Proceedings of the Linnean Society of New South Wales

Sy hate
ahi g.

e

PAPE Be
pon eee

ae (atiiad
Petty Wits dg
tenia flied

i i
rie Lhe

Hand
feted ty

on ats
Vad

ey

Ros Pepe ats

in

mee a, bait isi soho
e y aM Ti f, Buk ix ag ; | ; *%

MEM AA
pry es
Faby,
i
j
\
1) 7 i
oe q
1 = a
i
i
A
;
»
¥
vf
7
‘
rd
)
w
-
|
4
iy 4
‘ A
if A * At a |
' he + ‘ 4 uf J j
ue ib \ ‘ hy ee
Tae Hi

ee at
Ws 4 lai chante

=
% Po
5
i
)
4 ‘
Pair t
: 4
i 4° 4
Pe Seal
en td 4
‘ 7 4
re A Y

i
AL

PROCEEDINGS

INNEAN
OCIETY

of
NEW SOUTH WALES

VOLUME 127

INCLUDING: A SPECIAL SECTION CONTAINING PAPERS ON THE
BIOLOGY AND ECOLOGY OF GIBRALTAR RANGE NATIONAL PARK.

: ee

aA Ay

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 $480.00 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.P. Barkas, I.G. Percival

Treasurer: 1.G. Percival
Secretary: J-C. Herremans

Council: A.E.J. Andrews, M.L. Augee, J.P. Barkas, M.R. Gray, J-C. Herremans, M.A. Humphrey,
D. Keith, R.J. King, H.A. Martin, P.M. 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: P.R. Reeves

Auditors: Phil Williams Carbonara

The postal address of the Society is: P.O. 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 publications and
activities, see the Society's homepage

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

VOLUME 127
February 2006

5 i. i ko orem af Watutal huehary ta al) ne et

fa a, nveth niente aench GEM ad

wy ' WALES:

Low eet waged |

>, aut, wily ets iat, Oporto hc Danes Ohi es

mar i } A. atEsg ecg fae Jevec ee We er? wW
a aaet Lah SOPs = ennlty, Marrnc Vine) hte 88
: = Pt nem Ma toe Totheeekis foriirecas Cae
si P= ; F hy ‘ 7 (eS, Gti eae ae ie _iowtan Nei Sanat
Pea;

roto y et aire Waiversity vt vyne ye 4

timber the Pree diace V0 hold etd clei

AUC Tete maemiiny the iene Ser

Mi ache 34 Niet iia) Dechon deliwered Sy '@
MM on metas Zech ommend Sehetee

» Firm, beets Gee Se ath ly cantilie “te
Corres Spa a epee ab zulolW rim0@-sesid.torySiaeivieeansih rast St
syoquiriod e‘yisiood sill soa ,zontryi2e

mint sta Ae eden ae HOS Yee Wie ©

:
ni ill dvsceadiane ita
‘ f rane Virani ol alii

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 corner of the
state of New South Wales (29°31’ 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 Linné (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 anniversary in
the history of natural science. The Linnean Society of NSW is planning 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 Linnean Taxonomy, but also with the recent advances in systematics that have built upon the ground-
breaking work of Linnaeus. That issue of the journal will also be open to papers dealing either with Linnaeus
himself or systematics at any level whether they have been part of the symposium or not, so start now!

M.L. Augee
Editor

ce Ce Te. eee eT
JAUNOTION

yooloid adt filiw gailesh enqsq to bozoqmes monoos isiooqe 8 ahitinoo smuiov- 2idT
edt lo remtoo SM ott ci ar ogne wetletio ofT teh lenohelt sgaat silat) dt to a
ot 8h sotiai bos esen! asl to emwal oh asowted Orset *TEesy aoleW dune wor
=i eukeg joqad notauhentill lt al navig om elisiod vogasdt gaibredtt word) or to opbs
igucoweyM . ioisT bas oe! 1 sate? yd balibe-due need sved mga, setlendif) | aT
foliose off grruesia otal tog ovad ysl) ahow, ort eoinioorggs -WRVE to ‘Yioivoe neon
seuroo to at 8H (8001 -TOT 1) dail mov hel to did ol} to aevittas dborbaud soul ott) ai YOOS
avo) — bsrieiidug ad doicw sole orteee aid To norerey boamist ont yet zynbs oeartt mwond ont
Mi yueTeVitis MAH aie) wa OF sdolg ort 2eo1s8 elnove avon: xd iduob or Tie
gnizne brs srwizoqieye tsiopge a yrinnalg 4 Wet to yratooe nwoentit oT engine Te
eussnanid to vrroteit oft dtiw ¥fno ton bsemoncs od‘ liw diet damaumot elt Yo omee Isisoqe a
-brvorg ot noqu ind over wedt gsitenoleye ni esonevbe tasoot Si ihiw oats wd anion
yusenni.] riiw tortie gatlasb eraqsg OF ASO od ovle iw leno; oct Io suael ie? ssanin .
twort hia oe Jon 20 muiwonenve ait Yo heq ased. oved yadi asitedw lowell, sale te

il

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

PETER J. CLARKE! AND PETER J. MYERSCOUGHZ

"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 127, 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 of 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 Vernes
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 PAPERS

short fire frequencies 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 number 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. ligustrina),
New England Blackbutt (E. andrewsii) and Coast
Blackbutt (E. pilularis). Others with local occurrences
are Needle-leaved Stringybark (E. planchoniana),
Narrow-leaved Peppermint (E. radiata) 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 (EZ. approximans) in some areas,
and stunted trees of the Red Mahogany (E. notabilis).
Along watercourses in shallow valleys narrow bands
of Mountain Gum (E. dalrympleana) 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. campanulata),
Messmate (E. obliqua) 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
and Lepyrodia, 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. /aevopinea), Gum-topped Peppermint
(E. campanulata) and Brush Box (Tristania conferta).
Some of these trees are over 160 ft high. The
understorey in these forests contains wattles, treeferns,
some “rainforest” shrubs and vines, and some tall
flowering shrubs such as Nightshade (Solanum
cinereum), Mint-Bush (Prostanthera), Correa and
Tall Everlasting (Helichrysum rufescens).

Near the bottom of the range, the rainfall is much
lower, and OPEN-FPREST ofa 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, Sassafras, 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 turnoff).
Large epiphytes such as Birds-nest Fern (Asplenium
nidus), Elkhorns (Platycerium), 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

jis im ITCRpeI ies cad Teh (he PoyerrnibAg

oy of cep ehrobs. In short. 4 t | nat
t [icrej ee te r in

, ‘ wes Art st f (

a istics pentegviel Serine cote
ose) pes bou2 silat stall Ae eset
ft ewosid si eh ale ose Abt
vegviahon with sve resent sti gi
ses ori Whd: Danis RG 4a? Te weaOM Og
B. Wi Ayid Nkok lpbdr2 Ved emantept Ye
“OTS qd \vsitore gi Oia aay ya Aer seein bem
core @Y2d907-VEIO LAAT somos bud? out o7)
ithe New os Pan OKRA ererh Sage opnal pst
UPRMMO ae ee on.
agg atv po bieat panier mrt |
Gus Ausitaian slo ie pater pt typ

arty. 40 years Wadia sinter anneal
lyed in Hi Seo! thins Pte re rad >

bariy acaba Rt er hee pee renee
He OUTS evap aa Pee reeMte SA Get
Sool Aaron iat 1 Sat ols migne cram Ga
fh’ heat HUW “niente teh ToT 2R8 IA bet
ixpared Hl nbiaTe enligartai! wit pidtoal = G
if cits, Amotie the ighepepA hives
pucy | pita sah ‘Seene® thei ba in
wea ty lapis aeopey ley Re cohnaregy Lodole
rh hd ‘DAERO BTEC AOE We HSS Tol DEE
cerita ate i eae pee

dooce DUS TO? abe. Ree Nagitog
. tr eats ove bin SAT a
estas Wook teri Wide 18 Bla’ iarapumiclnith
cotta Sabie D Woo? SP Mii Mbidle ste
Woled? Viavifort (GAP aioe indabeabqsaie sion
TeAHOAMEAT START Ba A bem
bhewhGSY wolggetinO! WeRasee2 .boowrboso dl
SvodW tivet ae Rest veuhMe® ‘bee! jaten od Dalkey 4
Soririies Winsswlepe BANG dQ! mcrae SHE
athennsns be Wort amen otete! larch aM bA
WMSIGEH) Mio tarsebie Heitiote Ggtiee
etterts Be olen Re UI hoa
uth Hiieqaorigia \eonoippriod Dan Henne sie dh
weatls ranienbsni Tasicgronte say yt

pewry f ® fuul-re iifall grea
hl mate plants Gy thee vidincavieicu
fough-)caved herbs, chang restisds sca ae K
; cx ath! Lenyrodia, and ange, to fied sodges-ci0h as Bu
ane plate willie Ou (tags. (Ceri thane h- Spe sedge (Sehe /
meer red. a oe Viiely mad Raz irate (Lepulosperate). [Bowart at Ruz
sedges, the fal, eneltow, heaves and stems have!
00 wf iarie sages which exp uate deepens.) A Late teat

don ptcd ivy iin watercowraes 37 teesWatnpes ape dcveesl etal Wheat

wa rhe mites iw wh Very which flower well in iste earned

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

Ropney 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, Ferntree 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 127, 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 A. 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 (Quinn et al. 1995),
two specimens of an Acacia housed in the N.C.W.
Beadle Herbarium (NE) that had been determined
variously as Acacia ruppii 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.Cunn. ex
Benth., A. gordonii (Tindale) Pedley, A. ruppii, 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).
Field trips were undertaken in Gibraltar Range National
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.
(1990) [http://207.156.243.8/emu/ih/index.php].

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

ACACIA BEADLEANA, ANEW AND 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,
CANB, HO, K, MEL, MO, NE, PERTH, PRE). Figs
1-2.

Description: Single to multi-stemmed, lignotuberous,
erect to spreading evergreen shrub, 0.42.5 m high.
Stems woody, terete, roughened by phyllode scars.
Branchlets terete with persistent, densely pilose
indumentum; trichomes simple, hyaline appearing
silver to white, antrorse to retrorse. 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, straight 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, 32-46
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,
straight. Seeds of transverse 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. Williams
& 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 restricted 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 A. conferta (Jones 1997; Maslin 2001).

Figure 1 (right). Isotype of 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

N.C.W. Beadic Herhariom (NE)
University of New England

ISOTYPE
Reacca Deadleana OH ones + 4. 1. B.uh|

Dets J 3b ht 29 Jus 2005

N.C.W. Beadle Herbarium (NE)
The University of New England
Armidale NSW 2351 Australia
Notification of change of determination would be appreciated by NE
NE 85360
Fabaceae subfam. Mimosoideae

Acacia sp. (Gibraltar Range)

Australia. New South Wales: Northern Tablelands:

Moderate rocky slope, mid-slope, N aspect. Grey, skeletal
sandy loam on granite between boulders and in rock crevices.
Patchy Eucalyptus willilamsiana layered open woodland with
Leptospermum trinervium, Callitris monticola, Allecasuarina
rigida, Acacia sp. nov., A. baeverlenii, Boronia anethifolia,
Leucopogon neo-anglica, Mirbelia speciosa, Calyinx

tetrag sopogen petiolaris, Lepidosperma gunnii,

L. viseidum, Caustis flexuosa, Schoenus turbinatus,
Cenospermum burgessiorum, Trachymene incisa.

Cornmon at site, focalised (c. 120 plants seen). Shrubs
to 2 x 2 m. Flowers golden yellow.

Coll.: J.J. Bruhi 1548 28 Jan, 1996
Det.;

Rep(s) to: BRI, CANB, HO, K, MEL, MO, NSW, PERTH, PRE

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

100 mm

ACACIA BEADLEANA, ANEW 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 =
1mm; 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 Ecology: Plants of Acacia beadleana grow in
Noel C.W. Beadle (1914-1998), foundation Professor skeletal to deep sandy soils on granite in layered
of Botany at The University of New England, noted eucalypt woodland and heath. The type locality is
ecologist and taxonomist. 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 morphological features of Acacia beadleana, A. brunioides and A. gordonii
* Additional observations provided by P. Kodela (NSW)

Character/Taxon

Branchlet hair density

Phyllode base
Phyllode indumentum
Pulvinus indumentum
Petal indumentum

Sepal indumentum

Flowers per head

Flower colour

A. beadleana

Dense

Cuneate

Abaxial margin only
Usually present
Present

Present

32-46

Bright golden yellow

A. brunioides subsp.
brunioides

Absent, isolated or
sparse

Obtuse

Absent

Absent

Absent

Sparse or absent
21-26

Pale creamy yellow

A. gordonii

Dense

Cuneate to obtuse
Over whole phyllode
Present

Absent

Sparse or absent
(12—-)21—25(-34)*
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

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

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 fruit 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 A. 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.

ACACIA BEADLEANA, ANEW AND 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 (BRI)
and Peter Clarke (UNE) and Philip Kodela (NSW) for
personal communications; directors of herbaria BRI,
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.

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).

Holmgren, P.K., Holmgren, N.H. and Barnett, 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 T.D. 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.
(1995). 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

Radford, A.E., Dickison, W.C., Massey, J.R. and Bell,
C.R. (1974). Vascular Plant Systematics. (New
York, Harper & Row).

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. Cappy & 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
Wales 127, 11-18.

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 from 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 unburnt 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 within 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; fruit 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.

METHODS

Species distribution

Specimen data for Grevillea rhizomatosa
from the New England Herbarium and flora survey
data from 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 further 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

U2

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 100
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).

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 fruit-set over the 2000 and 2001 flowering season
(see Table 1 for sample sizes). Bags were placed
over developing fruits to reduce fruit-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-tetrazolrum 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 from
Dandahra Granite complex within Gibraltar Range and
Washpool National Parks.

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

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

com Innes

SEZ CG yi Highway
——— unsealed road

1 ~~ walking track
—“—— _ creek

2km

4 study populations
© other populations

Cascade

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 FE. 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

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

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 (MHut)

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

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.

Number of
plants
(% seedlings)

Latitude
longitude
altitude

Site

29° 28’ 03”S
152° 18’ 18"E
870 m ASL

Wash c. 225 (0)

29° 31’ 00”S
152° 21’ 39"E
910 m ASL

MHut c. 250 (0)

29° 31’ 42”S
152° 20’ 30”E
980 m ASL

Dand c. 250 (0)

29° 317 58”S
152° 20’ 27°E
960 m ASL

Swamp 165 (0)

29° 32° 37°S
Cas SP Ake Sy
830 m ASL

4] (0)

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 rufa, 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

14

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

10.19 (206)

13.35 (337)

% seed viability
(n = number of
seed treated)

100% (7)

0 (62) ‘

7.08 (367) 2

not quantified but none
observed from 2000-2005

100% (7)

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.

MEClimonts Swamp (Swamp)

The Swamp population is _ located
approximately 500 m down-slope from 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.

Murrumbooee 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

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 unburnt
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 unburnt individuals at Wash and in the unburnt
population of MHut.

Fecundity

Fruits were only detected in Wash, Dand
and Cas (Table 1, Figure 3a, 2b). Flowers have two
ovules, but fruits mainly contained one seed. Fruit
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

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 fruit in the fertile
populations. No seed was produced from tagged
flowers at MHut and no fruit 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 (Figures 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
fruit on any flowering plant. Seedlings were never

16

encountered, even after fire had burnt populations
containing fruit-bearing plants. These three fertile
populations (Wash, Dand and Cascades) are widely
separated and thus valuable for the conservation of
the species. Natural fruit-set was low (< 0.14 fruit
to flower ratio) but higher than that found 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 from 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.

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 pronounced vegetative
reproduction (Fischer and Stécklin 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, found 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 shown 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 fruit-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 2001 work was undertaken by HAR Caddy as
part of an Honours dissertation. The project was supported
by funds to C.L. Gross from NSW National Parks and
Wildlife Service, Glen Innes District and by the University

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

of New England. Peter Croft is thanked for suggesting
the project, arranging funding and for providing access
to flora and fire 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 Stécklin, J. (1997). Local extinctions of
plants in remnants of extensively used calcareous
grasslands 1950—85. Conservation Biology 11 (3),
(imi Bue

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 fruit: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). Telopea 5, 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.

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 pollination
ecology of Grevillea beadleana McGillivray
(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 mallee
eucalypt using RAPDs. Australian Journal of Botany
46,161-172.

Vesprini J.L. and Galetto L. (2000) The reproductive
biology of Jaborosa 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 from 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 unreliable (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 Wooller 2001, 2002, 2003). Self-com-
patibility has also been demonstrated in B. ilicifolia,
although fruit 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 BANKSIA 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

20

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 from 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
function. 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 from the final
model to increase the degrees of freedom for testing
the main effects. When the interaction was significant,

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 (%) 1997 SC 94 (16) 92 (25)
1998 SC 94 (35) 94 (36)
1999 SC 83 (57) 93 (45)
1999 WT 85 (20) 89 (44)
Number of follicles per cone 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)
Number of seeds per cone 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.

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

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 = ws/Wo,
where ws and Wo, 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

Pil

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 34.7 + 2.9 (16) 49.1+4.2 (16) 0.73 + 0.05
Seed mass (mg) 6.24 + 0.10 (320) 7.32 + 0.07 (20) 0.85 + 0.03
Seed germination (%) 97.2 (320) 97.8 (320) 0.99 + 0.01
Seedling survival (%) 79.1 (311) 84.0 (313) 0.95 + 0.04
Seedling mass (mg) 156.2 + 2.9 (246) 174.7 + 3.3 (263) 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 IB Hal) ley e—(0 074) Selfed
cones produced significantly fewer follicles than
open cones in 1997; differences in other years were
not significant (1997, F, ,. = 15.58, P < 0.001; 1998,
Bia O22 ee 0108 8591999 REFN 10 10 ve O58):
In addition, selfed cones produced significantly
fewer seeds than open cones in 1997 and 1998, but
NOt 99 (OO TLE, = oe 2U O00, 1998, cE Ne.
Sd! SH VW UsMy Tp SAU ig? a Uh8310))

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

Table 3. Results of two-way ANOVAs (F) for

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,

F, 44 < 0.26, P > 0.614). Numbers of follicles and

seeds per cone did not differ between treatments
or sites (Table 1, all ae Sy Oe Pein) 27/2).

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

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. +P < 0.08; ** P < 0.01; *** P < 0.001.

Trait Treatment Plant Interaction
df ForG df ForG df ForG

KK KK RK
Seed mass 1,15 23.31 15, 608 29.76 15, 608 5.98

* KK +
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

? KEK
Seedling survival 1 Si] 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 from 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 fruits and seeds were
produced following experimental self-pollination,
indicating that Banksia marginata plants at GRNP
are self-compatible. Studies of other banksias have
shown 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

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

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
beestimatedifweassume that levels ofself-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
from 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

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 were 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 B. 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.

24

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. 358, 991-1004.

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 fruit-set in four co-occurring Banksia species.
Journal of Ecology 77, 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 22, 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 1n 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.

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., Collins, 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 127, 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

JES)

SE FCORPRMASLIM Cid HOTKRUAMARGT

kets al oar atc thal wos pty

‘ote are (pe ore aigueat re rah)
commit: Cleese, peur’! be aces! i
iS =
7 a 2c deatiew 1! ® a iu
“a use . ey
9
4
a
4
r ba
os
tire i as =
hl od toe
Al
1) OLA ve
: st lap hous
4 11 serv
i clesh
’ Mt pop
! iv ire
+ OP of
i crt '
: i ry rin i<
. ' ’ ‘ i
1
“ i of
‘ 4 ~
Te
' Lh’
;
# j i] |
A
’ ’ -
1? - sj let
we } ri
_ ;
my
4 '
:
i
|
' ?
'
} ! ‘
af
Tv \ bj ry
W w «i bu
t} > = -
ac
oe]
7 (ae aa

Moi ik ihe

eenmn bond,

Tel Dy
Pp ‘ we ,

bait oe

ao

¥ vi one

- sien i)
saga ae a
eee hee
% aaa ara se P

Baltes ‘ me: ve
ae na ee SH

ba eel si ray an
outa i
Ae

~

nit pa gi Hod fy, ae ge Be
£08 a rpotoa’d is Besta so

«S99
‘Sundettg adds), ie ee ere i! “i lug
SRO PP RY es aiaipnd

% 2 aie a poppy Sips ty ft Hig van
x LE Mor ’
: am os $02 be? (ORs : ‘a %y * oe

raz fod. q bor Parts My —— ad conta

ee Rs 4 St re a La oly)
emmy ; iat 7 hee Toh

92 sho \ajngewe ee ath, SN ARR

-t} ia to. sles aT, AEROS) A; Lita Lk

pied, 19 a8 rat EAM
Be Fe 12 vtinot As mien ae c
ni Yaitdur those Bigs j i avlleg die hh flow
rete, OLA mM gil Blea Billog

wait bG HEAT aS {NPY Seta

Mie
ca é

wd Schult, 3.7. (1094),
ling depresaory § lective interference
orl the Oweshold selfing ste for pungin
wive Jothal mutations, Mvalarienr 48, 963-972
LOL (1979), Sule Feproddative factors Ay
he Unectiog of seifferttirention ti Manta American

Nuinratie WS. 6079

yy D094 1999) Sidf- and -crese-fernitization in p
1 Tie solection of voll-fertilieaion. deernational
Jr of Piawt Souencves 18, ITO-380

ante, FB Sclienuke. DB W
Lohwee

fiery linn

wai)

Fire History and Soil Gradients Generate Floristic Patterns
in Montane Sedgelands and Wet Heaths of Gibraltar Range
National Park

PAUL RICHARD WILLIAMS!” AND PETER JOHN CLARKE!

"BOTANY, SCHOOL OF ENVIRONM ENTAL SCIENCES AND NATURAL RESOURCES MANAGEM ENT, UNIVERSITY OF NEw ENGLAND,

ARMIDALE, NSW, 2351 (P CLARKE@UNE.EDU.AU).

SCHOOL OF TROPICAL BioloGy, JAMES Cook UNIVERSITY, AND QUEENSLAND Parks AND WILDLIFE SERVICE, PO Box

5597, TOWNSVIIE, QUEENSLAND 4810, AusiRALA (P AUL WILLIAM S@EP A.QID.GOV.AU).

WILLIAM S, P. AND CLARKE P.J. (2006). FIRE HISTORY AND SOIL GRADIENTS GENERATE flORISTIC P ATIERNS IN M ONTANE
SEDGELANDS AND WETHEATHS OF GIBRALTAR RANGE NATIONAL Park. Proceedings of the Linnean Society of
New South Wales 127, 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 free from
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 frequency. 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 15 years after fire. Fire
frequency 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 Range

an ideal location to assess the long-term effects of fire regimes in montane sedge-heaths.

Manuscript received 1May 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 free from
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 frequency.

28

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 frequency 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
unburnt since 1980 and were considered in this study
as long unburnt (i.e. 15 years since fire). Two sedge-
heaths were burnt in a planned burn in 1994 and
were considered regenerating communities, having
been burnt only half a year prior to this study. The
remaining two sedge-heaths had an intermediate age
since last fire. One was burnt in a wildfire in 1989, the
other in a planned burn in 1990. Therefore of the six
sedge-heaths surveyed, two were last burnt 15 years,
two 5-6 years, and two were burnt half a year prior
to the survey (Williams 1995). Following the 1995
survey all study sites were burnt 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 southern 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 channel. 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

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 from 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

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

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
channel 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 channels 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 (1 =
- 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

29

FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS

-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 declined 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 unburnt 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
unburnt 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 burnt sedge-heaths contained
atotal 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
unburnt 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 channel 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 forshrubspeciesinthe 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 P F ratio P F ratio P F ratio 12
Time-Since-Fire 2 141.7 felts 244.9 wih! 2.9 = 2.6 ns
Habitat 2 30.0 ae 27.0 ee 8.3 pa 65.9 aye
TSF x Habitat 4 2.3 ns 2.5, ns 3.7 ay 3.3 :
Residual 153

30

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

P. WILLIAMS AND P.J. CLARKE

| 0.5 Years
"| 5.0 Years
L] 15.0 Years

Canopy height (m)

=
az
=
8 a
6 2
4 R
2
0 C C
Channel Mid slope Outer edge Channel Midslope Outer edge
25 d)
20
15
10
5
4 4 — 5
Channel Midslope Outer edge Channel Midslope 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.

~ = Channel

30 x Mid slope
fe ® Outer edge
Ly
oe 25
om
Es
= 20
tay
es
Es
215
at
ow

10

5

O 5 10 15 20 25 30 26
Conductivity

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

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

31

Sys

yeugqey

C97 KEK 8 €7~ *RK 78 *KK 0 CET EK 90 #EK 9 LTI *EK 6 vrI C

2k 2k ok

FIRE HISTORY, SOIL GRADIENTS AND FLORISTIC PATTERNS

911J-29UIS-OUIL |

su 6 su 91 see 67I ek 1'9
onel 4

Te

2K KK

oer 4

d

Omen dt Osea ge Om — af Ole ay oe a

d

10}087

Ssouyoll ssouyors Ssouyort
SSOUYOL SSPID ssouyolt ApooA\ : ssouyor sorsadg
: prourureIn 7 Japoas 91e311qO Joinoidsoy to a

SSOUYOI GIo07

*poienbs-1yD wosiveg SUISN pazVUIIyso a[VIS ALY PUL UOTIUNY YUL-dO| & YI S9ANON.YS 10.119 UOSSIOg & 2ALY S[OpoUI
IV ‘sdnois Jeuonouny 10 yeyIquy puL eAY-9dUIS-9UNL) 10J S[opoul AvaUT] [e19UAS 10}9¥y OM) 1OJ s}[NSo.1 AreUMUING °7Z a[qQE |,

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 inrelation
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,

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

P. WILLIAMS AND P.J. CLARKE

a) Total b) Resprouters

28 28
24 Ee] 0.5 Years
5 5.0 Years
< 20 [_] 15.0 Years Pa
a IG
in 16 2
Fr a
a12 : ©
a
o 8 3
4 :
3, 3
(@) —————— a
Channel Midslope Outer edge Channel Mid slope Outer edge
14 c) Obligate seeders d) Woody species
12
© 10
e :
~
i? a
z 6 2
i 2
oe, =
2 =
2 =
0
Channel Midslope Outer edge
Channel Midslope Outer edge
14 e) Grasses and sedges 1.4 f) Forbs and ferns
12 12
ci 10 210
= In
In ol
= 8 ~~ 8
bi a
a 5
zo 6 a6
a Wi
“i va
‘4 ° 4
xi o
c= =
= 2 2
0 —————_———y 0
Channel Mid slope Outer edge 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

Table 3a. Summary results for linear models for fire frequency and habitat. All mod-
els have a Binomial error structures with a logistic link function for species data.

Factor df et Baeckea omissa ¥ a A Hibbertia rufa ee ai iin Ea ae
F IP 2 P x2 P 9 Ge IP 2 P 1 Ga P
Fire frequency 2 3.0 ns 2.0 ns 4.8 ns 6.1 x 30.3 pee OD
Habitat 1 te 8.2 fe 49 * - = : = :
Fire frequency x Habitat 2 5.2 % 2.0 ns 70.8 Fate: - - - © = =
Residual 30

Table 3b. Summary results for general linear models for fire frequency and habitat. All models
have a Binomial error structures with a logistic link function for species data. Species listed in
order of relative abundance.

anksia

Leptospermum ‘ Hibbertia Grevillea Hakea Epacris
Factor df i marginata NE: ees : en
arachnoides : riparia acanthifolia microcarpa obtusifolia
ere | Ne Se ee eS ES | a Sees eae
7 P Ge P Me P 7 P 0 a P 7 P
Fire frequency 2 8.5 as 1.1 ns 6.1 : 26.4 a 30.3 uae 24.6 oo
Residual 30
* P <0).05
** P <().01
**x* P <().001

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

34

P. WILLIAMS AND P.J. CLARKE

4 fires

|| 5 fires

Abundance

©
ate 2
a on
— =
} Ss = fo) iS
© = © = S
> rs) re) S
~ S © Q a
& x c & <
ro) 2) = = S)
PS = = S o
= > i Q
g ® 9 Ww
= jaa)
3S ©
lo)
a
c
S
5
Wy

Hibbertia rufa
Beackea omissa

Leptospermum greganum
Banksia marginata (seedlings)
Leptospermum arachnoides

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,

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

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-unburnt sedge-heaths
were halted in 2003 when all long-unburnt 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

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 frequency 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, P.A., Brown, L. and Morris, E.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 Bulli 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?

36

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, P.J. (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, P.J., Clarke, P.J. 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, P. 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 37, 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 of Sphagnum

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

P. WILLIAMS AND P.J. CLARKE

communities in NSW and the ACT and their Department of Botany, University of New England,

conservation management. Cunninghamia 7, 463- Armidale, New South Wales.

500. Williams P.R. and Clarke, P.J. (1997). Habitat segregation
Williams, P.R. (1995). Floristic Patterns Within and by serotinous shrubs in heaths: Post-fire emergence

Between Sedge-Heath of Gibraltar Range National and seedling survival. Australian Journal of Botany

Park, New South Wales. BSc Honours Thesis, 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

;

Allocasuarina littoralis Tree R
Amphipogon strictus Grass R
Aotus subglauca var. subglauca Shrub R
Aristida ramosa Grass R
Austrostipa pubescens Grass R
*Axonopus affinis Grass R
Baeckea omissa Shrub R
Baloskion fimbriatus 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 stricta Forb R
Dianella caerulea Graminoid R
Dillwynia phylicoides Shrub R
Drosera binata Forb R
Drosera spatulata Forb R
Empodisma minus Graminoid R
Entolasia stricta Grass R
Epacris microphylla vat. 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 af

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 Ss
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 Ss)
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
Xanthorrhoea johnsonii Graminoid R

Yuri I pet k

38 Proc. Linn. 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,

Armidale 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 B. 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 Charlesworth
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 (Lamont
et al. 1989; Mustart and Cowling 1993; Schiitz 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). 3

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 1s
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 B. 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

40

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

Banksia spinulosa Sm. var. neoanglica A.S.
George (B. 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 (1.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: 29929’ S, 152919’ E, 1050 m a.s.l.) and

Surveyors Creek (SC: 29932’ S, 152918’ 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 18 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 spanned 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

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

S. VIRGONA, G. VAUGHTON AND M. RAMSEY

ww) ae Any

PAA 1 AAR
0-20 21-40

41-60
Transect locations (m)

<+— Swamp —><— Woodland ————>

61-80

Figure 1. Vegetation profile of the swamp and woodland

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 mx 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 from 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 110 °C. Seeds were
deemed viable if they were intact and plump.

association and the placement of stratified belt-transects

used to examine the 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.

the swamp or woodland habitat (Fig. 1).

To assess habitat segregation of B. marginata and
B. spinulosa, we calculated an analysis of deviance
using a logit model with a binomial error term and
a logit link function. The three explanatory variables
were site (SC, WT), plant age (adults, seedlings at
18 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 from 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 3 x 2 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 3

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).

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

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 from each of 10 plants per
species from 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 log{9 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

41

HABITAT SEGREGATION OF BANKSIA SHRUBS

Table 1. Analysis of deviance (logit model) examining the dis-
tribution of Banksia marginata and B. spinulosa adults and
seedlings (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.

numerous than B. spinulosa plants
(F, 17 = 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:
F196 = 16.85, P< 0.001; seeds: Fy 96>
21.15, P<0.001), and we examined the
simple main effects of species at each

site. The number of cones produced

Source df A Deviance ; ae
by B. marginata was significantly
Site 1 20a <0.0001 greater than B. spinulosa at both sites
(WT: By 196 = 103.49, P < 0.001; SC:
Habitat 1 2505 < 0.0001 F, 195 = 19.08, P < 0.001; Table 3).
ee ) 11.33 0.003 For the number of seeds per cone, B.
marginata produced more seeds than
Age x habitat 2 12.34 0.002 B. spinulosa at WT (F, 76 = 24.25, P
< 0.001), but at SC seed Pigduction
Site x habitat 1 2.97 0.085

of both species was similar (Fy 75 =

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 singularly 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%).

For B. 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 7% 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

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
18-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: F, ,,,= 11.13,
P<0.001; seed area: F; ;;,= 15.06, P< 0. 001), and
we examined simple main effects of species at each
site. Seeds of B. spinulosa weighed at least 22% more
than B. marginata seeds at both sites (WT: F, 116 =

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 (S1
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.

swamp, whereas > 99% seedlings were

confined to the woodland. Population B. marginata B. spinulosa

Seed bank : Avs Sj vs 82 Avs Sj vs 82
For plant density (plants/m ), the site x sc G 17.54 10.07

species interaction was not significant (F116

= 0.48, P > 0.450), and it was pooled with P 0.0002 0.0065

the error term for the final analysis. Sites

did not differ (F, ,,= 0.07, P > 0.750), but WT G 23.27 0.15

B. marginata plants were about 10-fold more P < 0.0001 0.926

42

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

S. VIRGONA, G. VAUGHTON AND M. RAMSEY

SC adults

Proportion

SC seedlings 18 months WT seedlings 18 months

Proportion

WT seedlings 28 months

Proportion

a | 0 ical aie a
S$ gg S o 8 8 2 ©
aaa eanretY naka YS er
Intervals (m) 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 seedlings. 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 SC

B. marginata B. spinulosa
Planes me 0.9 + 0.2 0.1+0.1
Cones/plant 722 ES, O) 2 3e Ne
Seeds/cone 24.4 + 3.8 33.3 43.8
Seeds/plant 420+ 72 306 + 45
Total seedslc 386 + 53 AEE)

WT

B. marginata B. spinulosa

1.0+0.3 0.1+0.1
45.8+4.5 9.4+0.9
STRIE=SES 30.0 + 4.7
2642 + 220 282 + 47
Pylejoe= 27M yee 2

15.74, P< 0.001; SC: F, 116= 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: Fy 146 51.47, P < 0.001; SC: F, 116= 2.84, P<
0.01; Table 4).

For wing loading, the site x species interaction
was not significant (F, ,;¢= 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; };7= 2.04, P = 0.156). Wing loadings of B.
spinulosa seeds were significantly greater than the
wing loadings of B. marginata seeds, indicating that
B. spinulosa should disperse shorter distances than B.
marginata (Fi = 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; Schiitz et al. 2002).

In B. marginata, processes operating during
seedling establishment appear to mediate habitat
segregation. The wider distribution of B. 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).

F ; Seed mass Seed area Wing loading
Site Species 2 D
ini (cm_) (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. Linn. 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,
adult B. 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 18 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 from the woodland
is unlikely to account for the absence of seedlings in

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

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 B. 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.

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.
Oecologia 132, 582-591.

Clarke, P.J., Myerscough, P.J. 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, 762-774.

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.

46

(University of New South Wales Press, Sydney).

Harper, J.L. (1977). “Population biology of plants’.
(Academic Press: London).

Keith, D.A. and Myerscough, PJ. (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, PK. (1998). Seed and seedling
biology of the woody-fruited 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 frequency 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, P.J. and Cowling, R.M. (1993). The role of
regeneration stages in the distribution of edaphically
restricted Fynbos Proteaceae. Ecology 74, 1490-1499.

Myerscough, P.J., Clarke, P.J. and Skelton, N.J. (1995).
Plant coexistence in coastal heaths: Floristic patterns
and species attributes. Australian Journal of Ecology
20, 482-493.

Myerscough, P.J., Clarke, P.J. and Skelton, N.J. (1996).
Plant coexistence in coastal heaths: Habitat
segregation in the post-fire environment. Australian
Journal of Ecology 21, 47-54.

Quinn, G.R. and Keough, M.J. (2002). ‘Experimental
design and data analysis for biologists’. (Cambridge
University Press, Cambridge).

Schiitz, W., Milberg, P. 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, 211-224.

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

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

S. VIRGONA, G. VAUGHTON AND M. RAMSEY

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

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, 19-26.

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

Williams, PR. and Clarke, P.J. (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

ADAMOWLEDOEMEN)

NG! & MM. i aroptell acat ? C' lars (ov corned an the
te | wt few S, Care Prov ieny chitiatical
SRY <p? & revi i by ONE end 4
botane MCW Benall ~
:
ri
} bey of Sve MY
‘ ‘ é 1s Protcéactac to
7~ 199%
4 y \ ecru “rd ur

4
=
se

cm stratigy

Ww Sosuerv 46, 7*-10)
agreny P.M.) ican, A RC amd Cferwer, RK
iy \ eos Mi Dic Mew ian Of
aera. Aurion nara of Beolegy
rT
r + al nig VAAe / M
uh) 1 ct ab rcal ive
ppt ) 6 MOUINSTICIOR
Jui } CYL Lote AING my 4
§ M ivean hehytat: ere! Fealags
*) d
ue. I FOO). Habitat tomilarity and fire reapanec
! : from 'n ecleroph yl arcliyprddano
132 ‘1
mwoueh, Pl aad Shehon Nuk 1) 996)
wateuence i mio) heaths: Retwoen~an¢
bital aflects of wpenticn, disturbance
i ryt a the post-lire covirongienl, Anvrolian
f w 2t
tJ mon. BD zd Marcus, R (7SUG) *
wed ' | oft Aire repiine
highly <crotinodas alin hainhkstia hooker iaivi
j » Ka ;
ta iy ril-effat i HK) MOURA
’ : ‘ ‘ ny veces ‘a,
Ciyverdvel? LL. Pargihi. “4 Laxtyectat, ERS sored Shel,
y z in « } TiVWCOP a thrae
a i , Me akiee 4 tener pris 4 abent m fore
one ehrubtaids fevraal al Prodew 90, 762.7 8
1 Ry Matioké, &.M. snd Gumer,
i U ! ag-toven chu Ow I WIR! and
f ut 1 Aardksio and Atari over wry
‘ i) swan Ucoaeial Pinin, Journal of te
hisad Secdety of Mesiern Avetralig 84,63-49
a » PaAlsi7.. Th tau lonance of specice retnaat
t oe uittiiecs WMUFLAACE OF Une
Risleoical Repiew $2, 107-145
K., Brajletnd’. BM. ond Alloway, W.(TOORD Pat
Shares] HO FHC en re~al sty! pllerrevat on
; is bee { olin owned! Of omen ,
4
1 (1241) (.om b ice vy Seth Wiikes ve IT
AB
oe
>)

Fetes +h of S Dewan 6, ts ans, oe:
Lennon, BED med Wier, 12, (200). Pepe an j
qpebil antan sates always low among cpecieg that
resprout @ftar fine, ‘ant why? Bvolnatomery | is

17 aoe 907
Morvan, TRA, Cir: OL, Pangetly, SM, oe
Neato, BLS, Thana, 2282. ca Andre
oe Bifexts of fie pcm ee

“ane

ara rower: Inter~fine uniaregh aah 1s
Aurercdtan Journal of Bcobogy 38, sha
Mostar, PJ, and Cowling, RM. (1993) bbe te:
reps stages in the discibonion. dap
re riche Fynbte Proteagene, Eevloge 74,14
Myorsconih, PJ. Clare, BS and Seen, MS, ee
Pant coednence tn count elias Hunan
tan! peches witrsulew Asdetrlltn Josireet
29, 482-493; =
Myerscough: Ps, Gerke, WJ; and Shdten 8, 419
Plant oeerishence if coaxtel haathe, Habinat
sxereyagon.in the pom-fire_ environment.
Jamal af Krolpoy 21, 41-54~
Quan, OR and Keough, MJ (2002), * Exper
deviyn and dats enalysie far biologists’ iC
University Press, € ainbridye),
Schtitz, WS Mithetp, Pad Laniear BB: ony iii
Gerivination fequircienty and secdling Te
oe witer avadebiiity and sell supe in lwurs
spices, aioe Dwoologing 23, LanI,
Rheringhem, Tw Huater, + T (anOas ‘i
floristics.ot Gibralte Range, Nutipaoal
» Natlonal Parks god Wikdbite Soren, Ceti:
SOR, vt C. arotm, Tf and Anilerson, pS. 4(tS7
Srecfies iw the Caney of omstel fucver hh | in New Su
Wiles 1, Vegeta strectare, Procendepe atthe” ie
lees Nader of Nee Siva ' Tater OT, aepcr "
Stvertowey, | eel Churhesworth, 0, (2001) thitadactf
jo wet pegualation Preah tly econ (ae
‘ Somewre, tyafowih). ;
Lae pari G ead Cashew, 5M 98%), Eeaces fie
whats vhonticn li Mankria spinulece (Pieter
pe aye vel devapal of the Linneouts Sie ete 36, 34
Viewplican, aod Ritrrry, M4. 1190 Sores. wai
somaoqurtoacs if eed mune Warkieioe lis Buarkeeia
seers atube { Pretemieae) Warne wf Booliigy Me. mess?
sn
Venaghtie ©, and Wansory: va aren giheadbilag : —

Betwngn Gat Holts, aewd nutrients, anal

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 (pclarke1@une.edu.au),

“current address: Department for Environment and Heritage, PO Box 822, Clare, South Australia 5453.

Knox, K.J.E. and Clarke, P.J. (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 127, 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 frequency influences post-
fire mortality and if survival was related to the size of plant; 3) fire frequency 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 frequency 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 1May 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 from 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 ofa 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
burns 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 frequencies.

Evidence of an effect of fire frequency would
show that more frequently burnt sites had more dead
plants and surviving plants with reduced growth and
reproduction. If, however, these sites had smaller
plants, then these may appear to show reduced 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 frequency influences the resprouting ability
of plants 3) fire frequency 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.

50

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
1964 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 from 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
fruits. In addition, the post-fire flowering of the
grasstree Xanthorrhoea johnsonii was also recorded
along each transect. For 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 from 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 frequency, with lignotuber size as the covariate.
We also used ANCOVA in Xanthorrhoea to test if the
inflorescence length was reduced by fire frequency,
with the caudex size as a covariate. Plots of residuals

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 frequency 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 a) Banksia 25
45
100 A
35
eg 30
60 25
20
40 15
10
20 :
@) )
0 10 20 30 40 50 60 70 80 90 Om 20) 3040) 50) G0) 70 480 90
Girth (cm) Girth (cm)
8 90 b) Hakea F 60
” = A
3 e 2 50
2) 70 = °
= °
= oc + 30
os 40 2
is) 30 * 20
- 5
eo & 10
10 2
) 3 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
Girth (cm) Girth (cm)
Petrophile
60 Gs 35
50 30 c
25
40
20
30 A
20 16
10 5
0 0
OmeniOen 2ONTS0 wn40rauSOx s60 70 O pilOR 20m SO 40) 550K GO ni a70

Girth

Girth (cm)

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

>) Il

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 inflorescence (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

Factors B. spinulosa H. laevipes P. canescens X. johnsonii
Ht Shoot Ht Shoot Ht Shoot Inf.
Fire frequency NS NS oe NS NS NS NS
Size (covariate) eR Per RK Ee ER oR NS
Fire fg. x size NS NS NS NS NS s Ns

b) Hakea
24 [| 2 Fires

a) Banksia

20 4 Fires

5 Fires
= 7 16
= £ 12

— —
fa) Oo 3
4
0
Dead Alive Dead
34 c) Petrophile 30000 d) Xanthorrhoea
eSOOO fs 2 Fires

ha :

c 4 Fires
c= oa 20000 5 Fires
nea) E
= = 15000
fe $
= = 10000

72

=

7

Oo 5000

0 eee eee SS EE QE | 0)

Dead Alive
Not flowering Flowering

Figure 2. Mean (+ se) basal girth of lignotubers in each of three fire frequencies in Gibraltar Range National
Park fora) Banksiaspinulosa,b) Hakealaevipes,andc) Petrophilecanescens.Meanvolume ofthecaudex(+se)
for Xanthorrhoeajohnsoniifor each fire frequency wheresmaller plants flowerin sites with less frequentfires.

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 likelihood ratio tests. *** indicates
P<0.05, ** indicates P<0.01, and * indicates P<0.001

Species ee Fire frequency Neneh areca
frequency volume
2 4 5
B. spinulosa Dead 10 24
Alive 96 108 NOE 16.6***
H. laevipes Dead 4 3
Alive 116 115 116 0.5 NS Gi.
P. canescens Dead 9 4
Alive 120 121 117 0.8 NS Va
X. johnsonii Not flowering 103 110
Flowering 17 8 8.4* 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?
= (il)

Fire frequency did not reduce the height and
number of shoots resprouting when basal girth was
used as a covariate (Table 1). Fire frequency, however,
appeared to increase the height of Hakea laevipes
which is not consistent with the hypothesis that fire
frequency 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 frequency 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 frequency 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 frequency and size
on post-fire survival using logistic regression. All

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

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

Finally, we examined whether fire frequency
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
frequency (Fig. 2).

DISCUSSION

In this study we examined the influence of storage
organ size and fire frequency on post-fire mortality,
resprouting vigour and flowering. Generally, we
found that larger storage organ size was related to: (1)
greater post-fire survival, (11) more resprouting shoots
(iii) faster growing resprouting shoots, and (iv) the
presence, and size of inflorescences for Xanthorrhoea
johnsonii. Fire frequency influenced the post-fire
survival of Banksia spinulosa and also influenced the
presence of inflorescences of Xanthorrhoea johnsonii
following fire (Table 2).

53)

RESPONSE OF RESPROUTING 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.

Are storage organ size

Does increased fire

Does increased

; Does increased fire
storage organ size

Specie and post-fire growth frequency affect frequency affect
Pao P er q y affect post-fire 4 Me
response correlated? storage organ? : post-fire survival?
survival?
Banksia spinulosa Yes +ive Yes -ive Yes tive Yes -ive
_Hakea laevipes Yes +ive No evidence Yes tive No evidence
Petrophile canescens Yes +ive No evidence Yes + ive No evidence

Does increased

Does storage organ Does fire frequency ‘ Does fire frequency
: ; storage organ size
size affect presence of affect size of affect presence of
y : affect post-fire :
inflorescence? inflorescence? : inflorescence?
survival?
SOT NOEE Yes +ive No evidence No evidence Yes -ive
johnsonii
Post-fire survival for the species studied resprouting shoots. Although this pattern was not

ranged from 99% for Xanthorrhoea johnsonii to
82% for Banksia spinulosa, although we may have
underestimated mortality if small dead plants were
overlooked. For 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
fruits. 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 fruits 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

54

unexpected, we had hoped to be able to determine
whether carbohydrate storage or number 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 found 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 frequency 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 frequently 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

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, Telopeaspeciosissima) 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. Cary 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.

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

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. Jan 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 of the 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.
Oikos 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 72, 7-16.

Bowen, B.J. and Pate, J.S. (2004). Effect of season of
burn 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, P.J. (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.

Cary, 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, P.J. 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 48, 261-269.

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, P.J. (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 P.W. and Ward, D. (2000).
Plant size and season of burn affect flowering and
fruiting 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, 7DAMIEN 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., Hofmeyer, 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 burning. Acacia barringtonensis Vindale, Grevillea rhizomatosa P.M.Olde & N.R.Marriot,
Persoonia rufa 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 from
buds at the base of stems and from rhizomes but the resprouts were heavily browsed by insects and Swamp
Wallabies (Wallabia bicolor Desmarest). In Grevillea rhizomatosa, individuals survived and resprouted
from underground rhizomes and no seedlings were found after fires. After fire in Persoonia rufa, all
scorched plants died but seedling recruitment occurred from 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 burning 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 burning 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
burns in NSW National Parks in 2003-04 than in 2002-
03. Hazard-reduction burning may be initiated when
fuel begins to accumulate beyond specified thresholds
(Gill et al. 1987; Morrison et al. 1996; Morrison and
Renwick 2000; Fernandes and Botelho 2003). On
the Northern Tablelands and similar areas, fuel may
be kept at or below these thresholds by burning 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

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 Zelopea 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.

METHODS

Study Area

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

58

Study Area

The mean annual rainfall at Mulligan’s
Hut is 2100 mm. The study area has a
mean annual temperature of 13°C on the
plateau with a mean annual 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 Gibraltar Range National Park
where a small hazard-reduction burn
was scheduled in 1999. Mulligan’s Hut
camping area is towards the centre of the
Gibraltar Plateau, at an altitude of 900
m. This burn was planned to help protect
visitors and facilities in the camping area
from wildfire in the park.

The open forest community at this
locality is dominated by 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 burn in 1999,
NSW National Parks & Wildlife Service fire records
indicated two large wildfires had burnt the Mulligan’s
Hut area in 1964 and 1988 with fire-history maps
indicating the study site was burnt. Another fire
occurred after the project was initiated in 2002 and
the subject populations were burnt during back-
burning operations.

Target species

Acacia barringtonensis is an erect shrub endemic
to high altitude areas of northern 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).

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.

q o Lea 1 3 5 7 34
Species Number of
P month months months months months plants
cacia
i y 0-50 0% 100% 100% 100% 0% 2
barringtonensis
51-75 0% 100% 100% 100% 0% 1
76-100 0% 50% 57% 54% 7% 28
Grevillea
‘ 0-50 0% 100% 100% 100% 100% 2
rhizomatosa
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 Jelopea
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

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

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 burn. Measurements were taken of all tagged
plants before the experimental burn 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 burnt
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 burnt. 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

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 further 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 unburnt 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-burn. 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 burn. The sole means of survival
was by resprouting from the base/lignotuber. As a
consequence of hazard-reduction burning, mean
number of stems per 7elopea 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.

60

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 burn was resprouting. The
delay may have been in part due to increased grazing
pressure that immediately 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 unburnt 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 unburnt areas.

Noseedlings of Grevillearhizomatosawere 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 Knox and Clarke this volume).
The smaller Grevillea rhizomatosa plants in this
study may have depleted smaller quantities of starch
reserves. Following the second burn three years later
about 65% of the original survivors of the hazard-
reduction burn 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 burns. 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 burning. 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

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

P. CROFT, D. HOFMEYER AND T. HUNTER

fire increasing the number of stems per plant post-
burn. Of the four species monitored, Telopea aspera
was the first to show signs of recovery after fire but no
seedling 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, frequent
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 burn, 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 burns may have minimal
effects on the number of wildfire events particularly
in fire-prone vegetation (Turner 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
burns are planned, 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 burning can reduce the intensity of
a wildfire for several years (Raison et al. 1986) it may
not prevent the area from re-burning during a wildfire,
especially in severe fire weather (Turner et al. 2003;
Moritz et al. 2004). Additionally, hazard reduced
ground is often chosen as an area from which back
burns are planned during wildfires control operations
because of lower fuel levels. If hazard-reduction
burning 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.

