6% 15.5% 43.4% 88.4% 48.4% 85.8% 30.0% 29.9% 

SD 0.08 0.05 0.10 0.04 0.09 0.12 0.07 0.08 

MIN 65.8% 9.1% 26.9% 81.3% 36.7% 70.0% 19.2% 18.9% 

MAX 98.2% 28.6% 62.5% 93.8% 66.7% 116.7% 38.1% 43.8% 



212 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 




Figure 6. a-d. Artiotreta longisepta Valentine, Brock and MoUoy, 2003 both from sample MU 36. a. Con- 
joined valves AM F128334; lateral view. b-d. Dorsal valve AM F128335; b, exterior; c, detail of larval 
shell (scale bar equals 10 |im); d, detail of larval shell microornament (scale bar equals 10 |^m). e-y. 
Acrotretella dizeugosa sp. nov. all from sample MU 31 unless otherwise mentioned, e, f. Ventral valve 
paratype AM F128336, sample MU 35; e, exterior; f, detail of larval shell (scale bar equals 10 \x.va). g. 
Ventral valve paratype AM F128337; interior, h-k. Ventral valve paratype AM F128338, sample MU 
35; h, exterior; i, plan; j, posterior; and k, lateral views. 1. Dorsal valve paratype AM F128339; inte- 
rior, m, n. Dorsal valve paratype AM F128340; m, interior; n, lateral view, o, p. Dorsal valve para- 
type AM F128341; o, interior; p. lateral view, q, r. Dorsal valve paratype AM F128342; q, interior; r, 
lateral view. s. Dorsal valve paratype AM F128343, sample MU 35; interior, t. Dorsal valve paratype 
AM F128344, sample MU 32; interior, u-w. Dorsal valve holotype AM F128345, sample MU 32; u, in- 
terior; V, lateral; and w, anterior views, x, y. Dorsal valve paratype AM F128346; x, exterior; y, de- 
tail of larval shell (scale bar equals 10 )_im). Unless otherwise mentioned all scale bars equal 100 (im. 



Proc. Linn. Soc. N.S.W., 127, 2006 



213 



SILURIAN BRACHIOPODS AND CONODONTS 



long; 500 f^m wide) (Ireland 1961:1138). 

von Bitter and Ludvigsen (1979) documented 
two sizes of larval shell pits in A. parva — a 
larger set (3-6 |iim in diameter) with no cross-cutting 
relationships and a smaller set (-0.3 |Lim in diameter) 
located on flat areas between the larger pits. Artiotreta 
longisepta also possesses two sizes of larval shell 
pits — a larger set (4-5 ^im in diameter) with none to 
one (occasionally two) orders of cross-cutting and a 
smaller set (0.2-1 \xm in diameter) (Fig. 6d). A smaller 
set of larval shell pits has not been documented in A. 
krizi. 

Family Torynelasmatidae Ro well, 1965 
Acwtretella Ireland, 1961 

Type species 

Acwtretella siluriana Ireland, 1961. 

Emended diagnosis 

Ventral valve conical to subpyramidal with 
distinct larval shell and broad, procline to catacline 
pseudointerarea. Pedicle tube and apical process 
absent. Dorsal valve flat to weakly convex with 
distinct, bulbous larval shell. Pseudointerarea broad, 
anacline, occasionally weakly depressed medially. 
Median septum low to high with dorsally concave 
surmoimting plate on ventral margin. Anterior 
margin of median septum with variably developed 
number of spines and folds depending upon valve 
size and species. One to two pairs of lateral processes 
developed either side of dorsal valve median septum 
in some species. 

Discussion 

Previous authors (Biemat and Harper 1999, Mergl 
2001 and Valentine et al. 2003) have defined species 
of Acwtretella based upon the presence or absence of 
lateral processes (sensu Biemat and Harper 1999:88) 
in the dorsal valve (Table 5). Despite both forms 
having a similar stratigraphic range, no consideration 
has previously been given to the possibility that the 
development of lateral processes may be part of an 
ontogenetic growth continuum. It is only in the Cobra 
Formation and the Llandovery- Wenlock Boree Creek 
Formation near Orange in central-western New 
South Wales (Valentine et al. 2003), that Acrotretella 
with and without lateral processes occur in the same 
stratigraphic horizons (Tables 5, 6). Analysis of the 
ontogeny of Acrotretella has been prevented before 
because most occurrences are restricted to a handfiil 
of specimens (Table 5). A sufficient number of 
specimens assignable to Acrotretella dizeugosa sp. 
nov. (36 ventral valves and 39 dorsal valves; Table 



2) have been recovered fi^om the Cobra Formation to 
allow the first detailed ontogenetic investigation of 
Acrotretella. 

The dorsal valve ontogeny of ^4. dizeugosa can 
be divided into four overlapping developmental 
growth stages (Fig. 7): (1) development of a simple 
dorsal valve median septum with a dorsally concave 
surmounting plate; (2) growth of spines along the 
anterior margin of the dorsal valve median septum; 

(3) insertion of the first pair of lateral processes; and 

(4) insertion of a second pair of lateral processes 
posterior of, and parallel to, the first pair. Although 
considerable overlap exists in the size range of each 
growth stage, each growth stage generally corresponds 
with an increase in dorsal valve size (Fig. 7). 

During the first dorsal valve growth stage oi A. 
dizeugosa (413-725 |^m long; 488-788 (im wide) (Fig. 
7) a simple, low median septum with surmounting 
plate is developed (Fig. 6m, n). The surmounting 
plate originates slightly posterior of valve midlength 
as two ridges separated by a dorsally concave plate 
averaging 63 |am wide. A single dorsal valve of A. 
dizeugosa from sample MU 31 (628.6 m above the 
base of the section) (275 fim long; 288 ^m wide) has 
yet to develop a median septum with surmounting 
plate (Fig. 61). The surmounting plate coalesces into 
a single blade at 56%, and continues to 86%, valve 
length. The median septum reaches an average height 
of 75 |am at 81% valve length. The anterior margin of 
the median septum is steep, straight to weakly curved 
and smooth (Fig. 6n). 

Up to three spines, termed 'septal spines' by 
Popov (in Nazarov and Popov 1 980:75) are developed 
along the anterior margin of the median septum during 
the second dorsal valve growth stage (563-763 |_im 
long; 688-838 )j.m wide) (Fig. 7). The median septum 
of this growth stage is higher than in the proceeding 
stage, averaging 144 (j.m high at 75%, and extends 
to 87%, valve length. The anterior margin of the 
median septum is longer and more strongly curved 
than in the first growth stage (Fig. 6q, r). Septal spines 
also occur in numerous Ordovician acrotretoids (eg. 
Numericoma Popov, 1980 in Nazarov and Popov 
1980). Popov (in Nazarov and Popov 1980) and 
Holmer (1989) have linked the development of septal 
spines in such genera to the ontogenetic development 
of the median septum — from a simple, subtriangular 
blade in juveniles, to a complexly spinose structure in 
mature individuals. 

The first pair of lateral processes appear during 
the third dorsal valve growth stage (625-950 |j,m long; 
750-1325 (am wide) (Fig. 7). The lateral processes 
originate around valve midlength and are initially 
developed as low, short, anteriorly divergent rods 



214 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 






■= 2 

•I? 



s 

• MM 

>^ 

S 

cs 

a 



I ^ - 

^ .^ o 

■*« oc « 

^ Ji .a 

^ II ^ 

© — ■ 

w *• *C 

M « « 

>- - S 

-o ■= I 

— II « 

«a > 3 

a vSS « 

a» « . 

S ^ o 

^ > e 

TS — ' ^ 

a « > 

R V g 

•E > o 

« II -K 

a >. s 

s ?: i 

-§.2 c 



W « « 

aT X! _g 

OB .a ■" 

« ^ g 

o « «i 
^ l§ 

« 2 g 

XI V s 

H ,S ■« 



o 

c 
u 



£ 

to 

n 
o 



c 



OB 

> 
< 



.2 
.-^ ■*-» ^ 

2 S 



Oh -2 

cd > 



J3 

■♦-< 

o 



<L> 
OB 



CO 

u 

o 



to 

'o 

u 

C/3 



T3 

c 



as 

ON 



0) 






P; ON 



>/-i 



- ? 



CO 



n 
<u 

E 

ISO 

CO 



> 

c 

CO 



"o 
a. 

« 

c 

to 
(J 



CO 



ON 



ON 

a 

CO _ 

^ S 
> O 



r- o 



CO 

B 

=2 CQ 

cu -3 
■is r u 

O ii) CO 

X .5 -o 

nS C 

bo 






ON 

r-- 

ON 



to 

c 
o 



c 

CO 

Q 



00 <^ 

00 o^ 







u 


m 


c 


o 




o 


c 




(U 


w 






> 


13 

-4-t 




<u 




t~~ 


o 


(N 


o 




ON 


— 


II 


II 




& 



CJ 

o 
U 



CO 

■<— » 
CO 
T3 

3 



00 

t^ o 

00 ,-~ 



- ^ - > - 



00 
to 
< 



c 

CO 



C 

o 

to 

u 

s 

u 
So 

T3 






< 



tO u 



2 S 

^ -4— • 

c o 



o 



O 

CQ 



ON 






o 

u 

u 

u 
o 
CQ 



.2 

to 
3 
< 



CQ 



(N 



O 

o 

w 

"ob 



vo 00 

II II 






T3 
C 
CO 

m 
D 

00 

Q 



to 

3 



CO 



o 

<u 

o 

c 
u 

o 
CQ 

c 

CO 



CO 



o 

b 



(N 



n 
2 
'> 
o 



o 



3 

o, 

a: 

CJ 

(U 
N 

o 

c 
o 

to 



o 






<a 



Is 
o 



w 
o 

z 

o 

Q 

DQ 
P 

s 

H 

z 

o 
u 



w 
-co 



Proc. Linn. Soc. N.S.W., 127, 2006 



215 



SILURIAN BRACHIOPODS AND CONODONTS 



X 




iu 

Q 


o 

1 


5* c 


2 

TO 


s? 


■i"" ^ 




a 


o ^2 
o a 


CO 


V 




> 


n 


O 




3 


«5 


3 


5 


*n 


O 


o 

■-1 


o 


O 


D. 


3 


r 


r 


^ 


3" 


§■ 


5 




n 


3 


^^ 


w 




o 


o 


n 




s 


N 


n 


n 






o 


> 


> 


3* 


c 


c 


?o 


s. 


M 
^ 


n 
T3 


3 


— 


C 


5' 


5' 


o' 




^ 


^ 


-TS 


-n 


"-I 


"-I 








Q.' 


5^ 


ol 


o_ 


o_ 


o_ 














O 


o\ 


o 


a\ 


O 

^ - 

II II 

ON '^ 


1 


^ - 


^ - 


^ - 


II II 


II II 


II II 


^ 22 




2o 


<1 =» 


VO ®° 


o o 


1^ 


2 
« 




C 03 


^ 


P^ :^ 


, ^ 




^ « 


to 


t-^ 


o. 2:: 


o 


'vO 


S- 


o 


^o 


p 




-~J 


^ ' 







3 
CI. 



TO 



ON 



a 

S 

S' 
s 
a 



2 

o 

ft- 

o_ 

O 
& 

o' 

3 

n 

N 
0) 

o 

3- 

?0 
<Tl 
"O 

C 

a; 
o' 



ft 

3_ 

o" 
o 



— 00 
O 4^ 

O 



II II 

-I 
ji, 00 



ft 
OH, 

o 

o 



p 



p 



3_ 

o" 
o 

5^ 



00 



ON 



ON 
ON 
O 



II II 



03 

3 

CL 



OS 



< I 

C= 3 

TO 



3 
p. 



= 3- a- 



3^ 

o 



K) 



^ _ 



^- 00 
O L/1 
--1 O 



^ - 






n 

o 
o 



3 CI 

O" fD 

H. ^ 

w NO 



5' 

3 

3 

O 3 



o 

n 

n 
a 

pr 

Tl 
o 



> 

c 

5' 



o 

3 
I 

w 

o 

1-1 

<T> 
3 
O 
-I 

n 






ON 






NO ~^ 



II II 
NO ~J 



D 

3 
I 

O 
3 
<T> 



NO 
NO 



DD 



03 
o 
"1 
n 
n 

n 

a 
n 

?r 

■n 
o 






to 

ON 



4^ 



00 g 



t - 






03 



< 

fT 

3 



U> <T 



o 

c 

03 

3^ 

n n 

2 o> 

03 C. 

g. 3 

£. o 
2 =^ 

-1 CO 
O Ci. 



5' 



fU 



03 
3 
D- 
O 
< 



ON 



03 

3 

n> 

3 

03 



o 
o 



t - 



NO 

to 



^-^ 




X 


^ 




o 


TS 


CO 


3 


C 


Eo' 


O 


O" 


o 


"^ 




3- 


^,^1^^ 


CL 


O 


1—* 


03 


^ 


NO 

00 



TO 
TO 

ft- 

czi 

V 

03 



03 
a 

o 

r 
3' 

0) 
tn 
<-f 
O 
3 
n 

n 

3 



> 

3- 
OQ 



to 






to ^ 



o 


TO 


>! 


>! 


o 


o 


<»». 


«»*. 


>! 


^ 


TO 


TO 


"* 


»>. 


TO 


TO 


»-~ 


<-» 


^.- 




C 


a 


v> 


to 


TS 


>T3 



03 
(/] 

Tl 
O 



o" 

3 
?^ 

03 
N 
03 



3 



n 
p 

a. 
o 

o 



to 



^ - 



— -J 

O ON 

o 



o 

O 
< 

1o 

o 

o 
o 



O 
p 



o 

3 
CO 

3 



n 
p 

p 

o. 
o 
o 



KJ 



II II 

^ ft 

O ^ 

NO "^ 



o 

3" 
n 



NO 

00 

NO 



on 

fD 

o 



r 
o 
o 
p 






n 5 

c 

> Q 

^ p. B- 





P 


xi 




a; 


i-( 




cT 


o 
o 




3 


m 




p 


Vl 


CL 

< 


p.' 

> 

•1 


!/3 


< 


p 




< 


era 

CL 

s 

o' 

3 




Q. 


Cfl 




< 


9" 








3 
o 



a- 



o 

o 

B 

5' 

c 

&. 



216 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



B 

X 

H 

o 



1400- 



1200- 



1000- 



800- 



600- 



400- 



200- 



+ 



n 



D 



O 



+ 



V 



p\ 



+ 



"++ 






o^ 



200 



400 



600 



800 



1000 



1200 



1400 



WIDTH (^m) 



Figure 7. Length versus width for ventral valves and each dorsal valve growth stage of 
Acrotretella dizeugosa sp. nov. 

+ = ventral valves (n = 12); 

O = dorsal valve growth stage 1 (n = 4); 
■^ = dorsal valve growth stage 2 (n = 4); 
V = unknown dorsal valve growth stage (n = 7)*; 

Q = dorsal valve growth stage 3 (n = 5); 

• = dorsal valve growth stage 4 (n = 2). 

*Note that the dorsal valve median septum of some specimens of A. dizeu- 
gosa without lateral processes is not preserved. Such specimens, equiva- 
lent to dorsal valve growth stages 1 or 2, are presented here as a separate, com- 
bined group. See text for discussion on dorsal valve growth stages of A. dizeugosa. 



Proc. Linn. Soc. N.S.W., 127, 2006 



217 



SILURIAN BRACHIOPODS AND CONODONTS 



with rounded anterior margins (Fig. 6s). The median 
septum of this growth stage is higher again than in 
the previous growth stages (averaging 188 jam high) 
and has a long, curved anterior margin with up to four 
septal spines. A positive relationship between dorsal 
valve size and the development of septal spines can 
also be demonstrated in Acrotretella spinosa Mergl, 
2001 from the Pfidoli Formation of the Czech 
Republic (see Mergl 2001 :pl. 26, figs 4-6). 

The final dorsal valve growth stage of A . dizeugosa 
(1000-1275 nm long; 1025-1500 ^m wide) (Fig. 
7) is characterised by the development of a second 
pair of lateral processes that are inserted posterior 
of, and parallel to, the first pair at approximately 
one-third valve length (Fig. 6t, u). Two specimens 
assignable to this grovv1:h stage were recovered from 
Murruin Creek. The lateral processes of the smaller 
of these specimens (1000 i^m long; 1025 |nm wide) 
are developed as low, slightly elongate rods, with 
the posterior pair being slightly shorter than the 
anterior pair. The anterior ends of the first pair are 
weakly twisted and flattened. The median septum 
of this specimen was not preserved (Fig. 8t). The 
posterior pair of lateral processes in the larger of these 
specimens (1275 jim long; 1500 |Lim wide) are higher 
and longer than the anterior pair, and both pairs end 
in variably developed, stubby projections (Fig. 6u- 
w). The median septum of this specimen, although 
damaged, bears the remains of five septal spines along 
its anterior margin (Fig. 6v, w). Concurrent with this 

Table 6. Stratigraphic distribution and abundance of ventral valves 
and each dorsal valve growth stage oi Acrotretella dizeugosa sp. nov. 
recovered from productive samples along the MU section through the 
Cobra Formation in Murruin Creek. *Note that the dorsal valve medi- 
an septum of some specimens of A. dizeugosa without lateral processes 
is not preserved. Such specimens, equivalent to dorsal valve growth 
stages 1 or 2, are presented here as a separate, combined group. See 
text for discussion on dorsal valve growth stages of A. dizeugosa. 



Metres above base of MU section 




en 


0\ 




1 


Sample Numbers 
Acrotretella dizeugosa 


31 


32 


35 


36 


ventral valves 


25 


6 


3 


2 


36 


dorsal valve growth stage 1 


5 




2 


2 


9 


dorsal valve growth stage 2 


7 








7 


dorsal valve grovi1;h stage 3 




3 


2 




5 


dorsal valve grovi1;h stage 4 




2 






2 


unknown dorsal valve growth stage* 


11 


2 


1 


2 


16 



final dorsal valve growth stage is the initiation of 
folding in the median septum, with up to two folds 
being developed. Biemat (1973:43) demonstrated 
a positive relationship between valve size and the 
degree of folding in the dorsal valve median septum 
of Myotreta Gorjansky, 1969. Although only one 
specimen of A. dizeugosa was recovered with a folded 
median septum, this feature does occur in the largest 
specimen suggesting it is also related to valve size. 

Apart fi-om Acrotretella goldapiensis Biemat and 
Harper, 1999 from the Llanvim Baltic syneclise of 
northwest Poland, no ventral valves have previously 
been assigned to any species of Acrotretella with 
lateral processes (Table 5) (Mergl 2001; Valentine et 
al. 2003). Ventral valves assignable to Acrotretella 
from the Cobra Formation co-occur with, and overlap 
the size range of, each dorsal valve growth stage of 
A. dizeugosa (Fig. 7; Table 6). A similar trend is also 
observable in A. goodridgei Valentine, Brock and 
MoUoy, 2003 fi-om the Llandovery- Wenlock Boree 
Creek Formation near Orange in central-western New 
South Wales (see species discussion iox A. dizeugosa 
below) (Fig. 8). 

Therefore, a positive relationship can be 

demonstrated to exist between dorsal valve size 

and the development of lateral processes and septal 

spines along the anterior margin of the dorsal valve 

median septum in A. dizeugosa (Fig. 7). Additional 

material for most Acrotretella species is required to 

confirm if lateral processes, and/or septal spines, were 

developed in all species (Table 

5). Until such time, care must 

be exercised when relying upon 

the presence or absence of these 

features to define species. To 

this end, an ontogenetic growth 

continuum for each population 

of Acrotretella should be 

established prior to the erection 

of new species and the level of 

intraspecific variation present 

determined. 

Acrotretella dizeugosa sp. nov. 
Figs 6e-y, 9a-b 

Etymology 

Gr., di, two, double; Gr., 
zeugos, team, pair; in reference 
to the development of two pairs 
of lateral septa in the dorsal 
valve of mature individuals. 



218 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



o 



1400- 



1200- 



1000- 



800- 



600- 



400- 



200- 



D + 



D 




+ 



200 400 600 800 

WIDTH (^m) 



1000 



1200 



1400 



Figure 8. Length versus width for ventral valves and each dorsal valve growth stage 
of Acrotretellagoodridgei Valentine, Brock and MoUoy, 2003 from the BM section of 
Valentine et al. (2003) and the B section of Bischoff (1986) through the Llandovery- 
Wenlock Boree Creek Formation near Orange in central-western New South Wales. 

+ = ventral valves (n = 17); 

O = dorsal valve growth stage 1 (n = 10); 

■^ = dorsal valve growth stage 2 (n = 12); 

V = unknown dorsal valve growth stage (n = 20)*; 

G = dorsal valve growth stage 3 (n = 6); 

• = dorsal valve growth stage 4 (n = 2). 

*Note that the dorsal valve median septum of some specimens of A. goodridgei 
without lateral processes is not preserved. Such specimens, equivalent to dor- 
sal valve growth stages 1, 2 or 3, are presented here as a separate, combined 
group. See text for discussion on dorsal valve growth stages of A. goodridgei. 



Proc. Linn. Soc. N.S.W., 127, 2006 



219 



SILURIAN BRACHIOPODS AND CONODONTS 




Figure 9. a-b Acrotretella dizeugosa sp. nov. Dorsal valve paratype AM F128347, sample MU 31; a, ex- 
terior Figure; b, detail of larval shell (scale bar equals 10 ^m). c-n. Opsiconidion ephemerus (Mergl, 
1982) all from sample MU 32 unless otherwise mentioned, c, d. Dorsal valve AM F128348, sample MU 
35; c, exterior; d, detail of larval shell, e, f. Dorsal valve AM F128349; e, interior; f, lateral view, g, h. 
Dorsal valve AM F128350; g, interior; h, lateral view, i, j. Dorsal valve AM F128351; i, interior; j, detail 
of pseudointerarea. k-n. Ventral valve AM F128352; k, exterior; 1, anterior; m, posterior; and n, lateral 
views, o-q. Opsiconidion sp. o. Dorsal valve AM F128353, sample MU 31; interior, p, q. Dorsal valve AM 
F128354, sample MU 36; p, exterior; q, detail of larval shell, r-t. Siphonotretid gen. et sp. indet. 1. r, s. 
Dorsal valve AM F128355, sample MU 36; r, exterior; s, detail of spines (scale bar equals 10 jam), t. Dor- 
sal valve AM F128356, sample MU 34; interior. Unless otherwise mentioned all scale bars equal 100 [am. 



220 



Proc. Linn. Soc. N.S.W, 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



Table 7. Acrotretella dizeugosa sp. nov., ventral and dorsal valve dimensions (in [ini) and ratios. 





Acrotretella dizeugosa 


sp. nov., 


ventral valve dimensions i 


(|im) and ratios 








L 


W 


H 


FP 


Fa 


Fw 


M 


HP 


LP 


WP 




N 


12 


15 


15 


12 


17 


18 


15 


17 


19 


20 




MEAN 


505.2 


655.0 


344.2 


85.4 


36.8 


33.0 


117.5 


84.6 


164.3 


190.6 




SD 


182.35 


240.99 


171.45 


61.67 


8.53 


9.05 


62.11 


26.34 


20.55 


31.64 




MIN 


350 


375 


125 


37.5 


25 


25 


50 


50 


125 


150 




MAX 


912.5 


1150 


675 


275 


56.25 


50 


275 


150 


225 


262.5 






L/W 


H/L 


M/L 


LP/WP 


LP/L 


WP/W 


HP/H 


HP/LP 








N 


12 


9 


12 


18 


12 


13 


13 


14 




MEAN 


85.0% 


55.5% 


22.8% 


88.8% 


36.0% 


32.3% 


31.2% 


47.5% 








SD 


0.09 


0.13 


0.10 


0.07 


0.12 


0.12 


0.15 


0.11 








MIN 


71.4% 


35.7% 


9.6% 


75.0% 


17.8% 


15.2% 


10.3% 


28.6% 








MAX 


100.0% 


75.0% 


44.0% 


108.3% 


50.0% 


48.5% 


43.8% 


66.7% 






























Acrotretella dizeugosa sp. i 


nov., dorsal valve dimensions (fim) and ratios 






L 


W 


LI 


WI 


OSP 


LSP 


WSP 


LS 


MHS 


LP 


WP 


N 


22 


31 


27 


19 


29 


19 


28 


24 


13 


29 


31 


MEAN 


678.4 


763.3 


50.7 


388.2 


234.5 


375 


81 


592.2 


110.6 


163.4 


161.3 


SD 


218.57 


244.58 


28.61 


152.04 


28.08 


71.32 


27.82 


172.42 


54.21 


16.68 


25.28 


MIN 


275 


287.5 


12.5 


100 


187.5 


250 


50 


350 


50 


125 


112.5 


MAX 


1275 


1500 


125 


725 


275 


575 


150 


1075 


187.5 


200 


200 




LAV 


LI/WI 


WlAV 


LS/L 


LSP/L 


OSP/L 


LSP/LS 


LP/L 


WPAV 


LP/WP 




N 


22 


19 


17 


19 


14 


20 


16 


18 


24 


29 




MEAN 


84.1% 


13.7% 


47.7% 


82.8% 


55.6% 


37.0% 


69.7% 


26.7% 


22.6% 


104.6% 




SD 


0.09 


0.06 


0.07 


0.07 


0.09 


0.11 


0.11 


0.06 


0.06 


0.21 




MIN 


69.8% 


6.3% 


34.6% 


70.0% 


42.3% 


17.7% 


53.5% 


15.8% 


11.3% 


76.9% 




MAX 


97.9% 


30.8% 


54.8% 


90.2% 


68.4% 


52.6% 


89.3% 


68.2% 


60.9% 


133.3% 





Type material 

Holotype: AMP 128345 (Fig. 6u-w): dorsal valve, 
sample MU 32. Figured paratypes: AM F128336 (Fig. 
86, f): ventral valve, sample MU 35; AM F128337 
(Fig. 6g): ventral valve; AM F128338 (Fig. 6h-k): 
ventral valve, sample MU 35; AM F128339 (Fig. 
61): dorsal valve; AM F128340 (Fig. 6m, n): dorsal 
valve; AM F 128341 (Fig. 6o, p): dorsal valve; AM 
F128342 (Fig. 6q, r): dorsal valve; AM F 128343 (Fig. 
6s): dorsal valve, sample MU 35; AM F128344 (Fig. 
6t): dorsal valve, sample MU 32; AM F128346 (Fig. 
6x, y): dorsal valve; AM F128347 (Fig. 9a, b): dorsal 
valve. All from sample MU 31 unless otherwise 
mentioned (Table 1). 

Type horizon and locality 

Sample MU 32 (Fig. 3), Ludlow i^siluricus 
Zone), upper part of the Cobra Formation cropping 
out in Murruin Creek (Fig. 2), Taralga Group, 
southeastern New South Wales, Australia (Fig. 1). 



Diagnosis 

A species of Acrotretella with numerous, closely 
spaced growth lamellae on the ventral valve (eight 
per 100 |J,m), but more widely spaced on the dorsal 
valve (three to five per 100 \xm). Dorsal valve larval 
shell with rounded, variably developed, medially 
depressed ridge bounding anterior and anterolateral 
margins. Anterior margin of dorsal valve median 
septum bearing up to two folds and five septal spines. 
Two pairs of lateral processes inserted centrally, 
either side of dorsal valve median septum, in mature 
individuals. 

Description 

Ventral valve subpyramidal, subtending apical 
angle of 85° in anterior view. In lateral profile 
posterior slope straight to weakly convex; anterior 
slope long, flat to weakly convex. Valve height (Table 
7) averaging 56% valve length and 46% valve width. 



Proc. Linn. Soc. N.S.W., 127, 2006 



221 



SILURIAN BRACHIOPODS AND CONODONTS 



Beak directed ventrally. Pseudointerarea catacline 
to weakly apsacline, subtending apical angle of 
approximately 80°, separated from remainder of valve 
by gentle flexure. Intertrough vaguely defined in some 
specimens, subtending apical angle of approximately 
20°. Larval shell subcircular, averaging 164 \xm long, 
191 |am wide, 85 ^m high. Foramen confined to larval 
shell, subcircular, averaging 37 |am long and 33 |_im 
wide. Narrow, subparallel sulcus extending anteriorly 
from foramen, dividing larval shell into two lateral 
swellings, occasionally continuing into juvenile 
portion of post-larval shell. Larval shell bearing 
shallow, circular, flat-bottomed pits averaging 5 |j,m in 
diameter. Post-larval shell ornament of well-defined, 
closely spaced, continuous concentric lamellae 
(eight per 100 |im) with rounded crests. Concentric 
lamellae on juvenile portion of post-larval shell less 
well-defined. Lamellae becoming disordered and 
less distinct on pseudointerarea, especially across 
intertrough. 

Ventral valve interior of some specimens with 
weakly impressed, elongate adductor scars on 
posterior slope. No other muscle scars or mantle canal 
patterns observed. Pedicle tube and apical process 
absent. 

Dorsal valve outline subquadrate to transversely 
elongate, with straight posterior margin, weakly to 
strongly curved lateral margins, and straight to weakly 
curved anterior margin. Maximum width occurring 
slightly posterior of valve midlength. Anterior 
slope of juveniles long and flat in lateral profile, 
becoming depressed medially and raised anteriorly in 
mature specimens. In anterior view, lateral slopes of 
juveniles short and flat, developing raised margins in 
mature specimens. Larval shell bulbous, subcircular, 
averaging 163 |am long and 161 fim wide, with 
flattened lateral and posterior margins and separated 
from post-larval shell by raised rim. Larval shell with 
rounded, variably developed, medially depressed 
ridge bounding anterior and anterolateral margins. 
Pitted larval shell microomament similar to that of 
ventral valve. Post-larval shell ornament similar to 
that of ventral valve, but concentric lamellae more 
widely spaced (three to five per 100 |j,m) separated by 
flat interspaces bearing finer growth filae. 

Dorsal valve interior with anacline 
pseudointerarea extending approximately 50% valve 
width. Median plate broadly subtriangular, weakly 
depressed medially, merging almost imperceptibly 
with propareas laterally. Anterior margin of 
pseudointerarea raised slightly above valve floor 
medially. Cardinal muscles scars weakly impressed, 
suboval, located posterolaterally, extending 
anteriorly approximately one-third valve length. 



Anterocentral muscle scars and mantle canal patterns 
not observed. Median septum subtriangular in lateral 
profile, extending 83% valve length, bearing dorsally 
concave surmounting plate on posterior margin 
for 70% of length. Surmounting plate originating 
slightly posterior of valve midlength as two ridges, 
separated by dorsally concave plate, merging into 
single blade at 56%) valve length. Anterior margin 
of median septum bearing up to two folds and five 
hollow, septal spines. Two pairs of lateral processes 
developed centrally in mature individuals — first pair 
originating slightly posterior of valve midlength; 
second pair posterior of, and parallel to, first pair, 
originating at approximately one-third valve length. 
Both pairs of lateral processes initially developed as 
low, rounded, anteriorly divergent (at approximately 
90°) rods with rounded anterior margins. Lateral 
processes becoming longer and higher anteriorly with 
increasing valve size, extending to 60% valve length. 
Stubby projections variably developed along anterior 
margins of both pairs of lateral processes. Second pair 
of lateral septa in mature specimens longer and higher 
than first pair. 

Discussion 

Mature dorsal valves of Acrotretella dizeugosa 
are easily distinguished from other Acrotretella by 
the development of two pairs of centrally located 
lateral processes (Fig. 6t, u). In comparison, most 
other Acrotretella with lateral processes only 
possess a single pair. The closely spaced concentric 
lamellae (eight per 100 p,m) on the venfral valve of A. 
dizeugosa (Fig. 6e, h-k), and the variably developed 
low, rounded, medially depressed ridge bounding 
the anterior and anterolateral margins of the dorsal 
valve larval shell (Figs 6y, 9b), are also unique to the 
species. 

Wright and McClean (1991: fig. 1 H-I) figured 
a single acrotretoid dorsal valve from the Ashgill 
Kildare Limestone of Ireland with two pairs of lateral 
processes and a complexly folded median septum 
bearing a dorsally concave surmounting plate along its 
ventral margin. Although Wright and McClean ( 1 99 1 ) 
referred to this specimen as a new, but unnamed, 
acrotretid genus, these features suggest assigmnent 
to Acrotretella. However, unlike A. dizeugosa, the 
Irish taxon has a transversely oval dorsal valve 
outline and both pairs of lateral processes are located 
posteromedially (Wright and McClean 1991). 

Valentine et al. (2003) recognised two species 
of Acrotretella in the Llandovery- Wenlock Boree 
Creek Formation near Orange in central-western 
New South Wales — Acrotretella goodridgei (without 
lateral processes) and Acrotretella sp. A (with 



222 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



lateral processes) (Table 5). Recovery of additional 
specimens of A. goodridgei from the Boree Creek 
Formation and study of Dean- Jones' (1979) material, 
indicates that A. goodridgei passed through a similar 
ontogenetic growth continuum to A. dizeugosa 
(Fig. 8). Acrotretella sp. A is therefore considered 
synonymous with A. goodridgei herein. The first 
two ontogenetic dorsal valve growth stages of A. 
goodridgei are similar to those of ^. dizeugosa. The 
third dorsal valve growth stage of A. goodridgei 
differs in developing a folded dorsal valve median 
septum, prior to the development of lateral processes. 
Valentine et al. (2003) believed that the folded dorsal 
valve median septum oi Acrotretella was restricted to 
individuals with lateral processes, but the additional 
material from the Boree Creek Formation indicates 
this feature can also occur in specimens without 
lateral processes. Acrotretella goodridgei developed 
only a single pair of centrally located lateral processes 
during the fourth dorsal valve growth stage (Fig. 
10). However, one damaged, gerontic dorsal valve 
(1625 fxm wide) with a highly folded median septum 
and well-developed lateral processes, possesses 
a secondary pair of lateral processes inserted 
anteromedially, midway between the median septum 
and the first pair of lateral processes (see Valentine 
et al. 2003 :pl. 2, fig. 26). Acrotretella goodridgei is 
also distinguished by having up to six septal spines 
and four folds along the anterior margin of the dorsal 
valve median septum in mature individuals. 

Family Biematidae Holmer, 1989 
Opsiconidion l^VidVigscn, 191 A 

Type species 

Opsiconidion arcticon Ludvigsen, 1974. 

Opsiconidion ephemerus (Mergl, 1982) 
Fig. 9c-n 

Synonymy 

See Mergl (2001:33) plus the following: 
1984 Opsiconidion podlasiensis n. sp. Biemat, 

p. 97; pi. 26, figs la-c, 2; pi. 27, fig. la-e; 

pi. 28, figs la-c, 2a, b, 3; pi. 29, figs 2a, b, 

3; pi. 30, figs 1, 2, 3a, b; pi. 31, figs la-c, 2. 
2003 Opsiconidion ephemerus (Mergl); 

Williams; pi. 2, fig. 2. 

Description 

See Mergl (1982:115). 

Figured material 

AM F128348 (Fig. 9c, d): dorsal valve, sample 



MU 35; AM F128349 (Fig. 9e, f): dorsal valve; AM 
F128350 (Fig. 9g, h): dorsal valve; AM F 1283 5 1 (Fig. 
9i, j): dorsal valve; AM F128352 (Fig. 9k-n): ventral 
valve. All from sample MU 32 unless otherwise 
mentioned (Table 1). 

Discussion 

The Murruin Creek material is characterised by 
a subcircular dorsal valve outline with maximum 
width occurring around valve midlength. The dorsal 
valve pseudointerarea is anacline and broadly 
subtriangular with a shallowly depressed median 
plate bearing fine growth lines. The anterior margin 
of the pseudointerarea is weakly arcuate (occasionally 
straight) and raised above the valve floor (Fig. 9c, e, 
g, i, j). These features are identical to O. ephemerus 
(Mergl 1982, 2001) and O. podlasiensis from the 
Wenlock Podlasie Depression of Poland (Biemat 
1984). Biemat (1984) noted variations in the dorsal 
valve outline, and in the height and width of the dorsal 
valve pseudointerarea of O. podlasiensis. Similar 
variations also occur in the dorsal valve outline and 
pseudointerarea of the Australian (Fig. 9c, e, g, i, j) 
and Czech material (see Mergl 1982:pl. 1, figs 5, 6, 
8-11). 

The dorsal valve holotype of O. ephemerus is 
700 |im long and 700 jum wide, andMergl (1982:116) 
noted dimensional uniformity within his population. 
Biemat (1984:97) listed the dimensions of the dorsal 
valve holotype oiO. podlasiensis as 690 |im long and 
840 (im wide. On average, the Australian material is 
smaller, only 553 |xm long and 622 ^m wide, but its 
size range encompasses both the Czech and Polish 
material (Table 8). Ventral valves of the type material 
of O. ephemerus and O. podlasiensis are strongly 
conical and can reach over 1000 |.im in height. In 
comparison, the most complete ventral valve of the 
Australian material is only 475 fim high and not as 
strongly conical (Fig. 9k-n). 

The larval shell microomament of O. ephemerus 
and O. podlasiensis consists of circular, flat-bottomed 
pits (3-6 |am in diameter) with few, or no, cross- 
cutting relationships (Fig. lid). This differs from the 
more commonly observed cross-cutting type of larval 
shell pitting observed in Opsiconidion. The Czech 
and Polish material also possess a smaller set of 
pits (0.3-0.5 |Lim in diameter) located on the smooth, 
level areas between the larger pits. No evidence of 
a smaller set of pits were observed in the Murmin 
Creek specimens (Fig. 9d). 

Dorsal valves of Opsiconidion simplex Mergl, 
2001 from the Pridoli Pozary Formation of the Czech 
Republic, have a rounder outline than O. ephemerus 
and a median septum that is consistently shorter 



Proc. Linn. Soc. N.S.W., 127, 2006 



223 



SILURIAN BRACHIOPODS AND CONODONTS 



Table 8. Opsiconidion ephemeras (Mergl, 1982), ventral and dorsal valve dimensions 
(in jam) and ratios. 





Opsiconidion epmemerus (Mergl, 1982), 
ventral valve dimensions (|am) and ratios 
Fa Fw HP LP WP LP/WP 


HP/LP 






N 


17 


17 


20 


20 


18 


17 


19 




MEAN 


29.8 


29.8 


147.4 


159.4 


174.6 


91.9% 


92.4% 






SD 


7.82 


7.82 


26.65 


18.97 


20.4 


0.13 


0.15 






MIN 


18.75 


18.75 


110 


125 


137.5 


71.4% 


62.9% 






MAX 


50 


50 


187.5 


187.5 


225 


115.4% 


115.4% 
























Opsiconidion 
L 


ephemerus (Mergl, 1982), 
W LI WI 


dorsal valve dimensions (^m) and ratios 
LS MHS BS LP WP 


N 


54 


58 


53 


54 


51 


16 


61 


62 


61 


MEAN 


553.2 


622.0 


30.7 


214.1 


473.3 


204.7 


85.7 


180.4 


195.1 


SD 


94.86 


112.64 


9.90 


46.39 


85.92 


53.23 


15.95 


31.58 


25.95 


MIN 


287.5 


362.5 


12.5 


112.5 


250 


75 


50 


100 


150 


MAX 


750 


850 


50 


325 


675 


287.5 


125 


275 


250 




L/W 


LI/WI 


WI/W 


LS/L 


BS/L 


LP/WP 


LP/L 


WPAV 




N 


49 


49 


42 


48 


47 


59 


48 


47 




MEAN 


89.8% 


14.8% 


34.0% 


86.2% 


16.5% 


92.9% 


34.4% 


32.8% 




SD 


0.09 


0.05 


0.07 


0.05 


0.04 


0.12 


0.08 


0.07 




MIN 


71.9% 


6.3% 


24.0% 


65.2% 


9.6% 


57.1% 


19.1% 


17.4% 




MAX 


111.6% 


28.6% 


55.3% 


95.2% 


30.4% 


121.4% 


65.2% 


62.1% 





(only 65-70% valve length) and lower compared 
to other members of the genus (see Mergl 2001:pl. 
30, figs 6, 7, 9-13). Opsiconidion aldridgei (Cocks, 
1979) fi-om the Llandovery of the Welsh Borderlands 
(Cocks 1979), the Llandovery- Wenlock of Saaremaa 
Island, Estonia (Popov 1981) and the Llandovery- 
Wenlock of the Boree Creek Formation near Orange 
in central-western New South Wales (Valentine 
et al. 2003) has a similar subcircular dorsal valve 
outline to O. ephemerus, but has a shorter dorsal 
valve pseudointerarea with a straight anterior margin 
and a well-defined median plate. The dorsal valve 
pseudointerarea of O. angustus Valentine, Brock and 
MoUoy, 2003 from the Llandovery- Wenlock Boree 
Creek Formation near Orange in central-western New 
South Wales, extends approximately 40% valve width 
and has an arcuate anterior margin and an indistinct 
median plate. Opsiconidion angustus also has a 
transversely suboval dorsal valve outline (Valentine 
et al. 2003). 

Opsiconidion sp. 
Fig. 9o-q 

Synonymy 

cf 1999 Opsiconidion sp. Cockle; pi. 5, fig. 15. 



cf. 2003 Opsiconidion sp. A Valentine, Brock and 
MoUoy, p. 317; pi. 3, figs 16, 17. 

Figured material 

AM F128353 (Fig. 9o): dorsal valve, sample 
MU 31; AM F128354 (Fig. 9p, q): dorsal valve, 
sample MU 36 (Table 1). 

Discussion 

The Murruin Creek specimens differ firom most 
Opsiconidion by their transversely elliptical dorsal 
valve outline (Fig. 9o, p). The anacline dorsal valve 
pseudointerarea is broadly subtriangular with a 
weakly depressed median plate and a straight anterior 
margin that is raised above the valve floor (Fig. 
9o). The median septum is low and subtriangular 
in lateral profile. Elliptical Opsiconidion also occur 
in the Wenlock Borenore Limestone near Orange 
in central-western New South Wales (Cockle 1999; 
Valentine et al. 2003). The Borenore specimens have 
a dorsal valve pseudointerarea with a more strongly 
depressed median plate and a gently arcuate anterior 
margin, but insufficient material is currently available 
fi-om both localities to determine if these differences 
are significant. Mergl (2001) also documented an 
elliptical Opsiconidion, Opsiconidion sp. A, firom 



224 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



the Llandovery Zelkovice Formation of the Czech 
Republic. The Czech material is not as strongly 
elliptical as the Murruin Creek specimens (compare 
Fig. 9o, p with Mergl 2001 :pl. 29, figs 9, 12) and has 
a dorsal valve pseudointerarea with a more strongly 
depressed median plate. 

Order Siphonotretida Kuhn, 1949 

Superfamily Siphonotretidae Kutorga, 1 848 

Family Siphonotretidae Kutorga, 1848 

siphonotretid gen. et sp. indet. 1 
Fig. 9r-t 

Figured material 

AM F128355 (Fig. 9r, s): dorsal valve, sample 
MU 36 (Fig. 3); AM F 128356 (Fig. 9t): dorsal valve, 
sample MU 34 (Table 1). 

Discussion 

The Murruin Creek siphonotretid differs fi-om 
Orbaspina in lacking a pitted post-larval shell and 
possesses erect spines that are scattered evenly across 
the valve surface (Fig. 9r). Schizambonine sp. B from 
the Pragian Praha Formation of the Czech Republic 
also lacks a pitted post-larval shell, but is distinguished 
by its well-developed dorsal valve sulcus and prostrate 
spines that tend to be restricted to the valve margins 
(Mergl 2001:pl. 36, figs 11-13). Acanthambonine sp. 
fi-om the Pragian Dvorce-Prokop Limestone of the 
Czech Republic, whilst also lacking a pitted post- 
larval shell, has more widely spaced, suberect spines 
and a submarginal dorsal valve larval shell (Mergl 
2001 :pl. 36, figs 1,4, 5-7). Little is known concerning 
the internal morphology of any of these species. The 
apsacline dorsal valve pseudointerarea of the Murruin 
Creek siphonotretid is well-developed and shelf-like 
(Fig. 9t), similar to the dorsal valve pseudointerarea 
of Orbaspina. 

Phylum Conodonta Pander, 1856 

Genus 5e/o<ie//a Ethington, 1959 

Type species 

Belodus devonicus Stauffer, 1940. 