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

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 burn in 1999. The input
from the editors and referees was appreciated.

REFERENCES

Andrew, N., Rodgerson, L. and York, A. (2000). Frequent
fuel-reduction burning: 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 fire 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.P. 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). ‘i

Clark, S.S. (1988). Effects of hazard-reduction burning
on populations of understorey plant species on
Hawkesbury sandstone. Australian Journal of
Ecology 13, 473-484.

Clarke, P.J. 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, P.H. (1993). Geographic and
ontonogenic variation in morphology of Australian
Waratahs (Ze/opea: Proteaceae). Systematic Biology
42, 49-76.

Department of Environment and Conservation (2004).
Annual Report. Department of Environment and
Conservation (NSW). Sydney.

Fernandes, P.M. and Botelho, H.S. (2003). A review
of prescribed burning effectiveness in fire hazard

61

FIRE RESPONSES IN FOUR RARE PLANT SPECIES

reduction. International Journal 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 fuel
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 of Monotaxis
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 Schaffner,
A.A. (2004). Testing the basic assumption of
shrubland fire management: how important is fuel
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. \n ‘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, P.V. and Khanna, P.K. (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.P., Moore, D.M. and Andrews, S.P. (1992).
Proposed Forestry Operations in the Glen Innes
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.

62

Turner, M.G., Romme, W.H. and Tinker, D.B. (2003).
Surprises and lessons from 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 P.J (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.

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 (pclarkel@une.edu.au).

Campbell, M.L. and Clarke, P.J. (2006). Response of montane wet sclerophyll forest understorey species
to fire: evidence from high and low intensity fires. Proceedings of the Linnean Society of New South
Wales 127, 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 sclerophyl! 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.

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

INTRODUCTION

Wet sclerophyll forests typically form an interface
between two broad vegetation types, rainforest and
dry sclerophyll forest. These tall eucalypt forests have
been described as a stage in long-term succession as
their structure and composition approaches that of
rainforest in the prolonged absence of disturbance.
Hence, changes in community properties have been
closely linked to the passage of fire and length of
fire-free periods (Ashton and Attiwill 1994). Wet
sclerophyll forests are not highly flammable most of the
time, and as time-since-fire increases the probability
of a subsequent fire is reduced as mesophyllous taxa
become more dominant in the standing vegetation
(Unwin 1989; Adam 1992; Harrington and Sandersen
1994). Generally, one intense crown fire every 100-
200 years is thought to be sufficient to maintain
the sclerophyllous component and restrict the more
mesophyllous taxa of the wet sclerophyll ecotone
(Gilbert 1959; Chesterfield et al. 1991).

On the New England Tablelands many areas of

wet sclerophyll forest have been used for hardwood
timber production and cattle grazing. These forests
have consequently been exposed to a regime of
frequent, low-intensity fires that are often associated
with grazing (stimulation of green pick). A high fire-
frequency is thought to reduce diversity of woody
species in the forest understorey by removing
mesophyllous taxa and promoting growth of fire-
tolerant grasses and forbs (Binns 1991; Henderson
and Keith 2002). Currently there are few data on the
demographic links between disturbance frequency
and changes in understorey composition in the wet
sclerophyll ecotone.

The classification of plant species into functional
groups based on fire-response traits can be useful in
preliminary modelling of how vegetation will change
with more frequent or less frequent fires (Whelan
1995; Bond and van Wilgen 1996). Baseline data on
the response of plant populations to crown fire are
being sought to contrast the effects of seed-based
recruitment (persistence of populations) with those of
resprouting (persistence of individuals) (Clarke and
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
burn at Mummel Gulf National Park the previous
year. We addressed the following questions: 1) Do
fire-response traits 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-burn, initiated by the NPWS to contain a

64

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 recruits
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, 1.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

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

M.L. CAMPBELL AND P.J. CLARKE

Table 1. Summary table for

Attribute WPNP
Growth form
Shrub (<3 m) 1G
Small Tree (3-10 m) 13
Tree (>10m) 12
Leaf type
Sclerophyllous 15
Coriaceous 6
Mesophyllous 21
Dispersal syndrome
Vertebrate Pal
Invertebrate
Wind
Passive
Seed bank type
Soil 31
Canopy 3
Dispersed 8
Species richness 42

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 frequencies of persistence
syndromes ‘after crown and understorey fire were
compared with a G-test for independence (Sokal and
Rolf 1981).

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

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

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.

MGNP

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
19 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).

15

7 Life-history traits and environmental
variables

12 To examine the relationship between
plant traits and environmental variables,
constrained ordinations were derived from
the full floristic data set. Patterns of life-

21 history traits and environmental correlates

4 were tested with Canonical Correspondence

Analysis (CCA) using CANOCO™ 14.5 (ter
Braak and Smilauer 1992). A life-history trait
4 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.
26 Plant persistence syndromes were also
3 included in the analysis. An environmental
matrix was constructed using the variables

5
recorded in each vegetation survey
quadrat and climatic parameters extracted
34 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
environmental variables was determined using 499
Monte Carlo Permutations and the forward selection
option in CANOCO™ 14.5 (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

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)

Low intensity Fire (MGNP)

Ridge Slope Gully Total Ridge Slope Gully Total

Fire Response (II-VI)
II. Killed, soil-stored seed bank 5 5

IV. Resprouts via root suckers 3 4
V. Resprouts via basal stem 13 14
VI. Resprouts via stem bud bank 1

Both IV & V 5 4

Fire Response

Killed (> 70% killed) 5 5
Resprout (< 30% killed) Dil 23
Variable (30 - 70% killed) 1 0

the frequency of fire-response traits between crown
fire (WPNP) and understorey fire (MGNP) (G? =
2.68, P > 0.05). The frequency 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 from 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)

Resprouters

+ seedlings 9 10 12

- seedlings 7d) 15 2
Killed

+ seedlings 6 D 3

- seedlings 1 2 1
66

6 4 D
Z 4

12 16 10 14 133 18
1 1 1
5 8 3 4 4 3
5 6 3 1 1 2
20 BY 15 20 21 Ji)
1 1 1 1 1 3

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

Coriaceous sites and low-intensity fire
sites (G?= 2.84, P > 0.05).
1 Persistence syndromes and
29 leaf types

Sclerophyllous (dry forest
taxa) species had a higher

frequency of resprouting
species with post-fire
seedling recruitment than

mesophyllous and coriaceous

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

M.L. CAMPBELL AND P.J. CLARKE

Passive
CorriaceouS4

Shrub

A Ground cover

Wind

A

-1.0

1.0

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 (G? = 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

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

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

67

RESPONSES OF UNDERSTOREY 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 from 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

68

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 fire per se. 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

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 seedling
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; Williams 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
MLE 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 48, 1-26.

Ashton, D.H. and Attiwill, P.M. (1994). Tall open-forests.
In “Australian 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.
Oikos 89, 409-416.

Binns, D.L. (1991). Vegetation dynamics of Eucalyptus

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

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.F. (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, C.D. (1991).
Recovery after wildfire: warm temperate rainforest
at Jones Creek, East Gippsland, Victoria. Australian
Forestry 54, 157-173.

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.

Clarke, P.J., 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 national 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-
burn 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 dynamics 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, 71-87.

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.P. and Foster, R.B.
(2000). The demographics of resprouting in tree and
shrub species of a moist tropical forest. Journal of
Ecology 88, 765-777.

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
and Melbourne University Press, Melbourne).

Stocker, G.C. (1981). Regeneration of a north Queensland
rainforest following felling and burning. Biotropica
13, 86-92.

ter Braak, C.J.F. and Smilauer, P. (1992). “CANOCO
version 4.5 reference manual & users guide to
CANOCO for Windows’. (Microcomputer Power,
Ithaca, NY).

Turton, S.M. and Duff, G.A. (1992). Light environments
and floristic composition across open forest-rainforest
boundary in northeastern Queensland. Australian
Journal of Ecology 17, 415-423.

Unwin, G.L. (1989). Structure and composition of
the abrupt rainforest boundary in the Herberton
Highland, North Queensland. Australian Journal of
Botany 37, 413-428.

Vesk, P.A and Westoby, M. (2004). Global patterns of
sprouting ability: can all plant species be divided into
sprouters and non-sprouters? Journal of Ecology 92,
310-320.

Whelan, R.J. (1995). ‘Ecology of fire’. (Cambridge
University Press, Cambridge).

Williams, P.R. (2000). Fire-stimulated rainforest seedling
recruitment and vegetative regeneration in a densely
grassed wet sclerophyll forest of north-eastern
Australia. Australian Journal of Botany 48, 651-658.

Appendix 1. Life-history traits of 65 montane wet sclerophyll forest
understorey species at Washpool and Mummel Gulf National Parks on
the New England Tablelands. * indicates species not included in CCA;
** fire response of species recorded at both sites. Nomenclature follows
Harden (1990; 1991; 1992; 1993).

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

M.L. CAMPBELL AND P.J. CLARKE

dNdM - A® AI posiodsiq O}PIQOIOA, [[Aydosoyy qniys BulmMey eneuodng ovooeneuodng

dNdM ‘NOW - A [Ios O}PIQO}IOA, [[Aydosoyy 991], Wo][onw evipueipuq ovooriney]
xxdNdM dNOW + A [10S o}PIQO}IOA, snosoellod, (99TL []eUIS snjejnones sndieoosrjq svaoedivooar|q
dNdM - A® AI posiodsiq OAISSU SnodovlIOD, 901], seqjesses vroydA10g SLOOLILUTUO

dNdM + Il [10S dAISSed [Aydosoyy do] [ews ,eshzesoul vovuopog avoorpulors

dNOW - A [Ios END CRINCYN snosdellop, dol], [[eUls stjeqsne sorkdsoiq ovoovuog”

dNdM - A [10s SAISStd snosorllog do1], [[eUIS Soploqsofoo eiueyuod QPOdRISE[OD

dNdM ANOW - A [Ios O}LIQOLIOA, [Aqdosoyy do] [[eUS episi1 eA1e90}dA1> ovooumney]
dNdM - A [los QJLIGSIOA, [[Aydosoyy ddI], [[eus suoosooneys eAIrd0}dAID avoorineT]

dNOW - A [IOS a}eIGo}OA, [[Aydosoyy dol, [[eus Bye[OOAOJ vAIBOO}dAID, ovooviney

xxdNdM dNOW ‘ A LSS SRIQO}-IOA TAydorsjog quis epylpenb eursoido5 ASSIGNS
dNdM - A [los aeIQO}IOA, snosorllo09, 901], TOAT[O WINUIOWeUUID ovoovIney]

dNdM - A [los oAISseg [Aydosayy 201], PI[OJILIIOS BUIODTT[ED ovoovluounD

xxdNdM ANS 4 A [Ios dAISSed [Aydosopy 901], esonorued eianjopled ovooriuounD
dNdM - A® AI possodsiq OJVIGOIOA, [Aydosoyy qniys pnuieo elUuAoIG avooriquoydng

dNOW + A Adour) pulm [[Aqdors]9S qniys ,esornurds eisyueg avd0R9}01g

xxdNdM ANOW + IA Adouep pul [[Aydorsjos dol] pjoonuour ‘dsqns eljoylus9jur visyueg ova0ro}01g
dNdM\ 4 Il [Ios SAISSed [Aqdosoyy dO], [[eUIS BI[OJIOIIOS vISe[OIO|SV ovoorIny

xxdNdM dNOW z A [10S S}eIqoyen: 11[Aydosoyy Sor], T]eWS Hopyooq snuAwopoysyory ovooR}IAy
dNOW + A [10S QPIQOIOA, [[Aydorsjos qniys Ryesoldde OYOINOINY ovoorpuioed”

xxdNdM dNOW - A posiadsiq O}BIQOLIOA, [[Aydosoyy dol], TIUjTWIs eUOWOy ovooRyAy\
dNOW 7 II [l0S S}erqe HeAuy 11Aydorojog quays UT TOFTONM BIOBOV SuSE UH)

xxdNdM dNSW + I [los d}PIQOIOAUT [[Aydorsjos 901], uo|AxouROur vIoOROVy ovoorqey
«xd NdM\ dNOW + A AI 110S SEAGER CYA | 11Aydorsjog SERIE MOS ARE ACL ONY ovaorqeyy
dNdM 4. Il [10S 9}PIQO}IOAUT [[Aydorejos sol], [RWIS PIVIOME PIOLOV ovoorqey

dNdM 4 A® Al [Ios o}PIQOIOAUT [[Aydossjos sor], []eWIS STUIOJIO[LJ BIOROW ovaorgey

BIIKS eS oa . yueg poss opoy [esiodsiq adXy yeoT WHO YIMOIH satoadg Ayrue.y

via

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

RESPONSES OF UNDERSTOREY SPECIES TO FIRE

dNOW

xxdNdM dNOW

dNOW
dNdM dNOW
dNdM
dNdM\ dNDW
dNOW
dNdM\ dNOW
dNdM.
dNdM dNOW
dNdM
dNOW
dNOW
dNOW

xxdNdM dNoW

dNdM.
dNdM dNOW
dNdM

xxdNdM dNOW

dNOW
dNdM\ dNOW
dNdM\ dNOW
dNdM dNOW
dNdM dNOW

ails

= A® Al

o1-1SO, oll

IOS
[HOS
[10S
TOs
[10S
[LOS
[0S
[HOS
posiodsiq
possodsiq
[10S
pasiodsiq
posiodsiq
pasiedsiq
[LOS
[IOS
Adour
Adouepd
[0S
[10S
[LOS
posiodsiq
Adoued
10S

yueg pases

aJPIQOLIOAUT
EXTACEINEYN
OAISSeg
OAISSeg
d}PIQOLOA,
a}PIQOLIOA
aJeIQOLOA,
a}PIGOLIOA,
OAISSeg
Pulm
d}PIGO}IOA,
pul
Pur
Pulm
d}PIQOLOA,
d}PIQOLIOA,
dAISSeg
Pulm
O}PIQOLOA,
O}PIQOLOA,
d}PIQOLIOAUT
OIPIQOLIOA
Pulm

9}eIQO}OAUT

jAydorsjos
j[Aydosoyy
yJAqdorsjos
jAydosoyy
[[Aydorsyasg
[Aydorsjos
[Aydorsjos
JAydorejos
yAydosayy
JAydorsjos
yAydosayy
SnosoeII0D
snosoRI10D
snosorlI0d
SnoaoelI0D
jAydosoyy
|[Aqdorsjos
[Aydorsjos
[Aydorsjos
[[Aydorsjos
[[Aydosoyy
|[Audosoypy
[[Aydorsjag
yAydosayy

odky yeoT

qnays

doll [ews
qnays
qnays
qniys
qnays
qnays
qniys
qniys

soll

sol] []eWs

qnays
qnys
qniys

dol], [[eWIS
qniys

901],

qnzys
qniys
qnzys
qniys

sol] [[eus
qniys
qniys

WHO07 YMOIDH

UNTTOFTOT]T VNIGo[opod
winje[npun wn0dsojIg
vol] sUe-OOU voOUlIg
eULsnst] ‘dsqns euLjsnsi] eojowig
RJNI eIUOOSIOg

RIPOU BIUOOSIOg

Seoul] BIUOOSIOg
eyoun{uo0o eluoosIog
SNI[OJIWUSOIp snuuTeY}0zZO
BS[99XO SO}LIO

,SUS}IU snyjUe]eUIO
eljoynisoddo erres[O
TAUOAOD BIIeSTOC)

Rloordye eiealoO

Vy ‘ds vorjoiony
PODOSOIOIUI SdOoT[oI\
snjloyuod uoUIa}soydoT
BI[OFETIS elyeulo 7]
snjejoaour] ‘dsqns snjejosour] uosodooneT
layooy uosodoone
Sljeqsne eiozosIpuy
eljoyysnsue eAreokpoyy
eyUeLIO voyeyy

BIOFO] BIPOOS

soroadg

avoorqey
svoovlodsoyig
evoorlfouAy
ovaorl[ouAy TL,
ava0vojOlg
av90R9}01g
avaovajOlg
avadro}0lg
ava0rlojsy
Oe99R9}01g
ovooriqioydng
avaorlojsy
avo0R10jSV
ava0Rla}SV

Pe) he) )(6)
avoorny
OvooRLIAT\
avodRoj}Olg
ovoorpuioedy
evoorpliordg”
avoorgey
SvOORIUITUO/\
avooRoj}Olg

avoouqey

an MP a UU OD aan SUCCo = i <" ts = == “lL = a= SC EE nn
js poy [esiodsiq

Ayrouey

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

72

M.L. CAMPBELL AND P.J. CLARKE

dNdM
dNdM
dNOW
+xdNdM dNOW
dNOW
dNdM
«xdNdM dNDW
dNdM
xxdNdM dNOW
dNdM
+xdNdA\ dNOW
dNdM.
dNOW
dNdM
*xxdNdM dNOW

Il
Il

[10S
THOS
THOS
TOs
THOS
THOS
pasiedsiq
THOS
10S
[f0$
10S
THOS
[LOS
[LOS
[IOS

OJPIQOLIOAUT
a]PIQOLOAUT
O}PIQOVOA,
ayeIGoLoA,
EIU EIBEYN
ayeIqgoyoA,
a}BIGo}OA,
O}PIQOVOA,
d}PIGOLIOA,
EXVACEIREYN
a}eIgoLoA,
ayePIQo}IOA,
o}eIQS}OA,
9}PIQOVIOAUT

O}PIQOLIOA,

y[Aydosayy
|JAydosayy
[[Aydo1ejos
[Aydoreyos
y[Aydosayy
yJAudosoyy
yAudosoyy
yAudosayy
y[Aydosayy
jjAydosayy
snosoell09
snosoe1I09,
[Audosayy
[Aydorejas

JAudosayy

qniys

dol], [[eWs

qnzys
dol], [[ewWs
qnays
qnays
dal], [BUS
qniys
qniys
qniys

2a1, [Tews

del], [[euls

qnays
qnays

aol],

Tmyyrtus ‘dsqns pyyrus epi

SUODSOIOGIL LLIOIZ,
y ds edieosoyooly,
euliney] ediesoyooly,
PJeVIdYs elUUeUISe],
epidisut eruueusey,
uinsojnpueys winouAS
UWIN}}SoAosuep WNUL[OS
BICAO PLIOUIOZIYIS
,susosogni eIuWMepoyy
SljIqeuea vourdey
euelyiMoy eourdey
SoploraoluO] eLyOysASG

q ‘ds vovucyng

Vv ‘dsqns eljoyionquies seroskjog

avoorINy
avaoriny
avooupuoed”
ovooepuoerdg
QPOORIO}UI
SLOORIOIUI\
aeaorlpay|
aeaoRuLloS
avoorimound
avooryIAW]
SvooRUISIAY
OvaORUISIATY
evooriqny
avoorqey

aeooel[ely

73

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

ue

PS PONERB AD AVA

iw

ind Yi Ww

“hirtn

Bie acct
a

ees
iets
“Blix
ola SiN

nA

giggwM,.. «=a Jigme
ihedgueat4, , chewit.

Pre SOTLEY) Pry spacadye threenz) ey
TenawTee 5 wd] rpms yo EDS) FeO. ae.
1AQ Sys yy qrulepy SOA GUOS Thy Se | 7
iwi mt ee? tte} agony yg $
; = © = {
' ¥ j ¥ Sees oT aes
aS.
sAlp
t ! : te -
® iL t inl
| ih

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

Ross L. GOLDINGAY AND Davip 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 127, 75-81.

The significance 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 tracks. Habitat
disturbance was dependent on site category, occurring at 8 of 10 near sites compared to | 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 future 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. 1998; 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 1997; Shine et al. 1998). 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 further 19 reptile species under loose rocks (eight

DISTURBANCE TO ROCK OUTCROPS

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

snakes, three geckos, seven skinks and one dragon)
during a regional survey for the broad-headed snake.
Rock crevices are commonly used as retreat sites
by many species of reptile and frog, 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

76

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 bounded 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.

Survey Sites

Areas of rock habitat suitable for
survey were identified from 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) from
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 | for location
details).
We selected rock platforms that contained at

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

R.L. GOLDINGAY AND D.A. NEWELL

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 occurred. 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

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

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

Of 22 sites chosen for survey, 10
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
caim 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 instance 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 from
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. Eulamprus
tenuis was the most common species, being detected at
11 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

Wi

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

1 0433091 6736800

N

0433091 6736839
3 0432567 6736726
4 0433539 6737182
5 0433761 6737538
6 0434263 6737888
7 0433590 6730564
8 0433556 6730491
9 0430090 6732380
10 0429603 6732331
1] 0432309 6736252
12 0432232 6736539
13 0432268 6736821
14 0432140 6737080
15 0432161 6737590
16 0432542 6737299
17 0432776 6737429
18 0433024 6737329
19 0433834 6730303
20 0433924 6730049
21 0434210 6730149
DD, 0434288 6730664

Distance

(m)
Near (50)

Near (20)
Near (200)
Near (150)

Near (50)

Near (20)
Near (100)
Near (300)

Near (50)
Near (200)

Far
Far
Far
Far
Far
Far
Far
Far
Far
Far
Far

Far

Rocks

29

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.

78

Good

11

10

This

Reptiles Types of Disturbance
: Cairn (of 3 rocks), broken
rock, displaced rock
- Cairn (of 12 rocks)
2 Bt Cairn outside transect only
3 Em None
1Et Broken rock
- 3 cairns (13, 15,16 rocks)
6 Et Fire place, rock seat, 2 cairns
(3, 3 rocks)
1Et,2Ew None
- Fireplace, broken rock
Ew Cairn
1 Bp None
3 Em None
- None
1 Et None
- None
1 Et 1 flipped rock
- None
1 Et, Ec None
2 Et None
1 Et None
3 Et None
2 Et None
DISCUSSION
study has provided some important

insights that will extend our understanding of habitat
disturbance within protected areas. We detected

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

R.L. GOLDINGAY AND 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 from 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 from
access points on the likelihood that disturbance will
occur. This provides Park managers witha clear insight
for managing rock habitats. That is, areas within 500
m of existing tracks are likely to be associated with

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

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
useful 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 further 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

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 frequency 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

80

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
understood.

This study highlights that disturbance to loose
rock habitats is not confined to areas around Sydney.
We could generalize from 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.

ACKNOWLEDGEMENTS

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,
P. (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 14, 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.

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

R.L. GOLDINGAY AND 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. Lunney) 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).

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

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.

81

"=
4
“aT
7
:
'
/ hie —s
Ai rule ear nahy ths
t way to (he Weems
: ’ '
{ al
) | ive
{ 1 pai i
VA!
; ahiitat
: té
4
oy
iy
q
‘
‘ ; yt
]
| Mel
' te
cw
rt :
7 t his 4 oe ni
frequen i cdistirdanc i Tae
, A
th } ) CoM .30T TOOK
ec the munber Of POC Ks suitehié
| mew differ across oie Categeties
io Gites Was nol ditvorent
{ ; survey was nolopiities lar
reriiles that use loose cocks
mid. dtas Likeiy that somo soecies © tore sanswuve
yar thers. Surveys comuotid dunng

cece rer sir

oe

%

ars

hy
. as 2 nine MR
eee le rl
aria Bey, iat
nig bs bade thet, ay
a mderatacxiiog (the namnber BE spre :
depen gittrtsODCRMNOS) MN yegniblod a
eal dni? ywolt lo etzavdl sili wi 2gonl bane
“S Senet ten 2 aileneAIe agitav

nyoR pACORS my (ppm et hs)
dgonbye Wel be. vinioe? Inviguio

ark eed

lanciteVl wor wile oht luc} lanai lero s(e00as
to nel? ord nor med haw
RNY bone et noi

ao Se ‘emit CARE ‘aise M

abn pall eit? aint BN et i

nnerveatieut
0 Shi oh |
ghia tion ne ‘
Rive Cie, #5 fot
bint Waa Wri Sie 0 ey eg
: aeSt ar }

oe atte ‘duit title’
bre zat thlioiiobt"W2

it

wile SOP oid ty Bit She J

Socqenaweing Dl ty elhineZat BORD pe.
> arctan, sessaens eon

Uh yb q: * wap ie ei!

soloaihak Bis itis) fei 5 to 2ovitos
Goldings sae dai dapmbahaire diese eier
arora ntias PqeeCs yet), blening it bas
tub “oisege stagnabes rea Mp avi
sshivmaswaladelenacinnls reves
Crip OEMS Ribena
(8 ye GEE, | ie bite sigh
bemgnabns. But iond ; rs aie 0 Jay 0;
hfe 7

rahe
ih ene eA
pa at

dwelling Kamel ‘ical Co ”

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. 127, 83-91.

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 1s 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 stream 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.

KEY WORDS: amphibians, Gibraltar Range, Great Dividing Range, Great Escarpment, Hylidae, mesic

forests, Myobatrachidae, 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).

Anunderstanding of the composition and ancestry
of the amphibian fauna of the Gibraltar Range is
underpinned by interpretations of the geological
history of the landforms of the range and its climate.

The Gibraltar Range occurs on the eastern boundary
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. Ollier (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 refuges 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 streams
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 Mann 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

84

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 Environmental 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 (IUCN) categories
(Global Amphibian Assessment 2004).

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 further 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
myobatrachids. 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 distributional records, and even new
species, may be found there (Donnellan 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 Australia
(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

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

altitude in Washpool National Park (NP) and Ewingar
State Forest (SF).

Frogs of forest communities of the 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 stream habitats. Among these, three
species, Litoria subglandulosa, L. piperata and
Mixophyes balbus(Fig. 1b), are restricted to higher
altitudes on the escarpment. Vegetation community
is not the major factor determining their distribution;
they occur in streams within heath, dry forest, wet
forest and rainforest communities. One stream frog,
Mixophyes iteratus (Fig. 1c), 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 (L. 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 abundant 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 L. 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

85

AMPHIBIANS OF THE GIBRALTAR RANGE

Typical breeding location

Association with major vegetation community Pond

Conser- Stream and/or Ephemeral Toes!
eggs and

Species scientific and common name vation breeding swamp pool ‘
status (lentic) | breeding | breeding BOSON
(otic) stage.
Widespread occurring
in many vegetation
communities from
rainforest to grassland
Litoria caerulea Green Tree Frog
L. dentata Bleating Tree Frog
L. fallax Dwarf Tree Frog
L. latopalmata Broad-palmed Frog
L. peronii Peron’s Tree Frog
L. verreauxi Whistling Tree Frog
L. wilcoxi Rocky River Frog
Adelotus brevis Tusked Frog
Crinia signifera Eastern Froglet
Limnodynastes dumerillii Banjo Frog
L. ornatus Omate Burrowing Frog
L. peronii Striped Marsh Frog
L. tasmaniensis Spotted Grass Frog
Uperoleia fusca Dusky Toadlet
U. laevigata Smooth Toadlet

MP PS PS POS OM OO OO OS

belt. Wet forests excluding

rainforest specialists

Litoria barringtonensis Barrington Tree Frog
L. brevipalmata* Green-thighed Frog
L. chloris Red-eyed Frog

L. gracilenta Dainty Tree Frog

L. piperata Peppered Frog

L. revelata* Revealed Frog

L. subglandulosa Glandular Tree Frog
L. tyleri* Tyler’s Tree Frog
Mixophyes balbus Stuttering Frog

M. fasciolatus Great Barred Frog
M. iteratus Giant Barred River Frog
Pseudophryne coriacae Red-backed Toadlet

Rainforest specialists

Assa darlingtoni Hip-pocket Frog
Lechriodus fletecheri Sandpaper Frog
Philoria pughi Mountain Mist Frog

ablelands and western
species. Woodlands, dry
forest and grasslands
Litoria booroolongensis Booroolong Frog
L. castanea* Yellow-spotted Bell Frog
Crinia parinsignifera Beeping Froglet
Pseudophryne bibroni Brown Toadlet

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 frogs
from this group is not unexpected, given that there is
no tree frog 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 frogs
that are restricted to rainforest habitats.

Two of the ground frogs, Assa darlingtoni (Fig.
la) and Philoria pughi (Fig. 1d) 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 IUCN 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.

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

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
frogs (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.

87

AMPHIBIANS OF THE GIBRALTAR RANGE

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 refugial 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

88

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 Australian continent; the tree
frogs of the family Hylidae and the ground frogs
of the family Myobatrachidae. These families are
recognized as being of Gondwana origin (Tyler
1979); they are old endemics. Molecular genetic
evidence indicates that the ancestral tree and ground
frogs were already well differentiated at the time
Australia separated from Antarctica some 52 million

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 further species found in montane
rainforest in Papua New Guinea (Donnellan et al.
1990). Phylogenetic studies (Heyer and Leim 1996;
Kluge and Farris 1976) place this genus in a basal
position among the myobatrachids and their current
distribution and habitat preferences strongly suggest
they have had a long association with the wet forests
of the Great Divide and escarpment. Among the tree
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
distributed 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 strong 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

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

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 retreated further west the higher altitude
ranges of the Great Divide and their fauna and flora
were isolated (Ollier 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, Philoria pughi, and
Lechriodus fletcheri) 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 1991), 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 distribution
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
(Knowles 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).

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 frogs (Berger et al. 1998, 2004).
High altitude stream frogs 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 frog (L. piperata), deserves more detailed
consideration. This frog 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
from Diehard Creek, which drains south-west from
the Gibraltar Range to the Mann River. Conservation
assessments of the peppered frog have been fraught
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

90

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 frogs were found 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
from each of these sites with a larger collection
of specimens of members of the Litoria citropa
species group from 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
from 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 (NSW NPWS 2001).
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

lam 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.

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 38(3), 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.

Donnellan, S.C., McGuigan, K., Knowles, R., Mahony,
M.J. and Moritz, C. (1999). Genetic evidence for
species boundaries in frogs of the Litoria citropa
Species group (Anura: Hylidae). Australian Journal
of Zoology 47, 275-293.

Heatwole, H., de Bavay, J., Webber, P. 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 Zoology No. 233, 1-29.

Knowles, R., Mahony, M.J., Armstrong, J. and Donnellan,
S. (2004). Systematics of sphagnum frogs of the
genus Philoria (Anura: Myobatrachidae) in eastern
Australia. Records of the Australian Museum 56, 57-
74.

Mahony, M. J. and Knowles, R. (1994). “A taxonomic
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 Donnellan. 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 frog (Litoria

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

pearsoniana) from mid-east Australia. Molecular
Ecology 7,175-186.

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).

Ollier, C.D. (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 P.L. Beesley) pp. 35-40. (Australian
Government Printing Service, Canberra).

Robinson, M. (1993). “A field guide to Frogs of Australia.
From Port Augusta to Fraser Island including
Tasmania’. (Reed, Sydney).

Smith, A.P., 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). A new species of
Litoria (Anura: Hylidae) from New South Wales,

Australia. Copeia 1985(1),1145-1149.

91

ii ae

ria i

4
398d

ESPs eZ, | is

‘ ' *?
; i u an Ss

vathonow sith ty sr

Lay °¢ i ah tant me 1 < etn Es ay bya
Pope oe pba Llenosel gulgyl wa
hey Y = ot hig ehhzgevibyerE i =
‘% rs
we) SORT bab Sher onee 1
BN ons, Ae ginT Kak pause Bret be. fooggnrall
senor Nlnath iapaerheon ek bot ieenseeyo Tbk the
oateapaore hae. zaristieg qos A vistlinesid
wehad seus eV Inca siecmah seootaied) ibys.
lennite rmibepen: Ag CLL. an Comins, A dt ie
[ ariodsig > , 45] ivtoe SBUW ale ei <
“aie iene Alt omhnGey teat Sila sad Rub ROTC
“WS Rien ‘oyilase a Yoreaeol,
nino whist (MOS) deosd
hina, DeletreaA Wale vg
BL. chee: asespalen bth, ernie RI Ws
Pah ge ae ) He tga iy ou fewe 43 aft to OOVHELS
_—s | Gre ‘chat o> 2 rans
“pid Yager yhoyorH repay PP) oes toe GO po

Y map shosta scugs

2° BaUe SHISHA bind Demis Tontip FiN Abe wom
eetiwl fete, eu ORIG waitinene dd QA Avi ale tea
Kivh t Hob) smo A) eiqogT, WW A paleo A Me A,
wnat 5 WianSahe moet wan) se? 1S8ef)- die
(ESA iets
ye a rot Wig $41 f oeha walgodoa APRS Lyn
Sgbit vac us 3D) 2 i gs bve 2x “bay aoe st oF a

RA A Fo ad ORR COL fir AGA RO Ghai gltCT

(ini hb Rehhy iis sine eitinint be ie hav
reat tgreuh Oto Egy iyekoet al 1 bossro 8 thigg

\ (eet) 529° 6 pains SORE ROD) (le
silaveah Ae qBOvL oLoby pa trler AS SEMEL Ds M anenidod
spisband ot bangle | tear Oo) greens nef jon,
“a ve 50 we “Sif aril
iBT) Head spo) Mt ne Le itch bA hf 5h Aine
‘ Syae hee hits Hoe SW bh lattes Po
ih Dy a trot eA Rarenraiy tae des
tigen ths HD! lapse? uae iS oes ic
Ditirmndb yodreque® 219 uae nee tt
sole! sinigd welt te tesro Pane |
7" smugekl ig. Asvigal ;
ACR 1). eon! duane meld to sas Anz

JA voab

Tee a

ReRNOv

0820 qo 3*

uamegsnaM Digprsitg Lait 04 ARTI, WAP #10}, 7
JTatistel? tpg He bone the st e | Ke ae
igonecaes ; iqomistite Taugial equi

' Ao? Te sqifentstint tbo hates ty (eC OF) | CM YT

i SttA Hi SAPO GE atin a eine ein A tv er
nN aeqntltae cd But aie at Gules torre |
Mts Ae Tet: vty (rretised chi Benn
Saat. Aivbaorgonntil vee pe inden ie lay
eS} AL -crvew herpes wre ake Reptior 2Rb »
- LH ALOR 2 rie ‘BT 8)) Ma canbe ban Lv suit
bunton pied.) aygeg Dall sigs, dole uay Ath Lt
ye see scutes agua, ; egelont ‘ Ian \ mploiual a
AY Oo ke ae
st ste Woks A » ead Re iste ha suey
“aN be! Poet al ill rf FouBieel ity SP tepetie

Vital ity 4 hers
bie ate oh) 1 iyeper Bivgho - eiteban ®

So Spec PSAs RTS,

le et LAR VIS re y <ayt ats af iyo) weTe

vn tocar soe) tau
ward BPttee eB Ah aoenbetii
c Heat niki Rap

Ai 3 we

ve aoe oot Maat

oo Saati? ei 3°

SPECI CS yeh EN REF 1s

and cpasihl tel placed eR
species. ‘Yornnidideprih tire ulin Aris) Os
L ineorisa Te mmr nbet powcr To
a oat eomends a

2

ey a aoeinicdenetatoeee 1 at
0 <ClOF SURBBIVS Cita. nai dicho ran
wethic aaa paneer ae x omanaten eee

aa POSH

calli ap ate peagcamay leo.
histtricge it an eA cS RECO
Tp ees eth len eset
iS presumed .cxhiaot Red (RED ol ipl

esit een Cihoe, A erent afte wba meds A A

«83,192 Q gE asd oe de aciteapalaye
2 erate bn fompiiose tedoyM sewn A) cron aun
2 Az tumaté Nealiworh. oft \p avoo09A silevaud

rare ere im bas

Siig

AOA Le Peet) A 28 4 baie at M, Be
etesnat WHA seeSt ian Ravel ibete waive
af nv moqoh viinsviboid weno oe abi He

ANT W hie dike torigheo resi ae
Gelder bi ahiaanea Seve Dona

via tin Ano oraniannianrta car (088),
Sicha ir !

tactrrotaist Bink pesca
winotil) 2071 nite oe we

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 2351.

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 mammal 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 ground-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

Despite Gibraltar Range National Park being one
of the oldest parks in New South Wales (reserved in
1963; NSW NPWS 2005), relatively little published
information exists on the mammals in the park. Surveys
by Osbourne and Marsala (1982) and Pulsford (1982)
are summarised by Clancy (1999) who provided a list
of 25 native and six introduced species for the park,
but few other details about habitat associations or
relative abundances. More recently, records of flora
and fauna gathered over many decades by a number
of government agencies in New South Wales have
become available on a single web-accessible database
(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
faunal abundances, and because contributions to the
database can be made by any interested individual
(through submissions to the NSW Department of
Environment and Conservation Atlas), the veracity of

some records is difficult to confirm.

One of us (Vernes) has begun a long-term study
on the mammals of Gibraltar Range, particularly
those species that consume and disperse the spores
of hypogeous ectomycorrhizal fungi, otherwise
known as ‘truffles’. As a first step in understanding
the mammal community structure and dynamics in
the region, we have surveyed mammals at a range of
sites in Gibraltar Range, and present these data here.
In addition to providing a simple species list for the
park, we have also attempted to summarise the broad
habitat preferences of mammals that are present, and
to show how habitat heterogeneity in the park leads to
the structuring of distinct mammal communities.

MATERIALS AND METHODS

Study area
We undertook broad, observational surveys
throughout Gibraltar Range National Park in

MAMMALS IN GIBRALTAR RANGE NATIONAL PARK

northeast New South Wales, but focused our trapping
and spotlighting 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

94

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 fully the structure of the small
mammal community 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. 1a) 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 ground-dwelling
mammals in selected areas of the park (see following
sections, and Fig. 1b). 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. 1b). 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. 1b) 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. 1b)
for a total of seven clusters (63 traps) per transect.
The distance between each cluster was variable,

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

K. VERNES, S. GREEN, A. HOWES AND L. DUNN

Gibraltar
Range NP

Hphway
Minar mad

Valk ngitec ks 4
Part, boundary ; , Mulgan 's Hut

Spabiihg
Lersacts.

(b)

151 11S

1st 20or Sedge heath Canc mp)
a De apen Socaptus woodkind wih
hesth underslomey
Welapen Socapotos fost
wihfem understory

4 Rokfars! i
Tlapen Secalpotas fom wih
| mokl undertony

Rock auicmps.

== T2 Goadbnt larsact

—— Whore

: ———- Wakhglacts era

awe

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. 1a) 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 (G1 and G2; Fig. 1b) 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 future 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 11 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

96

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 (Zrichosurus vulpecula)
in eucalypt forest, but this species appears to be
considerably less common than 7: 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 boundary (near
that end of Cedar Track), and we have included it
in our species list because the Sydney blue gum (E.
saligna) 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 parma). 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

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

K. VERNES, S. GREEN, A. HOWES AND L. DUNN

40j01g vIqQU]IJDY = AA :viutsvd sndososvy = AI ‘SHay ajvsojAyy = LI ‘Snavusdd sajvqossp = qV ‘snuiuvd snansoysiuy = DI, svjnaad
-]NA SNANSOYIUT = AJ, *SNasauId SOJIADOISVYT = Jd *SdaIidadq Snanvjag = Ad *snulssasad snsiayIopnasd = dd ‘Suvjoa saploinvjad = Ad
‘ninspu Sajaubiag = Nd *}90S8ue.1) SuHYSIVOdS Jo wy 19d Uses sjeUUL JO S1BquUINU sv pojUdSeId o1¥ ddUEPUNGE Jo SajIEUISS DANLIOY

(8) “SIq UO UMOYS 9S0Y} 0} puodsa1109 sisquINU IESURIT, “SNOT Jequiej}deg pue ¢Q0z7 Av] U9OMJ0q (NAA) WV [VUONEN JoodyseAy
JO u0Ndas yUDde[pe UL pue YIvg [LUOHEN osULY 1vz[VIQI) UL PoJINpUOdD s}dJ0SUL.1) SUNYSITOds UO Uses [eUTUIEU JO SaIdadS “{[ 21qQ¥UI,

jsoloyurTe 0} SUIpe. yoely
£0 LI 60 Tl 90 6c IL 00L ¢ ? Ol
pue[poom ydAjeone uedO sjjeq eryepueq
jSOIOJUTEI 0}

vO 6¢ 19 L0 00L Vv 3 : yet] SSTPSON SY. L 6
Ulpeis “‘pue;poom ydAjeone
09 OT 08 00s c pue[poom ydéjeona uedQ =} INdAL_D S_uestTINYA| 8
80 a 76 009 G pue[poomjd4jeone uedO yorly, Yoo [Auy ti
val 00L I pue]poom ydAjeone uedO SALIC Sues NYY 9
LO OV 00ST I pue[poomjdAjeons uedQ@ =«=- or, SIOWIeNbpeoyy ¢

jSOOJUILI 0} SopeId

L0 LO LS Ce 00S € k eM [oodysen Vv
Ppornb yey} ysor10F ydATeone

(dN)
CL 008 I pue[poom jdAjeone usdQ €
peoy oytueID
(dNA\)
€0 LO £0 L8 0001 € JSoroj dATeono 49M IIe], F Z
BA [oodyseA\

Pus SpIeMO} JSoIOfUTeI

AG LO) 80 Wr Ce EO 0001 € E (dNM) PIL Jepapd I
0} SUIPeIs ‘jsoIOF JOM [BL

aM dN ILL dV OL AL Od dd dd Ad Wd a) poAoains yJoquinn,
ns 2 = LS... = SS ES ee So ro UIST UOT}diIOSep so10,7 oUIeN
why Jod uses sjeuttue Jo sloqunN SOUNT} “ON yoosuedy,

JOOsSueT]

Of

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

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

PROTOTHERIA (Monotremes)

Family Tachyglossidae

Short-beaked Echidna, Tachyglossus aculeatus
Family Ornithorhynchidae

Platypus, Ornithorhynchus anatinus
METATHERIA (Marsupials)

Family Dasyuridae

Spotted-tailed Quoll, 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

98

Detection
Method

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-kill”
Spotlight, heard, scats
Scat, seen, spotlight

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’

1Listed as ‘Vulnerable’ 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 (7:
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, our 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 introduced species of non-flying
mammal (Table 2).

The mammal community along the swamp-to-

rainforest gradient
Seven species of small mammal were detected

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

Detection method

Trap
Trap
Trap
Trap
Trap
Clancy (1999)

Scat, road-kill
BioNet

Seen
Seen

Diggings

in our traps across the habitat gradient (Sites T1—
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 introduced
house mouse (Mus musculus), the brown antechinus
(Antechinus stuartii) and the eastern pygmy possum
(Cercartetus nanus). Spotlighting yielded a further
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 spanning
swamp to rainforest (Fig. 2), despite this representing
a distance of only about 700 m. Amongst small
trappable mammals, several patterns in distribution
emerged. R. fuscipes and M. cervinipes changed
significantly in abundance (KTBA) between habitats
(P=0.004 and P= 0.001 respectively; Kruskal-Wallis
Nonparametric ANOVA), with abundance increasing
from the dry eucalypt woodland and the wet eucalypt
forest, peaking on the open forest/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 Dunn’s
Multiple Comparison Test), but the few captures
of M. musculus precluded statistical comparisons.

99

MAMMALS IN GIBRALTAR RANGE NATIONAL PARK

10.0

----f---- R. fuscipes |
a8 sha
---@-- M.cervinipes hai
na ‘

Mean KTBA per period

Swamp Ecotone Woodland Ecotone Wetforest Ecotone Rainforest

5.0
----M---- R. lutreolus
4.0
Ain ---@-- P.novaehollandiae
3.0 fi,
i \, ---h--- | M. musculus

Mean KTBA per period

Swamp Ecotone Woodland Ecotone Wetforest Ecotone Rainforest

5.0
40 ----f---- A. sfuartii
an ---@-- C.nanus

Mean KTBA per period

Swamp Ecotone Woodland Ecotone Wetforest Ecotone Rainforest

oS
2.0 ----m---- P. volans
15 ---®-- P. peregrinus

Mean rate of detection
(animals/transect)

Swamp Ecotone Woodland Ecotone Wetforest 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. VERNES, S. GREEN, A. HOWES AND L. DUNN

Likewise, our few captures of R. /utreolus 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 (T1 — 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 R. 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.

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

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. Barnett et al. (1976) surveyed
the mammal fauna in a 118-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 useful
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 Vernes (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,
Barnett ef 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 Vernes (in
press) noted that 12 species of macropod persist in the
New England Bioregion. Interestingly, two of these
species, the eastern grey kangaroo (M. giganteus) and
common wallaroo (MM. 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.

101

MAMMALS IN GIBRALTAR 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 1966; Barnett et a/. 1976; Jarman and Phillips
1989) we suspect that 7: thetis is more likely to
occur around the wet sclerophyll/rainforest ecotone,
whereas 7: 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 MZ. 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) for 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. parryi) also occur in the adjacent
Washpool National Park. Thus, the only macropod
species that occurs in New England (see Jarman
and Vernes in press), but does not occur locally in
the Gibraltar Range/Washpool region, is the black-
striped wallaby (M. dorsalis).

_ Another species-rich group in the park was
the possums and gliders (see Table 2). Of the eight

102

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 14 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 from swamp to rainforest, and found
that despite the short distance (~700 m) there were
significant and consistent changes in the structure
of the mammal community. 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 (Banksia 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

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

K. VERNES, S. GREEN, A. HOWES AND L. DUNN

detected. P volans 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 ef 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 7: 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. /utreolus 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

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

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

Barnett, 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,
PE., 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.

103

MAMMALS IN GIBRALTAR RANGE NATIONAL PARK

Possums, Gliders & Koala , P.volans

DRY FOREST WITH HEATH SWAMPS

& HEATH

\ Pp ‘ P.volans 4 |

oe ee P. peregnnus f !

; : T. caninus P. breviceps : : / i

T. caninus p i> Ses P. brevicaps J {

P. Peregnnus SS ee A. pygneeus

; : i | TF. vulpecula

C. nanus : ; T. vufpecula

C. nanus T. caninus @nctos H |

1 P. cinereus ; ' i

i f

i i

\ i

j

!

T
'
1
1
j
'
1
P. breviceps PP. australis d
1
'
\
'
i
1
i
i
i]

ise pew.

=

sia ee
tiarky cree ey
feeaedy teers
PRE

Monotrames O. anatinus

©. analinus
T. aculeatus

Bandicoots
P. nasufa
Das yurids

D. maculatus D. maculatus
A. stuart A. stuartis

T. aculeatus T. aculeatus T. aculeatus T. aculeatus

P. nasuta i mecrounss

A. stuart

i
Macropods . : ,
: : T. thetia ae aie WEebicolor se ee
T. sfigmetica P. tndactylus : M. rufognseus : aa
: A. parma : M. rufognseus Wi bicolor P. pentcillata?
T. thetis : WW bicolor Af. panne
P. indactyius s M. panne
s M. rufognseus A. rufescens?

f
Rodents \ P. novaehollanadiae

: , R. fuscipes R. fuscipes ;
R. fuscipes RA. fuscipes eens dennis M. musculus P. novaehollandias
} MM. musculus

i. cananipes iM. carnainipes
1

A futrecive A fufrecius {

Figure 3. Schematic representation of the broad mammal communities and their vegetation associations at Gibraltar Range National
Park based upon the data we have gathered. The vegetation types depicted in this figure has been used and modified, with permission
from NPWS, Glen Innes.

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

104

K. VERNES, S. GREEN, A. HOWES AND L. DUNN

Jarman, P.J. and Phillips, C.M., 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 Vernes, K. (in press). The Wildlife of New
England In “High lean country full of old stories”:
Environment, 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, Macropus parma, in
New South Wales. Australian Journal of Wildlife
Research 4, 109-125.

NSW NPWS (2004). Gibraltar Range National Park
— Climate. URL: http://www.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, P.S. 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

VE. OMA RBWOR (AS ATAIENG OA

it atsi€l .08ef M24
SD aati ol goiseqe bugen
alae ey eegenarsphe

BM tes

qe
ehyraree
ew Sc

:
|
= y t |
4 : 2
{
y aw ¥
| iw all } 4
A 2 ij ? : “a al | aw f
y | Ke @ ¥ ie an ;
: 4 | = . ie 7
> > ©
= aan. (?f (8 | then
i i 3 ‘ i
i d ba . ell
é : :
1 ri
tz : t
) >
> f
. ° é .
' s ¢ ; |
-
y i

.s q 7 a
2% -aen) : ‘
*. 7 © , rf
e s ~ » .
‘ 2 rd ( j
o
e
5 ;
. 5 i
~~
4
|
4 4
. ;
°
4
5
be
» r

o*heu~
ee

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 1 @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. Fran¢ois Péron 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 19% Century illustrations of C. nanus are reproduced.
This study documents the significance of the primary source 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. manus 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. Modern 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. manus 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 B. parvus 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 knownto Europeans
was collected by Francois Péron, 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; Horner
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. manus. Whilst on a short
stay on Maria Island, off eastern Tasmania between
19 and 27 February 1802, Péron traded with the
Aboriginal inhabitants (the Tyreddeme people; Ryan
1981) for a single small marsupial. Péron (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, from a savage who
was just going to kill and eat it’. In an unpublished
manuscript (now held in the Le Havre Museum in
France) Péron 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 Linnaeus’ (Observations zoologiques by Francois
Péron, on Maria Island, unpublished manuscript
# 18043:31). The specimen collected by Péron (a
juvenile male) was transported back to France, and
is now held in the Muséum 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
Péron 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 Warneke (1963). Waterhouse (1846)
interpreted Péron’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

108

1817, 1820; Cuvier 1826; Lesson 1827, 1838,
1842; Temminck 1827; Fischer 1829; Schinz 1844;
Iredale and Troughton 1934; Tate 1945; Wakefield
and Warneke 1963). The type specimens for both C.
nanus and A. minimus were collected by Péron, but
the latter is considered to have come from Waterhouse
Island, which lies close to the north-eastern coast of
Tasmania (Wakefield and Warneke 1963; Rounsevell
1989).

Taxonomy and nomenclature

Upon the return of the Baudin expedition to
France in 1804, several of the great French zoologists
of the period, including Anselme-Gaetan Desmarest
and Etienne Geoffroy Saint-Hilaire worked rapidly
describing and classifying the specimens collected
by Péron. 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 Iredale and Troughton 1934; Marlow 1962;
Wakefield 1963; Green 1974; McKay 1988; Turner
and McKay 1989; Flannery 1994; Menkhorst 1995;
Turner and Ward 1995). Following Desmarest
(1817), brief descriptions of the species appeared in

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

J.M. HARRIS

Ey
e
=
Se.
=
a

Figure 1. This illustration above of two Phalangista gliriformis (=Cercartetus nanus) appeared in an article
by Thomas Bell published in the Transactions of the Linnean 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

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. HARRIS

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 Linnean 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
careful 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
confidence in the distinction relied on the phrase ‘les
oreilles sont arrondies et couvertes de poils’ 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
(1841) 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 1843). 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 illustrations (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 confusion in the literature regarding
a statement made by Burmeister (1837) which
translates as ‘a specific genus (Cercaértus Glog.) 1s
formed by the common brush tailed Ph. vulpina’.
It has occasionally been presumed that Cercaértus
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 Cercaértus
was used in reference to Phalangista vulpina, which
is an absolute synonym for Trichosurus vulpecula, the

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

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
Cercaértus was not mentioned. Instead, Gloger (1841)
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 confusion, the name
Cercartetus was at that time basically disregarded for
P. nana. However, it is clear that the name Cercaértus
is a junior synonym of TJrichosurus 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 J.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 Krefft 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

apt

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 P. 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;

M2

Forbes-Leith and Lucas 1884) or from Tasmania
(as accepted by Iredale and Troughton 1934). He
compared the teeth of manus (Desmarest 1817),
gliriformis (Bell 1829), unicolor (Krefft 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 Australia, 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.

Confusion 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 (Krefft
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

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; Turnbull and Schram 1973; Green
and Rainbird 1983; Menkhorst and Knight 2001).

In terms of morphology, Desmarest (1817) made
a description from 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, from
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 under,
and almost entirely conceals it.
The lips are scantily covered with

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

soft short hair, of a whitish colour,
and are furnished 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 fur 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 fur as ‘a uniform mouse-colour lighter
on the sides and beneath, with a blackish patch in front
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 infusion
in the general body colour and are yellowish on the
sides and underneath’, while the mainland form (C.n.
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

113

EARLY NATURAL HISTORY OF CERCARTETUS NANUS

COMMON DORMOUSE-PHALANGER

PLATE XVII.

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.

114

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 ‘furnished 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

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 1854 (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 Warrnambool, Victoria, but no
specimen was collected. The first collected specimen
from Victoria appears to have come from 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 from Gembrook 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

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

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 from NSW include those
from 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). Krefft
(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. manus. 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 further 27
records, and subsequently the status of C. manus in
SA was changed to ‘Vulnerable’ (under Schedule 8
of the National Parks and Wildlife Act 1972) (van

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, fruits and insects (Lord and Scott
1924; Le Souef and Burrell 1926; Troughton 1931;
Bocking 1939; Conway 1939; Hickman and Hickman
1960). In the wild, the first feeding observation was
made by Krefft (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 1867). Gould (1863) 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 ef al. 1986; Dickman and Happold
1988; Tulloch 2004).

As early as 1863 it was recognised that ‘of all
trees it prefers banksias’ (Gould 1863), an observation
which is supported by modern 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 from the literature
about the nesting requirements of C. nanus. Le Souef

116

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 from the
weather, making their nest at an angle. The nest is
composed of scft 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. manus] 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 “from
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 (1829) 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

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’.

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

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 fully
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 1829; Lesson 1830; 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 Burrell 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; Krefft
1871; Thomas 1888). Bell (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, Bell (1829) was certainly mistaken
in asserting that C. nanus does not undergo 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

ALT

EARLY NATURAL HISTORY OF CERCARTETUS NANUS

DROS IA Tee ee eC

Figure 4. This charming illustration of a pair of Dromicia gliriformis (=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
Goulburn 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 Péron’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

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

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: Fran¢gois Péron’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 78, 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
mammiféeres du Muséum 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.

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 21, 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 47, 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 espéces de mammiféres. Encyclopédie

120

Méthodique 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 Frangaise 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.

Flannery, T. (1994). Eastern Pygmy-Possum. pp.
54-55. In ‘Possums of the world: a monograph
of the Phalangroidea’. (GEO Productions:
Chatswood).

Forbes-Leith, T.A. 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
indication de leurs moeurs, et de leurs rapports
avec les arts, le commerce et l’agriculture’. (L.
Curmer: Paris). 2, 275-276.

Giebel, C. (1859). Die Saugethiere in zoologischer,
anatomischer und palzeontologischer Beziehung
umfassend dargestellt von C.G. Giebel. Leipzig,
A. Abel. p. 699-700.

Gloger, C.W.L. (1841). Cercartétus. In
Gemeinniitziges 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
modern notes by Joan M. Dixon. (Macmillan
1977: South Melbourne).

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. Zasmanian 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 annotated lists of other vertebrate fauna.
Records of the Queen Victoria Museum 34, 1-
BP.

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 annotated 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.
Launceston. 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.

Gunn, 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 in 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,

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

176.

Harris, J.M (2005). Annotated 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 27, 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, V.V. 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 85, 163-95.

Horner, F.B. (1987). ‘The French reconnaissance:
Baudin 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
Hordern 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.

Julien-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.

121

EARLY NATURAL HISTORY OF CERCARTETUS NANUS

Krefft, G. (1868). ‘Notes on the fauna of Tasmania’.
(F.White Printer: Sydney). pp. 4-5.

Krefft, 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 ‘Dictionnaire 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 générale et
particuliére des mammiferes et des oiseaux
décoverts depuis 1788 jusqu’a nos jours.
Volume IV, Suite des mammiferes. (Baudouin
Freres: Paris). pp. 466-471.

Lesson, R.P. (1838). “Complément de Buffon,
races humaines et mammiféres’, 2™ edition. (P.
Pourrat Freres: Paris). 1, 447-448.

Lesson, R.P. (1842). Famille Phalangistae.

In ‘Nouveau Tableau du Régne 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. W.B.
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

22

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, P.W and Knight, F. (2001). ‘A field
guide to the mammals of Australia’. (Oxford
University Press: Melbourne).

Milne-Edwards, C.R. (1877). Note sur quelques
Mammifeéres noveaux provenant de la Nouvelle-
Guinée (“Dromicia caudata”’). Comptes Rendus
Hebdomadaires des Seances, Academie 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. (W.R.
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 | (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).
[On 2513),

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).
Francois Péron’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 séugethiere; oder,
Synopsis mammalium nach dem Cuvier’schen
system. Solothurn, Jent 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 from
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.

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

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.

Temminck, C.J. (1827). Phalangista nana. In
“Monographies de Mammalogie ou description
de quelques genres de mammiferes, don’t les
espéces ont été observées dans les différens
musées de l’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).

Turnbull, 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.

Tumer, 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. 1B Mammalia’.
(Eds D.W. Walton and B.J. Richardson) pp.

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: Burramyidae) 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. Goldfuss
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. Goldfuss.
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, 99-116.

Wakefield, N.A. (1970). Notes on Australian
pigmy possums (Cercartetus, Phalangeridae,
Marsupialia). Victorian Naturalist 87, 11-18.

Wakefield, N.A. and Warneke, R.M. (1963). Some
revision in Antechinus (Marsupialia) — 1.
Victorian Naturalist 80, 194-219.

Wallace, C. (1984). ‘The lost Australia of Fran¢ois
Péron’ (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). (W.H. 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’. (Bailliére:
London). pp. 307-317.

124

Whinray, J.S. (1971). The present distribution of
some mammals in the Furneaux Group, Bass
Strait. Victorian Naturalist 88, 279-286.

Wood Jones, F. (1925). A new South Australian

dormouse opossum. Jransactions of the Royal
Society of South Australia 49, 96-98..

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, Ostervoldgade 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 1984 (Marsupialia: Macropodinae). Proceedings of the Linnean Society of

New South Wales 127, 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 Hann
(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 of Baringa material from Nelson
Bay was reported by Herrmann (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 P173573) referred to Baringa by
Flannery and Hann (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, more 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 Flannery, 1989
Baringa nelsonensis Flannery and Hann, 1984

AN EARLY PLEISTOCENE WALLABY

Description
Deciduous Premolars:

Eight dP,s are known, three of which are
certainly associated with Baringa nelsonensis
molars (Fig 1j). 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 known, five of which are
certainly associated with B. nelsonensis molars (Fig
1k). 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 dP, 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) lingual view (h) NMV P200702 left I’ labial view, (i) lin-
gual view, (j) NMV P200410 left dP,, dP,, M,, in dentary fragment with associated I, occlu-

sal view, (k) NMV P201155 left associated dP*, dP*, M!? occlusal view. Scale bar =

126

10 mm.

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

K.J. PIPER AND N. HERRMANN

Table 1. Dimensions (mm) of Baringa nelsonensis deciduous premolars. L = length, AW = anterior

width, PW = posterior width.

PT PeRUAsinieiCluc Cusmite LoslTanrmeodaan
ioe SE ep a te i AW ee PW yee le AWE PW)

Specimen 2
NMV P200410 5.4 es)
NMV P200449 Sai] 2.4
NMV P200450 5,8) 2
NMV P200690 - -
NMV P201155 5.5 2.0
NMV P215777 Se) 2.6
NMV P215789 Sal 25)
NMV P215790 5.3 2.4
NMV P200444

NMV P200482

NMV P201155

NMV P215774

NMV P215777 (R)

NMV P215777 (L)

NMV P215966

NMV P216888

cusp by a deep groove. A weakly developed medial
posterior fossette is present (Figs 1d-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 1f-g; Table 2).
Unfortunately none of the I's have been found in

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

Dail

2.6

De)

2.4

2.8

Me

2.8
5.6 3.0 3)
5.8 all 4.1
6.3 3.0 4.1
Sef) Jae) 4.0
6.1 Se) 4.4
6.0 Syms) 4.4
6.5 31.3) 4.0
5.8 Sod) -

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 ofits description, only one moderately
worn I? (NMV P173591 originally identified as I°)
was known for B. nelsonensis, and was not associated
with any other specimens (Flannery and Hann 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 P173591, 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

2a

AN EARLY PLEISTOCENE WALLABY

Table 2. Occlusal length (mm) of Baringa nelsonensis upper incisors. OL = occlusal length, e = esti-

mated.

Co Ql (cone i) IST Scol i
SQ) SP QR PQs
CM Sow voy oul ouwhou gouMo
(SS rt ts pr rt) eS)
Se re I Seettny ern
AY AY Ay On AY AK Ay Au
> FF FF FF FF FF F 2
SSS CON Se MSR SINS Sy iS
Ziel (uw, te teen Coe

Il

OL G2 Onl ON 70) 6:9) eo Gol eA.

12

OL

the posterior end of the occlusal crest present in NVV
P173591 is variably developed in the present sample
(Figs lh-i; Table 2).

Remarks

The dP*s described above differ from NMV
P173573, the isolated dP? originally referred to
Baringa by Flannery and Hann (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
Flannery and Hann’s (1984) description.

The identification of the posterior incisor
as I? rather than I? is based on the following
observations. In all grazer and intermediate grazer-
browser macropodines, I? is relatively elongate
anteroposteriorly and divided into two lobes by a
labial groove, which is positioned approximately

NMV P216035

oO

NMV P216145a
NMV P173591
NMV P200702
NMV P215806
NMV P215810
NMV P215871
NMV P215992b
NMV P216202
NMV P216224

=
NS)

Sal aoe aS 5.4

centrally or towards the posterior (Ride 1957). In
contrast, I? 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 I? 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 ‘BARINGA’
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 P, (Pledge 1992). Our re-examination
of these specimens can confirm the possession of only
the first of the features used by Flannery and Hann
(1984) to diagnose Baringa (i.e. a well-developed
crest on the dentary just ventral 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 B. nelsonensis, and the I, is
relatively narrower dorso-ventrally, and possesses a
more horizontally-inclined wear facet.

Compared to B. nelsonensis, which possesses
only two intermediate cuspules, the P, of cf. Baringa
sp differs in possessing three intermediate cuspules
with more defined associated ridgelets. The lower
deciduous premolar, dP, also differs in lacking the
small anterior and posterolingual cuspules present
in the dP, 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 M, 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, than M, (Flannery and
Hann 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 7.
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 (Flannery 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

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

of the dentary. We believe these features differ
significantly from those of B. 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 I, that is
shallower dorsoventrally, and has a more horizontal
wear facet; shorter and broader anterior cingula on
the lower molars; a longer dP, that lacks the small
anterior cuspule; a larger P, 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 Hann (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 °°
(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.

129

AN EARLY PLEISTOCENE WALLABY

Observations on the stage of eruption and wear
of molars associated with lower incisors supports
Flannery and Hann’s (1984) hypothesis that the
majority of incisor wear occurs after the eruption of
P,, although some wear is seen to occur while dP, 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 Protemnodon,
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, 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

130

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 helpful. 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 Australia 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., Turnbull, 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). A new 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.

Proc. Linn. 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 Australia. 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 Australasia’. (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 Australia. Records of the South
Australian Museum 27, 79-93.

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

APPENDIX

Specimens of Baringa nelsonensis from Nelson
Bay examined and discussed in the text

NMV P173648 right I'; NMV P200659, partial right
I'; NMV P2006835, partial left I'; NMV P216022, 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 and right I'; NV P216035, worn
left I'; NMV P216036, right I'; NMV P216145a, right
I'; NMV P216221; partial left I'; NMV P173591,
right I?; NMV P200702, left I?; NMV P215806, left
I; NMV P215810, left ?; NMV P215811, right I’;
NMV P215992b left I?; NMV P216202, left I?; NUV
P216224, left 2; NMV P200444, left dP? (associated
with left dP?, M'?); NMV P200482, left dP?; NMV
P200490, anterior cusp of left dP?; NMV P215774,
left dP? (in maxilla fragment also containing left dP>);
NMV P215794, right dP?; NMV P215966, right dP?;
NMV P216888, right dP?; NMV P201155, isolated
left dP? and right dP, (associated with isolated left and
right dP?, M!~*, left and right I,, dP, M,_,, unerupted
left and right M,, right P,); NMV P215777, isolated
right and left dP, and left dP, (associated with
isolated left and right I,, isolated right P,, left dP,,
M,_, and unerupted P, and M, in partial dentary, night
M,_, and unerupted M, in partial dentary, isolated left
P? and left dP?, left M'? in maxilla fragment, right
dP?-M! in maxilla fragment); NMV P200410, left
dP, (in dentary fragment with dP,, M,_,, associated
with I,); NMV P200449, right dP,; NMV P200450,
partial left dP,; NMV P200690, partial left dP,; NVV
P215789, right dP,; NMV P215790, right dP,; NVV
P201156, partial dentary containing I,, P,, M, 4

131

oan llaraly ais maranrts eruption and wwe

¥ all

endalei _ ieineiltikd cai Ye presse
we we ut aang eve! bamteaine yeu

pet t Lie kk a? 4
ig tain n UO ROT "9 sg eteIg ONT. VA
fro SOR TEA Vilas 4) tok le irra ita OC TPE

NYE c oilab engl) 2900 Foe YM) idyn bas
(EER sigh ebea (OT Men) frat deRODes

lg erie VY oniyhe bre Retepaodegs
ieee EEG IST ites Se ep eT OO CO MO
fiat! 2heparcd VIM {feu aROo PRM 9 stot
Be Here Gat seyret TSE ARAORE ST
PPI 1 fel eroded Fae it

Op nie ALERT Vest eel eree bed Wiest!

MES Rath COSA MVM Titehentdoe MA ara
horaluners 9b oh ELOY VINES fel eget oA
ie eth onal ot-onis qv Mihe tM liege ey
ae Teecth ot Ay tel Ve eared hans (Re hOes4
“Ey ihsl Gniniatics wala insamgst sllixnm mi) "4b Rel
“Yh Wipe eee TS “ PRA Ob men Severgy Vine

baintoed (BR (68 VV Raa PS idan Reso Meh

tyrls Pha] eae ci saris here Abarighs Bus? Tbyisi

Hasire Low! oth eptwehgis Hes Wet Sha Shs vais
honuinel STTAIGG PAM P| WRIT, , Morinhr beta ffol
Myint fessioewny Tb fia) “bri r Sag feel | baie gdigts
hfe! dt beudeel y Pdghe he viel Sevabee:
Hiya’ 5 ier oko Siwy A teriiven bea’ eg
+ tp Olwpreiels rad ! Ete CIS (0 DAL ae
wet reindct atin ip ve 4b. Qish ipsre &7

il q fot ty

fale OT eld V/A conten wih ainda are

miek OM CHT aaah veel alga
ues Vt HAD wits AKSOOSY VnMivt =) ne "

Mi Voth Aad win ety VROSOVET VR 7b. Rob thier;

- ave ; . _ era od Bi » Thos nie Oar
4 ME OB ri OCR AT YM 5. Thodgheeeat cra
. q ; pale ps Diyin. : ive ee ao a
Mert bE uhtainney” yams hii nee ee 17089
i Tragttta vali
A i
:
I ¥
(’ ca
t ‘
te f Whee, Uy isi
4 i. |
me Th > {
Lhe ‘ &
: iff) Fie ‘ cry) f
*
bs ‘ it p
~ aL i hla > Ths mis: >
‘ t. ? . Tis bad thepial
leitte Cit ¢ sf}
ete ceygveel material, 1 thee minterta!
erary > i Ww ve jersan'ty, eatin)
ms

182

gemini ‘wiltba ine
of this é oni * +h ie
sdaptaataebcemvaiaah Faiutach marine.
\gatiewoud porspe quempenasr nasi tag

sasonisl) thet oat mot) cotayase goisers
38M bileuaEA

fagadngaa? J eh et
Yo werver A (1008) KA, asleshy Bak a
iw estate eye (iitta: <
PL GNP SMa we cena wea ROME Ac
CAD efibgeldnettinsla® desir ViREW \6 bir bao
Copeuhescs tor ihe providing of ibe faciysamd &
{abs ites ara neRgORO LAH EPH YW i .
M leer pre Bheshicaer ry ast ino bente

tienen eran oa
i tha prnoetrre, and fo (ie reve eS HO i
Wit _aenliertialn eh byron ss EPO HY
‘wees fadbtiod (oDbnk Ban emai AOaE 20
onaati 6 Hae UST aR LNT) FT PE OREN.
ln nn oe 7
OAS Dee byhet att retenas A aime E
rere Yar ZAG eased notre madre
SE-@IE@
set ar vinonoxet bas polesteaM (0201) J
sieomoxet bas noitomus, \. ‘Aud anibeqorasen
CREE eget ney en “sonenoqml
+E OV anita ihe mason sortancisteye)
Duce wee 36 mini grays oi Robs.
{oe -nlihdieiiadibaaiates Min 2h eb
olelaihkonan Sone $y os rset deeb oHboaal N
Crowe nd brs tee WRT TORO
raanard Yo aaonrawargefirs Semcon
meee eal HPD BE!
soieuban fees senbsloekiphoneth HORlt oD
Smamene Wile: atest vtotrurtoriycads tot) a
(anbibpran ‘nite ad remtiejowss Ne:
2A SPE EE Hoe ke EyrRh Mila |
nba thfedor nf gitttibes hate noma lov SBRT pith
oleids ree tzanbei uinebidbdéomanansom
M.A basa si i ebals ‘stele tA lock
ovina ——- RED ng ep Bl
Syaire Aocrmodiand
uirsan io. 2 ~ciicbigid Ledigobod ii) (888 Dud
ol .rsbibbieet Si) ofthaiigaibosttosdemib on.
2h) | védetearppelaiee init aadeline bootmgrsi end
8al-i28 “ (senesced ina chest (ag gD ener:
| Aqoithye- Ado? bra yes yore)
Vian vetyudnis sbseted noldassoa8 HRCE IY
» loeunnt gnc sasapiaistT 1b
Aan praernacarrmpinen prints Ss
Mg es 1) BRR TE mde Ne 204), 7
i. (120), Cin ie Teaiurn eran +
ciaDrie! 20s rele. eae eee ine ates '
18, 256-310, 7 :
M, (19834, The vertebra pi
of the Nelson Bay Formutwin, Porlend Vicionae
BSe(fiopoars) thet, bere sep

Nader Burt

b

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 127, 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.

KEY WORDS: 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 from 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 Gisbornian 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 P.L.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

694 695 696 697 698 699
6040

6039

6038

6037 fe

6036 f=

6035

6034

Cainozoic Ryrie Formation

Livingstone Porphyry Foxlow Beds

Black chert & slate

Devonian Granites (fossiliferous)

Colinton Volcanics Unconformity

Cappanana Formation Railway

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. RICKARDS

*Sp10991 194}0 WOAJ APYSI[S SAI JIP WsULA ULI[V.AYSNY UL 194A JJVOIPUT SopIA1D SAodvd sty} UL poqiiosep 10 poinsy JOU = SySl19}SsV
*SOUOZOIG [VQO[s pastusoda1-A[OPIA JO JOS B PUL SaSE}S ULI[BAISNY JSUIUSE pojjO]d syUDUAITO [eUNLY OSe[OYIIJA] 94} JO Sasuv. [eqo[y *[ IqQeL

so

UO0ZOIq

ce

135

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

EASTONIAN GRAPTOLITES FROM MICHELAGO

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

136

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

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

P.L. WILLIAMSON AND R.B. RICKARDS

hematite surrounding the graptolites. Of the hundreds
of graptolites 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.1.
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 fragments 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 the
uppermost circular structure for some further 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.1. 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.

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

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 of ?Climacograptus
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 from 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
from Gisbornian 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

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 /inearis
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 growth 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.
tubuliferus; 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 1873

Type species (by original designation) Graptolithus
flaccidus Hall 1865.

Diagnosis Biramous, occasionally multiramous
stipes, slender, flexed, often slightly reclined, with
simple, long, low-angled thecae mostly without
spines.

Leptograptus flaccidus (Hall, 1865) cf. macer Elles
and Wood 1903
Figures 3a-e

Cit.

1903 Leptograptus flaccidus vat. macer vat. Nov.;
Elles and Wood, pp. 110-111, pl. 15, figs 2a-i.

1934 Leptograptus flaccidus Hall var. macer Elles
and Wood 1903; Ruedemann and Decker, p.

138

306, pl. 40, figs 5-6.

21963 Leptograptus cf. L. flaccidus var. macer Elles
and Wood 1903; Ross and Berry, p. 101, pl. 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 Williams (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
from 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 LZ. 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

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 and L. 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 LZ. 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 ?flaccidus spinifer Elles and Wood
(1903)
Figures 3f-i

21903 Leptograptus flaccidus vat. spinifer var. nov;
Elles and Wood.

Holotype BU1037, figured by Elles and Wood
(1903) plate 14, 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 th1' and th1? 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).

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

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

21867 Didymograpsus flaccidus Hall; Nicholson, pp.
110-111, pl. 7, figs 1-3.

1868 Didymograpsus elegans Carruthers; Carruthers
(pars) pl. 5, figs 8b, c (non figs 8a, d= D.
elegans sensu stricto).

1871 Dicellograpsus morrisi sp. nov.; Hopkinson, p.
5, pl. 1, figs 2a-h.

1876 Dicellograptus morrisi Hopkinson; Lapworth,
pl. 4, fig. 85.

1877 Dicellograptus morrisi Hopkinson; Lapworth,
pl. 7, fig. 6.

1904 Dicellograptus morrisi Hopkinson; Elles &
Wood, pp. 155-157, pl. 21, figs 6a-d, text-figs
98a-e.

1904 Dicellograptus pumilus Lapworth: Elles &
Wood (pars), pl. 21, fig. 3c, (non p. 149, pl. 21,
figs 3a, b, d-f = D. pumilus sensu stricto).

1963 Dicellograptus morrisi Hopkinson; Skoglund,
pp. 31-32, pl. 1, figs 1, 2.

1970 Dicellograptus morrisi Hopkinson; Toghill, pp.
17-18, pl. 7, figs 1-4, text-figs 4d-f.

1976 Dicellograptus morrisi Hopkinson; Erdtmann,
pp. 92-93, pl. 5, figs L/2b, M/6a, pl. 11, fig.
K/2b, pl. 12, fig. K/4.

1982 Dicellograptus morrisi Hopkinson; Williams,
pp. 238-239, figs 7e, f, 8a-c.

21983 Dicellograptus morrisi Hopkinson; Williams
& Bruton, pp. 169-170, figs 10D, 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

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 fragments 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, pl. 7, fig. 3.

1904 Dicellograptus caduceus Lapworth; Elles and
Wood, pp. 161-3, pl. 23, figs 4a-c, text-figs 102
a-c.

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, from 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 thl? 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 from the original

140

descriptions in that the stipes distally seem less than
1 mm in our material. Some of the specimens figured
by Elles and Wood (1903, pl. 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 D. cf. caduceus and Williams
(1983, p. 36) implies, but does not state specifically,
that D. complexus is distinguished from 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 from Australia; D. caduceus has
(see VandenBerg and Cooper, 1992) but the species,
considered globally is rarely recorded. It may be that
further research will recognise more variation and
more species.

Dicellograptus sp.
Figures 5c, d

Material A single specimen, AMF 114903.

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 | 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 ora
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 fragment (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,

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 of D. ornatus.

Family Diplograptidae Lapworth 1873
Genus Climacograptus Hall 1865

Type species Graptolithus bicornis 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,
pl. 2, fig. 48.

1877 Climacograptus scalaris var. caudatus
Lapworth; Lapworth, pl. 6, fig. 34.

1906 Climacograptus caudatus Lapworth; Elles &
Wood, pp. 202-203, pl. 27, figs 7a-e, text-figs
133a-d.

1908 Climacograptus caudatus Lapworth;
Ruedemann, pp. 438-439, pl. 28, figs 17-18,
text-fig 405.

21913 Climacograptus caudatus Lapworth;
Hadding, pp. 49-50, pl. 3, figs 18-19, text-fig.
19.

21934 Climacograptus caudatus Lapworth;
Ruedemann & Decker, p. 319, pl. 43, figs 1-1a.

1947 Climacograptus caudatus Lapworth;
Ruedemann (pars), p. 424, pl. 72, figs 57-65
(non pl. 71, figs 51-52).

21955 Climacograptus caudatus Lapworth; Harris &
Thomas, pp. 38-39, pl. 1, figs 4-6.

1971 Climacograptus caudatus Lapworth; Strachan,
p. 32.

1981 Climacograptus? caudatus Lapworth
1876; Williams, pp. 135-6, pl. 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.

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

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 th1!
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 often 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 | and 2 for this species in Victorian strata,
although VandenBerg and Cooper (1992) give
Gisbornian 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

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. 10B). 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, pl. 2, fig. 49.

1877 Climacograptus scalaris vat. tubuliferus
Lapworth; Lapworth, pl. 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, pl. 27, figs 8a-d, text-figs
134a-c.

1947 Climacograptus tubuliferus Lapworth;
Ruedemann, p. 440, pl. 75, figs 54-56.

21948 Climacograptus styloideus Lapworth;
Henningsmoen, p. 404.

1955 Climacograptus tubuliferus Lapworth; Harris
& Thomas, p. 40, pl. 1, figs 10-12.

1960 Climacograptus tubuliferus Lapworth; Berry,
pPasomplelOlstiges:

1963 Climacograptus tubuliferus Lapworth; Ross &
Berry, p. 132, pl. 10, figs 1,2.

21963 Climacograptus styloideus Elles & Wood;
Skoglund, pp. 38-40, pl. 2, figs 1-4. pl. 3, fig. 3.

1969 Climacograptus tubuliferus Lapworth; Moors,
pp. 268-270, figs 3a-c.

1977 Climacograptus tubuliferus Lapworth; Carter
& Churkin, pp. 23-24, pl. 7, fig. 5.

1982 Climacograptus tubuliferus Lapworth;
Williams, pp. 245-246, figs 1la-n.

1983 Climacograptus tubuliferus Lapworth; Williams
& Bruton, pp. 170-172, figs 12c-e, 15a-n.

1983 Climacograptus tubuliferus Lapworth; Koren’

142

and Sobolveskaya (pars), pp. 139-141, pl. 41,
figs 1-3, (non pl. 40, figs 6-117).

1987 Climacograptus tubuliferus Lapworth, 1876;
Williams, p. 80, figs 4F, H, I, 6G, 70-Q.

21988 Scalarigraptus tubuliferus (Lapworth); Riva,
figs 2i, j (?=Normalograptus normalis).

1989 Normalograptus tubuliferus (Lapworth); Riva
(in Riva and Kettner), pp. 87-89, figs 10a-i, 11a-
es:

1991 Climacograptus tubuliferus (Lapworth, 1876);
Williams, pp. 593-4, pl. 1, figs 2-4, ?5, figs 8A-
GC:

1992 Climacograptus tubuliferus Lapworth;
VandenBerg and Cooper, p. 81.

1992 Normalograptus tubuliferus tubuliferus;
VandenBerg and Cooper, p. 50, fig 10A.

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.

?Climacograptus lanceolatus VandenBerg 1990
Figure 7g

21990 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

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. /anceolatus 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 (Ruedemann 1912)
Figures 7h, i

1906 Climacograptus minimus (Carruthers); Elles
and Wood, p. 191, pl. 27, figs la-g, text-figs
124a-d.

non 1868 Diplograptus minimus sp. nov.;
(Carruthers); p. 74, pl. 5, figs 12a, b.

1912 Diplograptus (Mesograptus) mohawkensis sp.
nov.; Ruedemann; pp. 80-2, pl. 2, figs 18, 19,
text-figs 19, 20.

1947 Diplograptus (Mesograptus) mohawkensis
Ruedemann; Ruedemann, pp. 419-20, pl. 71,
figs 24-6.

1948 Climacograptus cf. minimus (Carruthers);
Henningsmoen, pp. 404-5.

1960 Climacograptus minimus (Carruthers); Berry,
p. 80, pl. 19, fig. 2.

21963 Climacograptus minimus (Carruthers); Ross
and Berry, pp. 125-6, pl. 8, fig. 7.

1964 Climacograptus minimus (Carruthers); Obut
and Sobolevskaya, pp. 57-8, pl. 11, figs 8,9.

1969 Climacograptus minimus (Carruthers); Riva, p.
521, text-figs 3h-j.

non 1969 Climacograptus minimus (Carruthers);
Strachan, p. 191-2, pl. 4, fig. 3, text-figs 4a.

1977 Climacograptus mohawkensis (Ruedemann);
Walters, pp. 937-8, pl. 2, figs f, h, 1.

1982 Climacograptus mohawkensis (Ruedemann);
Williams, pp. 246-7, figs 10c-j.

2002 Climacograptus mohawkensis (Ruedemann
1912); Rickards, pp. 8-9, fig. 3N.

Holotype The specimen figured by Ruedemann,
1947, pl. 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

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

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 from “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.

?Climacograptus uncinatus Keble and Harris 1934
Figures 7c, d

21934 Climacograptus uncinatus sp. nov.; Keble and
Harris pp. 173-4, pl. 20, figs Sa-c.

21972 Climacograptus uncinatus, Keble and Harris
1934; Carter, pp. 48-9, pl. 1, figs 2-7, 10, text-
figs 2J, L-O.

Type specimen A type has never been designated.
Material Only the two specimens figured.

Remarks 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
?C. uncinatus, appear to have almost orthograptid
thecae, as does one of the Keble and Harris originals
(1934, pl. 20, fig. 5a). The pair of proximal spines is
clearest in scalariform views (Fig 7d. herein; Keble
and Harris 1934, pl. 20, figs 5b, c). In our material the
spines are 2.5 mm from the proximal end, but in the
types they are only 1.5 mm from the proximal end.
In this respect our specimens are closer to the Carter
(1972) specimens from Idaho than the specimens
from Victoria. The Idaho specimens are from the
linearis Biozone (approximately Eastonian 3) and the
Victorian specimens from Bolindian 1. There is also
the problem of the relationship, if any, of C? 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 wncinatus 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 1873

Type species Graptolithus quadrimucronatus Hall,

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, pl. 13, figs 1-10.

1876 Diplograptus aculeatus Lapworth; Lapworth,
pl. 2, fig. 44.

1877 Diplograptus quadrimucronatus Hall;
Lapworth, p. 133, pl. 6, fig. 20.

1906 Diplograptus (Orthograptus)
quadrimucronatus (Hall); Hall, T.S. p. 277, pl.
34, figs 10, 11.

1907 Diplograptus (Orthograptus)
quadrimucronatus (Hall); Elles and Wood, pp.
223-4, pl. 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, pl. 78, figs 1-5.

1948 Diplograptus (Orthograptus) quadrimucronatus
(Hall); Henningsmoen, pp. 403-4.

1955 Diplograptus (Orthograptus)
quadrimucronatus (Hall); Harris and Thomas. p.
37, pl. 2, figs 37.

1970 Orthograptus quadrimucronatus (Hall);
Toghill p. 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, pl. 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, pl. 2, figs 1-4, figs
90-q.

1992 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, from the Utica
Shale east of Pointe Bleue, Lake St. John, Quebec as

syntypes.

144

Material Only two specimens, both figured.
Diagnosis Wide rhabdosome with dorsoventral width
in excess of 3 mm within 5 mm of the proximal end
from 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.

Remarks The thecal apertures appear to be not quite
so inturned as in the O. calcaratus groups (see below);
but the presence of spines along the rhabdosome is
sufficient to distinguish O. quadrimucronatus from
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

1876 Diplograptus foliaceus Murchison v.
calcaratus Lapworth; Lapworth pl. 1, fig. 30.

1907 Diplograptus (Orthograptus) calcaratus
Lapworth; Elles and Wood, pp. 239-241, pl. 30,
figs la-c, text-figs 159a-c.

1960 Orthograptus calcaratus; Thomas; pp. 12, 19,
pl. 10, fig. 132.

1992 Orthograptus calcaratus calcaratus
(Lapworth, 1876); VandenBerg and Cooper, p.
82.

2001 Orthograptus calcaratus calcaratus
Lapworth); Rickards et al. p. 82, figs 11H-J.

Holotype Specimen figured by Elles and Wood 1907,
pl. 30, fig. 1b.

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
th1?; thecal apertures very slightly everted proximally
and more or less horizontal distally; thecal spacing 11-
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 th1!
and which th1?. 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

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., 2001).
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 /inearis Biozone, that is from Gisbornian 2 to
Eastonian 3.

Orthograptus calcaratus ?priscus (Elles and Wood
1907)
Figure 8f

21907 Diplograptus (Orthograptus) calcaratus vat.
?priscus var. nov.; Elles and Wood, pp. 244-5,
pl. 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 th1'/th1* 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 fragment 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 from 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 from Australia before, unlike O.

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

c. priscus (see VandenBerg and Cooper 1992), and it
occurs in Gisbornian 2 and Estonian 1. Orthograptus
calcaratus priscus is thought to be earlier, around the
gracilis Biozone (approximately Gisbornian 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.?acutus.

Orthograptus calcaratus cf. vulgatus (Lapworth
1875)
Figures 8c-e

cf. 1907 Diplograptus (Orthograptus) calcaratus
var. vulgatus var. nov.; Elles and Wood, pp. 241-
2, pl. 30, figs 5a-d, text-figs 160a-c.

cf. 1992 Orthograptus calcaratus vulgatus
Lapworth; VandenBerg and Cooper, p. 82, fig.
8M.

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 th1'
and at the base of thl*; 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 th1! is subapertural and the
spine associated with th1? is either at the base of th1?
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
from Gisbornian 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. 1907 Diplograptus (Orthograptus) calcaratus

var. tenuicornis, var. nov.; Elles and Wood, pl.
30, figs 4a-c, text-figs 163a,b.

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 th1' and
one associated with thl*; 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
th1'/th1? 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. Th]? 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 114913,
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 thl* 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 th1*. 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 doubtful 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

1907 Diplograptus (Orthograptus) truncatus vat.
pauperatus var. nov.; Elles & Wood, p. 237, pl.
29, figs Sa-d.

1915 Diplograptus truncatus Lapworth var.
pauperatus Lapworth mscr.; Hadding, p. 15, pl.
2, figs 8-11.

1948 Diplograptus truncatus pauperatus Elles &
Wood; Henningsmoen, p.403.

1963 Orthograptus pauperatus Elles & Wood;

146

Skoglund, pp. 45-46, pl. 1, fig. 11.

1970 Orthograptus truncatus pauperatus Elles &
Wood; Toghill, p. 24, pl. 16, figs 1,2.

1976 Orthograptus amplexicaulis pauperatus Elles
& Wood; Erdtmann, pp. 113-114, pl. 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 21P, 22A-C,
23E.

Type species Not designated according to Strachan
(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; th1' and th1? 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. Th1' has a small mesial spine and thl? 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 Gisbornian 2 to Eastonian 3.

Orthograptus amplexicaulis intermedius (Elles and
Wood 1907)
Figure 9f

1907 Diplograptus (Orthograptus) truncatus vat.
intermedius var. nov.; Elles and Wood, p. 236,
pl. 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 dorsoventral
width of 2.50 mm by the 11" thecal pair and

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). Th1' has a sub-apertural
spine and thl? a spine low on the free ventral wall,
close to the sicula.

Remarks Orthograptus truncatus intermedius was
recorded from Australia 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 famariscus (1) 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 daviesi Williams 1982
Figure 9a

1982 Glyptograptus daviesi sp. nov.; Williams pp.
251-2, figs 14b-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 ventral wall of thl' is relatively short at 0.50
mm compared with that of thl? 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.

Remarks The best specimen is identical to those

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

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, T.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 from 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, P.R., 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), 1-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 Universitat 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 Dicellograptusskiffern i Skane.
Lunds Universitet Arsskrift, new series, 9, 1-91.

147

EASTONIAN GRAPTOLITES FROM MICHELAGO

Hadding, A. 1915. On Glossograptus, Cryptograptus och
trenne dem narstaende graptolitstlakten. Geologiska
FGreningens i Stockholm Férhandlingar, 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 xi. 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
Kullatorp Cove. Jn Waern, B. Thorslund, P. 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. 1871. 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.

148

Lloydell, D.K., Orr, P.J., and Kearns, 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,
SSpp.

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, P.-L. 2001.
Gisbornian (Caradoc) graptolites from New South
Wales, Australia: 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.

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 Vasterg6land 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, P. 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. Jn 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 (D. 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 Linn, 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.

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

Williams, S.H. 1991. Stratigraphy and graptolites of the
Upper Ordovician Point Leamington Formation,
central Newfoundland. Canadian Journal of Earth
Sciences, 28, 581-600.

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.

149

EASTONIAN GRAPTOLITES FROM MICHELAGO

Figure 3 a-e Leptograptus flaccidus cf. macer Elles and Wood, respectively AMF114895, 114938, 114939,
114934, 114892; f-i Leptograptus ?flaccidus spinifer Elles 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 cf. 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.

SZ 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

Figure 7 a, b Climacograptus sp., respectively 114952, 114919; c,d Climacograptus? uncinatus Keble
and Harris, respectively AMF 114950, 114889; e,f Orthograptus quadrimucronatus (J. Hall), respectively
AMF 114916, 114917; g ?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; f Orthograptus calcaratus ?priscus (Elles and Wood), AMF 114937; scale bars 1

mm.

Proc. Linn. Soc. N.S.W., 127, 2006 5S)

EASTONIAN GRAPTOLITES FROM MICHELAGO

Na:

Figure 9a Glyptograptus daviesi Williams, AMF 114946; b,c Orthograptus calcaratus aff. tenuicornis
(Elles and Wood), respectively AMF 114933, 114888; d, e Orthograptus amplexicaulis pauperatus (Elles
and Wood), respectively AMF 114901, 114898; f Orthograptus amplexicaulis intermedius (Elles 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@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 from the overgrazing of
catchments associated with European settlement.

Larger lakes with inflowing streams fill in El Nifio 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

KEY WORDS: biodiversity, El Nifio, lake compartmentilisation, lake migration, lake origins, lake

sedimentation, lunette dunes, saline lakes, spits.

INTRODUCTION

In most hot arid lands, geomorphic processes and
resultant landforms are dominated by wind action on
unconsolidated surfaces (Thomas, 1989). Therefore
depressions and their lakes are likely to owe their
origin to aeolian processes, or at least have their
basins and shorelines modified by wind. Furthermore,
because drainage is often uncoordinated, most lakes
are closed hydrologically (Cole, 1968, 1983), so that
saline waters abound. Lakes fill and dry intermittently
(Williams, 1984), either seasonally or episodically
according to prevailing climate. The extent of
filling is influenced by the interaction between
rainfall, evaporation, lake basin geomorphology and
hydrological character of the catchment. In total, the
geomorphology and hydrology of arid-zone lakes,
particularly if saline, are likely to be distinctive.

In the Australian context, these issues have been

partly explored at the large scale (lake areas > 100 km?
) on Lake Eyre (Kotwicki, 1986) and its predecessor
Lake Dieri (de Vogel et al, 2004), on Lake Victoria in
southwestern New South Wales (Gill, 1973; Lees and
Cook, 1991; Chen, 1992), and on lakes of Salinaland
in Western Australia (Van de Graaf et al., 1977). The
SLEADS program on large salina playas in Australia’s
arid and semi-arid inland (Chivas and Bowler, 1986),
besides its main aim of interpreting past climates from
lake sediments and lunettes, has confirmed the role of
wind in lake basin evolution. Besides these studies on
large salinas, much can also be learnt from comparative
studies on smaller lakes (A < 50 km’, often < 5 km?
) of a confined area where the hydrological pattern
is known. The middle Paroo of northwestern New
South Wales and southwestern Queensland has many
small saline lakes and a few freshwater lakes (Fig. 1)
that salinise as they dry, and moreover hydrological
data covering many years (up to 18) are available for
most. It is the aim this contribution to explore role

SALINE LAKES OF THE MIDDLE PAROO

>z

/ f{ Currawinya

/

: Horseshoe Lake
rasa o ,« Bloodwood
wet Gidgee Lake

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 limnology 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 ‘full’ 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

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

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 (P.and 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).

159

SALINE LAKES OF THE MIDDLE PAROO

800
700
600
500
400
30
20
100

Oo oO

Annual Rainfall (mm)

lind

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 moving

average shown by solid line.

Werewilka Ck.

Benanga Ck.

Pelican
Island

Youlainge Ck. 5°"

/

if
4 7 /beach
4 4

z- “ | ridges
-_-— 7

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).

160

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 ‘lakefull’ 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

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

B.V. TIMMS

88 89 90 91 92 93 94 95 96 97 98 99 00 O01 O02 03 04

Wyara 3

Numalla 2.5

Karatta 8 <1 1 <4

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

1 : f
CO T= = > beach inundated at higher water
North Blue 114 I, 2 <1 31 2 : : &
? ——_ — = levels. Major spits occur at sudden
Mid Blue 61 40 2 2 102 4 ; :
ar ? Eas aCe a - changes in shore _ orientation
a ? - (B on Fig. 5) and in two places
Wombah 2 1 26 2 2 30 2
Re 1S apes oe Sa a ae ee ee these almost occlude two large
lagee
: — a ae Gee backwaters, the Northwest Arm
Ow ;
RDS te eee a ks and a lakelet north of the Public
Oe conecuc a gee Beach (C on Fig. 5). Smaller
Bells Bore x) 131 30 «6171 209 89 69

Figure 4. Comparsions of wet (black line) and dry periods
(blanks) of Paroo lakes. Some salinities (TDS in gL") are given.

in the last 118 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
v
D
CY
SS
areas The Point
Des p_t N
Public a+ ]
Beach | B 0 1km
—EE—,!
D Carwarra Ck

Figure 5. Lake Numulla showing beaches (A),
spits (B), major occluded bays (C), and minor oc-
cluded bays (D).

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

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 corner 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 from 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 S/em in May 2005). Inflowing water is of
very low salinity (<100 pS/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

1).

Lake Yumberarra

This lake is a triangular-shaped, 170 ha in
area and 3.4 m deep when full (Fig. 6). It lies in a
depression in Quaternary alluvium at the edge of the
Paroo floodplain. It is fed by Paroo floodwater via

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 small numbers.

Years 1995-2001 2002 Jul03 & Feb04 Nov03 &Nov04
Conductivities (mS/cm) <3.5 3.6-4.2 6.4 - 9.6 ike} = 417
Number of lake visits n=60 n=12 n=6 n=6
Species

Boeckella triarticulata Thomson XX XXX x x
Calamoecia canberra Bayly x

Calamoecia lucasi Brady XXX XX

Apocyclops dengizus Lepeschkin X XX

Metacyclops sp. x

Mesocyclops cf woutersi Van de Velde XXX
Cletocamptus deitersi Richard x Xx

Diaphanosoma unguiculatum Gurney

Moina australiensis Sars X XX

Moina micrura Kurz XX XX

Bosmina meridionalis Sars r

Daphnia carinata s.|.King r

Ceriodaphnia cornuta Sars r

chydorids (mainly Alona spp.) r r

Heterocypris sp. r x XX
Mytilocypris splendida (Chapman) x 4
Asplanchna sieboldi (Leydig) X X X X
Brachionus calyciflorus Pallas X XXX x xX
Brachionus ibericus Ciros-Perez et al. XX XX
Filinia australiensis Koste Xx

Filinia cf pejleri Hutchinson x

Hexarthra sp. x

Keratella sp. x

Macrobrachium australiense Holthuis XX XX

Cherax destructor Clark r

Cloeon sp. x

Tasmanocoenis tillyardi (Lestage) x

Xanthoagrion erythroneurum Selys x

Diplacoides spp. r

Hemianax papuensis (Burmeister) r r

Hemicordulia tau (Selys) r r

Austrogomphus sp. X

Agraptocorixa eurynome Kirklady XX XX XX XX
Agraptocorixa parvipunctata Hale x x x
Micronecta sp. 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 Jul03 & Feb04 Nov03 &Nov04
Conductivities (mS/cm) <3.5 3.6-4.2 6.4 - 9.6 A oll fh
Number of lake 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 Xx
unident. ceratopogonind larva X Xx
unident. tabanid larva r

Arrenurus sp. x r

Elyais sp. x r

Corbiculina sp. Xx r

Alathyria sp.

>

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 corner totally (at 0.5 m depth) to partially
(at 2 m depth). No enhanced sedimentation in the
main lake was detected. A lunette only 1.5m higher
than the full shoreline flanks the eastern shore.

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

Lake Yumberarra had three filling-drying cycles
during the 17 years of study. It usually fills from 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 freshwater species persist (see
Timms and McDougall, 2005).

163

SALINE LAKES OF THE MIDDLE PAROO

witrom
Carwarra
Ck.

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 channel
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 > | m thick near
the creek mouth (corer could not penetrate
coarser bottom sediments). There is a
broad, low lunette up to | m high abutting
much of the eastern and southeastern
shoreline.

164

Stinking >

Well Creek

Lake Karatta generally fills from
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
fresh, 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 full, but
usually depths are < 1m (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 from ~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

——— ToLake
Yumberarra

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.

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

B.V. TIMMS

Number 10 Ck.

465
e

Figure 8. Bathymetric map of North Blue Lake with
heights of the lunette dune on the eastern and southern
shores and location of cliffs on the western shore. Contour

Overflow to
Mid Blue Lake

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 further
data.

intervals 25 cms. Key: beach ridges— dot and dashed lines.

to south; one cuts off the southeast corner 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 | m contour, a inner
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 corners. 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

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

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 northern shoreline is cliffed (up to 7.5 m high),

3.3m ~ _ from North
% Blue Lak
2.5m ON it pte
1m \\. beach Q
cliffs to 4m \.. ridge
above high \
shoreline

position of
transect in
Fig.10
outer
lunette

ai \. 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 lines.

165

SALINE LAKES OF THE MIDDLE PAROO

‘full’ water level

Height (m)
o2NnN |W Pf U1

500 - 750
Distance across lake (m)

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 from
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.11
& 12). In normal fillings it is160 ha in area and ca 5
cm deep, but in unusually 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 corner is cut off
by a recurved spit; this spit
and adjacent southern beach
are each overlaid with a small
lunette (Fig. 11A). 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

166

the lake’s deepest point to
g the south and halved the

outer normal filling depth (Fig.
ae: lunette 11A & B).
lunette S 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 from 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 from the
northwest. The lake is wedge-shaped with the main
axis SW-NE and the creek entering in a wide channel
at the southeastern corner (Figs 12 & 13). When full,
the lake is 185 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

Figure 11. A, Bathymetric 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.

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

Lower
Bell Lake

Dungarvon — Bloodwood Rd

routes \, ..-”

6. transect
beach ridges
outer

B.V. TIMMS

Horseshoe Lake

Horseshoe Lake (A = 746 ha) has a
flat floor with slightly deeper parts at

the southern end of each arm (Fig.12),
‘Freshwater :
Bloodwood’ 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 ofa
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 corner (Figs 12 and 14),
where there are three ridges increasing

f§ ‘Palaeolake’

‘Bells Bore Salt Lake

& Ms C

goa es in average height landwards. There is

_— no marked vertical differentiation in
the bottom sediments, but those of the

Figure 12. Map showing streams and lakes in the vicinity of the alluvial fan of Bartons Creek are more

terminus of Bells Creek.

silty and give the appearance of recent
deposition. Lake Horseshoe now never

comer of the lake. The rudiments of another beach overflows, but two former pathways
further into the lake and at lower elevation is marked are evident to Lower Bell Lake (Fig. 12). Better
by two low gypseous mounds and slight elevations evidence for a drainage change in this area is seen
in the lake floor as evidenced in the bathymetric nearby at Palaeolake and Freshwater Lake (Fig. 12) —
map (Fig.13 A). The lake basin extends further east, once Palaeolake with its older gypseous lunettes was

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 1m 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).

Bells Ck.

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

0 250 500

Distance from lake 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 inner clayey lunette. Sometimes water
flows on to Palaeolake, creating an unusual situation
of water abutting an older gypeous lunette.

Horseshoe Lake fills from 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 corner to the blind southwestern corner,
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
& 15). When full, 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 southeastern shore. The inner lunette
is of clayey silt and the higher outer lunette is
of gypsum. 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.

Lake Burkanoko

Lake Burkanoko on Wangamanna Station is
oval shaped with a N-S major axis and is 280 ha
in area and ca 40 cm deep when full (Fig. 16A).
It is the terminus for a creek flowing from the

168

750 1000

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
further 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
southwestern 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).

lunette dune

on fF OD ©
Height above shoreline (m)

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.

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.

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

B.V. TIMMS

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 —218 gL
' and median salinity of 115 gL"! (Timms,
1993, 1998b).

Taylors Lake

Taylors Lake on Ballycastle Station
is a relatively deep (1.2 m) hypsosaline
lake in a hollow among dunes (Fig 18),
probably made smaller by an advancing
transgressive dune from the northwest.
The lake is orientated SW — 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

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

: inner
<n
10 ~S

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 18 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.).

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 further 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.