Belodella anomalis Cooper, 1974 
Fig. lOa-i 

Synonymy 

See Farrell (2004:947) plus the following: 
1993 Belodella sp. aff. B. anomalis Cooper; 
Simpson et al., p. 153; fig. 4J. 



Figured material 

AMF128283 (Fig 10a): Sb element; AM F 128284 
(Fig. 10b): Sb element; AM F128285 (Fig. 10c): Sb 
element; AM F128286 (Fig. lOd): Sd element; AM 
F128287 (Fig. lOe): Sc element; AM F128288 (Fig. 
lOf): t element; AM F128289 (Fig. lOg): fragment of 
?t element; AM F128290 (Fig. lOh): M element; AM 
F 128291 (Fig. lOi): M element. All from sample MU 
34 (Table 2). 

Description 

See Farrell (2004:948). 

Discussion 

This species from Murruin Creek is reconstructed 
recognising all five elements recorded by Farrell 
(2004), ie. Sa, Sb, Sc, Sd and 't' or tortiform elements, 
plus an adenticulate M element. The presence of the 
adenticulate M element in reconstructions of the genus 
has been discussed by Barrick and Klapper (1992) 
and is often still recognisable in small collections (eg. 
Mawson et al. 1995). The M element of 5. anomalis 
(Fig. lOh, i) is more strongly curved than the M 
element of Belodella resima (see Mawson et al. 1995: 
pi. 4, fig 1.) and Belodella cf B. resima (see Barrick 
and Klapper 1992:pl. 1, fig. 7) and is more robust and 
broad-based than the M element of Belodella anfracta 
(see Barrick and Klapper 1992:pl. 1 fig. 9). 

Cooper (1974) established the diagnostic 
characteristic of B. anomalis as the denticulated 
anterior margin, but also noted the distinctive apical 
'fan-like' structure of denticles. Simpson et al. (1993: 
fig. 4J) illustrated a specimen from the Cowombat 
Formation at Cowombat Flat in eastem Victoria 
lacking the distinctive fan-like denticulation near the 
cusp and assigned it, with some doubt, to B. anomalis. 
It is now included within the species concept because 
the serrated nature of the anterior margin represents 
putative denticulation. 

Genus Coryssognathus Link and Druce, 1972 

Type Species 

Cordylodusl dubius Rhodes, 1953. 

Coryssognathus dubius (Rhodes, 1953) 
Fig. lle,f 

Synonymy 

See Simpson and Talent (1995:163) and Farrell 

(2004:959), plus the following: 
2002 Coryssognathus dubius (Rhodes); 

Talent et al; pi. 2, figs U-W; pi. 4, figs F, G. 



Proc. Linn. Soc. N.S.W., 127, 2006 



225 



SILURIAN BRACHIOPODS AND CONODONTS 




Figure 10. a-i. Belodella anomalis Cooper, 1974, all from sample MU 34, a. Sb element AM F128283; 
lateral view. b. Sb element AM F128284; lateral view. c. Sb element AM F128285; lateral view. d. Sd 
element AM F128286; lateral view. e. Sc element AM F128287; lateral view. f. t element AM F128288; 
lateral view. g. fragment of ?t element AM F128289; lateral view; h. M element AM F128290; later- 
al view. i. M element AM F128291; lateral view. j-1. Dapsilodus obliquicostatus (Branson and Mehl, 
1933) all from sample MU 37 unless otherwise mentioned, j. M element AM F128292; lateral view, 
k. M element AM F128293, sample MU 38; lateral view. 1. M element AM F128294; lateral view, 
m. Decoriconus fragilis (Branson and Mehl, 1933). Sc element AM F128295, sample MU 34; lat- 
eral view, n, o. Panderodus recurvatus (Rhodes, 1953), both from sample MU 34. n. Sc element AM 
F128296; lateral view. o. Sb element AM F128297; lateral view. p. Panderodus unicostatus (Branson 
and Mehl, 1933). M element AM F128302, sample MU 34; lateral view. All scale bars equal 100 p^m. 



Description 

See Miller and Aldridge (1993:246). 

Figured material 

AM F 128303 (Fig. lie): Pa element; AM 



F128304 (Fig. llf): partly preserved Sb element. 
Both from sample MU 34 (Table 2). 

Discussion 

The partially preserved Sb element from Murrain 
Creek has a prominent cusp and the remains of a lateral 



226 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 




Figure 11. a, b. Panderodus unicostatus (Branson and Mehl, 1933), both from sample MU 34. a. Sa ele- 
ment AM F128300; lateral view. b. M element AM F128301; lateral view, c, d. Panderodus serratus 
Rexroad, 1967, both from sample MU 34. c. Sc element AM F128299; lateral view. d. Sb element AM 
F128298; lateral view, e, f. Coryssognathus dubius (Rhodes, 1953), both from sample MU 34 e. Pa ele- 
ment AM F128303; lateral view. f. Partly preserved Sb element AM F128304; lateral view, g, h. Oulodus 
sp. cf. Oulodus elegans (Walliser, 1964), both from sample MU 34. g. Sb element AM F128305; lat- 
eral view. h. Sa element AM F128306; lateral view. i-m. Ozarkodina excavata excavata (Branson and 
Mehl, 1933) all from sample MU 34 unless otherwise mentioned, i. M element AM F128307, sample 
MU 38; inner lateral view. j. Sa element AM F128308; lateral view. k. Sb element AM F128309; inner 
lateral view. 1. Sc element AM F128311; inner lateral view. m. Pa element AM F128310; lateral view, 
n, 0. Kockelella maenniki Serpagli and Corradini, 1998, both from sample MU 34. n. Sc element AM 
F128312; inner lateral view. o. Pa element AM F128313; oblique upper view. All scale bars equal 100 |am. 



process bearing a single denticle adjacent to the break 
in the lateral process. The cusp is evenly curved toward 
the posterior and tapers evenly toward the apex. The 
denticulate process projects downward from a broad 
'dished' area at the base of the cusp. The basal margin 
of the process curves toward the anterior from the 



'dished' area (Fig. 1 If). The poorly preserved scaphate 
Pa element has an erect, triangular cusp only slightly 
larger than the other denticles. No denticles were 
observed on the lateral process and it may therefore 
represent a juvenile Pa element (Fig. He) (Miller and 
Aldridge 1993). 



Proc. Linn. Soc. N.S.W., 127, 2006 



227 



SILURIAN BRACHIOPODS AND CONODONTS 



Genus Dapsilodus Cooper, 1976 

Type species 

Distacodus obliquicostatus Branson and Mehl, 
1933. 

Dapsilodus obliquicostatus (Branson and Mehl, 1933) 
Fig. lOj-1 

Synonymy 

See Armstrong (1990:70), plus the following: 
1990 Dapsilodus obliquicostatus (Branson and 

Mehl) Uyeno, p. 98; pi. 2, figs 11-16. 
71992 Dapsilodus sp. Barrick and Klapper, p. 44; 

pl.2,fig.2. 
1994 Dapsilodus obliquicostatus (Branson and 

Mehl); Sarmiento et al.; pi. 1, figs 1, 6. 
1999 Dapsilodus obliquicostatus (Branson and 

Mehl); Cockle, p. 119; pi. 4, figs 13-19. 

Description 

See Cooper (1976:212). 

Figured material 

AMF128292 (Fig. 10j):M element; AMF128293 
(Fig. 10k): M element, sample MU 38; AM F128294 
(Fig. 101): M element. All fi-om sample MU 37 unless 
otherwise mentioned (Table 2). 

Discussion 

It has not been possible to separate the Sb and Sc 
elements fi-om Murruin Creek as morphologies appear 
gradational and they have therefore been tabulated 
together (Table 2). The M elements recovered (Fig. 4j- 
1) are recurved with a prominent costa almost centrally 
positioned in lateral view. Oblique striations are present 
along the anterior margin in some elements. Among the 
M elements, the point of maximum curvature shows 
some variability in relation to the generally shallow 
basal cavity. 

Genus Z)econco«M.s Cooper, 1975 

Type species 

Paltodus costulatus Rexroad, 1967. 

Decoriconus fragilis (Branson and Mehl, 1933) 
Fig. 10m 
Synonymy 

See McCracken (1991:79), Zhang and Barnes 
(2002:11) and Farrell (2004:958). 

Description 

See Barrick (1977:53). 



Figured material 

AM F 128295 (Fig. 10m): Sc element, sample 
MU 34 (Table 2). 

Discussion 

The Sc elements of D. fragilis fi-om Murruin 
Creek are of the typical 'drepanodonti-form' first 
identified by Cooper (1975). These distinctive 
elements are inclined, with an almost straight anterior 
margin and are generally compressed, but expanded 
around the small basal cavity (Fig. 10m). 

Genus ^ocA:e/e//a Walliser, 1957 

lype species 

Kockelella variabilis Walliser, 1957. 

Kockelella maenniki Serpagli and Corradini, 1998 
Fig. 1 In, o 

Synonymy 

See Serpagli and Corridini (1999:284). 

Description 

See Serpagli and Corradini (1999:286). 

Figured material 

AM F128312 (Fig. lln): Sc element; AM 
F128313 (Fig. llo): Pa element. Both from sample 
MU 34 (Table 2). 

Discussion 

The laterally compressed Pa element of this 
taxon fi-om Murruin Creek is curved, slightly arched 
and narrow with a strongly asymmetrical platform. 
The anterior portion of the blade has ten closely 
packed, compressed denticles. The posterior portion 
of the blade arches downwards and bears six closely 
spaced, laterally compressed denticles. The outer 
lateral process has five aligned, but slightly proclined 
denticles. The shorter inner lateral process appears to 
bear a single small denticle fiised to the cusp (Fig. 
lln). It should be noted that not all of Serpagli and 
Corradini's (1999:pl. 3, fig. 10) specimens have 
denticulate lateral processes. The Sc element is 
slender and has well-spaced denticles with a slightly 
twisted and downwardly deflected, antero-lateral 
process (Fig. lln). 

The stratigraphic range of K. maenniki is 
interpreted as restricted to the lower to middle part of 
the P. siluricus Zone (Corradini and Serpagli 1999; 
Serpagli and Corradini 1999). Corradini et al. (1998) 
reported that the genus Kockelella became extinct 



228 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



before the close of the siluricus Zone and that K. 
maenniki therefore represents the terminal taxon of the 
genus. Kockelella maenniki also occurs in the Ludlow 
Coral Gardens sequence of the Jack Formation in 
northern Queensland, where it occurs just below the 
youngest occurrence of P. siluricus. 

Genus Oulodus Branson and Mehl, 1933 

Type Species 

Oulodus serratus Staufifer, 1930. 

Oulodus sp. cf. Oulodus elegans (Walliser, 1964) 
Fig. llg, h 

Figured material 

AM F128305 (Fig. llg): Sb element; AM 
F 128306 (Fig. llh): Sa element. Both from sample 
MU 34 (Table 2). 

Discussion 

The ramiform elements possess discrete, peg-like 
denticles and a prominent cusp that curves toward the 
lateral view. The anterior process of the Sb elements 
bear six or seven denticles and the postero-lateral 
process six denticles (Fig. llg). The Sa elements are 
bilaterally symmetrical about the lateral processes 
which possess four denticles (Fig. llh). 

Genus Ozarkodina Branson and Mehl, 1933 

lype species 

Ozarkodina typica Branson and Mehl, 1933. 

Ozarkodina excavata excavata (Branson and Mehl, 

1933) 

Fig. lli-m 

Synonymy 

Simpson and Talent (1995:147) and Farrell 
(2003:123) covered the majority of published 
accounts. However, at least an additional 20 illustrated 
records pre- and postdating the synonymies cited above 
exist, but due to space limitations it was not possible 
to include them. This will be undertaken in another 
publication where the primary focus is conodont 
taxonomy. 

Description 

See Simpson and Talent (1995:152). 

Figured material 

AM F128307 (Fig. Hi): M element, sample MU 
38; AM F 128308 (Fig. llj): Sa element; AM F 128309 
(Fig. Ilk): Sb element; AM F128311 (Fig. Ill): Sc 



element; AM F 1283 10 (Fig. 11m): Pa element. All 
from sample MU 34 unless otherwise mentioned 
(Table 2). 

Discussion 

This species from Murruin Creek shows the long, 
discrete denticles and well-developed basal cavity 
typical of this ubiquitous Silurian to Early Devonian 
taxon (Fig. lli-m). The Sa elements show some 
variation in the angle between the processes (Fig. Hi), 
but Farrell (2003, 2004) reported similar variations in 
his material from the Late Silurian to Early Devonian 
Camelford Limestone and the Early Devonian Garra 
Limestone at Wellington in central-western New 
South Wales. The Pa and Pb elements display the 
typical anterior and posterior process morphology with 
closely packed compressed denticles and a prominent 
cusp (Fig. 11m, o). 

Genus Panderodus Ethington, 1959 

Type Species 

Paltodus unicostatus Branson and Mehl, 1933. 

Panderodus recurvatus (Rhodes, 1953) 
Fig. lOn, o 

Synonymy 

See Simpson and Talent (1995: 1 1 7) and Farrell 

(2003:122), plus the following: 
1 995 Panderodus recurvatus (Rhodes); 

Colquhoun, p. 354; pi. 3, fig. 4. 
1999 Panderodus recurvatus (Rhodes); Cockle, 

p. 120; pi. 5, figs 9-14. 
2002 Panderodus recurvatus (Rhodes); Aldridge; 

pi. 4, figs 4-7. 
2002 Panderodus recurvatus (Rhodes); Talent et 

al.; pi. 2, figs J, K. 
2002 Panderodus recurvatus (Rhodes); Zhang 

and Barnes, p. 3 1 ; figs 16.1-16.27. 
2004 Panderodus recurvatus (Rhodes); Farrell, p. 

958;pl.3,figs9, 12, 13. 

Description 

See Barrick (1977:54). 

Figured material 

AM F128296 (Fig. lOn): Sc element; AM 
F 128297 (Fig. lOo): Sb element. Both from sample 
MU 34 (Table 2). 

Discussion 

The available elements of P. recurvatus from 
Murruin Creek are all broken to a greater or lesser 
extent, but are distinctly recurved, lack ornament and 



Proc. Linn. Soc. N.S.W., 127, 2006 



229 



SILURIAN BRACHIOPODS AND CONODONTS 



possess a longitudinal groove developed along the 
middle to posterior portion of one lateral surface (Fig. 
lOn, o). 

Panderodus serratus Rexroad, 1967 
Fig. lie, d 
Synonymy 

1997 Panderodus serratus Rexroad; Jeppsson, p. 
107; fig. 7.4. 

Description 

See Jeppsson (1997:107). 

Figured material 

AM F 128299 (Fig. lie): So element; AM 
F128298 (Fig. lid): Sb element. Both fi-om sample 
MU 34 (Table 2). 



Discussion 

Despite being a ubiquitous component of 
many Silurian conodont faunas, the taxonomy of P. 
unicostatus is poorly understood. Specimens typically 
vary morphologically in terms of shape, total height 
and in the degree and location of strongest curvature 
(Jeppsson 1975; Simpson and Talent 1995). This 
variation between elements is such that distinction 
between S series elements is often problematic and 
intergradational morphologies possibly exist (Dzik 
and Drygant 1986; Sweet 1988). Jeppsson (1997) 
considered that intemal structures of Panderodus, 
such as the form of the basal cavity and white matter 
distribution, to be taxonomically significant. Although 
over a thousand elements of this taxon were recovered 
fi-om Murruin Creek (Table 1), many are broken at, or 
near the basal cavity termination. 



Discussion 

Jeppsson (1997:107) noted a close similarity 
between P. serratus and P. unicostatus, and indicated 
they could only be separated by the serrate posterior 
margin of the arcuatiform (Sc) element of P. serratus. 
He did not, however, indicate whether serrations 
were present on other elements. The Murruin Creek 
specimens are rare (Table 2), but there are clear 
examples of a serrate Sc element (Fig. lie) and one 
interpreted as a Sb element (Fig. lid). 

Panderodus unicostatus (Branson and Mehl, 1933) 
Figs lOp; 11a, b 

Synonymy 

See Simpson and Talent (1995: 1 1 8) and Farrell 

(2004:959), plus the followmg: 
1997 Panderodus unicostatus (Branson and 

Mehl); Jeppsson, p. 107; fig. 7, 7.3. 
1999 Panderodus unicostatus (Branson and 

Mehl); Cockle, p. 120; pi. 5, figs 1-8. 
2002 Panderodus unicostatus (Branson and 

Mehl); Aldridge; pi. 4, figs 8-17. 
2002 Panderodus unicostatus (Branson and 

Mehl); Talent et al.; pi. 2, fig. I. 
2002 Panderodus unicostatus (Branson and 

Mehl); Zhang and Barnes, p. 32; figs 15.1- 

15.24. 

Description 

See Cooper (1976:213). 

Figured material 

AMF128302(Fig. 10p):Melement;AMF128300 
(Fig. 11a): Sa element; AM F128301 (Fig. lib): M 
element. All fi-om sample MU 34 (Table 2). 



ACKNOWLEDGEMENTS 

The authors greatfully acknowledge the assistance 
provided in the field by John Talent, Ruth Mawson and 
Warrick Try. Peter Molloy assisted with acid processing of 
samples and shared his knowledge of Silurian conodonts. 
Heidi-Jane Caldon helped pick samples. John Paterson 
kindly identified trilobite remains recovered during this 
study. Peter Cockle and John Farrell generously donated 
specimens of Acrotretella fi-om their conodont residues and 
Patrick Conaghan provided access to the linguliformean 
brachiopod fauna of the late Gunther Bischoff from the 
Boree Creek Formation. Dean Oliver skillfully drafted the 
geological maps and stratigraphic columns. This project 
would not have been possible without the generosity of 
Leo and Robin Chalker who kindly granted permission 
to collect samples on their property on several occasions. 
This manuscript benefited greatly fi-om the constructive 
comments made by Glenn Brock (Macquarie University, 
Sydney) and two anonymous reviewers. 



REFERENCES 

Aldridge, R.J. (2002). Conodonts from the Skomer 
volcanic group (Llandovery Series, Silurian) 
of Pembrokeshire, Wales. Special Papers in 
Palaeontology 61 , 15-28. 

Armstrong, H.A. (1990). Conodonts from the Upper 
Ordovician-Lower Silurian carbonate platform 
of North Greenland. Grenlands Geologiske 
Unders0gelse Bulletin 159, 1-151. 

Barrande, J. (1879). 'Systeme Silurien du centre de la 
Boheme. lere partie. Recherches paleontologiques, 
vol. 5. Classe des Mollusques. Ordre des 
Brachiopodes.' (Published by the author: Prague and 
Paris). 



230 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



Barrick, J.E. (1977). Multielement simple-cone conodonts 

from the Clarita Formation (Silurian) Arbuckle 

Mountains, Oklahoma. Geologica et Palaeontologica 

11, 47-68. 
Barrick, J.E. and Klapper, G. (1992). Late Silurian-Eariy 

Devonian conodonts from the Hunton Group (Upper 

Henryhouse, Haragan, and Bois d'Arc Formations), 

south-cenfral Oklahoma. Oklahoma Geological 

Survey Bulletin 145, 19-65. 
Bassett, M.G. (1986). Brachiopodes inarticules. 

Biostratigraphie du Paleozoique 3, 85-96. 
Bednarczyk, W. and Biemat, G. (1978). Inarticulate 

brachiopods from the Lower Ordovician of the Holy 

Cross Mountains, Poland. Acta Palaeontologica 

Polonica 23, 293-316. 
Biemat, G. (1973). Ordovician inarticulate brachiopods 

from Poland and Estonia. Palaeontologica Polonica 

28, 1-120. 
Biemat, G. (1984). Silurian inarticulate brachiopods from 

Poland. Acta Palaeontologica Polonica 29, 99-103. 
Biemat, G. and Harper, D.A.T. (1999). A Ungulate 

brachiopod Acrotretella: new data from Ordovician 

of Poland. Acta Palaeontologica Polonica 44, 83-92. 
Bischoff, G.C.O. (1986). Early and Middle Silurian 

conodonts from midwestem New South Wales. 

Courier Forschungsinstitut Senckenberg 89, 1-337. 
Branson, E.B. and Mehl, M.G. (1933). Conodonts studies 

no. 1: conodonts from the Harding Sandstone 

of Colorado; Bainbridge (Silurian) of Missouri; 

Jefferson City (Lower Ordovician) of Missouri. The 

University of Missouri Studies 8, 39-53. 
Chatterton, B.D.E. and Whitehead, H.L. (1987). Predatory 

borings in the inarticulate brachiopod Artiotreta from 

the Silurian of Oklahoma. Lethaia 20, 61-1 A. 
Chems, L. (1979). The environmental significance of 

Lingula in the Ludlow series of the Welsh Borderland 

and Wales. Lethaia 12, 35-46. 
Cockle, P. (1999). Conodont data in relation to time, space 

and environmental relationships in the Silurian (late 

Llandovery-Ludlow) succession at Boree Creek 

(New South Wales, Australia). Abhandlungen de 

Geologischen Bundesanstalt 54, 107-133. 
Cocks, L.R.M. (1979). New acrofretacean brachiopods 

from the Palaeozoic of Britain and Austria. 

Palaeontology 22, 93-100. 
Colquhoun, G.P. (1995). Early Devonian conodont faunas 

from the Capertee High, NE Lachlan Fold Belt, 

southeastern Ausfralia. Courier Forschungsinstitut 

Senckenberg 182, 347-370. 
Cooper, B.J. (1974). New forms of Belodella (Conodonta) 

from the Silurian of Ausfralia. Journal of 

Paleontology 48, 1120-1125. 
Cooper, B.J. (1975). Multielement conodonts from the 

Brassfield Limestone (Silurian) of southern Illinois. 

Journal of Paleontology 49, 984-1008. 
Cooper, B.J. (1976). Multielement conodonts from the St. 

Clair Limestone (Silurian) of southem Ohio. Journal 

of Paleontology 50, 205-217. 



Cooper, G.A. (1956). Chazyan and related brachiopods. 

Smithsonian Miscellaneous Collections 127, 1-1024, 

1025-1245. 
Corradini, C. and Serpagli, E. (1999). A Silurian 

conodont biozonation from late Llandovery to end 

Pfidoli in Sardinia (Italy). Bollettino della Societd 

Paleontologica Italiana 37, 275-298. 
Corradini, C, Ferretti, A., Serpagli, E. and Barca, 

S. (1998). The Ludlow-Pridoli Section "Genna 

Ciuerciu" west of Silius. Giornale di Geologia 60, 

112-118. 
Davidson, T. (1848). Memoire sur les Brachiopodes du 

Systeme Silurien superieur de I'Angleterre. Societe 

Geologique de France, Bulletin 5, 309-338, 370-374. 
Dean-Jones, G. (1979). Late Cambrian to Early Devonian 

inarticulate brachiopods from Ausfralia: their 

classification, ontogeny, functional morphology and 

ulfrastmcture. Unpublished MSc thesis, Macquarie 

University, Sydney. 
d'Orbigny, A. 1847. Considerations zoologiques et 

geologiques sur les Brachiopodes ou Palliobranches. 

Comptes Rendus Hebdomadaires des Seances de 

r Academic des Sciences 25, 193-195, 266-269. 
Dzik, J. and Drygant, D.M. (1986). The apparatus of 

panderodontid conodonts. Lethaia 19, 133-141. 
Ethington, R.L. (1959). Conodonts from the Ordovician 

Galena Formation. Journal of Paleontology 33, 257- 

292. 
Farrell, J.R. (2003). Late Pridoli, Lochkovian and early 

Pragian conodonts from the Gap area between Larras 

Lee and Eurimbla, central westem NSW, Australia. 

Courier Forschungsintitut Senckenberg 245, 107- 

181. 
Farrell, J.R. (2004). Siluro-Devonian conodonts from the 

Camelford Limestone, Wellington, New South Wales, 

Ausfralia. Palaeontology 47, 937-982. 
Foerste, A. (1888). Notes on Paleozoic fossils. Bulletin of 

the Scientific Laboratories ofDenison University 3, 

117-137. 
Gorjansky, V.I. (1969). Bezzamkovye brakhiopody 

kembriiskikh i ordovikskikh otlozhenii sever-zapada 

Russkoi platformy. Materialy po geologii ipoleznym 

iskopaemym severo-zapada R.S.F.S.R. 6, 1-173. 
Hall, J. (1872). Notes on some new or imperfectly known 

forms among the Brachiopoda, etc. New York State 

Cabinet of Natural History, Annual Report 23, 244- 

247. 
Haug, E. (1883). Ueber sogermannte Chaetetes aus 

mesozoischen Ablagemngen. Neues Jahrbuchfuer 

Miner alogie, Geologic und Paldeontologie 1, 171- 

179. 
Holmer, L.E. (1986). Inarticulate brachiopods around the 

Middle-Upper Ordovician boundary in Vastergotland. 

Geologiska Foreningens i Stockholm Forhandlingar 

108, 97-126. 
Holmer, L.E. (1989). Middle Ordovician phosphatic 

inarticulate brachiopods from Vastergotland and 

Dalama, Sweden. Fossils and Strata 26, 1-172. 



Proc. Linn. Soc. N.S.W., 127, 2006 



231 



SILURIAN BRACHIOPODS AND CONODONTS 



Holmer, L.E. (1991). The systematic position of 

Pseudolingula Mickwitz and related lingulacean 
brachiopods. In 'Brachiopods through time' (Eds D.I. 
MacKinnon, D.E. Lee and K.S.W. Campbell) pp. 15- 
21. (A. A. Balkema: Rotterdam). 

Holmer, L.E. and Popov, L.E. (2000). Lingulata. In 
'Treatise on Invertebrate Paleontology, part H, 
Brachiopoda (revised), vol. 2' (Ed. R.L. Kaesler) 
pp. 35-146. (Geological Society of America and The 
University of Kansas: Boulder and Lawrence). 

Huleatt, M.B. (1969). The geology of Palaeozoic 

sediments south-east of Taralga, N.S.W. Unpublished 
BSc Thesis, Australian National University, Canberra. 

Ireland, H.A. (1961). New phosphatic brachiopods from 
the Silurian of Oklahoma. Journal of Paleontology 
35, 1137-1142. 

Jeppsson, L. (1975). Aspects of Late Silurian conodonts. 
Fossils and Strata 6, 1-54. 

Jeppsson, L. (1989). Latest Silurian conodont fauna 
from Klonk, Czechoslovakia. Geologica et 
Palaeontologica 23, 21-37. 

Jeppsson, L. (1997). Anew latest Telychian, Sheinwoodian 
and Early Homerian (Early Silurian) standard 
conodont zonation. Transactions of the Royal Society 
of Edinburgh: Earth Sciences 88, 91-114. 

Jongsma, D. (1968). Geology of the upper Murruin Creek: 
an area between Mnt. Werong and Mnt. Shivering, 
N.S.W. Unpublished BSc Hons Thesis, University of 
New South Wales, Sydney. 

Klyza, J.S. (1997). The Tamworth Belt in the Tamworth- 
Attunga area: stratigraphy, structure, biochronologic 
and palaeoenvironmental analysis. Unpublished MSc 
thesis, Macquarie University, Sydney. 

Kutorga, S.S. (1846). Uber das silurische und devonische 
Schichten-System von Gratschina. Russisch- 
Kaiserliche Mineralogische Gesellschaft zu St. 
Petersbourg, Verhandlungen 1845-1846, 85-139. 

Kutorga, S.S. (1848). Uber die brachiopoden-familie 
der Siphonotretaeae. Russisch-Kaiserliche 
Mineralogische Gesellschaft zu St. Petersbourg, 
Verhandlungen 1847, 250-286. 

Link, A.G. and Druce, E.C. (1972). Ludlovian and 

Gedinnian conodont stratigraphy of the Yass basin. 
New South Wales. Australian Bureau of Mineral 
Resources, Geology and Geophysics Bulletin 134, 
1-136. 

Ludvigsen, R. (1974). A new Devonian acrotretid 

(Brachiopoda, Inarticulata) with unique protegular 
ultrastructure. Neues Jahrbuchfur Geologic und 
Paldontologie Monatschefte 1974, 133-148. 

Matthews, K.M.C. (1985). The nature of the contact 
between the Burra Burra Creek Formation and 
the Cobra Formation, N.E. of Taralga, N.S.W. 
Unpublished BSc Hons Thesis, Macquarie 
University, Sydney. 

Mawson, R., Talent, J. A. and Furey-Greig, T.M. (1995). 
Coincident conodont faunas (late Emsian) from 
the Yarrol and Tamworth belts of northern New 
South Wales and central Queensland. Courier 
Forschungsinstitut Senckenberg 182, 421-445. 



McCracken, A.D. (1991). Taxonomy and biostratigraphy 
of Llandovery (Silurian) conodonts in the Canadian 
Cordillera, northern Yukon Territory. Geological 
Survey of Canada Bulletin 417, 65-95. 

Mergl, M. (1982). Caenotreta (Inarticulata, Brachiopoda) 
in the Upper Silurian of Bohemia. Vestnik Ustfedniho 
ustavu geologickeho 57 , 115-116. 

Mergl, M. (1995). New Ungulate brachiopods from 
the Milina Formation and the base of the Klabava 
Formation (late Tremadoc-early Arenig), cenfral 
Bohemia. Vestnik Ceskeho geologickeho ustavu 70, 
101-109. 

Mergl, M. (1996). Discinid brachiopods from the 

Kopanina Formation (Silurian) of Amerika quarries 
near Mofina, Barrandian, central Bohemia. Casopis 
Ndrodniho Muzea Rada Pfirodovednd 165, 121-126. 

Mergl, M. (1999a). Genus Paterula (Brachiopoda) in 
Ordovician-Silurian sequence of cenfral Bohemia. 
Vestnik Ceskeho geologickeho ustavu 74, 347-362. 

Mergl, M. (1999b). Genus Lingulops (Lingulata, 

Brachiopoda) in Silurian of the Barrandian. Journal 
of the Czech Geological Society 44, 155-158. 

Mergl, M. (2001). Lingulate brachiopods of the Silurian 
and Devonian of the Barrandian (Bohemia, Czech 
Republic). Acta Musei Nationalis Prague, Series B, 
Historia Naturalis 51 , 1-49. 

Mergl, M. (2002). Linguliformean and craniiformean 

brachiopods of the Ordovician (Tfenice to Dobrotiva 
Formations) of the Barrandian, Bohemia. Acta Musei 
Nationalis Prague, Series B, Natural History 58, 1- 
82. 

Miller, C.G. and Aldridge, R.J. (1993). The taxonomy and 
apparatus structure of the Silurian distomodontid 
conodont Coryssognathus Link and Druce, 1972. 
Journal of Micropalaeontology 12, 241-255. 

Morritt, R.F.C. (1979). Structural analysis of Palaeozoic 
rocks near Taralga, N.S.W. Unpublished BSc Hons 
Thesis, Macquarie University, Sydney. 

Munson, T.J., Pickett, J.W and Strusz, D.L. (2000). 

Biosfratigraphic review of the Silurian tabulate corals 
and chaetetids of Australia. Historical Biology 15, 
41-60. 

Naylor, G.F.K. (1937). Preliminary note on the occurrence 
of Palaeozoic strata near Taralga, N.S.W. Journal and 
Proceedings of the Royal Society of New South Wales 
71,45-53. 

Nazarov, B.B. and Popov, L.E. (1976). Radiolyarii, 

bezzamkovye brakhiopody i organizmy neyasnogo 
sistematicheskogo polozheniya iz srednego ordovika 
vosttochnogo Kazakhstana. Paleontologicheskii 
Zhurnal 4, 33-42. 

Nazarov, B.B. and Popov, L.E. (1980). Sfratigrafiya i 
fauna kremnisto-karbonatnykh tolshch ordovika 
Kazakhstana (radioliarii i bezzamkovye 
brakhiopody). Geologicheskiy Institut Akademii Nauk 
SSSR, Trudy 33, 1-190. 

Philip, G.M. (1966). Lower Devonian conodonts from the 
Buchan Group, eastern Victoria. Micropaleontology 
12, 441-460. 



232 



Proc. Linn. Soc. N.S.W., 127, 2006 



J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON 



Pickett, J.W. (Ed.) (1982). The Silurian System in New 
South Wales. Bulletin of the Geological Survey of 
New South Wales 29, 1-264. 

Pickett, J.W. (1985). Silurian corals from north of 

Wombeyan Caves. Unpublished Geological Survey 
of New South Wales Palaeontological Report 85/7. 

Pickett, J.W., Burrow, C.J., Holloway, D.J., Munson, T.J., 
Percival, I.G., Rickards, R.B., Sherwin, L., Simpson, 
A.J., Strusz, D.L., Turner, S. and Wright, A.J. (2000). 
Silurian palaeobiogeography of Australia. Memoirs of 
the Association of Australasian Palaeontologists 23, 
127-165. 

Popov, L.E. (1981). Pervaia nakhodka mikroskopicheskikh 
bezzamkovye brakhiopod semeistva Acrotretidae 
V silurie Estonii. Eesti NSV Teaduste Akadeemia 
Toimetised (Geologia) 30, 34-41. 

Popov, L.E. (2000). Late Ordovician linguliformean 
microbrachiopods from north-central Kazakhstan. 
Alcheringa 24, 257-275. 

Popov, L.E. and Holmer, L.E. (1994). Cambrian- 

Ordovician Ungulate brachiopods from Scandinavia, 
Kazakhstan, and South Ural Mountains. Fossils and 
Strata 25, 1-156. 

Powell, C.McA., Edgecombe, D.R., Henry, N.M. and 

Jones, J.G. (1976). Timing of regional deformation of 
the Hill End Trough: a reassessment. Journal of the 
Geological Society of Australia 23, 407-421. 

Powell, C.McA. and Fergusson, C.L. (1979a). Analysis 
of the angular discordance across the Lambian 
Unconformity in the Kowmung River - Murruin 
Creek area, eastern N.S.W. Journal and Proceedings 
of the Royal Society of New South Wales 112, 37-42. 

Powell, C.McA. and Fergusson, C.L. (1979b). The 
relationship of structures across the Lambian 
unconformity near Taralga, New South Wales. 
Journal of the Geological Society of Australia 26, 
209-219. 

Rexroad, C.B. (1967). Stratigraphy and conodont 
paleontology of the Brassfield (Silurian) in the 
Cincinnati Arch area. Indiana Geological Survey 
Bulletin 36, 1-69. 

Rhodes, F.H.T. (1953). Some British Lower Palaeozoic 
conodont faunas. Philosophical Transactions of the 
Royal Society of London B 237, 261-334. 

Roots, W.D. (1969). The geology of the area around 

Bindook, New South Wales. Unpublished BSc Hons 
Thesis, University of New South Wales, Sydney. 

Sarmiento, G., Mendez-Bedia, I., Arbizu, M. and Truyols, 
J. (1994). Early Silurian conodonts from the 
Cantabrian Zone, NW Spain. Geobios 27, 507-522. 

Satterfield, I.R. and Thompson, T.L. (1969). Phosphatic 
inarticulate brachiopods from the Bainbridge 
Formation (Silurian) of Missouri and Illinois. Journal 
of Paleontology A3, 1042-1048. 

Scheibner, E. (1973). 'Geology of the Taralga 1:100,000 
sheet 8829'. (Geological Survey of New South 
Wales: Sydney). 

Serpagli, E. and Corradini, C. (1998). New taxa of 

Kockelella (Conodonta) from late Wenlock-Ludlow 
(Silurian) of Sardinia. Giornale di Geologia 60, 79- 
83. 



Serpagli, E. and Corradini, C. (1999). Taxonomy and 
evolution of Kockelella (Conodonta) from the 
Silurian of Sardinia (Italy). Bollettino delta Societd 
Paleontologica Italiana 37, 275-298. 

Sherwin, L. (1969a). Report on fossils from the Bindook 
1:36,680 sheet (Burragorang IV 1:50,000) submitted 
by D. Roots. Unpublished Geological Survey of New 
South Wales Palaeontological Report 69/15. 

Sherwin, L. (1969b). Fossils from the Little Wombeyan 
Creek Limestone, Bindook 1:31,680 sheet 
(Burragorang IV 1:50,000). Unpublished Geological 
Survey of New South Wales Palaeontological Report 
69/17. 

Sherwin, L. (1970). Lower Palaeozoic fossils from the 
Burragorang, Taralga and Katoomba 1:100,000 
sheets. Unpublished Geological Survey of New South 
Wales Palaeontological Report 70/24. 

Sherwin, L. (1979). Late Ordovician and Late Silurian 
fossils from the Taralga district. Unpublished 
Geological Survey of New South Wales 
Palaeontological Report 79/14. 

Simpson, A.J. (1995a). Silurian conodont biosfratigraphy 
in Australia: a review and critique. Courier 
Forschungsinstitut Senckenberg 182, 325-345. 

Simpson, A.J. (1995b). Silurian conodont studies in 

eastern Ausfralia. Unpublished PhD thesis. University 
of Queensland, Brisbane. 

Simpson, A.J., Bell, K.N., Mawson, R. and Talent, 
J. A. (1993). Silurian (Ludlow) conodonts and 
foraminiferas from Cowombat, southeastern 
Australia. Memoirs of the Association of Australasian 
Palaeontologists, 15, 141-159. 

Simpson, A.J. and Talent, J.A. (1995). Silurian conodonts 
from the headwaters of the Indi (upper Murray) 
and Buchan rivers, southeastern Australia, and 
their implications. Courier Forschungsinstitut 
Senckenberg 182, 79-2 17. 

Sowerby, J. de C. (1839). Shells. In 'The Silurian System, 
part II. Organic remains' (Ed. R.I. Murchison) pp. 
579-712. (John Murray: London) 

Stauffer, C.R. (1930). Conodonts from the Decorah Shale. 
Journal of Paleontology!, 121-128. 

Stauffer, C.R. (1940). Conodonts from the Devonian 
and associated clays of Miimesota. Journal of 
Paleontology 14, 417-435. 

Strusz, D.L. and Munson, T.J. (1997). Coral assemblages 
in the Silurian of eastern Ausfralia: a rugosan 
perspective. Boletin de la Real Sociedad Espanola de 
Historia Natural Seccion Geologica 92, 3 1 1-323. 

Sweet, W.C. (1988). 'The Conodonta, morphology, 
taxonomy, paleoecology, and evolutionary history 
of a long-extinct animal phylum'. (Clarendon Press: 
New York). 

Talent, J.A., Berry, W.B.N., Packham, G., Bischoff, 
G.C.O. and Boucot, A.J. (1975). Correlation of the 
Silurian rocks of Ausfralia, New Zealand, and New 
Guinea. Geological Society of America Special Paper 
105, 1-108. 

Talent, J.A., Mawson, R., Simpson, A.J. and Brock, G.A. 
(2002). Palaeozoics of NE Queensland: Broken River 
Region: Ordovician-Carboniferous of the Townsville 



Proc. Linn. Soc. N.S.W., 127, 2006 



233 



SILURIAN BRACHIOPODS AND CONODONTS 



hinterland: Broken River and Camel Creek regions, 
Burdekin and Clark River basins. IPC2002 Post-5 
Field Excursion Guide Book. Macquarie University 
Centre for Ecostratigraphy and Palaeobiology 
Special Publication 1, 1-82. 

Temple, J.T. (1987). Early Llandovery brachiopods of 

Wales. Monograph of the Palaeontographical Society 
139, 1-137. 

Uyeno, T.T. (1990). Biostratigraphy and conodont faunas 
of Upper Ordovician through Middle Devonian 
rocks, eastern Arctic Archipelago. Geological Survey 
of Canada Bulletin ^{Sl, 1-211. 

Valentine, J.L. and Brock, G.A. (2003). Anew 

siphonotretid from the Silurian of cenfral-westem 
New South Wales, Australia. Records of the 
Australian Museum 55, lil-lAA. 

Valentine, J.L., Brock, G.A. and Molloy, RD. (2003). 
Linguliformean brachiopod faunal turnover across 
the Ireviken Event (Silurian) at Boree Creek, central- 
western New South Wales, Australia. Courier 
Forschungsinstitut Senckenberg 242, 301-327. 

von Bitter, P.H. and Ludvigsen, R. (1979). Formation 
and fiinction of protegular pitting in some North 
American acrotretid brachiopods. Palaeontology 11, 

von Eichwald, E. (1829). 'Zoologia Specialis, quam 
expositis animalibus turn vivis, tum fossilibus 
potissimum Rossiae in universum, et Poloniae in 
specie, in usum lectionum publicarum in Universitate 
Caesarea Vilnensi habendarum'. (Josephi Zawadzki: 
Vilniae). 

Walliser, O.H. (1957). Conodonten aus dem oberen 
Gotlandium Deutschlands und der Kamischen 
Alpen. Notizblatt des hessischen Landesamtes fur 
Bodenforschung, Wiesbaden 85, 28-52. 

Walliser, O.H. (1964). Conodonten des Silurs. 

Abhandlungen des Hessische Landesamtes fur 
Bodenforschung 41, 1-106. 

Williams, A. (2003). Microscopic imprints on the juvenile 
shells of Palaeozoic linguliform brachiopods. 
Palaeontology 46, 67-92. 

Wright, A.D. (1963). The fauna of the Portrane Limestone 
1 . The inarticulate brachiopods. Bulletin of the British 
Museum (Natural History), Geology 8, 221-254. 

Wright, A.D. and McClean, A.E. (1991). 

Microbrachiopods and the end-Ordovician event. 
Historical Biology 5, 123-129. 

Zhang Shun-xin and Barnes, C.R. (2002). Anew 

Llandovery (Early Silurian) conodont biozonation 
and conodonts from the Becscie, Merrimack, and 
Gun River formations, Anticosti Island, Quebec. 
Journal of Paleontology, Supplement 16, 1-46. 



234 Proc. Linn. Soc. N.S.W., 127, 2006 



Late Ordovician Faunas from the Quandialla-Marsden District, 

South-central New South Wales 

Ian G. Percival^ Yong Yi Zhen^ and John Pickett^ 

'Geological Survey of New South Wales, Department of Primary Industries, Londonderry Geoscience Centre, 

947-953 Londonderry Road, Londonderry NSW 2753, Australia; 

^ Palaeontology Section, Australian Museum, 6 College St, Sydney NSW 2010, Australia. 



Percival, I.G., Zhen, Y.Y. and Pickett, J.W. (2006). Late Ordovician faunas from the Quandialla-Marsden 
disfrict, south-central New South Wales. Proceedings of the Linnean Society of New South Wales 127, 
235-255. 

Two Late Ordovician faunas, one from shallow water limestones and the other from deep water spiculitic 
siltstones, are documented from the southern Macquarie Arc in south-central New South Wales. Limestone 
encountered in the subsurface dmng exploration drilling in the Barmedman Creek area (midway between 
Marsden and West Wyalong) yields Eastonian conodonts including Aphelognathus cf webbyi, Belodina 
compressa, Phragmodus undatus, Tasmanognathus cf borealis and Yaoxianognathusl tunguskaensis. 
Associated macrofauna includes the corals Tetradium tenue, Bajgolial cf grandis, Propora bowanensis, 
Paleofavositesl , Cystihaly sites, Halysites and Palaeophyllum, stromatoporoids Labechiella variabilis, 
Stratodictyon ozakii, Clathrodictyon cf microundulatum and Ecclimadictyon, and sponge Cliefdenella cf 
perdentata. The Jingerangle Formation, exposed between Caragabal and Quandialla, may be as young as 
Bolindian 2 on the basis of some poorly preserved graptolites. Associated nektic nautiloids and sponges 
(Hindia) represent components of Benthic Assemblage 4-5, suggesting a deep water environment. The 
limestones at Barmedman Creek, and the spiculitic clastic rocks of the Jingerangle Formation, are associated 
(although exact relationships are unclear) with two separate volcanic complexes in the Macquarie Arc. Late 
Ordovician successions exposed further north in the area west of Parkes and Forbes, where early to late 
Eastonian limestones are overlain by early Bolindian deep water sediments, provide the closest regional 
analogues to the fossiliferous sfrata documented in the paper. 