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 innermost 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

170

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 inner 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 Australia, 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
strongest 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
Australia, 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, 1976). The same environmental

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 gypseous 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 full, 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 et al. 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

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

episodic, with only Lake Numalla almost permanent.
This contrasts with saline lakes in southern Australia,
where some are permanent (Timms, 1976; Williams,
1995), 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 full to dry.

Eastern and northern Australia, including the
inland, is affected by the El Nifio/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
full and dry periods over 118 years in Lake Wyara
and the SOI. For the shorter period covéred 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/

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. 11B); 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. 18) 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 friable 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 Bindegolly
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 unknown. 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

72

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
18 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 further 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
corner 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
Nifio 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

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 helpful comments on the manuscript
I am grateful 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).

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

Cole, G.A. (1983). ‘Textbook of Limnology’. (Waceland
Press, Prospect Heights, Illinois).

De Deckker, P. 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). A GIS-based reconstruction of late
Quaternary paleohydrology: Lake Eyre, arid central
Australia. Palaeogeography Palaeoclimatology.
Palaeoecology 204: 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. Western 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. y

Kingsford, R. T. and Porter, J.L. (1994). Waterbirds on
an adjacent freshwater lake and salt lake in arid
Australia. 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 Nifia 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, P.G. (1991). A conceptual model
of lake barrier and compound lunette formation.
Palaeogeography Palaeoclimatology Palaeoecology
84: 271-284.

Madigan, C.T. (1944). ‘Central Australia’. 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.
Corella: 25: 25-31.

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: “Environmental 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 267: 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: 287-313.

Timms, B.V. (1998a). A study of Lake Wyara, an
episodically filled saline lake in southwest
Queensland. Jnternational Journal of Salt Lake
Research 7: 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 limnology 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 Australia, with special
reference to community structure and succession of
invertebrates. Proceedings of the Linnean Society of
New South Wales 123: 193-213.

174

Timms, B.V. (2001c). Wetlands of Currawinya National
Park: Conservation and Management. In “Research
needs for managing a Changed Landscape in the
Hungerford/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 Australia: 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 Australia. Hydrobiologia.

Timms, B.V. and Boulton, A. (2001). Typology of arid-
zone floodplain wetlands of the Paroo River, inland
Australia 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) 22:
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 Australia. Zeitschrift fiir
Geomorphologie 21: 379-400.

Wasson, R.J. and Galloway, R.W. (1986). Sediment yield
in the Barrier Range before and after European
settlement. Rangeland Journal 8: 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, Australia,

a permanent saline lake: Conservation and
management issues. Lakes and Reservoirs: Research
and Management 1:54-64.

Zenkovich, V.P. (1959). On the genesis of cuspate spits
along lagoon shores. Journal of Geology 67: 269-
2iiie

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

Féldvary, 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 127, 175-189.

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 within 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 (Lindstrém,
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 central 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 Synclinortum
in the Karaganda region (composed mainly of
Silurian to Lower Devonian formations). Kovalevsky
(1965) described further 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 further 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 (Regnéll, 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. shigianensis 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 P. turpanensis Deng, 1997, P. 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 tubulli.

Amongst the youngest species referable to
the genus are those known from central 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 Carnic Alps in Austria.

Species originally placed in Plasmopora, such
as P. follis (Milne-Edwards and Haime, 1851) and P.
heliolitoides (Lindstrém, 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). Féldvary (2000) illustrated P. sp. from
central New South Wales. Useful 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

176

from Australia include P. follis (Milne-Edwards and
Haime, 1851), P. heliolitoides (Lindstrém, 1899), P.
gippslandica (Chapman, 1914), and two informally-
designated species illustrated by Hill et al. (1969).
Kaljo and Klaamann (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 distribution 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
Australia, 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).

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:

u =P follis, A = P. heliolitoides, + = P. gippslandica, O = 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, + = P. 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

("]_Girlambone Group metamorphics 4 Volcanics

Lilly

Ut 4

NY
S
N

NY
> :
! A ox
7 ( =
»

of
i

i
CONDOBOLINSA_— >

a =

LEGEND

-Hervey Group

Yarra Yarra Creek Group
Jerula Limestone Member
Inverleith Sandstone

q7 Myamley Sandstone

Trundle Group

Troffs Formation
Connemarra Formation
Derriwong Group
Yarrabandai Formation
Cookeys Plains Formation

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

Table 1 provides a concise summary of the

principal distinguishing features of those species

NOTE: TABLE I AND FIGURES 4-8 ARE ATTHE __ described below.

END OF THE PAPER

Suborder Heliolitina Frech, 1897

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 from the New South Wales Geological Genus Pseudoplasmopora Bondarenko, 1963

Survey are prefixed by MMF, those from the University
of Queensland are designated UQF. Stratigraphical Type species

Superfamily Heliolitoidea Lindstrém, 1876
Family Pseudoplasmoporidae Bondarenko, 1963

and locality details for the Central West area of Pseudoplasmopora conspecta Bondarenko, 1963.
New South Wales are given in Féldvary (2000) and Late Silurian (Ludlow) age, from the top of the

shown herein in Figure 3. The classification follows
Hill (1981) with updated zoological nomenclature as

Isen Suite, southern border of the Karaganda Basin,
Akbastau, Central Kazakhstan [Note that Hill (1981)

recommended by the ICZN (4" edn. 2000). 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)
(Figures 4, A—F; 8, E - F)

Synonymy

Plasmopora follis Milne-Edwards and Haime, 1851,
p. 223, pl. 16, figs. 3, 3a.

Plasmopora follis Lindstrém, 1899, p. 82, pl. 7, figs.
19-20.

Pseudoplasmopora sp. nov. Hill et al. 1969, pl. I],
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 Lindstrém’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 (Regnéll, 1941)

180

has tabularial diameters of 0.8 — 1.2 mm with 2 to
4 tubuli (4 to 6-sided) interposed between tabularia,
whereas P. shigianensis 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
Corner, 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 from Siluro-Devonian strata in the Trundle
— Condobolin district of central western New South
Wales. A species from Queensland, unnamed 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

AMF105567 (cf. Figures 4 and 8), a complete colony,
12x12x6 cm in size, and AMF105568 are from Loc.
XXX; AMF116146, AMF116148, AMF116149 and
AMEF116150 are from Loc. XX,5kmSW ofLoc. 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 (Lindstrém, 1899)
(Figure 5, A—F; 6, A - F)

Synonymy

Plasmopora heliolitoides Lindstrém, 1899, p. 86, pl.
7, figs. 32-33.

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

G.Z. FOLDVARY

Heliolites distans Dun, 1927, p.258, pl. XIX, figs.
3-6.

Heliolites distans var. humewoodensis Dun, 1927, p.
261, pl. XX, figs. 3, 4.

Heliolites distans var. intermedia Dun, 1927, p. 261,
pl. XX, figs. 5, 6.

Heliolites distans var. minuta Dun, 1927, p. 262, pl.
XXI, figs. 1-4.

Plasmopora heliolitoides Jones and Hill, 1940, pl.
IX, figs. 4 and 5; pl. X, figs. 1-4.

Pseudoplasmopora sp. Féldvary, 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 from 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 from Loc. XX,

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

33 km north-north-east of Condobolin, situated east
of the road, and 0.5-1 km east of ‘Moorefield’ Station
(Féldvary, 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 Linné, var. gippslandica, vat.
nov. Chapman, 1914, Pl. 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.10a, b.

Pseudoplasmopora gippslandica (Chapman, 1914),
Ghassan, 1971, p. 593, pls. 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
from Queensland to Victoria. Although documented
from 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 from Lower Devonian beds, and in Victoria
it is known from 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 from 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? from the

181

SILURO-DEVONIAN TABULATE CORALS

Middle Devonian (Eifelian?) of the Carnic 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. II, 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 of P. sp. A.
The only confirmed occurrence of P. sp. Ais 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, pl. III, fig. 7) here re-
illustrated on Figure 8, A— B; and F11587 (Geological

182

Survey of Queensland collection).

Pseudoplasmopora sp. B
(Figure 8, C — D)

Synonymy
Pseudoplasmopora sp. Hill et al. (1969), p. 6, pl. II,
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 — 30).

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, pl. 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 from the same organization for his

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

G.Z. FOLDVARY

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, Ser. 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, pls. 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
from the Tarim Basin and adjacent areas. Acta
Palaeontologica Sinica 36 (Supplement), 116-135.

Deng, Z.-Q. and Zheng, C.-Z. (2000). Tabulatomorphic
corals from 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; Pls. XVIII-XXI.

Féldvary, 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

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

evidence bearing on global Ordovician-Silurian
continental reconstructions. Earth-Science Reviews,
61, 245-307.

Frech, F. (1897). Refarat, J. Wentzel: Zur Kenntniss
der Zoantharia tabulata. Neues Jahrbuch der
Mineralogie, Geologie und Palaeontologie, 1897,
Part 2, 212-214.

Ghassan, K.M. (1971). Korallen aus dem Unterdevon der
Karnischen 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 Australian
Palaeozoic corals. Papers, Department of Geology,
University of Queensland 8 (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 from 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 geliolitode1
Karaesinskogo gornzonta. V. Kn.: Stratigrafiya
nizhnepaleozoishikh i siluriyskikh otlozhenty
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
Corals 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, XXXII, No.1, 1-140.

Milne-Edwards, H. and Haime, J. (1849). Mémoire sur
les polypiers appurtenant aux groupes naturels des
Zoanthaires perforés et des Zoanthaires tabulés.
Académie Science Paris, Comptes Rendus, 29, 257-
263.

183

SILURO-DEVONIAN TABULATE CORALS

Milne-Edwards, H. and Haime, J. (1851). Monographie
des polypiers fossils des terrains paléozoiques.
Museum Histoire Naturales, Paris, Archives 5, 1-502,
pl. 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.mgq.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.

Regnéll, 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 heliolitid corals.
Palaeontological Atlas of Northwest China Sinkiang
Autonomous Region, 39-72. Geological Publishing
House, Being (in Chinese).

Yang, S. et al (1978). Tabulata. In: Palaeontological
Atlas of Southwest China, Guizhou volume, Part
I, Cambrian to Devonian, 161-250. Geological
Publishing House, Beijing (in Chinese).

Table 1

Comparison of parameters distinguishing species of Pseudoplasmopora
discussed in text

Diameter No. of tubuli Tabularial Spacing of Spacing of

of surrounding

tabularium the aureole
P. follis Os0- 161 12
P. heliolitoides 1.0-1.75 12
P. gippslandica 1.2-1.5 12
P. sp. A 1.0-1.1 12
P. sp. B 0.4-0.5 12

184

spacing tabulae diaphragms Septa

percm? inSmm inSmm

25-30 =. 10-15 15-16 absent
5-7 12-16 15 lump
4-5 10-15 15-20 absent
12-15 14-16 8-10 _ blunt
50-60 20-25 16-18 spines

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, 1.55 km ENE of “Meloola” Homestead, about 40 km NNE of Condobolin, 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. Other 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. FOLDVARY

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 of H. distans chosen by Jones and Hill, 1940. E. Longitudinal section
and F. Transverse section of AMF4082 [AM 140 (AM65)], paralectotype of H. distans. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 127, 2006 187

SILURO-DEVONIAN TABULATE CORALS

? 4+ i
+4

a rqiaus

se
6

@,

%4

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), pl. III, 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), pl. III, 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), pl. III, fig. 6.

Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 127, 2006 189

aoaseaainied a lite man

EOSBEIO' to voltoea taalbutigaot 4 bar agitase seseveneil A A ge ovqomentqohuarl &
° of 0 tq (0801) Je to HH yd cele bough ;neiwllé sites dso) brecsre nies

w

a1 mon MVOGAOYT e woltse2 fang epi a bne maitoee peices 2 4 fe plete

Yast) ores mel pa 410% Ie neitose sapildO a nee sattobe oe arene ‘a <Ahat wvoqorean plage
3 .gft LI Iq ,(@801) In to HH yd o@le bowel jaxtli2 noites
a2 I = aed § stare

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 127, 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 transects adjacent to pine plantations, and a second suite of native
species that were reduced in abundance in transects 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

The timber of Pinus radiata (D. Don) is
valuable because of its numerous applications in
manufacturing and construction industries (Sutton
1999). Consequently, extensive areas now support
plantations globally, making P radiata the most
commonly grown conifer in the world (Lavery
and Mead 1998). In Australia, pine plantations are
commonly bordered by native vegetation (Williams
and Wardle 2005), and are frequently established
amongst native vegetation and/or contain areas of
remnant vegetation within the plantation(Lindenmayer
et al. 2002, Lemckert et al. 2005). Of particular
concern is the occurrence of plantations in close
proximity to areas set aside for conservation purposes.
For example, to the west of Sydney, pine plantations
border Blue Mountains National Park, Kanangra-
Boyd N.P., and Jenolan Caves Karst Conservation
Reserve. The close proximity of plantations to native
vegetation is problematic because P. radiata is highly

invasive and can readily escape and establish in native
vegetation, beyond the boundaries of the plantation
(Richardson and Higgins 1998). For instance, prior
to removal of the Jounama Pine Plantation (southern
NSW, Australia), an estimated 16 000 Pinus spp.
individuals were growing in 24 000 ha of adjacent
native vegetation in Kosciusko National Park (Leaver
1983).

The establishment of P radiata within
remnant native vegetation has been linked to the
displacement of native plant species (Richardson et
al. 1994, Richardson and Higgins 1998, Holmes et
al. 2000, Morgan et al. 2000). When pines become
well established, as in plantations, the richness of
bryophytes, vertebrates and invertebrates is reduced
relative to undisturbed native forest (Bonham et al.
2002, Lindenmayer and Hobbs 2004, Parris and
Lindenmayer 2004, Pharo et al. 2004). In this study,
we compare patterns of plant distribution between
patches of remnant vegetation that are adjacent to, and
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 further 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
remnant 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

192

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 transect 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 controls, and temporal trajectories
can be confusing, 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

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+1)
transformed prior to analysis and all other
recommended settings were used (van den Brink and
Ter Braak 1999). Because plants multiply or die, count
data are naturally modelled by proportional changes,
1.e. by a multiplicative model. We used a logarithmic
transformation, 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 et al. 1995).

In a PRC diagram, the horizontal axis represents
the distance along transects of the experiment and the
vertical axis represents the treatment 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

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

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 trail (distance = 0 m) and closest to the
plantations, the disturbed and undisturbed 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) treatment effects at 0 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 strongly driven by differences in
the abundance of P. radiata, which had a strongly
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
penetration of pine litter.

193

VEGETATION RESPONSES TO PINUS RADIATA

@— 8 Undisturbed areas not adjacent ta plantations

O— Disturbed areas adjacent to plantations

Distance (m}

0 10 20 30

0.0 @——__—_____6______# se 9g-

“2 we

03 ¥
4 p=0.19

0.4

Canonical coeficen (Cat

0.5

0.6

a Persoonia acuminata
Monatoce scoparia

Clematis arstata

Stellana pungens

Mois barntchale
sperala scoporea

, Olehandra repens

= Seneciospp.

5 Leueopogen juniperinus

rass 2

Veronica calycine
| - Pos sieberana
A-— —— 1 - Plantego debiis

Dianella sp. 2
_ Gilarciera scandens
Lomandra filfformis
Cassina aculeata
Poranthera micraphyia
Leucopogon lanceoiatus
Lomanara longifolia

=a
S
=]

Species Weight (b.)

-{ *~ Pinus radiata

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 from 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 from 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

194

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 Ter Braak 1999).

Accompanying the increase in pine abundance,
the vegetation of disturbed sites contained a greater
abundance of Lomandra longifolia, Leucopogon

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 abundance 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

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

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 Environmental Sciences
(UTS), the Jenolan Caves Reserve Trust, and the
Linnean 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.

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 African
fynbos vegetation following alien woody plant
clearing and fire: implications for restoration. Austral
Ecology 25, 631-639.

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 frogs. 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 from 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., Cunningham, R.B., Donnelly, 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

196

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 environmental
quality: a novel approach. Marine Ecology-Progress
Series 267, 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 25, 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, P.A. and Hobbs, R.J.

(1994). Pine invasions in the Southern Hemisphere:
determinants of spread and invadability. Journal of
Biogeography 21, 511-527.

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, T.E. (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 C.M. (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, P.J. 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

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

~ eo
- to Mans
u
ef
, are as oo
Mnnnaly ericwe
j MVS | nfs wf tho
ies Gare hd) Read
t ih Walk
(4 mate hater
t reiive PeteractiiGs = an
Tal sure Var ve fhnwelé ied
iM veo Ween HW one
Hw) Resovery of Mouth Atrwat
raion Cllowing alien axand) pi
ilu elinaie tor i ce. Aaairw
Z i
Ve , ahs) é ¢
i cy } ize an)
At ; Lal Vi 7
’ rary h °
wt) SYN) Pin udi-vty tAkE® of)
Ex ivlong mf Progeowraedy of Flee
harden) pp. €92-449 (0 wee
“ ( arr
H eating ats Retell Of
im ion, Kaleo varvomm| Purt
ith) tmipact Maten (Pwortiemadl Parks
Tvu iN « Soth Wale Statnev)
: ‘\ n wn A. CoS) Poetry
tu! roy fi tho Pare? peariatice
1 i vir ; ae
‘ ‘| T 5a
: nm wy : . a
awn! j wa iho
‘ ' ie * Reve
—— ® 4 F 5} J
o PAu hevwrtain wel yew
el ARMA «4 EDT ixige
' t sported
ly u hiva, H MAM). F autia
irahay ARADO TOF Get 4
muro 1998, 181-168
gout 33. Likes ally ‘ i >
f ‘ a ) Pexito 1)
wee Ss 6 hiris
> >) cw
‘ ‘ b pn
‘ , Pr 7
aires Of Ain De # a
?

Pree, KM. and Ladera, Dee (2000).

aheer presiag "Abdegtndl Coe

Pandaoten ya
echeraent ia
lle y

NCTa's 167, i- pit

Pens ATL nse

they orethan ef Pans nactiater piaxtanions 3a sc
canter Avewalin fas eeduced tabites for tres de
Ueculogiem 34, 93+ Te

Prue, 1 .. Lonhenraayer on en aN ton
efFoein orf Lange aisle fag areedtatieny Gra bryan
in ternearne Qarste Jiamal of Appllad Foe
G{o-d2) 5

ManthPt (2004), New Suuth Weles Mom Online Nag
péartnet. shan yd.ugw pov “Viewed 1e@e200S,

Dackerdan, 2.M; Williants, PA, aod Hetts, Be oe
(1904), Plog nvashans wn the Sautherm ont ap

he:
=

dvieanioeante of spread and invodubsfiity, ‘: , iat f
Hingeegy aay st, 51-557, ; ~ =
Hrcherien, DM. end Aiggina, Sil. (i998), Finen :
riwader: in the svutherl beminnbere, d
Biogengraphy of Pfam tah BC Riche
132-446) (Ciaptins dye Laiventity Press: ery ee
Scholes, 6 ad Nowicki, TE 11998). Effects oi
on sr] peypertion dod protean. fo Ceology aa
Bhoyoography of Mime’ (Ed. DM. Bicherdoge
432-448, (Cmbridge Latversizy Preaa mat
berermartio, My, quiny. M. 2. Distel, LA,
x Sa Gah (4, (2006), Liner uly af
woven epeling wlfecled by graziny-intluond Ag
(patients ahing procitnhen gradien. 6
1e7, hte

om. & 29. (1999) Tha need for plan feieine
meaner of radiate pine: New Foraker 17, 95+
wr Teen, CTP, and Sunilauer, P. (1998). “GANG ee

efercuco mama! end user's guide to Crpucal
windows: Sofeware for Canemcel endtontion ¥4

4)." (Micrecomeuter Power, New. York),
van den Bonk, #1 and Tee Aeaak CF (1999), Pr cp
curve’ anulvsis cf § rnehependant tS
orultivariae respoedes ofa binlogibal commun BA:
eipeae. krvivenmenid Tomentogy and Chemistry 2
aH. 148 a
ven des Mink) van Doak, FE. , Ulystra, kv Cw
S31. and Bek, CM (1994) Elect a
low Copereninerniw of the Psicrigewinen
nen) scence oe indoor Sreshwater HCTR ICE,
i» were peg re vi, 4181-2200, © :
vows dan Brink, PL, wan WAjrgsanden, RPA. Lice
WT, Brack. T GM and
t es Se P (Teen. seed olin
E (acto igre “ein

—
+ 4
Me a
i

on:

fi
pom
a
yy

Tf

he

se

(copunec

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 127, 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), acrotretoids (four taxa) and a siphonotretoid.
These are the first Late Silurian linguliformean brachiopods to be documented from eastern Australia.
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 mid-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
Australia 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 (Naylor 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 et al. 1975; Powell et al. 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
Cowhorn 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

200

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
from 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 from 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
(1973) 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 from 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 Corner 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 et al. 1976).

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

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

t+ et tteet+
+++ ete tet
++ tttee¢ +

34°00'S ‘Ati: 34°00'S

Lake
Burragorang

Abercrombjs

Permian Sydney Basin

/ Late Carboniferous Granite Plutons

/ Late Devonian to
34°30; Early Carboniferous

Lambie Group

Early to Middle
Devonian

? Early Silurian to
Early Devonian

? Late Ordovician
to ? Early Silurian

Bindook Porphyry
Complex and equivalents

Taralga Group

Triangle Group

Anticline

Syncline

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

a
le

samp!

LIMEBURNERS FLAT

WOMBEYAN *
CAV :

AVES oe 3
‘ : LWA~ruts
)

Lao
MU 24

a
5S

Tertiary cover
Whipbird Creek Formation

no outcrop and/or mined area

Taralga
Group

Cobra Formation
conglomerate

Triangle

Group Burra Burra Creek Formation

MU Section line
Inferred geological boundaries

Fault

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 lithology 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,

202

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.

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

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

perigee

a Flaggy sandstones and shales

Massive black limestones

Formation

Calcite bands

No outcrop

Dark - grey to black limestone beds
Cobra
Formation 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

RTE

°
°

Burra Burra Conglomerate
Creek

Formation Black, thinly bedded shales

Dr
°
°
°

Sample numbers

Proc. Linn. Soc. N.S.W., 127, 2006 203

204

SILURIAN BRACHIOPODS AND CONODONTS

Sale | se
Metres above base of MU section Sailers =
— Ke) ‘©

Sample Numb

1

(vo)
Lan}
o
©
=a
fe)
ae)
fe)
|
©
Pl
©

osagittella? sp.

Q.
<
~

Rowellella? sp.

Paterula sp.

inguloid gen. et sp. indet. 1
inguloid gen. et sp. indet. 2

chizotreta sp.

— ),— SG
S z
S <
5 o.
S 1S)
= =
S
® e
S ="
n [on
1) aq
(q@)
5
(q?)
oe
DN
io)
p—_ie
5
OQ.
cq?)
>

=? > =p =?

we PN eae ea

4
Opsiconidion sp. cf. O. ephemerus 79
79

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: vv = ventral valve(s); dv = dorsal valve(s); frag = fragment(s); co =
conjoined specimen(s).

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

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

CC Spee
ows] TLL Tei ETT Els ET efefelelel=| | I
eee] Te ele | ielstelst | TT feleleletel | I
ipa a abe P| =| [| ofells fells helidilisbslsde-l-le[>| | fol
Recall | elotelod alfa elspdob Plata] | TT |
pes] & | felstsfel-lea/-[la-| [ale] [iS Els{[-[-]-[>/2/2/-[=[-[--[5|
eee PEEELEEEEE CEE
Po 2S Gon aaa eee GOSSUoRE

Sample Numbers

eS

Metres above base of MU section

arkodina excavata excavata

Oulodus sp. cf. Oulodus elegans

Conodont Taxa
Belodella anomalis
Dapsilodus obliquicostatus
Decoriconus fragilis
Panderodus recurvatus
Panderodus serratus
Panderodus unicostatus
Coryssognathus dubius
O

Kockelella maenniki
Unassigned elements

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

Creek.

a

in Murrui

10n In

Table 2. Distribution and abundance of each conodont species recovered from productive samples along the MU section through the
Cobra Format

205

SILURIAN BRACHIOPODS AND CONODONTS

agreement with the Ludlow age
determination for this part of the
MU section.

Conformably overlying
the Cobra Formation in
Cowhorn, 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

Abbreviation Explanation
IL, valve length
W valve width
H valve height
WI width of pseudointerarea
LI maximum length of pseudointerarea
Fa length of pedicle foramen
Fw 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 surmounting 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
maximum value
minimum value

Table 3. Abbreviations used for measurements (in um) 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:
pl. 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 mid-si/uricus 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

206

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
cannot 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 Dumeéril, 1806

Measurements

Measurements (in pm) 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.

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, 1828
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 um) (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 Pseudolingula quadrata (von
Eichwald, 1829). ‘Lingula’ lewisii differs by being
more rectangular with sharper cardinal angles and
is larger (average length 11.5 mm) (Cherns 1979;
Bassett 1986). ?Wadiglossa 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, 1846
Kosagittella Mergl, 2001

Type species
Kosagittella clara Mergl, 2001.

Kosagittella? sp.
Fig. 4m-o

Figured material
AM F128321 (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

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

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, 2001 from the Lochkovian Lochkov Formation
of the Czech Republic. However, the ventral valve
pseudointerarea of the Murruin Creek specimens
differ from Kosagittella in lacking laterally inclined
propareas (Fig. 40).

Family Zhantellidae Koneva, 1986
Rowellella Wright, 1963

Type species
Rowellella minuta Wright, 1963.
Rowellella? sp.
Fig. 4p-r

Figured material

AM F128322 (Fig. 4p): dorsal? valve; AM
F128323 (Fig. 4q, r): ventral? valve. Both from
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 um) separated by flat interspaces bearing filae
that are initially discontinuos laterally, but become
concentric during later growth stages (Fig. 4q, r). The
post-larval shell microornament of Rowellella cf. R.
lamellosa Popov, 1976 (in Nazarov and Popov 1976)
from Middle Ordovician strata in Sweden (Holmer
1989) consists of similar sets of discontinuous
filae, but these are developed over the entire shell.
Rowellella distincta Bednarezyk and Biernat, 1978
from the lower Arenig of the Holy Cross Mountains
in Poland (Bednarczyk and Biernat 1978) and the
Arenig Klabava Formation of the Czech Republic
(Mergl 1995, 2002), also has a similar post-larval
shell microornament, but has more prominent and
widely spaced concentric lamellae. The post-larval
shell microornament of Rowellella sp. from 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

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
F 128315, 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 F 128317; 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. |. 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; 0, anterior view. p-r. Rowellella? sp. both from sample MU 36 p. Dorsal? valve AM F128322;
interior (scale bar equals 1000 um). q, r. Ventral? valve AM F128323; q, exterior; r, detail of post-larval
shell microornament (scale bar equals 10 um). Unless otherwise mentioned all scale bars equal 100 um.

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 Rowellella?
parvicapera Valentine, Brock and Molloy, 2003 from
the Llandovery-Wenlock Boree Creek Formation
near Orange in central-western New South Wales.
The Murruin Creek specimens, however, lack the
microormment of irregularly arranged wrinkles
possessed by R? parvicapera on the interspaces
between the concentric ridges on the post-larval shell
(Valentine et al. 2003).

Family Paterulidae Cooper, 1956
Paterula Barrande, 1879

Type species
Paterula bohemica Barrande, 1879.
Paterula sp.
Fig. 4g, h

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
1999a). 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 ventral
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 F 128320 (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 wm. The ridges, and
concave interspaces, bear closely spaced, rounded
concentric filae (six per 100 um) (Fig. 41). This is

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

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 um 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. 41-k

Figured material
AM F128319 (Fig. 41-k): dorsal valve, sample
MU 35 (Table 1).

Discussion

The well-developed limbus, ?elongate outline
and lack of post-larval shell pitting (Fig. 41, 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, 1840
Family Discinidae Gray, 1840
Orbiculoidea d’ Orbigny, 1847

Type species
Orbicula forbesii Davidson, 1848.
Orbiculoidea sp.
Fig. 5a-g

Figured material

AM F128324 (Fig. 5a): ventral valve fragment,
sample MU 34; AM F128325 (Fig. 5b, c): dorsal
valve, sample MU 35; AM F128326 (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).

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 F 128325; b, exterior;
c, lateral view. d, e. Dorsal valve AM F 128326; 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
F 128328; h, exterior; i, detail of larval shell. j. Ventral valve AM F128329; interior. k-n Ventral valve AM
F 128330; k, exterior; |, detail of larval shell; m, detail of pedicle track and foramen; n, interior. 0, p. Disci-
nid gen. et sp. indet. 1. o. Dorsal valve AM F128331, sample MU 36; interior. p. Dorsal valve AM F 128332,
sample MU 35; exterior. q-w. Artiotreta longisepta Valentine, Brock and Molloy, 2003. q-t. Dorsal valve AM
F128333, sample MU 32; q, interior; r, exterior; ands, lateral views; t, interiorin lateral view. u-w. Conjoined
valves AM F128334, sample MU 36; u, plan; v, anterior; and w, posterior views. All scale bars equal 100 um.

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. Biernat 1984; 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 (Merg] 2001).

Schizotreta Kutorga, 1848

Type species
Orbicula elliptica Kutorga, 1846.

Schizotreta sp.
Fig. Sh-n

Figured material

AM F128328 (Fig. 5h, i): dorsal valve; AM
F128329 (Fig. 5j): ventral valve; AM F128330 (Fig.
5k-n): ventral valve. All from sample MU 36 (Table

1).

Discussion

Both valves of this species from 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
uum) that become more prominent toward the valve
margins (Fig. 5h, k). The large larval shell, located
submarginally in the ventral valve (averaging 438
um long; 500 um wide) and marginally in the dorsal
valve (averaging 354 um long; 399 um 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

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

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 from 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). Biernat (1984) assigned a single dorsal? valve
fragment 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. from 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. 50, p

Figured material

AM F128331 (Fig. 50): 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 um long; 500 um 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)-

PA

SILURIAN BRACHIOPODS AND CONODONTS

Order Acrotretida Kuhn, 1949
Superfamily Acrotretoidea Schuchert, 1893
Family Scaphelasmatidae Rowell, 1965
Artiotreta \reland, 1961

Type species
Artiotreta parva Ireland, 1961.

Artiotreta longisepta Valentine, Brock and Molloy,
2003
Figs 5q-w, 6a-d

Synonymy
2003 Artiotreta longisepta sp. nov. Valentine,
Brock and Molloy, p. 314; pl. 2, figs 9-18.

Description
See Valentine et al. (2003:314).

Figured material

AM F128333 (Fig. 5q-t): dorsal valve, sample
MU 32; AM F128334 (Figs Su-w, 6a): conjoined
valves; AM F128335 (Fig. 6b-d): dorsal valve. All
from 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 Merg], 2001 from the Llandovery
Zelkovice and Wenlock Motol formations of
the Czech Republic, has a similar dorsal valve
outline to A. /ongisepta, although its anterior margin
tends to be more rounded. The median septum of A.
krizi is also shorter, arising around valve midlength
(Mergl 2001:33, pl. 28, fig. 3). Artiotreta krizi attains
a larger maximum size than A. longisepta (up to 1100
um 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 um long; 708
uum wide; Table 4) than A. parva (averaging 400 um

Table 4. Artiotreta longisepta Valentine, Brock and Molloy, 2003, dorsal valve

dimensions (in tm) and ratios.

Artiotreta longisepta Valentine, Brock and Molloy, 2003,

dorsal valve dimensions (um) and ratios

L WwW LI WI
N V7 43 17 15
MEAN 533.8 694.8 40.1 255.8
SD 106.32 140.85 10.14 46.74
MIN 375 400 18.75 150
MAX 675 975 SH) 23
L/W LI/WI WI/W_ LS/L
N 17 15 13 15
MEAN 83.6% 15.5% 43.4% 88.4%
SD 0.08 0.05 0.10 0.04
MIN 65.8% 9.1% 26.9% 81.3%
MAX 98.2% 28.6% 62.5% 93.8%

22

LS MHS_ BS LP WP
18 42 16 17 18
475.0) Wo405 245.35 150s 8 Wo:
89.22 63.03 32.56 20.48 26.94
325 PPS eae — PAR) MAS) 150
600 287.5 325 175 Jie)
BS/L LP/WP LP/L WP/W
12 17 12 14
48.4% 85.8% 30.0% 29.9%
0.09 012 0.07 0.08
36.7% 70.0% 19.2% 18.9%
66.7% 116.7% 38.1% 43.8%

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 Molloy, 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 um); d, detail of larval shell microornament (scale bar equals 10 pum). 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 um). 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. 0, p. Dorsal valve para-
type AM F128341; o, interior; p. lateral view. q, r. Dorsal valve paratype AM F 128342; 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 um). Unless otherwise mentioned all scale bars equal 100 pm.

Proc. Linn. Soc. N.S.W., 127, 2006 213

SILURIAN BRACHIOPODS AND CONODONTS

long; 500 um wide) (Ireland 1961:1138).

von Bitter and Ludvigsen (1979) documented
two sizes of larval shell pits in A. parva—a
larger set (3-6 um in diameter) with no cross-cutting
relationships and a smaller set (~0.3 um 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 uum in diameter) with none to
one (occasionally two) orders of cross-cutting and a
smaller set (0.2-1 um in diameter) (Fig. 6d). A smaller
set of larval shell pits has not been documented in A.
krizi.

Family Torynelasmatidae Rowell, 1965
Acrotretella Ireland, 1961

Type species
Acrotretella 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
surmounting 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 (Biernat and Harper 1999, Mergl
2001 and Valentine et al. 2003) have defined species
of Acrotretella based upon the presence or absence of
lateral processes (sensu Biernat 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 handful
of specimens (Table 5). A sufficient number of
specimens assignable to Acrotretella dizeugosa sp.
nov. (36 ventral valves and 39 dorsal valves; Table

214

2) have been recovered from the Cobra Formation to
allow the first detailed ontogenetic investigation of
Acrotretella.

The dorsal valve ontogeny of A. 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 of A.
dizeugosa (413-725 um long; 488-788 um 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 um wide. A single dorsal valve of A.
dizeugosa from sample MU 31 (628.6 m above the
base of the section) (275 um long; 288 um 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 um 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 1980:75) are developed
along the anterior margin of the median septum during
the second dorsal valve growth stage (563-763 um
long; 688-838 um wide) (Fig. 7). The median septum
of this growth stage is higher than in the proceeding
stage, averaging 144 um 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 um long;
750-1325 um wide) (Fig. 7). The lateral processes
originate around valve midlength and are initially
developed as low, short, anteriorly divergent rods

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

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

215

91C ADVd NO GHNNILNOO S$ ATEaVL

1002 ‘13191

319 -
(1002) 119 Z 0 URIAOYYIOT — tJ Qnday yooezD ‘auoysou] soy pIvIdasi.4) DIJ2Ja4JOLIP

BITesNy “(suoljoes [YOR

eyep ‘qndun =—=M
( ae ) sa * - Z 0 yoo[ua AA pue g/OSq) UonjeUuL0,; s10U010g pue
a. (uoN99S YAM) UOHeULIO,] Yoo1D so10g
i On ie IEEE SEROTEC CU) EE) Se
(€00Z) Te 19 Lcll = F 6 0 d0[U9 AA elyjensny “(uonses Wg) V ‘ds vjjajaso419p
ourjuaye A, 006 =I cee UONCULIO,J YOO 9010g
€96=M yoojue Bljeysny ‘(UoT}Oes gq) UOTeULIO
souor-ue - ; : ;
(6L61) fed 808 =I ul e - AIQAOpury’T dIOUDIOG - UOTJVULIOY YIO1ID 9010g
OT LLL
(6661) uea[DW LEOl =™ é (3x9) 998) “Ao
pue 13 919 =| I 0 Isysy puejos] SUuOj}SoUN’] Sreply ‘ds 19 ‘uo8 pnonoioy
6661 Jodiey pue
(6661) tedieyH 16s =™ Areyuowise.y ;
pue qeworg opm a6) € I UMAUP]'T puvjog ‘asi[seuAs one g yeUlolg sisuaidppjo3
DIJ2JaAJOLIP
EEN SSS
Ap AA Ap AA
0UdIAJOY ,(UIN) SUOISUSUNIpP OSeIOAY [PIIO}VU 9] qQuTIeAY o3V Ayljeo0'T solsedg

SSod0Id [e1o}e] YIM D]JaJa4jO/Ip

a SSS

aIUdAIJO1 SUIAULCUIOIE 9Y} Ul ‘[eLIgjIVU Pozv.AZSN][I UO 10 “USAIS SUOISUZWIP UO peseq WIN UT) SUOISUdWIp
ISBADAV, “YIPIM = M SY}SUI] = | S(S)oATVA [USIOP = AP S(S)aATVA [VAQUDA = AA :SUONVIAIIQGY ‘sossad0ad [v.19}e]
NOY PUL YBIM YZajajOsJIp JO SaId9ds 10} JEP [VUOISUIUTIP pUL [eL1o}VU sIqQuUIIVAL ‘ase ‘AQITLIOT “s a1qQUI,

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

SILURIAN BRACHIOPODS AND CONODONTS

Table 5 continued

Acrotretella without lateral processes

Species Locality Age Available material Average dimensions (um)* Reference
VV dv VV dv
: 1= 439
Acrotretella sp. Dalby Limestone, Sweden Caradoc 0 2 - w= 509 Holmer (1989)
Acrotretella sp. Mayatas Formation, Kazakhstan Caradoc 0 2 - : a ie 0 Popov (2000)
Acrotretella sp. a Bestrop Limestone, Sweden Ashgill 3 2 pei Lae Holmer (1986)
4 w = 480 w = 520
Quarry Creek Limestone, Bridge Creek is 1= 558 Bischoff
Limestone and Cobbler's Creek Llandovery 1 6 fees 50D) (anpubeante)
Limestone (E-E' section), Australia 6 wy ee Bit
ease. Boree Creek Formation Llandovery - 1= 648 = SI Valentine
Il drid; ‘
Achoiretelia” govdridgel (BM section), Australia Wenlock He: w= 859 w=914 et al. (2003)
Walemtine Brock Boree Creek Formation - Borenore Llandovery - 1= 730 l=799
and Molloy, 2003 Formation (B section), Australia Wenlock 26 oH w = 952 w = 937 PeansJones (122)
Sel BOrENE Formation (OSC ARE ~ WEHISEK = LEST ee eae.” * RED,
ee oc a w=1075 w=1113 unpub. data)

BOR/1 sections), Australia

Chimney Hill Limestone, Hunton 1= 660 1 = 626

Acrotretella siluriana Formation, Oklahoma, USA Menlock 3 te w = 782 w = 689 declan eel)
Ireland, 1961 2 d 1= 843 1= 828

Motol Formation, Czech Republic Wenlock 5 10 w= 1030 eer Mergl (2001)

Acrotretella sp. A Lynore Limestone, Australia Pridoli 0 2 - es Klyza (1997)
Acrotretella spinosa ’ : eee = 435) 1= 808

Za 1(@2
Mergl, 2001 Pozary Formation, Czech Republic Pridoli 6 20 = fe IG w= 939 Mergl (2001)
: : ee 1= 808 Farrell
Acrotretella sp. Camelford Limestone, Australia Pridoli 0 6 - w= 937 (unphodcata)

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

216

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

1400-4
mn @
1200-
1000— @
nly :1008 Sp al
=
5 4 Ors
Z, -
600 Rolain
V.
‘ ce
oY
Sie
400
Tet Aa
200-4
Za | | | | J ]
0 200 400 600 800 1000 1200 1400
WIDTH (um)

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);
4 = dorsal valve growth stage 2 (n = 4);
V = unknown dorsal valve growth stage (n = 7)*;

CJ = 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

2G,

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 pm 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 Pridoli Formation of the Czech
Republic (see Mergl 2001:pl. 26, figs 4-6).

The final dorsal valve growth stage of A. dizeugosa
(1000-1275 wm long; 1025-1500 um 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 growth stage were recovered from
Murruin Creek. The lateral processes of the smaller
of these specimens (1000 um long; 1025 um 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 um long; 1500 um 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 of 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

Sample Numbers
Acrotretella dizeugosa

entral valves

218

final dorsal valve growth stage is the initiation of
folding in the median septum, with up to two folds
being developed. Biernat (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 from Acrotretella goldapiensis Biernat and
Harper, 1999 from the Llanvirn 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
Molloy, 2003 from the Llandovery-Wenlock Boree
Creek Formation near Orange in central-western New
South Wales (see species discussion for 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.

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

LENGTH (um)

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

1400-4
1200
@
= fee ol
10004 e
800— O al @ fu
" . ye
one U
6004
8
| +
+
400—
+
200—
Me Mie Seis ase eecoe ie eee
0 200 400 600 800 1000 1200 1400
WIDTH (um)

Figure 8. Length versus width for ventral valves and each dorsal valve growth stage
of Acrotretella goodridgei Valentine, Brock and Molloy, 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);

4 = dorsal valve growth stage 2 (n = 12);

V = unknown dorsal valve growth stage (n = 20)*;
L) = 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

209

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 um). c-n. Opsiconidion ephemerus (Mergl,
1982) all from sample MU 32 unless otherwise mentioned. c, d. Dorsal valve AM F 128348, 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 F 128350; 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. 0-q. Opsiconidion sp. 0. 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 F 128355, sample MU 36; r, exterior; s, detail of spines (scale bar equals 10 xm). t. Dor-
sal valve AM F128356, sample MU 34; interior. Unless otherwise mentioned all scale bars equal 100 um.

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 p.m) and ratios.

Acrotretella dizeugosa sp. nov., ventral valve dimensions (um) and ratios

L W H IE Fa Fw M HP 1} WP
N 12 15 15 12 17 18 15 17 19 20
MEAN 505.2 655.0 3442 854 368 33.0 117.5 846 1643 190.6
SD 182.35 240.99 171.45 61.67 853 9.05 62.11 2634 20.55 31.64
MIN 350 375 125 Bia 25 25 50 50 125 150
MAX 912.5 1150 675 DIS S025 50 275 150 22 O2
L/W HL ~~ 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 O09. O13 O10 007 O12, O12 O15 O11
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. nov., dorsal valve dimensions (um) and ratios
L W et WI OSP LSP WSP LS MHS LP WP
N 22 31 Di, 19 29 19 28 24 13 29 31
MEAN 678.4 763.3 50.7 388.2 234.5 375 81 592.2 1106 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
MINGuae2 ome 287-500 12:5 100 187.5 250 50 350 50 125h24n 12:5
MAX 1275 1500 125 725 275 575 150 1075 187.5 200 200
L/W LI/WI_ WI/W_ LS/L_ LSP/L_ OSP/L LSP/LS_ LP/L WP/W_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 0.06 007 +®.007 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 Diagnosis

Holotype: AM F128345 (Fig. 6u-w): dorsal valve,
sample MU 32. Figured paratypes: AM F 128336 (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 F128341 (Fig. 60, p): dorsal valve; AM
F128342 (Fig. 6q, r): dorsal valve; AM F128343 (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 (?siluricus
Zone), upper part of the Cobra Formation cropping
out in Murruin Creek (Fig. 2), Taralga Group,
southeastern New South Wales, Australia (Fig. 1).

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

A species of Acrotretella with numerous, closely
spaced growth lamellae on the ventral valve (eight
per 100 um), but more widely spaced on the dorsal
valve (three to five per 100 um). 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.

pis)

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 um long,
191 um wide, 85 um high. Foramen confined to larval
shell, subcircular, averaging 37 um long and 33 um
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 um in
diameter. Post-larval shell ornament of well-defined,
closely spaced, continuous concentric lamellae
(eight per 100 um) 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 um long and 161 um 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 microornament 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 um) 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.

2

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 um) on the ventral 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 (1991)
referred to this specimen as a new, but unnamed,
acrotretid genus, these features suggest assignment
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

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 A. 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 of 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 um 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 Biernatidae Holmer, 1989
Opsiconidion Ludvigsen, 1974

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. Biernat,
p. 97; pl. 26, figs la-c, 2; pl. 27, fig. la-e;
pl. 28, figs la-c, 2a, b, 3; pl. 29, figs 2a, b,
3; pl. 30, figs 1, 2, 3a, b; pl. 31, figs la-c, 2.

2003 Opsiconidion ephemerus (Merg});
Williams; pl. 2, fig. 2.

Description
See Mergl (1982:115).

Figured material
AM F128348 (Fig. 9c, d): dorsal valve, sample

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

MU 35; AM F128349 (Fig. 9e, f): dorsal valve; AM
F128350 (Fig. 9g, h): dorsal valve; AM F128351 (Fig.
91, 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, 1, J). These features are identical to O. ephemerus
(Mergl 1982, 2001) and O. podlasiensis from the
Wenlock Podlasie Depression of Poland (Biernat
1984). Biernat (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, 1, j)
and Czech material (see Mergl 1982:pl. 1, figs 5, 6,
8-11).

The dorsal valve holotype of O. ephemerus is
700 um long and 700 um wide, and Mergl (1982:116)
noted dimensional uniformity within his population.
Biernat (1984:97) listed the dimensions of the dorsal
valve holotype of O. podlasiensis as 690 um long and
840 um wide. On average, the Australian material is
smaller, only 553 um long and 622 um 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 um in height. In
comparison, the most complete ventral valve of the
Australian material is only 475 um high and not as
strongly conical (Fig. 9k-n).

The larval shell microornament of O. ephemerus
and O. podlasiensis consists of circular, flat-bottomed
pits (3-6 um in diameter) with few, or no, cross-
cutting relationships (Fig. 11d). 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 um in diameter) located on the smooth,
level areas between the larger pits. No evidence of
a smaller set of pits were observed in the Murruin
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

223

SILURIAN BRACHIOPODS AND CONODONTS

Table 8. Opsiconidion ephemerus (Mergl, 1982), ventral and dorsal valve dimensions

(in pm) and ratios.

Opsiconidion epmemerus (Mergl, 1982),
ventral valve dimensions (um) and ratios

Fa Fw HP LP

N 17 17 20 20
MEAN) 29.8 29.8 147.4 159.4
SD 7.82 7.82 26.65 18.97

MIN 18.75 18.75 110 125
MAX 50 50 187.5 187.5

WP LP/WP_ HP/LP
18 17 19
174.6 91.9% 92.4%

20.4 0.13 0.15
137.5 71.4% 62.9%
225 115.4% 115.4%

Opsiconidion ephemerus (Mergl, 1982), dorsal valve dimensions (um) and ratios

L W LI WI

N 54 58 53 54
MEAN 553.2 622.0 30.7 214.1
SD 94.86 112.64 9.90 46.39
MIN} 9287-55 362:5.01 12 Sine LIS

MAX — 750 850 50 325
L/W LI/WI_ WI/W_ LS/L

N 49 49 42 48
MEAN 89.8% 14.8% 34.0% 86.2%

SD 0.09 0.05 0.07 0.05
MIN 71.9% 6.3% 24.0% 65.2%
MAX 111.6% 28.6% 55.3% 95.2%

(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) from 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
Molloy, 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. 90-q

Synonymy
cf. 1999 Opsiconidion sp. Cockle; pl. 5, fig. 15.

224

LS MHS BS EP WP
51 16 61 62 61
473.3 204.7 85.7 180.4 195.1
De PIT Peas we See bye SHES) | Baye)s)
250 WS 50 100 150
Sie) PASTY PS) 215 250
BS/L_ LP/WP LP/L_ WP/W
47 59 48 47
16.5% 92.9% 344% 32.8%
0.04 0.12 0.08 0.07
9.6% 57.1% 19.1% 17.4%
30.4% 121.4% 65.2% 62.1%

cf. 2003 Opsiconidion sp. A Valentine, Brock and
Molloy, p. 317; pl. 3, figs 16, 17.

Figured material

AM F128353 (Fig. 90): dorsal valve, sample
MU 31; AM F128354 (Fig. 9p, q): dorsal valve,
sample MU 36 (Table 1).

Discussion

The Murruin Creek specimens differ from most
Opsiconidion by their transversely elliptical dorsal
valve outline (Fig. 90, 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.
90). 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
from both localities to determine if these differences
are significant. Mergl (2001) also documented an
elliptical Opsiconidion, Opsiconidion sp. A, from

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. 90, 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, 1848
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 F128356 (Fig. 9t): dorsal valve,
sample MU 34 (Table 1).

Discussion

The Murruin Creek siphonotretid differs from
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.
from 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 Belodella Ethington, 1959

Type species
Belodus devonicus Stauffer, 1940.

Belodella anomalis Cooper, 1974
Fig. 10a-i

Synonymy
See Farrell (2004:947) plus the following:
1993 Belodella sp. aff. B. anomalis Cooper;
Simpson et al., p. 153; fig. 4J.

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

Figured material

AM F128283 (Fig 10a): Sb element; AM F128284
(Fig. 10b): Sb element; AM F128285 (Fig. 10c): Sb
element; AM F128286 (Fig. 10d): Sd element; AM
F128287 (Fig. 10e): Sc element; AM F128288 (Fig.
10f): t element; AM F128289 (Fig. 10g): fragment of
?t element; AM F128290 (Fig. 10h): M element; AM
F128291 (Fig. 101): 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 B. anomalis
(Fig. 10h, i) is more strongly curved than the M
element of Belodella resima (see Mawson et al. 1995:
pl. 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 eastern 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
Cordylodus? 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.; pl. 2, figs U-W; pl. 4, figs F, G.

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-l. 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, 0. Panderodus recurvatus (Rhodes, 1953), both from sample MU 34. n. Sc element AM
F128296; lateral view. 0. 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 um.

F128304 (Fig. 11f): partly preserved Sb element.

Description
Both from sample MU 34 (Table 2).

See Miller and Aldridge (1993:246).

Figured material Discussion —
AM F128303 (Fig. Ile): Pa element; AM The partially preserved Sb element from Murruin
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. 0. Pa element AM F128313; oblique upper view. All scale bars equal 100 um.

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

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

‘dished’ area (Fig. 11f). 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. 1le) (Miller and
Aldridge 1993).

227

SILURIAN BRACHIOPODS AND CONODONTS

Genus Dapsilodus Cooper, 1976

Type species
Distacodus obliquicostatus Branson and Mehl,
1933.

Dapsilodus obliquicostatus (Branson and Mehl, 1933)
Fig. 10)-1

Synonymy

See Armstrong (1990:70), plus the following:

1990 Dapsilodus obliquicostatus (Branson and
Mehl) Uyeno, p. 98; pl. 2, figs 11-16.

21992 Dapsilodus sp. Barrick and Klapper, p. 44;
pl. 2, fig. 2.

1994 Dapsilodus obliquicostatus (Branson and
Mehl); Sarmiento et al.; pl. 1, figs 1, 6.

1999 Dapsilodus obliquicostatus (Branson and
Mehl); Cockle, p. 119; pl. 4, figs 13-19.

Description
See Cooper (1976:212).

Figured material

AM F128292 (Fig. 10j): M element; AM F128293
(Fig. 10k): M element, sample MU 38; AM F128294
(Fig. 101): M element. All from sample MU 37 unless
otherwise mentioned (Table 2).

Discussion

It has not been possible to separate the Sb and Sc
elements from 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 Decoriconus 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).

228

Figured material
AM F128295 (Fig. 10m): Sc element, sample
MU 34 (Table 2).

Discussion

The Sc elements of D. fragilis from 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 Kockelella Walliser, 1957

Type species
Kockelella variabilis Walliser, 1957.

Kockelella maenniki Serpagli and Corradini, 1998
Fig. 11n, 0

Synonymy
See Serpagli and Corridini (1999:284).

Description
See Serpagli and Corradini (1999:286).

Figured material

AM F128312 (Fig. 11n): Sc element; AM
F128313 (Fig. 110): Pa element. Both from sample
MU 34 (Table 2).

Discussion

The laterally compressed Pa element of this
taxon from 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 fused to the cusp (Fig.
11n). 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. 11n).

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

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

J.L. VALENTINE, D.J. COLE AND A.J. SIMPSON

before the close of the si/uricus 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 Stauffer, 1930.

Oulodus sp. cf. Oulodus elegans (Walliser, 1964)
Fig. 11g, h

Figured material

AM F128305 (Fig. 11g): Sb element; AM
F128306 (Fig. 11h): 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. 11g). The Sa elements are
bilaterally symmetrical about the lateral processes
which possess four denticles (Fig. 11h).

Genus Ozarkodina Branson and Mehl, 1933

Type species
Ozarkodina typica Branson and Mehl, 1933.

Ozarkodina excavata excavata (Branson and Mehl,
1933)
Fig. 11li-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 F 128307 (Fig. 111): M element, sample MU
38; AM F128308 (Fig. 11j): Sa element; AM F128309
(Fig. 11k): Sb element; AM F128311 (Fig. 111): Sc

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

element; AM F128310 (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. 111),
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, 0).

Genus Panderodus Ethington, 1959

Type Species
Paltodus unicostatus Branson and Mehl, 1933.

Panderodus recurvatus (Rhodes, 1953)
Fig. 10n, 0

Synonymy

See Simpson and Talent (1995:117) and Farrell
(2003:122), plus the following:

1995 Panderodus recurvatus (Rhodes);
Colquhoun, p. 354; pl. 3, fig. 4.

1999 Panderodus recurvatus (Rhodes); Cockle,
p. 120; pl. 5, figs 9-14.

2002 Panderodus recurvatus (Rhodes); Aldridge;
pl. 4, figs 4-7.

2002 Panderodus recurvatus (Rhodes); Talent et
al.; pl. 2, figs J, K.

2002 Panderodus recurvatus (Rhodes); Zhang
and Barnes, p. 31; figs 16.1-16.27.

2004 Panderodus recurvatus (Rhodes); Farrell, p.
958; pl. 3, figs 9, 12, 13.

Description
See Barrick (1977:54).

Figured material

AM F128296 (Fig. 10n): Sc element; AM
F128297 (Fig. 100): 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

229

SILURIAN BRACHIOPODS AND CONODONTS

possess a longitudinal groove developed along the
middle to posterior portion of one lateral surface (Fig.
10n, 0).

Panderodus serratus Rexroad, 1967
Fig. llc, d
Synonymy
1997 Panderodus serratus Rexroad; Jeppsson, p.
107; fig. 7.4.

Description
See Jeppsson (1997:107).

Figured material

AM F128299 (Fig. 1lc): Sc element; AM
F128298 (Fig. 11d): Sb element. Both from sample
MU 34 (Table 2).

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. 11c) and one
interpreted as a Sb element (Fig. 11d).

Panderodus unicostatus (Branson and Mehl, 1933)
Figs 10p; lla, b

Synonymy

See Simpson and Talent (1995:118) and Farrell
(2004:959), plus the following:

1997 Panderodus unicostatus (Branson and
Mehl); Jeppsson, p. 107; fig. 7, 7.3.

1999 Panderodus unicostatus (Branson and
Mehl); Cockle, p. 120; pl. 5, figs 1-8.

2002 Panderodus unicostatus (Branson and
Mehl); Aldridge; pl. 4, figs 8-17.

2002 Panderodus unicostatus (Branson and
Mehl); Talent et al.; pl. 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

AM F 128302 (Fig. 10p): M element; AM F128300
(Fig. lla): Sa element; AM F128301 (Fig. 11b): M
element. All from sample MU 34 (Table 2).

230

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 internal 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
from Murruin Creek (Table 1), many are broken at, or
near the basal cavity termination.

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 from 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 from 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 67, 15-28.

Armstrong, H.A. (1990). Conodonts from the Upper
Ordovician-Lower Silurian carbonate platform
of North Greenland. Gronlands Geologiske
Undersggelse Bulletin 159, 1-151.

Barrande, J. (1879). ‘Systeme Silurien du centre de la
Bohéme. Jére partie. Recherches paléontologiques,
vol. 5. Classe des Mollusques. Ordre des
Brachiopodes.’ (Published by the author: Prague and
Paris).

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-Early
Devonian conodonts from the Hunton Group (Upper
Henryhouse, Haragan, and Bois d’Arc Formations),
south-central Oklahoma. Oklahoma Geological
Survey Bulletin 145, 19-65.

Bassett, M.G. (1986). Brachiopodes inarticules.
Biostratigraphie du Paleozoique 3, 85-96.

Bednarczyk, W. and Biernat, G. (1978). Inarticulate
brachiopods from the Lower Ordovician of the Holy
Cross Mountains, Poland. Acta Palaeontologica
Polonica 23, 293-316.

Biernat, G. (1973). Ordovician inarticulate brachiopods
from Poland and Estonia. Palaeontologica Polonica
28, 1-120.

Biernat, G. (1984). Silurian inarticulate brachiopods from
Poland. Acta Palaeontologica Polonica 29, 99-103.

Biernat, G. and Harper, D.A.T. (1999). A lingulate
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 midwestern 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, 67-74.

Cherns, 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 acrotretacean 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 Australia. Courier Forschungsinstitut
Senckenberg 182, 347-370.

Cooper, B.J. (1974). New forms of Belodella (Conodonta)
from the Silurian of Australia. 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 southern Ohio. Journal
of Paleontology 50, 205-217.

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

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
Pridoli in Sardinia (Italy). Bollettino della Societa
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). Mémoire sur les Brachiopodes du
Systeme Silurien supérieur de |’ Angleterre. Société
Géologique de France, Bulletin 5, 309-338, 370-374.

Dean-Jones, G. (1979). Late Cambrian to Early Devonian
inarticulate brachiopods from Australia: their
classification, ontogeny, functional morphology and
ultrastructure. Unpublished MSc thesis, Macquarie
University, Sydney.

d’Orbigny, A. 1847. Considérations zoologiques et
géologiques sur les Brachiopodes ou Palliobranches.
Comptes Rendus Hebdomadaires des Séances de
|’ Académie 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 western NSW, Australia.
Courier Forschungsintitut Senckenberg 245, 107-
181.

Farrell, J.R. (2004). Siluro-Devonian conodonts from the
Camelford Limestone, Wellington, New South Wales,
Australia. Palaeontology 47, 937-982.

Foerste, A. (1888). Notes on Paleozoic fossils. Bulletin of
the Scientific Laboratories of Denison University 3,
117-137.

Gorjansky, V.I. (1969). Bezzamkovye brakhiopody
kembriiskikh i ordovikskikh otlozhenii sever-zapada
Russkoi platformy. Materialy po geologii i poleznym
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, ete. New York State
Cabinet of Natural History, Annual Report 23, 244-
247.

Haug, E. (1883). Ueber sogennannte Chaetetes aus
mesozoischen Ablagerungen. Neues Jahrbuch fuer
Mineralogie, Geologie und Paldeontologie 1, 171-
179.

Holmer, L.E. (1986). Inarticulate brachiopods around the
Middle-Upper Ordovician boundary in Vastergotland.
Geologiska Féreningens i Stockholm Forhandlingar
108, 97-126.

Holmer, L.E. (1989). Middle Ordovician phosphatic
inarticulate brachiopods from Vastergétland and
Dalarna, Sweden. Fossils and Strata 26, 1-172.

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). A new 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. Newes Jahrbuch fur Geologie 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 northem New
South Wales and central Queensland. Courier
Forschungsinstitut Senckenberg 182, 421-445.

232

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. Véstnik Ustredniho
ustavu geologického 57, 115-116.

Mergl, M. (1995). New lingulate brachiopods from
the Milina Formation and the base of the Klabava
Formation (late Tremadoc-early Arenig), central
Bohemia. Véstnik Ceského geologického ustavu 70,
101-109.

Mergl, M. (1996). Discinid brachiopods from the
Kopanina Formation (Silurian) of Amerika quarries
near Morfina, Barrandian, central Bohemia. Casopis
Narodniho Muzea Rada Prirodovédnd 165, 121-126.

Mergl, M. (1999a). Genus Paterula (Brachiopoda) in
Ordovician-Silurian sequence of central Bohemia.
Véstnik Ceského geologického tistavu 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 57, 1-49.

Mergl, M. (2002). Linguliformean and craniiformean
brachiopods of the Ordovician (Trenice 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).
Biostratigraphic 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 1 organizmy neyasnogo
sistematicheskogo polozheniya iz srednego ordovika
vosttochnogo Kazakhstana. Paleontologicheskii
Zhurnal 4, 33-42.

Nazarov, B.B. and Popov, L.E. (1980). Stratigrafiya i
fauna kremnisto-karbonatnykh tolshch ordovika
Kazakhstana (radioliarii 1 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.

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.

Popoy, L.E. (1981). Pervaia nakhodka mikroskopicheskikh
bezzamkovye brakhiopod semeistva Acrotretidae
v silurie Estonii. Eesti NSV Teaduste Akadeemia
Toimetised (Geologia) 30, 34-41.

Popoy, L.E. (2000). Late Ordovician linguliformean
microbrachiopods from north-central Kazakhstan.
Alcheringa 24, 257-275.

Popoy, L.E. and Holmer, L.E. (1994). Cambrian-
Ordovician lingulate brachiopods from Scandinavia,
Kazakhstan, and South Ural Mountains. Fossils and
Strata 35, 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 43, 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.

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

Serpagli, E. and Corradini, C. (1999). Taxonomy and
evolution of Kockelella (Conodonta) from the
Silurian of Sardinia (Italy). Bollettino della Societa
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 biostratigraphy
in Australia: a review and critique. Courier
Forschungsinstitut Senckenberg 182, 325-345.

Simpson, A.J. (1995b). Silurian conodont studies in
eastern Australia. 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 F OE
Senckenberg 182, 79-217.

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 2, 121-128.

Stauffer, C.R. (1940). Conodonts from the Devonian
and associated clays of Minnesota. Journal of
Paleontology 14, 417-435.

Strusz, D.L. and Munson, T.J. (1997). Coral assemblages
in the Silurian of eastern Australia: a rugosan
perspective. Boletin de la Real Sociedad Espanola de
Historia Natural Seccion Geologica 92, 311-323.

Sweet, W.C. (1988). ‘The Conodonta, morphology,
taxonomy, paleoecology, and evolutionary history
of a long-extinet 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 Australia, 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

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 401, 1-211.

Valentine, J.L. and Brock, G.A. (2003). A new
siphonotretid from the Silurian of central-western
New South Wales, Australia. Records of the
Australian Museum 55, 231-244.

Valentine, J.L., Brock, G.A. and Molloy, P.D. (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 function of protegular pitting in some North
American acrotretid brachiopods. Palaeontology 22,
705-720.

von Eichwald, E. (1829). ‘Zoologia Specialis, quam
expositis animalibus tum 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 Karnischen
Alpen. Notizblatt des hessischen Landesamtes fur
Bodenforschung, Wiesbaden 85, 28-52.

Walliser, O.H. (1964). Conodonten des Silurs.
Abhandlungen des Hessische Landesamtes ftir
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). A new
Llandovery (Early Silurian) conodont biozonation
and conodonts from the Becscie, Merrimack, and
Gun River formations, Anticosti Island, Quebec.
Journal of Paleontology, Supplement 76, 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
district, 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 during 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 Yaoxianognathus? tunguskaensis.
Associated macrofauna includes the corals Tetradium tenue, Bajgolia? cf. grandis, Propora bowanensis,
Paleofavosites?, Cystihalysites, 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 strata documented in the paper.

Manuscript received 13 September 2005, accepted for publication 7 December 2005

KEYWORDS: biostratigraphy, conodonts, corals, Macquarie Arc, nautiloids, Ordovician, palaeoecology,

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 macrofossils. 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 fault
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 Gisbornian 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 (Gisbornian
age) in the vicinity of Gunningbland, northward
along the Junee-Narromine Volcanic Belt (Pickett

236

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.

The presence of Yaoxianognathus? tunguskaensis,
Aphelognathus cf. webbyi and Tasmanognathus cf.
borealis in the assemblage also supports an Eastonian
age assignment. Yaoxianognathus? 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 Yaoqupognathus, 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 stromatoporoids. These are illustrated in
Figs 5-9, and their occurrences are detailed in Table
2s

Three samples from borehole DDMN 042
all yielded a single species of coral, TYetradium
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.

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

I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT

X Va
Gunningbland Va
_

{/
Parkes |

6320000mN
=Parkes My

|
West Wyalong> +Grenfell JsyDNEY
~, a

ee
6300000mN

LS

\ Lake
| Cowal

Cowal Mine &

NS

6280000mN —

/

|
6260000mN

Marsden

eS
Prospect _Caragabal y Ler, (

West
Wyalong \ yh Highway
= Sis Grenfell

Gibber ‘
| j ~ 6240000mN

Quandialla

520000mE 580000mE 600000mE 620000mE
|

REFERENCE

Jingerangle Formation subcrop inferred SSS ROHS

Jingerangle Formation outcrop 7 ~~. Waterways
--4-- Thrust Fault inferred

0 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

DDMN 042

ACDMN 043
> GL GL

C2102
C2103

50m som

i

Cainozoic regolith cover

i
Cainozoic regolith cover

a

a
Pil
LY Lo:

400m 100m

Volcaniclastic
sandstone &
conglomerate

C2073

Volcaniclastic 150m
sandstone &
conglomerate

Ea3? FHP? C2074
HTR or PEC Ea1-2

il

nD
Ia
entnee

Bad

HL

C2075
HTR Eat

Monzodiorite

il
UI 9

C2104
C2105
C2106

]
L

iW
tH

C2093
HTR Eat C2094
C2095

Ea1? C2096
HTR Eat? C2072 :

PEC/FHP C2076
Ea2 PEC

238.7m

2a 2

C2098 go

C2099 35

HTR = Hillophyllum-Tetradium-Rosenella Assemblage Zone Eai/2 C2100 = =
PEC = Propora-Ecclimadictyon-Cliefdenella Assemblage Zone C2101 450m stag

FHP = Favistina-Halysites-Plasmoporella Assemblage Zone
2005_10_0203

Figure 2. Diagrammatic representation of major lithologies intersected in Newcrest boreholes DDMN
042 and ACDMN 043 in the Marsden prospect, with enlargement of limestone-dominated intervals to
show sampled horizons and faunas recovered. GL = ground level; F = fault; Ea = Eastonian Stage, with
subdivisions Kal (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 pm. 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. Z 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 Creek; scale bars 100 pm. 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-
anognathus? 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,

MMM(C4146, 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.

Aphelognathus cf. webbyi
Belodina compressa
Panderodus gracilis
Panderodus sp.

Phragmodus undatus
Plectodina tenuis ?
Rhipidognathus sp.
Tasmanognathus cf. borealis
Yaoxianognathus tunguskaensis

Labechiella variabilis, Bajgolia? 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 from 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 and microfaunal 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 from
two of the cored holes. Though the conodont faunas
recovered lack Taogupognathus, 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

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

: — N (92) —t ~ (se) st LQ ice} ~ >
~ ~ ~ = ~ for) for) fer) fox) fer) =
Meee EEE EE Se SB S|
CONODONT TAXA He. Oe OO Ono OMe Ce eoy©)
. 1

= 00 ~/C2075

A=
ONANAAES ©| Total

only makes sense when details from the lithology log
are integrated with the palaeontological sampling
(Fig. 2). Samples from 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-
Ecclimadictyon-Cliefdenella Assemblage Zone.
Above a barren interval, samples from 189-193 m
yield both conodonts and stromatoporoids indicative
of an earlier, Eal, age. This succession 1s 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 from a
sample extending over the interval 179.5-183.9 m that
includes the aforementioned macrofossil assemblages,
are definitely of Eal-2 age — confirming structural
interleaving of fault slices. Samples from shallower
depths exhibit a similar intermixing of ages, with
macrofossils from 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 cn 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

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 min DDMN 0472, and longitudinal section from 276 min ACDMN 045. E, F,
Bajgolia? cf. grandis Webby, 1977, transverse and oblique longitudinal sections from 232 minACDMN 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

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

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 (GR 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

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, Paleofavosites? 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, 24 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

244

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 type in

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

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
1mm. A-D, Stratodictyon ozakii Webby, 1969, from 182 min 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 min ACDMN 043. F, Ecclimadictyon sp., longitudinal section, from 229 min 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) |

DDMN 042 399
405
406

ACDMN 043 173

Halysites sp

indeterminate
Te

'| ACDMN 045 27
2h

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 unmeasurable
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 from these outcrops were
provided by Sherwin (1982, 1985), Pickett (1986)
and Percival (1999). Faunal lists from 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

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

Tetradium tenue Webby & Semeniuk, 1971
Tetradium tenue Webby & Semeniuk, 1971
Tetradium tenue Webby & Semeniuk, 1971; Tetradium sp.

174 | Paleofavosites? sp

179-180 | Cystihalysites sp
181 | Cliefdenella cf. perdentata Webby & Morris, 1976

192
229 | Bajgolia cf. grandis Webby, 1977; Palaeophylium sp.; Propora
bowanensis Hill, 1957; heliolitid indet.; Clathrodictyon cf.

microundulatum Nestor, 1964; Ecclimadictyon sp.

| ATL

232 | Bajgolia? cf. grandis Webby,

1969; Cliefdenella sp.

182 | Stratodictyon ozakii Webby,

tradium? sp.

Assemblage

the exception of a nautiloid depicted by Percival and
Lyons (2000) (here re-illustrated in Figure 10E),
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 (Elles and Wood),
Normalograptus angustus (Perner), Orthograptus ex.
gr. amplexicaulis (J. Hall), together with Prilograptus
sp., and an unidentified climacograptid. Overlap of
the published ranges (VandenBerg and Gooper 1992)
suggests an early Bolindian (Bo 2) age is most probable
for the formation. Coiled tarphyceratid nautiloids,
referred to Discoceras? 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

247

QUANDIALLA-MARSDEN DISTRICT LATE ORDOVICIAN FAUNAS

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 siltstones
with a faunal association of graptolites, trilobites,
and diminutive lingulate and plectambonitoid
brachiopods. In the case of the Jingerangle Formation,
the Hindia-dominated 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 (*°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 from
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 120 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 from 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 of Hindia sphaeroidalis Duncan, 1879, MMF36600, from roadbase quarry, immedi-
ately south of Jingerangle State Forest. D-J, nautiloids from Jingerang!e 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 from 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 Vauréal Formation of Anticosti Island,
Quebec (Nowlan and Barnes 1981, pl. 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 from
adjacent denticles (Fig. 4G, H). The Sb element

250

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 from the Barmedman
Creek limestone. The form determined as Halysites
sp. (Fig. 5A, B) is definitely not the same as the only
other true Halysites known from 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
Tetradium? 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 1s also ruled
out by the absence of septa. Tetradium? sp. may
be related to a form referred to Tetradium sp. A by
Webby and Semeniuk (1971, pl. 17, figs 4, 5), that
has inconspicuous septa, angular corallites, and thin
walls. Cerioid or sub-cerioid corals known from 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 Survey of NSW collections
include a large number (MMF 29519-29538, 36592-
36608) of well preserved individuals of Hindia
sphaeroidalis Duncan, 1879 from the Jingerangle
Formation (Pickett 1986). The largest of these is
illustrated (Fig. 10C). This species was also reported
by Rigby and Webby (1988) from three of their four
horizons in the Malongulli Formation near Cliefden
Caves, and additionally from Late Ordovician strata

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

I.G. PERCIVAL, Y.Y. ZHEN AND J.PICKETT

at “Currajong Park”, Gunningbland, 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
182 min 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,
pl. 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. 10A-
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
omament 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 fulcral 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 from the Jingerangle Formation,
with the exception ofa section ofa large indeterminate
orthocone (Fig. 10D), are preserved as external moulds
or an internal mould impression. The position of the
siphuncle, and shape of the septa crossing the dorsal

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

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 Discoceras? (Fig. 10E), although Trocholites? or
Hardmanoceras? may be equally valid identifications.
Several large slowly expanding conchs (Fig. 10G-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 Gunningbland, west of
Parkes (Stait et al. 1985). A small individual (Fig.
10F) 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, Shaanx1
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: Beying) (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 studies.
University of Missouri Studies 8, 1-349.

Downes, P.M. and Burton, G.R. (1999). Mineral
occurrences in the Forbes district, pp. 37-52 in Lyons,

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"4 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 Geologn, Tallinn. 112 pp. (in Russian).

Nowlan, G.S. and Barnes, C.R. (1981). Late Ordovician
conodonts from the Vauréal Formation, Anticosti
Island, Quebec. Geological Survey of Canada,
Bulletin 329, 1-49.

Packham, G.H. (ed) (1969). The Geology of New South
Wales. Journal of the Geological Society of Australia
16(1), xx + 654 pp.

Percival, I.G. (1978). Inarticulate brachiopods from the
Late Ordovician of New South Wales, and their
palaeoecological significance. A/cheringa 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
GS1999/560 (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.

DS)

Pickett, J.W. (1986). Fossil sponges from Jingerangle.
Palaeontological Report 1986/02. Geological
Survey of New South Wales, Report GS1986/010
(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 GS1985/187 (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 central New South Wales,
Australia. Alcheringa 9, 143-157.

VandenBerg, A.H.M. and Cooper, R.A. (1992). The
Ordovician graptolite sequence of Australasia.
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 + 160 pp. Geological Survey
of New South Wales, Sydney.

Webby, B.D. (1969). Ordovician stromatoporoids 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, P.D. (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,

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 7etradium 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, 351-361.

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, Australia. 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, central New South Wales, Australia.
Geobios 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 IIT) 1:50 000 sheet
Jingerangle Formation in roadbase quarry, immediately south of Jingerangle State Forest.
Brachiopod: dorsal valve of small dalmanelloid?
Sponge: Hindia sphaeroidalis Duncan, 1879

indeterminate small conical form

gigantic monaxons (several cms in length)

echinoderm: crinoid ossicle
graptolite: indeterminate climacograptid?
Ptilograptus sp

Grid Reference 560500mE 6243280mN, 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 (Elles and Wood)

Normalograptus angustus (Permer)

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: Discoceras? 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 387m _ microfossil sample C2090 barren
389 m C2091 barren
391.2m 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)
410m C2097 conodont
412m C2098 ostracode and bryozoa
414m C2099 _bryozoa and lingulate brachiopod fragment
416m 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
173m macrofossil sample: coral (Halysites)
174m macrofossil sample: coral (Paleofavosites?)
179.5-183.9 m C2074 conodonts, silicified corals
179-180 m macrofossil sample: coral (Cystihalysites)
181m macrofossil sample: sponge (Cliefdenella)
182m sponge (Cliefdenella), stromatoporoid (Stratodictyon)
189-192.8 m C2075 conodonts
192 m macrofossil sample: stromatoporoid (Labechiella)
212m C2104 barren
214m C2105 barren
216m 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

- a
i :

PVAROMDOMAMIHXY VATA

ssheetae amt Su ERISD «—ssssslqunas line
| \ooubity dimadd FESDIsts © aren a
zinoboans ETOSD mr BP -TONE
0 ata bi Debate edgireee (Ged tae’ ED) 2:50 006 init me
(rove) dexom nigdie-lizedtatoadvingeranuiic Sint Fora. ea 8h
dmseiesiitelingatibdiics TOS)” meeei-eerr
(cae) howolgeweizzdionem = me ORL-OTE
folie peorqga tdiemes fizezdiomsm — - sea
(nowrysitegnaled Ser RR TLR Nliig t= ft $84 oS
2inobonos eTOs> im BO0T-C8E p
\sfoni harasses lizzotowanm . moet
nened $0159 a ae
comad 2018) | RTS
m= Meauk 6143 2wN, Marvdeented 2001S 11) 14 = G40 sheet — ma io
daubbgonntebjaicicitesergiri lant GS” howiestedd, ‘mi teeaen®
moulotariol) sedtitkeatticsd ebsthicdqemtemmoriqvals109 sztavib m
Luaracpaiieriegs: oiqnidddizestoroeny on SES
; Liles and Weaxl>
wie ancustius (Pecner)

F., | a ee. comer icy ott tlall:

et Belervace SdSimE 624310004, Marston 1-50 060-sheel
ry reé Gn i de behind “Bland Farni” aoe vicad
ruinate laswe multiecwtate urthide’

1 4 : Wak VCO
Prermiiiic OTTO {
‘ewerwsi Gril hale DDMN. 042, Marsden peaenert (iememeet BLSS24) ;
} Vip hotes 2 VAM. Li) eek 7 om
Sho mN (GDA eo-rdineter
iG at epee wd macrotis, rele te Tiles | and 7
ni roloesi! eerngrle t. tA) Paes |

MBIVaN

ary

TY 7 mloints ‘ : : PP ke ;
eS) > Cones ard oe maces |

(94 copadowt, anacrofesadl:comsl ‘indians

ue minkoras
pe | 4
intracades, seolecidints
mentite seme: corm! (Fetradiue)
amie. coral ( Jetruniver)
oe Jeaattoni
Ls -satrecode and hryvagi
Lier? trweroe and ling Wate nasiopod rage
‘Tn shuts v
Cle) > Berend i

Wi! bole ACDMN ON), Meoruder prosnect (feuement cS)
siete 7/47/2002, TY) 538.7 m 7
NS SGOA Co-arcenares

j

re! (iecTofaum, refer ie Tales | peek 7

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
number 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 from 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 from 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. Parnaby 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, P.C. Catling and
N.C. Coops give examples of the use of airborne
videography in forest management. C.P. Catterall et
al. deal with quantification, including design issues, of
the biodiversity values of reforestation. D. Milledge
suggests an innovative approach to conservation
planning 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 quolls,
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 fruit-eaters,
two frogs (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 fungus 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 fuel-reduction burns on fungi has,
in my Own experience, been used as a justification of
the practice.

258

Dan Lunney closes the book with a summary
entitled “The future of Australia’s forest fauna
revisited” in which he states the aim of this second
edition is to enhance the opportunities to communicate.
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

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 (M.L. 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, W.K. (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
from the website or from the Secretary.

260

"
Bs
atk

CONTENTS CONTINUED

83

93

Mahony, M.
Amphibians of the Gibraltar Range.
Vernes, K., Green, S., Howes, A. and Dunn, L.
Species richness and habitat associations of non-flying mammals in Gibraltar Range National Park

Section II: General papers.

107

125

133

157

175

191

199

235

256

Harris, J.M.
The discovery and early natural history of the Eastern Pygmy-possum, Cercartetus nanus (Geoffroy
and Desmarest, 1817).
Piper, K.J. and Herrmann, N.
Additions to knowledge of the early Pleistocene wallaby Baringa nelsonensis Flannery and Hann
1984 (Marsupialia, Macropodinae).
Williamson, P.L. and Rickards, R.B.
Eastonian (Upper Ordovician) graptolites from Michelago, near Canberra.
Timms, B.V.
The geomorphology and hydrology of saline lakes of the middle Paroo, arid-zone Australia.
Foldvary, G.
Pseudoplasmopora (Cnidaria, Tabulata) in the Siluro-Devonian of eastern Australia with comments
on its global biogeography.
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.
Valentine, J.L., Cole D.J. and Simpson, A.J.
Silurian linguliformean brachiopods and conodonts from the Cobra Formation, southeastern New
South Wales, Australia.
Percival, I.G., Zhen, Y.Y. and Pickett, J.
Late Ordovician faunas from the Quandialla-Marsden district, south-central New South Wales.
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 1.f. (Moracea).

PROCEEDINGS OF THE LINNEAN SOCIETY OF N.S.W.

INSTITUTION LIBRARIES

NLC

01210 2877

MITH

wi

Issued 23 February 2006
CONTENTS

Section I: The biology and ecology of Gibraltar National Park.

1 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.

11 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, P.R. and Clarke, P.J.
Fire history and soil gradients generate floristic patterns in montane sedgelands and wet
heaths of Gibraltar Range National Park.

39 ~=Virgona, S., Vaughton, G. and Ramsey, M.
Habitat segregation of Banksia shrubs at Gibraltar Range National Park.

49 Knox, K.J.E. and Clarke, P.J.
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.
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

Printed by Southwood Press Pty Ltd,
76-82 Chapel Street, Marrickville 2204

PROCEEDINGS

me INNEAN
aSOCIETY

of
NEW SOUTH WALES

VOLUME 128

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 Wiliam
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: D.R. Murray

Vice-presidents: M.L. Augee, I.G. Percival, J.P. Barkas

Treasurer: 1.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: P.R. Reeves

Auditors: Phil Williams Carbonara

The postal address of the Society is: P.O. 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: Myrtaceae pollen from Robbie and Martin, page 65, this volume

PROCEEDINGS
of the

LINNEAN
SOCIETY

of
NEW SOUTH WALES

SMITHSON A NV

MAR 15 2007 ‘

LIBRARIES

For information about the Linnean Society of New South Wales, its publications and
activities, see the Society’s homepage

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

VOLUME 128
February 2007

se OLMIS

Poamled 1874
lex pornqul L684:
: i.
MiCIcty ¢£ics fe pram ote the cultivation and

the sciende ot neal SRC r aeny

he Socwty ewerds pitearch granieeach vear 4
“ida of Life Sovnoes Tie Jayos Vickery te
wth Sciences aha Berry Mayon fend) offen
uvican Macteay Fellowship fos research, Con

te (ic etrheid of the Lisean Niaciray td tM
Mies hilo ws it ibe | niversity ot Syet ay
wibhisthew the Preseli. 1 tyohds fei chee
.lenifttc pectings, avclucdiee: the become Set
Mocleey Memortil Loctere deiiesead oF a iM
rhiveud i ware Eiagoh of snllfal ENE 7

.

ben eaotinoilducpan woke. adtteediw tif fie sarod anotuatlont 4 ;
> waqamod diytsaoe shops: eeanreion} aiBacripticin ih

mat. smodiaw\oaennt- “\ut. (G9. VBI8. vewrw
c & Bey eur Tolhione Bis) recent eR

i}

f the Preceed ar ge silable ie punches

bel f = out, MM .R. tay,
; nial ire 1 Clty 2) by ermmmenagi CST

- rom i S4: %.'. Gaver nS Ww Fa bo Anneke”
"y Ve Bepea § Cia 2) tae

—— orw ‘4 herd

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 (anita.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 Corymbia 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
under 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; Cockburn 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 etal. 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 (T7richosurus
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 strenua) 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 living 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, fungi 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 fungi, 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 FE. 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; 11) 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 frequency 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

OURIMBAH

SF
oN

POPRAN
NP

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

Newcastle

Wyong
®

Gosford

Figure 1. Location of the 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 Target species in each Frequency* Land Tenure No. of Sites
community* vegetation community

1 Angophora costata 36% Ourimbah State Forest 2
Eucalyptus punctata 31% Olney State Forest 2

y Eucalyptus punctata 68% Dharug National Park 3

3 Angophora costata 74% Ourimbah State Forest 2
Eucalyptus pilularis 40% Bouddi National Park 4
Corymbia gummifera 48%

4 Angophora costata 45% Ourimbah State Forest 4
Eucalyptus haemastoma 50% Popran National Park 4
Corymbia gummifera 75%

5 Eucalyptus punctata 52% Dharug National Park 1
Corymbia gummifera 40%

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 from 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 | 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 trees 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 tree was
determined from the ground using 10 mm x 25 mm
binoculars. A hollow was defined as any cavity with
an entrance > 2 cm in diameter and occurring > 3m
above the ground. Entrances 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) entrance 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

Proc. Linn. Soc. N.S.W., 128, 2006

P. TODARELLO AND A.

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

A total 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 from | to 37 trees
ha” (mean of 16.6 ha‘') and the number of hollow-
bearing trees ranged from 0 to 27 ha! (mean of
6.2 har'). 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 A.
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 A. costata (40 %) > E. punctata (26
%) > C. gummifera (24 %) > E. pilularis (22 %).
At the stand level (1.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-

Proc. Linn. Soc. N.S.W., 128, 2006

Table 2. The range of values for various attributes of tree stems and hollows for each of the five species examined in the current study. SE

standard error of the mean.

CHALMERS

Corymbia
gummifera

Eucalyptus
haemastoma

Eucalyptus
vailevats

tus
ta

fa

Eucaly
punc

Angophora
costata

Variable

No. of | ha quadrats with target species

181

175 157 82 103
255
39.8 + 1.35

211

35.7 + 1.28

Total no. of trees sampled

113
33.8 + 1.14

97
51.8 + 3.61

105
38.4 + 1.45

Total no. of hollows observed

Mean (+ SE) tree diameter (cm)

20 - 86

21 - 85
12.7£0.27

20 - 124 20-151
21.7+0.41

27.6 + 0.43

20 - 147
23.3 + 0.53

Range of tree diameter (cm)

18.1 £0.34

Mean (+ SE) tree height (m)

18 - 39 16 - 32 7-21 11 - 32

10 - 39

Range of tree height (m)

0.6 + 0.10

0.7 +£0.12 1.2+0.30 2.5 + 0.20
16.2 + 3.89

18.1 + 1.78

12+0.15
17.0 + 3.40

Mean (+ SE) no. of hollows per tree

Mean (+ SE) no. of trees ha!

16.5 +£1.77
4.1 + 1.34

14.7 = 1.44

3.8 + 1.88 11.0 + 1.20

5.9 + 1.16

6.5 + 2.48

Mean (¢ SE) no. of hollow-bearing trees ha"!

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).

Spearman’s rho

The mean number of
hollows per tree increased
with increasing trunk
diameter and all of the five
species had, on average,

n five or more hollows per

Tree diameter Tree height tree once their diameter

was = 84 cm (Fig. 5).

Angophora costata 175 0.61** OSes With the exception of E.
= te haemastoma, trees with

Eucalyptus punctata 157 0.44 0.32 Slanneionms basnosn 20 en
; : # ae and 51 cm had, on average,

Eucalyptus pilularis 82 0.67 0.40 foveuihadinwaibolloneinen
Eucalyptus haemastoma 103 0.58** 0.23* treen(lige<d): Compated
to the other species

Corymbia gummifera 181 0.53** 0.06" examined, £. haemastoma

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
15 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.

had a relatively high mean

number of hollows per

tree in the three smallest
diameter classes (Fig. 5). In contrast, EF. pilularis had
a relatively low mean number of hollows in all but the
largest size class. For 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

Proc. Linn. Soc. N.S.W., 128, 2006

P. TODARELLO AND A. CHALMERS

(a)
®
E E
2 2
x) So
a D
2
E E
z z
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
Trunk diameter at breast height (cm) Trunk diameter at breast height (cm)

®
£
—
Q
fs
re)
£
6
3
&
=]
Fa
0 2 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160

Trunk diameter at breast height (cm) Trunk diameter at breast height (cm)

Number of hollows / tree

0 20 40 605 80 100 120 140 160

Trunk diameter at breast height (cm)

Figure 2. Scatterplots of number of hollows per tree against trunk diameter at breast height for (a) An-
gophora costata; (b) Eucalyptus punctata; (c) Eucalyptus pilularis; (d) Eucalyptus haemastoma; and (e)
Corymbia gummifera.

Proc. Linn. Soc. N.S.W., 128, 2006 f

HOLLOW-BEARING TREES

(a)
16 (0)
16
14
®
© @ 14
= 12 5 12
n j a a
= 10 * a g 10
Oo 0
— 8 xe)
= o o o To) 2 : i
£&
— 6 @) Gi e 6 o
(o) o o o °
oO 4 oo o o oo -
() 4 o o
ne} o oo o oo oo oo o a o o oo
£ 2 oooo o o ooogooooo o i= 2 oo oo o
3 o o oo 8000 o ooo =) o Do o000 o oooo
0 DB 8 S80009000000500005 oooo0 o Zz 0 poocooc0oocCoooooooooooo
-2

Tree height (m) Tree height (m)

(c) (d)

16

2 16

14
® . @ 14
eS =S 12
o —~
= 10 o 2) 10
= 8 o 3 iB e
2 sf ron ;
te 6 o = 6 o ooo o
© a i) ie) o oo ooo oo o
a + Ajay Br the: incrennedsazeth
E 2 3) = E 2 o000 oo
Zz o o o z opoooooo

0 500 oooo0000 0 6 coOoo0000

-2 2

0 10 20 30 40
Tree height (m) Tree height (m)

(e)

16
@ 14
2
ce 12
¥ 10 °
xe)
oO 8
At = o
5 6 oo
o 4 oo ]
Oo o ooo oo oo
E 2 oo oo o
2 0 ooooo0ooooooooooooo o Oo

2

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

1.0 - d c
0.9 -
0.8 -
0.7 + d
0.6 -
0.5 -
0.4 -
0.34 a

0.24 []b =
0.1 - c
0.0 4

T T
n= 111 9937 41 124 43 28 16 42 28 2 16 715 21 59 536 4 5A?) he

20-35 36-51 52-67 68-83 84+

one hollow

C C

Proportion of trees with at least

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 = Angophora costata; b = Eucalyptus
punctata; ¢ = Eucalyptus pilularis; d = Eucalyptus haemastoma; e = Corymbia gummifera.

Number of hollows / tree

n= 111 99 37 41 124 43 28 16 42 28 12 6 7 15 21 59 5 3 6 457 2 2

20-35 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 = Angophora costata; b = Eucalyptus punctata; c = Eucalyptus pilularis; d = Eucalyptus haemastoma; e
= Corymbia gummifera.

E. pilularis (9 %). The distribution of hollow entrance | medium (6-10 cm) diameter entrance and 47% had a
diameter sizes between branches and main stems was _ small (2-5 cm) diameter entrance; of those occurring
similar. Of the hollows occurring in branches, 26% in the main stem, 24% had large, 21% had medium
had a large (> 10 cm) diameter entrance, 27% had a and 55% had small entrances.

Proc. Linn. Soc. N.S.W., 128, 2006 9

HOLLOW-BEARING TREES

Number of hollows / tree
(o>)

abc] le be cde

0 ty es ' 24 01 77 93 77 35 48 4 57 23 59 33 0 26 50 63 0 O 5

5-10 11-16 17-22 23-28 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.

Proportion of trees with hollows
SSS99S99D9090900=
O-]-NWAhUDN OOO

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 growth 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 trees 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 trees 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 tree (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 trees may also have a smaller
number of hollows with large entrances (Wormington
et al. 2003) and therefore will suit a narrower range

Proc. Linn. Soc. N.S.W., 128, 2006

of fauna. The relatively small eucalypt species, E.
haemastoma, had as many as 60% of trees 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 trees
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
tree 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 trees 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 trees were sampled and the shorter trees
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 trees 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 trees (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
trees that “do not reach large diameters, regardless of
longevity, are only infrequently observed to contain
hollows” and trees < 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 tree would be
less than that of the other species surveyed. Therefore,
smaller diameter E. haemastoma trees are likely to
have greater susceptibility to fungal decay.

The proportion of A. costata trees 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 hollow-

1]

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 A. 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 EF. 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 EF. 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.

12

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 (1.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

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, 222-235.

Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall,
N., Hyland, B.P.M., Johnston, R.D., Kleinig, D.A.
and Turner, J.D. (1992). ‘Forest Trees of Australia’.
(CSIRO Publishing: Victoria).

Cockburn, 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, P.C. (1996). Modelling
distributions of arboreal and ground-dwelling
mammals in relation to climate, nutrients, plant
chemical defences and vegetation structure in the

Proc. Linn. Soc. N.S.W., 128, 2006

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 Publishing: 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 southeastern 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 southeastern 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 27, 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 Strategy (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 southeastern
Australian forests and some implications for
conservation. Journal of Mammalogy 78, 1117-1127.

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
eastern 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

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, 207-221.

Lindenmayer, D.B., Cunningham, R.B. and Donnelly, 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 137, 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.

Lindenmayer, D.B., Cunningham, R.B., Tanton, M.T.,
Smith, A.P. 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, C.M. (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, fungal 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.

14

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.

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 Australia

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.

Manuseript received 23 January 2006, accepted for publication 18 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/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

THIRLMERE -
LAKES :

Wesco

Burragorang

Valley
<—————_

m surface sample
TS: Forest sample
LS Lake sample

& Vegetation quadrat

erreerr Railway
=-——=— Roads
0 0.5

SCALE

Sandstone feat

The Oaks

Y
Penrith SOE

@ Campbelltown,
The Oaks (

@
Thirlmere ,
@Picton

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

16

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

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

OTe

Forest

(P = Pine)
(ia Cleared

rare We 2ece.
eee6
c

a Pee
.

a
by
4

e
e

@
e

=
\ e

ee fey

eke:

“2
a ---* 4
_*> a

% Core for pollen analysis

AW
MN}

Transect
Vegetation quadrats

Auger holes
Pits

Reed beds

Channel

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 deaneiand Eucalyptus elatatall 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

Proc. Linn. Soc. N.S.W., 128, 2007

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

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 A/nus of a known concentration, treated
with hydrofluoric acid to remove siliceous material,
boiled in 10% sodium hydroxide to remove of
humic acids, disaggregated with ultrasonic 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 A/nus 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 0 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 160-200 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.

18

RESULTS

The Lake Sediments.
The sequence of sediments in the central part of
Dry Lake is as follows, from top to bottom (Fig. 3):

1)Alayer of little decomposed fibrous peat, composed
of rhizomes, root and stems of cyperaceous reeds
(probably Lepironia articulata).

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) streaks or small iron stained clay
concretions (<< 3 mm diameter) appeared along
old roots or root channels. 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
15% 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

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

KEY
Black clayey peat
WD
Sandy peat
| Sandy peaty clay
Medium organic clay
Ht}

a Iron /clay concretions

Depth

(em) Auger 1

Brown fibrous peat

Peaty sand
Dark sand
Pallid sand

Sandy clay

Fine sandy clay ‘diatomite’

100

Medium grey/brown clay

Mottied clay

Heavy pallid clay VeD

140

Auger 2

20

Auger 4

% loss on ignition

32 40 50

Pit 2b

I
bpp rll)

©! 1o | 10
1
a

®& Radiocarbon years.
. 920 + 80

- 1120 + 80

. 2170 +100

. 5820 + 130

. 8780 + 160

Ak WNH =

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
Pit 1
Black clayey peat (humic
Lae acid fraction)
Charcoal
22-24 Humic acid in charcoal
Peat (around charcoal)
74-32 Diatomite (humic acid
fraction)
38-45 Organic clay (humic acid
fraction)
Pit 2
10 cm length of wood
73-83 Wood (humic acid
fraction)

Proc. Linn. Soc. N.S.W., 128, 2007

Radiocarbon years BP

920+80 (NSW 375)

1,120+£80 (NSW 381)
1,560+120 (NSW 380)
1,660+90 (NSW 384)

2,170+£100 (NSW 376)

5,820+130 (NSW 377)

8,780+160 (NSW 387)
8,060+300 (NSW 388)

Calibrated years BP

795

986
1,417
1,499

2,101

6,570

SID
8,899

19

VEGETATION HISTORY OF DRY LAKE, NSW

1

a
i
a

ey a

GZ Closed sedgeland

open forest
Melaleuca linariifolia
low closed or open forest Eel Cleared land
7 Angophora floribunda *-| Gully forest
J woodland/open forest
[TI] Mixed Eucalyptus/Corymbia mm pe
woodland 11 111 Ridgetop contour

Figure 4. The vegetation of the Thirlmere Lakes region.

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 from the

20

C] Mixed Eucalyptus/Corymbia

littoral zone reflects deeper water where
the dominant process is the settling of
fine 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 layermustremain 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.

The Vegetation

Appendix | 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/

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
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. 6A). 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

Proc. Linn. Soc. N.S.W., 128, 2007

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 environmental 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 moister foot
slopes.

3. 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

ih

VEGETATION HISTORY OF DRY LAKE, NSW

site drainage is moderate to poor.

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
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 stamps 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 of Imperata 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 L. 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

Proc. Linn. Soc. N.S.W., 128, 2007

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 surrounded
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. /ongitudinale, 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.

23

VEGETATION HISTORY OF DRY LAKE, NSW

2
_ Water level \|
Ss Tee bie Ge
Gx)

7

3m
2m 1 0

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 S. melanostachys.

Modern Pollen Deposition

Pollen is produced by the contemporaneous
vegetation, but a multitude of factors affect the
representation of pollen in the sediments (Birks and

24

Birks, 1980; Dodson, 1983; Moore etal., 1991), hence
it is not possible to relate fossil pollen assemblages
directly to the vegetation which produced it.

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 modern 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 longitudt-

nale. The surrounding slopes are cleared.

Proc. Linn. Soc. N.S.W., 128, 2007

US

VEGETATION HISTORY OF DRY LAKE, NSW

%. 3, &
G6. Oe. Ie, Ey “4 os Co Ss, 9 Se,
9B, Fa, °C My, POG, #2 fo) 36 25, & eo, We, A> %o, 885. “Soy, © CD
28, "Uo, eo YI, us Oe ayy 8, Oy, es, bse “Oy, Bu, 929, ap by, “89 yy), "Len, Pip, Yaz,
“Mo Oo. Se "9 Xe Sy He, Y, So, 2p, M4409. Ven, Poy, oy, og, Ota Bg Hes a Ye “oy ors Qo, "My

ob Bae

(G-8 21) 2 X27 Siow, Wed) S;JUEWIPES

a :
t+ | BSRRRce@ = file cid

|
IAA SN

orn ow

F Ss
%0L § 0 ales (9-1 SL) SemlunWwos pue|poom pue }sel04

rT NOT WY O

Sy

4Q,

Veo,
y

(S-8 Z1) 2 ayeT Grau) WOY SJUaWIPeES

u
= a FF

(p-L 1q) sejdwes eyeq Aug

2 2, F SOI}IUNWLWOD pue|poO, e 1Ssa10 V
%OL S oO 2989S (9-1 SL) seniu Ppue|PooM pue } 4

TN OM g on oOo Ww

TF NO THM ©

Proc. Linn. Soc. N.S.W., 128, 2007

26

S. ROSE AND H.A. MARTIN

T PLABL Ul OS 94} 38 WOeIIS9A JY) YIIM oAedUI0D “pound jou ynq juasoad ‘aaa ‘+ *¢ xIpusddy
998 ‘UO}B}999A 9Y} UT UaT[Od ay} JO adAno0s afquqord oy) 104 “| aANSIy aas ‘sajts Supduies dIBJANS JO UOILIO] 94} 104 ‘UOHIsodap ualjod usspoyy [] aNsI1y

9,
8, Loy
a S95 MO, S5 ey i “Ying, 7 Mop ‘ 3
tip 9, Po, 8 Go a5, ng So "tp," Ban Bg
Sy, Yo, $306, Uo, Yoo, M6,, "Soy Bs Sou, Vey, oy, 5, “Bq
"yy, py, 045 "Sa, ag, 785 NW My yy Oy 5 4S, b

or o wo

(S-8 71) Z eye] G1EW}IY) Wos) SJUAWIPES

Wi pea: rT 444 33

(v-L 1G) sejdwes aye q Aq

Tr NO

Es ae

+
——— (9-1 SL) SejluNWCS pue|pooM pue }ses04 ]
% OL S 0 8/E9S

Di.

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION HISTORY OF DRY LAKE, NSW

Table 2 Surface sample collecting sites and their vegetation. Compare with pollen representation in

Fig. 11.

Sample

TS 1

TS

TS 3

TS 4

TS 5

TS 6

DL 1

DL 2

DIES

DL 4

28

Site details

Vegetation

Forest and woodland sites

Lake margin, on

alluvium

Colluvial footslope

Seep rocky slope

Ridgetop plateau

Colluvial footslope,

recenty burnt

Gully

Dry Lake sites
Colluvial slope

Colluvial slope

Margin of lake

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.

Fernland 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.

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

Thirlmere Lake sites

Closed sedgeland of Lepironia articulata, Eleocharis spacelata,

ILS 50 m from lake margin, | Open water, no vegetation.
open water 3 m deep

L26 20 m from lake margin, | Open water, no vegetation.
open water 3 m deep

Ey 5 m from lake margin,
water 2 m deep Brasenia schreberi.

128 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, Allocasuarina, 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 spectra 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 spectra 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

Proc. Linn. Soc. N.S.W., 128, 2007

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 unburnt 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. Ina 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

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
palynological 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

30

(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.