Manuscript received 13 September 2005, accepted for publication 7 December 2005 

KEYWORDS: biosfratigraphy, conodonts, corals, Macquarie Arc, nautiloids, Ordovician, palaeoec(3logy, 
sponges, stromatoporoids 



INTRODUCTION 

Late Ordovician shelly fossils documented in 
this paper are the most southerly known from the 
Junee-Narromine Volcanic Belt of the Ordovician 
Macquarie Arc in central and southern New South 
Wales (Glen et al. 1998). The study area, between 
Marsden and Quandialla, is not far from the Cowal 
Mine, presently under development near Lake Cowal 
(Fig. 1). The impetus of mineralisation potential in 
the area led to a recent drilling program by Newcrest 
Exploration at the company's Marsden prospect at 
Barmedman Creek, 25 km northeast of West Wyalong, 
disclosing the presence of a previously unsuspected 
limestone that has proven to be of Eastonian age. A 
moderately diverse Late Ordovician fauna has long 
been known from the Jingerangle Formation in the 



Quandialla district, approximately midway between 
West Wyalong and Grenfell (Fig. 1), but has remained 
undescribed until now. This fauna is somewhat 
younger than that obtained from the limestone at 
Barmedman Creek. Together, these Late Ordovician 
fossils provide important palaeontological constraints 
in an area of poor or hidden outcrop, and enable 
correlation with better-known and well-exposed 
successions further north along this belt in the region 
west of Parkes. 



STRATIGRAPHIC SETTING AND 
BIOSTRATIGRAPHIC DETERMINATIONS 

Unnamed limestone, Marsden prospect at 
Barmedman Creek 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 



Newcrest Exploration encountered limestone in 
three cored drill holes at their Marsden prospect in 
the Barmedman Creek area, located in the vicinity 
of the Mid Western Highway 22 km west of Grenfell 
(Fig. 1). Cores from two of these holes, DDMN 042 
and ACDMN 043, were extensively sampled for 
conodonts and macro fossils. Logs of these cored 
intersections (kindly provided by Irvine Hay of 
Newcrest) are shown in Figure 2. The remaining 
hole, ACDMN 045, was spot-sampled for macro- 
and microfossils over the depth interval 273-276.5 
m. This interval yielded sparse conodonts (sample 
C2077) and one coral from 276 m. Total thickness 
of the limestone cannot be accurately determined 
due to the prevalence of faulting in the other two 
cores. In ACDMN 043, a 15 m-thick zone of fauh 
gouge cuts through the middle of a limestone interval 
approximately 42 m in thickness. This faulted zone 
coincides with a series of intermixed and out-of- 
sequence biostratigraphic determinations (Fig. 2). The 
33 m of apparently continuous limestone intersected 
in DDMN 042 is faulted at its top. 

Composition and age significance of the conodont 
fauna 

Limestone intersected in the Newcrest drilling 
program yielded 96 identifiable conodont specimens 
recovered from 12 samples. Sample weights varied 
from 900 g to 3.9 kg (average 1.7 kg), with the larger 
samples being obtained from intersections of several 
metres. Limestone samples were dissolved in dilute 
acetic acid and separated using sodium polytungstate. 
The conodonts, illustrated in Figs 3-4, are referrable 
to nine species (Table 1) which indicate a Late 
Ordovician (Eastonian) age. 

Species of biostratigraphic significance include 
Belodina compressa, Plectodina tenuis? and 
Phragmodus undatus. All three are zonal index species 
of the North American Mid-continent biostratigraphic 
scheme, though it has been recognised that 
differences in local ranges and relative abundances 
present difficulties in precisely correlating with the 
North American zonation (Zhen and Webby 1995, 
Zhen et al. 1999). Belodina compressa first appears 
in NSW in the late Gisbomian upper part of the 
Wahringa Limestone Member (Zhen et al. 2004) and 
was replaced by B. confluens in limestones of early 
Eastonian age throughout the Macquarie Arc. Though 
mostly confined to slightly younger (Ea2-3) horizons 
where previously recorded in these limestones, 
Phragmodus undatus also is rarely present within 
the lower Billabong Creek Limestone (Gisbomian 
age) in the vicinity of Gurmingbland, northward 
along the Junee-Narromine Volcanic Belt (Pickett 



and Percival 2001, appendix 1). Plectodina tenuis"?, 
only tentatively identified in the Marsden core from 
a couple of elements, is elsewhere in NSW restricted 
to early Eastonian (Eal-2) strata. Co-occurrence 
of P. undatus with B. compressa and P. tenuis? in 
sample C2077 (273-276 m in borehole ACDMN 045) 
therefore most likely implies a basal Eastonian age 
for this level. 

Thepresence of Yaoxianognathusl tunguskaensis, 
Aphelognathus cf. webbyi and Tasmanognathus cf. 
borealis in the assemblage also supports an Eastonian 
age assignment. Yaoxianognathusl tunguskaensis is 
widely distributed in limestones of this age throughout 
the Macquarie Arc. Aphelognathus webbyi is common 
in the early Eastonian Fossil Hill Limestone of the 
Cliefden Caves Limestone Group (Savage 1990, 
Zhen and Webby 1995). Tasmanognathus borealis 
was recorded from the Yiaoxian Formation (mid- 
early Eastonian age equivalent) of North China, 
where it is associated with Phragmodus undatus and 
Taoqupognathus blandus (An and Zheng 1990). 

Apparently absent from the fauna are any 
examples of Taoqupognathus, species of which 
are biostratigraphically significant in Eastonian 
limestones in central NSW and China (Zhen et al. 
1999, Zhen 2001). Another characteristic feature of 
the Barmedman Creek limestone is the occurrence of 
Rhipidognathus, which has not previously been noted 
from NSW 

Coral and stromatoporoid assemblages 

All three of the Newcrest boreholes at the 
Marsden prospect yielded corals, and one also 
included sfromatoporoids. These are illustrated in 
Figs 5-9, and their occurrences are detailed in Table 
2. 

Three samples from borehole DDMN 042 
all yielded a single species of coral, Tetradium 
tenue, which is known only from the Hillophyllum- 
Tetradium-Rosenella Assemblage Zone (Pickett 
and Percival 2001) of Eal age. The material from 
borehole ACDMN 045 is poorly preserved, with 
only a provisional determination of Tetradium? sp., 
implying a generalised early Eastonian age, which is 
in accord with the conodont-based age from sample 
C2077 (273-276.5 m in ACDMN 045) previously 
mentioned. 

The most abundant material is from borehole 
ACDMN 043, which, in addition to forms already 
known from NSW, also includes a number of unusual 
occurrences. Stratodictyon ozakii was previously 
recorded only from the Hillophyllum-Tetradium- 
Rosenella Assemblage Zone, but is here associated 
with forms characteristic of younger levels. 



236 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 




5200CI0nnE 



6240000mN — 



620000mE 

I 



REFERENCE 

Jingerangle Formation subcrop inferred 

Jingerangle Formation outcrop 
Thrust Fault inferred 



Roads 



Waterways 



10 



20 



30 km 



Figure 1. Locality map of south-central New South Wales showing places mentioned in the text. 
Simplified geological data, including location of Marsden prospect drill sites, the regional thrust 
fault, and outcrop and subsurface extent of the Jingerangle Formation (incorporating the Cur- 
rumburrama volcanics) are derived from the Forbes 1:250 000 Geological Map (second edition). 



Proc. Linn. Soc. N.S.W., 127, 2006 



237 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 



ACDMN 043 



DDMN 042 



C2102 
C2103 H 



Ea3? FHP? 



. C2073 
Ea1-2 



Ea3? FHP? C2074 
HTR or PEC Ea1-2 



C2075 
HTR Eal 



C2104 
C2105 ~-\ 
C2106 



PEC/FHP C2076 
Ea2 PEC 




- 238 7m 



Volcaniclastic 
sandstone & 
conglomerate 

Volcaniclastic 
sandstone & 
conglomerate 



C2090 
C2091 
C2092 

Eal C2071 



C2093 
HTR Ea1 C2094 

C2095 - 

Ea1?C2096 - 

r 

HTR Ea1? C2072 \ 



C2097 
C2098 
C2099 -t 



HTR = HiUophyllum-Tetradium-RoseneUa Assemblage Zone Ea1/2 C2100 
PEC = Propora-Ecclimadictyon-Cliefdenella Assemblage Zone >^^ i u i 

FHP = Favistina-Halysites-Plasmoporella Assemblage Zone 



385m 



50m 




100m - 



150m 



200m 



250m 



300m — 




450m 



o 

T3 
O 
N 

c: 
o 



2005 10 0203 



Figure 2. Diagrammatic representation of major litliologies intersected in Newcrest borelioles DDMN 
042 and ACDMN 043 in tiie Marsden prospect, witli enlargement of limestone-dominated intervals to 
show sampled horizons and faunas recovered. GL = ground level; F = fault; Ea = Eastonian Stage, with 
subdivisions Eal (oldest), Ea2, Ea3 (youngest). See text for discussion of age relationships of macrofos- 
sil Assemblage Zones (defined in Pickett and Percival 2001) and conodonts (C samples). 



238 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 




Figure 3. SEM photographs of conodonts from Eastonian limestone in core, Marsden prospect 
at Barmedman Creek; scale bars 100 \im. A-D, Belodina compressa (Branson and Mehl, 1933). A, B, 
grandiform elements, from C2071, A, MMMC4122, outer lateral view; B, MMMC4123, inner lateral 
view; C, D compressiform elements, from C2072, inner lateral views, C, MMMC4124, D, MMMC4125. 
E-H, Panderodus gracilis (Branson and Mehl, 1933). E, F, tortiform element, MMMC4126, from C2071, 
E, outer lateral view, F, inner lateral view; G, falciform element, MMMC4127, from C2095, outer 
lateral view; H, falciform element, MMMC4128, from C2096, outer lateral view. I, J, Panderodus sp. I, 
"b" element, MMMC4129, from C2072, outer lateral view; J, "a" element, MMMC4130, from C2073, 
outer lateral view. K, L, Plectodina tenuis? (Branson and Mehl, 1933). K, S element, MMMC4133, from 
C2077, posterior view; L, Pb element, MMMC4132, from C2071, inner lateral view. M, Tasmanognathus 
sp. cf. T. borealis An, in An et al., 1985. Pa element, MMMC4131, from C2075, inner lateral view. 



Proc. Linn. Soc. N.S.W., 127, 2006 



239 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 




Figure 4. SEM photographs of conodonts from Eastonian limestone in core, Marsden prospect at 
Barmedman Creelc; scale bars 100 ^m. A-E, Phragmodus undatus Branson and Mehl, 1933. A, Pa 
element, MMMC4134, from C2100, anterior view; B, Sc element, MMMC4135, from C2077, in- 
ner lateral view; C, Sc element, MMMC4136, from C2077, outer lateral view; D, Sb element, 
MMMC4137, from C2094, outer lateral view; E, Sb element, MMMC4138, from C2094, inner later- 
al view. F-I, Rhipidognathus sp. F, Sb element, MMMC4139, from C2075, posterior view; G, Pa el- 
ement, MMMC4140, from C2094, inner lateral view; H, Pa element, MMMC4141, from C2072, in- 
ner lateral view; I, Pb element, MMMC4142, from C2073, inner lateral view. J, Aphelognathus sp. 
cf. A. webbyi Savage, 1990. Pa element, MMMC4143, from C2075, outer lateral view. K-N, Yaoxi- 
anognathusl tunguskaensis (Moskalenko, 1973). K, Sd? Element, MMMC4144, from C2074, in- 
ner lateral view; L, Sb element, MMMC4145, from C2075, inner lateral view; M, Sc element, 

MMMC4146, from C2073, inner lateral view; N, M element, MMMC4147, from C2074, posterior view. 



240 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 



Table 1. Distribution of conodonts in Newcrest boreholes, Marsden prospect, Barmedman Creek; 
C2077 from ACDMN 045, 273-276.5 m depth; for details of other samples see Appendix. 



CONODONT TAXA 


SAMPLES 




o 
O 


CO 
O 


CD 
CNJ 
O 


o 

CNI 

O 


CM 

O 


CO 
CT> 
CD 
CNI 
O 


C3 
CM 
O 


o 


CD 

g 
CM 

o 


CJ> 

o 

3 


C3 
C3 

3 


.2 


Apholognathus cf. webbyi 






1 


1 


7 
















9 


Belodina compressa 




3 


2 






3 


2 








2 






12 


Panderodus gracilis 




5 


6 


1 


8 


8 


5 






7 


2 


1 


1 


44 


Panderodus sp. 






1 


1 


5 


1 
















8 


Phragmodus undatus 














2 




2 








1 


5 


Plectodina tenuis ? 




1 










1 














2 


Rhipidognathus sp. 






1 


1 


1 


1 






1 










5 


Tasmanognathus cf. 


borealis 










1 










1 






2 


Yaoxianognathus tunguskaensis 






2 


3 


3 




1 












9 


Total 


9 


10 


6 


18 


24 


10 


1 


3 


7 


5 


1 


2 


96 



Labechiella variabilis, Bajgolial cf. grandis, the 
heliolitids and Ecclimadictyon are typical of the 
Propora-Ecclimadictyon-Cliefdenella Assemblage 
Zone, of Ea2 age, while Paleofavosites and halysitids 
are only known fi-om the next-youngest (Ea3-4) 
Favistina-Halysites-Plasmoporella Assemblage Zone 
(Pickett and Percival 2001). Cystihalysites has so far 
not been reported from strata younger than Early 
Silurian, so its occurrence here, apart from being the 
first report of the genus in Australia, is outside its 
known range. 

Synthesis: age of the Barmedman Creek limestone 

Regional biostratigraphic zonation of Upper 
Ordovician limestones within the Macquarie Arc of 
central NSW is well-established, based on integrated 
macrofaunal andmicrofaunal assemblages. Diagnostic 
taxa from three of the coral-stromatoporoid faunas 
first recognised by Webby (1969), updated by Webby 
et al. (1997) and more recently formalised by Pickett 
and Percival (2001), are identified in limestone fi-om 
two of the cored holes. Though the conodont faunas 
recovered lack Taoqupognathus, a key component of 
the local zonation (Zhen 2001), sufficient associated 
species are present to confirm the ages of most 
individual samples. 

When plotted against the log of the Newcrest 
drill hole DDMN 042 (Fig. 2), occurrence of a coral 
species restricted to the Hillophyllum-Tetradium- 
Rosenella Assemblage Zone (Pickett and Percival 
2001) is consistent with presence of early Eastonian 
(Eal) age conodonts. Indeed, the identification of 
Belodina compressa in three samples from this core, 
which are closely associated with the levels that 
produced the coral Tetradium tenue, implies a basal 
Eastonian age. 

The sequence of macrofaunal assemblages and 
conodonts in ACDMN 043 is more problematic, and 



only makes sense when details fi-om the lithology log 
are integrated with the palaeontological sampling 
(Fig. 2). Samples fi^om the deepest limestone 
intersected (229 and 232 m) are consistent with 
an Ea2 age, based on presence of a diverse suite 
of corals and stromatoporoids of the Propora- 
Ecdimadictyon-Cliefdenella Assemblage Zone. 
Above a barren interval, samples firom 189-193 m 
yield both conodonts and stromatoporoids indicative 
of an earlier, Eal, age. This succession is at variance 
with what would be expected, and may imply the 
presence either of a fault (unrecognised in the core) 
or an overturned sequence. Within the interval 174- 
182 m, ages are mixed in a zone identified on the 
log as extensively faulted. The lowermost sample 
from this faulted zone contains sponges (including 
stromatoporoids) consistent with an early Eastonian 
age (Eal -2), that is overlain by limestone with 
sponges and corals (including halysitids) suggesting 
a younger, Ea3, age. However, conodonts firom a 
sample extending over the interval 179.5-1 83.9 m that 
includes the aforementioned macrofossil assemblages, 
are definitely of Eal -2 age — confirming structural 
interleaving of fault slices. Samples firom shallower 
depths exhibit a similar intermixing of ages, with 
macrofossils fi-om 173 and 174 m characteristic of the 
Favistina-Halysites-Plasmoporella Assemblage Zone 
(Ea3) associated with Eal -2 conodonts from sample 
C2073. The two highest samples unfortunately yield 
no biostratigraphically useful information. 

The Marsden prospect is located on the western 
(hangingwall) block of a major regional thrust fault 
(Fig. 1). Although the faulting has disrupted the 
normal biostratigraphic succession in the drill core, it 
has had the fortunate effect of demonstrating — even 
in a relatively short intersection of limestone — that 
the limestone at Barmedman Creek commenced 
deposition in basal Eastonian time and continued 



Proc. Linn. Soc. N.S.W., 127, 2006 



241 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 




Figure 5. Tabulate corals from Eastonian limestone in core, Marsden prospect at Barmedman 
Creek. Scale bar shown in photos (A), (C) and (F) represents 10 mm. Scale bar in (B), (D) and (E) 
represents 1 mm. A, B, Halysites sp. from 173 m in ACDMN 043; A, transverse section; B, enlarge- 
ment to show macro- and microcorallites. C, D, E, Cystihalysites sp. from 179-180 m in ACD- 
MN 043; C, oblique transverse and longitudinal section; D, enlargement of the lower left corner of 
B, showing cystose coenenchymal tubules; E, detail of upper right corner of D, clearly displaying 
cyst. F, Bajgolia? cf. grandis Webby, 1977, oblique longitudinal section from 229 m in ACDMN 043. 



242 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 




Figure 6. Corals from Eastonian limestone in core, Marsden propect at Barmedman Creek. Scale bar shown 
in photo (B) represents 10 mm and applies also to photos (A) and (C-F). A, B, Tetradium tenue Webby and 
Semenuik, 1971, longitudinal and transverse sections, from 399 m in DDMN 042. C, D, Tetradium? sp., ob- 
lique transverse section from 406 m in DDMN 042, and longitudinal section from 276 m in ACDMN 045. E, F, 
Bajgolial ctgrandisWebhy, 1977, transverse and oblique longitudinal sections from 232 m in ACDMN 043. 



into the late Eastonian (Ea3). This age determination 
is significant in correlating the succession with 
limestones of Late Ordovician age elsewhere in the 
Macquarie Arc. 

Jingerangle Formation 

In the southernmost area of the Forbes 1 :250 
000 map sheet, south of the Mid Western Highway 
between Grenfell and West Wyalong, Ordovician 



formations are mostly hidden beneath alluvial cover 
of Cainozoic age. Very few outcrops stand above the 
plain, and fossiliferous strata are almost absent. The 
sole exception is the Jingerangle Formation which is 
best exposed in two road aggregate quarries in the 
vicinity of Gibber Trig (OR 560200mE 6244600mN, 
Marsden (8430 II and III) 1:50 000 sheet). This low 
hill is located immediately south of the Jingerangle 
State Forest, which is itself situated south of the 



Proc. Linn. Soc. N.S.W., 127, 2006 



243 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 




Figure 7. Tabulate and rugosan corals from Eastonian limestone in core, Marsden propect at 
Barmedman Creek. Scale bar represents 10 mm. A, B, Paleofavositesl sp., transverse and lon- 
gitudinal sections, from 174 m in ACDMN 043. C, D, E, Propora bowanensis Hill, 1957, lon- 
gitudinal, transverse and oblique sections from 229 m in ACDMN 043; note also trans- 
verse section through partial corallite of Palaeophyllum sp. in upper right corner of photo (E). 



Mid Western Highway between Grenfell and West 
Wyalong about 37 km east of the latter town (Fig. 1). 
Further locality details are given by Lyons and Wallace 
(1999). Warren et al. (1995) named the unit and 
provided its formal description [despite their assertion 
that Bowman (1976) first described the Jingerangle 
Formation, no such name or distinguishing description 
appears either on the Forbes 1:250 000 metallogenic 
map or in the accompanying explanatory notes]. An 
up-to-date description of the Jingerangle Formation 
appears in the Explanatory Notes to the Forbes 1 :250 
000 Geological Sheet, 2"'' edition (Percival and Lyons 
2000). 

The Jingerangle Formation is significant 
in containing the youngest, most diverse. Late 
Ordovician shelly macrofauna in central NSW, near 
the southernmost extent of outcrop of sediments 



associated with the Junee-Narromine Volcanic Belt. 
In this belt, only the lower section of the Cotton 
Formation (Sherwin 1973, Sherwin et al. 1987), on 
trend to the northeast just west of Forbes, appears to 
be of broadly comparable (Bolindian) age. 

Lithologies in the lower Jingerangle Formation, 
exposed in the working road base quarry, mostly 
consist of a succession of thinly bedded siltstones 
and mudstones, the latter generally weathered into 
multicoloured clays (pink and white) and orange- 
brown ochres. The siltstones are more resistant as 
they are largely composed of sponge spicules, which 
provide a tightly interlocking meshwork of silica. 
Fresh recently exposed material is relatively dense 
and mostly dark grey in colour, but natural outcrops 
are weathered to a lighter biscuit-like texture, of grey- 
white appearance. The other major sediment tj^e in 



244 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 




Figure 8. Sponges, including stromatoporoids, from Eastonian limestone in core, Marsden prospect 
at Barmedman Creek. Scale bar represents 10 mm. A, B, Cliefdenella cf. perdentata Webby and Mor- 
ris, 1976, transverse and longitudinal sections, from 181 m in ACDMN 043. C, D, Labechiella vari- 
abilis (Yabe and Sugiyama, 1930), longitudinal and transverse sections from 192 m in ACDMN 043. 



Proc. Linn. Soc. N.S.W., 127, 2006 



245 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 





\ "i'^i" ; 


"' ■• ^ 


^r^**^^ 


' '■ -=?^"" 


* ' '"■['■ ■/, \% 


..-.--'^.^^ 






%^^\ :::,'' H.^ 


iS^- 


.^•"^ 


.^.'' " r. lisl?;--- 


Q,: 




Figure 9. Stromatoporoids from Eastonian limestone, Marsden prospect at Barmedman Creek. Scale bar 
beneath (C) represents 10 mm and applies to (A), (C), (E) and (F); scale bars in photos (B) and (D) represent 
1 mm. A-D,Stratodictyon ozakiiWebhy, 1969, from 182 m in ACDMN 043; longitudinal (A) and transverse 
(C) sections, with respective enlargements (B) and (D); note columns spanning up to seven laminae in lower 
left corner of (B), and astrorhizal canal in upper centre of (D). E, Clathrodictyon cf. microundulatum Nestor, 
1964, from 229 m in ACDMN 043. F, Ecdimadictyon sp., longitudinal section, from 229 m in ACDMN 043. 



246 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 



Table 2. Distribution of coral and sponge species in Newcrest boreholes, Marsden prospect, Barmed- 
man Creek. 



Borehole 


Depth (m) 


Assemblage 


DDMN 042 


399 


Tetradium tenue Webby & Semeniuk, 1971 


405 


Tetradium tenue Webby & Semeniuk, 1971 


406 


Tetradium tenue Webby & Semeniuk, 1971; Tetradium sp. 


ACDMN 043 


173 


Halysites sp. 


174 


Paleofavosites? sp. 


179-180 


Cystihalysites sp. 


181 


Cliefdenella cf. perdentata Webby & Morris, 1976 


182 


Stratodictyon ozakii Webby, 1 969; Cliefdenella sp. 


192 


Labechiella variabilis (Yabe & Sugiyama, 1930) 


229 


Bajgolia cf. grancf/s Webby, 1977; Palaeophyllum sp.; Propora 
bowanensis Hill, 1957; heliolitid indet.; Clattirodictyon cf. 
microundulatum Nestor, 1 964; Ecclimadictyon sp. 


232 


Bajgolia? cf. grandis Webby, 1977 


ACDMN 045 


275 


indeterminate 


276 


Tetradium? sp. 



the quarry occurs in stratigraphically higher beds, 
composed of coarser silts to fine sands that are partly 
silicified. These strata are distinguished by the high 
concentration of siliceous sponges (predominantly the 
spheroidal Hindia) which are clustered on the surface 
of beds. Thin maroon-coloured medium to coarse 
grained sandstone layers are rarely interspersed in the 
siltstone succession towards the basal beds exposed 
in the working quarry. Most beds at this locality dip 
towards the east at variable angles, from nearly zero 
to about 30 degrees. Only an estimated 20-30 metres 
of continuous section is exposed in the floor of the 
working quarry; the true thickness of the formation 
is considerably in excess of this, but is urmieasurable 
due to structural complexity. Siltstone beds on the 
western side of this quarry are erratic in trend, but 
generally dip towards the southwest at low angles. 
In the wall and floor of the disused quarry to the 
south, folding and associated faulting is particularly 
prominent. 

Fossils from the Gibber Trig outcrop, identified 
by K. Sherrard, were first referred to by Wynn (1961), 
with this information republished by Moye et al. (in 
Packham 1969, p. 98). Subsequent unpublished reports 
on the faunal assemblage fi-om these outcrops were 
provided by Sherwin (1982, 1985), Pickett (1986) 
and Percival (1999). Faunal lists fi-om the earlier 
of these reports were subsequently published in the 
palaeontological appendix to the Cootamundra 1 :250 
000 Geological Sheet Explanatory Notes (Warren 
et al. 1995). Percival's (1999) identifications were 
incorporated into the Forbes 1:250 000 Geological 
Sheet Explanatory Notes (Lyons et al. 2000). With 



the exception of a nautiloid depicted by Percival and 
Lyons (2000) (here re-illustrated in Figure lOE), 
none of the fauna has previously been illustrated or 
described. 

The graptolite assemblage indicates species 
that range in age from middle Eastonian to middle 
Bolindian; they include Dicellograptus gravis Keble 
and Harris, Dicellograptus ornatus (EUes and Wood), 
Normalograptus angustus (Pemer), Orthograptus ex. 
gr. amplexicaulis (J. Hall), together with Ptilograptus 
sp., and an vinidentified climacograptid. Overlap of 
the published ranges (VandenBerg and Cooper 1992) 
suggests an early Bolindian (Bo 2) age is most probable 
for the formation. Coiled tarphyceratid nautiloids, 
referred to Discocerasl sp. and a large indeterminate 
genus, are the most spectacular component of the 
shelly fauna (Fig. 10), but cannot be precisely 
identified due to their preservation mainly as external 
moulds. Other faunal elements include indeterminate 
cyrtoconic brevicone and orthoconic nautiloids, a 
small dalmanelloid? and large multicostate orthide? 
brachiopods, and the siliceous sponges previously 
mentioned. 

Palaeoecological interpretation 

The association in the Jingerangle Formation of 
graptolites, nautiloids of nektic habit (particularly the 
proliferation of tarphyceratids, which are thought to 
have been strong swimmers), and lithistid sponges 
is interpreted to indicate deep water environments at 
depths typical of Benthic Assemblage 4-5 (perhaps 50- 
200 m). A somewhat comparable faunal association 
is present in the basal Malongulli Formation in the 



Proc. Linn. Soc. N.S.W., 127, 2006 



247 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 




248 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y. Y. ZHEN AND J.PICKETT 



Cliefden Caves area between Orange and Cowra 
(Webby 1992, Percival and Webby 1996). Here, a 
diverse suite of sponges (including Hindia) populated 
the periplatformal zone, between the shelf edge and the 
deep basin (Rigby and Webby 1988). The Malongulli 
sponge assemblage was subsequently dislodged 
as debris flows or slumps into the lower slope and 
basinal sediments (equivalent to Benthic Assemblage 
6), which are largely comprised of spiculitic sihstones 
with a faunal association of graptolites, trilobites, 
and diminutive lingulate and plectambonitoid 
brachiopods. In the case of the Jingerangle Formation, 
the Hindia-diom\mX.Qd fauna is preserved in laminated 
sediments that are not slumped and are interpreted to 
have formed in situ. Fauna in the Bolindian section 
of the Cotton Formation consists only of graptolites, 
orthoconic nautiloids, ostracodes (Sherwin 1973) 
and the lingulate brachiopod Paterula (Percival 
1978); presumably these sediments were deposited 
at depths slightly greater than that interpreted for the 
Jingerangle Formation. 



REGIONAL CORRELATION 

Volcanic and intrusive host rocks of the Marsden 
copper-gold prospect belong to the Cowal volcanic 
complex, the geology of which is known only from 
exploratory drilling (Miles and Brooker 1998, 
Downes and Burton 1999). Beneath Lake Cowal 
this complex consists of calc-alkaline to shoshonitic 
volcanics and associated sedimentary rocks, including 
volcaniclastics, mass-flow deposits, and laminated 
mudstones and siltstones of deeper water origin. This 
succession is apparently older than early Darriwilian 
age, as it is intruded by diorites and granodiorites, 
including one dated C^Ar/^'Ar) at 465.7 ± 1 Ma (Miles 
and Brooker 1998). Rocks of the Cowal volcanic 
complex could therefore have formed the basement 
on which the Eastonian limestones (not recognised 
despite extensive exploratory drilling at Lake 
Cowal) accumulated in shallow water environments. 



Stratigraphic relationships in the cored holes fi"om 
the Marsden prospect are not clear due to structural 
complications. Newcrest DDMN 042 intersected 
approximately 245 m of monzodiorite above the 
Eastonian limestone, with evidence from the core log 
that these units are fault-juxtaposed. Beneath 1 20 m of 
regolith (an average thickness for this area), ACDMN 
043 passed through 20 m of volcaniclastic sandstone 
and siltstone (undated) before intersecting limestone 
that continued to the bottom of the hole. 

In an adjacent tectonic block separated from the 
Lake Cowal-Marsden region by a major thrust fault 
(Fig. 1), the Jingerangle Formation is also associated 
with igneous rocks, known as the Currumburrama 
volcanics. Here, however, relationships are even 
more obscured by the fact that this buried igneous 
complex has thus far only been recognised on the 
basis of its distinctive geophysical response. Age 
and composition of the Currumburrama volcanics is 
unknown, and their stratigraphic position relative to 
the Jingerangle Formation is uncertain. 

Thus the only significant information to 
assist regional correlation with other areas of the 
Macquarie Arc derives fi'om the fossiliferous rocks 
documented in this study. The Eastonian limestones 
from the Marsden prospect contain some macrofossils 
and conodonts that have not previously been 
recognised in the Junee-Narromine Volcanic Belt. 
For example, the coral Tetradium tenue, prominent 
in DDMN 042, is elsewhere known only from the 
Daylesford Limestone of the Bowan Park Group on 
the western flank of the Molong Volcanic Belt. Such 
differences are probably environmentally controlled. 
Overall, the Barmedman Creek limestones most 
closely correspond to the succession in the vicinity 
of Gunningbland (west of Parkes), through the upper 
part of the Billabong Creek Limestone (Eal-2) and 
into the overlying Gunningbland Formation which 
includes intermittent limestones of Ea3 age (Pickett 
and Percival 2001). The Gunningbland area lies 90 
km to the northeast of the Marsden prospect, along 
the trend of the Junee-Narromine Volcanic Belt (Fig. 



Figure 10 LEFT. Fossils from the Jingerangle Formation. Scale bars represent 10 mm. 
A-B, indeterminate dalmanelloid brachiopod. A, dorsal valve internal mould, and B, external mould of 
same individual MMF36611a-b, from roadbase quarry, immediately south of Jingerangle State Forest. C, 
natural cross section oi Hindia sphaeroidalis Duncan, 1879, MMF36600, from roadbase quarry, immedi- 
ately south of Jingerangle State Forest. D-J, nautiloids from Jingerangle Formation collected from scree 
on hill behind "Bland Farm" homestead; original specimens in possession of landholder. D, weathered 
profile of large indeterminate orthocone. E, Discoceras? sp., latex impression from external mould. F, la- 
tex impression from external mould of micromorphic or juvenile individual of indeterminate tarphycer- 
atid, G-I, large indeterminate tarphyceratid; G, latex replica of internal mould showing septa and living 
chamber. H, latex impression from partial external mould. I, latex impression from external mould. J, 
latex impression from external mould of indeterminate cyrtoconic brevicone, living chamber uppermost. 



Proc. Linn. Soc. N.S.W., 127, 2006 



249 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 



1). Also on this trend to the west of Forbes, about 75 
km northeast of the Marsden prospect, are outcrops 
of the lower (Bolindian age) Cotton Formation 
which — as previously observed — is the closest 
analogue to the Jingerangle Formation in terms of 
lithology, depositional environment and age. It would 
be reasonable, given the relatively well-documented 
Late Ordovician succession in the Forbes-Parkes 
region, to interpret the Jingerangle Formation as a 
similarly widespread deep water unit overlying the 
older limestone and volcanic rocks encountered in 
the Barmedman Creek area. However, as these units 
are presently separated by a major thrust fault, this 
relationship remains conjectural. 



TAXONOMIC NOTES 

Responsibility for palaeontological discussion 
is indicated for each phylum. Some taxa have been 
documented by illustration only where material 
is insufficient for comment or where species are 
well known. Specimens are catalogued in the 
Palaeontological Collection of the Geological Survey 
of NSW (prefix MMF for macrofossils, MMMC for 
conodonts), housed in the Londonderry Geoscience 
Centre in western Sydney. 

Conodonts [Zhen] 

Only grandiform and compressiform elements of 
B. compressa were recovered from the Barmedman 
Creek samples (Fig. 3A-D). Both illustrated 
specimens of compressiform elements show a straight 
section of anterior margin near the antero-basal 
comer, recognised as the most distinctive character to 
differentiate B. compressa fi-om the stratigraphically 
slightly younger B. confluens (Zhen et al. 2004). 

Two specimens are doubtfully referred to 
Plectodina tenuis. One, identified as the Pb element 
(Fig. 3L), bears anterior and posterior processes more 
or less equal in length, but has a shorter posterior 
process in comparison with elements reported from the 
Cliefden Caves Limestone Group (Zhen and Webby 
1995). A Pb element of P. tenuis with a similarly short 
posterior process was also reported from the Late 
Ordovician Vaureal Formation of Anticosti Island, 
Quebec (Nowlan and Barnes 1981, pi. 4, fig. 20). 

Rhipidognathus sp. (Fig. 4F-I) with only Pa, 
Pb, and Sb elements recovered may represent a new 
species. Both Pa and Pb elements are angulate with 
denticulate anterior and posterior processes, but the 
Pb element bears a large robust cusp (Fig. 41), whereas 
the cusp in the Pa element is indistinguishable fi^om 
adjacent denticles (Fig. 4G, H). The Sb element 



is palmate digyrate, slightly asymmetrical with a 
prominent basal tongue on the anterior face that 
extends below basal margin (Fig. 4F). 

Corals [Pickett] 

The halysitids represent the most unusual 
elements of the coral fauna fi-om the Barmedman 
Creek limestone. The form determined as Haly sites 
sp. (Fig. 5A, B) is definitely not the same as the only 
other true Halysites known fi-om the Ordovician, H. 
praecedens Webby and Semeniuk 1969; that species 
has subrounded corallites 1.2 - 2.0 mm long, with 
tabulae at 6 - 7 in 5 mm, whereas in the present 
material the corallites are elongate and the palisades 
only slightly wider at their widest point, and the 
tabulae are much more frequent: up to 8 in 2 mm. 
The Cystihalysites has clearly developed cystose 
coenenchymal tubules (Fig. 5C-E), but the material is 
too scant for proper description. 

The poorly preserved specimen designated 
Tetradiuml sp. (Fig. 6C), from 406 m depth in 
borehole DDMN 042 at Barmedman Creek, is cerioid 
or subcerioid in habit, with complete, distant tabulae, 
a double-layered wall, and apparently without either 
mural pores or septa. The absence of mural pores and 
presence of a double-layered wall suggest an early 
stauriid rugosan such as Favistina or Crenulites, but 
both of these have septa, and the tabulae of Crenulites 
are distinctively shaped. Foerstephyllum is also ruled 
out by the absence of septa. Tetradiuml sp. may 
be related to a form referred to Tetradium sp. A by 
Webby and Semeniuk (1971, pi. 17, figs 4, 5), that 
has inconspicuous septa, angular corallites, and thin 
walls. Cerioid or sub-cerioid corals known firom near 
this level in NSW include a variety of auloporoid 
forms described by Webby (1977). However, none of 
these appears to show the thin walls of the present 
specimens. 

The material of Propora bowanensis Hill, 1957 
(Fig. 7C-E) falls within the variation reported for this 
species by Webby and Kruse (1984), though rather 
more consistently with the Heliolites end of the 
spectrum. 

Sponges [Pickett] 

The Geological Siirvey of NSW collections 
include a large number (MMF 29519-29538, 36592- 
36608) of well preserved individuals of Hindia 
sphaeroidalis Duncan, 1879 fi-om the Jingerangle 
Formation (Pickett 1986). The largest of these is 
illustrated (Fig. IOC). This species was also reported 
by Rigby and Webby (1988) firom three of their four 
horizons in the MalonguUi Formation near Cliefden 
Caves, and additionally from Late Ordovician strata 



250 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 



at "Currajong Park", Gumiingbland, west of Parkes. 

The stromatoporoid identified as Labechiella 
variabilis (Yabe and Sugiyama, 1930) (Fig. 8C, 
D) has pillars up to 0.6 mm in diameter, somewhat 
stouter than those reported for this species by Webby 
(1969). Stratodictyon cf ozakii Webby, 1969 from 
1 82 m in ACDMN 043 shows well-developed pillars, 
stouter than the laminae, which may cross up to seven 
laminae (Fig. 9B). The transverse section (Fig. 9C, D) 
shows a distinct astrorhizal canal. Stromatoporoids 
from the 229 m level in this drill hole include 
Ecclimadictyon sp. (Fig. 9F) and Clathrodictyon cf 
microundulatum Nestor, 1964. The latter, represented 
by a single longitudinal section (Fig. 9E), accords 
well with the specimen figured by Webby (1969, 
pi. 127, fig. 3). As noted by Webby, his specimens 
of C. cf. microundulatum are associated with, and in 
some cases resemble, Ecclimadictyon. Co-occurrence 
of these forms in the Marsden prospect core is 
reminiscent of this situation. 

Brachiopods [Percival] 

Brachiopods are uncommon in the Jingerangle 
Formation. Most specimens are poorly preserved 
external impressions of weakly biconvex multicostate 
valves with wide hingelines, probably referable to an 
indeterminate orthide? 

The only example presently known from the 
Jingerangle Formation of a small dalmanelloid?, 
represented by a dorsal valve, is illustrated (Fig. lOA- 
B) as it is better preserved than the other brachiopods. 
The valve is transversely quadrate in outline, of low 
convexity with a narrow median sulcus; the distinctive 
ornament is interpreted from the exterior mould 
as comprising closely spaced coarse exopunctae 
regularly distributed between the fine multicostellae. 
Internally, the cardinalia consist of a simple blade- 
like cardinal process, with rod-shaped brachiophores 
apparently supported by delicate fiilcral plates. 
A median septum is not developed, although the 
narrow median sulcus is ventrally directed to mimic a 
raised ridge. Without details of the ventral valve it is 
impossible to assign this specimen at family level, but 
general affinities with the paurorthids are suggested. 
No comparable shells have been noted elsewhere in 
the Late Ordovician brachiopod faunas from central- 
western N.S.W. 

Nautiloids [Percival] 

All nautiloids fi-om the Jingerangle Formation, 
with the exception of a section of a large indeterminate 
orthocone (Fig. 1 OD), are preserved as external moulds 
or an internal mould impression. The position of the 
siphuncle, and shape of the septa crossing the dorsal 



whorl profile, are unable to be determined, making 
generic identification uncertain if not impossible. 
Nevertheless, two genera of coiled nautiloids can be 
readily distinguished. A tightly coiled form is referred 
to Discocerasl (Fig. lOE), although Trocholitesl or 
Hardmanocerasl may be equally valid identifications. 
Several large slowly expanding conchs (Fig. lOG-I) 
with coarse ribbing appear to be broadly externally 
similar to an indeterminate tarphyceratid illustrated 
from the Gunningbland Shale Member (Ea 3 age) of 
the Goonumbla Volcanics at Gurmingbland, west of 
Parkes (Stait et al. 1985). A small individual (Fig. 
lOF) may represent a juvenile or micromorphic form 

of this tarphyceratid. 



ACKNOWLEDGMENTS 

We are grateful to the owners of "Bland Farm" who 
permitted us access to study the fossils in the Jingerangle 
Formation, and to Irvine Hay of Newcrest Exploration who 
alerted us to the presence of limestone in core from their 
Barmedman Creek prospect in the Marsden area. Gary 
Dargan (NSW Geological Survey) processed the conodont 
samples and prepared thin sections of fossils in the limestone. 
Photography of macrofossils was undertaken by David 
Barnes (NSW Department of Primary Industries) and the 
figures were drafted by Cheryl Hormann (NSW Geological 
Survey). Scanning electron microscope illustrations of the 
conodonts were prepared in the Electron Microscope Unit 
of the Australian Museum. Reviews by two anonymous 
referees assisted us in polishing the manuscript. This paper 
is a contribution to IGCP Project No. 503: Ordovician 
Palaeogeography and Palaeoclimate. Ian Percival and John 
Pickett publish with the permission of the Deputy Director- 
General, NSW Department of Primary Industries - Mineral 
Resources Division. 



REFERENCES 

An, T.X., Zhang, A.T and Xu, J.M. (1985). Ordovician 

conodonts from Yaoxian and Fuping, Shaanxi 

Province, and their stratigraphic significance. Acta 

Geologica Sinica 59, 97-108 (in Chinese with 

English abstract). 
An, T.X. and Zheng, S.C. (1990). 'The conodonts of the 

marginal areas around the Ordos Basin, north China'. 

199 pp. (Science Press: Beijing) (in Chinese with 

English abstract). 
Bowman, H.N. (1976). Forbes 1:250 000 metallogenic 

map. Geological Survey of New South Wales, 

Sydney. 
Branson, E.B. and Mehl, M.G. (1933). Conodont stodies. 

University of Missouri Studies 8, 1-349. 
Downes, PM. and Burton, G.R. (1999). Mineral 

occurrences in the Forbes district, pp. 37-52 in Lyons, 



Proc. Liim. Soc. N.S.W., 127, 2006 



251 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 



p. and Wallace, D. (eds). Geology and Metallogenesis 
of the Parkes-Grenfell-Wyalong-Condobolin Region, 
New South Wales. Forbes 1:250 000 Geological 
Sheet and Conference Guide 11-16 April 1999. 
AGSO Record 1999/20. 

Duncan, P.M. (1879). On some spheroidal lithistid 

spongida from the Upper Silurian formation of New 
Brunswick. Annals and Magazine of Natural History, 
series 4 4, 84-91. 

Glen, R.A., Walshe, J.L., Barron, L.M. and Watkins, J.J. 
(1998). Ordovician convergent-margin volcanism and 
tectonism in the Lachlan sector of east Gondwana. 
Geology 26, 751-754. 

Hill, D. (1957). Ordovician corals from New South Wales. 
Journal and Proceedings, Royal Society of New 
South Wales 91, 97-107. 

Lyons, P., Raymond, O.L. and Duggan, M.B. (eds). 
(2000). Forbes 1:250 000 Geological Sheet SI55- 
7, 2""* edition, Explanatory Notes. AGSO Record 
2000/20. 

Lyons, P. and Wallace, D. (eds) (1999). Geology and 
Metallogenesis of the Parkes-Grenfell-Wyalong- 
Condobolin Region, New South Wales. Forbes 1:250 
000 Geological Sheet and Conference Guide 11-16 
April 1999. AGSO Record 1999/20. 

Miles, I.N. and Brooker, M.R. (1998). Endeavour 42 

deposit. Lake Cowal, New South Wales: a structurally 
controlled gold deposit. Australian Journal of Earth 
Sciences 45, 837-847. 

Moskalenko, T.A. (1973). Conodonts of the Middle 
and Upper Ordovician on the Siberian Platform. 
Akademiy Nauk SSSR, Sibirskoe Otdelenie, Trudy 
Instituta Geologii i Geofiziki 137, 1-143 (in Russian). 

Nestor, H.E. (1964). Stromatoporoidei Ordovika i 

Llandoveri Estonii. Academiya Nauk Estonskoy SSR, 
Institut Geologii, Tallinn. 112 pp. (in Russian). 

Nowlan, G.S. and Barnes, C.R. (1981). Late Ordovician 
conodonts from the Vaureal Formation, Anticosti 
Island, Quebec. Geological Survey of Canada, 
Bulletin 3,19, 1-49. 

Packham, G.H. (ed) (1969). The Geology of New South 
Wales. Journal of the Geological Society of Australia 
16(1), XX -H 654 pp. 

Percival, I.G. (1978). Inarticulate brachiopods from the 
Late Ordovician of New South Wales, and their 
palaeoecological significance. ^/c/2m«ga 2, 117-141. 