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

*po}UNOd J0U yng UESoId ‘o1v1
‘+ °¢ xipueddy des ‘wie1serp usfjod oy} Uo auIeU Yo¥a AO UaT[od dy} Jo ad.1NOs d[quqord 94} 10,7 ‘s}UdWIpEs oye] Aa ay) JO v.Q90dS uaTjod oy], 7] OANSIY

OL X Woullpas jo wojsueiby Zz oO % OL suiei6 Ubu} ssayyunoD x
e jo, + SR o S$ 6 60 IED OO} UEL} Ssaj juNDD
SUOHBNUSSUOS UBijod ‘ajEDS sebeyusoled uaijod ‘9] jUNOD Us}|Od |B}0} = wns Ua}jOg

wee =I
i

oe 0
prarrra

a

=a ial

t

1
ead
oO

q
i

31

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION HISTORY OF DRY LAKE, NSW

*pojunod jou nq jUVSe.1d ‘a1v.1 ‘+ *¢ xIpusddy
298 ‘WikASEIpP Ua][od ay} UO sUIeU YdLO 10J UaT[od ay) Jo 2d.1NOs a[quqoAd ay) 104 “syUaWIpas aye] Arq ey} Jo eQoeds uatjod ayy ((penuyUOd) ZT ans

@ Ls ______j, __,
UM les eat iid

Se,
My
oy,
Sy. >

Pe,
*F8y,
7 4
#9
Gey,
Yo ik
"82y,

Proc. Linn. Soc. N.S.W., 128, 2007

32

S. ROSE AND H.A. MARTIN

*pajunod jou ynq yUNSo1d ‘a1v.1 ‘+ *¢ xIpueddy
998 “Wv.ASvIp Ua][od 94} UO sUIvU YOvS 105 Uat[od ay} Jo 9dAN0s a[quqoid ay) 104 *sjusUMIpes aye] A1q 94} JO v.ayd0ds UaT[od ay, “(panuNuOd) Z] 21n31q

<OLX JUaWIpes jo ,wo/surei5 Jo Jequinn
Ove of8) > F 2 HLE Te 0

UO}B.jUBIUD Ueljod [B}0),

33

Proc. Linn. Soc. N.S.W., 128, 2007

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

34

deposited at this site. There are two peaks in the values
for megascleres: at 31 cm and a much larger one at
5lcm, 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 found 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 from 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 R. 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 environment
depositing clay. The margins probably supported
some cyperaceous reeds and Gonocarpus, but a fringe
of paperbarks (Melaleuca linariiforia), similar to the

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

35

*SI1I[ISVBIUI 0} SAAI[ISOWIUIIS JO ONLY ‘q ‘Sd1I[ISUSOUI B[OYA 0} SJUDUISBAY
d19[ISESOUI JO OLA ‘> *,O] X JUSUMIPAS ,WId/ So1d[ISOWIUIAS JO ABQUINN “_ *,.O] X JUSUIIPas EUID/sar1a[ISEsoUI JO JaqUINN] “y “s[Issojorstu asuods “Ey a1ns1]

OZ't 08'0 #0 0 00r 00g 0 009 OOF 00z 0

£S@'1
a ozd
sejdwes 10D (wo) inde

Sc evaeses seg~ ode ewe Coe ete

sajdwes soeyng

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION HISTORY OF DRY LAKE, NSW

Thirlmere Lakes today, was lacking. The freshwater
sponge, Radiospongilla sceptroides, \ived round the
margin of the lake, amongst fallen debris. Eucalyptus
spp. and A//ocasuarina were the dominant trees in the
surrounding vegetation and Asteraceae (Tubulifiorae)
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 Me/aleua 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

36

and Asteraceae (Liguliflorae: probably Hypochoeris
radicata) denote the zone of European influence. All
the trees, viz. Eucalyptus spp, Angophora/Corymbia
and A/locasuarina 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

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
tolerant 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 A//ocasuarina in this
study because there are only two species in the area
today: A. littoralis and A. torulosa. A. littoralis 1s
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. Jittoralis (Plantnet,
2005). The alluvial flat around Dry Lake would have
been suitable for A. torulosa, but at Lake Baraba, in a
sandstone valley, A. littoralis 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.

Proc. Linn. Soc. N.S.W., 128, 2007

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 Armold: 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 115,163-191.

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 27, 1079-1094.

Dodson, J.R. (1983). Modern 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).

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.P., 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). Modern 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.P. (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, P.D., Webb, J.A., Collison, M.E. (1991). ‘Pollen
Analysis, second edition’. (Blackwell Scientific
Publications: London).

NPWS, 1997. Thirlmere 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, 113-137.

Plantnet (2005). National Herbarium website (http://
plantnet.rbgsyd.nsw.gov.au). Accessed October 2005.

38

Racek, A.A. (1969). The freshwater 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. and Reimer, P.J. (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, P. (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).

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

APPENDIX 1A

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, prostrate herb or shrub. TW, twiner, climbing plant.

Pollination mechanisms (Poll’n Mech), from Faegri and van de Pijl (1971), Dodson (1979), Arm-
strong (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).

Species ie oll’'n ccurrence.
MYRTACEAE

Angophora floribunda (Sm.) Sweet T A,E,O Very abundant, lower slopes only
CPR Meniaeunts (OCRETEE) Hou) EU ee ir A,E,O Very abundant, esp. ridgetop plateaux
. Cac (Gaertn.) K.D. Hill & L.A. T ABO. Nenad

Eucalyptus agglomerata Maiden alt A,E,O Occasional

E. botryoides Sm. TT AN IB,©Q) *

E. oblonga Blakely. T A,E,O Occasional, mostly steeper slopes

E. piperita Sm T A,E,O Very abundant

E. punctata DC T A,E,O Occasional

E. racemosa Cav. T A,E,O Occasional

E. resinifera Sm. T A,E,O Rare

E. sieberi L. Johnson T A,E,O Occasional, esp. near ridgetop plateaux
Leptospermum trinervium J. Thompson S A. E. Common, mostly on slopes

L. polygalifolium Salsb. S) A,E Rare, mostly on slopes

L. juniperium Sm. S p18 Occasional, mostly lake margins
Kunzea ambigua (Sm.) Druce S Aes, (O) *

Melaleuca linariifolia Sm T A,E,O Abundant, mainly lake margins

M. thymifolia Sm. S A,E,O Occasional, along lake margins

PROTEACEAE

Banksia integrifolia L. f. Mens) BE. © =

B. serrata L.f. TorS E,O Abundant, mainly lake margins
B. spinulosa Sm. S E,O Very abundant

Grevillea arenaria R. Br. S E, O. zs

G. mucronulata R. Br. Ss E,O Very abundant

Hakea dactyloides Cav. S lag @ Occasional, moist sites

H. salicifolia (Vent.) B.L. Burtt. S) E,O *

Proc. Linn. Soc. N.S.W., 128, 2007 39

VEGETATION HISTORY OF DRY LAKE, NSW

H. sericea Schrad. & J.C. Wendl. S E, O Rare

Isopogon anemonifolius Knight S E Occasional

Lambertia formosa Sm. S) B@ abundant

Persoonia lanceolata Andrews S E *

P. laurina Pers. S E Occasional, esp. ridgetop plateaux
P. levis (Cav.) Domin S E Occasional

P. linearis Andrews S E Occasional

Petrophile pedunculata R. Br. S) E Occasional

P. pulchella R. Br. S E

P. sessilis Sieber ex Schult. S E ~

Telopea speciosissima R. Br. Ss E Occasional

Xylomelum pyriformis Sm. T B Occasional

FABACEAE

1) MIMOSOIDAE

Acacia decurrens Willd. T E Occasional esp. moist gullies/slopes
A. falcata Steud. S E z

A. falciformis DC. T E *

A. floribunda Willd. S E i

A. implexa Benth. S E ~

A. linifolia Willd. S E Abundant

A. longifolia (Andrews) Willd. S E Abundant esp. after fire

A. myrtifolia Willd. S E Occasional, ridgetop plateaux
A. parramattensis Tindale S) E Very abundant

A. suaveolens (Sm.) Willd. N) E Occasional

A. terminalis J.F. MacBr. S) E Occasional ridgetop plateaux
A. ulicifolia Court S) E Abundant

2) FABOIDAE

Bossiaea buxifolia A.Cunn. S E Rare

B. heterophylla Vent. S) B Occasional

B. lenticularis DC. S E sg

B. neo-anglica F. Muell. S E =

B. obcordata Druce S) E Occasional

B. rhombifolia Sieber ex DC. S E Occasional

Daviesia corymbosa Sm. S E Rare

Desmodium rhytidophyllum F. Muell ex S or E -

Benth. TW

D. varians (Labill.) G. Don TW E *

Dillwynia glaberrima Sm. S) E i
40 Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

D. parvifolia R.Br.

D. phylicoides A. Cunn. sp. complex
Glycine clandestina J.C. Wendl.
Gompholobium grandiflorum Sm.

G. latifolium Sm.

G. minus Sm.

Hardenbergia violacea (Schneev.) Stearn
Hovea linearis (Sm.) R. Br.

Indigofera australis Willd.

Kennedia rubicunda Vent.

Mirbelia rubiifolia (Andrews) G. Don

Podolobium ilicifolium (Andrews) Crisp &

P.H. 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

Proc. Linn. Soc. N.S.W., 128, 2007

NnNanan gyn

N

An AHRA DV

les} les} esl tes} [eel teal test Jes! lea! les! les!
(e)

tH

lel esl Jes} Jes!

leol Tes! Jes! Jes|

*

Abundant
Abundant

*

*

Occasional
Very abundant
Occasional
Abundant

Occasional

*

Occasional esp. rocky slopes

Very abundant
*
Occasional esp. alluvial fans

Occasional esp. damp sites

Occasional, esp. rocky slopes
*
Occasional

Abundant

*

Rare
Abundant

Rare
Occasional
Occasional

Occasional

*

Very abundant
Occasional
Occasional, moister slopes

Occasional, moister slopes

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 (Aiton) 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
FH. laxiflora DC.

H. peduncularis. A. Rich.

Platysace linearifolia 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.

42

a, DQ A B®

TmANANUR D

or

2rUe ft

A fac} ja aE, (OY BE

TW

TW

H
H

> mm ow

les! [esl [esl lesl lesl fel

Occasional
Occasional, moister slopes
Occasional, lake margins

Occasional

Very abundant, lake margins,
footslopes

Occasional, upper slopes, ridgetops

Occaisional
*

Abundant
Abundant

*

2

*

*

Abundant on lake margin, moist areas

*

Abundant on lake margin, moist areas

Very abundant

Occasional

*
*
*

*

Abundant.

One specimen observed

*

Occasional, moist areas

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

V. hederacea Labill.
CRASSULACEAE

Crassula sieberiana (Schultes & Schultes.

f.) Druce

DROCERACEAE

Drosera spathulata Labill.
POLYGONACEAE
Persicaria hydropiper (L.) Spach.
Acetosella vulgaris Four.
OXALIDACEAE

Oxalis corniculata* L.
GERANIACEAE

Geranium homeanum Turcz.
HALORAGACEAE
Gonocarpus micranthus Thunb.
G. tetragynus Labill.
Myriophyllum variifolium Hook. f..
THYMELEACEAE

Pimelea linifolia Sm.
PITTOSPORACEAE
Billardiera scandens Sm.
Bursaria spinosa Cav.
PASSIFLORACEAE
Passiflora edulis * Sims
HYPERICACEAE

Hypericum gramineum G. Forst.
ELAEOCARPACEAE
Elaeocarpus reticulatus Sm.
Tetratheca thymifolia Sm.
MALVACEAE

Sida rhombifolia * L.
CUNONIACEAE
Ceratopetalum gummiferum Sm.
ROSACEAE

Rubus parvifolius L.

R. fruticosus * species complex
STACKHOUSIACEAE

Stackhousia monogyna Labill.

Proc. Linn. Soc. N.S.W., 128, 2007

AqH

TW

TW

Occasional, moist areas

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

a

Rare, disturbed sites

43

VEGETATION HISTORY OF DRY LAKE, NSW

S. viminea Sm.
LORANTHACEAE
Unidentified

SANTALACEAE

Exocarpos cupressiformis Labill.
E. strictus R. Br.

Leptomeria acida R. Bt..
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

Lonicara japonica * Thunb. ex Murray
PLANTAGINACEAE
Plantago lanceolata * L.
CAMPANULACEAE
Wahlenbergia graniticola Carolin
W. stricta (R. Br.) Sweet

W. communis Carolin
LOBELIACEAE

Isotoma axillaris Lindl.

Pratia purpurascens (R. Br.)F. Wimmer.
STYLIDIACEAE

Stylidium graminifolium Willd.
S. laricifolium Rich.

S. lineare Sw. ex Willd.
ASTERACEAE

Bidens pilosa* L.

Brachycome aculeata R. Br.

B. angustifolia Cunn. ex DC.

44

an

an

les] [esl lest eal

Rare, host Eucalyptus spp.

Occaisional
Rare, moist gullies

Rare

Occasional, moist slopes

Occasional

Rare

Rare, moist, shaded positions

Occasional, esp. shaded slopes

*

* near water

Occasional esp. disturbed sites

Occasional

*

*

*

Occasional esp. moist sites

Occasional esp. open sunny sites

*

*

*

abundant

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.

H. 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

Solanum 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.

Proc. Linn. Soc. N.S.W., 128, 2007

or

ey jE, SE] jae; acy ae, AN WA A AN ay, fay jac) WANjae) foe} jan} jae) fae} jae, WN ale

en)

aqH

A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E
A,E

A,E

A,E
A,E
A,E

Occasional, shady slopes

*

Occasional, shady slopes
*

Weed on disturbed sites
Weed on disturbed sites

*
*

*

Occasional esp. gullies, very moist
slopes
Occasional

Abundant esp. moist disturbed sites
Shady slopes

Occasional, mostly shady slopes

Pa
*

*

Occasional
Occaisional

Weed, on disturbed ground

*

Weed, on disturbed ground

*

*

Occasional, esp. moist areas

Occasional esp. lake margins

*

Rare, floating on open water > 2m deep

Occasional, disturbed sites

45

46

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

aqH

TW
TW

ae, SE} jae] SB] ae, a} ac} ja

lesl esl lesl les! les! |esl es)

Rare in damp places

Occasional

*

Occasional, only northern end of park

Occasional

Occasional, moist slopes

Occasional, moist slopes

Rare, moist slopes only

Occasional

Occasional

*

*

Occasional

Abundant esp. footslopes
Rare

Abundant

Occasional along ridgetop plateaux

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.
Johnson & D.F. Cutler

Lepyrodia mulleri Benth.

L. scariosa R. Br.

CYPERACEAE

Baumea arthrophylla (Nees) Broeck.
B. teretifolia Palla

Baumea sp. nov.

Bolboschoenu fluviatilis (Torrey) Sojak
Caustis flexuosa R. Br.

Cyperus laevis R. Br.

Eleocharis atricha R. Br

E. sphacelata R. Br

Isolepis inundatus Hook. f.

Proc. Linn. Soc. N.S.W., 128, 2007

lesl Jes} Jes} les! {es} [esl les! les! esl esl esl
n

> > > YS

Occasional, moist slopes
*
Occasional, moist slopes

Occasional, only moist places

*
*
Occasional, shady rock outcrops
*
*

*

Occasional, lake margins

Occasional, lake margins

*

Ea

*
*
*

*

Occasional, rocky slopes, plateau tops
Occasional, moist sites
Occasioal

Abundant, in open water

*

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,) P. Beauv.
Entolasia marginata (R. Br) Hughs

E. stricta (R. Br) Hughs

Eragrostis leptostachya Steud.

Imperata cylindrica P. Beauv. var.
major (Nees) C.E. Hubb.

Microlaena stipoides (Labill.) R. Br
Panicium simile Domin
Paspalidium gracile (R. Br.) Hughes
Paspalum dilatatum ~* Poi.
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.

48

©
+2
an)

ee) SE, acy SE, SE, SE, SE] BE, SE, SE) SE) AE, SEY SE] FEY SE, YE, SE, SE] SE) a

A,E
A,E
A,E
A,E
A,E
A,E

ies]

rrr rrr re F&F Fe eee ere er 5e F&F ee PS

Occasional

Abundant, margins of lakes

Very abundant, mostly open water
Occasional, lake margins

Occasional, lake martins

*

Occasional, open water only

*

Occasional

*

*
*
*

Occasional esp. footslopes

**

Occasional esp. open areas

*
*

*

Abundant esp. after burning

*
*

*

Occasional esp. disturbed footslopes

*

Occasional

*

Abundant

Occasional, damp places

Proc. Linn. Soc. N.S.W., 128, 2007

S. ROSE AND H.A. MARTIN

DICKSONIACEAE

Calochlaena dubia (R.Br.) M. Turner Ones sonal
& R. White nly in very moist guilies

CYATHACEAE

Cyathea sp. Rare, moist gullies
DENNSTAEDTIACEAE

Hypolepis muelleri N.A. Wakef. Occasional, moist creek banks

eniatieia sou lena (HONS 6) Abundant, esp. disturbed areas

Cockayne

LINDSAEACEAE

Lindsaea microphylla Sw. Rare, moist gullies
ADIANTACEAE

Adiantum aethiopicum L. Occasional, moist gullies

A. hispidulum Sw. <

SINOPTERIDACEAE

Cheilanthes distans (R.Br.) Mett.

C. austrotenuifolia Quirk & Chambers Occasional, esp. rock outcrops
DAVALLIACEAE

Davallia pyxidata Cav. Rare, on rockfaces in moist areas
BLECHNACEAE

Blecknum cartilagineum Sw. abundant, rocky, shaded slopes
Doodia aspera R.Br. Occasional, moist slopes

APPENDIX 1B

The species in the gully forest (site TS 6, Fig. 1).
a ee ee gg Presence Oulsidg

Species Family

fa ae ee 40 | LO

Trees, 10-30 m
Eucalyptus deanei Maiden Myrtaceae -

E. elata Dehnh. : =
E. piperita Sm. es ats

Small trees and shrubs < 10 m
Doryphora sassafras Endl. Monimiaceae -

Grevillea mucronata R. Br. Proteaceae Te
Hakea salicifolia (Vent.) B.L. Burtt. 4 =
Lomatia silaifolia (Sm.) R. Br ze
Persoonia levis (Cav.) Domin.

P linearis Andrews. ff

+ + +

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION HISTORY OF DRY LAKE, NSW

P. mollis R. Br.

Stenocarpus salignus R. Br.
Pittosporum revolutum Dryand.
Elaeocarpus reticulatus Sm.
Lasiopetalum ferrugineum Sm. vat, 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.

50

oe

19

Pittosporaceae
Elaeocarpaceae
Sterculiaceae
Euphorbiaceae

Cunoniaceae

oe

oe

Fabaceae

Casuarinaceae
Rhamnaceae

(T4
Santalaceae
Rutaceae

(14
Sapindaceae
Aralaceae
Ericaceae

(<4
Loganiaceae
Oleaceae
Myrsinaceae
Goodeniaceae

Asteraceae

Violaceae
Droseraceae
Solonaceae

Orchidaceae

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.

Sticherus 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.

Proc. Linn. Soc. N.S.W., 128, 2007

Cyperaceae

Dilleniaceae

Cassythaceae
Menispermaceae
Smilacaceae
Vitaceae

Luzuriagaceae

ce

Cyatheaceae

Osmundaceae
Gleicheniaceae
Hymenophyllaceae
Cyatheaceae
Dennstaediaceae
Adiantaceae
Sinopteridaceae
Polypodiaceae
Aspleniaceae

Blechnaceae

(73

3

Dil

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 um (mean in brackets), with x (rarely w)
type pore and a thick exine, 2-4 um.

Eucalyptus group: Medium-sized grains, 18 (21-25) 28 um, with x (rarely w) type pore and medium-thick-
ness exine, 1.5-3.0 um.

Melaleuca/Leptospermum group: Small sized grains, 10 (13-19) 23 um, with only y and z type pore and thin
exine, <1-1 ym.

A POLAR VIEW
Rounded angle of amb

Colpus Biunt angle of amb

yy

B TYPE OF POLE

Jeihek
fet

© PORE TYPES

D SIDES OF AMB

ae hee

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.

‘ Equatorial diameter Type of Type of Amb Amb Exine . ae
Species ; thickness
Mean (um) Range pole pore angle sides pattern
Angophora/Corymbia pollen group
Angophora floribunda 3041.7 (26-33) f(e)a x(w) round straight faint 2-3
Corymbia eximia 40.742.8 (35-45) a,c(f) x round straight - 2-4
C. gummifera 40.043.0 (33-45) b,a(f) x round convex - 2.5
Eucalyptus pollen group
straight &
Eucalyptus punctata 23.841.3 (18-27) a w (x) round - 1.5-2
concave
ioe straight &
E. piperita 21.0£1.3 (18-24 a x round
concave
E. tereticornis 23.4+£1.3 a x round straight - 1.5-2
: straight &
E. globoidea 24.842.0 (20-28) a x (w) round - 1.5-3
concave
Melaleuca/Leptospermum pollen group
Melaleuca thymoides 19.641.4 (18-23) f (a) Z round concave - to 1
M.linariifolia 15.4+1.4 (13-18) f Z round concave - to 1
Lept
eS Pe aie 13.0+1.8 (10-19) f Z round concave - to 1
juniperinum
: coarse,
L. trinervium 13.5+1.4 (12-16) f y blunt straight to 1
granular
Tristaniopsis sp. straight &
ee a 14.441.1 (12-17) az blunt : 3 tol
laurina concave
straight &
Acmena smithii 14 f - round - to 1
concave
Backhousia myrtifolia 18.0 (14-21) fi 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 1A

Probable source in present day vegetation
Pollen type

Pollen types found on both surface sample and fossil pollen diagrams.

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 cf varia
Chenopodiaceae
Caryophyllaceae
Brassicaceae
Asteraceae Tubuliflorae
Asteraceae Liguliflorae
Zygophyllaceae
Polygonaceae
Gonocarpus

Trilete spores
Monolete spores
Sellaginella
Hydrocotyle

54

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.

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

Proc. Linn. Soc. N.S.W., 128, 2007

Platysace linearifolia
Leucopogon spp

Liliaceae (sensu. lat.)
Exocarpus spp.

Scaveola ramossisima
Parsonsia straminea
Portulaca oleracea
Mistletoe on Eucalyptus spp.

Lomandra spp

35

Pollen type nate oat pollon-Cimginuna wool OBS E PS

cus, ce Appendix 1A

PollSHBS nist oratieg mulunsig seo dtiw |
Patien (ypes found an both surface sacophossmnbipwnik RGURD diagrams.

Fam pelyyran

d sry whoru' (or miia

ipishea {Vw aervinyt

Ba vig fh)
rontnoratie 2 (15 um grate)

Prmeleg

1 bed

Poarear

Mantago cl loncesiata

j Teint e: Varia
Chen 4x) NWaceae
ary OM vlinc cae
Brassicaceae
Asteracese Tuduliflarse

Liguiliflarae

Asiermcrat
Zygophyliacead
Polyponaceae
Gonovanpus nig?
Trilete spores
Monoicte spares

Secllerginelle

Hydrocotyle

=

— ee tlhe lS

muriloyinns rulkelqorvad

Vt! Mae Vhat yi Ou oats
wat i tg CLR ‘. aN in preset

if a Ss ui rrsqun

2uisTy bo iptralainy mactien pA
A forilunda, C. eximig, C gummifera

Me ial STRD gallog slamaee, 228t tus, 89, 29600.

Tristaniopd:s 8%, Backhouria mavertifial tea
oily foot tstontent
ny other species in the fantily
Fiat Ai At A Mivoral
Bimes my ny mee} likely P radiata

Natt 74 C iris or (her imtoduced species

Bs ep wend ond

at cies in the family, excluding Banksia 6p.
Merve.

pe

curnadeng giana

akin sooo tee pasate t eochesing Acacia,
Dilleniaceac, Goodenia hederaceae, Ampera
Violoceac, Bursaria spinosa, Stylidiaceag
Mainly B/aocarpus reviculatue, Ceratopetalion spp.
Pimeleu lintfolia )
Acacia spp,

Poaceag kpecies

Mlantayo lancamiata (introduced):

“yi -

(tal

Plamtage varia (aative)
Chenopediacese species (not in Appendix 1) mane
Caryephyacead, as above re bs
Brawsicacond, as above

Astermocae species, excluding Hypochoeris radiata.
Probatly only 2fypockoeris raddiciata

Probmitty Tribulus ferrestris, but the plant was not obs
Doge 4 on-time — P Aydropipan;: 7 orientale

ner hh ia
chitersy Parties, hence menial n

Mortem app, Davellia sp,, Doodia sp. * ihe
Svlhenginiiting altyinnnnd "ak Main «Werte we!

at

The History of the Vegetation from the Last Glacial Maximum

at Mountain Lagoon, Blue Mountains, New South Wales.

ANTHONY Ropsste! AND HELENE A. MARTIN?

'St. James College, 25 Mary St Cygnet, Tas. 7112
* 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 spectra 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 18 October 2006.

KEY WORDS: 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 xerophytic 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 11,000
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

Mt. WilsoneT
r

A
Lake Baraba
Wollongong

Great

<=

SYDNEY BASIN

prot

Vv Westerm_,
Katoomba \-7

fast, Bilpip La

. 7

~ —-7 5. Bey...
OG ~=.s US Ling

ZiSd,

Bs
Bo

tA
ZY Mt Tomah
7

Richmond

/'® Springwood
oe > }
Se tiaKe Wye € ;
~wv Penrith Lakes
@

“EE
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).

58

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
etal., 1996). 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 51-70 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

Proc. Linn. Soc. 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 from
20,000 to14,000 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 1s 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

Proc. Linn. Soc. N.S.W., 128, 2007

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 from 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 from 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 from 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

59

VEGETATION HISTORY OF MOUNTAIN LAGOON

=
[22] Lagoon [ {| Exposed escarpment
[i] Melaleuca Forest [IT] Protected escarpment
[L]] Tall Forest ia Warm temperate rainforest

L] Cleared

Figure 2. Vegetation survey sites (numbers) and
map of the vegetation types.

[- J] Hilltop vegetation

(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

60

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. 0 coverslips.

Pollen was identified by comparison with modern
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 180 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 type 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 4
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.

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 Calib Rev5.0.2 (Stuiver and Reimer, 1986-2005)

Depth (cm) Sample number Technique io apeon ee ee Sea ee
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 AMS C"* 19,350 + 220 23,036
90-100 OZD667 AMS C¥# 19,700 + 390 23,484

3. The tall forest has a 60-70%
canopy cover of Eucalyptus deanei
and a sub-canopy of Syncarpia
glomulifera and Pittosporum

NW
SE e Depth (cm)

20
undulatum. Leucopogon spp. and 3
climbers such as Smilax form lal
much of the understorey in this 60
forest and a variety of ferns form a

thick ground cover. In the cleared 2
areas to the north-east of the 100

onl Excavation
(] Deposition
@ Core site

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.Ontheexposed, 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. Figure 3. Mountain Lagoon, depicting disturbed sites, strati-
Banksia spinulosa and Telopea graphic transects and the site of the core for this study.
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

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

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
19,3502220

6)

18,660 + 150-7
Dark greyish brown herbaceous
peat and lake mud

8Q

Dark grey lake mud, trace of sand

19,700+360

100 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.

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 beginning 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

Stratigraphy
The stratigraphic transects and cross sections of
the lagoon are shown in Fig. 3. Part of the lagoon

62

that some instability of the fault escarpment probably
contributed to the rapid rate of clay accumulation at
the base of the profile.

Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN

ENVIRONMENT ZONE VEGETATION

Sedgeland

Casuarinaceae/Myrtaceae

Cc woodland/forest

40

Sedgeland

Shrub/herblands

Age, ka

| Radiocarbon date

eM 1 eros 2 ial Clay @ Calibrated age
WAS peat on f L A,

Figure 5. Summary diagram of the history of Mountain Lagoon. This model assumes continuous depo-
sition (see text). For sedimentary symbols, see Figure 4.

MICROSCOPIC MACROSCOPIC
CHARCOAL CHARCOAL CARBON CONTENT

ia
9,040 |LS< | . +
++++
y ++
18,660 +
19,350 + Uf +
++
Les : i
19,700 +
oo

— |
20 %

e-_—_—_—" e———
5 x A.m2.kg" X10° 20 mm?/slide

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 5 10 20
Acmena smithii SW) ate 4.9
Angophora costata 38) +f
Corymbia gummifera +
Eucalyptus creba DIAL i: 4.9
E. punctata Sy) ae
E. deanei Te 30.3 42.6
E. piperata 16.4
E. haemostoma Oli
Syncarpia glomulifera =P
Leptospermum spp 5.9 6.1 8.2
Mytyaceae type I 3
Mytyaceae type II 6.1
Mytyaceae type III 13 Jana Gal 11.5
Myrtaceae type IV +
Mytyaceae type V +
Unidentified Myrtaceae PIMA WS TS)
64

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 from 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

50

70

4.0

3) Osc meme

12.0

ar to some extent (Rummery 1983). In

these sediments, fire seems to have
had a greater influence on the SIRM
values than the mineral content.

7.4 16.0

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

14.7 24.0

19.1

32.0

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. 7: 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.

Zone D: 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

Proc. Linn. Soc. N.S.W., 128, 2007

Scale bar =5 um

Figure 7. Common, unidentified Myrtaceae pollen

types.

2006). The habit of the species cannot be determined
from 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 from 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 from 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. 7
pleistocenicus, (Cassinia arcuata, Calomeria and
probably others) is found throughout the profile and it
may be common in some glacial and older sediments

65

"[ I[QUy, 908 ‘sose pozyeiqyes 10,4 “s1voA

soyeq ‘¢ xIpueddy aas ‘weaserp oy) uo oueU 9d4q UaTjod oY} UL papnysUT solseds 104 “p SI 208 ‘s[OquIAS dISO[OUIT] 10,7 “SuUOT

UOGAVIOIPL.A UI USAIS o1¥

66

4Un0d UdT]Od [e}0} JO soseJUeI10d SUIMOYS WeAseIpP UIT[Od § eNSIy

-v.1)U99U09 UaT[od [v}0} pue sadA} ueT[od UOUTUIOD soul dy} 10J SUOIL.ZUEDU0D UdT[od

, woysures6 0o0'0g HS

JUNOD UAI|Od [e}0} jO%07 ———

O6E F 00Z'61

VEGETATION HISTORY OF MOUNTAIN LAGOON

A
@o@
c*)
on
fo)
ba

+

+

TF? 7 7 . woysuei6 000'008

HY 19 fe 1
a6 %
2
ry re %¢ tebe,

x

oe
2

Oy

YS
& ee ee
&
&
Roe | | é
S8NYHS/SSsyL

x=
m
D
D

a

—ese =

@=---——

+
o Q
Os
FS ls
mo
l=
2
|
fe)
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 fragments
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

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
common 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
fringe 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.
yt, 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

Proc. Linn. Soc. N.S.W., 128, 2007

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 fresh 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 climatic interpretation for the changes at
Mountain Lagoon is somewhat uncertain. A lake
implies water too deep for rooted swamp plants. It
is estimated that rainfall would have been up to half
of the present values during the last glacial period
(Thom et al. 1994; Allan and Lindsay 1998) and
the warmest month was up to 9 °C less than today,
(Galloway 1965; Allan and Lindsay 1998). With
a lower rainfall and less evaporation, and if the
dominants were shrubs, evapotranspiration would be
less also, then free water may have been available for
the lake. Mottling indicates a fluctuating water table
and there were probably dry spells when the lake
dried up, but probably not long enough for sedgelands
to become established. Perhaps water balance was
too variable for the development of sedgelands.
However, sedgelands colonized the lagoon surface
during the glacial period, inferring that the water
balance had become favourable or stable enough for
rooted vegetation, at a time that other sites record dry
and cold conditions for southern Australia. Moreover,

67

VEGETATION HISTORY OF MOUNTAIN LAGOON

the sedgelands may have persisted from the height of
the glacial period to the beginning of the Holocene,
and the pollen concentration and amount of organic
matter suggest a quite productive ecosystem.

The climatic tolerances of most of the myrtaceous
species identified are presented in Table 3 and it can
be seen that the ranges of mean annual precipitation
under which these species are found is quite large. If
the rainfall of Mountain Lagoon is halved, as would
have been likely during the glacial maximum, then the
site would have been at or close to the lower limits for
all of the Angophora/Corymbia/Eucalyptus species.
Consequently, all of these species could have been
present throughout the glacial period. Unfortunately,
temperature data are not available for a similar
analysis since the closest meteorological stations are
at such different altitudes.

In reviewing the studies of dune building in
southeastern Australia, Thom et al. (1994) found that
extensive aeolian deposits in the Shoalhaven River
Catchment were dated to two periods of dune building
between 19,000-6,000 yr BP, with a period of stability
between 18,000 and 14,000 yr BP. This evidence
implies an increase in vegetation cover during this
period, and there may have been fluctuations in
climate within the overall general trends. Mountain
Lagoon data would support this view.

DISCUSSION

There are three other sites in the Sydney Basin
with a vegetation history going back to the last
glacial period. Lake Baraba, one of the Thirlmere
Lakes (Fig. 1), is an upland fluviatile system
contained in an entrenched meander (Timms 1992).
Here, Casuarinaceae was dominant from >34 kyr
to the Holocene, when the sediments were clay.
Myrtaceae became co-dominant about 8 kyr, when
peat formation began (Black et al. 2006). In contrast,
both Casuarinaceae and Myrtaceae are prominent
through the glacial period and the whole of the time
at Mountain Lagoon. Sclerophyll communities with
minimal Poaceae were present at Lake Baraba the
whole time, just as they are at Mountain Lagoon.

The second site, Penrith Lakes (Fig.1), was an
abandoned meander in the flood plain of the Nepean
River (Chalson 1991). Clay was deposited from
>33 kyr, changing to peat only about 3 kyr. During
the glacial period, there were minimal Myrtaceae
and Casuarinaceae, some Poaceae and relatively
abundant T. pleistocenicus, the Cassinia arcuata type.
The shrubby C. arcuata may become abundant on
disturbed mineral soils, and this habitat was probably
common on the floodplain, which would have been
a shrubland at the time of the glacial period. Some

Table 3. Climatic tolerances of Myrtaceous species identified in sediments from Mountain Lagoon
(Boland et al. 2002) and climatic data for the closest meteorological stations. Richmond (average of Uni-
versity of Western Sydney and RAAF base), Katoomba (BoM 2006) and Mountain Lagoon (Hungerford
1995. The altitude for the meteorological stations is included.

Species Mean max. hottest Mean min. coldest Max. frost Mean annual
month (°C) month (°C) days/year ppt (mm)
Acmena smithii 26-32 5-15 few 1000-2000
Angophora costata 25-35 0-8 0-50 600-1200
Corymbia gummifera 24-32 1-8 0-30 700-1800
Eucalyptus creba 26-36 0-17 0-50 550-2000
E. punctata 26-33 1-6 0-40 700-1200
E. deanei 25-30 0-5 0-50 750-1500
Richmond, alt.~ 20m 24 10-11 800-810
Mt. Lagoon, alt ~ 500 m L257
Katoomba, alt. ~ 1040 m 16.6 7.9 1400

68

Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN

Myrtaceae and Casuarinaceae are found during the
upper Holocene, and Poaceae became prominent,
unlike Mountain Lagoon.

The third site, Redhead Lagoon (Fig.1), a now
near-coastal site south of Newcastle (Williams et al.
2006) has a long record. During the height of the
glacial period, there was a very high Casuarinaceae
pollen content with lesser amounts of Angophora/
Corymbia, and Eucalyptus. The environment was not
treeless, although it is possible that the Casuarinaceae
may have been small trees or shrubs. It is thought
that local conditions may have enhanced the moisture
relationships (Williams et al. 2006).

The rapid rate of clay deposition during the glacial
period at Mountain Lagoon is an unusual feature,
but it is recorded in at least one other site. Burraga
Swamp on the Barrington Tops (Sweller and Martin
2001) is situated at the base of Mount Lumeah hence
has a topographically similar situation to Mountain
Lagoon. There, from from 38-21,000 years BP., the
rate of sedimentation was low. Then followed a much
higher rate of sedimentation during the height of the
glacial period, attributed to catchment instability
caused by periglacial activity. Mountain Lagoon was
probably at or above the treeline during the glacial
period (Hope 1989) and if the vegetation cover was
disrupted by the harsh climate of the glacial period, it
may have caused some instability of the easily eroded
Wainamatta Shale escarpment, with a consequent
higher rate of sedimentation in the lake. Movement
along the Kurrajong Fault line, even if slight, would
be another cause of instability that could occur
at any time and contribute to an increased rate of
sedimentation.

The deposition of peat during the glacial period
is also unusual. As discussed previously, clay was
being deposited at both Penrith Lakes (Chalson
1991), and Thirlmere Lakes (Black et al. 2006), and
silt at Redhead Lagoon (Williams et al. 2006), the
only other sites in the Sydney Basin with records
going back to the last glacial maximum. Further
afield, at Lake George (Singh and Geissler 1985)
and Burraga Swamp on the Barrington Tops (Sweller
and Martin 2001), inorganic sediments were being
deposited during the last glacial period. On the
Barrington Tops, the change from inorganic to peat
sedimentation occurred at least close to or during the
Holocene (Dodson 1987). Mountain Lagoon must
have been particularly favourable for plant growth
during the glacial period. Perhaps its location on
the western side of the Kurrajong block, which may
have provided some protection, was advantageous.
The lagoon would have received run-off from the
surrounding slopes, and the warming of the Block by

Proc. Linn. Soc. N.S.W., 128, 2007

the afternoon sun may have meant that temperatures
were less extreme. In any case, a comparison of the
range of precipitation where the dominant species are
found today and the probable precipitation during the
glacial period show that they could have been present
at Mountain Lagoon through the height of the glacial
period, albeit at the lower end of their range. The
shrubs and herbaceous species found at Mountain
Lagoon are mainly widespread taxa and the results of
this study would suggest that they too existed at the
site through the glacial maximum.

A decline in Casuarinaceae and its replacement
with Myrtaceae about mid Holocene time may be
found in a number of sites, and likely causes for this
feature have been suggested, as discussed in Rose
and Martin (this volume). Mountain Lagoon does
not show any decline in Casuarinaceae and both
Casuarinaceae and Myrtaceae were well represented
the whole time. Casuarinceae was not found in the
survey of the vegetation for this study, hence its only
decline would have been the result of logging by
European settlers. The wood of Casuarinaceae was
prized by Europeans for firewood, building and tool
making (Entwisle 2005).

The history of the vegetation at Mountain Lagoon
suggests very little change in the vegetation from the
glacial period to the present and even some to the
Eucalyptus species are found throughout the profile.
Unfortunately, the palynology cannot determine if
a species assumed a different lifeform during the
glacial period. The only change in the vegetation is
associated with the change from lake to sedgelands
(Fig. 5), controlled by hydrological changes. Even
these changes do not fit the traditional view of a harsh
glacial climate, slowly improving to a climate like
the present about the time of the Holocene. Mountain
Lagoon may either have been a refugium or there
was more variation in the vegetation during the last
glacial period than previously thought. As discussed
previously, evidence from ‘coastal’ dunes during the
last glacial period (Thom et al. 1994) suggested that
there was a ‘...greater concentration of forests in more
discrete, protected sites along the eastern escarpment
than was previously considered by palaeoecologists’,
and this view may be applicable to other regions such
as the Blue Mountains.

In a study of aeolian dunes on the Newnes
Plateau in the Blue Mountains (altitude 1000 m),
Hesse et al. (2003) came to the conclusion that
unrealistically drier conditions were necessary to
allow wind transport at this site. They have suggested
that it would require additional impediments to
plant growth, such as lower temperatures and lower
atmospheric carbon dioxide concentrations during

69

VEGETATION HISTORY OF MOUNTAIN LAGOON

the height of the glacial period to disrupt the sparse
vegetation and allow the necessary conditions for
dune formation. This interpretation is at odds with the
story from Mountain Lagoon. Today, periodic drought
will disrupt the vegetation and allow inactive dunes to
become mobile, and this could have happened during
the height of the glacial period also. The two sites
are not comparable: palynology requires sites which
remain permanently wet and are thus probably the
most hydrologically favourable in the landscape, in
contrast to sites that allow aeolian transport, such as
an exposed plateau. There must have been a mosaic of
environments during the glacial period, just as there
is today and these two studies have sampled different
environmental settings.

CONCLUSIONS

The rapid accumulation of clay in the lake at
22-23 k cal. yr is thought to have been caused by
vegetation and soil instability, the result of the harsh
climate at the height of the glacial maximum.

Sedgelands colonized the lake surface about 22 k
cal. yr and then followed peat deposition from what
must have been a productive ecosystem, comparable
to today, until the Holocene.

The Lagoon reverted to a lake and clay deposition,
c. 10-8 k cal. yr, probably because of slight instability
of the fault-line and a burst of accelerated erosion.

Sedgelands re-colonized the lake surface again
and remained to the present day.

The dryland vegetation appears remarkably
similar through the whole time: it was sclerophyllous
shrubland/woodland/forest.

When Myrtaceae grains are identified to species,
they show that there has been change in the species,
but some species have been present the whole time.

The species found in the locality today could
have been present through the glacial period, albeit at
the lower end of their range of precipitaiton.

Both Casuarinaceae and Myrtaceae are prominent
the whole time, and Casuarinaceae only declines with
European settlement.

ACKNOWLEDGEMENTS

We are indebted to an ANSTO grant, 98/069R, for
carbon dating and a grant from the Joyce W. Vickery
Research Fund to make this project possible. We would
like to thank Prof. John Dodson who suggested we study
Mountain Lagoon and who supplied a core. Mr. P. Hind
generously assisted with the flora of the region. We
appreciate the encouragement and assistance provided by

friends and colleagues.

70

REFERENCES

Allan, R., Lindsay, J. (1998). Past climates of Australasia.
In: ‘Climates of the Southern Continents’ (Eds J.E.
Hobbs, J.A. Lindesay, H.A. Bridgman) pp. 208-247.
(Wiley & Sons, Chichester).

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).

Birks, H.J.B. and Birks, H.H. (1980) ‘Quaternary
Palaeoecology’ (Edward Arnold: London)

Black, M.P., Martin, H.A.. and Mooney, S.D. (2006). A
> 43,000 year vegetation and fire history from Lake
Baraba, New South Wales. Quaternary Science
Reviews 25, 3003-3016.

Black, M.P. and Mooney, S.D. (2006). Holocene fire
history from the Greater Blue Mountains World
Heritage Area, New South Wales, Australia:
the climate, humans and fire nexus. Regional
Environmental Change 6, 41-51.

Boland, D.J., Brooker, M.I.H., Chippendale, G.M. et al.
(2002). ‘Forest Trees of Australia’ Fourth Edition.
(CSIRO Publishing: Melbourne)

Botanic Gardens Trust (2006). Mount Tomah Botanic
Gardens website http://rbgsyd.nsw.gov.au/mount
tomah_botanic_gardens. Accessed 10-2-2006.

Branagan, D.F. (1969). The Lapstone Monocline and
associated structures. In “The Proceedings of
the Advances in the study of the Sydney Basin
Symposium’ pp. 61-62. (Dept. of Geology, University
of Newcastle: Newcastle).

Chalson, J.M. (1991). Late Quaternary vegetation and
climatic history of the Blue Mountains, N.S.W.,
Australia. PhD Thesis, University of New South
Wales.

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 115,163-191.

Churchill, D.M. (1960). Living and fossil unicellular algae
and aplanospores. Nature 186, 493-494.

Clarke R.L. (1984). Point count estimation of charcoal in
pollen preparations and thin sections of sediment.
Pollen et Spores 24, 523-535.

Conyers, (1987). The Aborigines of the Mount Tomah
District. In “The Mount Tomah Book’ pp 12-14.
(The Mount Tomah Society and the Royal Botanic
Gardens: Sydney).

David, T.W.E. (1902). An important geological fault at
Kurrajong, N.S.W. Proceedings of the Royal Society
of New South Wales 36, 359-370.

Dodson, J.R. (1987). Mire development and environmental
change, Barrington Tops and Upper Hunter regions of
New South Wales. Quaternary Research 27, 73-81.

Dodson, J.R. (1994). Quaternary vegetation history. In
“Australian Vegetation, Second Edition’ (ed. R.H.
Groves) pp. 37-56. (Cambridge University Press:
Cambridge).

Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN

Edney, P.A., Kershaw, A.P. and de Deckker, P. (1990).

A Late Pleistocene and Holocene vegetation and
environmental record from Lake Wangoom, Western
Plains of Victoria, Australia. Palaeogeography,
Palaeoclimatology, Palaeoecology 80, 325-343.

Entwisle, T. (2005). She-oak up in smoke. Nature
Australia Spring 2005 28(6), 72-73.

Floyd, A.G. (1989). “Rainforest Trees of Mainland South-
eastern Australia’. (Inkata Press: Melbourne).

Galloway, R.W. (1965). Late Quaternary climates in
Australia. Journal of Geology 73, 603-618.

Grady, A. and Hogbin, H. (1926). Mountain Lagoon and
the Kurrajong Fault. Proceedings of the Royal Society
of New South Wales, 60, 119-126.

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 Press: Sydney).

Hesse, P.P., Humphreys, G.S., Selkirk, P.M. et. al.

(2003). Late Quaternary aeolian dunes on the
presently humid Blue Mountains, Eastern Australia.
Quaternary International 108, 13-32.

Hope, G.S. (1989). Climatic implications of timberline
changes in Australasia from 30,000 years to
present In “CLIMANZ 3: Proceedings of the Third
Symposium on the Late Quaternary History of
Australasia’ (eds. T.H. Donnelly and R.J. Wasson) pp.
91-99. (CSIRO: Melbourne).

Hope, G.S. (1994). Quaternary vegetation. In * History
of the Australian Vegetation: Cretaceous to Recent’
(Ed. R.S. Hill), pp. 368—389. (Cambridge University
Press: Cambridge)

Hungerford, M. (1995). “Bilpin the Apple Country,
including Mount Tomah, Mount Tootie and Mountain
Lagoon: a Local History’. (University of Western
Sydney-Hawkesbury: Richmond NSW.)

Livingstone, D.A. (1955). A lightweight piston sampler
for lake deposits. Limnology and Oceanography 12,
346-348.

Macphail, M.K. and Martin, T. (1991). ‘Spineless
Asteraceae’ (episode two). Palynological and
Palaeobotanical Association of Australasia
Newsletter 23, 1-2.

Martin, H.A. (1973). The palynology of some Tertiary
Pleistocene deposits, Lachlan River Valley, New
South Wales. Australian Journal of Botany,
Supplementary Series 6, 1-57.

Maher, L.J. (1972). Nomograms for computing 0.95

confidence limits of pollen data. Review of

Palaeobotany and Palynology 13, 85-93.

Moore, P.D., Webb, J.A., Collison, M.E. (1991). “Pollen

Analysis, Second Edition’. (Blackwell Scientific

Publications: London).

Neale, J.L. and Walker, D. (1996). Sampling sediments
under warm deep water. Quaternary Science Reviews
15, 581-590.

Pickett, E.J., Harrison, S.P., Hope, G. et al. (2004). Pollen-
based reconstructions of biome distributions for
Australia, Southeast Asia and the Pacific (SEAPAC

Proc. Linn. Soc. N.S.W., 128, 2007

region) at 0, 6000 and 18,000 C yr BP. Journal of
Biogeography 31, 1381-1444.

Pentecost, A. (1984). “Introduction to Freshwater Algae’.
(Richmond Publishing Co Ltd.: Richmond, England).

Plantnet (2005). National Herbarium website http://
plantnet.rbgsyd.nsw.gov.au/, accessed April 2006.

Rose, S. and Martin, H.A. (this volume), The vegetation
history of the Holocene at Dry Lake, Thirlmere, New
South Wales. Proceedings of the Linnean Society of
New South Wales.

Rummery, T.A. (1983). The use of magnetic
measurements in interpreting the fire histories of lake
drainage basins. Hydrobiologia 103, 53-58.

Ryan, A., Fisher, M. and Schaeper, L. (1996). The natural
vegetation of the St. Albans 1:100,000 map sheet.
Cunninghamia 4, 433-482.

Singh, G. and Geissler, E.A. (1985). Late Cainozoic
history of vegetation, fire, lake levels and climate
at Lake George, New South Wales, Australia.
Philosophical Transactions of the Royal Society of
London B 311, 379-447.

Stockton, E. (1993). Archaeology of the Blue Mountains.
In ‘Blue Mountains Dreaming — The Aboriginal
Heritage’ (ed. E. Stockton) pp.23-55. (Three sisters
Production: Winmalee, N.S.W.)

Stuiver, M. and Reimer, P.J. (1986-2005). Radiocarbon
calibration program Calib. Rev 5.0.2. http://calib.qub.
ac.uk/calib/calib.html (accesed May 2006).

Sweller, S. and Martin, H.A. (2001) A 40,000 year
vegetational history and climatic interpretation of
Burraga Swamp, Barrington Tops, New South Wales.
Quaternary International 83-85, 245-256.

Taylor, Griffith (1970). ‘Sydneyside Scenery, 2™¢ Edition’.
(Angus and Robinson: Sydney).

Thom, B., Hesp, P. and Bryant, P. (1994). Last glacial
“coastal” dunes in Eastern Australia and implications
for landscape stability during the Last Glacial
Maximum. Palaeogeography, Palaeoclimatology,
Palaeoecology 111, 229-248.

Thompson, R. (1990). Palaeomagnetic dating. In
“Handbook of Holocene Palaeoecology and
Palaeohydrology’ (Ed. B.E. Berglund.) pp. 313-327.
(Wiley and Sons: Chichester).

Thompson, R. and Oldfield, F. (1986). “Environmental
Magnetism.’ (Allen and Unwin: London).

Timms, B.V. (1992). ‘Lake Geomorphology” (Gleneagles
Publications: Adelaide).

van Geel, B. (1978). A palaeoecological study of Holocene
peat bog sections in Germany and the Netherlands.
Review of Palaeobotany and Palynology 25, 1-120.

Van Geel , B. and Grenfell, H.R. (1996). Spores of
Zygnemetaceae. In “Palynology: Principles and
Applications’ (Eds J. Jansonius and D.C. McGregor)
Vol. 1, pp 173-179. (American Association of
Stratigraphic Palynologists Foundation: Texas).

Williams, N.J., Harle, K.J., Gale, S.J. and Heijnis, H.
(2006). The vegetation history of the last glacial-
interglacial cycle in eastern New South Wales,
Australia. Journal of Quaternary Science 21, 735-
750.

71

VEGETATION HISTORY OF MOUNTAIN LAGOON

APPENDIX 1.

Mountain Lagoon species lists obtained from the vegetation survey. Sites: 1, Lagoon. 2, Melaleuca forest. 3,
Tall forest. 4, Exposed escarpment (full species list not obtained and omitted here. For vegetation description,
see text). 5, Protected escarpment. For site 6, the Green Scrub rainforest, see Appendix 2. For a full list of
species in the region, see Ryan et al. (1996). Nomenclature follows Harden (1992; 1993; 2000:; 2002) and
Plantnet (2005). *Indicates introduced species.