Percival, I.G. (1999). Bolindian (Late Ordovician) fossils 
from the Jingerangle Formation, near Quandialla, 
New South Wales. Palaeontological Report 1999/03. 
Geological Survey of New South Wales, Report 
GS 1999/560 (unpublished). 

Percival, I.G. and Lyons, P. (2000). Jingerangle Formation, 
pp. 33-35 in Lyons, P., Raymond, O.L. and Duggan, 
M.B. (eds). Forbes 1:250 000 Geological Sheet SI55- 
7, 2°'' edition. Explanatory Notes. AGSO Record 
2000/20. 

Percival, I.G. and Webby, B.D. (1996). Island Benthic 
Assemblages: with examples from the Late 
Ordovician of Eastern Australia. Historical Biology 
11, 171-185. 



Pickett, J.W. (1986). Fossil sponges from Jingerangle. 

Palaeontological Report 1986/02. Geological 

Survey of New South Wales, Report GS 1986/0 10 

(unpublished). 
Pickett, J.W. and Percival, I.G. (2001). Ordovician faunas 

and biostratigraphy in the Gunningbland area, central 

New South Wales. Alcheringa 25, 9-52. 
Rigby, J.K. and Webby, B.D. (1988). Late Ordovician 

sponges from the Malongulli Formation of central 

New South Wales, Australia. Palaeontographica 

Americana 56, 1-147. 
Savage, N.M. (1990). Conodonts of Caradocian (Late 

Ordovician) age from the Cliefden Caves Limestone, 

southeastern Australia. Journal of Paleontology 64, 

821-831. 
Sherwin, L. (1973). Stratigraphy of the Forbes-Bogan 

Gate district. Records of the Geological Survey of 

New South Wales 15, 47-101. 
Sherwin, L. (1982). Fossils from the Marsden district. 

Palaeontological Report 1982/04. Geological 

Survey of New South Wales, Report GS1982/136 

(unpublished). 
Sherwin, L. (1985). Fossils from the Marsden and Bogan 

Gate 1:100 000 sheets. Palaeontological Report 

1985/08. Geological Survey of New South Wales, 

Report GS 1985/1 87 (unpublished). 
Sherwin, L., Clarke, I. and Krynen, J.P (1987). 

Stratigraphic units in the Forbes-Parkes-Tomingley 

district. Geological Survey of New South Wales, 

Quarterly Notes 67 , 1-23. 
Stait, B., Webby, B.D. and Percival, I.G. (1985). Late 

Ordovician nautiloids from cenfral New South Wales, 

Australia. Alcheringa 9, 143-157. 
VandenBerg, A.H.M. and Cooper, R.A. (1992). The 

Ordovician graptolite sequence of Ausfralasia. 

Alcheringa 16, 33-85. 
Warren, A.Y.E., Gilligan, L.B. and Raphael, N.M. (1995). 

Cootamundra 1:250 000 Geological Sheet SI/55-11: 

Explanatory Notes, vii -i- 160 pp. Geological Survey 

of New South Wales, Sydney. 
Webby, B.D. (1969). Ordovician sttomatoporoids from 

New South Wales. Palaeontology 12, 637-662. 
Webby, B.D. (1977). Upper Ordovician tabulate corals 

from central-western New South Wales. Proceedings 

of the Linnean Society of New South Wales 101, 167- 

183. 
Webby, B.D. (1992). Ordovician island biotas: New South 

Wales record and global implications. Journal and 

Proceedings, Royal Society of New South Wales 125, 

51-77. 
Webby, B.D. and Kruse, PD. (1984). The earliest 

heliolitines: a diverse fauna from the Ordovician of 

New South Wales. Palaeontographica Americana 54, 

164-168. 
Webby, B.D. and Morris, D.G. (1976). New Ordovician 

stromatoporoids from New South Wales. Journal and 

Proceedings, Royal Society of New South Wales 109, 

125-135. 
Webby, B.D. and Semeniuk, V. (1969). Ordovician 

halysitid corals from New South Wales. Lethaia 2, 



252 



Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT 



345-360. 

Webby, B.D. and Semeniuk, V. (1971). The Ordovician 
coral genus Tetradium Dana from New South Wales. 
Proceedings of the Linnean Society of New South 
Wales 95, 246-259. 

Webby, B.D., Zhen, Y.Y. and Percival, I.G. (1997). 

Ordovician coral- and sponge-bearing associations: 
distribution and significance in volcanic island shelf 
to slope habitats, Eastern Australia. Boletin de la Real 
Sociedad Espanola de Historia Natural 92, 163-175. 

Wynn, D.W. (1961). Notes on the geology of Bland Shire 
with special reference to deposits of road materials. 
NSW Department of Mines, Technical Reports (for 
1958) 6, 93-96. 

Yabe, H. and Sugiyama, T. (1930). On some Ordovician 
stromatoporoids from South Manchuria, North China 
and Chosen (Corea) with notes on two new European 
forms. Science Reports Tohoku Imperial University, 
ser. 2 (Geology) 14, 47-62. 

Zhen, Y.Y. (2001). Distribution of the Late Ordovician 
conodont Taoqupognathus in Eastern Australia and 
China. Acta Palaeontologica Sinica 40, 35 1-361 . 

Zhen, Y.Y, Percival, I.G. and Webby, B.D. (2004). 

Conodont faunas from the Mid to Late Ordovician 
boundary interval of the Wahringa Limestone 
Member (Fairbridge Volcanics), central New South 
Wales, Australia. Proceedings of the Linnean Society 
of New South Wales 125, 141-164. 

Zhen, Y.Y. and Webby, B.D. (1995). Upper Ordovician 
conodonts from the Cliefden Caves Limestone 
Group, central New South Wales, Ausfralia. Courier 
Forschungsinstitut Senckenberg 182, 265-305. 

Zhen, Y.Y, Webby, B.D. and Barnes, C.R. (1999). 

Upper Ordovician conodonts from the Bowan Park 
succession, cenfral New South Wales, Australia. 
GeoWo5 32, 73-104. 



Proc. Linn. Soc. N.S.W., 127, 2006 253 



QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS 



APPENDIX 
locality data and faunal lists 

Grid Reference 560500mE 6244240mN, Marsden (8430 II and III) 1:50 000 sheet 

Jingerangle Formation in roadbase quarry, immediately south of Jingerangle State Forest. 
Brachiopod: dorsal valve of small dalmanelloid? 

Sponge: Hindia sphaeroidalis Duacan, 1879 

indeterminate small conical form 

gigantic monaxons (several cms in length) 
echinoderm: crinoid ossicle 

graptolite: indeterminate climacograptid? 

Ptilograptus sp 

Grid Reference 560500mE 6243280niN, Marsden (8430 II and III) 1:50 000 sheet 

Jingerangle Formation in disused quarry, just west of "Bland Farm" homestead. 
Graptolites: indeterminate small climacograptid 

Dicellograptus gravis Keble and Harris 

Dicellograptus ornatus (EUes and Wood) 

Normalograptus angustus (Pemer) 

Orthograptus ex. gr. amplexicaulis (J. Hall) 

(centred on) Grid Reference 560500mE 6243100mN, Marsden 1:50 000 sheet 

Jingerangle Formation, scree on hillside behind "Bland Farm" homestead. 
Brachiopod: indeterminate large multicostate orthide? 
Nautiloids: Discocerasl sp 

indeterminate tarphyceratid 

indeterminate cyrtoconic brevicone 

indeterminate orthocone 

core from Newcrest drill hole DDMN 042, Marsden prospect (tenement EL5524) 

commenced 14/3/2002, completed 27/3/2002, TD 460.7 m 
GR 541658 mE 6256524 mN (GDA co-ordinates) 
for further details of micro- and macrofauna, refer to Tables 1 and 2 
Depth 387 m microfossil sample C2090 barren 

389 m C2091 barren 

391.2 m C2092 barren 
391.2-395.6 m C2071 conodonts 

397 m C2093 conodonts and ostracode 

399 m C2094 conodonts, macrofossil: coral {Tetradium) 

401 m C2095 conodonts 

403.9 m C2096 conodonts 

403.9-408 m C2072 conodonts, ostracodes, scolecodonts 

405 m macrofossil sample: coral {Tetradium) 

406 m macrofossil sample: coral {Tetradium) 
410 m C2097 conodont 

412 m C2098 ostracode and bryozoa 

414 m C2099 bryozoa and Ungulate brachiopod fragment 

416 m C2100 conodonts 

418.3 m C2101 barren 

core from Newcrest drill hole ACDMN 043, Marsden prospect (tenement EL5524) 

commenced 2/4/2002, completed 7/4/2002, TD 238.7 m 

GR 542347 mE 6255784 mN (GDA co-ordinates) 

for further details of micro- and macrofauna, refer to Tables 1 and 2 



254 Proc. Linn. Soc. N.S.W., 127, 2006 



I.G. PERCIVAL, Y. Y. ZHEN AND J.PICKETT 



Depth 141m microfossil sample C2102 benthic forams, ostracodes 

143 m C2103 barren 

170.7-175 m C2073 conodonts 

173 m macrofossil sample: coral (Halysites) 

174 m macrofossil sample: coral (Paleofavositesl) 
179.5-183.9 m C2074 conodonts, silicified corals 
179-180 m macrofossil sample: coral (Cystihalysites) 
181m macrofossil sample: sponge (Cliefdenella) 

1 82 m sponge (Cliefdenella), stromatoporoid {Stratodictyori) 

189-192.8 m C2075 conodonts 

192 m macrofossil sample: stromatoporoid (Labechiella) 

212 m C2104 barren 

214 m C2105 barren 

216 m C2106 barren 

227.8-232 m C2076 fragment of indet. conodont 

229 m diverse corals, stromatoporoids Ecclimadictyon, Clathrodictyon 

232 m macrofossil sample: coral {Bajgolia!) 



Proc. Linn. Soc. N.S.W., 127, 2006 255 



256 



Conservation of Australia's Forest Fauna (Second Edition) 

Daniel Lunney (editor) 

Royal Zoological Society of New South Wales 

PO Box 20, Mosman NSW 2088 

RRP $75.00 (plus $8.80 postage in Australia) 
Order form can be found at www.rzsnsw.org.au 



The first thing anyone will notice about 
this book is its size; 1070 pages weighing in at 3.36 
kilograms. I have been a bit tardy and more than a 
little hesitant to write a review of this book, since 
I have always made it a strict point to only write a 
review if I had read the entire publication. With this 
book, in some cases I did not get beyond the abstract. 
Like most people approaching a multi-author volume 
of wide scope, I first read those papers dealing with 
my own speciality (mammals), then looked for 
reviews of broader areas and finally at papers with 
catchy titles (of which there are an extraordinary 
niomber in this book). Some of those titles can be 
a bit misleading. I went straight to "Echidnas and 
archaeology: understanding the Aboriginal values of 
forests in NSW" only to find the echidna got only a 
brief mention. Most of the essay was concerned with 
exploring "... recent developments in the management 
of Aboriginal values in (forests of NSW)". That 
doesn't really make much sense, nor does a concluding 
observation that "Research and planning cannot be 
divorced fi-om the reality of people's strong feelings 
about social justice". I think I prefer Lord Kelvin's 
remark (as quoted by W Braithwaite on p. 524) that 
"If you can't measure it, it's not science". 

Many of the accounts are essays rather than 
'papers' in the research sense. I suspect the editor 
probably encouraged a less formal approach, which 
can lead, especially in reviews, to a much more 
readable work. 

The book is divided into five sections, and I 
will deal with them in sequence. 

IDENTIFYING THE ISSUES. 

M. Calver and G. Wardell- Johnson probably 
identify the underlying issue apparent throughout the 
book in one sentence - "ESFM cannot be achieved 
... without a ... will to assert long-term sustainable 
practice in the face of short-term goals" ESFM 
is, by the way. Ecologically Sustainable Forest 
Management. This section of the book contains many 
acronyms, arising no doubt firom the fact that many 
of the authors are working in governmental units of 
ever-changing acronyms (does DNR= DPNIR and 



what is NP&WS today?). I have always, as an editor, 
been very suspicious of any manuscript submitted 
that contained more than four acronyms. There is 
one essay here, which I shall kindly not name, that 
manages four in one sentence. 

H. Pamaby and E. Hamilton-Smith manage 
to encapsulate in one sentence, without a single 
acronym, the point of several entire essays that 
follow. They write: " ... conservation of Australia's 
forest bats has everything to do with cultural, political 
and corporate influences, and very little to do with 
biological 'facts'". They go on to describe the strange 
phenomenon of the "Adaptable bat". 

Other highlights in the section include a 
discussion of "predictor sets" of invertebrates by R.L. 
Kitching. A very different type of research to that 
employed by most biologists is used by S.M. Legg, 
who examined 19,000 newspaper items in order to 
determine how wildlife was portrayed in Victoria 
1839-1948. 

Surprisingly, my personal award for the 
most interesting, and perhaps the most significant for 
conservation, essay in this section goes to a lawyer. 
I am sure J. Prest is a lawyer because the essay uses 
footnotes instead of the usual Harvard system of 
citation. And in true legal style they often take up 
half the page. However the topic is vital in regard to 
the 87% of NSW native vegetation that is on private 
land and to the lack of control of deforestation on 
private land as opposed to crown land. This is the best 
coverage of the legislation (and lack of legislation) 
relating to private native forestry I have seen. The vital 
point is made that environmental laws remain mere 
words on paper without sufficient implementation and 
enforcement. Certainly in western NSW, what little 
legislation that is applicable is rarely applied to rural 
landholders. Many rural landholders can of course 
make effective use of public and political avenues of 
resistance to anything that seems to endanger their 
short-term interest. A good example in NSW is the 
reaction to the Native Veg. Act. 

Harry Recher, for example, has long argued 
that wildlife management and conservation must 
be extended to private land, an important aspect of 



BOOK REVIEW: FOREST FAUNA 



forest conservation that is examined in several places 
in this book. There is, by the way, a very interesting 
contribution by H. Recher at the start of this section 
(on eucalypt forest birds). 

LOOKING ACROSS THE LANDSCAPE 

The title of this section doesn't really tell 
what it contains, which is probably reasonable as it is a 
very mixed bag. A lot of information about techniques 
can be found herein. For example, RC. Catling and 
N.C. Coops give examples of the use of airborne 
videography in forest management. C.R Catterall et 
al. deal with quantification, including design issues, of 
the biodiversity values of reforestation. D. Milledge 
suggests an innovative approach to conservation 
plaiming in forests based on large owl territories. 

This section also includes a really good 
review of the role of nutrition in conservation of 
marsupial folivores by B.D. Moore et al. 

SINGLE SPECIES STUDIES 

The papers in this section are mostly reports 
of the kind of studies familiar to field biologists. 
Species covered are koalas (of course), tiger quoUs, 
brush-tailed phascogales, western ringtail possums, 
squirrel gliders and swift parrots. 

Subsequent papers don't really deal with 
single species but with larger groups. Individual 
papers deal with 26 species of feathered fi-uit-eaters, 
two fi-ogs (southern barred and giant burrowing), 
a small mammal community of nine species, two 
gliders (yellow-bellied and mahogany) and the entire 
mammal fauna in SE forests. A paper on bats in state 
forests is probably out of place here since it deals with 
management and really belongs in the next section. 

MANAGING FOREST FAUNA 

Having found some of the essays related 
to management in the first two sections of the book 
heavy going, I approached this final section with 
considerable trepidation. However, many of the 
papers in this section contain an amazing amount of 
information and are oriented more towards the data 
on which management should be based rather than 
the management process itself. Two very interesting 
sets of data concern the effects of Phytophthora 
dieback on forest fauna (M.J. Gerkaldis et al.) and 
the effects of fire on fiangus species which are an 
important component of the diet of many forest 
animals (A.W. Claridge and J.M. Trappe). The latter 
is very much a management issue in that an assumed 
beneficial effect of fiiel-reduction bums on fimgi has, 
in my own experience, been used as a justification of 
the practice. 



Dan Lunney closes the book with a siimmary 
entitled 'The fixture of Australia's forest fauna 
revisited" in which he states the aim of this second 
edition is to enhance the opportunities to commimicate. 
The book has achieved that aim admirably and the 
credit for that must go to the editor. 

I strongly recommend this book to 
conservationists, biologists and especially forest 
and fauna managers. After all, it is only $25 a kilo 
including postage; I've paid more than that for 
cheese. 

M.L. Augee 

Sydney 

20 December 2005 



258 



Proc. Linn. Soc. N.S.W., 127, 2006 



INSTRUCTIONS FOR AUTHORS 

{this is an abbreviated form - the full instructions can be obtained from our web site or from the Secretary) 



1 . The Proceedings of the Linnean Society of New South Wales publishes original research papers dealing 
with any topic of natural science, particularly biological and earth sciences. 

2. Manuscripts should be submitted to the Editor (MX. Augee, PO Box 82, Kingsford NSW 2032). All 
manuscripts are sent to at least two referees and in the first instance three hard copies, including all figures 
and tables, must be supplied. Text must be set at one and a half or double spacing. 

3. References are cited in the text by the authors' last name and year of publication (Smith 1987, Smith and 
Jones 2000). For three of more authors the citation is (Smith et al. 1988). Notice that commas are not used 
between the authors' names and the year. The format for the reference list is: 

Journal articles: 

Smith, B.S. (1987). A tale of extinction. Journal of Paleontological Fiction 23, 35-78. 
Smith, B.S., Wesson, R.I. and Luger, WK. (1988). Levels of oxygen in the blood of dead Ringtail 
Possums. Australian Journal of Sleep 230, 23-53. 
Chapters or papers within an edited work: 

Ralph, P.H. (2001). The use of ethanol in field studies. In 'Field techniques' (Eds. K. Thurstle and P.J. 
Green) pp. 34-41. (Northwood Press, Sydney). 
Books: 

Young, V.H. (1998). 'The story of the wombat'. (Wallaby Press, Brisbane). 

4. An abstract of no more than 200 words is required. Sections in the body of the paper usually include: 
INTRODUCTION, MATERIALS AND METHODS, RESULTS, DISCUSSION, ACKNOWLEDGEMENTS 
and REFERENCES. Some topics, especially taxonomic, may require variation. 

5. Subheadings within the above sections should be in the form: 

Bold heading set against left margin 

This is the form for the first level headings and the first line of text underneath is indented 
Underlined heading set against left margin 

This is the next level, and again the first line of text underneath is indented. 
Further subheadings should be avoided. 
Italics are not to be used for headings but are reserved for genus and species names. 

6. Up to 10 KEYWORDS are required. These are often used in computer search engines, so the more specific 
the terms the better. 'Australian' for example is useless. Please put in alphabetical order. 

7. Paragraphs are to be set off by a tab indentation without skipping a line. Do not auto-format the first line 
(i.e. by using the "first line" command in WORD). All auto-formatting can be fatal when transferring a 
manuscript into the publisher platform. 

8. Details of setting up the manuscript: 

Use 12 point Times New Roman font. 
Do not justify 

Margins should be: 3 cm top, 2.5 cm bottom, 3 cm left and 2.5 cm right. This is the area available for text; 
headers and footers are outside these margins. 

9. 3. The final version, incorporating referees' and editor's comments, must be supplied on floppy disc or 
CD in WORD for PC format (Mac discs will not be accepted). The text file must contain absolutely no auto- 
formatting or track changes. 



FIGURES: 

Figures can be line drawings, photographs or computer-generated EXCEL or WORD files. No 
figures will be accepted larger than 15.5 x 23 cm. Width of lines and sizes of letters in figures must be large 
enough to allow reduction to half page size. If a scale is required, it must be presented as a bar within the 
figure and its length given in the legend. It is the editor's prerogative to reduce or enlarge figures as necessary 
and statements such as "natural size" or "4X" in the legend are unacceptable. 

Photographs must be supplied as black and white prints or as .TIF files (Jpeg is not acceptable) 
scanned at 600 dpi. Line drawings must be supplied as original copies or as .TIF .files scanned at 1200 dpi. 
Other figures must be in hardcopy. 

While there is no objection to full page size figures, it is journal policy to have the legend on the 
same page whenever possible and figures should not be so large as to exclude the legend. Figure legends 
should be placed together on a separate page at the end of the manuscript. 

TABLES 

Because tables may need to be re-sized, it is essential that table legends are not set within the table 
but are supplied separately with the figure legends. Legends need to be the same font and size as the rest of 
the manuscript. 

While the text of the legend is expected to be in 12 point type, it may be necessary to use a smaller 
font size for large tables. It is journal policy to accept tables that run over more than one page only in 
exceptional circumstances. 

Do not use vertical lines in tables unless absolutely necessary to demark data columns. Keep 
horizontal lines to a minimum and never put a border around tables. 

WORD or EXCEL tables are acceptable, but EXCEL is preferred. 

Tables and/or figures must be separate from the text file. Never embed figures or tables in the text. 

10. Details of punctuation, scientific nomenclature, etc. are to be found in the complete instructions available 
fi-om the website or from the Secretary. 



260 



CONTENTS CONTINUED 

83 Mahony, M. 

Amphibians of the Gibraltar Range. 
93 Vemes, K., Green, S., Howes, A. and Dunn, L. 

Species richness and habitat associations of non-flying mammals in Gibraltar Range National Park 

Section 11: General papers. 

107 Harris, J.M. 

The discovery and early natural history of the Eastern Pygmy-possum, Cercartetus nanus (Geoffrey 

and Desmarest, 1817). 
125 Piper, K.J. and Herrmann, N. 

Additions to knowledge of the early Pleistocene wallaby Baringa nelsonensis Flannery and Hann 

1984 (Marsupialia, Macropodinae). 
133 Williamson, PL. and Rickards, R.B. 

Eastonian (Upper Ordovician) graptolites from Michelago, near Canberra. 
157 Timms, B.V. 

The geomorphology and hydrology of saline lakes of the middle Paroo, arid-zone Australia. 
175 Foldvary,G. 

Pseudoplasmopora (Cnidaria, Tabulata) in the Siluro-Devonian of eastern Australia with comments 

on its global biogeography. 
191 Baker, A.C., Hose, G.C. and Murray, B.R. 

Vegetation responses to Pinus radiata (D. Don) invasion: a multivariate analysis using principal 

response curves. 
199 Valentine, J.L., Cole D.J. and Simpson, A.J. 

Silurian linguliformean brachiopods and conodonts from the Cobra Formation, southeastern New 

South Wales, Australia. 
235 Percival, I.G., Zhen, Y.Y. and Pickett, J. 

Late Ordovician faunas from the Quandialla-Marsden district, south-central New South Wales. 
256 Book review: Conservation of Australia's forest fauna. 

A recent expansion of its Queensland range by Eupristina verticillata Waterston (Hymenopera, 
Agaonidae, Agaoninae), the pollinator of Ficus microcarpa l.f. (Moracea). 



GS OF THE LINNEAN SOCIETY OF N.S.W. 



SMITHSONIAN INSTITUTION L^RARJES 

3 9088012102877 



THE 



Y\ new south /'^J 

\ /fV WALES y^ y 



Issued 23 February 2006 - 

CONTENTS 

Section I: The biology and ecology of Gibraltar National Park. 

I Clarke, P.J. and Myerscough, P.J. 

Introduction to the biology and ecology of Gibraltar Range National Park and adjacent 
areas: patterns, processes and prospects. 
5 Jones, R.H. and Bruhl, J.J. 

Acacia beadleana (Fabaceae: Mimosoideae), a new, rare, localised species from Gibraltar 
Range National Park, New South Wales. ^^^g 

II Caddy, H.A.R. and Gross, C.L. 

Population structure and fecundity in the putative sterile shrub, Grevillea rhizomatosa Olde 

& Marriott (Proteaceae). 
19 Vaughton, G. and Ramsey, M. 

Selfed seed set and inbreeding depression in obligate seeding populations of Banksia 

marginata. 
27 Williams, RR. and Clarke, RJ. ^ 

Fire history and soil gradients generate floristic patterns in montane sedgelands and wet 

heaths of Gibraltar Range National Park. i^m 

39 Virgona, S., Vaughton, G. and Ramsey, M. ^M 

Habitat segregation of Banksia shrubs at Gibraltar Range National Park. ^M 

49 Knox, K.J.E. and Clarke, RJ. ^ — 

Response of resprouting shrubs to repeated fires in the dry sclerophyll forest of Gibraltar 

Range National Park. 
57 Croft, P., Hofmeyer, D. and Hunter, J.T. 

Fire responses in four rare plant species at Gibraltar Range National Park, Northern 

Tablelands, NSW. 
63 Campbell, M.L. and Clarke, P.J. L 

Response of montane wet sclerophyll forest understorey species to fire: evidence from high 

and low intensity fires. 
75 Goldingay, R.L. and Newell, D.A. ' ., 

A preliminary assessment of disturbance to rock outcrops in Gibraltar Range National Park '. 

CONTINUED INSIDE BACK COVER 1 



Printed by Southwood Press Pty Ltd, 
76-82 Chapel Street, Marrickville 2204 




NEW SOUTH WALES 



VOLUME 128 




^^m,iiMm 



NATURAL HISTORY IN ALL ITS BRANCHES 



THE LINNEAN SOCIETY OF 

NEW SOUTH WALES 

ISSN 0370-047X 




Founded 1874 
Incorporated 1884 

The Society exists to promote the cultivation and study 
of the science of natural history in all its branches. 
The Society awards research grants each year in the 
fields of Life Sciences (the Joyce Vickery fund) and 
Earth Sciences (the Betty Mayne fund), offers annually 
a Linnean Macleay Fellowship for research, contributes 
to the stipend of the Linnean Macleay Lecturer in 
Microbiology at the University of Sydney, and 
publishes the Proceedings. It holds field excursion and 
scientific meetings, including the biennial Sir William 
Macleay Memorial Lecture delivered by a person 
eminent in some branch of natural science. 



Membership enquiries should be addressed in the first instance to the Secretary. Candidates for elec- 
tion to the Society must be recommended by two members. The present annual subscription is 
$A56.00. 

The current subscription rate to the Proceedings is set at $A80.00 per volume. In recent years a 
volume consists of a single annual issue. 

Back issues of all but a few volumes and parts of the Proceedings are available for purchase. Prices 
are Usted on our home page and can also be obtained from the Secretary. 

OFnCERS AND COUNCIL 2005/2006 

President: D.R. Murray 

Vice-presidents: M.L. Augee, I.G. Percival, J.P. Barkas 

Treasurer. I.G. Percival 

Secretary: J-C. Herremans 

Council: A.E.J. Andrews, M.L. Augee, J.P. Barkas, M. Cotton, G. Edgecombe, M.R. Gray, 

J-Cl. Herremans, D. Keith, R.J. King, H.A. Martin, E. May, D.R. Murray, P.J. Myerscough, I.G. Percival, 

S. Rose and K.L. Wilson 

Editor. M.L. Augee 

Assistant Editor: Elizabeth May 

Linnean Macleay Lecturer in Microbiology: PR. Reeves 

Auditors: Phil Williams Carbonara 

The postal address of the Society is: RO. Box 82, Kingsford NSW 2032, AustraUa 

Telephone: (International) 61 2 9662 6196; (Aust) 02 9662 6196 

E-mail: linnsoc@acay.com.au 

Home page: www.acay.com.au/~linnsoc/welcome.html 

© Linnean Society of New South Wales 



Cover motif: Myrtaceae pollen from Robbie and Martin, page 65, this volume 



PROCEEDINGS 
of the 

LINNEAN 
SOCIETY 



of 

NEW SOUTH WALES ^''^W^^^ONi^ 

\ t 5 2007 ) 



For information about the Linnean Society of New South Wales, its pubhcations and 

activities, see the Society's homepage 

www.acay.com.au/~hnnsoc/welcome.htm 



VOLUME 128 

February 2007 



The Characteristics of Five Species of Hollow-Bearing Trees on 

the New South Wales Central Coast 

p. TODARELLO' AND A. ChALMERS^ 

Centre for Sustainable Coasts and Catchments, The University of Newcastle, PO Box 127 Ourimbah, 

NSW 2258; 'Present address: Ku-ring-gai Council, Bushland Operations, 818 Pacific Highway 

Gordon, NSW 2072; ^Corresponding author (amta.chalmers@newcastle.edu.au) 



Todarello, P. and Chalmers, A. (2007). The characteristics of five species of hollow-bearing trees on the 
New South Wales central coast. Proceedings of the Linnean Society of New South Wales 128, 1-14. 

Five native eucalypt species were examined to investigate the abundance, entrance size diameter and type 
(e.g. trunk, branch) of hollows present. A total of 698 living trees were sampled within 22 one hectare plots. 
The trees were distributed across five open forest or woodland communities on the Central Coast of NSW; 
these communities were underlain by Narrabeen or Hawkesbury sandstone. The number of hollows per tree 
was positively correlated with the diameter of the tree and, with the exception of Coiymbia gummifera, with 
the height of the tree. The smallest species examined, Eucalyptus haemastoma, contained a high proportion 
(60%) of small diameter (20-35 cm) hollow-bearing trees, confirming that hollow availability is more 
strongly related to species characteristics rather than to absolute diameter Eucalyptus haemastoma had 
the highest proportion of hollow-bearing trees (78%) followed by Angophora costata (40%), Eucalyptus 
punctata (26%), C. gummifera (24%)) and Eucalyptus pilularis (22%). The results obtained for E. pilularis 
may not be a true reflection of the propensity of this species to form hollows, as the sampled population 
may have been affected by timber removal. Most hollows had small (2-5 cm) diameter entrances (47%)) and 
occurred in branches (84%) rather than in main stems (16%)). 

Manuscript received 1 June 2005, accepted for publication 18 May 2006. 

KEYWORDS: Angophora costata, Corymbia gummifera, Eucalyptus haemastoma. Eucalyptus pilularis. 
Eucalyptus punctata, cavities, habitat trees, hollows 



INTRODUCTION 

Gibbons and Lindenmayer (2002) estimate that 
there are over 300 native vertebrate species that use 
tree hollows within Australia. On the Central Coast 
of NSW there are at least 54 fauna species that use 
tree hollows, 13 of which are listed as threatened 
vmder the NSW Threatened Species Conservation Act 
1995. For example. Smith and Murray (2003) found 
that the abundance of all possums and gliders in the 
Wyong region of the NSW Central Coast increased 
with the number of hollow-bearing trees, particularly 
in areas where the average diameter at breast height 
was greater than 80 cm. They also found that the 
highest estimated density of squirrel gliders (Petaurus 
norfolcensis) occurred in associations of Scribbly 
Gum {Eucalyptus haemastoma). Smooth-barked 
Apple {Angophora costata) and Red Bloodwood 
{Corymbia gummifera). 



Many authors (Lindenmayer et al. 1991, 1993b, 
1994; Cockbum and Lazenby-Cohen 1992; Eyre and 
Smith 1997; Lindenmayer 1997; Wormington et al. 
2003) have shown that different species of arboreal 
marsupials exhibit preferences for hollow-bearing 
trees with different characteristics. Occupation of 
hollows by fauna is associated with hollow entrance 
diameter and hollow depth as these characteristics 
influence the degree of protection from predators, the 
micro-climate and the provision of sufficient space for 
sleeping and nesting (Gibbons et al. 2002; Gibbons and 
Lindenmayer 2002). While small animals generally 
prefer hollows with small entrances, they may also 
use hollows with large entrances. For example, 
Antechinus spp., feathertail gliders {Acrobates 
pygmaeus) and sugar gliders {Petaurus breviceps) 
prefer hollows with entrance widths of 2-5 cm, but will 
use hollows with entrance widths > 5 cm (Gibbons et 
al. 2002). Larger species such as the common ringtail 



HOLLOW-BEARING TREES 



possum {Pseudocheirus peregrinus), greater glider 
{Petauroides volans), yellow-bellied glider (Petaurus 
australis) and common brushtail possum (Trichosurus 
vulpecula) are restricted to hollows with a minimum 
entrance width of > 5 cm (Gibbons et al. 2002). Large 
forest owls and cockatoos require large hollows for 
breeding, which tend to only occur in large diameter 
trees (Gibbons and Lindenmayer 2002). For example, 
Gibbons et al. (2002) only recorded the Powerful Owl 
{Ninox strenud) in hollows with a minimum entrance 
diameter of > 10 cm. Fauna are more likely to occupy 
trees with many hollows because it is more likely 
that these trees will have a least one suitable hollow 
(Gibbons et al. 2002). 

The combined factors of clearing for agriculture, 
forestry and urbanisation have all contributed 
significantly to the reduction of the forest estate 
(Lindenmayer et al. 1993a; Cork and Catling 1996). 
Unfortunately many of these cleared forests supported 
optimal habitat for hollow-dependent fauna (Norton 
1987; Bennett et al. 1994). Further, of those species 
that inhabit wood production forests, arboreal fauna 
are considered the most vulnerable to the impacts 
of timber harvesting (Ball et al. 1999). Bennett et 
al. (1994) argue there is growing evidence that the 
availability of suitable hollows is a limiting factor for 
most hollow-dependent fauna. Hollow-bearing trees 
in managed stands may be reduced by about 50 - 90% 
of that found in 'natural' stands, a reduction predicted 
to reduce populations of hollow-using fauna as well as 
faunal diversity (Gibbons and Lindenmayer 2002). 

Hollow formation in eucalypts results from 
a series of abiotic and biotic events following the 
wounding of Hving stem tissue (Wilkes 1982). 
Wounding can occur in a number of ways including 
branch breakage due to high winds and exposure 
to high temperatures during fire (Gibbons and 
Lindenmayer 2002). After wounding, the process of 
wood decay follows a complex succession of micro- 
organisms including bacteria, fiingi and insects such 
as termites (Wormington et al. 2003; Wilkes 1982; 
Perry et al. 1985; Gibbons and Lindenmayer 2002). 
A hollow eventually forms when decay undermines 
the strength of a branch, or when a branch has broken 
off during strong winds and/or fire; and the hollow is 
subsequently excavated by fiangi, termites and other 
invertebrates and animals (Gibbons and Lindenmayer 
2002). Physiological stress and fire predispose trees 
to attack by fungi and termites, while fire is also 
directly involved in excavating hollows (Gibbons et 
al. 2002). It may take 120-220 years for hollows to 
form (Gibbons and Lindenmayer 2002). 

The number of hollows in individual trees 
and the size of hollows generally increase as tree 



diameter increases (Bennett et al. 1994; Williams 
and Faunt 1997; Gibbons et al. 2000; Lindenmayer 
et al. 2000; Whitford 2002). Larger trees tend to have 
a greater number of hollows because trees become 
physiologically weaker and shed more branches as 
they age and are more likely to have been exposed 
to stochastic events (e.g. fire) that facilitate hollow 
formation (Gibbons et al. 2002; Gibbons and 
Lindenmayer 2002). 

Despite the large number of fauna species that 
rely on tree hollows, there is a paucity of data on 
the distribution and abundance of hollows within 
Australia (Gibbons and Lindenmayer 2002). Little 
is known about the hollow characteristics of specific 
tree species occurring on the Central Coast of NSW. 
An understanding of the propensity of different tree 
species to form hollows in any given area or region 
is essential to manage and maintain the hollow tree 
resource for that area. Thus, the main aim of this study 
was to examine the number and type of hollows in five 
tree species {Angophora costata Britten, Corymbia 
gummifera (Gaertn.) K.D.Hill and L. A. S.Johnson, 
Eucalyptus haemastoma Sm., Eucalyptus pilularis 
Sm. and Eucalyptus punctata DC. subsp. punctata) 
common on the NSW Central Coast. With the 
exception of E. pilularis, no previously published 
information on hollows could be found for these 
species. More specifically, the study asked: 

1 . Does hollow abundance depend on tree size 
(diameter and height)? 

2. Are there differences in the propensity of the 
species examined to form hollows? 

3. Are there differences between the species in 
the location (main stem versus branch) and 
entrance size of hollows? 



MATERIALS AND METHODS 

Study area 

The Central Coast region of NSW has a warm 
temperate climate and supports closed forests, tall 
open forests, open forests, woodlands and heath 
(Murphy 1993). Rainfall ranges from a high of 
1310 mm along the coast at Gosford to a low of 740 
mm at Bucketty in the northwest (Murphy 1993). 
In summer, the average monthly temperatures are 
highest (27.2°C) on the coast and lowest (15.2°C) on 
the plateau, while in winter, average temperatures are 
highest (19.7°C) on the coast and lowest (4.2°C) in 
the valleys (Murphy 1993). 

The five vegetation communities sampled in 
this study were open forests or woodland found on 
infertile soils underlain by Narrabeen or Hawkesbury 



Proc. Linn, Soc. N.S.W., 128, 2006 



p. TODARELLO AND A. CHALMERS 



Sandstone. They were: i) Coastal Foothills Spotted 
Gum- Ironbark Forest; ii) Dharug Roughbarked Apple 
Forest, which is found over a number of topographic 
positions on Narrabeen Sandstones and within the 
rain shadow of the Watagan Ranges; iii) Coastal 
Narrabeen Shrub Forest, which occurs on skeletal 
ridge-top soils often near or with outcroppings of 
Hawkesbury and Narrabeen Sandstone; iv) Exposed 
Hawkesbury Woodland, which generally occurs on 
crests, ridges and exposed slopes on sandy soils of 
the Hawkesbury Sandstone series; and v) Exposed 
Yellow Bloodwood Woodland, which is found on dry 
exposed, infertile ridges and slopes on Hawkesbury 
Sandstone (LHCCREMS 2000). 

Site selection 

Sites were selected based on the fi'equency of the 
target species within the 55 vegetation communities 
that occur within the Central Coast and Lower Hunter 
Region (LHCCREMS 1: 100 000 Vegetation Map 
Sheet 2003). To minimise sampling time and effort, 
preference was given to those vegetation communities 



that contained more than one of the target species at 
frequencies greater than 30% (Table 1). A total of 
five vegetation communities fulfilled this criterion. 
Sampling of the five target species was undertaken 
at 22 sites distributed within three National Parks 
and two State Forest reserves on the Central Coast 
(Fig. 1; Table 1). State Forest logging history records 
indicate that the four sites sampled in vegetation 
community 1 had been logged between 1966 and 
1980-82. The two State Forest sites in vegetation 
community 3 had been logged between 1966 and 
1999. For vegetation community 4, two of the State 
Forest sites had been logged between 1966 and 1984- 
85, whilst the other two sites were last logged in 1962 
and 1965-66. Sites sampled at Bouddi National Park 
(vegetation community 3) may have been subject to 
timber removal by subsistence farmers prior to the 
land being added to the Park between 1938 and 1967 
(Strom 1986). 

The location of each site was randomly 
selected within each vegetation community using the 
following procedure. The distance of the main access 




Figure 1. Location of tiie Central Coast of New South Wales (inset) and the three Na- 
tional Parks and two State Forests sampled in the current study. 



Proc. Linn. Soc. N.S.W., 128, 2006 



HOLLOW-BEARING TREES 



Table 1. Vegetation communities sampled in the study, target species and their expected frequencies 
and number of sites by land tenure within each vegetation community. 1 - Coastal Foothills Spotted 
Gum-Ironbark Forest; 2 - Dharug Roughbarked Apple Forest; 3 - Coastal Narrabeen Shrub Forest; 
4 - Exposed Hawkesbury Woodland: 5 - Exposed Yellow Bloodwood Woodland. * based on LHC- 
CREMS (2000) 



Vegetation 
community* 


Target species in each 
vegetation community 


Frequency* 


Land Tenure 


No. of Sites 


1 


Angophora costata 
Eucalyptus punctata 


36% 
31% 


Ourimbah State Forest 
Olney State Forest 


2 
2 


2 


Eucalyptus punctata 


68%^ 


Dharug National Park 

r 


^^3^ 



Angophora costata 
Eucalyptus pilularis 
Corymbia gummifera 



74% 
40% 
48% 



Ourimbah State Forest 
Bouddi National Park 



2 
4 



Angophora costata 45% 

Eucalyptus haemastoma 50% 

Corymbia gummifera 75% 



Ourimbah State Forest 
Popran National Park 



4 
4 



Eucalyptus punctata 
Corymbia gummifera 



52% 
40% 



Dharug National Park 



road running through the area to be sampled (portion 
of reserve containing one of the five vegetation 
communities) was measured from its entry to its exit 
point on a topographic map. Each 1 km section of the 
access road was allocated a number and numbers were 
randomly chosen to determine how many kilometres 
the site would be fi^om the entry point of the reserve. 
A 100 m section of road was then randomly chosen 
from that 1 km section using the same procedure (with 
100 m sample lengths). At each survey point a 1 ha 
(100 m X 100 m) quadrat was established 50 m off the 
access road. The side of the access road to be sampled 
was determined by flipping a coin. Quadrats were 
placed 50 m away from any existing road or track to 
minimise the influence of edge effects and disturbance 
created by road construction and maintenance. All 
quadrats were established at least 1 km apart to 
ensure the samples were independent of each other 
and would be representative of any variation within 
the vegetation. The placement of quadrats 1 km apart 
and 50 m from the road is consistent with the methods 
used by Gibbons et al. (2000). 

Data collection 

All living trees of the target species with a 
diameter at breast height (dbh) > 20 cm were sampled 



in each 1 ha quadrat. The lower limit of 20 cm 
dbh was chosen because previous studies in other 
regions (Williams and Faunt 1997; Whitford 2003; 
Wormington et al. 2003) have shown that hollow- 
bearing frees of this size contain hollows that may be 
used by the smaller marsupials. The diameter of each 
tree was measured using a diameter tape at a height of 
1.3 m over bark and allocated to one of the following 
diameter classes: 20-35, 36-51, 52-67, 68-83 or >84 
cm. Tree height was determined using a clinometer 
and each tree sampled was allocated to one of the 
following height classes: 5-10, 11-16, 17-22, 23-28 
or >29 m. The number of hollows in each free was 
determined from the ground using 10 mm x 25 mm 
binoculars. A hollow was defined as any cavity with 
an enfrance > 2 cm in diameter and occurring > 3m 
above the ground. Enfrances that were obviously 
'blind' were not counted. 'Blind' was defined as "a 
branch stub or area of damage that does not lead to a 
cavity" (Gibbons and Lindenmayer 2002). Hollows 
in stumps or large fire scars (fissures) were not 
included. Each hollow was assigned as either having 
a small (2-5 cm), medium (6-10 cm) or large (>10 
cm) enfrance based on a visual estimate from the 
ground. The location of each hollow was recorded as 
either occurring in a branch or main stem. The lower 



4 



Proc. Linn. Soc. N.S.W., 128, 2006 



p. TODARELLO AND A. CHALMERS 



size limit for sampling and the diameter, height 
and hollow classes were consistent with previous 
studies by Gibbons and Lindenmayer (1997), 
Williams and Faunt (1997), Gibbons et al. (2002) 
and Wormington et al. (2003). 

Statistical analyses 

The data were not normally distributed and 
transformation did little to improve normality. 
Therefore the Kruskall-Wallis test was used 
to determine whether there were significant 
differences between species in the ranked averages 
of the number of hollows per tree, tree density and 
density of hollow-bearing trees. Spearman rank 
correlations were used to test for an association 
between number of hollows and diameter, as well 
as between hollow number and tree height. All 
statistical analyses were conducted with SPSS 
version 11.5. 

RESULTS 

Atotal of 698 living trees were sampled across 
the five species, with 254 of these trees (36%) being 
hollow-bearing and 781 hollows being observed. 
Tree density of those species examined (i.e. not 
total tree density of a site) ranged firom 1 to 37 trees 
ha' (mean of 16.6 ha') and the number of hollow- 
bearing trees ranged fi-om to 27 ha' (mean of 
6.2 ha'). Due to the composition of the vegetation 
communities sampled, there were considerably 
fewer data collected for Eucalyptus pilularis 
than for the other species (Table 2). Angophora 
costata and E. pilularis showed a similar range 
of tree diameters, but the mean diameter of E. 
pilularis was considerably larger than that of ^. 
costata (Table 2). Eucalyptus haemastoma was the 
shortest species investigated, whilst Eucalyptus 
punctata was the tallest (Table 2). Of those species 
examined, Angophora costata showed the greatest 
range in height (Table 2). 