Species Site 1 Site 2 Site 3 Site 4 Site 5
BRYOPHYTA
Sphagnum sp. 5

FERNS/FERN ALLIES

Adiantaceae

Adiantum aethiopicum L. ats

Blechnaceae

Blechnum cartilagineum Sw. °

Dennstaedtiaceae

Hypolepis meulleri N.A. Wakef.

Pteridium esculentum (Forst. f.) Cockayne =F a +
Gleicheniaceae

Gleichenia dicarpa R.br. ats

Pteridaceae

Pteris tremula R. Br oF +
Schizaeaceae

Schizaea dichotoma Sm. +

DICOTYLEDONS

Apiaceae

Hydrocotyle peduncularis A. Rich. +
Araliaceae

Polyscias sambucifolia Harms +
Apocynaceae

Tylophora barbata R. Br. an
Convolvulaceae

Cuscuta australia R. Br. 4.
Cunoniaceae

Ceratopetalum apetalum D. Don | =F +
Dilleniaceae

Hibbertia dentata DC. ar

72 Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN

H.. hermanniifolia DC.
Elaeocarpaceae
Tetratheca ciliata Lindl. +
Ericaceae
Leucopogon ericoides R. Br.
L. juniperinus R. Br.
L. lanceolatus (Sm.) R. Br. + +
Fabaceae - faboideae
Dillwynia retorta Druce
Gompholobium latifolium Sm.
Pultenaea flexilis Sm. +
P. linophylla Schrad. and J.C. Wendl.
Fabaceae - mimosoideaea
Acacia elata Benth.
A. filicifolia M.B. Welch, Coombs & McGlynn
A. longifolia (Andrews) Willd.
Haloragaceae
Myriophyllum latifolium F. Muell.
Myriophyllum cf M. variifolium Hook. f.
Menyanthaceae
Nymphoides geminata (R. Br.) Kuntze =F
Myrtaceae
Angophora costata Britten

Eucalyptus agglomerata Maiden

E. deanei Maiden at
E. elata Debnh.
E. piperita Sm. a

E. saligna Sm.

Leptospermum polygalifolia Salib. =F +
Melaleuca linariifolia Sm. ae
Syncarpia glomulifera (Sm.) Nied. =e
Pittosporaceae

Billardiera scandens Sm.

Pittosporum revolutum Dryand.

P. undulatum Vent. ae
Proteaceae

Banksia serrata L. f.

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION HISTORY OF MOUNTAIN LAGOON

B. spinulosa Sm.

Lomatia silaifolia (Sm.) R. Br.
Persoonia laurina Pers.
Telopea speciosissima R. Br.
Santalaceae

Exocarpus strictus R. Br.
Sapindaceae

Dodonaea triquetra J.C. Wendl.
Violaceae

Viola hederaceae Labill.

MONOCOTYLEDONS
Alismataceae

Alisma plantago-aquatica L.
Damasonium minus Buchenau

Cyperaceae

Baumea articulata (Nees) Broeck.

Lepidosperma laterale R. Br.
L. longitudinale Labill.
Schoenus melanostachys R. Br.
Juncaceae

Juncus usitatus L.A.S. Johnson
Lomandraceae

Lomanara longifolia Labill.
Luzuriagnaceae

Eustrephus latifolius Ker Gawl.
Phormiaceae

Dianella longifolia R. Br.

D. caerulea Sims

Poaceae

*Echinopogon caespitosus C.E. Hubb.

*F. ovatus (G. Forst.) P. Beauv.

Entolasia marginata (R. Br.) Hughs
Microlaena stipoides (Labill.) R. Br.

*Paspalum distichum L.
Smilacaceae

Smilax austalis R. Br.

74

+
+
+
+
+
+ +
- +
+
+
+
+
+
+
+ -
+
4 +
- +
+
+
+
+

Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN
S. glyciphylla Sm.

Xanthorrhoeaceae

Xanthorrhoea arborea R. Br.

Proc. Linn. Soc. N.S.W., 128, 2007

75

VEGETATION HISTORY OF MOUNTAIN LAGOON

APPENDIX 2

Species found in Green Scrub Rainforest, Site 6 (P. Hind, pers. comm.). Nomenclature follows Harden (1992;

1993; 2000:; 2002) and Plantnet (2005). * denotes an introduced species.

FERNS/FERN ALLIES
Adiantaceae

Adiantum aethiopicum L.

A. diaphanum Blume

A. formosum R. Br.

A. hispidulum Sw.

A. silvaticum Tindale (R. Br.) F, e
Pellaea falcata (R. Br.) Fée

P. nana (Hook) Bostock

Aspleniaceae

Asplenium attenuatum R. Br.

A. australasicum (J. Sm.) Hook.
A. flabelllifolium Cav.

Athyriaceae
Diplazium australe (R. Br.) N.A. Wakef.

Blechnaceae

Blechnum ambiguum (C. Presl.) Kaulf. ex C. Chr.
B. cartilagineum Sw.

B. nudum (Labill.) Mett. ex Luerss.

B. patersonii (R. Br.) Mett.

B. wattsii Tindale

Doodia aspera R. Br.

Cyatheaceae
Cyathea australis (R. Br.) Domin.
C. leichhardtiana (F. Muell.) Copel.

Davalliaceae
Arthropteris tenella (G. Forst.) J. Sm. ex Hook. f.

Dennstaedtiaceae

Dennstaedtia davallioides (R. Br.) T. Moore
Histiopteris incisa (Thunb.) J. sm.
Pteridium esculentum (G. Forst.) Cockayne

Dicksoniaceae
Calochlaena dubia (R. Br.) M.D. Turner& R.A.
White

Dryopteridaceae

Lastreopsis acuminata (Houlston) C.V. Morton
L. decomposita (R. Br.) Tindale

L. microsora (Engl.) Tindale

Polystichum australiense Tindale

76

Gleicheniaceae
Sticherus flabellatus (R. Br.) H. St John

Grammitaceae
Grammitis billardieri Willd.

Hymenophyllaceae
Hymenophyllum australe Willd.
H. cupressiforme Labill.

Lindsaeaceae
Lindsaea microphylla Sw.

Osmundaceae
Leptopteris fraseri (Hook. & Grev.) C. Presl.
Todea Barbata (L.) T. Moore

Polypodiaceae
Platycerium bifurcatum (Cav.) C. Chr.
Pyrrosia rupestris (R. Br.) Ching

Pteridaceae
Pteris tremula R. Br.
P. umbrosa R. Br.

Schizaeaceae
Cheilanthes distans (R. Br.) Mett.
C. sieberi Kunze

Tmesipteridaceae
Tmesipteris truncata (R. Br.) Desv.

DICOTYLEDONS
Amaranthaceae
Deeringia amaranthoides (Lam.) Merr.

Aphanopetalaceae
Aphanopetalum resinosum Endl.

Apocynaceae

Melodinus australis (F. Muell.) Pierre
Marsdenia flavescens A. Cunn. ex Hook.
Parsonsia straminea Pichon

Tylophora barbata R. Br.

Araliaceae
Astrotricha latifolia Benth.

Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN

Polyscias murayi (F. Muell.) Harms.

Asteraceae

*Conza albida Willd. ex Spreng].
Olearia tomentosa (J.C. Wendl.) Benth.
Senecio linearifolius A. Rich.

Bignoniaceae
Pandorea pandorana (Andrews) Steenis

Boraginaceae
Ehretia acuminata R. Br.

Austrocynoglossum latifolium (R. Br.) R. Mill.

Caprifoliaceae
Sambucus australasicus (Lindl.) Fritsch

Caryophyllaceae
Stellaria flaccida Hook.

Chenopodiaceae
Einadia hastata (R. Br.) J. Scott

Convolvulaceae
Calystegia marginata R. Br.
Dichondra repens J. Forst. & G. Forst.

Cunoniaceae

Callicoma serratifolia Andrews
Ceratophyllum apetalum D. Don
C gummiferum Sm.

Schizomeria ovata D. Don

Dilleniaceae
Hibbertia dentata R. Br. ex D.C.
Hibbertia sp.

Ebenaceae
Diospyros australis (R. Br.) Hiern

Elaeocarpaceae
Elaeocarpus reticulatus Sm.

Ericaceae
Acrotriche divaricata R. Br.
Trococarpa laurina R. Br.

Euphorbiaceae

Breynia oblongifolia F. Muell.
Claoxylon australis Baill.
Omalanthus populifolius Graham

Eupomatiaceae
Eupomatia laurina R. Br.

Proc. Linn. Soc. N.S.W., 128, 2007

Fabaceae

Desmodium varians (Labill.) C. Don
Glycine clandestina J.C. Wendl.
Indigofera australis Willd.
Kennedia rubicunda Vent.
Pultenaea flexilis Sm.

Gesneriaceae
Fieldia australis A. Cunn.

Geraniaceae
Geranium homeanum Yurcz.

Goodeniaceae
Goodenia ovata Sm.

Lamiaceae

Chloanthes stoechadis R. Br.
Plectranthus parviflorus Willd.
Prostranthera rhombea R. Br.
Teucrium corymbosum R. Br.

Malvaceae

Abutilon oxycarpum (F. Muell.) F. Muell. ex Benth.

Howittia trilocularis F. Muell.

Meliaceae
Toona ciliata (F. Muell.) Harms

Menispermaceae
Sarcopetalum harveyanum F. Muell.

Stephania japonica (Thunb.) Miers var. discolor

(Blume) Forman

Mimosaceae

Acacia elata Benth.

A. implexa Benth.

A. maidenii F. Muell.

A. oxycedrus Dieber ex DC.
A. parramattensis Tindale
A. saliciformis Tindale

Monimiaceae

Doryphora sassafras Endl.
Hedycarya angustifolia A. Cunn.
Palmeria scandens F. Muell.
Wilkiea huegeliana (Tul.) A. DC.

Moraceae
Ficus coronata Spin
F. rubiginosa Desf. ex Vent.

qi]

VEGETATION HISTORY OF MOUNTAIN LAGOON

Rubiaceae
Myrsinaceae Galium binifolium N.A. Wakef.
Rapanea howittiana Mez. G. propinquum A. Cunn.
R. variabilis (R. Br.) Mez. Morinda jasminoides A. Cunn.
Psychotria loniceroides Sieber ex DC.
Myrtaceae
Acmena smithii (Poir.) Merr. & L.M. Perry Rutaceae
Angophora costata Britten Melicope microccoca (F. Muell.) T.G. Hartley
Backhousia myrtifolia Hook. & Harv.
Eucalyptus agglomerata Maiden Sapindaceae
E. piperata Sm. Guioa semiglauca R. Br.
E. saligna Sm.
Rhodamnia rubescens (Benth.) Migq. Scrophulariaceae
Syncarpia glomulifera (Sm.) Nied. Veronica plebeia R. Br.
Tristaniopsis laurina (Sm.) Peter G. Wilson & J.T.
Waterh. Solanaceae
Solunum aviculare G. Forst.
Oleaceae S. americanum Miller
Notolea ovata R. Br. S. stelligerum Sm.
N. venosa F.Muell. Solanum sp.
Passifloraceae Thymelaeaceae
Passiflora cinnabarina Lindl. Pimelea latifolia R. Br. var. hirsuta (Meissner)
P. herbertiana Ker. Gawl. Threlfall
*P._ subpeltata Ortega
Pittosporaceae Ulmaceae
Bursaria spinosa Cav. Trema aspera (Brongn.) Blume
Pittosporum multiflorum (A. Cunn. ex Loudon) L.
Cayzer, Crisp & I. Telford Verbenaceae
P. revolutum Dryand. Clerodendrum tomentosum R. Br.
P. undulatum Vent.
Violaceae
Plantaginaceae Hymenanthera dentata R. Br. ex DC.
Plantago debilis R. Br. Viola hederacea Labill.
Proteaceae Vitaceae
Persoonia levis (Cav) Domin. Cayratia clematidea (F. Muell.) Domin
P. linearis Andrews Cissus antarctica Vent.
P. pinifolia R. Br C. hypoglauca A. Gray
Stenocarpus salignus R. Br
Winteraceae
Tasmannia insipida R. Br. ex DC.
Ranunculaceae
Clematis aristida R. Br ex Ker. Gawl. MONOCOTYLEDONS
Araceae
Rhamnaceae Gymnostachys anceps R. Br.
Alphitonia excelsa (Fenzl.) Benth.
Arecaceae
Rosaceae Livistonia australia (R. Br.) Mart.
Rubus moluccanus L. var. trilobatus A.R. Bean
R.rosifolius sm. Asteliaceae
Rubus sp. aff. R. moorei F. Mull. Cordyline stricta (Sims) Endl.

78 Proc. Linn. Soc. N.S.W., 128, 2007

A. ROBBIE AND H.A. MARTIN

Commelinaceae
Commelina cyanea R. Br.

Cyperaceae

Carex appressa R. Br.

Cyperus imbicillis R. Br.

Gahnia aspera (R. Br.) Spreng.

G. melanocarpa R. Br.

G. sieberiana Kunth.

Gymnoschoenus sphaerocephalus (R. Br.) Hook. f.
Lepidosperma laterale R. Br.

L. urophorum N.A. Wakef.

Lomandraceae
Lomanadra longifolia Labill.
L. montana (R. Br.) L.R. Frazer & Vickery

Luzuriagaceae
Eustrephus latifolius Ker. Gawl.
Geitonoplesium cymosum R. Br.

Orchidaceae

Bulbophyllum shepherdii (F. Muell.) F. Muell
B. exiguum F. Muell.

Cymbidiium suave R. Br.

Dendrobium aemulum R. Br.

D. pugioniforme A. Cunn.

D. speciosum Sm.

D. striolatum Rchb. f.

D. teretifolium R. Br.

Liparis reflexa (R. Br.) Lindl.
Plectorrhiza tridentata (Lindl.) Dockrill
Pterostylis curta R. Br.

P. grandiflora R. Br.

P. longifolia R. Br.

P. nutans R. Br.

P. obtusa R. Br.

P. pedunculata R. Br.

Sarcochilus falcatus R. Br.

S. hillii (F. Muell) F. Muell.

S. olivaceus Lindl.

Phormiaceae
Dianella caerulea Sims

Poaceae
*Echinopogon ovatus (G. Forst. P. Beau.
Entolasia stricta (R. Br.) Hughes

Imperata cylindrica var major S.W.L. Jacobs & C.A.

Wall

Oplismenus aemulus (R. Br.) Roem. & Schult.
O. imbecillis (R. Br.) Roem. & Schult.
Microlaena stipoides (Labill.) R. Br.

Panicum pygmaeum R. Br.

Proc. Linn. Soc. N.S.W., 128, 2007

Ripogonaceae
Ripogonum album R. Br.

Smilacaceae
Smilax australis R. Br.

79

VEGETATION HISTORY OF MOUNTAIN LAGOON

Appendix 3

Pollen type name on pollen diagram (Fig. 8) and probable source in the vegetation, taken from the vegetation
survey and Ryan et al. (1996).

Pollen type on pollen diagram Probable source in the vegetation

TREES AND SHRUBS

Casuarinaceae Casuarina and Allocasuarina. A. torulosa most likely
All species in the family. For identifications and proportions,

Myntaceae see Table 2
Podocarpus Probably Podocarpus spinulosus
Monotoca Probably Monotoca scoparia

Other Ericaceae

Other species in the family

Banksia All species in the genus
Hakea All species in the genus
Acacia All species in the genus

Asteraceae T. pleistocenicus

A ‘spineless’ Asterceae, thought to be a Cassinia

HERBS

Asteraceae (Tubuliflorae) All species in the Tubuliflorae
Gonocarpus Gonocarpus/Haloragis
Chenopodiaceae All species in the family
Poaceae All species in the family
Brassicaceae All species in the family
Caryophyllaceae All species in the family
AQUATICS

Cyperaceae All species in the family
Myriophyllum All species in the genus
Restionaceae All species in the family
Potamogeton All species in the genus
?Convolvulaceae ?Convolvulaceae
FERNS AND ALLIES

Trilete/monolete spores

Unknowns

80

All ferns and their allies

Unidentified pollen types

Proc. Linn. Soc. N.S.W., 128, 2007

A Revision of the Cryptandra propinqua Complex
(Rhamnaceae: Pomaderreae)

JURGEN KELLERMANN!? AND FRANK Upovicic!

‘National Herbarium of Victoria, Royal Botanic Gardens Melbourne,
Birdwood Avenue, South Yarra, Victoria 3141 Guergen.kellermann@rbg.vic.gov.au)
School of Botany, The University of Melbourne, Victoria 3010.

Kellermann, J. & Udovicic, F. (2007). A Revision of the Cryptandra propinqua complex (Rhamnaceae:
Pomaderreae). Proceedings of the Linnean Society of New South Wales 128, 81-98.

Four species are recognised in the Cryptandra propinqua complex: C. propinqua A. Cunn. ex Fenzl,
C. ciliata A.R. Bean, C. speciosa A. Cunn. ex Kellermann & Udovicic, here as new described, and C.
magniflora F.Muell., here re-instated. Two subspecies are recognised and described as new: C. propinqua
subsp. maranoa Kellermann & Udovicic and C. speciosa subsp. strigosa Kellermann & Udovicic. The
recently named taxon C. rigida A.R. Bean is reduced to synonymy under C. propinqua subsp. propinqua.
Descriptions, illustrations of flowers and distribution maps are provided for each taxon. A lectotype is

designated for C. magniflora.

Manuscript received 11 July 2006, accepted for publication 13 December 2006.

KEYWORDS: Australia, flora, Cryptandra, New South Wales, Pomaderreae, Queensland, Rhamnaceae,

South Australia, taxonomy, Victoria.

INTRODUCTION

Cryptandra Sm. is the second largest genus of
Australian Rhamnaceae. It occurs mainly in the
heathlands and woodlands of temperate to semi-arid
Australia and extends from south-western Western
Australia to south-eastern Australia, some species
occur in subtropical and tropical Queensland and there
are scattered occurrences in the Kimberley, Pilbara
and northerly part of the Northern Territory (Ladiges
et al. 2005, Kellermann et al. 2005, Kellermann 2006).
Key synapomorphies for the genus are one-flowered
inflorescences, imbricate rows of bracts surrounding
the base of the flower, a densely tomentose disk
surrounding the base of the ovary, fruitlets that
dehisce by a slit to release the seed, and stipules that
are fused around the base of the petiole (Thiele and
West 2004, Thiele 2007).

The species of the Cryptandra propinqua
complex are widely distributed from the mallee
regions of South Australia and Victoria to inland and
coastal New South Wales and Queensland. Taxa in
the complex have relatively large flowers, with sepals

usually longer than the free part of the hypanthium.
The base of the flowers is surrounded by many (up to
11) spirally arranged bracts, which often cover part
of the floral tube as well. In Western Australia, the
closest relatives on morphological grounds appear
to be C. aridicola Rye and C. minutifolia Rye (Rye
1995).

Previous research on species in the C. propinqua
complex has so far been focussed on state floras, and
no attempt had been made since Bentham (1863)
to examine specimens over their whole range of
distribution. In preparation for the “Flora of Australia’
treatment of Rhamnaceae, herbarium specimens
were examined from all major Australian herbaria,
allowing a comprehensive study of C. propinqua and
related taxa.

TAXONOMIC HISTORY

Allan Cunningham was the first botanist to collect
specimens of the Cryptandra propinqua complex,
during expeditions in New South Wales and southern
Queensland in 1823, 1825 and 1827. Cunningham

CRYPTANDRA PROPINQUA COMPLEX

apparently realised that there were two different
species present in New South Wales. He gave the
taxon he collected as number 24 the manuscript name
‘C. speciosa’ and described it as a ‘shrub of rigid
habit frequent in the barren rocky situations in various
parts of the Interior from the latitude of 29 to 33 S.
& Long. 151-148 flowering actually in May & June’
(note on BM 50750). He distinguished this inland
species from the tablelands from a closely related
coastal taxon with collecting number 22, which was
named by him ‘C. propinqua’. Fenzl published C.
propinqua in a footnote in ‘Enumeratio plantarum ...
Hiigel’ in 1837, taking up Cunningham’s manuscript
name. However, he could not have received material
of Cunningham’s collection of ‘C. speciosa’, and this
taxon remained unnamed. Fenzl did not indicate a
collecting number or precise locality in the protologue.

82

This led to confusion about the circumscription of C.
propinqua, since some authors assumed that both
Cunningham collections represent the same species.
An examination of the type at W (Fig. 1) revealed
that the material Fenzl used to name the species was
indeed the coastal form collected by Cunningham
under number 22 (Judy G. West, pers. comm., July
2005).

In 1862 Ferdinand von Mueller published C.
magniflora, a species that occurs in the Victorian and
South Australian mallee region and extends into New
South Wales. This was later reduced by Bentham
(1863) to a variety of C. propinqua under the name C.
propinqua var. grandiflora. Bentham’s concept of var.
grandiflora also included the unnamed *C. speciosa’
of Cunningham. He writes in his description of the
variety that it ‘is also amongst Cunningham’s plants
who had given it the name C. speciosa,
and designated the smaller variety by
that of propinqua, as being near the
larger one. Unfortunately this latter
name was the only one in the Vienna
herbarium, and was thus, although
inappropriate, adopted by Fenzl for the
species’ (Bentham 1863: 442).

Bentham’s separation of the species
into two varieties was not followed by
subsequent botanists, who adopted the
name C. propinqua for all taxa involved,
mainly due to a lack of ‘good characters’
(note on NSW 386701, dated 14 Sep.
1962). For example, Burbidge (1970),
Canning and Jessop (1986) and Harden
(1990) stated that C. propinqua was
distributed in South Australia, Victoria,
New South Wales and Queensland (the
addition of Western Australia to the area
of distribution by Harden was due to a
mis-identification of specimens). Other
botanists had a different concept of the
species, depending on the geographic
area in which they worked. Black
(1926, 1952) stated that the species
occurred in South Australia, Victoria
and western New South Wales only,
therefore excluding the coastal taxon.
Beadle et al. (1962) remarked that C.
propinqua was found at the coast, but
Beadle (1980) stated its distribution to
be in the N.S.W. Tablelands. Walsh and

Figure 1. Holotype of Cryptandra

propinqua A. Cunn. ex Fenzl. A.
Cunningham 22 (W).

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

Udovicic (1999) wrote that there were two forms of
C. propinqua in Victoria, one in the northern mallee
region, and a second form in eastern Victoria that
continues into south-eastern N.S.W.

The situation in Queensland is more complex.
C.T. White collected specimens of the coastal taxon
at the Glasshouse Mountains, mistakenly identifying
it as C. spinescens Sieb. ex DC. (White 1917). He
later coined the manuscript name ‘Cryptandra
ramosissima’ for that taxon, but never published this
species. His manuscript description is attached to
sheet BRI AQ109442. Since then, specimens at BRI
were housed under the phrase name ‘Cryptandra sp.
Q4 (ramosissima C.T. White ms)’. Stanley and Ross
(1986) used the name Cryptandra sp. | for that taxon
and C. propinqua for the remainder.

Bean (2004) accepted three species in the C.
propinqua complex in Queensland. He named a
very distinct species, which is allied to C. propinqua
as C. ciliata. It was first collected in the 1960s and
occurs in the area west of Theodore in south-eastern
Queensland. Botanists at BRI referred to it tentatively
as ‘Cryptandra sp. Q3 (aff. propinqua)’. Bean named
White’s coastal species as C. rigida and used the
name C. propinqua for an inland entity, which has its
main area of distribution in the district of Maranoa
and the western Darling Downs.

However, Cryptandrarigida from the Queensland
coast is indistinguishable from C. propinqua from
coastal New South Wales; as such, C. rigida is a new
synonym of C. propinqua. The taxon that Bean (2004)
referred to as C. propinqua is an inland form of C.
propinqua, and particular to southern Queensland. It
is here described as C. propinqua subsp. maranoa.
Cryptandra speciosa from the New South Wales
Tablelands is described in this paper after being
identified by Cunningham 180 years ago. A taxon
close to C. speciosa from the districts of Leichhardt
and South Kennedy in Queensland is described as C.
speciosa subsp. strigosa.

The species C. ciliata from south-eastern
Queensland is accepted and C. magniflora from the
mallee regions of South Australia, Victoria and New
South Wales is re-instated. Key characters of the four
species are listed in Table 1.

TAXONOMY
Key to Cryptandra propinqua and allied species

1 Stipule apices attenuate; hypanthium tube 0.7—1.2
mm long, petals 0.7—0.8 mm long, stamens 0.5—0.7

Proc. Linn. Soc. N.S.W., 128, 2007

mm long; floral bracts papery with margins reflexed
and flexuous cilia 0.3—0.6 mm long; fruit torus in
Upper Maltese eae eee sericea 4. C. ciliata
1: Stipule apices acute; hypanthium tube 1.2—3.5 mm
long, petals (0.7—) 0.9—1.6 mm long, stamens
(0.6—) 0.8-1.5 mm long; floral bracts with flat
margins and regular cilia (1.e., parallel and straight
or slightly curved), cilia 0.1—-0.3 mm long; fruit
torus equatorial or in lower half ..............ce 23
2 Leavessubsessile; bracts acuminate; adaxial surface
of bracts and stipules with coarse simple haits ......
Seeds: eee kata genie 8 3. C. magniflora
2: Leaf petioles apparent, (0.1—) 0.2-0.8 mm long;
bracts obtuse and glabrous adaxially; stipules
PlADTOMS AG axa liveries eerste. es tee ts. eeeea cena Ae 3
3 Stem indumentum of dense stellate hairs, with
very few simple hairs; bracts light brown, obovate
to elliptic; hypanthium tube 1.2—2.5 mm long ......
eee code nomennereler eh POE L403: 1. C. propinqua
3: Stem indumentum of antrorse, moderate to dense,
closely appressed simple hairs, sometimes also
with sparse stellate hairs underlying; bracts dark
brown to black, broadly ovate to broadly elliptic;
hypanthium tube 2.2—-3.5 mm long ... 2. C. speciosa

1. Cryptandra propinqua A. Cunn. ex Fenzl in S.F.L.
Endlicher et al., Enum. Pl. 23 (1837). Type citation:
‘New South Wales (Cunningham)’. Holotype: New
South Wales, 1825, A. Cunningham 22 (W n.v., photo
seen).

Shrub 0.2—1.5 m high, often intricately branched,
not spinescent, with a dense grey indumentum of
stellate hairs and sometimes also simple hairs on
young stems; leaves clustered in fascicles. Stipules
persistent, narrowly triangular, 0.9—1.5 (-2) mm long,
apex acute, connate around the base of the petiole;
abaxial side moderately pubescent or glabrous;
adaxial side glabrous. Petioles 0.1—0.8 mm long. Leaf
blades narrowly elliptic to linear, sometimes ovate to
broadly ovate, 0.8—5 (—11) mm long, 0.4—-1.7 (2.2)
mm wide, entire; margins revolute; base cuneate or
obtuse; apex acute, obtuse or occasionally shortly
mucronate; lower surface partly visible or not visible,
densely grey-stellate-hairy, sometimes glabrescent,
midrib with simple hairs; upper surface glabrous,
smooth or often tuberculate. Conflorescences
axillary, 1-2 cm long, consisting of 1—10 sessile to
shortly pedicellate flowers arranged in few branched
elongated pseudoracemes; axes densely stellate-
pubescent. Bracts 5—11, persistent, obovate or elliptic,
1.34.2 mm long, 0.9—2 mm wide, apex obtuse, light
brown; abaxial surface with few hairs or glabrous;
adaxial surface glabrous; cilia regular, 0.1—0.3 mm

83

CRYPTANDRA PROPINQUA COMPLEX

Table 1. Key characters distinguishing the species of the Cryptandra propinqua complex.

Distribution

Indumentum of
young stems

Leaf surface

Stipule apex

Petiole

Bract shape

Bract colour
Bract apex

Bracts and
stipules, adaxial
indumentum

Bract cilia

Bract abaxial
indumentum

Hypanthium tube
length

Hypanthium tube
indumentum

Hypanthium tube
hair types
Sepal length

Sepal
indumentum

Petal length
Stamen length

Style
indumentum

Style length
Fruit length

Torus position

84

C. ciliata

C. propinqua

Qld (Districts
Darling Downs and
Leichhardt)

Small stellate hairs
underlying coarse,
antrorse simple or
multiarmed hairs
that spread about 30
degrees to stem

Smooth or tuberculate

Attenuate

Sessile—subsessile,
0.10.3 (—0.5) mm

Broadly obovate or
orbicular

Light brown
Obtuse

Glabrous

Long flexuose

+Glabrous

0.7-1.2 mm

Upper 1/2 to 1/5 hairy

Dense stellate

1.5—2.2 mm

Densely stellate hairy,
very few simple hairs
at apex

0.7—0.8 mm
0.5—0.7 mm

Glabrous

0.7-1.1 mm
2.7-3.0 mm
Upper half

Coastal regions of
N.S.W. and south-
eastern Qld (subsp.
propinqua); inland
regions of southern
Queensland (subsp.
maranoa)

Dense stellate hairs,
some have very few
simple hairs

Smooth or often
tuberculate
Acute

0.1—0.8 mm

Obovate / elliptic

Light brown
Obtuse

Glabrous

Short regular

+Glabrous

1.22.5 mm

Upper 1/2 to 1/3 hairy

Sparse to dense
stellate

2.0-3.4 mm.

Dense simple at
apex and dense to
moderately simple
hairy in middle,
rest sparse to dense
stellate hairy

0.7-1.4 mm
0.6—-1.1 mm

Lower 1/4 to 1/3 hairy

0.5—3.7 mm
2.5-3.5 mm

Lower half or
equatorial

C. speciosa

N.S.W. Tablelands
(subsp. speciosa);
Districts Leichhardt
and South Kennedy,
Qld (subsp. strigosa)

Antrorse, closely
appressed, moderate
to dense simple hairs,
subsp. speciosa also
with sparse stellate
hairs

Smooth

Acute
0.2—0.7 mm

Broadly ovate /
broadly elliptic

Dark brown to black
Obtuse

Glabrous

Short regular

Often hairs on upper
middle or glabrous

2.2-3.5 mm

All hairy, on some
upper 1/2 hairy

Simple (mostly at top)
and sparse to rarely
dense stellate

2.24.0 mm

Dense simple hairs
overlying sparse
stellate hairs

1.0-1.6 mm
0.9-1.5 mm

Base to lower 1/3
hairy

2.6—3.7 (—5) mm
3.0-4.0 mm

Equatorial or lower
half

C. magniflora

Mallee regions of
S.A., Vic. and N.S.W.

Intertwined, matted or
loosely appressed fine
stellate and simple
hairs

Smooth or tuberculate

Acute

Sessile—subsessile,
0-0.2 mm

Ovate / elliptic, rarely
obovate

Dark brown to black
Acuminate

Coarse simple hairs

Short regular

Usually on upper
middle

1.5—2.6 mm

All hairy

Simple at top, mostly
dense stellate
1.8-3.5 mm

Dense simple hairs
with dense stellate
hairs

0.9-1.5 mm
0.8-1.4 mm

Glabrous or hairy at
base

1.7—2.6 mm
2.8-3.0 mm

+Equatorial

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

long. Pedicels 0-1.1 mm long, densely pubescent.
Flowers white or cream, sometimes pinkish with age.
Hypanthium tubular, tube 1.2—2.5 mm long, 1.32.4
mm wide; upper 1/2 to 1/3 of tube covered with sparse
to dense stellate hairs. Sepals erect or spreading, 2—
3.4 mm long, with an indumentum of sparse to dense
simple and stellate hairs, simple hairs mainly at apex
and midrib. Petals erect, 0.7—1.4 mm long; claw short
or absent, 0—0.2 mm long. Stamens erect, 0.6—1.1 mm
long; anthers 0.3—0.4 mm long. Disc a sinuate ring,
densely stellate-pubescent. Ovary inferior to semi-
inferior, 3-carpellate; summit densely stellate-hairy.
Style 0.5—3.7 mm long, lower 1/4 to 1/3 hairy; stigma
minutely 3-lobed. Schizocarp obovoid or ellipsoid,
2.5-3.5 mm long, brown, splitting into 3 dehiscent
fruitlets; apex acute or obtuse, torus position in lower
half or equatorial. Seeds 0.8—-1.9 mm long, reddish-
brown, somewhat darker in middle, base dark brown;
aril pale yellow-translucent.

Typification: The holotype (Fig. 1) consists of
two flowering branches. This sheet is the only
Cunningham collection of the species that could be
traced at W, which is annotated by Fenzl. It also bears
a label in Cunningham’s hand. The remaining original
collections (listed below) bear more exact information
about the collecting locality, e.g., the hills around the
Hunters River near Sydney. These specimens were
most likely collected during Cunningham’s expedition
to the Liverpool Plains in 1825 (Curry et al. 2001).

Notes and affinities: This species can be distinguished
from C. speciosa by its light brown bracts, the shorter
hypanthium tube, and a stem indumentum of dense
stellate and occasional simple hairs. It is closely
related to C. ciliata with which it shares the light
brown bracts.

Schlechtendal (1847) misapplied the name C.
propinqua to a collection of C. tomentosa Lindl. by
H. Behr from South Australia. Bentham (1862) and
Index Kewensis attribute the species to that author
and quote it as ‘C. propinqua Schltdl.’; however, this
is incorrect.

Original collections: NEW SOUTH WALES. ‘On
barren rocky hills on the north western branches of
the Hunters’ River’, Apr. 1825, A. Cunningham 22
(BM 50748, left specimen); Hunters River, May
1825, A. Cunningham 22 (K ex herb. Robert Heward,
top specimen of sheet with loan stamp ‘H/1310/95
54/76’). N.S.W., s. dat., A. Cunningham s.n. (MEL
238175).

Proc. Linn. Soc. N.S.W., 128, 2007

Key to subspecies of Cryptandra propinqua

1 Stem hairs stellate with occasional simple hairs;
stipules glabrous or sparsely hairy; leaves with
revolute margin, but lower surface usually visible;
bracts not covering sepals, entire; sepals with
stellate and simple hairs, especially along midrib;
style (1.5—) 1.7—3.7 mm long; coastal regions of
INSWerandi@lde la. subsp. propinqua

1: Stem hairs intertwined stellate and simple hairs;
stipules hairy at least on midrib; leaves with
margins closely revolute, lower surface not visible;
bracts partly hiding sepals, very fragile and easily
torn; sepals with small dense appressed stellate
and long simple hairs; style 1.5—1.8 mm long; Qld,
Maranoa district and adjacent regions ....................
ORC oe Eee eee roeernc aecrnaceaarcocas lb. subsp. maranoa

la. Cryptandra propinqua A. Cunn. ex Fenzl subsp.
propinqua

Cryptandra rigida A.R. Bean, Austrobaileya 6:
927 (2004). Holotype: Qld, Burnett District, “Cooya”,
W of junction of Barambah and Boonara Creeks,
17 July 1996, P. Grimshaw 2486 & R. Price (BRI
AQ641398). Jsotype: MEL 2263653.

Cryptandra sp. 1 sensu T.D. Stanley & E.M.
Ross, Fi. S.E. Queensl. 2: 46 (1986)

Cryptandra propinqua A. Cunn. ex Fenzl sensu
G.J. Harden, Fi. N.S.W. 1: 371 (1990), pro parte.

Cryptandra sp. (Ngungun L.S. Smith 13973)
sensu A.R. Bean in R.J.F. Henderson, Names distrib.
Queens. pl. algae lich. (2002)

“Cryptandra sp. Q4 (ramosissima C.T. White
ms)’ (BRI herbarium phrase name). Y

Cryptandra spinescens auct. non Sieber ex DC.:
C.T. White, Queensland Naturalist 2: 65 (1917).

Illustrations: S.G.A.P, Logan River Branch,
Mangroves to mountains 103 (2002), photograph, as
Cryptandra ‘sp. Ngungun’; A.R. Bean, Austrobaileya
6: 927, Fig. 4 (2004), photograph, as C. rigida.

Shrub 0.2-1.5 m high; young branches with
a dense grey indumentum of short stellate hairs,
sometimes with sparse simple hairs. Stipules 1—1.5
mm long, glabrous or sparsely hairy. Petiole 0.2—0.8
mm long. Leaf blades narrowly elliptic to ovate or
broadly ovate, (1.0—) 1.5—5 (11) mm long, 0.4—1.7
(—2.2) mm wide; margins revolute, but lower surface
usually visible. Bracts 6-11, 1.3-3.6 mm long, 0.9—2
mm wide, covering hypanthium tube, entire. Pedicels
0.3—1.1 mm long. Hypanthium tube 1.2—2.5 mm long,
1.32.4 mm wide. Sepals 2—3.3 mm long, with dense

85

CRYPTANDRA PROPINQUA COMPLEX

simple hairs at the apex, dense to moderately dense
simple hairs along the midrib, and sparse to dense
stellate hairs on the rest of the sepal. Petals 1—-1.4 mm
long, claw c. 0.2 mm long. Stamens 0.8—1.1 mm long.
Style (1.5—) 1.7—3.7 mm long. Schizocarp 2.5—3.2 mm
long, torus position in lower half. Seeds 0.8-1.9 mm
long. Figs 1, 2A.

Distribution and Habitat: The subspecies occurs
between Bundaberg (Qld) and the area around Jervis
Bay (N.S.W.), and grows in heathlands on rocky
outcrops, hillsides or gullies; it is recorded from sandy
soils or sandy loam on sandstone, and from granite
and rhyolite at 120-800 m altitude. Fig. 3A.

Phenology: Flowers Apr.—Sep.; fruits May—Nov.

Notes: The recently published Cryptandra rigida
from Queensland is conspecific with C. propinqua
subsp. propinqua. Specimens with and without ngid
habit can be found in Queensland and New South
Wales. Other characters mentioned by Bean (2004) to
be unique for C. rigida can be found in material of C.
propinqua subsp. propinqua from New South Wales,
such as branchlets with an indumentum of stellate
hairs only, a glabrous calyx tube and a mostly stellate
indumentum on the calyx lobes. However, they
distinguish the typical subspecies from C. propinqua
subsp. maranoa (see below). Some specimens from
northern N.S.W. have a tendency to glabrescent or
glabrous lower surfaces of the leaves.

Specimens examined: NEW SOUTH WALES:
Central Coast. Parramatta River, Apr. 1903, JL.
Boorman s.n. (NSW); Londonderry, 4 Feb. 1962,
C. Burgess s.n. (CBG at CANB); Blakehurst, Apr
1897, J.H. Camfield s.n. (NSW); Hurstville, Apr.
1898, J.H. Camfield s.n. (NSW); Como, May 1898,
J.H. Camfield s.n. (NSW); Maroota, 31 May 1961, E.
Gordon s.n. (NSW); Peats Ferry, Hawkesbury River,
14 May 1887, J.H. Maiden s.n. (NSW); Revesby to
Georges River, 9 Apr. 1956, K. Mair s.n. (NSW).
North Coast. 1 km NE of Nymboida, 25 Apr. 1994,
A.R. Bean 7647 (BRI); Grafton-Glenreagh road, near
Mt Kremnos, 5 Mar. 1997, A.R. Bean 11719 (BRI);
Mt Mullengen 4 miles [6 km] E of Ramornie, July
1922, WE: Blakely & D.W.C. Shiress s.n. (NSW);
Orara River, 10 miles [16 km] S of Ramornie, July
1922, WF. Blakely & D.W.C. Shiress s.n. (NSW);
Shore of Port Macquarie, 1819, A. Cunningham 16
(CBG at CANB); Rocky Creek, 30 km N of Grafton
on road to Coaldale, 23 Aug. 1985, D.B. Foreman
921 (MEL); Alum Mountain, Apr. 1924, H.R. Rupp
s.n. (NSW); Bulahdelah, Apr. 1924, H.M.R. Rupp

86

s.n. (NSW). South Coast. Jervis Bay, 1 Sep. 1977,
G.W. Althofer 6247 (NSW); Swan Lake, Cudmirrah,
20 miles [32 km] S of Nowra, 14 Apr. 1967, ELF
Constable 7371 (NSW); Budawang Range, N of
Currockbilly Mountain, 20 Sep. 1967, E.F. Constable
7458 (NSW); Yalwal Road near Nowra, Aug. 1922,
FA. Rodway 1394 (NSW); Sassafras, 11 May 1946,
FA. Rodway 14168 (NSW); Cross Road, Tomerong
to Turpentine, S of Nowra, 30 May 1934, J. Rodway
1397 (NSW). QUEENSLAND: Burnett. ‘Melrose’,
15 km W of Eidsvold, 15 Sep. 1990, A.R. Bean 2292
(BRI); State Forest 132, 9 km ESE of Brovinia, 7
June 1997, A.R. Bean 12037 (BRI); Mount Lorna,
3 km W of ‘Toondahra’, 3 Aug. 1988, PI. Forster
PIF4637 (BRI); 5.5 km W of ‘Toondahra’, 5 Apr.
1988, PI. Forster PIF4672 (BRI, MEL); ‘Cooya’,
E of Boonara Creek, 17 July 1996, P. Grimshaw
PG2492 (BRI); Campbell Creek, W of Mt Brian, 5
Nov. 1996, P Grimshaw PG2621 (BRI); Timber
Reserve 766, Abercorn, June 1971, G. Leiper
s.n. (BRI). Moreton. Top of Glass House, s. dat.,
EM. Bailey 6 (BRI, MEL); Mount Edwards, near
Aratula, 19 June 1990, A.R. Bean 1636 (BRI); Mt
Tunbubudla, W of Beerburrum, 17 May 1993, A.R.
Bean 6047 (BRI); near Picnic Creek and Surprise
Rock, Lamington, 26 Apr. 1958, S.7. Blake 20357
(BRI); NW of Ngungun, Glasshouse Mountains, 21
May 1985, A.M. Buchanan 6714 (HO); Turtle Rock,
15 June 1966, H.S. Curtis 292 (BRI); Rupari Hill,
1.8 km SW of Beerwah, 16 Apr. 1973, R. Dowling
15 (BRI); Mountain behind Esk, 12 July 1985, PL.
Forster PIF2055 & P.D. Bostock (BRI); Summit of
Mt Bangalora, 6 May 1990, PI. Forster PIF6785,
L.H. Bird & A.R. Bean (BRI); Mt Edwards Nat.
Park, 14km W of Boonah, 16 Sep. 1992, PI. Forster
PIF11467 & R. Reilly (BRI); Campbells Folly, 4km
SW of Tylerville, 19 Sep. 1992, PI. Forster PIF 11508
& G. Leiper (BRI, MEL); Mt Ernest, 24 Apr. 1993,
PI. Forster PIF13256 & G. Leiper (BRI, CANB,
NSW); Coochin Hills, Beerwah, stony section on top
of the hill, 3 June 1967, J.D. Hockings s.n. (BRI); Mt
Esk, approx 4 km NE of Esk on lower slopes, 20 May
1973, F.D. Hockings s.n. (BRI); Mount Edwards,
near Moogerah Dam, Sep. 1990, G. Leiper s.n. (BRD);
Plunkett, Timber Reserve 766, 10 Nov. 1990, G.
Leiper s.n. (BRI); Glen Rock, Esk, s. dat., G. Leiper
s.n. (BRI); Glen Rock, 9 Aug. 1988, E.M. Ross & PI.
Forster s.n. (BRI, NSW); Moonview Treviot Gorge,
4 Apr. 1999, MJ. Russel s.n. (BRI); Mt Edwards, 1
June 1938, E.J. Smith s.n. (BRI); Mt Gillies, 20 km
SW of Rathdowney on Mt Lindsay Hwy, 13 Oct.
1974, P. Sharpe 1105 (BRI); Mt Esk, s. dat., J. Shirley
s.n. (BRI); Ngungun, half way up S track of fairly flat
shoulder, 8 July 1968, L.S. Smith 13972 (BRI); NE

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

Figure 2. Flowers of taxa in the Cryptandra propinqua complex. A, C. propinqua subsp. propinqua (A.R.
Bean 12037); B, C. propinqua subsp. maranoa (D.M. Gordon 35); C, C. speciosa subsp. speciosa (L.A.S.
Johnson 7840); D, C. speciosa subsp. strigosa (E. McRobert s.n.); E, C. magniflora (N.G. Walsh 5090); F, C.

ciliata (V. Hando 214). All scale bars 1 mm.

Proc. Linn. Soc. N.S.W., 128, 2007 87

CRYPTANDRA PROPINQUA COMPLEX

Figure 3. Distribution map for taxa in the Cryptandra propinqua com-

cles).

comer of Ngungun below summit basin, 8 July 1968,
L.S. Smith 13973 (BRI, NSW, MEL); Coochin Hills
near summit of W peak on N side, 24 Aug. 1968,
L.S. Smith 14045 (BRI, CANB, NSW); Woodford,
Glasshouse Mts, 6 June 2002, J. Thompson 57 (BRI);
between Plunkett and Hopedale, 26 Aug. 1923, C.T.
White s.n. (BRI); Glasshouse Mountains, May 1910,
C.T. White s.n. (BRI); White Rock, S of Redbank
Plains, 8 June 1984, K.A. Williams 5043 (BRI). Wide
Bay. 1.5 km SSE of Biggenden Bluff, 1 Sep. 2002,
A.R. Bean 19229 (BRI); Summit of Mt Walsh, near
Biggenden, 17 Sep. 1983, 7. Bean s.n. (BRI); Head
of Stoney Creek, NW Boundary of Mt Walsh NP, 31
Oct. 1995, P. Grimshaw PG2224 & R.J. Price (BRI);
The Gorge, Biggenden Bluff, s. dat., C.T: White 7687
(BRI).

lb. Cryptandra_ propinqua subsp.
Kellermann & Udovicic, subsp. nov.

maranoa

A subspecie typica indumento caulium pilis
stellatis stmplicibusque dense-implicatis compositis,
foliis linearibus margine arcte revoluto, bracteis
marginibus fragilibus hypanthio sepalisque partim
tegentibus, stylo breviore differt.

Holotype: Qld, Maranoa District, St. George, 21 July
1949, D.M. Gordon 35 (BRI AQ109430).

88

plex. A, C. propinqua subsp. propinqua (circles), C. propinqua subsp.
maranoa (squares); B, C. speciosa subsp. speciosa (squares), C. speciosa

subsp. strigosa (triangles), C. ciliata (open circles), C. magniflora (cir-

Cryptandra propinqua subsp.
propinqua auct. non A. Cunn.
ex Fenzl: T.D. Stanley &
E.M. Ross, Fi. S.E. Queensl.
2: 46 (1986); A.R. Bean,
Austrobaileya 6: 926 (2004).

Shrub 0.2-1 m high; young
stems with dense intertwined
or matted shorter and longer
stellate hairs and occasional
simple hairs. Stipules 0.9-1.5
(—2) mm long, pubescent at
least on midrib. Petiole 0.1—-0.4
mm long. Leaf blades narrowly
elliptic to linear, 0.8-4 mm
long, 0.4—0.6 mm wide; margins
revolute, lower surface not
visible. Bracts 5-10, 2.24.2
mm long, 1.5—2 mm wide, very
fragile and easily torn, covering
hypanthium tube and at least
part of sepals. Pedicels 0-0.3
mm long. Hypanthium tube 1.3—
2.2 mm long, 1.5—2 mm wide.
Sepals 2.23.4 mm long, with small dense appressed
stellate and long simple hairs along midrib and apex.
Petals 0.7—1 mm long, claw absent. Stamens 0.6-0.9
mm long. Style 0.5—1.8 mm long. Schizocarp c. 3.5
mm long (2.8-4 mm according to Bean 2004), torus
position +equatorial. Seeds not seen. Figs 2B, 4.

Distribution and Habitat: The subspecies grows in
open woodlands with cypress pine (Callitris spp.)
or Angophora floribunda on sand, sandy loam and
sandstone in inland regions of southern Queensland.
Fig. 3A.

Phenology: Flowers recorded in May, July and Aug.;
fruits recorded in July

Etymology: The subspecific epithet refers to the
Queensland district of Maranoa, since the new
subspecies occurs in this district and adjacent regions.
‘Maranoa’ was the aboriginal name for the Maranoa
River, which was retained by Mitchell when he
discovered it in 1846 (Mitchell 1848); the district is
named after the river. (Although the epithet has the
form of a feminine Latin adjective, it is derived from
English and does not have any meaning in Latin.)

Notes: The taxon differs from the typical subspecies
in having long, intertwined stellate and simple hairs

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

AND HERBARNOL BOTANIC
GARDENS PRIDANE

ANNOY, J. Kellermann (MEE

~

Austrailan National Herbarium (CANB)
£ pai Comp

109430

Peptandi seogsimpsay commgbes.
Lemporary sorting tabet
Crplundre propingia Ma
$ July 2095

ay

HERGARIUM
| 320841 |
BRISBANE |

52/9e

Figure 4. Holotype of Cryptandra propinqua subsp. maranoa Kellermann &

Udovicic. D.M. Gordon 35 (BRI).

on the stem, very closely revolute, linear leaves, floral
bracts with fragile margins that cover the hypanthium
tube and part of the sepals, long simple hairs overlying
small dense appressed stellate hairs on the sepals, and
a shorter style.

Specimens examined: QUEENSLAND: Darling

Proc. Linn. Soc. N.S.W., 128, 2007

Downs. Barakula, 1959, D.M. Cameron OFD No.
59/272 (BRI); Brigalow logging area, 25 June 1997,
W. Drury 5 (BRI); near Nudley Tower, Barakula State
Forest, 27 July 1981, Vv. Hando 213 (BRI); Cecil Plains,
June 1962, FD. Hockings s.n. (BRI). Maranoa. Mt
Moffatt Nat. Park, 1 May 1997, E. Addicott MM45
(BRI, NSW); Bollon and St George, 21 July 1949,

89

CRYPTANDRA PROPINQUA COMPLEX

G.W. Althofer 31 (BRI); 87 km from Bollon on road
to St George, 17 Aug. 1979, F’ McKenzie CT22 (BRI);
15 miles [25 km] from St George, along Bollon Rd,
25 Aug. 1961, ME. Phillips s.n. (CBG at CANB).
Warrego. Near Boudens Dam, Chesterton Range, 14
Aug. 2001, C. Dollery 280 (BRI).

2. Cryptandra speciosa A. Cunn. ex Kellermann &
Udovicic, sp. nov.

A Cryptandra propinqua A. Cunn. ex Fenzl bracteis
fuliginosis ovatis-ellipticis et saepe paginis abaxialis
pubescentibus, pilis caultum simplicibus antrorsis,
hypanthio longiore differt.