The mean number of hollows per tree differed 
significantly between the tree species (K = 107.5; 
4 df; p < 0.0001). Eucalyptus haemastoma had the 
highest mean number of hollows per tree, followed 
by A. costata and E. pilularis, while E. punctata 
and Corymbia gummifera had the fewest number 
of hollows (Table 2). Eucalyptus haemastoma had 
the highest proportion of hollow-bearing trees (78 
%) followed by ^. costata (40 %) > E. punctata (26 
%) > C. gummifera (24 %) > E. pilularis {11 %). 
At the stand level (i.e. per ha), tree density did not 
differ significantly between the species (K = 1.5; 
4 df; p= 0.820), although the density of hollow- 



II 






■Sg 


>. 


If 


3 


■^^ 




^^ 


e 




a> 




u 




u 




S 




u 




V 


li 

ti 

^1 


5 


i 

e 
S 


c« 




>*. 




a> 




Ml 




.Si 






|.a 


o. 

Vi 


-1;^ 


> 


8:5 


« 


aci. 


aj 


tt) 


J= 




■^ri 




t<-i 




o 




J= 




u 






Is 


;. 
£ 


J:i 




a c 


(« 


u a 


^ 


SCi, 


_o 


til 










o 




.e 




"O 




e 




M 






a 


V3 


s ^ 


E 

<U 


OS 
"^1 


«1 


S^S 




^§ 


1. 

-*- 


^ 


5« 




O 




K 




01 




-w 




3 




£ 








'C 




-w 




■fcri 




A 




Vi 




S 




o 








'C 




c« 




> 




k. 




o 




<s 




») 




a* . 




2 B 

cs 5! 




^i 




© a, 




a Ji 




ej)-t- 




C "2 




S3 « 




u u 




^l 




H « 




. ■« 




r^ )i 


1) 




1 


■fi S 


;-i 


ea w 


cd 




> 



_ ^^ rn 

;z, 00 -^ 



m 
O 



IT) 

cs 



00 



ON 



in 



in 

o 



IT) — . 



+1 

oo 
rn 
m 



lO 



+1 
oo 

0\ 
CI 



+1 

od 



00 



+1 
i/-i 



00 



o 



oo 



<N 



O 



t^ 
■* 



O 



O 
+1 






<N 

O 

+1 



fN 



IT) m 
^ ^ '^ 



o 

+1 

(N 



lO 

o 
+1 

rn 



ON 



ON 



o 
+1 
1^ 
o 



IT) 






^ 
■* 



+1 



^H 




. — 1 




o 
d 


ON 




•T) 


d 




00_ 


+1 


1 


+1 


1 


+1 


+1 


OO 


o 


r-; 


VO 


(N 


tN 


l— t 


CN 


.— < 




' 


^d 



oo 



+1 



d ^ 

+1 +1 

















1 




-4-> 






^.— N. 








t_i 




'9^ 






s 








1) 




&P 




T3 








cx 






■a 

t 


0) 

o 


•3 


? 
^ 

tM 
U 

o 


J3 


op 


tji 

1 

1 

O 


C3 

CO 

(l> 

o 


■1 


c/l 


_o 


U 


ca 


O 










"o 


te 


^ 


^ 


1> 


d 


d 


i3 


jz 


^<-^ 


(U 


•— S 


u 








!+-! 


!*-l 


w 


2 


m 


<+-c 


m 


W 


o 


O 


on 


<4-l 


C/3 


00 


c« 


^H 


6 


d 


+1 


O 


+1, 


O 


+1, 


+1 


^ 


c 


c 


^ — ^ 


m 




flj 




^ — ^ 


o 


, 


1-^ 


a 


1 


f3 


Pi 


f2 


n 




«J 


(X3 


cS 


rt 


cd 


ca 


d 

2; 


■♦-» 

f2 


4-> 

f2 




1 


1 


2 






o -^ 

+1 +1 +1 

VO -^ -H 

O S -t 



o 



O 

d 
+1 
lo 

^" 2 :: 



+1 

o 



oo 

00 



+1 

00 



+1 

ON 



00 

+1 
in 
vd 



C3 

J3 






so 

_c 

'C 

ca 
to 

I 

_o 

"o 

Si 

o 

d 

c 

W 

00 

i 



Proc. Linn. Soc. N.S.W., 128, 2006 



5 



HOLLOW-BEARING TREES 



Table 3. Spearman rank correlations between hollow number and tree 
diameter and hollow number and tree height for each individual species 
* p<0.05; ** p< 0.01; ns = not significant (p>0.05). 



Tree diameter 



Spearman's rho 

Tree height 



Angophora costata 175 

Eucalyptus punctata 157 

Eucalyptus pilularis 82 

Eucalyptus haemastoma 1 03 

Corymbia gummifera 181 



0.61 



** 



0.44 



** 



0.3P 



0.32** 



0.67 



** 



0.58 



** 



0.53 



** 



0.40 



0.23=" 



0.06° 



** 



bearing trees did (K = 10.7; 4 df; p < 0.05). The mean 
density of hollow-bearing trees was highest for E. 
haemastoma followed by A. costata, E. punctata, C. 
gummifera and E. pilularis (Table 2). 

Hollow number and tree size 

There was a significant positive association 
between tree diameter and number of hollows per tree 
for each of the five species examined (Table 3; Fig. 
2). Eucalyptus pilularis formed few hollows in trees 
< 80 cm dbh (Fig. 2c) but had the highest count of 
hollows occurring in any one tree. That tree contained 
1 5 hollows and occurred in the > 84 cm diameter class 
(Fig. 2c). With the exception of Corymbia gummifera, 
there was a significant positive association between 
tree height and number of hollows per tree for each 
of the species (Table 3). However, the relatively low 
Spearman's rho values (Table 3) and the scatter plots 
(Fig. 3) illustrate that the relationship between tree 
height and number of hollows was weak. 

The proportion of trees that were hollow- 
bearing (i.e. at least one hollow) increased with 
increasing trunk diameter, and at least 80% of trees 
with diameters > 84 cm were hollow-bearing (Fig. 
4). The proportion of Eucalyptus haemastoma trees 
that were hollow-bearing was always greater than 
60%, irrespective of diameter size class (Fig. 4). For 
A. costata and E. haemastoma all trees > 68 cm in 
diameter were hollow-bearing (Fig. 4). Eucalyptus 
pilularis was the only species examined that had 
no hollow-bearing trees in the smallest (20-35 cm) 
diameter class (Fig. 4). Only a small number of E. 
pilularis individuals were sampled in the 52-67 cm 
and the 68-83 cm diameter classes (Fig. 4), thus any 
inferences about the lack of hollow-bearing trees in 
these size classes are tentative. 



The mean number of 
hollows per tree increased 
with increasing trunk 
diameter and all of the five 
species had, on average, 
five or more hollows per 
tree once their diameter 
was > 84 cm (Fig. 5). 
With the exception of E. 
haemastoma, trees with 
diameters between 20 cm 
and 5 1 cm had, on average, 
fewer than two hollows per 
tree (Fig. 5). Compared 
to the other species 
examined, E. haemastoma 
had a relatively high mean 
number of hollows per 
tree in the three smallest 
diameter classes (Fig. 5). In contrast, E. pilularis had 
a relatively low mean number of hollows in all but the 
largest size class. Vox Angophora costata. Eucalyptus 
punctata and Corymbia gummifera, the greatest 
increase in the mean number of hollows (more than 
double) occurred between the 68-83 cm and the > 84 
cm diameter classes (Fig. 5). The number of hollows 
per tree increased with tree height for A. costata, 
E. punctata and C. gummifera (Fig. 6). All of the 
Eucalyptus haemastoma trees that reached > 16 m in 
height had an average of five hollows per tree (Fig. 
6). 

Entrance size and location of hollows 

Overall (across all species) there was a much 
higher prevalence of hollows with small entrances 
(47%)) compared to those with medium (26%) and 
large (26%) entrances. Angophora costata had the 
highest proportion of hollows with small entrances, 
but the lowest proportion of hollows with medium 
and large entrances (Fig. 7). Eucalyptus pilularis 
had the lowest proportion of hollows with small 
entrances, while it had the highest proportion of 
hollows with large entrances (Fig. 7). However, the 
number of hollow-bearing E. pilularis trees sampled 
was relatively small. The other three species had 
similar hollow entrance size distributions, except that 
E. punctata had a higher proportion of hollows with 
large entrances compared to that of E. haemastoma 
and C. gummifera (Fig. 7). 

There was a much higher prevalence of hollows 
in branches (84%) than in main stems (16%). 
Angophora costata (26 %) and E. punctata (24 %) had 
the highest proportion of main stem hollows, followed 
by C. gummifera (19 %), E. haemastoma (10 %) and 



6 



Proc. Linn. Soc. N.S.W., 128, 2006 



(a) 

16i 

■b 12 

I 10 



a> 

X) 

E 

Z3 
Z 



4- 

2 


-2 



(c) 



o 



o 



CD 

XI 

E 

3 
Z 



(e) 

ISt 

4s 12 






10 
84 
6 



4 

E 

3 

=^ 04 

-2 



P. TODARELLO AND A. CHALMERS 



(b) 

16- 



0, 14 
*s 12 



a B 



20 40 60 80 100 120 140 160 

Trunk diameter at breast hdght (cm) 



lb 






B 






14 












12 












10 






a 




a 


8 






a 






6 








mo 


S 




4 




B 


a 






2 






a B 






n 




B 


B B 






-2 













20 40 60 80 100 120 140 160 

Trunk diameter at breast height (cm) 



^ 10 

o 



o 

x: 

JQ 

E 

Z 



a 09 
B amao 



20 40 60 80 100 120 140 160 

Trunk diameter at breast height (cm) 



(d) 






o 

I 

E 

3 



16 

14 

12 

10 

8 

6 

4 

2 


-2 



O D a O D 

» DO OD 

cm m a 
oa c M a 
m flovM «n o 
■DK mm a 



20 40 % 80 100 120 140 160 

Trunk diameter at breast height (cm) 



20 40 60 80 100 120 140 160 

Trunk diameter at breast height (cm) 



Figure 2. Scatterplots of number of hollows per tree against trunlt diameter at breast lieight for (a) An- 
gophora costata; (b) Eucalyptus punctata; (c) Eucalyptus pilularis; (d) Eucalyptus haemastonta; and (e) 
Corymbia gummifera. 



Proc. Linn. Soc. N.S.W., 128, 2006 



HOLLOW-BEARING TREES 



(a) 





16 


0) 


14 


0) 




is 


12 


*-* 




(0 

5 


10 


o 






B 


o 




^ 




>*- 


fi 


o 




k. 




O) 


4 


^ 




E 


? 


3 




7- 






U 




-2 



(c) 



0) 
0} 



(0 

o 



0) 

E 

3 



16 

14- 

12- 

10- 

8- 

6 

4 

2 


-2 



(e) 





0} 



(0 

$ 

o 
o 

E 

3 



16-r 
14- 

12- 
10- 

8 

6 

4 

2 


-2 





O DDD OOOa DDO 

ODDO DO DODDODD O 
O D DD OOOOOOOO 

o o ODoaoaoDOODODOoao oooo o 



10 20 30 

Tree height (m) 



40 



OOO OOODOOD 



10 



20 



30 



40 



Tree height (m) 



O D D 

a ODD DO OD 

OOOO o 

O 00 o o o o 

D O O OO O O OD O O OD O O OO O D O 



10 



20 



30 



40 



(b) 



0) 
0) 



o 
o 

k- 
0) 

E 

3 



16 

14 

12 

10 

8 

6 

4 

2 


-2 



(d) 





le-j 


0) 

2 


14 
12 


o 
o 


10 
8 


o 


6 


k. 




4- 


E 

3 


2 


Z 







-? 



D O D D 

O D D 

DO DDDD DODO 

DOODDOODOaOGaDDOODOOO 



10 20 30 

Tree height (m) 



40 



O OOO 
D O a 000 DD D 
ODD OOODDD 
D OOOO 
DODO O D 
OO D O O 
I ODODODDO 



10 



20 



30 



40 



Tree height (m) 



Tree height (m) 



Figure 3. Scatterplots of number of hollows per tree against tree height for (a) Angophora costata; (b) 
Eucalyptus punctata; (c) Eucalyptus pilularis; (d) Eucalyptus haemastoma; and (e) Corymbia gummifera. 



8 



Proc. Linn. Soc. N.S.W., 128, 2006 



p. TODARELLO AND A. CHALMERS 



w 




(TJ 




0) 








■I-' 


^ 


(/) 


o 





o 


(1) 


sz 


*-> 


<D 


O 


C 

o 



c 
g 

o 

Q. 

o 



1.0 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1 
0.0 



d e 



a 



1= 111 99 "37 41 124 43 28 16 42 28 12 16 7 15 21 



20-35 



36-51 



52-67 



5 9 5 3 6 

68-83 



5 17 2 2 

84+ 



Trunk diameter size class (cm) 



Figure 4. Proportion of hollow-bearing trees in each diameter size class for the five species examined 
on the Central Coast of NSW (n = number of trees sampled). a=Angoph0ra costata; b = Eucalyptus 
punctata; c = Eucalyptus pilularis; d = Eucalyptus haemastoma; e = Corymbia gummifera. 



12 

0) 
0) 

-i= 10 



o 
o 

o 

E 

3 



8 
6 
4 
2 




f*!-. 



a 



d 



5i 



6 



i 



fe 



d 

if 
4 



b c « 

TTd* 



n= 111 99 37 41 124 

20-35 



43 28 16 42 28 12 16 7 15 21 5 9 5 3 6 



5 17 2 2 



36-51 



52-67 



68-83 



84+ 



Trunk diameter size class (cm) 



Figure 5. Mean number of hollows per tree in each diameter size class for the five species examined on 
the Central Coast of NSW (n = number of trees sampled). Vertical bars represent ± one standard error. 
a = Ang0phora costata; b = Eucalyptus punctata; c = Eucalyptus pilularis; d = Eucalyptus haemastoma; e 
= Corymbia gummifera. 



E. pilularis (9 %). The distribution of hollow entrance 
diameter sizes between branches and main stems was 
similar. Of the hollows occurring in branches, 26% 
had a large (> 10 cm) diameter entrance, 27% had a 



medium (6-10 cm) diameter entrance and 47% had a 
small (2-5 cm) diameter entrance; of those occurring 
in the main stem, 24% had large, 21%) had medium 
and 55%) had small entrances. 



Proc. Linn. Soc. N.S.W., 128, 2006 



HOLLOW-BEARING TREES 



(D 
0) 

-J= 5 



(0 

o 



0) 
Si 

E 



4 
3 

2 

i 
1 J 







d 



a be 



a 



be 



e 

a 



a c 
1 b 



e 1 



6 



c 
a 


r- S 

e ■. 


1 
]b 


I b 


J- 


L 


i 


d 





ed e 



1 22 

5-10 



24 1 77 93 

11-16 



77 35 48 4 57 ^ 23 59 33 26 



17-22 



23-28 



50 63 5 

29+ 



Figure 6. Mean number of hollows per tree in each height size class for the five species examined on the 
Central Coast of NSW (n = number of trees sampled). Vertical bars represent ± one standard error, a = 
Angophora costata; b = Eucalyptus punctata; c = Eucalyptus pilularis; d = Eucalyptus haemastoma; e = 
Corymbia gummifera. 



I 

O 







O 

c 
o 

■c 
o 

Q. 
O 



1.0 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1 
0.0 



70 



i^ 



^ 



Y- 



41 






18 



^• 



^• 



.^^ 



\^' 



80 



V« 



10^ 



^' 






^' 



s> 



3 a 

a a ^ 

b _ 

_ -L ~ V ^ 

" c •= _ - 

T c <= 



45 



# 



V<^^ 



^ 



^ 



Figure 7. The proportion of hollow-bearing trees with hollows in each entrance size class for the five 
species examined on the Central Coast of NSW (n = number of trees sampled), a = small (2-5 cm); b = 
medium (6-10 cm); c = large (>10 cm). See text for full specific names. 



10 



Proc. Linn. Soc. N.S.W., 128, 2006 



p. TODARELLO AND A. CHALMERS 



DISCUSSION 

Abundance of hollows 

Most studies of tree hollows use ground-based 
surveys because climbing trees to measure and record 
hollow dimensions is impractical (Lindenmayer et al. 
1990b; Gibbons et al. 2002) unless a double sampling 
method is employed (see Harper et al. 2004). Many 
entrances in trees observed from the ground are blind 
(i.e. not leading to a cavity suitable for occupation) 
and thus it is likely that the number of hollows, 
especially small hollows, is often overestimated 
(Lindenmayer et al. 1990b). On the other hand. Harper 
et al. (2004) demonstrated that, on average, ground- 
based observers correctly identify hollow-bearing 
trees (where hollows are at least 5 cm deep and have 
an entrance diameter > 1 cm) 82 % of the time and 
that hollow frequency is likely to be systematically 
underestimated. In the current study, it is likely that 
the number of hollows suitable for fauna have been 
overestimated because of the large proportion of small 
hollows encountered and the greater likelihood that 
small hollows are blind. Therefore, counts of hollows 
should only be regarded as an "index of hollow 
availability" (Gibbons and Lindenmayer 2002). 

Consistent with previous studies on eucalypts 
(Lindenmayer et al. 1993a, 2000; Bennett et al. 
1994; Gibbons 1994; Gibbons and Lindenmayer 
1996, 2002; Williams and Faunt 1997; Gibbons et 
al. 2000; Wormington et al. 2003), hollow number 
per tree increased with increasing tree diameter. 
Older, larger trees are more likely to contain hollows 
because they are more likely to be repeatedly exposed 
to events that encourage hollow development, while 
the decline in grovi^h rate with age is associated with 
branch shedding, a reduced ability to occlude wounds 
and an increased chance of heartwood being exposed 
as sapwood thickness decreases (Gibbons et al. 2000; 
Gibbons and Lindenmayer 2002). 

All of the five species in the current study, with 
the exception of Eucalyptus pilularis, had hollow- 
bearing frees in the smallest (20-35 cm) diameter size 
class. However, this does not mean that these hollows 
are suitable for occupation by fauna. For example, 
hollow-bearing frees with many hollows are more 
likely to be occupied by fauna (Gibbons et al. 2002). 
Thus, smaller diameter trees may be less likely to be 
occupied because they are more likely to have a lower 
mean number of hollows per free (less than two in 
this study) compared to that of the larger diameter 
trees (five or more hollows per tree, when dbh > 84 
cm). Small diameter frees may also have a smaller 
number of hollows with large entrances (Wormington 
et al. 2003) and therefore will suit a narrower range 



of fauna. The relatively small eucalypt species, E. 
haemastoma, had as many as 60% of frees being 
hollow-bearing in the 20-35 cm diameter class. This 
result for E. haemastoma supports Gibbons and 
Lindenmayer (2002) who stated that in regard to 
hollow formation it is the relative diameter of frees 
within a species that is important rather than absolute 
diameter. 

The current study found a weak positive 
association between number of hollows per tree and 
free height in four of the five species examined. In 
contrast, Lindenmayer et al. (2000) found that hollow 
number decreased with increasing tree height. Their 
findings may be due to frees in the later stages of 
senescence having a large number of hollows but 
were shorter because the tops of their main stem had 
broken off (Lindenmayer et al. 2000). In our study, 
only live frees were sampled and the shorter frees 
belonged to species typically found in nutrient-poor 
habitats. 

Differences in the propensity of species to form 
hollows 

Similar to previous studies (Bennett et al. 
1994; Gibbons 1994; Lindenmayer et al. 1993a, 2000; 
Gibbons and Lindenmayer 1996; Gibbons et al. 2000; 
Wormington et al. 2003), our study found differences 
between tree species in abundance of hollows. 
Species differed in the proportion of trees that were 
hollow-bearing, the density of hollow-bearing frees at 
the stand level and in the mean number of hollows per 
hollow-bearing tree. The proportion of trees (stems > 
68 cm) that were hollow-bearing ranged from 63 % 
for C. gummifera to 100 % for E. haemastoma and A. 
costata. Similarly, Bennett et al. (1994) found that the 
proportion of hollow-bearing frees (stems > 70 cm) 
in the six eucalypts they examined on the northern 
plains of Victoria ranged from 55 % to 100 %. 

Gibbons and Lindenmayer (2002) suggest that 
frees that "do not reach large diameters, regardless of 
longevity, are only infrequently observed to contain 
hollows" and frees < 30 cm dbh rarely contain 
hollows. This was not the case for E. haemastoma in 
the current study. Eucalyptus haemastoma was the 
shortest of the five species examined, and its diameter 
did not exceed 85 cm. Being a smaller species, the 
diameter of a mature E. haemastoma free would be 
less than that of the other species surveyed. Therefore, 
smaller diameter E. haemastoma frees are likely to 
have greater susceptibility to fungal decay. 

The proportion of A. costata frees that were 
hollow-bearing was relatively high. The heartwood 
of A. costata is "not durable" (Boland et al. 1984), 
which is consistent with the high number of hoUow- 



Proc. Linn. Soc. N.S.W., 128, 2006 



11 



HOLLOW-BEARING TREES 



bearing trees observed in this species. Gibbons and 
Lindenmayer (2002) argue that trees with a poor 
resistance to decay may not be good hollow producers, 
largely because "a rapid progression of decay may 
reduce the length of time that hollows persist before 
the supporting branches fail". Low resistance to decay 
in^. costata and E. haemastoma may explain the low 
proportion of large hollows in these two species, as 
branches may fail before the small hollows have time 
to enlarge. 

Eucalyptus pilularis was the largest species in 
this study, but it was also the most variable in size 
due to few individuals being sampled in the 52-67 cm 
and 68-83 cm diameter classes. Eucalyptus pilularis 
had a relatively low density of hollow-bearing 
trees, a low proportion of hollow-bearing trees and 
a moderate number of hollows per tree. None of the 
sampled E. pilularis trees that were < 36 cm dbh were 
hollow-bearing, while most trees > 84 cm dbh were 
hollow-bearing and often contained many hollows. 
Similarly, Mackowski (1984) found that E. pilularis 
individuals with a diameter less that 100 cm have 
very few holes, while the number of hollows per 
tree increases above this size. Heartwood decay is 
one of the essential precursors for hollow formation 
(Gibbons and Lindenmayer 2002) and the durability 
of the heartwood of E. pilularis is reported to be 
"moderate to good" (Boland et al. 1984). 

Similar to E. pilularis, E. punctata had a 
relatively low proportion of hollow-bearing trees 
and a relatively high proportion of hollows with 
large entrances. Boland et al. (1984) report that the 
heartwood of E. punctata is "extremely durable", 
which may explain its low proportion of hollow- 
bearing trees and paucity of trees with multiple 
hollows. High resistance to decay may also explain 
the higher proportion of hollows with large entrances, 
as branches may be less likely to fail before the 
hollows have time to enlarge. 

Corymbia gummifera is a medium-sized tree 
(11-22 m), which was similar to E. haemastoma in 
that its diameter did not exceed 86 cm. Corymbia 
gummifera had a relatively low proportion of hollow- 
bearing trees and a low number of hollows per 
tree. The heartwood of C. gummifera is "extremely 
durable" and the species also has flaky tessellated 
bark (Boland et al. 1984). These characteristics may 
protect C. gummifera from damage by fire and decay 
processes that lead to hollow formation and at least 
partially explain its lower propensity to form hollows. 
In a study in south-eastern Queensland, Wormington 
et al. (2003) suggested that the good occlusion 
ability of Corymbia citriodora may explain the low 
number of hollows observed in trees < 90 cm dbh. 



The occlusion ability of the species in our study is 
not known. 

Differences between species in the location and 
entrance size of hollows 

Consistent with Gibbons and Lindenmayer 
(2002), most hollows in this study occurred in 
branches rather than in main stems. Gibbons and 
Lindenmayer (2002) reported that main stem hollows 
accounted for 21-47% of hollows in open forest, 32% 
of hollows in tall, open forest and rarely occurred in 
woodlands. Hollows in branches accounted for 49- 
69% of hollows in open forest, 65% of hollows in 
tall, open forest and 91% of hollows in woodlands 
(Gibbons and Lindenmayer 2002). The distribution 
of hollow locations observed in this study (i.e. 16% 
of hollows in main stems and 84% in branches) is 
consistent with the mix of woodland and open 
forest habitats that were sampled. In agreement with 
Lindenmayer et al. (2000), branches and main stems 
in this study supported a fairly even distribution of 
hollows with small (2-5 cm), medium (6-10 cm) and 
large (> 10 cm) entrances. 

While both E. pilularis and E. punctata had a 
relatively low proportion of hollow-bearing trees 
they had a relatively high proportion of hollows with 
large entrances suitable for large owls, cockatoos 
and both large and small marsupials. Further, for E. 
pilularis those trees that were hollow-bearing tended 
to be of large diameter and have multiple hollows. 
Thus not choosing a species for retention or planting 
because of its lower propensity to form hollows may 
bias against certain groups of fauna; in this case 
the larger fauna species. Angophora costata had 
a relatively high proportion of hollows with small 
entrances that would suit smaller marsupials such as 
squirrel gliders, feathertail gliders and sugar gliders. 
Although the number of species that can use hollows 
with small entrance widths (2-5 cm) is limited, a 
study by Gibbons et al. (2002) in East Gippsland 
Victoria showed that they were an important hollow 
resource as they represented 25 % of all occupied 
hollows. Corymbia gummifera had a relatively even 
distribution of hollows with small, medium or large 
entrances and therefore had hollows with entrances 
suited to a wide range of fauna species. 

Hollow characteristics (i.e. number, density, size, 
spacing, location) are not the only factor to consider 
when choosing habitat trees for retention or planting. 
Although E. punctata was one of the species with 
a lower propensity to form hollows, it did contain 
hollows and is an important sap tree for the yellow- 
bellied glider (Goldingay 2000). Similarly, E. pilularis 
provides winter nectar and C. gummifera provides 



12 



Proc. Linn. Soc. N.S.W., 128, 2006 



p. TODARELLO AND A. CHALMERS 



sap and summer nectar for squirrel gliders (Smith and 
Murray 2003). 

In conclusion, the number of hollows per tree 
was positively related to tree size and clearly hollow 
abundance will be low where few large trees are 
found. Timber removal prior to 1967 in Bouddi 
National Park and prior to 1999 in Ourimbah State 
Forest may have affected the E. pilularis population 
sampled in this study. Thus the data shown here for E. 
pilularis is not representative of hollow availability in 
'undisturbed' vegetation, particularly as this species is 
likely to have been preferentially removed. The five 
tree species examined did differ in their propensity to 
form hollows. The relative diameter at which hollows 
form was shown to be important, with the smallest 
eucalypt species {E. haemastoma) observed to have 
a substantial number of hollows at tree diameters 
less than those often considered in hollow resource 
assessments. Given the diversity of hollows required 
by hollow-dependent fauna, and the many variables 
affecting the development of hollows, the retention 
of a mix of tree species should be favoured to supply 
this critical resource. 



ACKNOWLEDGMENTS 

We would like to thank the NSW National Parks 
and Wildlife Service and Forests NSW for permission to 
sample within their areas of jurisdiction. Particular thanks go 
to Adam Fawcett of Forests NSW The manuscript benefited 
from the comments of two anonymous reviewers. 



REFERENCES 

Ball, I.R., Lindenmayer, D.B. and Possingham, H.P. 

(1999). A tree hollow dynamics simulation model. 
Forest Ecology and Management 123, 179-194. 

Bennett, A.F., Lumsden, L.F. and Nicholls, A.O. (1994). 
Tree hollows as a resource for wildlife in remnant 
woodlands: spatial and temporal patterns across 
the northern plains of Victoria, Australia. Pacific 
Conservation Biology 1, lll-lTiS. 

Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall, 
N., Hyland, B.PM., Johnston, R.D., Kleinig, D.A. 
and Turner, J.D. (1992). 'Forest Trees of AustraHa'. 
(CSIRO Publishing: Victoria). 

Cockbum, A.C. and Lazenby-Cohen, K. (1992). Use of 
nest trees by Antechinus stuartii, a semelparous 
lekking marsupial. Journal of Zoology 226, 657-680. 

Cork, S.J. and Catling, PC. (1996). Modelling 

distributions of arboreal and ground-dwelling 
mammals in relation to climate, nutrients, plant 
chemical defences and vegetation structure in the 



eucalypt forests of southeastern Australia. Forest 
Ecology and Management 85, 163-175. 

Eyre, T.J. and Smith, A.P. (1997). Floristic and structural 
habitat preferences of yellow-bellied gliders 
(Petarus australis) and selective logging impacts in 
southeast Queensland, Australia. Forest Ecology and 
Management 98, 281-295. 

Gibbons, P. (1994). Sustaining key old growth 

characteristics in native forests used for wood 
production: retention of trees with hollows. In: 
'Ecology and Sustainability of Southern Temperate 
Ecosystems' (Eds S.R. Dovers and T.W. Norton) pp. 
59-84. (CSIRO PubUshing: Tasmania). 

Gibbons, P. and Lindenmayer, D.B. (1996). Issues 
associated with the retention of hollow-bearing 
trees within eucalypt forests managed for wood 
production. Forest Ecology and Management 83, 
245-279. 

Gibbons, P. and Lindenmayer, D.B. (2002). 'Tree Hollows 
and Wildlife Conservation in Australia'. (CSIRO 
Publishing: Victoria) 

Gibbons, P., Lindenmayer, D.B., Barry, S.C. and Tanton, 
M.T (2000). Hollow formation in eucalypts from 
temperate forests in southeastem Australia. Pacific 
Conservation Biology 6, 218-228. 

Gibbons, P., Lindenmayer, D.B., Barry, S.C. and Tanton, 
M.T. (2002). Hollow selection by vertebrate fauna 
in southeastem Australia and implications for forest 
management. Biological Conservation 101, 1-12. 

Goldingay, R.L. (2000). Use of sap trees by the yellow- 
bellied glider of the Shoalhaven region of New 
South Wales. Wildlife Research 11, 217-222. 

Harper, M.J., McCarthy, M.A., van der Ree, R. and Fox, 
J.C. (2004). Overcoming bias in ground-based 
surveys of hollow-bearing trees using double 
sampling. Forest Ecology and Management 190, 
291-300. 

Harper, M.J., McCarthy, M.A. and van der Ree, R. (2005). 
The abundance of hollow-bearing trees in urban dry 
sclerophyll forest and the effect of wind on hollow 
development. Biological Conservation 122, 181- 
192. 

Lower Hunter Central Coast Regional Environmental 
Management Sfrategy (2000). 'Vegetation Survey, 
Classification and Mapping: Lower Hunter and 
Central Coast Region'. Version 1.2. (National Parks 
and Wildlife Service: Sydney). 

Lindenmayer, D.B. (1997). Difference in the biology 

and ecology of arboreal marsupials in southeastem 
Australian forests and some implications for 
conservation. Journal of Mammalogy 78, 1 1 17-1 127. 

Lindenmayer, D.B., Cunningham, R.B., Donnolly, C.F. 
and Tanton, M.T. (1993a). The abundance and 
development of hollows in Eucalyptus trees: a 
case study in the montane forests of Victoria, south 
eastem Australia. Forest Ecology and Management 
60, 77-104. 

Lindenmayer, D.B., Cunningham, R.B. and Donnolly, C.F. 
(1993b). The conservation of arboreal marsupials in 
the montane ash forests of the central highlands of 



Proc. Linn. Soc. N.S.W., 128, 2006 



13 



HOLLOW-BEARING TREES 



Victoria, south-eastern Australia. The presence and 
abundance of arboreal marsupials in retained linear 
habitats (wildlife corridors) within logged forests. 
Biological Conservation 66, 207-22 1 . 

Lindenmayer, D.B., Cunningham, R.B. and Dormelly, C.F. 
(1994). The conservation of arboreal marsupials in 
the montane ash forests of the central highlands of 
Victoria, south-eastern Australia. The performance 
of statistical models of the nest tree and habitat 
requirements of arboreal marsupials applied to new 
survey data. Biological Conservation 70, 143-147. 

Lindenmayer, D.B., Cunningham, R.B., Pope, M.L., 

Gibbons, P. and Donnelly, C.F. (2000). Hollow sizes 
and types in Australian eucalypts from wet and dry 
forest types - a simple rule of thumb for estimating 
size and number of hollows. Forest Ecology and 
Management Ml , 139-150. 

Lindenmayer, D.B., Cunningham, R.B., Tanton, M.T and 
Smith, A.P. (1990a). The conservation of arboreal 
marsupials in the montane ash forests of the central 
highlands of Victoria, south-east Australia. The 
loss of trees with hollows and its implications 
for the conservation of Leadbeater's Possum 
Gymnobelideus leadbeateri McCoy (Marsupialia: 
Petauridae). Biological Conservation 54, 133-145. 

Linderunayer, D.B., Cunningham, R.B., Tanton, M.T., 

Smith, A.R and Nix, H.A. (1990b). The conservation 
of arboreal marsupials in the montane ash forests 
of the central highlands of Victoria, south-east 
Australia. Factors influencing the occupancy of trees 
with hollows. Biological Conservation 54, 111-131. 

Lindenmayer, D.B., Cunningham, R.B., Tanton, M.T, 

Smith, A.P. and Nix, H.A. (1991). The conservation 
of arboreal marsupials in the montane ash forests 
of the central highlands of Victoria, south-east 
Australia. The habitat requirements of Leadbeater's 
Possum Gymnobelideus leadbeateri and models of 
the diversity and abundance of arboreal marsupials. 
Biological Conservation 56, 295-315. 

Mackowski, CM. (1984). The ontogeny of hollows in 

Blackbutt {Eucalyptus pilularis) and its relevance to 
the management of forests for possums, gliders, and 
timber. In: 'Possums and Gliders' (Eds A.P. Smith 
and I.D. Hume) pp. 553-567. (Surrey Beatty and 
Sons; Sydney). 

Murphy, C.L. (1993). 'Soil Landscapes of the Gosford- 
Lake Macquarie 1:100 000 Sheet'. (Department of 
Conservation and Land Management: Sydney). 

Norton, T.W. (1987) The ecology of small mammals in 
north-eastern Tasmania. Rattus lutreolus. Australian 
Wildlife Research 14, 415-433. 

Perry, D.H., Lenz, M. and Watson, J.A.L. (1985) 

Relationship between fire, fiingal rots and termite 
damage in Australian forest trees. Australian 
Forestry 48, 46-53. 

Smith, A.P. and Murray, M. (2003). Habitat requirements 
of the squirrel glider {Petaurus norfolcensis) and 
associated possums and gliders on the NSW Central 
Coast. Wildlife Research 30, 291-301. 



Strom, B. (1986) 'Bouddi Peninsular Study'. (Association 
for Environmental Education (NSW), Central Coast 
Region: Gosford). 

Whitford, K.R. (2002). Hollows in jarrah {Eucalyptus 
marginata) and marri {Corymbia calophylla) 
trees. Hollow sizes, tree attributes and ages. Forest 
Ecology and Management 160, 201-214. 

Wilkes, J. (1982) Stem decay in deciduous hardwoods 
- an overview. Australian Forestry 45, 42-50. 

Williams, M.R. and Faunt, K. (1997). Factors affecting 
the abundance of hollows in logs in jarrah forest 
of south-western Australia. Forest Ecology and 
Management 95, 153-160. 

Wormington, K.R., Lamb, D., McCallum, H.I. and 
Moloney, D.J. (2003). The characteristics of six 
species of hollow-bearing trees and their importance 
for arboreal marsupials in dry sclerophyll forests of 
southeastern Queensland, Australia. Forest Ecology 
and Management 182, 75-92. 



14 



Proc. Linn. Soc. N.S.W., 128, 2006 



The Vegetation History of the Holocene at Dry Lake, Thirlmere, 

New South Wales 

Suzanne Rose and Helene A. Martin 

School of Biological, Environmental and Earth Science, University of New South Wales, Sydney AustraHa 

2052 (h.martin@unsw.edu.au) 



Rose, S. and Martin, H.A. (2007). The vegetation history of the Holocene at Dry Lake, Thirlmere, New 
South Wales. Proceedings of the Linnean Society of New South Wales 128, 15-55. 

At the beginning of the Holocene, Dry Lake was a lake, with a fringe of cyperaceous reeds. Eucalyptus 
and Allocasuarina were the dominant trees and Asteraceae Tubuliflorae were prominent in the understorey. 
Between 8 ka and 2 ka, the lake became shallower, and the reeds grew over the surface of the developing 
swamp, forming peat. An hiatus in peat deposition between 5 ka and 2ka was followed by the formation 
of a thin layer of diatomite. Eutrophic conditions would be required to allow large populations of diatoms 
and burning seems the most likely way of increasing the nutrient mobility on the poor sandstone soils of 
the catchment. 

By 2 ka, the lake had become a peat swamp. Angophora/Corymbia pollen had increased dramatically, 
most likely representing Angophora on these alluvial flats. The shrub layer had also become more diverse. 
Allocasuarina did not decrease through the Holocene, unlike the record of many other Holocene sites. 
The likely reasons for this difference are probably related to site-specific environmental conditions. With 
European settlement, all trees decreased dramatically and grasses increased. Today, Dry Lake only contains 
water in exceptionally wet periods. 

Manuscript received 23 January 2006, accepted for publication 1 8 May 2006. 

KEYWORDS: Casuarina/Allocasuarina decline, freshwater sponge, Holocene, palynology, Thirlmere 
Lakes, vegetation history. 



INTRODUCTION 

Dry Lake is one of a series of freshwater lakes 
associated with an incised former river valley at 
Thirlmere. The Thirlmere Lakes are rare examples of 
very old, small lakes that have aged very slowly as a 
result of the stable geological nature and small size 
of the catchment (Horsfall et al., 1988). The initial 
development of the lakes was related to tectonic 
activity associated with the formation of the Lapstone 
Monocline, Kurrajong and Nepean Faults which 
beheaded a river that probably originally flowed 
westwards, leaving the isolated, sinuous channel that 
now contains the lakes (Timms, 1992). At this time, 
or sometime later, the drainage direction changed and 
today Dry lake drains in a north-easterly direction 
along Cedar Creek and the Thirlmere Lakes drain 
westwards along Blue Gum Creek (Fig. 1). Presently, 
Dry Lake only contains water intermittently in wet 
years, when the water depth approximates 60 cm. 

The basal sediments of Dry Lake have been 
radiocarbon dated at about 10,000 years before 



present which approximates the beginning of the 
Holocene when the climate had mainly recovered 
from the peak of the glacial period but there was a lag 
in the recovery of the vegetation. The Holocene thus 
records the establishment of the present vegetation. 
Allocasuarina/Casuarina was usually more 
prominent following the last glacial period but during 
the Holocene, it declined and Eucalyptus I Corymbia 
rose to dominance (Clarke, 1983). 

This paper presents the Holocene history of the 
vegetation at Dry Lake and compares it with Lake 
Baraba, one of the Thirlmere Lakes (Black et al., in 
press), some 4 km to the south. 



THE ENVIRONMENT 

The former river valley that includes Dry Lake is 
incised in Hawkesbury Sandstone, but the surrounding 
higher plateau surfaces retain cappings of Ashfield 
Shale, the lower member of the Wianamatta Group, 
which overlies the Hawkesbury Sandstone. Both 



VEGETATION HISTORY OF DRY LAKE, NSW 




Figure 1. Locality map. 



formations are of Triassic age (Herbert, 1980). 

Although Hawkesbury Sandstone dominates 
the landforms and soils occurring in the study area, 
there is a shale outcrop on the ridge to the east of Dry 
Lake (Fig. 1). This may be either the Ashfield Shale 
or a shale lens in the sandstone, as it is very close 
to the contact between the two formations. The soils 
developed on the sandstone are uniform sandy loams 
with some organic staining in the upper horizons. 
They are acid, of low nutrient status and low water 
retaining capacity, and vary in depth and drainage, 
depending on the topography. 

Thirlmere has warm to hot summers and cool to 
mild winters (Bureau of Meteorology website, BoM, 
2005). The average annual rainfall of the nearest 



station, Picton, is 820 mm and it is received in two 
relatively wet periods from January to March and in 
June. The median rainfall for each month is greater 
than 25 mm (BoM, 2005). There is considerable 
variation in rainfall and after long dry spells, the 
Thirlmere Lakes dry out. Known dry lake stages 
occurred about 1902, 1929 and 1940 (Rose, 1981). 

The Thirlmere Lakes were first sighted by 
Europeans in 1798. By 1833, many settlers had 
arrived at the Oaks, north of Thirlmere, and mixed 
farming flourished. Timber cutters logged mostly 
Eucalyptus deanei, especially along Blue Gum Creek 
(Woods, 1980). Most of these activities were more 
intense on the better soils of the shale areas and on the 
alluvial flats. Present land uses consists of residential 



16 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



Roads 

Streams 

Fence 






1 


• 




\ 


:/ 


y^ 




• 


/ 


1 "X 


i/ 




<< 

< 

v< 


V, *?=\ 

V', P2a, ?b 3 41 . 
A5x'i,a Jf 


a;/ 




t \ 


A6 \",.„ \ 

xr \ 

s', \ 


:/ 


-I ~^ 

■if. Core for pollen analysis 


V', \ 

\ \Va8 \ 


\/ 


Transect 


\ V» \ 




▲ Vegetation quadrats 


\ ^'" \ 


■■/ 


• Auger holes 
D Pits 


\ A9«"4, \ 


V 


V'v Reed beds 




• 


•f^, Channel 


^\ ""-'', 1 


• 
• 


1W1 :vi 


Om ^s/^jX 






, 







Figure 2. Dry Lake and environs. 



housing and small farms. 

Four main vegetation units described by Pidgeon 
(1937; 1941) apply to this area and they are: 1), the 
Eucalyptus deanei and Eucalyptus data tall open forest 
in the gullies; 2), the Mixed Eucalyptus Association of 
the ridges and slopes; 3), the Angophora floribunda/ 
Melaleuca linariifolia forest of the alluvial fans and 
4), the aquatic vegetation of swamps and lakes. The 
mixed Eucalyptus Forest Association constitutes 
the major part of this study area. The National Park 
was completely burnt in 1955 and has suffered 
considerable damage from local fires since then (R. 
Kinntish, pers. comm.). Dry Lake has been cleared 
of native vegetation and is predominantly a grassland 
but it is assumed that the native vegetation would 
have been much the same as that in the surrounding 
areas. 



METHODS 

Field work was carried out during 1981 when 
the vegetation survey was undertaken. (Appendix 1). 
Quantitative data on plant distributions were obtained 



from quadrats along transects (Rose, 1981). The 
vegetation map was prepared using aerial photographs 
and the field survey (Figs 1,2). 

The stratigraphy of Dry Lake sediments was 
investigated by auger holes and two pits (Fig. 2) 
which were limited to a depth of 1.5 m by heavy 
clay. A core, located near the centre of the lake but 
where there was minimal disturbance and away from 
local pollen sources, was chosen for pollen analysis. 
This core was taken using a Hiller corer. Samples 
for radiocarbon dating were taken from the pits and 
analysed by the then Radiocarbon Laboratory of the 
University of New South Wales. All of the samples 
were stored in a 4°C cold room to suppress microbial 
growth until work could proceed. 

Surface samples of soils, mosses and lake 
sediments were collected (Figs 1, 2) to study pollen 
deposition under the present vegetation and assist in 
the interpretation of pollen in the core. Lake surface 
sediments were sampled using a weighted cylinder on 
a line. 

Sediment samples 2 cm in length and 3 cm apart 
were taken from Pit 1 for organic matter analysis. 
Duplicated samples were oven-dried (105°C) and 



Proc. Linn. Soc. N.S.W., 128, 2007 



17 



VEGETATION HISTORY OF DRY LAKE, NSW 



ignited in a muffle furnace to 500°C. During ignition, 
structurally bound water is lost, but in highly organic 
sediments, the major loss on ignition is from the 
organic matter (Bengtsson and Enell, 1990). 

Pollen preparations from the sediment core were 
spiked with Alnus of a known concentration, treated 
with hydrofluoric acid to remove siliceous material, 
boiled in 10% sodium hydroxide to remove of 
humic acids, disaggregated with ulfrasonic vibration, 
followed by standard acetolysis (Moore et al., 1991). 
Surface samples were treated in the same way, with 
the addition of sieving to remove sand, leaves, twigs 
etc. and omitting the Alnus spike. Reference pollen 
used for identification was only treated with standard 
acetolysis. The residues were mounted in silicone oil 
(viscosity of 2,000 centistokes) or glycerine jelly, 
using grade coverslips. 

Siliceous fossils were recovered from a 
known volume of sediment using an acid sequence 
(hydrochloric, nitric and sulphuric acids) and then 
dehydrating the residue in absolute alcohol. The 
residue was made up to a known volume, and with 
constant agitation, a known aliquot was extracted, 
the alcohol allowed to evaporate and mounted in 
Napthrax in toluene (Lacey, 1963). 

Pollen was identified by comparison with a 
reference collection using the x 1000 magnification 
objective. Where it was not possible to identify some 
grains, they are listed as unknowns. The pollen of 
members of the family Myrtaceae is similar and it 
requires a careful analysis of the finer morphological 
features to separate them (Chalson and Martin, 
1995). In this study, three groups were distinguished: 
Angophora/Corymbia, Eucalyptus and Melaleuca/ 
Leptospermum (Appendix 2). The name on the pollen 
diagram and probable source in the vegetation is 
listed in Appendix 3. 

Pollen was counted using the x 400 objective of 
a Zeiss microscope. Tests to assure an adequate count 
showed 1 60-200 grains was sufficient. Some samples 
had insufficient pollen for an adequate count, and this 
is indicated on the pollen diagram. 

Counts were made of sponge spiclues on a Zeiss 
microscope, using the x 250 and x 400 objectives. 
Only one species of sponge was present and the 
spicules consisted of megascleres, gemmoscleres 
and fragments of both. Three quarters of a sclere was 
counted as a whole sclere and more than one quarter as 
a fragment. Counts were made along transects spaced 
evenly across the slide to ensure a representative 
count. Knowing the ratio of the area counted to total 
area of the slide, and the volume of the aliquot to 
total volume of residue, the counts were converted to 
numbers of scleres per volume of sediment. 



RESULTS 

The Lake Sediments. 

The sequence of sediments in the central part of 
Dry Lake is as follows, from top to bottom (Fig. 3): 

l)Alayer of little decomposedfibrous peat, composed 
of rhizomes, root and stems of cyperaceous reeds 
(probably Lepironia articulatd). 

2) Well humified black clayey peat with abundant 
roots, rhizomes and seeds of cyperacous reeds. 

3) A fine sandy clay, yellowish in colour and with 
sharp upper and lower boundaries. When dry, this 
material was light and powdery. This description 
resembled that of diatomite (Birks and Birks, 
1980) and microscopic examination showed it 
consisted almost entirely of siliceous sponge 
spicules, diatoms and sand grains. Organic content 
was minimal. 

4) A sandy peaty clay may or may not be present, 
was usually less than 3 cm thick and had a diffuse 
lower boundary. 

5) A black clay layer with light to medium texture 
appeared highly organic. Boundaries between all 
clay layers were diffuse. 

6) This medium textured clay formed the bulk of the 
sediment sequence. It was a greyish brown colour, 
with frequent yellowish-red mottles. At certain 
depths, red (iron) sfreaks or small iron stained clay 
concretions (< 3 mm diameter) appeared along 
old roots or root chaimels. At greater depths, the 
concretions formed continuous blocky structures 
or large single blocks (< 2 cm diameter). 

7) The clay layer below was distinguished by its 
heavy texture and pallid colour, although the 
transition was diffuse. The clay may be mottled, 
but no concretions were found here. 

Each layer was represented across the whole of 
the lake and only changed significantly at the lake 
margins where the sediments showed alternate layers 
of sand, clay and peat, although some sand was almost 
always present (Fig. 3). 

Loss on ignition was an approximate guide to the 
organic matter content of the sediment. The values for 
the peat were high, around 45-60%, but fell to about 
1 5% in the diatomaceous layer. Values peaked at 25%) 
for the sandy peaty clay layer, then declinde to 10%) 
in the clay layers and finally dropped to 5-7% in the 
pallid clay (Fig. 3). 

Radiocarbon dates are shown in Table 1 and on 
Fig. 3. The lowermost date is 8,780 ±160 radiocarbon 
years, which corresponds to 9,791 calibrated years 
BP, the whole of the Holocene. Surface peat was 



18 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



KEY 



^ 



ooo 
ooo 



l-l 



• • • 



Black clayey peat 

Brown fibrous peat 

Sandy peat 

Peaty sand 

Dark sand 

Pallid sand 

Sandy clay 

Sandy peaty clay 

Fine sandy clay 'diatomKe' 

Medium organic clay 

Medium grey/brown clay 

Mottled clay 

Heavy pallid clay 

iron /day concretions 



Depth 

(cm) Auger 1 

rO 



20 



40 



60 



80 



100 



120 



-140 



^ 



rq 




S 



Auger 2 
3<3 




Auger 3 




— I 

3. 



1-1 
"i-i" 



D^ntifcm) 



% loss on ignition 



Auger 4 



i 



r •-•■ 







Pit 2a 



m 



ij 

_ •_ 
*■_■• 



Pit 2b 



m 



1-1 






^ Radiocarbon years. 

1. 920 ±80 

2. 1120 ±80 

3. 2170 ±100 

4. 5820 ±130 

5. 8780 ±160 



Piti 




00 



12 
3 



• 7 



Figure 3. The stratigraphy of the sediments of Dry Lake. 



Table 1. Radiocarbon dates. Calibrated years has been calculated according to the Radiocarbon 
Calibrated Program Calib Rev5.0.2 (Stuiver and Reimer, 1986-2005) 



Sample Depth (cm) 



Material dated 



Radiocarbon years BP 



Calibrated years BP 



Pitl 
10-20 



Black clayey peat (humic 
acid fraction) 



920±80 (NSW 375) 



795 



22-24 


Charcoal 

Humic acid in charcoal 

Peat (around charcoal) 


1,120±80(NSW381) 

1,560±120(NSW380) 

1,660±90(NSW384) 


986 

1,417 
1,499 


24-32 


Diatomite (humic acid 
fraction) 


2,170±100(NSW376) 


2,101 


38-45 


Organic clay (humic acid 
fraction) 


5,820±130(NSW377) 


6,570 


Pit 2 








73-83 


10 cm length of wood 
Wood (humic acid 
fraction) ^^H 


8,780±160(NSW387) 
8,060±300 (NSW 388) 


9,791 

8,899 ^^^^^^ 



Proc. Linn. Soc. N.S.W., 128, 2007 



19 



VEGETATION HISTORY OF DRY LAKE, NSW 




^ Closed sedgeland 



P^ 



Melaleuca linariifolia 
low closed or open forest 

Angophora fJoribunda 
woodland/open forest 



rni Mixed Eucalyptus/Corymbia 
woodland 



I 1 Mixed Eucalyptus/Corymbia 
— open forest 

FO Cleared land 



Gully forest 



^k Lake 

1 1 1 1 1 Ridgetop contour 



Figure 4. The vegetation of the Thirlmere Lakes region. 



littoral zone reflects deeper water where 
the dominant process is the settling of 
tine particle sizes. The peat indicates 
organic material accumulated more 
rapidly than it decomposed, reflecting a 
consistently high water table. 

From the beginning of the Holocene 
up to about 6-5,000 years ago, the site 
was a lake depositing clay. The pallid 
clay, the deepest layer, the mottling and 
the iron concretions in the layer above 
the pallid clay suggest a fluctuating 
water table and the lake may have dried 
out periodically. It is not clear what 
happened in the period 5-2,000 years 
BP, represented by the hiatus in the 
sediments. 

An explanation for the 5 cm. thick 
diatomite layer must remain speculative. 
An explosion in the diatom population 
would require a considerable quantity 
of nutrients, and it is unlikely that the 
sandstone substrate of the catchment 
could supply these nutrients. Burning 
appears the most likely way of 
increasing the nutrient mobility. 
Unfortunately, an hiatus provides no 
evidence at all. 

For the last 2-1,000 years, the 
lake has been shallow enough to allow 
the rooted swamp vegetation. The 
Holocene history is thus the evolution 
of a lake gradually filling up with 
sediments. 



deposited most rapidly, i.e., 24-26 cm in 1,100 years 
and the diatomaceous earth (about 5 cm in depth) 
represents another 1,000 years. There appears to be 
an hiatus in the sediments, representing a period of 
zero or minimal deposition, or a period of erosion, 
between the diatomaceous earth (2,170±100 C'* 
years, see Table 1 for calibrated years) and the organic 
clay (5,820±130 '''years) immediately below it. Wood 
and charcoal dates are regarded as the most reliable, 
whereas humic acids may move from their place of 
origin and contaminate material elsewhere. Table 1 
reveals that where charcoal and humic acids have 
been dated from the same stratigraphic layer, there is 
relatively little difference. 

Sedimentary history 

The alternation of fine clay, peat and coarse 
sediments on the lake margin reflects the advance and 
retreat of the littoral zone in response to fluctuating 
water levels. Increasing clay content away firom the 



The Vegetation 

Appendix 1 lists all the species in the study area 
and Fig. 4 shows the general distribution and extent 
of the vegetation units which are as follows: 
1 . Low closed forest with emergent trees dominated 
by Eucalyptus deanei (Fig. 5), up to 35 m tall. This 
unit is restricted to the floors and steep-sided gullies. 
Below this tall open forest canopy is a low closed 
forest with a great diversity of small rainforest 
trees, including Pomaderris spp., Backhousia 
myrtifolia, Acmena smithii, Doryphora sassafras, 
Ceratopetalum gummiferum. C. apetalum and 
Stenocarpus salignus. Below this is a closed scrub 
with many sclerophyllous species, including 
Grevillea mucronata, Leptospermum trinervium, 
Persoonia levis and Lomatia silaifolia. Abundant 
twiners are also present, including Smilax 
australis, Cissus antarctica and Sarcopetalum 
harveyanum. The ground cover is a closed fern/ 



20 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 




Figure 5. Low closed forest with emergent Eucalyptus 
species is restricted to the floor of steep gullies. 

herbland with Gleichenia microphylla, Blechnum 
nudum, Sticherus flabellatus, Drosera auriculata 3 
and many orchids. 

The gully is protected from wind and fire, and 
is moist and well shaded. The soils are of variable 
thickness and are highly organic. Other more open 
gullies have some of these characteristics but are 
dominated by sclerophyllous shrubs and do not 
have such a complex structure. 
2. Mixed Eucalyptus/Corymbia forest is the most 
extensive unit occupying the well drained slopes 
and ridges. The structure of the tree canopy is 
variable, with open forest on the more sheltered 
sites and south-facing ridges, with Eucalyptus 
piperita, E. resinifera, E. punctata, Corymbia 
gummifera and C. eximia (Fig. 6Aj. Low open 
forest and woodland occupies the steep slopes, 
especially those with a northern or westerly 
aspect and along stony areas of the central ridge. 
Woodlands occur on the most extreme sites with 



greatest exposure to westerly winds and 
excessive drainage and here C. eximia and E. 
racemosa are the main species, with minor 
occurrences of the species mentioned above. 
Small trees of Persoonia levis, P. linearis, 
Allocasuarina torulosa and Xylomelum 
pyreformis are occasionally found here. 

The understorey is typically an open 
heath, dominated by the families Proteaceae 
and Fabaceae (especially Acacia spp.). Other 
common species include Pimelea linifolia, 
Platysace linearifolia and Eriostemon spp. 
The shrub layer is diverse and highly variable, 
due to a complex of envirormiental factors. At 
sites impacted upon by recent fire, the shrub 
layer has reduced diversity and density. The 
main species are Acacia spp., Indigofora 
australis and Hibbertia aspera (Fig. 6B). The 
groundcover is a dense sward of Imperata 
cylindrica and Pteridium esculentum. 

The groundcover is generally open 

on ridgetops and steep rocky slopes and 

closed on the footslopes and nearer the lake 

margins. The herbs include Opercularia spp., 

Viola betonicifolia, Pratia purpurascens, 

Gonocarpus tetragynus and climbers Glycine 

clandestina and Kennedia rubicunda are 

more important on moister ground, including 

the alluvial areas adjacent to the lakes 

and southern facing slopes. In most other 

situations, grasses and Lomandra species 

predominate. Lomandra obliqua is common 

on well drained slopes and ridges and L. 

longifolia is abundant on the inoister foot 

slopes. 

Angophora floribunda dominated woodland and 

open forest is found on alluvial fans adjacent to 

and between lakes. A. floribunda is not common on 

other sites. Other tree species which are common 

at these sites include Eucalyptus resinifera, E. 

piperita and Corymbia gummifera. Smaller trees 

include Allocasuarina littoralis, Banksia serrata 

and Persoonia levis. 

The shrub layer of this woodland is an open 
heath similar to that of the Mixed Eucalyptus/ 
Corymbia Forest but Banksia spinulosa and 
Pultenaea villosa are often important components. 
P. villosa is generally restricted to these alluvial 
areas and may be locally dominant. 

Groundcover is usually closed grass/ 
herbland. On the most poorly drained sites, 
Lepidosperma longitudinale, Schoenus spp. and 
Baloskion gracilis are important. Alluvial fans are 
characterised by deep soils, gentle slopes and the 



Proc. Linn. Soc. N.S.W., 128, 2007 



21 



VEGETATION HISTORY OF DRY LAKE, NSW 



site drainage is moderate to poor. M 

4. Melaleuca linariifolia low closed or low 
open forest is mostly confined to a narrow 
area fringing the lake margins and along 
the swampy parts of Blue Gum Creek. It 
is most extensive on flat, low lying and 
periodically inundated sites adjacent to 
the lakes (Fig. 7). 

The canopy is dominated by M 
linariifolia, open or closed and 6 to 7 
m high. A. floribunda is often present, 
usually as saplings on the drier landward 
margins of the unit. A few shrubs occur, 
e.g. Viminaria juncea. Acacia longifolia 
and Pultenaea villosa. At some sites, this 
unit and the A . floribunda woodland/open 
forest are difficult to distinguish and the 
M. linariifolia low open forest gives way 
to Angophora forest as the site becomes 
less subjected to periodic inundation. 

The groundcover is closed or 
open sedgeland with Schoenus spp., 
Lepidosperma longitudinale and Juncus 
spp. The lakes contain occasional dead 
M linariifolia stumps in 2 m of water, 
indicating a period of low water level 
which must have existed long enough for 





Figure 6. Mixed Eucalptus/Corymbia forest. A, more sheltered site. B, a site with a reduced shrub layer 
and a ground cover oilmperata cylidrica and Pteridium esculentum, the result of fire in recent years. 



22 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 




Figure 7. Melaleuca linariifolia low closed or open forest at the lake edge. 



M linariifolia to become well established. 

5. A closed sedgeland occurs as a continuous fringe 
around and between each lake. Aquatic vegetation 
is usually distinctly zoned according to water 
depth, however, there is considerable overlap 
between species distribution as shown on Fig. 8 
and some species may change their distribution 
over time. For example, Vorst (1974) noted that 
Elaeocharis sphacelata grew on the landward side 
of Lepironia articulata, but it now has a patchy 
distribution on both the landward and lakeward 
sides of Z,. articulata. This may have been caused 
by the lowering of water levels since 1974. 

The distribution of aquatic plants, especially 
rhizomatous sedges is probably constantly 
changing with water level fluctuations and possibly 
competition. Lake 3 of Thirlmere Lakes (Lake 
Baraba) is almost dry and covered by an extensive 
sedgeland of Lepidosperma longitudinale onto 
which Melaleuca linariifolia is encroaching (Fig. 

9). 

6. Dry Lake. The land around Dry Lake is mostly 
cleared (see Fig. 2). The vegetation on the slopes 
adjacent to the lake is largely grassland/herbland 
with native species, e.g. Imperata cylindrica, 
Themeda australis, Goodenia hederacea, 
Pratia purpurescens and Wahlenbergia spp., 
and introduced species, including Paspalum 
dilatatum, Echinopogon spp., Setaria geniculata, 
Hypochoeris radicata, Plantago lanceolata, 
Conyza spp. and Verbena bonariensis. Regrowth of 



trees and shrubs is occurring over most of the land 
and is most advanced nearest the swamp where 
the soils are moister. 

Dry Lake itself is a swamp but it has 
surface water in wetter periods. It is surroimded 
by a discontinuous fringe of Leipidosperma 
longitudinale which appears to be advancing onto 
the swamp (Fig 10). A patchy cover of herbs on the 
lake includes Gonocarpus micranthus,Dichondria 
repens and some grasses. The wettest patches are 
almost bare apart from the occasional Polygonum 
decipiens and Hypochoeris radicata. In 1981, a 
dead reed, probably Lepironia articulata, covered 
most of the lake basin. Live rhizomes of the reed 
are abundant in the peat. A channel dug through 
the centre of the basin and containing about 60 
cm of water has some Eleocharis sphacelata, 
Potamogeton tricarinatus, Persicaria orientale 
and several other sedges. 

Towards the eastern margin of Dry Lake, 
inside the fringe of L. longitudinale, there are 
a number of old tree stumps, some of which 
are quite large (up to 40 cm in diameter). They 
are not Melaleuca linariifolia but are possibly 
Eucalyptus/Corymbia or Angophora spp. They 
probably represent a period of reduced moisture 
balance which was long enough and suitable for 
the growth of trees. 



Proc. Linn. Soc. N.S.W., 128, 2007 



23 



VEGETATION HISTORY OF DRY LAKE, NSW 





■-3 m 



2m 



Figure 8 Aquatic vegetation. 1, Brasenia schreberi. 2, Lepironia articulata. 3, Melaleuca linariifolia 
stump. 4, Eleocharis sphacelata. 5, Lepidosperma longitudinale. 6, Baloskion gracilis. 7, Schoenus brevi- 
folius and 5. melanostachys. 

Modern Pollen Deposition Birks, 1980; Dodson, 1983; Moore etal., 1991), hence 

Pollen is produced by the contemporaneous it is not possible to relate fossil pollen assemblages 

vegetation, but a multitude of factors affect the directly to the vegetation which produced it. 
representation of pollen in the sediments (Birks and 



24 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 




Figure 9. An extensive sedgeland on an almost dry Thirlemere lake. 



Samples taken from beneath the present vegetation 
give some information about modem pollen 
deposition which may be used for interpretation of 
the fossil assemblages. Table 2 describes the surface 
sample sites and the associated vegetation growing 
there, and Fig. 11 shows the surface sample pollen 
spectra. 



Comparison of the representation of the pollen 
with the taxon in the vegetation allows recognition 
of well-represented taxa, where pollen and vegetation 
representation are similar, over-representation, 
where pollen abundance exceeds abundance in the 
vegetation and under-representation, with pollen 
abundance less than abundance in the vegetation. This 




Figure 10. Dry Lake, showing the central drainage channel and marginal Leipidosperma longitudi- 
nale. The surrounding slopes are cleared. 



Proc. Linn. Soc. N.S.W., 128, 2007 



25 



VEGETATION HISTORY OF DRY LAKE, NSW 



1 I I . I 



LUJ 1 



O . 



O . 

o 

s 

CO 



(O 



CO 

(0 
0) 



E 
E 



x> 

c 

T3 

o 
o 

T3 

C 
(0 

o 

U- 



(0 

a) 
a. 

E 

CD 

CD 

_l 

o 



if 









'- CM CO t lO <D 



tx to ^ 




<0 U) 



26 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



j_L 



o , 



o 

s 

CO 



l_L 



2 
CO 

CO 

c 

3 

E 
E 

8 

•o 

c 
iS 
■o 

s 



■a 
c 

(0 

w 
2 



L± 



u. 



LO 
CO L 

CM ^ 

(D 
.^ 

(B 

_J - 

£ 

0) 

E - 



0) 
0) 
Q. 

E 

(0 
CO 



CO 



E 
2 

B 

c 

E 
1 

CO 



I I I I I 



i_4-^ LU4 m 



/. 
// 



/ 



o^ 












/ 



/ 



.^ cf 



'- CM CO •<»• to 



»- CM to ^ 



00 r>- <o u> 



en 

«\ 

e 
o 
•G 

A 
V 

bc 

> 



a 




Oi 












o 




a. 


V 




JS 




■w 




Cm 




O 




V 




u 




)m 




a 




o 




09 




U 








JD 




CS 




XI 




o 




h 




a 




(U 




pfi 




-w 




o 


ri 


to 


4^ 


1-H 




SI 


H 


s 


s 


DXI 


•■■ 




01 


to 


■M 


0) 


c« 


4> 


4> 


M 


.r: 




-M 


M 
^ 


-M 






ce 


s 


DC 


o 


e 


■^ri 




"1 


a 


44 


c« 






u 

f 


cs 






,fi 


3 


•pM 


CA 


f 


Cm 

O 


2 


S 


ss 


o 


o. 


■M 


a 


cs 


o 


o 


U 






V 


"O 


,s 


0) 


^^ 




o 


a 

3 


to 


o 


e 
o 


o 
e 


-M 


■M 


(« 


3 


o 


n 




e 


e 


2J 


o 


a 


a 


4J 


s 
u 


c« 


o 


+ 


S 


• 


1-H 


mm 

•a 


?' 


a 


b 


0) 


3 


Q, 


OX) 


f^ 


to 


< 



Proc. Linn. Soc. N.S.W., 128, 2007 



27 



VEGETATION HISTORY OF DRY LAKE, NSW 



Table 2 Surface sample collecting sites and their vegetation. Compare with pollen representation in 
Fig. 11. 



Sample Site details 



Vegetation 



Forest and woodland 

TS 1 Lake margin, on 

alluvium 



TS 2 Colluvial footslope 



TS 3 Seep rocky slope 



TS 4 Ridgetop plateau 



TS 5 Colluvial footslope, 

recenty burnt 



sites 



TS 6 Gully 





Dry Lake sites 


DLl 


Colluvial slope 


DL2 


Colluvial slope 


DL3 


Margin of lake 



DL 4 Centre of Dry Lake 



Low, open forest of Melaleuca linariifolia and Angophora floribunda 

saplings, 

Closed sedge and cyperaceous reeds. 

Open forest dominated by Corymbia spp. 

Open heath dominated by Lambertia formosa. Acacia linifolia and 

Platysace linearifolia. 

Open forest dominated by Eucalyptus spp., some Corymbia spp. Open 
heath dominated by Banksia spinulosa, Platysace linearifolia. Open 
grassland with Lomandra cylindrica. 

Low open forest dominated by Corymbia spp. 

Open heath dominated by Lambertia formosa, Grevillea mucronata, 

Eriostemon spp. Open grassland. 

Open forest dominated by Angophora floribunda. 

Low shrubland dominated by Acacia spp. 

Closed grassland dominated by Imperata cylindrica, Pteridium 

esculentum. 

Tall open forest dominated by Eucalyptus deanei, 

Low closed forest dominated by Pomaderris spp., Ceratopetalum 

gummiferum, Tristaniopsis sp. aff laurina. 

Femland ground cover. 



Closed grassland dominated by Conyza parva, Eragrostis brownii. 

Low shrubland with Leptospermum juniperinum 
Closed grassland as above. 

Low shrubland as above 

Grassland/sedgeland dominated by Eragrostis brownii, Lepidosperma 

longitudinale, Selaginella uliginosa. 

Herbfield with Gonocarpus micranthus, Persicaria decipiens, and 
Hypochoeris radicata. 



28 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



L2 5 



L2 6 



L2 7 



L2 8 



Thirlmere Lake sites 

50 m from lake margin, Open water, no vegetation, 
open water 3 m deep 

20 m from lake margin, Open water, no vegetation, 
open water 3 m deep 



5 m from lake margin, 
water 2 m deep 



Closed sedgeland of Lepironia articulata, Eleocharis spacelata, 
Brasenia schreberi. 



0.5 m from lake margin, Closed sedgeland as above, 
water 40 cm deep 



study did not reveal any well represented taxa. Over- 
represented taxa include Eucalyptus, Angophora/ 
Corymbia, AUocasuarina, Poaceae and Gonocarpus. 
Under-represented taxa were by far in the majority, 
since pollen of many species in the vegetation was 
never observed. Under-represented taxa include 
Proteaceae, Acacia, Leucopogon, Restionaceae, 
Platysace, Pimelea, Melaleuca, Leptospermum and 
pteridophytes. Some taxa are difficult to classify into 
any group, e.g. Monotoca, which was mainly under- 
represented, except in one sample, where it was over- 
represented. Given limited pollen dispersal (Birks 
and Birks, 1980; Dodson, 1983; Kodela, 1990), 
unusually high concentrations of under-represented 
taxa probably indicate that the plant was growing at 
the site. 

Pollen specfra from the Dry Lake samples were 
very different from other spectra in this study (Fig. 11). 
The differences are related to the clearance of native 
vegetation on and around Dry Lake, and reflects the 
more distant source of Eucalyptus. Dry Lake samples 
were dominated by Poaceae and Gonocarpus pollen 
and had slightly higher proportions of "weedy" taxa, 
e.g. Asterceae Liguliflorae (probably Hypochoeris 
radicata) and Plantaago lanceolata type which 
all grow at the site of deposition. Thus the advent 
of European clearing at Dry Lake should be easily 
recognized in the fossil pollen profile using these 
criteria. 

Most samples from forest or woodland sites 
produced similar pollen specfra but there were two 
noticeably distinct samples. Surface sample TS 5 was 
collected from a recently burnt site and is distinctive 
with a shrub layer of reduced diversity and density, 
dominated by Acacia spp. and a dense groundcover 
of Imperata cylindrica and Pteridium esculentum. 
Sites which had not been burnt so recently have a 
denser and more diverse shrub layer and a more 
diverse groundcover (see Table 2). The pollen spectra 



reflect these differences: recently burnt sites have 
a lower percentage of tricolporate grains (which 
includes many shrub taxa), no proteaceous pollen and 
much higher percentages of Poaceae pollen (perhaps 
Imperata) and trilete spores (probably Pteridium) 
than unbumt sites. Acacia pollen was not recorded 
from the surface samples although Acacia spp. were 
dominant in the understorey: it is highly under- 
represented. 

At sites where Angophora/Corymbia were the 
canopy dominants (samples TS 1, TS 2, and TS 5), 
their percentages in the spectra was 15% or greater 
and usually higher than Eucalyptus. Thus values of 
15% or more may infer a dominance of Angophora/ 
Corymbia at the site. Melaleuca/Leptospermum was a 
poorly represented group. In a sample from Melaleuca 
low open forest (TS 1), this group reached 15%, the 
highest at any site, hence this value may be used to 
identify dominance in the vegetation. 

Surface sample TS 6 was collected in a gully 
with tall open forest dominated by E. deanei, a 
dense understorey of several rainforest species and 
a dense ground-cover of ferns. Eucalyptus pollen 
still dominated the spectrum at 30%, but a greater 
diversity of palynomorphs are present. Pteridophyte 
spores, especially the monoletes, characterise the 
gully sample with a value of 15%), compared with 0- 
4% in other samples. Apart from these differences, the 
tall open forest pollen spectrum is hardly any different 
to those of open forest. The small rainforest trees 
were very poorly represented in the surface pollen 
spectrum. These results are in accord with those of 
Ladd (1979) and Kodela (1990) who found that small 
pockets of rainforest amongst widespread eucalypt 
vegetation were hardly detected by the palynological 
method. 

The samples from Thirlmere Lake 2 show that 
pollen from taxa growing on the lake margin, e.g. 
Gonocarpus and poorly dispersed types, e.g. the 



Proc. Linn. Soc. N.S.W., 128, 2007 



29 



VEGETATION HISTORY OF DRY LAKE, NSW 



tricolporate group are more important in the spectra 
near the lake margins (L2 7 and L2 8) where on site 
production is important. Well dispersed types, e.g. 
Eucalyptus, Allocasuarina, Poaceae and Cyperaceae 
are more important in pollen spectra from the lake 
centre (L2 5 and L2 6) which tends to accumulate 
wind dispersed pollen. 

These surface pollen spectra provide some 
basis for interpretation of the fossil pollen spectra. 
However, the sensitivity of the palynological approach 
usually only allows identification of vegetation units 
to the formation level. It is unlikely that different 
Eucalyptus associations can be distinguished by the 
pal)Tiological method. 

History of the Vegetation 

Fig. 12 a-c presents the fossil pollen diagram 
from the profile through the lake sediments. Pollen 
content is expressed as percentages of total count for 
all taxa identified and as pollen concentration for the 
most abundant taxa. Because percentages must add 
up to 100, a change in abundance of just one taxon 
will influence the percentages of all other taxa. On the 
other hand, the pollen concentration of each taxon is 
independent of all other taxa, hence a change in just 
one taxon will be apparent. On the whole, the curves 
for pollen concentrations reveal similar information 
to those of percentages, with some exceptions. 

The total pollen concentration (Fig. 12 c) is low 
in the lower part of the profile, below 40 cm, and 
generally higher in the upper part of the profile. 
The low pollen concentration section is found in the 
mottled clay which was subjected to a fluctuating 
water table that could have had a destructive effect 
on pollen. Thus the lower concentrations could be 
the result of destruction of some pollen or a lower 
pollen producing plant community. Cyperaceae is a 
thin-walled grained which may be easily destroyed 
in adverse environments, but it is found throughout 
this lower section of the profile. Indeed, the pollen 
spectra of the low pollen content section of the profile 
are generally comparable with the spectra in the high 
pollen content part of the profile, inferring that any 
pollen destruction has not appreciably distorted the 
spectra. An examination of the surface sample spectra 
together with the fossil spectra reveals the following: 

Eucalyptus and Allocasuarina/Casuarina 
are the main trees in the lower part of the profile. 
Allocasuarina/Casuarina would have been 
more abundant in the vegetation than it is today. 
Angophora/Corymbia would have been a relatively 
minor part of the vegetation. Thus the vegetation 
would have been much the same from the beginning 
of the Holocene (10 ka ago) until the mid Holocene 



(5 ka). Unfortunately, there was an hiatus and hence 
no record from about 5 ka until 2 ka. By then, 
Angophora/Corymbia had increased and it suggests 
that Angophora (more common on alluvial soils) 
may have been a canopy dominant (values > 15%) 
for much of the time. Eucalyptus percentages were 
a little less and Allocasuarina was still abundant, 
although it fluctuated somewhat. Melaleuca/ 
Leptospermum would have been minimal in the early 
Holocene and more abundant in the late Holocene, 
but it was insufficient (< 15% ) to indicate a fringing 
Melaleuca forest, similar to the Thirlmere Lakes 
today. As discussed previously, stumps on Dry Lake 
were unlikely to be Melaleuca, thus supporting this 
interpretation of the pollen content. 

In the European zone (above 10 cm depth), all 
of these trees decrease, doubtless the result of timber 
cutting and agriculture. 

Poaceae was moderate in the early Holocene, 
decreasing in the late Holocene when the tree cover 
increased, but increasing markedly in the European 
zone, no doubt due to forest clearing and agriculture. 
Asteraceae, which may have represented herb(s) or 
shrub(s) was noticeably more abundant in the early 
Holocene, decreasing after 5 ka and remaining low 
for the rest of the time. Gonocarpus is more abundant 
in the early Holocene, decreasing after 5 ka and 
remaining lower until the European zone where it 
was much reduced. Some Dry Lake samples show 
more abundant Gonocarpus which grows on the 
surface today. The abundance of Cyperaceae was 
moderate in the early Holocene, increasing after 2 
ka and remaining high, although values fluctuated. 
Cyperaceae would have grown on the peat surface 
during the late Holocene, just as it does today. 

The only shrubby and herbaceous taxa recorded 
in the early Holocene were Proteaceae, tricoporates, 
Chenopodiaceae and Brassicaceae. Since it was 
not possible to count sufficient grains required for 
adequate sampling from this low pollen concentration 
zone, and these other shrubs and herbs were under- 
represented taxa, their absence may be the result of 
inadequate sampling. Spores of ferns (monolete and 
trilete) and Selaginella were not recorded. From 
2 ka to the present, shrubs and herbs were better 
represented. Spores of ferns and Selaginella were 
also present in this zone and they indicate conditions 
were somewhat wetter after 2 ka. 

Some exotics are present in the European zone, 
viz. Pinus, Plantago cf lanceolatus and Asteraceae 
Liguliflorae, assuming this pollen type represents 
Hypochoeris radicata and not a native species. As 
discussed previously, all the tree taxa are reduced and 
Poaceae increases markedly. 



30 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



«>< 



■^A- 






H.,, 



% 



"^. 






^i 



4 ■<*.. 



H !- 



■^ ,4„ 



,HL.,„j |_ 






^h 



%.. .1 I * 



(M 



-il- 



iUL 



-Ih 



IK 

ea 

61) 



■E 
e 

E 



s 

4> 



(0 

c 
o 



c 

8 



c: 
1 



"S f 



I I 



JL±_ 



'S^lJjJJL 



LJUi 



(i til 



jJL 



I I 



ut 



f I 



! O 



-IM- 



^^- 



J H 



o 
a 

a 
o 

B 

C 

JS 
u 
e3 

u 

a 

s 

■O S 

o 



u 

s 
o 

<u 

"Z 

o 
;• 

a 






-<io 



Li. 



J L 



■it- 



% 



iUi 



o 



-if? 



•4t 



^ 



feif 



I 



8 



1 



E 

3 
(0 

c 



c 



o 

c 
m 






2 



am- 



f f:>^^^s<ryyyy y y jooj ^^.M.nnnu.M, ,1 i ■ < 



S 

CM 






c- - 



1—1 _ I _ I 
_< — I _r .. 
I I I » 






eo 



o 



cs 



9i 

o 
« -o 

a o 
« g 

B ^ 
? « 

a s 

;^ *^ 

H a 

V a. 
s « 

OS k. 



Proc. Linn. Soc. N.S.W., 128, 2007 



31 



VEGETATION HISTORY OF DRY LAKE, NSW 



X 



I „,■»„. It,. 






Lili 



%^ 






IJ-J- 



%1 



V 



.._Li 



J L 



%. % 



-LXJ lit 



K 



jj__,f-j_ 



X,. '^^ 



H>^1 



MJ- 



A 



'^. 



% 



^^ ( 111 



"«</J 



^i- 



JLl 



ii riN! (, 






^ *J_J Lii 



L-JL 






% 



X 



V 

'% 



!M- 



J_tJL 



XJkJL. 



%i 



-II- 



QQ 



%to 



J U_ 



t-t- 



-II- 



it- 



■^^h 



-A^ 



■i\- 



^t 



-II- 



-If- 



Hi- 



-tt- 



1 M f t Iff ,J_jt_Li^ 



Hl- 







9t 
C3 



"© 
& 

a 
o 

s 

e 

.a 
u 



C 

a 



S 
O 
M 

.a 
o 
u 
a. 

u 



o 



a 

g 

• mm 



JS 
•^^ 

o «^ 

e« C 

^ S 

u o 

« « 

«) o 

o ,a 

.S 2 

§ + 

V e 

s a. 

M a 



32 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



c 
o 



c 

JO 

I 



X 



*bj 



x 



^z?. 






'% 



*^ 



\ 



'O -fc 



'<%>. 



X 



'"% 



% 



o 



^, 



'^> 



'IV 



%> 



«>, 



8 






-u-Uk 



I. II . J „ L 



-14- 



^t- 



^^-J— •- 



Ht- 



i 



J ±. 



J t.t-l 



-It- 



-*— I*- 



Hf- 



UL 



^ 



-11- 



-Ljt *_ 



I ' » I 1 1- 



-th 



J Ldk_ 



-iV 



-\>- 



O^-L-L 



^t- 



-ii- 



H-l I 



-«I-J — L 



Hf- 



« ' ' 



-t1- 





04 




4> 




t» 






^0 


E 


'i^TJ 


I. 




01) 


CN E 


-a 


"1 


s 


*- 








0. 

4> 




e 


.(o °S 





■^ O) 


0) 


•s 


s 


100 -Q 


C4 

B 




.S 


2 


u 


Z 


e5 




;> 





,0 




«M 




B 




4i 

















a 




V 




.s 




-«rf 




«M 









01 




u 




L. 




S 









v> 




44 








A 




CS 




.£ 









;. 




D. 




u 




.fl 




-*■* 




i> 









Ea^ 




VI 




■*^ 




e 




a> 




E 








-B 




u 




l« 


^ 


04 




.:;£ 




<a 




nJ 




>» 




i> 




Q 




0) 




.s 




■4ii« • 




;« -a 




04 




-M 




es S 




« 




04 W 




a -M 




M 




c e 




04 -w 




s s 




.a 




a. ^ 




2 04 




H « 




. u 




e> 




TJ ,. 




04 04 




3 h 




.5 2 




-t-i 




§ + 




sSir^ 




J2 •- 




0) e 




u a; 




?. o- 




OD p. 




E < 



Proc. Linn. Soc. N.S.W., 128, 2007 



33 



VEGETATION HISTORY OF DRY LAKE, NSW 



In summary, the early Holocene vegetation was 
a Eucalyptus/Allocasuarina/Casuarina woodland/ 
forest, with Asteraceae (Tubuliflorae) prominent in 
the understorey. Gonocarpus was probably common 
around the lake. In the late Holocene, Angophora 
woodland was present also and there was a diversity of 
shrubs in the understorey. The lake had become a peat 
swamp and Cyperaceae grew on and/or around the 
swamp. After Europeans arrived, the trees decreased 
and, grasses increased markedly. Today, Angophora 
woodland is found on the deeper, moister soils of 
the alluvial fans, hence its development around Dry 
Lake in the late Holocene probably indicates a wetter 
climate at that time. 

Siliceous microfossils 

Treatment for the recovery of siliceous 
microfossils yielded sponge scleres, diatoms, plant 
phytoliths and sand grains. Only sponge scleres 
were studied in detail, however some observations of 
diatoms or plant phytoliths are reported here. 

Freshwater sponges occur in most semi- 
permanent and permanent inland waters of Australia. 
Distribution of the species are not uniform and is 
largely governed by physiochemical properties 
of the environment. Only one species of sponge, 
Radiospongilla sceptroides Haswell is present in the 
Thirlmere Lakes system (NPWS, 1997). This species 
has a wide but scattered distribution east of the 
Dividing Range and has a preference for non-alkaline 
environments (Racek, 1969). R. sceptroides produces 
a vivid green pigment and lives mainly on fallen logs, 
branches and leaves in the littoral zone where water 
fluctuations are most frequently experienced (Racek, 
1969). It is thought that the relative abundance of 
sponge scleres could be used as an indicator of water 
depth and lake level fluctuations. 

Two surface samples were studied: SS 1 from the 
lake margin and SS 2 from the lake centre (Fig. 13). 
The lake margin had 20 times the megasclere (body 
scleres) numbers than found at the lake centre. The 
lake margin is the habitat of the sponge and few scleres 
are apparently transported to the lake centre in this 
low energy environment. There were also appreciable 
numbers of gemmoscleres (carried on the gemmules), 
showing that the sponges were gemmulating in the 
not too distant past. Today, R. sceptroides does not 
form gemmules which are produced in response to 
adverse environmental conditions (Racek, 1969; 
NPWS, 1997). 

Few scleres were found in the clays at the base 
of the sediments. The lake was probably much larger 
and deeper then and the sponge habitat would have 
been too far away for many of their scleres to be 



deposited at this site. There are two peaks in the values 
for megascleres: at 31 cm and a much larger one at 
51cm, the latter at the base of the diatomite. These 
peaks suggest that the lake had become shallower and 
smaller, such that the sponge habitat was close to this 
site. The megasclere content was moderate through 
most of the profile, declining towards the surface. 

Gemmoscleres were also foimd throughout 
the profile, and in appreciable numbers. The ratio 
of gemmoscleres to megascleres is an indicator of 
the harshness of the conditions (Racek, 1969). The 
two surface samples fi-om Thirlmere Lakes have 
extremely small ratios, which indicate that conditions 
today are not often harsh enough to induce the 
sponges to form gemmules. The highest ratio of 
gemmoscleres to megascleres were found at the base 
of the diatomite. The ratio of megasclere fragments to 
entire megascleres was quite high, especially in the 
lake margin sample. This ratio may indicate how well 
the sponge remains were preserved in the sediments, 
especially at depth. However, the large number of 
fragmented sponge remains in the shallowest depths 
probably indicate mechanical breakage and being 
silica, the scleres do not decompose. 

The high concentration of megascleres and 
gemmoscleres at the base of the diatomite were 
associated with a high concentrations and diversity of 
diatoms. Phytoliths were abundant also, and all these 
siliceous microfossils, being comparable to sand 
grains which are more common in the littoral zone, 
suggest that the lake was shallow and swamp plants 
were growing on or close to the site. The diatomite 
also contains a higher content of sand grains than any, 
which is indicative of shallow water of the littoral 
zone, the habitat of/?, sceptroides. 

In general, high concentrations of phytoliths 
were found from 50 cm upwards, or in that part of the 
profile that is organic. This high concentration implies 
that swamp vegetation had colonized the lake surface 
from the 50 cm level upwards. Below this level, the 
abundance of phytoliths was low, suggesting that 
little swamp vegetation was nearby and the lake was 
too deep for its growth. 



DISCUSSION 

The history of Dry Lake 

In the early Holocene (from 10 ka). Dry Lake was 
relatively deep, with a calm, low energy envirormient 
depositing clay. The margins probably supported 
some cyperaceous reeds and Gonocarpus, but a fringe 
of paperbarks {Melaleuca linariiforia), similar to the 



34 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 






d 



o 

d 



.8 



IS 

d 



d 



o 






-4>- 



m 



J_ 



^§ 



CM 1- 

(0 



a 

E 

(D 
(0 

S 
o 
O 



14- 



a 



8 



S 



-ti- 



? 



8 






8 



a. 

CD 



m^^^^^^^z^^^^Ai 



1 1 1 1 1 1 1 I _ I _ I _ i _ I _ I 

H H I I I I _!. J J 1 

t I I I M I I ~ I . . I 'I ' 



I 






4> 

la 

0) 



o 



e 

s 

■3 

a 

en 
V 

O) 



o 

g 
s 

bxi 

Cm 

O 

u 
a 
XI 

E 

s 



S ^ 

S w 

^ 0X1 
f»i E 

E o 
e£ E 

SB 

«M 61) 

O c« 
i- © 

9 Pi 
Z - 

•^ i 

o ^ 

9- o 
CO -S 

S DC 



Proc. Linn. Soc. N.S.W., 128, 2007 



35 



VEGETATION HISTORY OF DRY LAKE, NSW 



Thirlmere Lakes today, was lacking. The freshwater 
sponge, Radiospongilla sceptroides, Hved round the 
margin of the lake, amongst fallen debris. Eucalyptus 
spp. and Allocasuarina were the dominant trees in the 
surrounding vegetation and Asteraceae (Tubuliflorae) 
was prominent in the understorey, probably as a shrub. 
Some grasses (Poaceae), Proteaceae, tricolporates 
(most likely shrubs) and Chenopodiaceae were also 
present in the early Holocene. Although generally 
high, lake levels must have been very variable and 
the lake probably dried up for extended period(s), 
causing the formation of the pallid and mottled clays 
at the deepest parts of the lake. 

Between 8 ka and 5 ka, the lake became shallower 
and the fringing swamp vegetation grew over much 
of the lake surface. In the catchment, density of 
the Eucalyptus and Allocasuarina trees increased 
somewhat, Asteraceae was much reduced and shrubs 
and herbs increased in diversity, with Dodonaea, 
Monotoca, Pimelea, Brassicaceae and Portulacaceae 
being recorded. 

From 5 ka to 2 ka, there was an hiatus in the 
deposition of the sediment, or the sediments that were 
deposited were subsequently eroded. Unfortunately, 
this means there is no information for this period. 

About 2 ka, a shallow lake returned, probably 
covered with swamp vegetation and sufficiently 
nutrient rich to support large populations of diatoms 
and sponges. It is not clear how this nutrient rich status 
was achieved, given the nutrient-poor sandstone of 
the catchment. Decaying swamp vegetation would 
increase the nutrient status, but it would require a 
high nutrient status to produce a good plant cover in 
the first place. Burning may also mobilize nutrients. 
Unfortunately, there is no evidence about which is 
the more likely hypothesis in this case. This enriched 
nutrient status did not last long, and the diatom and 
sponge populations decreased to 'normal' levels. 
The lake remained shallow and probably supported 
swamp vegetation over most of its surface. 

After 2 ka, the Angophora/Corymbia group 
increased dramatically. Today, Angophora dominates 
the alluvial fans and soils adjacent to the lakes, and 
Corymbia is more common on well-drained slopes 
and ridges, hence this increase around Dry Lake 
was more likely to have been Angophora. This 
change suggests a somewhat moister environment. 
Eucalyptus and Allocasuarina probably decreased 
slightly, Melaleuca/Leptospermum increased, but not 
sufficiently to indicate a fringe of Melaleua around 
the lake. The diversity of shrubs and herbs increased 
further, and there was a considerable increase in 
cyperaceous swamp cover. 

The introduced Pinus, Plantago cf lanceolata 



and Asteraceae (Liguliflorae: probably Hypochoeris 
radicata) denote the zone of European influence. All 
the trees, viz. Eucalyptus spp, Angophora/Corymbia 
and Allocasuarina decreased markedly, no doubt the 
result of timber cutting. Grasses and the tricolporates, 
which could include any number of crop plants and 
weeds, would have been the result of agriculture. The 
cyperaceous reeds around the swamp remained, much 
the same as previously. 

Comparisons with Other Studies 

The history of the vegetation from a core in Lake 
Baraba (Thirlmere Lake 3 of this study, see Fig. 1), 
has been reported by Black et al. (in press). Lake 
Baraba is some 4 km south of Dry Lake. Peat began 
forming in the early Holocene, ~8.5 ka, earlier than at 
Dry Lake. Thus in contrast to Dry Lake, Lake Baraba 
had become shallow enough for the growth of swamp 
vegetation. At Lake Baraba, the dominant trees were 
Casuarinaceae which declined in the early Holocene, 
with a concurrent increase in Myrtaceae, thought to 
be the development of the fringing Melaleuca forest 
(Black et al., in press) which is present around the 
lake today. In contrast, at Dry Lake, Allocasuarina 
did not decrease, a fringing Meleleuca forest did 
not develop, and Angophora became prominent by 
the mid Holocene. At Dry Lake, the lake became 
shallow enough to support swamp vegetation and 
peat formation about the mid Holocene, later than 
at Lake Baraba. Allocasuarina remained prominent 
at Dry Lake until the European zone, unlike Lake 
Baraba where it remained low through most of the 
Holocene. 

These differences between the two sites may 
be attributed to the differences in local topography. 
Lake Baraba is confined within a relatively narrow 
valley which is likely to afford some protection and 
provide more favourable moisture relationships than 
the Dry Lake locality, which is more open, in a broad 
alluvial flat. This topographic difference may explain 
why Dry Lake did not develop a fringing Melaleuca 
forest. Although these two sites are only 4 km apart, 
the limited nature of pollen dispersal, where most 
pollen falls close to the source (Birks and Birks, 
1980; Dodson, 1983; Kodela, 1990) ensures that 
these local differences in the vegetation are recorded 
in the sediments. 

A decline of Casuarinaceae, when Eucalyptus 
replaced Casurarinaceae, may be found in a number 
of Holocene sites in southern Australia and is usually 
dated between 7.5 and 4.5 ka. Importantly, not 
all Holocene sites show this decline (Clark, 1983; 
Dodson, 1994; 2001; Lloyd and Kershaw, 1997). It 
has been suggested that anthropogenic fire may have 



36 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



been the cause of this change in dominant species, 
but the charcoal records generally do not support 
this hypothesis (Dodson, 2001; Kershaw et al.; 2002; 
Black et al., in press). Another likely cause, a rising 
water table or salinity, may be supported by an increase 
in Chenopodiaceae pollen (Crowley, 1994; Cupper 
et al., 2000), but studies in the vegetation show that 
some species of Casuarinaceae are more salt tolerant 
than certain species of Eucalyptus (Ladd, 1988). The 
anatomy of the branchlets of Casuarinaceae, with 
their restricted photosynthetic tissue, make it a poor 
competitor with broad leaved species. Anatomicaly, 
Casuarinaceae species are very xeromorphic and in 
comparative studies, Casuarinaceae is more drought 
tolQxant than Eucalyptus (Ladd, 1988). 

As the climate ameliorated after the last glacial 
period, the grasslands/shrublands were invaded 
by Casuarinaceae which were in turn replaced by 
Eucalyptus in the Holocene (Clarke, 1983). The 
climate in the last glacial period was much drier, hence 
the change in vegetation parallels the climatic change, 
viz. the increase in moisture. Casuarinaceae remained 
on poor or harsh sites as it appears to tolerate these 
conditions better than Eucalyptus (Ladd, 1988). 

At Dry Lake, Eucalyptus and Angohpora/ 
Corymbia increase, but Casuarinaceae does 
not decrease until it was logged by Europeans. 
Casuarinaceae was prized by the early settlers as 
firewood and it was the fuel of choice for bakeries. Its 
timber was in demand for shingles, tool handles, beer 
barrels and many other used (Entwisle, 2005). Indeed, 
the Oaks, some 15 km to the north of Thirlmere 
(Fig. 1) was so named for the abundant sheoaks 
{Allocasuarina torulosa). When the botanist George 
Caley passed through the district in 1804, he saw 'a 
large tract of grazing land abounding with sheoaks' 
(Woods, 1982). 

The Casuarinaceae pollen has not been identified 
further; but it is assumed to be Allocasuarina in this 
study because there are only two species in the area 
today: A. Uttoralis and A. torulosa. A. littoral is is 
an understorey tree in woodland or occasionally tall 
heath, on sandy or otherwise poor soils. A. torulosa, 
also an understorey tree, is found in open forests to 
tall open forests, generally on higher nutrient soils 
and moisture situations than A. Uttoralis (Plantnet, 
2005). The alluvial flat around Dry Lake would have 
been suitable for^^. torulosa, but at Lake Baraba, in a 
sandstone valley, A. Uttoralis seems more likely. Thus 
the different history of Casuarinaceae at the two sites 
may have been the consequence of different species, 
as well as the different topography and soils. 



ACKNOWLEDGEMENTS 

We would like to thank the National Parks and 
Wildlife Service for permission to undertake this study in 
the Thirlmere Lakes National Parks. Special thanks go to 
Mr. Ross Kinnish, National Parks Ranger at Thirlmere, 
who assisted with some field work. We are grateful to Mr. 
and Mrs. Lipping for allowing us onto their land to work 
at Dry Lake. Our special thanks go to family and friends 
who assisted with this study and provided moral support. 

We are indebted to Professer Carswell and Mr. V. 
Djohadze, of the then School of Nuclear and Radiation 
Chemistry, University of New South Wales, for the 
radiocarbon dates. We also thank Officers of the National 
Herbarium for assistance with plant identification and Mr. 
John Stanisic, of the Queensland Museum who kindly 
identified the sponge remains. 

Special thanks go to Dr. Scott Mooney, of the 
University of New South Wales, who read the manuscript 
and offered invaluable comments. 



REFERENCES 

Bengtsson, L., Enell, M, (1990) Chemical analysis. 

In 'Handbook of Holocene Palaeoecology and 

Palaeohydrology' (Ed. B.E. Berglund.) pp. 423-453. 

(Wiley and Sons: Chichester). 
Black, M.P., Mooney, S.D. and Martin, H.A. (in press). A 

> 43,000 year vegetation and fire history from Lake 

Baraba, New South Wales, Australia. Quaternary 

Science Reviews. 
Birks, H.J.B. and Birks, H.H. (1980) 'Quaternary 

Palaeoecology' (Edward Arnold: London) 
BoM, (2005). Commonwealth Bureau of Meteorology 

Website (www.bom.gov.au). Accessed 1-7-05. 
Chalson, J.M. and Martin, H.A. (1995). The pollen 

morphology of some co-occurring species of 

the family Myrtaceae from the Sydney Region. 

Proceedings of the Linnean Society of New South 

Wales nS,\6'i-\9\. 
Clark, R.L. (1983). Pollen and charcoal evidence for 

the effects of Aboriginal burning on Australia. 

Archaeology in Oceania 18, 32-37. 
Crowley, G.M. (1994). Quaternary soil salinity events and 

Australian vegetation history. Quaternary Science 

Reviews 13, 15-22. 
Cupper, N.L., Drinnan, A.N. and Thomas, I. (2000). 

Holocene palaeoenvironments of salt lakes in the 

Darling Anabranch region, south-western New South 

Wales. Journal of Biogeography 11, 1079-1094. 
Dodson, J.R. (1983). Modem pollen rain in southeastern 

New South Wales, Australia. Review of Palaeobotany 

and Palynology 38, 249-268. 
Dodson, J.R. (1994) Quaternary Vegetation. In 'Australian 

Vegetation' (Ed. R.H. Groves) pp. 37-54. (University 

of Cambridge Press: Cambridge). 



Proc. Linn. Soc. N.S.W., 128, 2007 



37 



VEGETATION HISTORY OF DRY LAKE, NSW 



Dodson, J.R. (2001) Holocene vegetation change in the 
mediterranean type climate regions of Australia. The 
Holocene 11, 673-680. 

Entwisle, T. (2005). She-oak up in smoke. Nature 
Australia Spring 2005 28(6), 72-73. 

Harden, G.J. (1992, 1993, 2000, 2002). 'The Flora of 
New South Wales, Vol. 3, Vol. 4, Vol.1 (revised 
edition) and Vol. 2. (revised edition)', respectively. 
(University of New South Wales: Sydney). 

Herbert, C. (1980). Wianamattta Group and the Mittagong 
Formation. In 'A Guide to the Sydney Basin' (Eds 
C. Herbert and R. Helby) pp. 254-272. Geological 
Survey of New South Wales Bulletin 26. (D. West, 
Government Printer, New South Wales: Sydney). 

Horsfall, L., Jelinek, A., Timms, B., (1988). The influence 
of recreation, mainly power boating on the ecology 
of Thirlmere Lakes, NSW, Australia. Vereinigung 
fiir Theretische und Angewandte Limnologie 23, 580 
-587. 

Kershaw, A.R, Clark, J.S., Gill, A.M. and D'Costa, D. 
(2002). A history of fire in Australia. In 'Flammable 
Australia: the fire regimes and biodiversity of a 
continent' (Eds R. Bradstock, J. Williams and 
A.M. Gill) pp 3-25. (Cambridge University Press, 
Cambridge). 

Kodela, P.G. (1990). Modem pollen rain from forest 
communities on the Robertson Plateau, New South 
Wales. Australian Journal of Botany 38, 1-24. 

Lacey, W.S. (1963) Palaeobotanical techniques. In 

'Viewpoints in Biology 2' (Eds J.D. Carthy and G.L. 
Duddington) pp. 202-243. (Butterworths: London). 

Ladd, P.G. (1979). A short pollen diagram from rainforest 
in highland eastern Victoria. Australian Journal of 
Ecology 4, 229-237. 

Ladd, P.G. (1988). The status of Casuarinaceae in 
Australian forests. In 'Australia's ever changing 
forests. Proceedings on the First National Conference 
on Australian Forest History' (Eds K.J. Frawley 
and N. Semple) pp 63-85. (Special Publication No. 
1 , Department of Geography and Oceanography, 
Australian Defence Force Academy, Campbell ACT) 

Lloyd , P.J. and Kershaw, A.R (1997). Late Quaternary 
vegetation and early Holocene quantitative climate 
estimates from Morwell Swamp, Latrobe Valley, 
south-eastern Australia. Australian Journal of Botany 
45, 549-563 

Moore, PD., Webb, J.A., ColHson, M.E. (1991). 'Pollen 
Analysis, second edition'. (Blackwell Scientific 
Publications: London). 

NPWS, 1997. Thirimere Lakes National Park New Plan of 
Management. (National Parks and Wildlife Service: 
Sydney). 

Pidgeon, I.M. (1937). The Ecology of the Central Coastal 
of New South Wales. I. Proceedings of the Linnean 
Society of New South Wales 62, 315-340. 

Pidgeon, I.M. (1941). The Ecology of the Central Coastal 
of New South Wales. IV. Proceedings of the Linnean 
Society of New South Wales 66, 1 13-137. 

Plantnet (2005). National Herbarium website (http:// 

plantnet.rbgsyd.nsw.gov.au). Accessed October 2005. 



Racek, A. A. (1969). The fi^eshwater sponges of Australia. 

Australian Journal of Marine and Freshwater 

Reaearch 20,267-310. 
Rose, S. (1981). Palynology and history of the Holocene 

at Dry Lake, Thirlmere, N.S.W. B.Sc. Hons. thesis. 

University of New South Wales, Sydney. 
Stuiver, M. andReimer, PJ. (1986-2005). Radiocarbon 

calibration program Calib.Rev 5.0.2. http://calib.qub. 

ac.uk/calib/calib.html. 
Timms, B., (1992). 'Lake geomorphology'. (Gleneagles 

Publishing; Adelaide). 
Vorst, R (1974). Thirlmere Lakes, NSW: Geomorphic 

environment and evolution. B. A Hons. thesis, 

Macquarie University, Sydney. 
Woods, D. (1982). 'A short history of the Oaks (third 

edition)', (the Oaks Historical Society: Camden, New 

South Wales). 



38 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



APPENDIX lA 

Species list of plants in the Thirlmere National Park, compiled from field work and augmented from 
a list prepared by the National Herbarium of New South Wales (*). +, introduced species. 

Life forms: T, tree with single stem, > 8 m tall. S, shrub, woody plant < 8 m tall. H, herbs, non- 
woody plants. epiH, epiphitic herbs. aqH, aquatic herbs, growing in wet or periodically wet areas. C, 
creeper, prosfrate herb or shrub. TW, twiner, climbing plant. 

Pollination mechanisms (Poll'n Mech), from Faegri and van de Pijl (1971), Dodson (1979), Arm- 
sfrong (1979), Ford et al. (1979) and Pyke (1981): A, anemophilous, wind pollinated. E, entomophilous, 
insect pollinated. O, pollination by other animals, e.g. birds, mammals. S, self pollinated. H, hydrophilous, 
water pollinated. 

Nomenclature follows Harden (1992; 1993; 200; 2002) and Plantnet (2005). 



Occurrence/ 
distribution 



Species 



Life 
form 



Poirn 
Mech. 




MYRTACEAE ^^ 

Angophorafloribunda (Sm.) Sweet 

Corymbia eximia (Schauer) K.D. Hill & L. 

A. S. Johnson 

C gummifera (Gaertn.) K.D. Hill & L.A. 
S. Johnson 

Eucalyptus agglomerata Maiden 

E. botryoides Sm. 

E. oblonga Blakely. 

E. piperita Sm 

E. punctata DC 

E. racemosa Cav. 

E. resinifera Sm. 

E. sieberi L. Johnson 

Leptospermum trinervium J. Thompson 

L. polygalifolium Salsb. 

L. juniperium Sm. 

Kunzea ambigua (Sm.) Druce 

Melaleuca linariifolia Sm 

M. thymifolia Sm. 

PROTEACEAE 

Banksia integrifolia L. f. 

B. serrata L.f. 
B. spinulosa Sm. 
Grevillea arenaria R. Br. 
G. mucronulata R. Br. 
Hakea dactyloides Cav. 

H. salicifolia (Vent.) B.L. Burtt. 



T 
T 



A, E, O Very abundant, lower slopes only 
A, E, O Very abundant, esp. ridgetop plateaux 



T 


A,E,0 


Abundant 


T 


A,E,0 


Occasional 


T 


A,E,0 


* 


T 


A,E,0 


Occasional, mostly steeper slopes 


T 


A,E,0 


Very abundant 


T 


A,E,0 


Occasional 


T 


A,E,0 


Occasional 


T 


A,E,0 


Rare 


T 


A,E,0 


Occasional, esp. near ridgetop plateaux 


S 


A.E. 


Common, mostly on slopes 


S 


A,E 


Rare, mostly on slopes 


S 


A,E 


Occasional, mostly lake margins 


S 


A,E,0 


* 


T 


A,E,0 


Abundant, mainly lake margins 


S 


A,E,0 


Occasional, along lake margins 


TorS 


E,0 


* 


TorS 


E,0 


Abundant, mainly lake margins 


S 


E,0 


Very abundant 


S 


E,0. 


* 


S 


E,0 


Very abundant 


S 


E, 


Occasional, moist sites 


.^ 


E,0 


* 




Proc. Linn. Soc. N.S.W., 128, 2007 



39 



VEGETATION HISTORY OF DRY LAKE, NSW 



H. sericea Schrad. & J.C. Wendl. 

Isopogon anemonifolius Knight 

Lambertia formosa Sm. 

Persoonia lanceolata Andrews 

P. laurina Pers. 

P. levis (Cav.) Domin 

P. linearis Andrews 

Petrophile pedunculata R. Br. 

P. pulchella R. Br. 

P. sessilis Sieber ex Schult. 

Telopea speciosissima R. Br. 

Xylomelum pyriformis Sm. 

FABACEAE 

1) MIMOSOIDAE 
Acacia decurrens Willd. 
A. falcata Steud. 

A. falciformis DC. 

A. floribunda Willd. 

A. implexa Benth. 

A. linifolia Willd. 

A. longifolia (Andrews) Willd. 

A. myrtifolia Willd. 

A. parramattensis Tindale 

A. suaveolens (Sm.) Willd. 

A. terminalis J.F. MacBr. 

A. ulicifolia Court 

2) FABOIDAE 
Bossiaea buxifolia A.Cunn. 

B. heterophylla Vent. 
B. lenticularis DC. 

B. neo-anglica F. Muell. 

B. obcordata Druce 

B. rhombifolia Sieber ex DC. 

Daviesia corymbosa Sm. 

Desmodium rhytidophyllum F. Muell ex 
Benth. 

D. varians (Labill.) G. Don 

Dillwynia glaberrima Sm. 



S 


E, 





Rare 


s 


E 




Occasional 


s 


E, 





abundant 


S 


E 




* 


S 


E 




Occasional, esp. ridgetop plateaux 


s 


E 




Occasional 


S 


E 




Occasional 


S 


E 




Occasional 


S 


E 




* 


S 


E 




* 


s 


E 




Occasional 


T 


E 




Occasional 



T 


E 


Occasional esp. moist gullies/; 


S 


E 


* 


T 


E 


* 


S 


E 


* 


s 


E 


* 


S 


E 


Abundant 


s 


E 


Abundant esp. after fire 


s 


E 


Occasional, ridgetop plateaux 


S 


E 


Very abundant 


S 


E 


Occasional 


S 


E 


Occasional ridgetop plateaux 


S 


E 


Abundant 


S 


E 


Rare 


S 


E 


Occasional 


S 


E 


* 


S 


E 


* 


S 


E 


Occasional 


S 


E 


Occasional 


s 


E 


Rare 


Sor 
TW 


E 


* 


TW 


E 


* 


S 


E 


* 



40 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



D. parvifolia R.Br. 

D. phylicoides A. Cuirn. sp. complex 
Glycine clandestina J.C. Wendl. 
Gompholobium grandiflorum Sm. 

G. latifolium Sm. 

G. minus Sm. 

Hardenbergia violacea (Schneev.) Steam 

Hovea linearis (Sm.) R. Br. 

Indigofera australis Willd. 

Kennedia rubicunda Vent. 

Mirbelia rubiifolia (Andrews) G. Don 

Podolobium ilicifolium (Andrews) Crisp & 
RH. Weston. 

Pultenaea flexilis Sm. 

P. linophylla Schrad. & J.C. Wendl. 

P. villosa Andrews. 

Viminaria juncea (Schrad,) Hoffsgg.. 

RUTACEAE 

Boronia ledifolia (Vent.) J. Gray ex. DC. 

B. polygalifolia Sm. 

Eriostemon australasius Pers. 

E. hispidula. (Spreng.) Paul G. Wilson 
ERICACEAE 

Astroloma humifusum R. Br. 

Epacris pulchella Cav. 

Leucopogon lanceolatus (Sm.) R. Br. var. 
lanceolatus 

Lissanthe sapida R. Br. 

L. strigosa R.Br. 

Monotoca elliptica R.Br. 

M scoparia R.Br. 

Styphelia angustifolia DC. 

DILLENIACEAE 

Hibbertia aspera DC. 

H. diffusa DC. 

H. obstusifolia DC. 

H. serpyllifolia DC 

GOODENIACEAE 



s 


E 




* 


S 


E 




Abundant 


TW 


E 




Abundant 


S 


E 




* 


S 


E 




* 


S 


E 




Occasional 


c 


E 




Very abundant 


s 


E 




Occasional 


s 


E 




Abundant 


TW 


E, 





Occasional 


s 


E 




* 


S 


E 




Occasional esp. rocky slopes 


s 


E 




Very abundant 


s 


E 




* 


s 


E 




Occasional esp. alluvial fans 


s 


E 




Occasional esp. damp sites 


s 


E 




Occasional, esp. rocky slopes 


s 


E 




* 


s 


E, 





Occasional 


s 


E, 





Abundant 


s 


E, 





* 


s 


E, 





Rare 


s 


E 




Abundant 


s 


E 




Rare 


s 


E 




Occasional 


s 


E 




Occasional 


s 


E 




Occasional 


s 


E 




* 


s 


E 




Very abundant 


s 


E 




Occasional 


s 


E 




Occasional, moister slopes 


s 


E 




Occasional, moister slopes 



Proc. Linn. Soc. N.S.W., 128, 2007 



41 



VEGETATION HISTORY OF DRY LAKE, NSW 



Coopernookia barbata (R. Br.) Carolin 

Dampiera purpurea R. Br. 

Goodenia hederacea Sm. 

Scaevola ramosissima K. Krause 

CASUARINACEAE 

Allocasuarina littoralis (Salisb.) L. Johson 

A. torulosa (Alton) L. Johnson 

EUPHORBIACEAE 

Amperea xiphoclada (Spreng.) Druce 

Breynia oblongifolia Muell. Arg. 

Phyllanthus gasstroemii Muell. Arg. 

P. occidentalis J.T. Hunter & J.J. Bruhl 

Poranthera ericifolia Rudge 

P. microphylla Brongn. 

RUBIACEAE 

Galium binifolium N.A. Wakefield 

G. propinquum A. Cunn. 

Opercularia aspera Gaertn. 

O. diphylla Gaertn. 

O. varia Hook. f. 

Pomax umbellata Benth. 

APIACEAE 

Actinotus helianthi Labill. 

Centella asiatica Urb. 

Hydrocotyle acutiloba. N.A. Wakefield 

H. laxiflora DC. 

H. peduncularis . A. Rich. 

Platysace linear ifolia C. Norman 

LAURACEAE 

Cassytha glabella R. Br. 

C. pubescens R. Br. 

Cinnamomum camphora ^T. Nees & C.H. 
Eberm 

RANUNCULACEAE 

Clematis aristata R. Br. ex Ker Gawl. 
VIOLACEAE 

Hybanthus monopetalum (Schultes) Domin. 
Viola betonicifolia Sm. 



s 


E 


Occasional 


s 


E 


Occasional, moister slopes 


c 


E 


Occasional, lake margins 


H 


E 


Occasional 


T 
T 


A 
A 


Very abundant, lake margins, 

footslopes 

Occasional, upper slopes, ridgetops 


S 


E 


Occaisional 


S 


E 


* 


S 


E 


Abundant 


S 


E 


Abundant 


S 


E 


* 


H 


E 


* 


H 


A 


* 


H 


A 


* 


H 


A 


Abundant on lake margin, moist areas 


H 


A 


* 


H 


A 


Abundant on lake margin, moist areas 


Hor 
S 


A 


Very abundant 


H 


E 


Occasional 


C 


E 


* 


H 


A,E 


* 


H 


A,E 


* 


H 


A,E 


* 


S 


E 


Abundant. 


TW 




* 


TW 




* 



TW 

H 
H 



A,E 



One specimen observed 



E 
E 



Occasional, moist areas 



42 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



V hederacea Labill. 

CRASSULACEAE 

Crassula sieberiana (Schultes & Schultes. 
f.) Druce 



H 



H 



Occasional, moist areas 



DROCERACEAE 






Drosera spathulata Labill. 


H 


E 


POLYGONACEAE 






Persicaria hydropiper (L.) Spach. 


H 


E 


Acetosella vulgaris Fourr. 


H 


E 


OXALIDACEAE 






Oxalis corniculata * L. 


H 


E 


GERANIACEAE 






Geranium homeanum Turcz. 


H 


E 


HALORAGACEAE 






Gonocarpus micranthus Thunb. 


C 


E 


G. tetragynus Labill. 


H 


E 


Myriophyllum variifolium Hook. f.. 


AqH 


A 


THYMELEACEAE 






Pimelea linifolia Sm. 


S 


E, 


PITTOSPORACEAE 






Billardiera scandens Sm. 


TW 


E 


Bursaria spinosa Cav. 


S 


E 


PASSIFLORACEAE 






Passiflora edulis ^ Sims 


TW 


E 


HYPERICACEAE 






Hypericum gramineum G. Forst. 


H 


E 


ELAEOCARPACEAE 






Elaeocarpus reticulatus Sm. 


T 


E 


Tetratheca thymifolia Sm. 


S 


E 


MALVACEAE 






Sida rhombifolia '^ L. 


S 


E 


CUNONIACEAE 






Ceratopetalum gummiferum Sm. 


T 


E 


ROSACEAE 






Ruhus parvifolius L. 


S 


E 


R. fruticosus ^ species complex 


S 


E 


STACKHOUSIACEAE 






Stackhousia monogyna Labill. 


H 


E 



Occasional, damp places 



Occasional, moist sites 



Occasional, esp. disturbed sites 



Occasional esp. near lake margins 
Occasional, moist places 



Very abundant 

Occasional, shady slopes 
Occasional 

One specimen observed 



Occasional, only moist gullies 
Rare 



Occasional, only moist gullies 



Rare, disturbed sites 



Proc. Linn. Soc. N.S.W., 128, 2007 



43 



VEGETATION HISTORY OF DRY LAKE, NSW 



S. viminea Sm. 
LORANTHACEAE 

Unidentified 
SANTALACEAE 

Exocarpos cupressiformis Labill. 

E. strictus R. Br. 

Leptomeria acida R. Br.. 

SAPINDACEAE 

Dodonaea triquetra Benth. 

LOGANIACEAE 

Mitrasacme polymorpha R. Br. 

APOCYNACEAE 

Parsonsia straminea F. Muell. 

Marsdenia flavescens A. Cunn. 

M suaveolens R. Br. 

Tylophora barbata R. Br. 

MENYANTHACEAE 

Villarsia exaltata G. Don 

CAPRIFOLIACEAE 

Lonicarajaponica ^ Thunb. ex Murray 

PLANTAGINACEAE 

Plantago lanceolata ^ L. 

CAMPANULACEAE 

Wahlenbergia graniticola Carolin 

W. striata (R. Br.) Sweet 

W. communis Carolin 

LOBELIACEAE 

Isotoma axillaris Lindl. 

Pratia purpuras cens (R. Br.)F. Wimmer. 

STYLIDIACEAE 

Stylidium graminifolium Willd. 

5'. laricifolium Rich. 

S. lineare Sw. ex Willd. 

ASTERACEAE 

Bidens pilosa ^ L. 

Brachycome aculeata R. Br. 

B. angustifolia Cunn. ex DC. 



H 



TW 



H 



C 



O 



A,E 



Rare, host Eucalyptus spp. 



T 


A,E 


Occaisional 


T 


A,E 


Rare, moist gullies 


S 


A,E 


Rare 



Occasional, moist slopes 



s 


E 




Occasional 


TW 


E 




Rare 


TW 


E 




Rare, moist, shaded positions 


TW 


E 




Occasional, esp. shaded slopes 


TW 


E 




* 


H 


H, 


E 


* near water 



Occasional esp. disturbed sites 



H 


E 


Occasional 


H 


E 


* 


H 


E 


* 



Occasional esp. moist sites 



H 


E 


Occasion 


H. 


E 


* 


H 


E 


* 


H 


A,E 




H 


A,E 


* 


H 


A,E 


abundant 



44 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



Cassinia aculeata R. Br. 

C. aureonitens N. A. Wakefield 

C. longifolia R. Br. 

C. quinquefaria R. Br. 

Conzya albida "^ Willd. ex Sprengel 

C. parva ^ Cronq. 

Coreopsis lanceolata "^ L. 

Facelis retusa ^ Sch. Bip. 

Gnaphalium gymnocephalum DC. 

Helichrysum elatum DC. 

//! scorpiodes Labill. 

Hypochaeris radicata ^ L. 

Lagenophora stipitata (Labill.) Druce 

Olearia microphylla Maiden & Betche 

O. viscidula Benth. 

Ozonthamnus adnatus DC. 

O. diosmifolium (Vent.) DC 

Podolepis jaceoides Voss 

Pseudognaphalium. luteoalbum ^ (L.) 
Hillard & B.L. Burtt 

Senecio lautus ^ G. Forst. ex Willd. 

S. linearifolius A. Rich. 

S. quadridentatus Labill. 

S. velleioides A. Cunn. ex DC. 

Sigesbeckia orientalis L. 

SOLANACEAE 

Solarium pungetium R. Br. 

CONVOLVULACEAE 

Dichondria repens J.R. Forst. & G Forst 

Polymeria calycina R. Br. 

SCROPHULARIACEAE 

Veronica plebeia R. Br. 

LENTIBULARIACEAE 

Utricularia australis R. Br. 

ACANTHACEAE 

Brunoniella pumilio (R. Br.) Bremek. 

VERBENECEAE 

Verbenea bonariensis L. 



s 


A,E 


Occasional, shady slopes 


s 


A,E 


* 


s 


A, E 


Occasional, shady slopes 


s 


A,E 


* 


H 


A,E 


Weed on disturbed sites 


H 


A,E 


Weed on disturbed sites 


H 


A,E 


* 


H 


A, E 


* 


H. 
Hor 

S 
H 


A,E 
A,E 
A,E 


Occasional esp. guUies, very moist 

slopes 

Occasional 


H 


A,E 


Abundant esp. moist disturbed sites 


H 


A,E 


Shady slopes 


S 


A,E 


Occasional, mostly shady slopes 


S 


A,E 


* 


s 


A,E 


* 


s 


A,E 


* 


H 


A,E 


Occasional 


H 


A,E 


Occaisional 


H 


A,E 


Weed, on disturbed ground 


H 


A,E 


* 


H 


A,E 


Weed, on disturbed ground 


H 


A,E 


* 


H 


A,E 


* 



H 

C 
C 

H 

aqH 
H 



E 
E 



Occasional, esp. moist areas 

Occasional esp. lake margins 
* 



Rare, floating on open water > 2m deep 



H 



Occasional, disturbed sites 



Proc. Linn. Soc. N.S.W., 128, 2007 



45 



VEGETATION HISTORY OF DRY LAKE, NSW 



LAMIACEAE 

Ajuga australis R.Br. 

Scutellaria humilis R. Br. 

POTAMOGETONACEAE 

Potamogeton tricarinatus A. Benn. 

XYRIDACEAE 

Xyris complanata R. Br. 

ANTHERICACEAE 

Arthropodium milleflorum (DC.) J.F. Macbr. 

Laxmannia gracilis R. Br 

Tricoryne simplex R. Br. 

PHORMIACEAE 

Dianella caerulea Sims 

D. revoluta R. Br. 

Stypandra glauca R. Br. 

Thelionema caespitosum (R. Br.) R.J.F. 
Hend. 

SMILACACEAE 

Smilax glyciphylla Sm. 

LUZURIAGACEAE 

Eustrephus latifolius Ker Gawl. 

Geitonoplesium cymosum R. Br. 

IRIDACEAE 

Patersonia glabrata R. Br. 

P. sericea R. Br. 

LOMANDRACEAE 

Lomandra confertifolia (F.M. Bailey) 
Fahn ssp. rubiginosa R.T. Lee 

L. cylindrica R.T. Lee 

L.filiformis (Thunb.) J. Britten 

L. glauca Ewart 

L. gracilis R.T Lee 

L. longifolia Labill. 

L. multiflora (R. Br.) J. Britten 

L. obliqua J.R. Macbr. 

XANTHORRHOEACEAE 

Xanthorrhoea sp. 

HAEMODORACEAE 



H 
H 

aqH 

H 



TW 



E 
E 



Rare in damp places 



H 


E 


* 


H 


E 


* 


H 


E 


* 


H 


E 


Occasional 


H 


E 


* 


H 


E 


Occasional, only northern end of park 


H 


E 


Occasional 



Occasional, moist slopes 



TW 


E 


Occasional, moist slopes 


TW 


E 


Rare, moist slopes only 


H 


E 


* 


H 


E 


* 


H 


E 


Occasional 


H 


E 


Occasional 


H 


E 


* 


H 


E 


* 


H 


E 


Occasional 


H 


E 


Abundant esp. footslopes 


H 


E 


Rare 


H 


E 


Abundant 



Occasional along ridgetop plateaux 



46 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



Haemodorum planifolium R. Br. 

PHILYDRACEAE 

Philydrum lanuginosum Banks & Sol. 
ex Gaertn. 

ORCHIDACEAE 

Acianthus caudatus R. Br. 
A. exsertus R. Br. 

A. fornicatus R. Br. 
Chiloglottis formicifera FitzG. 
C. reflexa Druce 

Corybas aconitiflorus K.D. Koenig & Sims 

Dendrobium speciosum Sm. 

Diuris maculata Sm. 

Liparis reflexa (R. Br) Lindl. 

Microtis unifloia (Forst. f.) Reichb. f 

Pterostylis sp. 

JUNCACEAE 

Juncus articulatus ^ L. 

Juncus continuus L.A.S. Johnson 

J. planifolius R. Br. 

J. prismatocarpus R. Br. 

RESTIONACEAE 

Baloskion gracilis (R. Br.) B.G. Briggs 
& L.A.S. Johnson 

Empodisma minus (Hook.f.) L.A.S. 
Jotmson & D.F. Cutler 

Lepyrodia mulleri Benth. 

L. scariosa R. Br. 

CYPERACEAE 

Baumea arthrophylla (Nees) Broeck. 

B. teretifolia Palla 
Baumea sp. no v. 

Bolboschoenu fluviatilis (Torrey) Sojak 

Caustis flexuosa R. Br. 

Cyperus laevis R. Br. 

Eleocharis atricha R. Br 

E. sphacelata R. Br 

Isolepis inundatus Hook. f. 



H 



aqH 



H 


E 


Occasional, moist slopes 


H 


E 


* 


H 


E 


Occasional, moist slopes 


H 


E 


Occasional, only moist places 


H 


E 


* 


H 


E 


* 


epH 


E 


Occasional, shady rock outcro 


H 


E 


* 


H 


E 


* 


H 


E,S 


* 


H 


E 


* 


H 


A 


* 


H 


A 


* 


H 


A 


* 


H 


A 


* 



H 



H 



Occasional, lake margins 



Occasional, lake margins 



H 


A 


* 


H 


A 


* 


H 


A 


* 


H 


A,E 


* 


H 


A,E 


* 


H 


A,E 


* 


H 


A,E 


Occasional, rocky slopes, plateau tops 


H 


A,E 


Occasional, moist sites 


H 


A,E 


Occasioal 


aqH 


A,E 


Abundant, in open water 


H 


A,E 


* 



Proc. Linn. Soc. N.S.W., 128, 2007 



47 



VEGETATION HISTORY OF DRY LAKE, NSW 



Lepidosperma laterale R. Br 

L. longitudinale Labill. 

Lepironia articulata Domin. 

Schoenus brevifolius R. Br 

S. melanonstachys R. Br 

S. villosus R. Br 

CABOMBACEAE 

Brasenia schreberi Gmelin 

POACEAE 

Anisopogon avenaceus R. Br 

Aristida ramosa R. Br 

A. vagans Cav 

Austrostipa rudis Spreng. ssp. nervosa 
(J. Vickery) J. Everett & S.W.L. Jacobs 

Briza maxima ^ L. 

Cymopogon refractus (R. Br) A. Camus 

Dichelachne rara (R. Br) J. Vickery 

Digitaria ramularis (Trin.) Henrad. 

Echinopogon caespitosus ^ C.E. Hubb. 

E. ovatus ^ (G. Forst,) R Beauv. 

Entolasia marginata (R. Br) Hughs 

E. stricta (R. Br) Hughs 

Eragrostis leptostachya Steud. 

Imperata cylindrica R Beauv. var. 
major (Nees) C.E. Hubb. 

Microlaena stipoides (Labill.) R. Br 

Panicium simile Domin 

Paspalidium gracile (R. Br.) Hughes 

Paspalum dilatatum ^ Poir. 

Pseudoraphis paradoxa (R. Br.) Pilger 

Setaria gracilis ^ Kunth. 

S. pubescens R. Br. 

Themeda australis (R.Br.) Stapf. 

FERNS/FERN ALLIES 

SELAGINELLACEAE 

Selaginela uliginosa (Labill.) Spring 

OPHIOGLOSSACEAE 

Botrychium australe R.Br. 



H 


A,E 


Occasional 


H 


A,E 


Abundant, margins of lakes 


aqH 


A,E 


Very abundant, mostly open water 


H 


A,E 


Occasional, lake margins 


H 


A,E 


Occasional, lake martins 


H 


A,E 


* 



aqH 



H 



Occasional, open water only 



H 


A 


* 


H 


A 


Occasional 


H 


A 


* 



H 


A 


* 


H 


A 


* 


H 


A 


* 


H 


A 


Occasional esp. footslopes 


H 


A 


* 


H. 


A 


Occasional esp. open areas 


H 


A 


* 


H 


A 


* 


H 


A 


* 


H 


A 


Abundant esp. after burning 


H 


A 


* 


H 


A 


* 


H 


A 


* 


H 


A 


Occasional esp. disturbed footslopes 


H 


A 


* 


H 


A 


Occasional 


H 


A 


* 


H 


A 


Abundant 



Occasional, damp places 



48 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



DICKSONIACEAE 

Calochlaena dubia (R.Br.) M. Turner 
& R. White 

CYATHACEAE 

Cyathea sp. 
DENNSTAEDTIACEAE 

Hypolepis muelleri N.A. Wakef. 

Pteridium esculentum (Forst f.) 
Cockayne 

LINDSAEACEAE 

Lindsaea microphylla Sw. 
ADIANTACEAE 

Adiantum aethiopicum L. 

A. hispidulum Sw. 

SINOPTERIDACEAE 

Cheilanthes distans (R.Br.) Mett. 

C. austrotenuifolia Quirk & Chambers 

DAVALLIACEAE 

Davallia pyxidata Cav. 

BLECHNACEAE 

Blecknum cartilagineum Sw. 

Doodia aspera R.Br. 



Only in very moist gullies 

Rare, moist gullies 

Occasional, moist creek banks 
Abundant, esp. disturbed areas 

Rare, moist gullies 
Occasional, moist gullies 



Occasional, esp. rock outcrops 

Rare, on rockfaces in moist areas 

abundant, rocky, shaded slopes 
Occasional, moist slopes 




APPENDIX IB 



The species in the gully forest (site TS 6, Fig. 1). 



Species 




Family 




Presence outside 
gully 



Trees, 10-30 m 

Eucalyptus deanei Maiden 

E. elata Dehnh. 
E. piperita Sm. 

Small trees and shrubs < 10 m 

Doryphora sassafras Endl. 

Grevillea mucronata R. Br. 
Hakea salicifolia (Vent.) B.L. Burtt. 
Lomatia silaifolia (Sm.) R. Br 
Persoonia levis (Cav.) Domin. 
P. linearis Andrews. 



Myrtaceae 



+ 



Monimiaceae 
Proteaceae 



+ 

+ 
+ 

+ 



Proc. Linn. Soc. N.S.W., 128, 2007 



49 



VEGETATION HISTORY OF DRY LAKE, NSW 



P. mollis R. Br. 

Stenocarpus salignus R. Br. 

Pittosporum revolutum Dryand. 

Elaeocarpus reticulatus Sm. 

Lasiopetalum ferrugineum Sm. var, ferrugineum 

Bertya pomaderriodes F. Muell. 

Callicoma serratifolia Andrews 

Ceratopetalum apetalum D. Don 

C. gummiferum Sm. 

Acacia decurrens Willd. 

A. elata Benth. 

A. parramattensis Tindale 

Pultenaea flexilis Sm. 

Acmema smithii (Poir.) Merr. & Perry 

Backhousia myrtifolia Hook. f. & Harv. 
Tristaniopsis sp aff. laurina (smith) Peter G. Wilson 

& Waterhouse 

Leptospermum trinervium (Sm.) J. Thompson 

Allocasuarina torulosa (Aiton) L. Johnson 

Pomaderris intermedia Sieber 

Pomaderris sp. unidentified 

Exocarpos strictus R. Br. 

Correa reflexa Vent. var. reflexa 

Nematolepis squameum (Labill.) Eng. 

Dodonaea triquetra J.C. Wendl. 

Astrotricha latifolia Benth. 

Dracophpyllum secundum R. Br. 

Leucopogon lanceolatus (Sm.) R. Br. var. lanceolatus 

Logania albiflora Druce 

Notelea sp. unidentified 

Rapanea variabilis Mez. 

Dampiera purpurea R. Br. 

Cassinia aculeata R. Br. 

Ground cover, herbs and shrubs < 1 m 

Viola bentonicifolia Sm. 

Drosera auriculata Backh.ex Planch. 

Solanum sp. unidentified 

Corybas frimbriatus (R. Br.) Rchb. f. 



Pittosporaceae 

Elaeocarpaceae 

Sterculiaceae 

Euphorbiaceae 

Cunoniaceae 



Fabaceae 



Myrtaceae 



+ 



+ 



+ 
+ 



(6 


+ 


(C 


+ 


Casuarinaceae 


+ 


Rhamnaceae 


- 


Santalaceae 


- 


Rutaceae 

£6 


- 


Sapindaceae 


+ 


Aralaceae 


+ 


Ericaceae 


- 


tc 


+ 


Loganiaceae 


- 


Oleaceae 


- 


Myrsinaceae 


- 


Goodeniaceae 


+ 


Asteraceae 


+ 


Violaceae 


+ 


Droseraceae 


+ 


Solonaceae 


+ 


Orchidaceae 


+ 



50 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



Gahnia sp unidentified 
Hibbertia obtusifolia DC. 

Climbers 

Cassytha glabella R. Br. 

Sarcopetalum harveyanum F. Muell. 
Smilax australis R. Br. 
Cissus antarctica Vent. 
Eustrephus latifolius Ker Gawler 
Geitonoplesium cymosum R. Br. 

Pachycauls 

Cyathea australis (R. Br.) Domin. 

Ground ferns 

Todea barbara (L.) T. Moore 

Gleichenia microphylla R. Br. 

Stichems sp. 

Hymenophyllum cupressiforme Labill. 

Calochlaena dubia (R. Br.) M. Turner 

Pteridium esculentum (Forst.f.) Cockayne 

Adiantum aethopicum L. 

Cheilanthes austrotenuifolia Quirk & Chambers 

Pyrrosia rupestris (R. Br.) Ching 

Aspenium flabellifolium Cav. 

Blechnum cartilagineum Sw. 

B. nudum (Labill.) Mett. ex Luerssen. 

Doodia aspera R. Br. 



Cyperaceae 


- 


Dilleniaceae 


+ 


Cassythaceae 


- 


Menispermaceae 


- 


Smilacaceae 


+ 


vitaceae 
Luzuriagaceae 


+ 


a 


+ 



Cyatheaceae 


~ 


Osmundaceae 


_ 


Gleicheniaceae 


- 


Hymenophyllaceae 


- 


Cyatheaceae 


- 


Dennstaediaceae 


+ 


Adiantaceae 


+ 


Sinopteridaceae 


+ 


Polypodiaceae 


- 


Aspleniaceae 


- 


Blechnaceae 


+ 




Proc. Linn. Soc. N.S.W., 128, 2007 



51 



VEGETATION HISTORY OF DRY LAKE, NSW 



APPENDIX 2 

The identification of pollen of the family Myrtaceae. 

Fig. 14 illustrates the pollen characters used and Table 3 presents the distribution of these characters amongst 
the species. These morphological characters were insufficient to reliably identify species but they have been 
used to place the species in distinctive groups which are defined thus: 

Angophora/Corymbia group: Large-sized grains, 26 (30-40) 45 |a.m (mean in brackets), with x (rarely w) 
type pore and a thick exine, 2-4 \xm. 

Eucalyptus group: Medium-sized grains, 18 (21-25) 28 (am, with x (rarely w) type pore and medium-thick- 
ness exine, 1.5-3.0 |iim. 

Melaleuca/Leptospermum group: Small sized grains, 10(13-1 9) 23 jam, with only y and z type pore and thin 
exine, <1-1 |j,m. 



A POLAR VIEW 
Pots' island 
Thickness of 




Rounded angie of arr^ 
Colpus ^'""^ ^"^^ ^ '"^ 

Pwe 




i TYPE OF POUS 




C PORE TYPES 






Straight 



D SIDES OF AMB 







Concave 



Convex 



Figure 14. The morphological characters used to identify pollen grains of Myrtaceae 



52 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



Table 3. Pollen morphological characters used to identify myrtaceous pollen groups. ( ) infrequent 
occurrence. 



Exine 

thickness 

— Oim) — 



Species 



Equatorial diameter 
Mean (|am) Range 



Type of Type of Amb Amb Exine 

pole pore angle sides pattern 



Angophora/Coiymbia pollen group 



Angophora floribunda 30±1.7 (26-33) f (e) a x (w) round straight 



Corymbia eximia 
C. gummifera 



40.7±2.8 (35-45) a, c (f) x 

40.0±3.0 (33-45) b, a (f) x 



round straight 

round convex 



faint 2-3 

2-4 
2.5 



Eucalyptus pollen group 



Eucalyptus punctata 23.8±1.3 (18-27) 



E. piperita 



E. tereticornis 



E. globoidea 



21.0±1.3 (18-24 



23.4±1.3 



24.8±2.0 (20-28) 



w(x) 



x(w) 





straight & 




concave 




straight & 




concave 


round 


straight 


round 


straight & 
concave 



1.5-2 



1.5-2 



1.5-3 



Melaleuca/Leptospermum pollen group 



Melaleuca thymoides 


19.6±1.4 


(18-23) 


f(a) 


z 


round 


concave 


- 


tol 


M. linariifolia 




15.4±1.4 


(13-18) 


f 


z 


round 


concave 


- 


tol 


Leptospermum 




13.0±1.8 


(10-19) 


f 


z 


round 


concave 


_ 


tol 


jumperinum 
















coarse, 




L. trinervium 




13.5±1.4 


(12-16) 


f 


y 


blunt 


straight 


granular 


tol 


Tristaniopsis sp. 


ajf 


14.4±1.1 


(12-17) 


a 


z 


blunt 


straight & 


. 


tol 


laurina 














concave 
straight & 






Acmena smithii 




14 




f 


" 


round 


concave 


" 


tol 


Backhousia myrtifolia 


18.0 (14-21) 


f 


z 


round 


concave 


faint 


<1 



Proc. Linn. Soc. N.S.W., 128, 2007 



53 



VEGETATION HISTORY OF DRY LAKE, NSW 



APPENDIX 3 

Pollen type name on pollen diagrams and probable source in the vegetation. For full lists of species in each 
genus, see Appendix lA 



Pollen type 



Probable source in present day vegetation 




Pollen types found on both surface 

Eucalyptus 
Angophora/Corymbia 

Melaleuca/Leptospermum 

Myrtaceae 

Allocasuarina 

Pinus 

Cupressaceae 

Dodonaea 

Proteaceae 

Banksia 

Monotoca 

Tricolporates 

Tricolporate 2 (15 um grains) 

Pimelea 

Acacia 

Poaceae 

Plantago cf lanceolata 

Plantago cfvaria 

Chenopodiaceae 

Caryophyllaceae 

Brassicaceae 

Asteraceae Tubuliflorae 

Asteraceae Liguliflorae 

Zygophyllaceae 

Polygonaceae 

Gonocarpus 

Trilete spores 

Monolete spores 

Sellaginella 

Hydrocotyle 



sample and fossil pollen diagrams. 

Eucalyptus spp. 

A. floribunda, C. eximia, C. gummifera 

Melaleuca spp. and Leptospermum spp., Acmena smithii, 

Tristaniopsis sp., Backhousia myrtifolia. 

Any other species in the family 

A. torulosa, A. littoralis 

Pinus spp., most likely P. radiata 

Native Callitris or other intoduced species 

Dodonaea triquetra 

All species in the family, excluding Banksia spp. 

Banksia spp. 

Monotoca spp. 

Includes species from Fabaceae (excluding Acacia), Rutaceae, 
Dilleniaceae, Goodenia hederaceae, Ampera xyphoclada, 
Violaceae, Bursaria spinosa, Stylidiaceae 

Mainly Elaeocarpus reticulatus, Ceratopetalum spp. 

Pimelea linifolia 

Acacia spp. 

Poaceae species 

Plantago lanceolata (introduced) 

Plantago varia (native) 

Chenopodiaceae species (not in Appendix 1) probably herbs 

Caryophyllaceae, as above 

Brassicaceae, as above 

Asteraceae species, excluding Hypochoeris radicata 

Probably only Hypochoeris radicata 

Probably Tribulus terrestris, but the plant was not observed 

Persicaria decipiens, P. hydropiper, P. orientale 

Gonocarpus spp. 

Cyathea sp, Pteridium sp. Adiantum spp. Cheilanthes spp. 

Blechnum spp., Davallia sp., Doodia sp. 

Sellaginella uliginosa 

Hydrocotyle spp. 



54 



Proc. Linn. Soc. N.S.W., 128, 2007 



S. ROSE AND H.A. MARTIN 



Restionaceae 
Cyperaceae 
Myriophyllum 
Potomogeton 

Unknown 1 (inaperturate) 

Unknowns 



Restionaceae species 

Cyperaceae species 

Myriophyllum variifolium 

Potomogeton tricarinatus 

Inaperturate grain with coarse granular pattern, thin exine, 

20-25 um diameter. 

All other unidentified grains 



Other pollen types on surface sample pollen diagram 



Platysace 

Leucopogon 

Monosulcate 

Exocarpus 

Goodeniaceae 

Apocynaceae 

Portulacaceae 

Loranthaceae 

Lomandra 



Platysace linearifolia 
Leucopogon spp 
Liliaceae (sensu. lat.) 
Exocarpus spp. 
Scaveola ramossisima 
Parsonsia straminea 
Portulaca oleracea 
Mistletoe on Eucalyptus spp. 
Lomandra spp 



Proc. Linn. Soc. N.S.W., 128, 2007 



55 



56 



The History of the Vegetation from the Last Glacial Maximum 
at Mountain Lagoon, Blue Mountains, New South Wales. 

Anthony Robbie' and Helene A. Martin^ 

'St. James College, 25 Mary St Cygnet, Tas. 7112 
2 School of Biological and Environmental Sciences, University of New South Wales Sydney 2052 

(h.martin@unsw.edu.au) 



Robbie, A. and Martin, H.A. (2007). The history of the vegetation from the last glacial maximum at 
Mountain Lagoon, Blue Mountains, New South Wales. Proceedings of the Linnean Society of New South 
Wales 128, 57-80. 

Mountain Lagoon in the Blue Mountains west of Sydney provides a sedimentary record of 23,000 years, 
thereby including the Last Glacial Maximum. Initially, the site was a lake where clay was being deposited 
and the vegetation was probably shrubland/herbfields. About 18-19 kyr, the lake became shallow enough 
for sedgelands and peat formation. At this time, pollen concentrations were high and both Casuarinaceae 
and Myrtaceae are prominent. In the early Holocene, about 10 kyr, the swamp became a lake again, perhaps 
because of some minor movement of the fault-line which could have caused a burst of accelerated erosion 
and clay deposition. The lake surface was re-colonized by the sedgelands again about 7-8 kyr, when the 
vegetation was woodland'forest. 

The vegetation surrounding the site was sclerophyllous throughout the last 23 kyr, as would be 
expected on these low nutrient soils. In contrast to the likely marked climatic changes during this period, 
the pollen specfra show remarkably little change in the major taxa. However, variations of some of the 
Myrtaceae pollen show that there were species changes, although some taxa were present the whole time. 
Casuarinaceae was prominent throughout and did not decline until European settlement. 

Manuscript received 3 July 2006, accepted for publication 1 8 October 2006. 

KEYWORDS: Blue Mountains, Holocene, last glacial maximum, Mountain Lagoon, 
palynology, vegetation history. 



INTRODUCTION 

Mountain Lagoon (Fig. 1), in a small enclosed 
basin, was formed following subsidence along the 
Kurrajong Fault line and the subsequent disruption 
to the established drainage patterns. The sediments 
extracted for this study record at least 23,000 years 
of deposition, which includes the Last Glacial 
Maximum (LGM) period at about 18,000 years ago. 
Estimates of the LGM from other records indicate that 
temperatures were some 4-8°C lower than today and 
the climate was also more arid (Dodson 1994) with 
up to 50 % less precipitation (Thom et al. 1994). The 
site stands at just over 500 m elevation today, but with 
lowered sea levels during the glacial period, it would 
have been about 100 m higher in elevation at that 
time. Hope (1989) estimates that this altitude would 
have been near or at the treeline during the last glacial 
period. Studies of the glacial period in southeastern 
Australia show that the vegetation of the time was 



more open, with few trees and more grasslands and 
shrublands (Dodson 1994; Hope 1994), but Pickett 
et al. (2004) think that xerophjdic woods and scrubs 
were more extensive in south-western and south- 
eastern Australia. 

There are few histories of the vegetation 
extending back beyond the last glacial period in the 
Sydney Basin. Chalson (1991) found that the Penrith 
Lakes Swamp (Fig. 1) provided a 33,000 year record 
and Black et al. (2006) present a >43,000 year history 
at Thirlmere Lakes, but these are both lowland sites. 
At Readhead Lagoon, a coastal site (Fig. 1), Williams 
et al. (2006) record a history that goes back well 
before the LGM. Chalson also presents a number of 
other sites in the Blue Mountains which are all 1 1 ,000 
years or younger in age and Black and Mooney (2006) 
present a 14,000 year history of Gooches Crater on 
the Newnes Plateau. Mountain Lagoon thus provides 
a record of the changes in the vegetation through 
the glacial period to the present at a relatively high 
altitude in the Blue Mountains. 



VEGETATION HISTORY OF MOUNTAIN LAGOON 




t 



B 



MOUNTAIN 

LAGOON ► 




AGOON^ / 



^r'^j-^^ Bi'P'" y'^ 






SYDNEY BASIN 



-I 1 

50 100 km 




Qreat /* Springwood 



Katoomba 



10 km 



' V.^ Penrith Lakes 

"^ X Penrith "'" 



Figure 1. Regional locality map showing study site and place names discussed in text. 



THE ENVIRONMENT 

Geology 

Mountain Lagoon is a shallow, swampy lake in 
a small basin-shaped valley (Ryan et al. 1996) 14 km 
north-east of Bilpin in the eastern Blue Mountains 
(St. Albans G.R. 663966), It was described by Grady 
and Hogbin (1926) as resembling an 'over-turned 
saucer", and therefore all sediment within the basin 
is derived from within its own catchment area which 
measures approximately 2 km^. The lagoon lies on 
top of a small, thin lens of Wainamatta Shale which 
is underlain by the Triassic Hawkesbury Sandstone 
(Grady and Hogbin 1926). It is a tectonic lake (Timms 
1992) abutting the Kurrajong Block, and was formed 
following the subsidence of the land to the west of the 
Kurrajong Fault (David 1902). The Kurrajong Block, 
which rises some 120 m above Mountain Lagoon on 
its eastern side and stretches south approximately 25 
km to Glenbrook (David 1902), is believed to have 
impeded the north-eastern progress of a small stream 
whose waters pooled at this barrier and formed the 
lagoon (Grady and Hogbin 1926), possibly in the late 
Tertiary (Branagan 1969). 



Climate 

Wedged between the coastal ranges, and the 
Upper Blue Mountains and Great Dividing range, 
the St Albans region is mostly in a rainshadow and 
is a relatively dry part of the Hawksbury-Nepean 
catchment. Rainfall is generally over 900 mm, but 
Bilpin, some 7-8 km WSW of Mountain Lagoon, 
seated in front of Mount Wilson, experiences higher 
orographic rainfall, and receives 1300 mm p.a. (Ryan 
et al., 1 996). Records kept by a landholder at Mountain 
Lagoon for the period 1952-1994 show an average 
annual rainfall of 1257 mm, with January-February 
the wettest months, with an average of 157-181 mm 
per month, and August-October the driest months, 
with an average of 5 1-70 mm per month (Hungerford 
1995). 

Average maximum temperature for January is 
28° C and average minimum temperature for July is 
2-3°C (Ryan et al. 1996). 

The Vegetation 

Prior to historic land clearance for forestry 
and agriculture, the rich, moist soils of the shale 
lens supported a tall open forest dominated by 
Eucalyptus deanei, E. cypellocarpa and Syncarpia 



58 



Proc. Linn. Sec. N.S.W., 128, 2007 



A. ROBBIE AND H.A. MARTIN 



glomulifera, specimens of which have survived 
in small patches of forest which remain in the area 
(Ryan et al. 1996). Two significant species with very 
restricted distribution are found in these forests, viz. 
Acacia pubescens and Alania endlicheri. The lagoon 
itself supports a freshwater reed swamp dominated 
by sedges, with a main canopy of Lepidosperma 
longitudinale and a fringing Melaleuca linariifolia 
forest. The sheltered western slope of the Kurrajong 
Block supports a Sydney Sandstone Gully Forest 
dominated by Angophora costata. Eucalyptus 
piperita, E. agglomerata and Syncarpia glomulifera. 
On the exposed ridges at the top of the Kurrajong 
Block, Corymbia eximia, Angophora bakeri, C. 
gummifera, A. costata and Eucalyptus punctata are 
dominant (Ryan et al. 1996). 

Three small patches of basaltic soils at Green 
Scrub to the south of the lagoon support a warm 
temperate rainforest (Floyd 1989). Prior to European 
arrival the forest was most likely dominated by 
Dorifera sassafras, Acmena smithii, Toona ciliata 
and Ceratopetalum spp., but repeated firing and 
logging have greatly altered the forest and continue 
to threaten the floral composition of this forest (Floyd 
1989). 

Human history and land use 

Archaeological evidence tends to suggest that 
Aboriginal people first settled in the region firom 
20,000 to 14,000 years before the present (BP) and 
that many sites may have been abandoned at 12,000 
years BP, to be followed by 'a more intensive phase of 
occupation' beginning around 10,000-5,000 years BP 
(Conyers 1987). Accounts by early European settlers 
suggest the region was well known to the Dharruk 
and possibly Wiradjuri groups, who had traditional 
names for prominent landforms such as Mt. Tomah, 
and advised on the more accessible routes over the 
mountains The area is culturally significant to the 
Dharruk and the raised area to the immediate west 
of the lagoon was used as a bora ground as late as 
the 1890s (Hungerford 1995). The region is encircled 
by sites with rock engravings, cave paintings and axe 
grinding sites (Stockton 1993). 

Europeans such as Mathew Etheringham and 
the botanist George Caley began exploring the 
mountains from Kurrajong Heights around the turn 
of the 19* Century (Hungerford 1995). The existence 
of Mountain Lagoon was known to Europeans before 
1830 and the area was frequented by shooters. Later 
timber extraction and milling became important in 
the area with the removal of 'wattle barks, blue-gum 
and other hardwoods', and it is likely that the lagoon 
formed part of a stock route linking the Hunter region 



with Bathurst. The land to the west of the lagoon was 
first squatted and was later purchased in 1868, and 
a mixed orchard of oranges, lemons, cherries, and 
apples was established, along with maize, oats and 
potatoes. Orchards spread in popularity across the 
region throughout the 20* century, and strawberries 
were introduced in the area in the early 1970s. 
Orchards have largely disappeared from the area 
since 1975, and the land surrounding the lagoon 
supports mostly cattle grazing with some citrus and 
apple growing (Hungerford 1995). 



METHODS 

Six sites, each within different environments 
in the vicinity of the lagoon (Fig. 2) were chosen to 
determine the major variations in vegetation, using 
aerial photographs and onsite inspections. A full list 
of species at Site 6 (Green Scrub rainforest) was 
obtained from P. Hind of the Royal Botanic Gardens, 
Sydney. 

In the latter part of the 1980s the lagoon was 
greatly modified in the hope that it would become a 
permanent source of water for cattle. Sediment was 
excavated from the northeastern end of the lagoon 
and deposited towards the south-western end (Fig. 3). 
The results of the excavation were obvious in 1991 
when the original core was taken, (Mr. C. Myers, pers. 
comm. 1996) and an undisturbed site was chosen. 

The stratigraphy along two transects at right 
angles was evaluated using a Hiller corer and the 
sediments were described using the Troels-Smith 
method for sediment description (Moore et al. 1991). 
Two cores for further analysis were taken fi-om a place 
as close as possible to the original site (cored by C. 
Myers), using a Livingstone type corer (Livingstone, 
1955) with modifications (Neale and Walker 1996). 

Two peat sediment samples taken at depths of 
33-38 cm and 59-68 cm fi-om the original core were 
radiocarbon dated by the Beta Analytic Company in 
1991 (Table 1). The top 15 cm of the original core was 
discarded in the belief that this section was disturbed. 
Two samples fi-om the clay extracted in later cores, at 
depths of 60-70 cm and 90-100 cm were dated by the 
Accelerated Mass Spectrometry method at ANTSO 
(Table 1). 

Organic matter was estimated on oven-dried 
(105°C) samples fired to 550°C, at 10 cm intervals. 
During ignition, structurally bound water is lost also, 
but in highly organic sediments, the major loss is 
from the ignition of organic matter (Bengtsson and 
Enell 1990). 

The saturated isothermal remnant magnetism 



Proc. Linn. Soc. N.S.W., 128, 2007 



59 



VEGETATION HISTORY OF MOUNTAIN LAGOON 




H Lagoon 
(TH Melaieuca Forest 
O Tall Forest 
[Z3 Hilltop vegetation 



FTl Exposed escarpment 
im] Protected escarpment 
Q Wami temperate rainforest 
LJ Cleared 



Figure 2. Vegetation survey sites (numbers) and 
map of the vegetation types. 

(SIRM) was measured on sub-samples taken at half 
centimetre intervals from the original core. The 
sediment was dried at 50°C, ground and treated in 
a magnetic field of 1.0 Tesla (Thompson 1990) and 
measured with a Molspin Magnetometer. 

For pollen extraction, sediment samples 
of 1 cm^ were taken at 10 cm intervals along the 
second core and were spiked with an exotic pollen 
suspension {Alnus rhombifolia was used) of known 



concentration (Birks and Birks 1980). Humic acids 
were removed with cold 10% potassium hydroxide 
and mineral matter was removed using hydrochloric 
and hydrofluoric acids. The residue was treated with 
acetolysis to clear remaining humic material (Moore 
et al. 1991). The residues were mounted in glycerine 
jelly, using No. coverslips. 

Pollen was identified by comparison with modem 
reference pollen. A minimum of 180 pollen grains 
were counted along traverses on the slides of the 
fine sediment residues. Where there was insufficient 
pollen to count 1 80 grains, those pollen types present 
were scored as 'present'. The number of the exotic 
Alnus grains encountered along the traverses were 
counted also, allowing calculation of the pollen 
concentration. The abundance of each pollen t5^e was 
expressed as percentages and as pollen concentration. 
Confidence limits for percentages were calculated 
following Maher (1972). The amount of charcoal in 
each preparation was determined as the area of the 
slide it covered, following the point count method 
(Clark 1984). 



RESULTS 

The vegetation 

Six vegetation units in the vicinity of the lagoon 
were defined and a list of species found in each is 
presented in Appendix 1. The vegetation units are 
shown in Fig. 2 and were defined as follows: 

1 . The swamp vegetation of the lagoon itself is a 
fen which becomes dry periodically. The centre 
of the lagoon is dominated by Baumea articulata, 
with Nymphoides geminata and Myriophyllum 
variifolium at the margins. 

2. The lagoon fen is ringed by a Melaleuca swamp 
forest, c. 5 m tall, with a 50% cover of Melaleuca 
linariifolia and an understorey of Leptospermum 
polygalifolium and Acacia filicifolia. The ground 
cover consisted of Lepidosperma longitudinale. 
Sphagnum sp. and Viola hederacea. North-east 
of the lagoon, the vegetation has largely been 
cleared for grazing and in this area, small M 
linariifolia emerge above an understorey of 
Acacia longifolia and L. polygalifolium. The 
ground cover in this area consists of Gleichenia 
dicarpa, Hypolepis muelleri and V. hederacea on 
Sphagnum peat. 

The fen and Melaleuca swamp forest together 
make up the Lepidosperma longitudinale- 
Melaleuca linariifolia Sedgeland (Ryan et al., 
1996) which is related to other low nutrient 
wetlands, such as the Thirlemere Lakes. 



60 



Proc. Linn. Soc. N.S.W., 128, 2007 



A. ROBBIE AND H.A. MARTIN 



Table 1. Radiocarbon dates. Calibrated years has been calculated according to the Radiocarbon 
Calibrated Program CaUb Rev5.0.2 (Stuiver and Reimer, 1986-2005) 



Depth (cm) Sample number Technique 



Age (Radiocarbon 
years) BP 



Calibrated years BP 
(cal. yr) 



33-38 



Beta 43680 



Standard C'^ 



9,040 ± 90 



10,079 



59-68 



Beta 43681 



Standard C'^ 



18,660 ±150 



22,230 



60-70 



OZD666 



AMSC* 



19,350 ±220 



23,036 



90-100 



OZD667 



AMSC'^ 



19,700 ± 390 



23,484 



sw 

Depth (cm) 



3. The tall forest has a 60-70% 
canopy cover oi Eucalyptus deanei 
and a sub-canopy of Syncarpia 
glomulifera and Pittosporum 
undulatum. Leucopogon spp. and 
climbers such as Smilax form 
much of the understorey in this 
forest and a variety of ferns form a 
thick ground cover. In the cleared 
areas to the north-east of the 
lagoon, a few E. deanei and some 
Eucalyptus piperita were found 
on the drier soils near the lagoon. 
There was no understorey in this 
area. The ground cover consisted 
largely of introduced grasses, with 
Pteridium esculentum growing 
close to the lagoon. 

4. On theexposed, rocky escarpment 
of the Kurrajong Block, the 
well drained soils support an 
open woodland dominated by 
a 40-50% cover of Eucalyptus 
piperita. Eucalyptus agglomerata 
and Syncarpia glomulifera. The 
understorey components are 
chiefly sclerophyllous species, e.g. 
Banksia spinulosa and Telopea 
speciosissima with Acacia elata 
quite common. 

5. The protected gully of Gospers 

Creek, the outlet of the lagoon, supports a closed 
Turpentine {Syncarpia glomulifera) forest with a 
canopy cover of greater than 70%). It has both 
mesic and xeric components and is dominated by 
S. glomulifera, Angophora costata and E. elata. 
The understorey is dominated by tall Banksia 




Depth (cm) 



20 



40 



60 



80 



100 



30 m 



20 



40 ■ 



60 



80 



100 




Core site 
tor study 



Herbaceous 
peat 



|0\/] Lake mud 
rrn Clay 



Sand 



Figure 3. Mountain Lagoon, depicting disturbed sites, strati- 
graphic transects and the site of the core for this study. 



serrata and Pittosporum revolutum, with Lomatia 
silaifolia,Leucopogonjuniperinus, L. lanceolatus 
and Xanthorrhoea arborea. The ground cover 
consists of Viola hederacea, Dianella caerulea 
and Echinopogon ovatus. See Ryan et al. (1996) 
for further descriptions of gully forests in the 



Proc. Linn. Soc. N.S.W., 128, 2007 



61 



VEGETATION HISTORY OF MOUNTAIN LAGOON 



Depth 
(cm) 



C14 
dates 



20— 



9,040 ± 90 ■ 



-^ 



18,660 ±150- 













1- 




L 
L 




I- 




Sediment type 
C14 
dates 
Black moss peat 



Black herbaceous peat 

Dark brown herbaceous peat, 
minor lake mud 



Light brownish clay 



Black clay, minor lake mud 

Dark grey herbaceous peat 
trace of clay 

Dark reddish brown herbaceous 
peat 
19,350±220 



Dark greyish brown herbaceous 
peat and lake mud 



Dark grey lake mud, trace of sand 



19,700±360 



too -| * IP Greyish brown clay 



Greyish brown mottled clay 



Figure 4. The sedimentary column. Standard C14 
dates are on the left and AMS C14 dates are on the 
right. Dates are given in radiocarbon years. For cali- 
brated ages, see Table 1. 

region. 

6, Green Scrub, on a small lens of basalt soils in 
a protected gully south of the lagoon, supports 
warm temperate rainforest and is dominated 
by Doryphora sassafras, Acmena smithii and 
Syncarpia glomulifera. See Appendix 2 for a full 
list of species. 

Stratigraphy 

The stratigraphic transects and cross sections of 
the lagoon are shown in Fig. 3. Part of the lagoon 



in the north has been excavated and the spoil 
dumped in a patch on the western side. Only the 
north-western half of the SW-NE cross section of 
the lagoon is regarded as undisturbed. Sediment 
descriptions of the core are shown in Fig. 4. 

A layer of moss peat covers the lagoon to a 
depth of about 15 cm in most areas (Fig. 3). There 
are minor patches of herbaceous peat on top of 
the moss peat, but they are associated with the 
disturbed areas. Herbaceous peat underlies the 
moss peat in the study core (Fig. 4) but it is not 
evident in the cross sections. A layer of brownish 
clay is found across the whole of the lagoon, 
underlain by black clay and/or lake mud over 
part of the lagoon. A relatively thick layer of 
herbaceous peat underlies the clay, with a layer 
of lake mud (very fine organic matter), and at the 
base, clay. There is some mottling in the deepest 
layers of the basal clay layer. Traces of sand are 
found in some of the deeper clays and lake muds. 

Table 1 presents the radiocarbon dates. 
Assuming continuous sedimentation (Fig. 5), the 
Holocene extends down to about 40 cm in the 
study core, to the base of the upper clay layer 
(Fig. 4) and the overall rate of sedimentation 
approximates 4 cm per k cal. yr. The height of 
the last glacial period (18 k cal. yr) is recorded at 
about 60 cm depth, hence during the time from 
the last glacial maximum to the begirming of 
the Holocene, the rate of sediment accumulation 
was about 2.5 cm per k cal. yr. This latter rate 
continues till about 22 k cal. yr, after which rate 
of sediment accumulation, was rapid, about 10 
cm per k cal. yr. 

The SIRM, microscopic and macroscopic 

charcoal content and carbon content of the 

sediments is shown on Fig. 6. Peak values for all 

of these factors are found in the peat and values 

are lower in the clay. 



Sedimentary history 

Initially, Mountain Lagoon was a lake with 
water too deep for rooted vegetation. Clay is 
usually an indication of a low energy environment, 
but at this location, the Wainamatta Shale weathers 
to produce predominantly clay. Moreover, with 
the lagoon situated at the base of the escarpment 
of the Kurrajong Fault, any tectonic movement or 
vegetation disturbance, even if only slight, could cause 
instability, and the accelerated erosion may contribute 
to the deposition of clay in the lagoon. It is thought 
that some instability of the fault escarpment probably 
contributed to the rapid rate of clay accumulation at 
the base of the profile. 



62 



Proc. Linn. Soc. N.S.W., 128, 2007 



A. ROBBIE AND H.A. MARTIN 



Depth 
(cm) 



ZONE VEGETATION 




too 





Moss peat 


Herbaceous 
peat 


m 


Lake muds 


L L 

L 
L U 



Clay 



Radiocarbon date 
4 Calibrated age 



Figure 5. Summary diagram of tlie history of Mountain Lagoon. This model assumes continuous depo- 
sition (see text). For sedimentary symbols, see Figure 4. 



SIRM 



-^-> 



9.040 ± I 



^ 



18.660 ± 
19.350 ± 



19,700 ± 



a 



D 0] 



20 



40 • 




60 



80 



L t |> 



100 



MICROSCOPIC MACROSCOPIC rARBON rONTENT 

CHARCOAL CHARCOAL CARBON CONTENT 



5 X A.m^.kg-' X1 0^ 20 mm^/slide 








+ 
+ 



80% 



20% 



Figure 6. SIRM, microscopic and macroscopic charcoal content and carbon content. For the macro- 
scopic charcoal content, the more '+'s, the more the charcoal, 'o' equals zero macroscopic charcoal. For 
lithologic symbols, see Fig 4. Dates are given in radiocarbon years. For calibrated ages, see Table 1. 



Proc. Linn. Soc. N.S.W., 128, 2007 



63 



VEGETATION HISTORY OF MOUNTAIN LAGOON 



The mottled clay at the base of the profile indicates 
a fluctuating water table and occasional dry periods in 
the earlier part of the glacial period. Towards the end 
of the peak glacial period, the lake became shallow 
enough to allow rooted vegetation and the production 
and preservation of peat. 

At the beginning of the Holocene, there is a layer 
of light brown clay, which is unusual when compared 
to other sites. The colour is also unusual, for if clay 
is deposited slowly under the anaerobic conditions of 
a lake or swamp, it would become grey or black. It is 
possible that some instability of the escarpment may 
have triggered a short burst of intensified erosion and 
deposition of this material. With a return to stability, 
the vegetation recolonized the swamp surface and the 



Table 2. The identification of Myrtaceae pollen in Mountain 
Lagoon sediments. The unidentified Myrtaceae types are depicted 
in Fig. 7. 



Depth (cm) in profile 



10 



20 



50 



Acmena smithii 
Angophora costata 
Corymbia gummifera 
Eucalyptus creba 
E. punctata 
E. deanei 
E. piperata 
E. haemostoma 
Syncarpia glomulifera 
Leptospermum spp 
Mytyaceae type I 
Mytyaceae type II 
Mytyaceae type III 
Myrtaceae type IV 
Mytyaceae type V 
Unidentified Myrtaceae 



5.9 



3.9 



+ 



29.4 



5.9 



7.8 



+ 



+ 



4.9 



+ 



+ 



+ 



4.9 



+ 



7.4 



8.8 



16.4 



5.9 



9.1 



+ 



6.1 



+ 



6.1 



13.7 6.1 



+ 



8.2 



+ 



7.4 



11.5 14.7 



70 



30.3 42.6 36.8 + 



21.6 18.2 11.5 19.1 32.0 



deposition of peat continued through the rest of the 
Holocene. 

It has been suggested that the model of 
continuous deposition adopted above may not apply 
and the basal clay may have been deposited in the 
glacial period, with an hiatus irom about 17 kyr to 
the Holocene, when peat deposition commenced. It is 
difficult to rule out the discontinuous model with only 
four dates, but it is harder to accommodate the dating 
into a discontinuous model. The two dates of 18,660 
and 19,350 radiocarbon years (22,230 and 23, 036 
calibrated years, respectively, see Fig. 4) both come 
from within the base of the peat/lake muds, which the 
discontinuous model assumes is Holocene. Further 
implications of the two models are discussed below. 
Charcoal is found in all of the 
samples, suggesting that burning 
could have occurred at any time. 
Charcoal content, however, is higher 
in the peat, when the vegetation 
growing on site may have burned 
and deposited charcoal directly into 
the sediments. When the lagoon was 
a lake and depositing clay, charcoal 
would have to be transported into 
the site, either by wind or water. The 
higher macroscopic charcoal content 
of the peat probably indicates woody 
shrubs were growing very close to 
the site of deposition. 

The SIRM values of the 
sediments closely parallel the 
charcoal input and both are higher 
in the peat. Commonly, high SIRM 
values correspond to a high mineral 
content in the sediment (Thompson 
and Oldfield 1986) but fire has been 
found to increase the soil magnetism 
to some extent (Rummery 1983). In 
these sediments, fire seems to have 
had a greater influence on the SIRM 
values than the mineral content. 



Pollen Analysis 

In an attempt to identify the 
myrtaceous pollen, a reference set 
of eleven species from the study 
area was examined in detail, using 
the method outlined by Chalson and 
Martin (1995). Ten of the species 
could be identified specifically 
in the profile (Table 2) but five 
common types (Fig. 7) found in 
the profile were not amongst the 



+ 



4.0 



12.0 



8.0 



16.0 



24.0 



64 



Proc. Linn. Soc. N.S.W., 128, 2007 



A. ROBBIE AND H.A. MARTIN 



reference set. Specific identification of fossil grains 
was often limited by distortion, poor preservation 
or being obscured by extraneous matter, hence the 
high proportion of unidentifiable Myrtaceae pollen. 
Specific identification is time consuming hence only 
a few levels of the profile have been studied in this 
detail (Table 2). Appendix 3 presents the name of 
the pollen type on the pollen diagram and the likely 
source of the pollen in the vegetation. 

The pollen diagram (Fig. 8) shows the percentages 

of total pollen count, pollen concentrations for the 

most common specific pollen types and total pollen 

concentrations. The profile has been divided into four 

zones based on pollen content and concentrations: 

Zone A: 100-85 cm, c. > 23 k cal. yr (see Fig. 

5 for approximate dates). Pollen concentration 

is very low, with a moderate representation of 

Casuarinaceae and Myrtaceae. T. pleistocenicus 

Martin 1973, a 'spineless' Asteraceae, cf 

Cassinia arcuata, Calomeria and possibly others 

(Macphail and Martin 1991), has the highest 

percentage for the profile, which, however, 

is not much. Other shrubs are restricted to 

Monotoca and Hakea, and the herb group is well 

represented. The aquatic group of Cyperaceae 

and Myriophyllum have low representation. 

Zone B: 85-35 cm, c. 23-10 k cal. yr. Pollen 

concentration is the highest for the profile (except 

for the very top). Casuarinaceae and Myrtaceae 

have low percentages but the concentrations are 

high. Shrubs are well represented in the lower 

part of the zone and the herb content is slightly 

higher than the other zones. The aquatic group 

has high percentages and concentrations. There 

appears to be a negative correlation between 

Cyperaceae and Myriophyllum. 

Zone C: 35-10 cm, c. 10 k cal. yr - ? present. Total 

pollen concentration is low and percentages for 

Casuarinaceae and Myrtaceae are high. There 

is a poor representation of shrubs and aquatic 

percentages and pollen concentrations are low. 

ZoneD: 5 cm, ? present. Total pollen concentration 

is exceptionally high and Casuarinaceae and 

Myriophyllum have the highest concentrations. 

Shrubs are poorly represented, and herbs are 

diverse. 

From the glacial period to the present, trees would 
have been almost entirely species of Casuarinaceae and 
Myrtaceae. Both of these families, however, contain 
shrubby species and even the same tree species may 
assume a shrubby habit under harsh conditions, e.g. 
Eucalyptus stricta is a mallee and both Eucalyptus 
pulverulenta and Corymbia gummifera may be a 
tree or mallee in the Blue Mountains today (Plantnet 








Scale bar = 5 |im 



Figure 7. Common, unidentified Myrtaceae pollen 
types. 

2006). The habit of the species cannot be determined 
fi-om the pollen, but since Mountain Lagoon is likely 
to have been at or above the treeline during the glacial 
period (Hope 1989), shrubby species are a possibility. 
By the time of the Holocene, when the climate was 
more like that of today, they were probably trees. 

Taxa within the family Casuarinaceae are generally 
not identifiable firom their pollen. Casuarinaceae was 
not found in the survey of the vegetation (Appendix 
1), but the pollen is wind distributed and some of it 
may travel a long way. Today, Allocasuarina torulosa 
is most likely in this region (Ryan et al. 1996). 
Casuarinaceae pollen is present throughout the profile, 
with higher percentages in the Holocene. 

Percentages of Myrtaceae pollen in Zone B, 
the period between the glacial maximum and the 
Holocene, are moderate, increasing in the Holocene, 
and decreasing to the present. Concentrations in 
Zone B, however, are very high, and during this time, 
Eucalyptus deanei, E. piperita and Myrtaceae type III 
were prominent in the vegetation. In the Holocene, E. 
deanei was still the most prominent, but Eucalyptus 
creba is the most common in the top of the profile. 
Pollen of Melaleuca linariifolia was not identified 
in this study, but Rose and Martin (this volume) 
found that it was indistinguishable fi-om pollen of 
Leptospermum spp. and >15 % in surface samples 
was recorded where M. linairifolia was dominant 
in the vegetation. In this study, M. linairifolia and 
Leptospermum spp. are included in the Myrtaceae 
group. 

Shrub taxa are most diverse in the lower part 
of zone B, and being low pollen producers, only 
their presence is recorded for most of them. T. 
pleistocenicus, {Cassinia arcuata, Calomeria and 
probably others) is found throughout the profile and it 
may be common in some glacial and older sediments 



Proc. Linn. Soc. N.S.W., 128, 2007 



65 



VEGETATION HISTORY OF MOUNTAE^ LAGOON 



"^ a Ti 

^ O ,£!* 
"^ S c 



o o 



00 

- -« ^ 

« S". o 

S, « TO 

era s S 

i%> Cr on 

-* O ST" 

« 55- o 
ffi -- :^ 

H ffi B 
» ffi (TQ 

l_i TO .^ 
S? TO 

n a 
»» S- 

'-" ^ 

n T3 

P* O 



a " 

= i 

re " 

^1 



re 

s 



re 

s 

ST o 
s 

re 
re 

s s 
ss sj 

3 i 

re S^ 

o 
s 



re 



o 
s 



ST O' 

re 1-1 

St S^ 

S9 re 

TO s 

S « 

SS O 



re 


s 


> 3 


"O 


o 


•a 


s 


re 


■a 


a 


o 






>«! 


re 


iji 


9 


• 


'J- 


O 


TS 


p 


re 


rt- 


vs 


re 




VI 


&> 




9 

a 


re 


rt- 


TO 


% 


< 


» 


re 


"S 




o 


h^» 




9 


re 


•t 


9 


»9 


re 

o 


o 


9 


re 


re 


» 


re 


>i 


S 


o 
s 


&9 



Vt 

8 

W 



<0 OD 

CO 'm 

vn o> 

o o 

H- tf 



r 101<J^Ctf?<|- ^S/tfh) 



8 > 



9_ 
S 



re s 



<o + 



3 



■r 
+ 



e. ■ 



CD 






■ !♦ 



1 '♦ 'I 

'l l' 00 IT 

IT ^1 



Oil 



> ^ 



T- 

it 



:» 



-f — T" 



I* 



:♦ :♦ 

-f — r 



"♦ 



O 8 



4 



TT rw [« TF" 



> !^ !* 



-r 



I 

I 
I 






6: 



^M y Sn . £^ 



'4 ' 
•T li 



■^ 



■^f 



I I ■ 



f 






^oO> 



*«^ 

vs 






"«* 



'*<■ 



'^^ 



*« 



%, 






9ft 



% 



% 









J3 

m 
m 

CO 

55 

I 

JO 

c 

GO 
CO 



T" T 

■• !» > 



" '♦ 'f 



w 



■^ 






% 



c 
o 

CO 









o 3 

g 5 

Z r- 

2 -D 
m o 

3 m 

o 

z 



(Martin 1973; Edney et al. 1990; 
Macphail and Martin 1991). Today, 
Cassinia arcuata may be common 
on disturbed mineral soils (Macphai 
and Martin 1991) and Calomeria 
is found along streams and may be 
abundant after fire (Botanic Garden 
Trust 2005). 

The herbaceous taxa are 
present throughout the profile and 
concentrations are higher in B 
Zone. Asteraceae (Tubuliflorae), 
Gonocarpus, Chenopodiaceae and 
Poaceae are the main taxa, but 
individually, are only present in low 
abundance. 

The aquatics Cyperaceae and 
Myriophyllum indicate the extent of the 
sedgeland vegetation. They are low in 
Zone A (glacial period) and the lower 
levels of Zone C (early Holocene) in 
the clay sediments. At these times, the 
lagoon would have been more of a 
lake with a fringing sedgeland. Zone 
B (post-glacial, pre-Holocene) has 
high percentages of Cyperaceae and 
Myriophyllum, where the sediments 
are almost entirely organic, when 
the sedgeland would have covered 
most of the lake. The inverse 
relationship between Cyperaceae and 
Myriophyllum probably reflects subtle 
changes in the water depth and the 
vegetation mosaic. 

The high pollen concentrations in 
Zone B are found in the lake muds, 
very finely divided organic matter. 
In the lake muds, the plant fi-agments 
of the peat have been mostly broken 
down, probably compacting the 
original peat and at the same time, 
concentrating the pollen content. 
Percentages suggest that the pollen 
content of Casuarinaceae and 
Myrtaceae decrease in Zone B, but 
concentrations show that this was not 
so: in fact the pollen concentrations 
have increased considerably. The 
better representation of shrubs and the 
increase in the herbs and considerable 
increase in aquatic pollen would mean 
that proportionately, other groups 
are less well represented (in terms 



66 



Proc. Linn. Soc. N.S.W., 128, 2007 



A. ROBBIE AND H.A. MARTIN 



of percentages). The low pollen concentration of 
aquatics in Zone C, the early Holocene, is probably 
the result of the change of habitat caused by the 
clay deposition, making it unsuitable for aquatics. 
Towards the present, the sedgeland vegetation was 
re-established. 

Freshwater algal spores of species of 
Zygnemataceae were found in the sediments and 
Debarya sp., cf Mougetia viridis, cf M. elegatula, 
Spyrogyra sp. and Zygnema sp. were identified. 
Botryococcus braunii and spores of Cyanobacteria 
(Churchill 1960) were also present. Characeae 
oospores were found at 5 cm in sieved material, before 
treatment with acids. An unidentified dinoflagellate 
was also common in the sediments. 

In the clay of Zone A, algal spores were 
moderately represented. Botryococcus, cf Mougetia 
viridis and cf M elegantula were common in the 
shallower margins of the lake. In Zone B, all of the 
algal types increased at 70 cm depth, where Zygnema 
and Debarya were at their most abundant. Very few 
algal spores were found at 60 cm, and Debarya and 
Spirogyra were not found in this zone above 60 cm. 
The remaining Zygnemataceae and the unknown 
dinoflagellate increased in abundance at 40 cm. 
Botryococcus remained abundant throughout Zone 
B with high amounts at 50 cm. Cyanobacteria were 
abundant at 70 cm and 40 cm. 

In Zone C, Zygnemataceae spores were low at 30 
cm, in the clay, increasing to high levels at 20 cm, with 
the exception of Debarya. Zygnema was particularly 
high in abundance at 20 cm, and high amounts of this 
alga were maintained into Zone D. Botryococcus was 
present in very high amounts at 30 cm and amounts 
remained relatively high to the top of the core. The 
abundance of Cyanobacteria was moderate at 30 cm, 
increasing to a peak at 20 cm and remaining high to 
the top of the of the core. Oospores of Characeae were 
conmion in Zone D. 

History of the Vegetation 

During the late glacial period, the vegetation 
was probably a shrubland with a diversity of species. 
When clay was being deposited and the lagoon was 
a lake, the sedgeland would have been confined to a 
fiinge around the lake. When the water depth became 
shallow enough, the sedgelands encroached on the 
surface of the lake. Peat was forming at 23-22 k cal. 
yr, prior to the height of the glacial period, hence the 
lake had become shallow enough for a sedgeland at 
this time (Fig. 5). 

During the period preceding the Holocene, the 
sedgeland flourished and it was probably comparable 
with the sedgeland there today. Myrtaceae was 



also abundant, and it may have been similar to the 
Melaleuca and Leptospermum swamp forest seen 
there today. Herbs were well represented also. In the 
early Holocene, the lagoon reverted to a lake and the 
sedgelands were once again restricted in extent, but 
they returned later in the Holocene. Casuarinaceae 
and Myrtaceae were relatively the most abundant and 
they were probably trees. 

Algal spores are present through the profile and 
are abundant at some levels. Zygnemetaceae are found 
in oligotrophic waters, and shallow, stagnant pools 
of mesotrophic waters, less than half a metre deep, 
induce spore formation in spring (Van Geel 1978; van 
Geel and Grenfell 1996). Of the Characeae, Chara is 
typically found in hard waters, and secretes lime, but 
Nitella grows in soft water (Pentecost 1984). These 
water conditions could occur, even if for only a short 
time, given the right combination of fi-esh water input 
and evaporation. 

At this level of identification of the pollen, there 
appears relatively little change in the taxa present, but 
where a more precise identification is possible, e.g. 
with some Myrtaceae grains, changes at the species 
level were detected. Some species of Eucalyptus 
have been present the whole time. The major dryland 
vegetation type, viz. sclerophyllous shrublands/ 
woodlands/ forests, with both Casuarinaceae and 
Myrtaceae prominent, seem to have occupied the site 
for the whole of the time recorded here. On these poor 
nutrient soils, substantial grasslands are unlikely, even 
with climatic change. 

Climatic Implications 

A