Holotype: N.S.W., Northern Tablelands, Mt Kaputar
Nat. Park, summit area of The Governor, 15 Sep.
1998, B.J. Mole 56 & W.A. Gebert (MEL 2071544).
Tsotypes: NSW 501107 n.v., NE 75406 n.v.

Shrub 0.4—2 m high, not spinescent, with a moderate
to dense grey indumentum of antrorse, closely
appressed simple hairs overlying sparse small stellate
hairs on young stems; leaves usually clustered in
fascicles. Stipules persistent, triangular, 1.1—2 (—2.5)
mm long, apex acute, connate around the base of the
petiole; abaxial side moderately pubescent, especially
at midrib; adaxial side glabrous. Petioles 0.2—0.7 mm
long. Leaf blades linear to narrowly elliptic, (1.5—)
2.6-5.1 (-8) mm long, 0.4-0.8 (-2.8) mm wide,
entire; margins revolute; base cuneate or obtuse; apex
acute or obtuse, sometimes shortly mucronate; lower
surface usually not visible, densely grey-stellate-hairy,
rarely becoming glabrous, midrib with simple hairs;
upper surface glabrous, smooth. Conflorescences
axillary, 1-2 cm long, consisting of 1-10 almost
sessile flowers arranged in few branched elongated
pseudoracemes; axes densely stellate-pubescent.
Bracts 6-10, persistent, broadly ovate or broadly
elliptic, 1.4-4.6 mm long, 1.5—2.5 mm wide, apex
obtuse, dark brown to black; abaxial surface often
with hairs in the upper middle or glabrous; adaxial
surface glabrous; cilia regular, usually dense, (0.2—)
0.3 (—0.5) mm long. Pedicels 0.2—1 mm long, densely
pubescent. Flowers white. Hypanthium tubular, tube
2.2-3.5 mm long, 1.8—3.1 mm wide; the whole tube
or at least the upper half covered with simple hairs
(mostly towards the sepals) overlying sparse to rarely
dense stellate hairs. Sepals erect or spreading, 2.24
mm long, with an indumentum of dense simple hairs
overlying sparse stellate hairs. Petals erect, 1—1.6
mm long; claw 0.1-0.4 mm long. Stamens erect,
0.9-1.5 mm long; anthers 0.4—0.6 (-0.7) mm long.

90

Disc a sinuate ring, densely stellate-pubescent. Ovary
inferior to semi-inferior, 3-carpellate; summit densely
stellate-hairy. Style 2.6-3.7 (—5) mm long, base to
lower 1/3 hairy; stigma minutely 3-lobed. Schizocarp
obovoid or ellipsoid, 3-4 mm long, brown, splitting
into 3 dehiscent fruitlets; apex acute or obtuse, torus
equatorial or in lower half. Seeds 2.1—-2.5 mm long,
brown with a dark base; aril pale yellow-translucent.

Etymology: The epithet is derived from the Latin
speciosus (showy, splendid) and was applied to
the species by Cunningham, presumably because
of its conspicuous white and _ large-flowered
conflorescences.

Typification: There are several original collections
of Cunningham available (see below), which were
all collected in 1827 during the expedition that led
to the discovery of the Darling Downs. Although we
are using Cunningham’s manuscript name for the
species, we do not typify the taxon with one of his
collections, since there are better and more recent
collections available. We choose a collection from Mt
Kaputar National Park of which there are specimens
at three Australian herbaria. Cunningham also passed
Mt Kaputar while he was on the 1827 expedition
(McMinn 1970).

Notes and affinities: The species is closely related to
C. magniflora and is readily recognised by its dark
brown to black floral bracts and the usually densely
hairy flowers. The two subspecies of C. speciosa are
separated by c. 500 km.

Original collections: NEW SOUTH WALES.
‘New South Wales, frequent in the interior’, 1827,
A. Cunningham 24 (BM 50748, right specimen);
‘A shrub of rigid habit frequent in the barren rocky
situations in various parts of the Interior from the
latitudes of 29 to 33 S. & Long. 151-148, flowering
usually in May or June’, [1827,] A. Cunningham
24 (BM 50750); ‘New South Wales, Interior’, May
1827, A. Cunningham 24 (BM 50753, bottom left
specimen); “N.S. Wales, Interior, Lat 29 S Long. 151’,
May 1827, A. Cunningham 24 (BRI AQ109433); “N
Holld.’ , s. dat., A. Cunningham s.n. (MEL 2103518
ex Herb. Hooker).

Key to subspecies of Cryptandra speciosa
1 Stem with stellate and appressed simple hairs;
bracts 2.14.6 mm long, 1.8—2.5 mm wide; pedicels

0.5—1 mm long; sepals 2.6—4 mm long, with long
simple hairs overlying stellate hairs; petals 1.2—1.6

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

mm long; stamens 1—1.5 mm long; fruit 3.64 mm
long, torus position equatorial; N.S.W. Tableland
and Western Slopes ................ 2a. subsp. speciosa
1: Stem with simple, antrorse, appressed hairs, rarely
stellate hairs present; bracts 1.4—2.7 mm long, 1.5—
1.6 mm wide; pedicels 0.2—-0.3 mm long; sepals
2.2—2.8 mm long, with long simple hairs and very
few stellate hairs; petals 1—-1.1 mm long; stamens
0.9—1 mm long; fruit 33.5 mm long, torus position
in lower half; Qld, Leichhardt and South Kennedy
Cistricts:... semen tee ete. 2b. subsp. strigosa

2a. Cryptandra speciosa A. Cunn. ex Kellermann &
Udovicic subsp. speciosa

Cryptandra propinqua vat. grandiflora Benth.,
FI. Austral. 1: 442 (1863), pro parte.

Cryptandra propinqua auct. non A. Cunn. ex
Fenzl: N.T. Burbidge & M. Gray, Fl. A.C.T. 252
(1970); N.C.W. Beadle, Stud. fi. N.E. N.S.W. 4: 518
(1980); G.J. Harden, F/. N.S.W. 1: 371 (1990), pro
parte.

Illustration: N.C.W. Beadle, Students flora of north
eastern New South Wales 4: 519, Fig. 226 F4 (1980);
G.J Harden, Flora of New South Wales 1: 371 (1990);
both as C. propinqua.

Shrub 0.4—-1.5 m high, young stems with antrorse,
appressed simple and small stellate hairs. Stipules
1.1-2 mm long. Petioles 0.2-0.7 mm long. Leaf
blades 2.5—5.1 mm long, 0.4-0.8 mm wide. Bracts
6-10, 2.14.6 mm long, 1.8-2.5 mm wide, cilia
(0.2—) 0.3 (0.5) mm long. Pedicels 0.5—1 mm long.
Hypanthium 2.3-3.5 mm long, 1.8-3.1 mm wide.
Sepals 2.6—4 mm long, with an indumentum of long
simple hairs overlying stellate hairs. Petals 1.2—1.6
mm long. Stamens 1—1.5 mm long; anthers (0.4-)
0.5—0.6 (—0.7) mm long. Style 2.63.6 (-S) mm long.
Schizocarp 3.6—4 mm long, torus position equatorial.
Seeds c. 2.5 mm long. Figs 2C, 5.

Distribution and habitat: The subspecies grows in
Eucalyptus woodlands and cypress (Callitris spp.)
forests, on rocky slopes and ridges or the rims of
gorges in the Tablelands of New South Wales and the
Victorian alps near the border to N.S.W., in sandy soil
over sandstone or volcanic substrates. It is recorded
between 500—1380 m altitude. Fig. 3B.

Phenology: Flowers May, July—Oct.; fruits Sep.—
Nov.

Notes: Specimens collected around Canberra have
slightly larger flowers, but in every other aspect they

Proc. Linn. Soc. N.S.W., 128, 2007

are typical for C. speciosa subsp. speciosa.

Specimens examined: AUSTRALIAN CAPITAL
TERITORRY. Flints Crossing, Paddys River,
8 Sep. 1963, EF. D’Arnay 278 (CANB, NSW);
Uriarra Crossing, 23 Aug. 1964, J. Beeton s.n.
(CBG at CANB); Paddys River, 18 Sep. 1981, E.M.
Canning 5045 & M.C. Johnson (CBG at CANB);
Murrumbidgee River, 16 Aug. 1950, E. Gauba s.n.
(CBG at CANB); Molonglo River, 27 Sep. 1953, E.
Gauba s.n. (AD, CBG at CANB); Kowen, 8 Sep.
1962, H.S. McKee 9568 (NSW); above Paddys River,
0.5 km E of Murrays Corner, 1 Sep. 1983, JE. Ward
28 & A. Hughes (CBG at CANB). NEW SOUTH
WALES: Central Tablelands. Bathurst Plains, s.
dat., s. coll. (NSW). Northern Tablelands. Apsley
Falls, 21 Oct. 1900, E. Cheel s.n. (NSW); Tia Falls,
Oct. 1900, E. Cheel s.n. (NSW); between Jokers
Spring and English Spring, Mt Kaputar Nat. Park, 19
Nov. 1976, R. Coveny 8526 & S.K. Roy (NSW); Gara
River, 9 miles [14 km] E of Armidale, 2 Oct. 1955,
G.L. Davis s.n. (NSW); Dundee, June 1963, E.A.
Farleigh s.n. (NSW); Apsley Falls, Oct. 1900, W.
Forsyth s.n. (NSW); Yarrowyck-Bundarra, Sep. 1947,
L.A.S. Johnson 947/28 (NSW); Plains of Heaven,
3km SSW of Mt Kaputar, 1 Sep. 1974, L.A.S. Johnson
7840 (NSW); Dangar Falls, Armidale, 10 Sep. 1971,
E. McBarron 20300 (NSW); Mt Kaputar, Nandewar
Range, E of Narrabri, 25 Aug. 1973, B. Muffet
M3/132 (CBG at CANB). North Western Slopes.
Warialda, June 1905, J.L. Boorman s.n. (CANB);
walk track to the ‘Governor’, Mt Kaputar Nat. Park,
27 Nov. 1987, J.M. Fox 87/154 (CANB); Warialda,
s. dat., E.J. Hadlez s.n. (NSW); SSE of Bowling
Alley Point Cemetery, Sep. 1999, J.R. Hosking 1745
(CANB, MEL, NE, NSW); Plagyan State Forest, 5
July 1985, D.F.) Mackay 278 (NSW); Woods Reef,
Barraba, Oct. 1913, H.M.R. Rupp 7085/13 (NSW).
Southern Tablelands. Turpentine Ridge, 24 June
1962, C. Burgess s.n. (CBG at CANB); Queanbeyan,
12 Nov. 1996, I. Crawford 4001 (CANB, MEL);
Queanbeyan, Nov. 2000, 1. Crawford 4787 (CANB,
MEL, NSW); Yass River, c. 1.5 km NE of Yass Post
Office, 29 Aug. 1993, B.J. Lepschi 1058 (A, CANB,
HO, L); Queanbeyan, 17 Sep. 1960, H.S. McKee 7258
(NSW); 2.5 km NW of Mt Tianjara, 30 Apr. 1981, K.
Paijmans 3971 (CANB); Murrumbidgee River, 2 km
upstream from junction with Bredbo River, 27 July
1975, M. Parris 7520 (CBG at CANB). VICTORIA:
Snowfields. Snowy River above Willis, 29 June
1962, K. Rogers s.n. (MEL); Snowy River, 1.5 km
downstream from Sandy Creek, 15 Oct. 1989, J.
Turner s.n. (MEL).

91

CRYPTANDRA PROPINQUA COMPLEX

: fete Oe
Oe es
S oe renin Onlec
= ee “ ont

Satins HERBARIUM OF
RIA AUSTRALIA

FLORA OF AUSTRALIA PROIECT

Cyfonde spectesxn AC ap
Hocorepe, eRe ane A Uolevicte
Pet Jonge VED erenttnom: Date & Ae Sek

NATIONAL HERBARIUM OF VICTORIA (ME
MELUOURNE, Al is

MEL, 2071544

23 RHAMNACEAR p
Crygtondra proploqua Aon ea Fonsi
Call: Mate, BS > | 1S Sep :
al: Male, BA, $6. Dake: 15'Sep 1998

AUSTRALIA NEA Scr
Regioe/District: Nocthern Fablelands :
Lowality: spate es le Sant eo
risers S lee (orca
hes te
Adb: (4 30m .
Rote Mewttacibeind 9 skeletal on a
Eee re eo
CSOs GOH TNS A cenal anes
Det) Mole Bd. 1S hep ie
Dupe NEAR

Fig. 5. Holotype of Cryptandra speciosa A. Cunn. ex Kellermann & Udovicic.

B.J. Mole 56 & W.A. Gebert (MEL).

2b. Cryptandra speciosa subsp. strigosa Kellermann
& Udovicic, subsp. nov.

A subspecies typica caulibus maturis sine
pilis stellatis sed tantum pilis simplicibus antrorsis
adpressis, bracteis petalis et sepalis minoribus,
fructibus minoribus toris in dimidio inferiore differt.

OD

Holotype: Qld, Leichhardt District, Salvator Rosa
Nat. Park, 170 km SW of Springsure, Aug. 1983,
M.B. Thomas 241 (BRI AQ367623).

Cryptandra propinqua subsp. propinqua auct.

non A. Cunn. ex Fenzl: A.R. Bean, Austrobaileya 6:
926 (2004).

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

LERCHMARDT District

2O8EP 1397

annor. 2 Kellermann (MEL)

MA4TS 147 IE an m
Saivalor Rosa Nalional Park 170k SW of Springsure

‘On top of the while sandstone formation wrich

‘wretches away south from the Spygiese

ee

Ausiratian National Herbarium (CANB)
*

“Femeparery sorting label 36762;

Cryptandra propingua Le

3 July 2005

QUEENSLAND
[ape a

32/94- 24 | £09583 1
| BRISBANE :

Fig. 6. Holotype of Cryptandra speciosa subsp. strigosa Kellermann & Udo-

vicic. M.B. Thomas 241 (BRI).

Shrubs 0.6—2 m high, stems with strigose, antrorse,
appressed simple hairs, rarely small stellate hairs
present on young branches. Stipules 1.1—2 (—2.5) mm
long. Petioles (0.2—) 0.3—-0.6 mm long. Leaf blades
(1.5—) 2.54 (-8) mm long, (0.4—) 0.6-0.8 (—2.8)
mm wide. Bracts 8-9, 1.4-2.7 mm long, 1.5—1.6
mm wide, cilia c. 0.2 mm long. Pedicels 0.2-0.3 mm

Proc. Linn. Soc. N.S.W., 128, 2007

long. Hypanthium 2.2-3.3 mm long, 2—2.3 mm wide.
Sepals 2.2—2.8 mm long, with an indumentum of long
simple hairs and very few stellate hairs. Petals 1—1.1
mm long. Stamens 0.9—1 mm long; anthers c. 0.4 mm
long. Style 2.6-3.7 mm long. Schizocarp 3—3.5 mm
long, torus in lower half. Seeds 2.1-2.3 mm long.
Figs 2D, 6.

93

CRYPTANDRA PROPINQUA COMPLEX

Distribution and habitat: The subspecies is recorded
from forests and woodlands on poor soil on sandstone
and rocky outcrops in the Leichhardt and South
Kennedy districts of Qld, between the Narrien Range
and the Buckland Tablelands, at 500—600 m altitude.
Fig. 3B.

Phenology: Flowers May, Aug.; fruits Sep.

Etymology: The subspecific epithet 1s derived from
the Latin striga (a straight, rigid, close-pressed hair)
and refers to the characteristic indumentum on stems
and sepals of the taxon.

Notes: This subspecies differs from the typical
subspecies in having simple, antrorse, appressed hairs
on stems, only young branches bear occasionally a
few small underlying stellate hairs. It has smaller
bracts, sepals, petals and fruits, which also have the
torus in the lower half of the fruits (compared to
equatorial in subsp. speciosa).

Specimens examined: QUEENSLAND: Leichhardt.
Salvator Rosa Nat. Park, S of Mt Spyglass, 20 May
1986, ME. Ballingall 2157 (BRI); Top of Little
St Peter, 21 Aug. 1984, A.R. Bean 557 (BRI);
Cungelella, 1890, Mrs Biddulph s.n. (MEL); Mount
Zamia environmental park, overlooking Springsure,
8 May 1990, B. Davis 40 (BRI); St Peter, NNW of
Springsure, 27 Sep. 1984, B. O'Keefe 733 (BRI);
On top of Little St Peter, Springsure, 10 Sep. 1985,
B. O'Keefe 790 (BRI); Little St Peter, Sep. 1985, B.
O’Keefe 838 (BRI); Spyglass Peak, 1 Sep. 1992, B.
O’Keefe 985 (BRI); E of Tambo, adjoining Nat. Park,
Dec. 1995, E. McRobert s.n. (BRI). South Kennedy.
70km SW of Clermont in Narrien Range, 24 Aug.
1992, E.J. Thompson GAL82 & P.R. Sharpe (BRI).

3. Cryptandra magniflora F. Muell., Fragm. 3: 65
(1862). Cryptandra propinqua var. grandiflora Benth.
FI. Austral. 1: 442 (1863). Type citation: “Ad flumen
Murray passim in plagis undulato-arenosis confluxui
flumen Darling et Murray interjacentibus’. Lectototype
(here designated): [N.S.W. or Vic.,] Murray desert, s.
dat., s. coll. [possibly F. Mueller] (MEL 2103262).
Residual syntypes: [N.S.W. or Vic.,] Murray desert,
s. dat., s. coll. [possibly F. Mueller] (NSW 386703);
S.A., Mt Roebuck Station sand ridges, 1858 [?], s.
coll. (MEL 2103266). Possible syntype: s. loc., s.
dat., s. coll. (bottom fragment-pocket glued onto a
type sheet of C. propinqua; MEL 238175).
Cryptandra propinqua auct. non A. Cunn. ex
Fenzl in Endl.: JM. Black, F/. S. Austral. 3: 371
(1926); A.J. Ewart, Fl. Victoria 744 (1931); E.M.

94

Canning in J.P. Jessop & H.R. Toelken, F/. S. Austral.
2: 810 (1986); J.H. Willis, Handbk. pl. Victoria 2: 372
(1973); N.G. Walsh & F. Udovicic, F/. Victoria 4: 112
(1999).

Illustrations: G.R. Cochrane, B.A. Fuhrer, E.R.
Rotherham, J.H. Willis, Flowers and plants of Victoria
70, Fig. 186 (1968), photograph; E.M. Canning in J.P.
Jessop & H.R. Toelken (eds), Flora of South Australia
2: 809, Fig. 427F (1986); N.G. Walsh & F. Udovicic
in N.G. Walsh & T.J. Entwisle (eds), Flora of Victoria
4: 113, Fig. 19f (1999); all as C. propinqua.

Shrub 0.3-1.5 m high, spreading, intricately
branched, usually not spinescent, with a dense grey
indumentum of intertwined or loosely appressed
fine stellate and simple hairs on young stems; leaves
clustered in fascicles. Stipules persistent, triangular
or ovate, 1.2—-2 mm long, apex acute, connate around
the base of the petiole; abaxial surface moderately
pubescent, glabrescent, adaxial surface covered with
dense coarse simple hairs. Petioles very short or
absent, 0—0.2 mm long. Leaf blades narrowly elliptic,
1.5—5 mm long, 0.5—1 mm wide, entire; margins
revolute; base cuneate or obtuse; apex acute or obtuse
sometimes shortly mucronate; lower surface usually
not visible, densely grey-stellate-hairy, midrib with
simple hairs; upper surface glabrous, smooth to
tuberculate. Conflorescence axillary, 1-2 cm long,
consisting of 1-10 mostly sessile flowers arranged in
few-branched elongated pseudoracemes; axes densely
stellate-pubescent. Bracts 6-10, persistent, ovate or
elliptic, rarely obovate, 1.6—4 mm long, 1.3—2.5 mm
wide, apex acuminate, very dark brown to black;
abaxial surface sparsely to moderately pubescent with
minute stellate hairs, at least towards apex; adaxial
surface covered with dense coarse simple hairs; cilia
regular, 0.05—0.3 mm long. Pedicels 0—0.4 mm long,
densely pubescent. Flowers white or cream, becoming
pinkish after anthesis. Hypanthium tubular, tube 1.5—
2.6 mm long, 2—3.2 mm wide; covered with dense
stellate hairs and additional simple hairs in upper half.
Sepals erect or spreading, 1.8—3.5 mm long, with an
indumentum of dense stellate and closely appressed
simple hairs. Petals erect, 0.9-1.5 mm long, distinctly
clawed; claw 0.2—0.3 mm long. Stamens erect, 0.8—
1.4 mm long; anthers 0.3—0.5 mm long. Disc a sinuate
ring, densely stellate-pubescent. Ovary inferior to
semi-inferior, 3-carpellate; summit densely stellate-
hairy. Style 1.7—2.6 mm long, glabrous or with a few
stellate hairs on the base; stigma minutely 3-lobed.
Schizocarp obovoid or ellipsoid, 2.8-3 mm long,
brown, splitting into 3 dehiscent fruitlets; apex acute
or obtuse, torus tequatorial. Seeds 1.7—1.8 mm long,

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

MEL 2103262

National Herbanum of Victoria (MEL}

C tandem nayen A.Cann. &x
Soe

Determinavit

_ FLORA OF AUSTRALIA PROJECT
alr = Mek

\.Staiste 19 He[acot

Fig. 7. Lectotype of Cryptandra magniflora. Collector unknown, possibly F.
Mueller (MEL).

reddish-brown, +uniformly coloured or with pale
mottling, darker at base; aril pale yellow-translucent.
Figs 2E, 7.

Distribution and Habitat: This species occurs in
dune mallee communities and scrubs on sandstone
outcrops, on sand and sandy loamy soils in the far
north-west of Victoria, north-eastern South Australia,

Proc. Linn. Soc. N.S.W., 128, 2007

extending into south-western New South Wales;
recorded at 40—100 m altitude. Fig. 3B.

Phenology: Flowers June-Sep.; fruits Sep.—Nov.

Typification: There are few specimens labelled in
Mueller’s hand at MEL. All of them have very

95

CRYPTANDRA PROPINQUA COMPLEX

limited label information. The lectotype consists
of a flowering branch and is labelled by Mueller as
‘Cryptandra propinqua A. Cunn. var. magniflora’.
However, when Bentham reduced C. magniflora to a
variety of C. propinqua, he did not take up Mueller’s
epithet and named it C. propinqua var. grandiflora.
The specimen was collected in the plains near the
Murray River, a locality which corresponds well
with Mueller’s protologue. The specimen from NSW
bears the same label information, however, it was
not written by Mueller himself. A second specimen
at MEL, labelled in Mueller’s hand as ‘Cryptandra
propinqua var. grandiflora’ was collected near Mt
Roebuck in South Australia, possibly in 1858, but the
date is not clearly legible. There is also a fragment
pocket containing several flowers and leaves of C.
magniflora, which is glued onto a possible isotype
sheet of C. propinqua at MEL. This is labelled by
Mueller ‘Cryptandra magnifiora’, and ‘Basionym
for C. propinqua var. grandiflora’ in a later hand.
The presence of this pocket on the MEL type of C.
propinqua indicates that the specimen might have
been used by Mueller to compare it to the Cunningham
material. However, no collection information is
available about the specimen in this fragment pocket.
As such, it was not selected as a lectotype.

Notes: This species is closely related to C. speciosa
with which it shares the dark brown floral bracts.
Cryptandra magniflora can be distinguished from
all other species in the C. propinqua complex by the
presence of coarse simple hairs on the inner surfaces
of the bracts and stipules. The common name Silky
Cryptandra was applied to this taxon by Canning
and Jessop (1989) and Walsh and Udovicic (1999)
(both as C. propinqua).

Specimens examined: NEW SOUTH WALES:
South Far Western Plains. Garston Station, W side
of Darling River, 43km N Wentworth, 3 Oct. 1982,
J.H. Browne 123 (NSW); 40 km S of Pooncarie,
Nov. 1974, W.E. Mulham W799 (NSW); Tapalin mail
road, off Sturt Hwy, between Euston and Buronga,
14 Aug. 1977, WE. Mulham 1211 (NSW). SOUTH
AUSTRALIA: Eastern. Pualco Range, June 1970,
R. Bates 426 (AD). Eyre Peninsula. 20 km W of
Secret Rocks, 14 July 1993, R. Bates 33580 (AD);
c. 55 km N of Wirrulla, 1 Aug. 1969, B. Copley 2732
(AD); Below Narlara rockhole on the dog-proof
fence, 28 Nov. 1991, ML. Evans 40 (AD); Wudinna,
5 Sep. 1938, E.H. Ising s.n. (AD); Yumburra
Conservation Park, 4 Sep. 1984, D. Keane 33 (AD);
Munyaroo Conservation Park, 9 Aug. 1992, 4.G.
Spooner 13318 (AD); 15 km N of Koonibba, 11 Sep.

96

1960, D.J.E. Whibley 574 (AD). Flinders Ranges.
N slope Yankaninna Range, 25 Feb 1956, T-R.N.
Lothian 2080 (AD). Gairdner-Torrens. Mt Finke, 7
Oct. 1987, D.E. Symon 14730 (AD). Murray. River
Murray, 7 miles [11 km] W of Berri, 25 Aug. 1962,
J.B. Cleland s.n. (AD); Mantung District, 18 Aug.
1924, J.B. Cleland s.n. (AD); 2.4 km S of claypan
on Gluepot Calperum fence line, 21 Aug. 1997, S.
Donaldson 1245 & G. Flowers (AD, CBG at CANB);
Black Oak Plains, N of Murray River, 20 Aug. 1974,
N. Gemmell 283 (AD); SW of campsite in Pooginook
Conservation Park, 19 Aug. 2004, T. Hall 519 (AD,
DAO n.¥.); Upper Murray Mallee, Canegrass Station,
21 Sep. 1937, E.H. Ising s.n. (AD); c. 15 km W of
Chowilla wool-shed, 30 Aug. 1974, J.B. Paton s.n.
(AD); Calperum south-west, 26 Aug 1990, A.G.
Spooner 1215] (AD); Pooginook Conservation Park,
18 Aug. 1993, A.G. Spooner 14285 (AD); Calperum
Station, 18 Aug. 1996, A.G. Spooner 15878 (AD,
AK n.v.); 10 km NE of Taylorville, 29 Sep. 1976,
L.D. Williams 8685 (AD). Nullarbor. 1.5 km E of
Immarna, 29 Sep. 1975, R.J. Chinnock 2667 (AD).
VICTORIA: Murray Mallee. Northern Sunset
Country, 13 km NW of centre of Rocket Lake, 13
Sep. 1989, D.E. Albrecht 3872 (MEL); c. 3.7 km E
of Hattah, 13 Sep. 1989, D.E. Albrecht 3876 (MEL);
Hattah Lakes Nat. Park, Oct. 1948, A.C. Beauglehole
ACB 1116 (MEL); near junction of Murray Valley
Hwy and entrance to Hattah-Kulkyne, 30 Aug. 1977,
D.G. Cameron 8721 (MEL); Red Cliffs, Stewart, 5
Sep. 1961, L.C. Chandler & A.C. Beauglehole 19745
(MEL); Northern Sunset Country, 23 Aug. 1986, D.C.
Cheal s.n. (MEL); Pink Lakes, c. 15 km N of Linga,
28 Aug. 1979, MG. Corrick 6232 & B.A. Fuhrer (AD,
MEL); Redcliffs, western extremity of irrigation area,
1 Aug. 1981, MG. Corrick 7477 (MEL); 8 miles
[13 km] W of road junction 14 miles [22 km] N of
Birthday Tank, Sunset, 24 Sep. 1965, R. Filson 7418
(AD, MEL); Underbool N track, 5.2 km S of Rocket
Lake track, 23 Aug. 1986, G.R. Lucas 198 (CANB,
HO); S end of Hattah Kulkyne Nat. Park, s. dat., K.
Macfarlane 129 (AD, CANB, HO, MEL, NSW);
Sunset country, several miles NW of Mt Crozier,
10 July 1962, A. McEvey 35 (MEL); c. 17 miles [27
km] N of Ouyen of the Calder Hwy, 17 Aug. 1960,
TB. Muir 1195 (AD); Banneston, 13 Aug. 1960, E.
Rowlands s.n. (MEL); Hattah-Kulkyne Nat. Park,
0.2 km from Calder Hwy, 9 Sep. 1986, N.G. Walsh
2568 (MEL); Sunset Country, Werrimull South track
extension, 18 Sep. 1989, N.G. Walsh 2623 (MEL);
Murray Sunset Nat. Park, 4 Sep. 1999, N.G. Walsh
5090 (MEL); Ouyen, Sep. 1913, H.B. Williamson s.n.
(AD, CANB, HO, MEL, NSW); Kooloonong, Sep.
1924, A.B. Williamson s.n. (CANB); Kulkyne Nat.

Proc. Linn. Soc. N.S.W., 128, 2007

J. KELLERMANN AND F. UDOVICIC

Park, 1 Sep. 1941, J.H. Willis s.n. (MEL).

4. Cryptandra ciliata A.R. Bean, Austrobaileya 6:
927 (2004). Holotype: Qld, Leichhardt District, 28 km
from Cracow on Nathan Gorge road, 15 July 1990,
PI. Forster PIF7037 (BRI AQ627884). Isotypes: AD
n.v., CANB n.v., K n.v., MEL 2263651, NSW n.v.

Cryptandra sp. | sensu J.D. Briggs & J.H. Leigh,
Rare Threat. Austr. PI. (1995).

Cryptandra sp. (Gurulmundi G.W. Althofer
8418) sensu A.R. Bean in R.J.F. Henderson, Names
distrib. Queensl. pl. algae lich. (2002).

‘“Cryptandra sp. Q3 (aff. propinqua)’ (BRI
herbarium phrase name).

Illustrations: A.R. Bean, Austrobaileya 6: 927, Fig. 4
(2004), photograph.

Shrubs 0.25—1 m high, not spinescent, young
stems with a dense grey indumentum of small
stellate hairs underlying coarse antrorse simple or
multiarmed hairs that spread at c. 30° from the stem;
leaves clustered in fascicles. Stipules persistent,
narrowly triangular, 1-2 mm long, scarious, apex
attenuate, connate around the base of the petiole;
abaxial side sparsely to moderately pubescent on
midrib and margin; adaxial side glabrous. Petioles
very short, 0.10.3 (-0.5) mm long. Leaf blades
narrowly ovate to narrowly elliptic, 1.2—2 (2.5) mm
long, 0.3—0.5 mm wide, entire; margins revolute; base
cuneate; apex acute or obtuse; lower surface largely
obscured, densely grey-stellate-hairy, midrib with
sparse to moderately dense, loosely appressed simple
hairs; upper surface glabrous, smooth or sometimes
shortly scabrous. /nflorescence of individual, axillary
flowers, or these aggregated in hemispherical and
terminal conflorescences, 3—6 cm long, 5—8 cm wide,
consisting of 1-10 almost sessile flowers arranged
in few-branched contracted pseudoracemes; axes
densely stellate-pubescent. Bracts 7—10, persistent,
broadly obovate or orbicular, 1.8—2.6 mm long,
0.9-1.5 mm wide, obtuse, light brown with papery,
corrugated upper margins; abaxial surface usually
glabrous; adaxial surface glabrous; cilia long, flexuose,
0.3—0.6 mm long. Pedicels 0.2—0.3 mm long, densely
pubescent. Flowers white. Hypanthium tubular, tube
0.7—1.2 mm long, 1.5—2.3 mm in diameter; upper 1/5
to 1/2 covered with dense stellate hairs. Sepals erect
or slightly spreading, 1.5—2.2 mm long, with a grey
indumentum of dense stellate and very few simple
hairs at the apex. Petals erect, 0.7—-0.8 mm long,
indistinctly clawed (c. 0.1 mm long) or not clawed.
Stamens erect, 0.5—0.7 mm long; anthers c. 0.3 mm
long. Disc a sinuate ring, densely stellate-pubescent.

Proc. Linn. Soc. N.S.W., 128, 2007

Ovary inferior to semi-inferior, 3-carpellate; summit
densely stellate-hairy. Style 0.7—1.1 mm long, +entire,
glabrous. Schizocarp obovoid or ellipsoid, 2.7—3 mm
long, brown, splitting into 3 dehiscent fruitlets; apex
acute or obtuse, torus in upper half. Seeds 1.5—1.8
mm long, reddish-brown, +uniformly coloured, dark
brown at base; aril pale yellow-translucent. Fig. 2F.

Distribution and Habitat: Occurs in heathland,
shrubland or woodland on steep, rocky sandstone
slopes and ridges on sandy soil or sandy loam, from
Barakula State Forest to the area west of Theodore.
Fig. 3B.

Phenology: Flowers May—Aug. Fruits Sep.—Oct.

Notes: The species differs from C. propinqua in
having very long, flexuose cilia on the bracts and
attenuate stipules. One specimen, Brushe JB1518,
has very long hairs on the abaxial side of the bracts, in
addition to the long cilia. It might be an aberrant form
of the species or may prove to be a distinct taxon.
Further collections are warranted.

Specimens examined: QUEENSLAND. Darling
Downs. Gurulmundi, June 1978, G.W. Althofer 8418
(BRI); NW comer of Barakula Forestry, 24 Aug.
1980, V. Hando 151 (BRI); 15 miles [24 km] NW
of Barakula Forestry, 24 Aug. 1980, V. Hando 214
(BRI); Gurulmundi, May 1958, FD. Hockings s.n.
(BRI); Gurulmundi, June 1962, FD. Hockings s.n.
(BRI); Taroom-Cracow road, about 20 miles [32 km]
from Cracow, 26 June 1950, R. W. Johnson 809 (BRI).
Leichhardt. Cracow-Taroom road, S ‘of Fairyland,
16 Sep. 1990, A.R. Bean 2302 (BRI); Planet Downs
pastoral holding, 29 Mar. 1998, J. Brushe JBI518&
(BRI); Gwambagwine, 24 Sep. 1996, PI. Forster
PIF 19653 (BRI, CANB); Gwambagwine, 24 Sep.
1996, PI. Forster PIF1966& (BRI, CANB, NSW);
Gwambagwine, 11 Sep. 2000, PI. Forster PIF 26056,
R. Booth & F- Carter (BRI, MEL); Panda Corner,
Barakula State Forest, 24 Sep. 1978, FD. Hockings
s.n. (BRI); Retreat Creek Rd, 16 miles [26 km] from
Miles, 3 May 1960, R.W. Johnson 1628 (BRI); 16
miles [26 km] SSW of Cracow township, 10 July
1963, M. Lazarides 6948 (BRI, CANB, MEL, NSW);
18 miles [29 km] S of Cracow, 18 Feb. 1964, N.H.
Speck 1930 (BRI, CANB, NSW); Cracow-Taroom
road, 18 Aug. 1976, K.A.W. Williams 76009 (BRI).

97

CRYPTANDRA PROPINQUA COMPLEX

ACKNOWLEDGEMENTS

We thank Judy West (CANB) for examining and
photographing the type of Cryptandra propinqua at W. Uwe
Braun (HAL) sent images from Schlechtendal’s herbarium.
Bob Coveny and Ian Simpson provided information on
specimens from NSW and NE, respectively. We are grateful
to the directors of AD, BM, BRI, CANB, K and NSW for
the loan of specimens. Jo Palmer (CANB) organised the
transfer of several loans from Canberra to Melbourne.
Neville Walsh (MEL) commented on an earlier draft of the
paper. Two anonymous reviewers provided constructive
feedback and criticism. This paper is written in preparation
of the Flora of Australia treatment of Rhamnaceae,
supported by the Australian Biological Resources Study
(ABRS).

REFERENCES

Beadle, N.C.W. (1980). “Students flora of north eastern
New South Wales’ 4. (University of New England,
Botany Department: Armidale).

Beadle, N.C.W., Evans, O.D. and Carolin, R.C. (1962).
“Handbook of the vascular plants of the Sydney
district and Blue Mountains’. (Published by the
authors: Armidale).

Bean, A.R. (2004). New species of Cryptandra Sm. and
Stenanthemum Reissek (Rhamnaceae) from northern
Australia. Austrobaileya 6, 917-940.

Bentham, G. (1863). Rhamneae. In ‘Flora Australiensis,
a description of plants of the Australian Territory’ 1,
409-445. (L. Reeve & Co.: London).

Black, J.M. (1926). Rhamnaceae. In ‘Flora of South
Australia’ 3, 364-371. (R.E.E. Rogers: Adelaide).

Black, J.M. (1952). Rhamnaceae. In ‘Flora of South
Australia’ 3, 544-553. (K.M. Stevenson: Adelaide).

Burbidge, N.T. (1970). ‘Flora of the Australian Capital
Territory’. Australian National University Press:
Canberra.

Canning, E.M. and Jessop, J.P. (1986). Rhamnaceae. In
‘Flora of South Australia’ (Eds J.P. Jessop & H.R.
Toelken) 2, 807-821. (The Flora and Fauna of South
Australia Handbooks Committee: Adelaide).

Curry, S., Maslin, B. and Maslin, J. (2001). “Allan
Cunningham Australian collecting localities’.
(Australian Biological Resources Study: Canberra).

Harden, G.J. (1990). Rhamnaceae. In ‘Flora of New South
Wales’ 1, 354-373. (New South Wales University
Press: Kensington).

Kellermann, J. (2006). Cryptandra triplex K.R. Thiele
ex Kellermann, a new species of Rhamnaceae
(Pomaderreae) from Armhem Land, Northern
Territory, Austrobaileya 7, 299-303.

Kellermann, J., Udovicic, F. and Ladiges, PY. (2005).
Phylogenetic analysis and generic limits of the tribe
Pomaderrae (Rhamnaceae) using internal transcribed
spacer DNA sequences. Jaxon 54, 619-631.

98

Ladiges, P.Y., Kellermann, J., Nelson, G., Humphries, C.J.
and Udovicic, F. (2005). Historical biogeography of
Australian Rhamnaceae, tribe Pomaderreae. Journal
of Biogeography 32, 1909-1919.

McMinn, W.G. (1970). Allan Cunningham: botanist and
explorer. (Melbourne University Press: Carlton).

Mitchell, T.L. (1848) ‘Journal of an expedition into the
interior of tropical Australia in search of a route
from Sydney to the Gulf of Carpentaria’. (Longman,
Brown, Green & Longmans: London).

Rye, B.L. (1995). New and priority taxa in the genera
Cryptandra and Stenanthemum (Rhamnaceae) of
Western Australia. Nuytsia 10, 255-305.

Schlechtendal, D.F.L. von (1847). Bestimmung und
Beschreibung der vom Dr. Behr in Siidaustralien
gesammelten Pflanzen. Linnaea 20, 559-672.

Stanley, T.D. and Ross, E.M. (1986). Rhamnaceae.

In ‘Flora of south-eastern Queensland’ 2, 40-51.
(Queensland Department of Primary Industries:
Brisbane).

Thiele, K.R. (2007). Two new species of Australian
Stenanthemum (Rhamnaceae: Pomaderreae). Journal
of the Adelaide Botanic Gardens 21, 67-74.

Thiele, K.R. and West, J.G. (2004). Spyridium
burragorang (Rhamnaceae), a new species from New
South Wales, with new combinations for Spyridium
buxifolium and Spyridium scortechinii. Telopea 10,
823-829.

Walsh, N.G. and Udovicic, F. (1999). Rhamnaceae. In
‘Flora of Victoria’ (Eds N.G. Walsh & T.J. Entwisle)
4, 82-120. (Inkata Press: Port Melbourne).

White, C.T. (1917). Botanic Notes, No. 4. Queensland
Naturalist 2, 65-66.

Proc. Linn. Soc. N.S.W., 128, 2007

Observations of Insect Damage to Leaves of Woodland

Eucalypts on the Central Western Slopes of New South Wales:

1990 to 2004

W.S. SEMPLE! AND T.B. KOEN?

' NSW Department of Natural Resources, PO Box 53, Orange, NSW 2800 (bill.semple@dnr.nsw.gov.au)
* NSW Department of Natural Resources, PO Box 445, Cowra, NSW 2794 (terry.koen@dnr.nsw.gov.au)

Semple, W.S. and Koen, T.B. (2007). Observations of insect damage to leaves of woodland eucalypts on
the central western slopes of New South Wales: 1990-2004. Proceedings of the Linnean Society of New
South Wales 128, 99-110.

Damage to leaves of ~680 eucalypt trees at 17 paired sites, distributed across three soil landscapes near
Molong and Manildra (NSW), was monitored each autumn from 1990 to 2004. Insect damage was assessed
by estimating the proportion of damaged leaves on each tree. Across all species and sites, and for most
of the time, mean damage fluctuated between 10 and 25 % of leaves obviously damaged. Higher values
(30 % of leaves damaged) were recorded in 1990 and 1994, which coincided with increased abundance
of Scarabeidae. After c.1995 abundance of Scarabeidae declined and most leaf damage was due to feeding
by other insects. Relative damage levels to dividual species changed over time and for Eucalyptus albens
and E. melliodora was associated with the soil landscape in which the trees occurred. When Scarabeidae
were active, £. albens and/or E. blakelyi-dealbata showed higher leaf damage than E. bridgesiana, E.
microcarpa and E. melliodora although the last mentioned was damaged by insects other than Scarabeidae
during this period. Leaf damage across all trees and times was negatively correlated with warm season
rainfall 4 years previously. Contrary to expectations, most individual trees did not experience severe leaf

damage in consecutive years.

Manuscript received 17 October 2006, accepted for publicatio 15 January 2007.

KEYWORDS: dieback, Eucalyptus albens, E. blakelyi, E. bridgesiana, E. dealbata, E. melliodora, Perey S

rainfall index, Psyllidae, Scarabeidae.

INTRODUCTION

During the summer of 1989-90, eucalypts in the
Central West of NSW experienced a severe attack by
Christmas beetles (Anoploganthus spp., Scarabeidae).
It was reported (Dick 1990a) to be the worst attack
since the 1970s. They were also seen as a threat to on-
farm tree plantings, which were increasing in that area
at the time. A newspaper article (Anon. 1990, p. 25)
cited a CSIRO entomologist as saying that Eucalyptus
blakelyi was no longer a suitable tree for on-farm
planting and as waterlogging was a major factor
in eucalypt dieback, “...there may be a case for not
planting native trees in heavily stressed waterlogged
sites, where it may be better to plant exotics like
willows.” In another article, Dick (1990b) reported
that CSIRO researchers had found ‘insect-resistant

trees’, which they intended to clone for potential use
in on-farm tree plantings.

[This project was terminated in the mid
1990s but not before the proportion of cineole in
leaf terpenoids was identified as a major factor in
explaining differing resistances to beetle defoliation
(Edwards et al. 1990, 1993). Cloning was apparently
not achieved prior to cessation of funding though
orchards produced from seed of resistant trees were
established (Floyd and Farrow 1995)].

As defoliation by Scarabeidae was a major
contributor to dieback (a symptom of a disorder
with various causes, including insect attack) on the
Northern Tablelands in the 1970s (Nadolny 1995), the
beetle attack of 1989-90 appeared to be a portent of
worse to come in the Central West. One of us [WS], a
newly-appointed “investigations officer’, was asked:

INSECT DAMAGE TO WOODLAND EUCALYPTS

(a) was this event going to be repeated on a regular
basis as occurred in the Northern Tablelands and (b)
was it true that the same trees were attacked year after
year? Though not providing specific answers to these
questions, a considerable volume of work on insect
defoliation and eucalypt dieback in general was
being reported at the time. Landsberg et al. (1990),
for example, proposed five possible explanations for
chronic defoliation by insects:

1. It is a naturally-recurring phenomenon. For
example, Curtis’ (1989) review of dieback on
the Northern Tablelands of NSW noted that it
had occurred several times in the previous 100
years and that wet summers were associated with
greater insect abundance. After an extensive
review of dieback events in Australia and
overseas, White (1986) proposed that recurrent
dieback events were due to changes in weather
patterns. Where rainfall patterns resulted in
moisture stress, subsequent changes in eucalypt
physiology made them more attractive to insects
and/or pathogens.

2. ‘Stressed trees’ — trees stressed, e.g. by soil
degradation or waterlogging, have reduced
ability to recover from insect attack.

3. ‘Ecosystem imbalance’— some insects have
been favoured by tree clearing and pasture
improvement resulting in increased abundance
on the remaining trees.

4. ‘Release from natural enemies’ — abundance
of predators, including other insects and birds,
has declined due to reduced habitat resources in
degraded tree remnants.

5. ‘Nutrient (nitrogen) enrichment’ — increased
soil fertility due to pasture improvement has
increased the nutrient quality of tree leaves and
hence their attractiveness to insects.

6. “Maladapted trees’ — Edwards et al. (1993)
and Floyd and Farrow (1995) subsequently
suggested that reduced opportunities for eucalypt
reproduction in agricultural landscapes had not
allowed types resistant to insect attack to fully
develop.

From an investigation of soil factors, tree health,
insect abundance and herbivory in paired stands of
intact and degraded stands of Eucalyptus blakelyi
— E. melliodora woodland, Landsberg et al. (1990)
concluded that explanation #5 was the most likely
reason for enhanced abundance of insects on trees and
suggested that the accumulation of nitrogen in stock
camps around tree stands was implicated. The other
explanations were not entirely ruled out and, as noted

100

by Wylie and Landsberg (1990), any one or more may
have application in a particular situation.

It was also clear that some eucalypt species
were more likely to be defoliated than others. Curtis
(1989) reported that E. nova-anglica, E. blakelyi
and E. melliodora were more severely affected than
E. pauciflora on the Northern Tablelands of NSW.
However, relative damage between species appeared
to vary regionally. Fox and Morrow (1983) reported
that E. pauciflora and E. blakelyi were both heavily
damaged in southern NSW and that E. pauciflora
suffered greater levels of damage when trees where
growing in mixed rather than monospecific stands.
Wylie and Landsberg (1990) noted that dieback
affected trees of all ages but was more severe in older-
age classes than younger ones. It was also evident that
insect damage levels varied between individuals of
the same species in the same area and even between
different parts of the one tree (e.g. Lowman and
Heatwole 1992, Edwards et al. 1993).

Thus despite a decade or so of research into
eucalypt dieback and defoliation, the causative factors
were far from clear in autumn 1990 when the eucalypt
monitoring project described below commenced.
It was designed to answer the two questions posed
earlier and at the same time test some of the then
current explanations for chronic defoliation. Its
primary aim was to monitor damage by Scarabeidae,
which was seen as the major threat at the time. The
aims of the project were to:

e document fluctuations in the extent of leaf
damage to a large number of eucalypts over a
period of time;

e examine a subset of the above (tagged trees)
for evidence of consistent differences in the level
of leaf damage to individual trees over time;

e investigate whether damage was associated
with environmental factors such as _ those
suggested by Landsberg et al. (1990) above.

METHODS

Site selection and description

It was considered that about ten sites in each of
three main agricultural landscapes (Kovac et al.’s
(1990) “Manildra’, ‘Canowindra’ and ‘Black Rock’
soil landscapes; Table 1) in the Manildra—Molong
area would be sufficient for the survey. In early
1990, readily-locatable (e.g. road junctions, railway
crossings) groups of trees extending from road
reserves into adjacent paddocks were marked on aerial
photographs of the Molong 1:50 000 topographic

Proc. Linn. Soc. N.S.W., 128, 2007

W.S. SEMPLE AND T.B. KOEN

Table 1. Brief descriptions of the three Central Western NSW soil landscapes sampled (Kovac et al.
1990, B.W. Murphy pers. comm.).

er dscane Lithology Soils
Black Rock _— Sandstone, Red podzolics on upper to
(Br) conglomerate, midslopes. Yellow podzolics in
shale drainage lines.
Low to moderate fertility
Canowindra _ Porphyry, Non-calcic brown soils with
(Cd) shale, some red podzolics or red earths
limestone on upper-midslopes. Yellow and
brown solodics in some drainage
lines.
Moderate to high fertility.
Manildra Shale, Non-calcic brown soils on mid
(Mn) porphyry, to upper slopes. Red and brown
limestone podzolics with some red earths

Topography

Rolling low
hills. Relief 60-
80 m. Slopes
8-20%

Undulating low
hills. Relief 20-
60 m. Slopes
2-8%

Undulating to

rolling low hills.

Relief 20-80 m.

Main land use

Grazing
— often of native
pastures.

Broadacre crops
and pastures in
rotation.

Broadacre crops
and pastures in
rotation.

and euchrozems on higher slopes.

Slopes 6-16%.

Red brown earths on lower
slopes. Yellow and red solodics in

drainage lines.

Moderate to high fertility but
tending to be more ‘patchy’ than

Canowindra.

map. Each pre-selected site was then assessed in the
field for satisfying the following criteria: (a) presence
of at least 20 naturally-occurring eucalypts along the
roadside and 20 in the adjacent paddock, (b) not near
buildings where tree plantings may have occurred,
and (c) where possible, not in drainage lines (which
represented a different environment to most of the
country being surveyed). If a site was unsuitable, and
most were due to insufficient trees being present, the
surveyor proceeded to the next site.

Ultimately six pairs of sites in each landscape
were selected but one in the Black Rock landscape
was subsequently abandoned due to difficulty in
distinguishing two closely-related species. The
locations of the remaining paired sites, Black Rock
(Br) 1, 3, 4, 5 and 6, Canowindra (Cd) 1-6 and
Manildra (Mn) 1-6 are shown in Fig. 1. Apart from
Br3 (Fig. 2) and Br4, none of the stands was near-intact
and all had been thinned to varying extents. As roads
were often located in transition areas between two
vegetation communities, and often at lower elevations
than upslope paddocks, species composition of the
roadside community did not always match that of the
adjacent paddock, e.g. Br6, Cd6, Mn1, 2 and 3 (Table
2). Sizes of sites ranged from up to ~300 m of roadside
to ~1 hectare of paddock. In addition to soil landscape

Proc. Linn. Soc. N.S.W., 128, 2007

and location (roadside or paddock), the composition
of the groundstorey (native v. exotic grass dominant)
was also recorded at each site.

Trees were identified using local tree
identification guides but many specimens were
submitted to the Royal Botanic Gardens, Sydney,
for confirmation. Red gums presented particular
difficulty in the Molong—Manildra environment.
Both E. blakelyi and E. dealbata were present but
intergrades, sometimes identified as ‘E. blakelyi ssp.
irrorata’, were particularly common. For this reason,
all the red gums encountered in the survey have been
treated together as “E. blakelyi-dealbata’ .

Assessment of insect damage

Assessments of the ‘dieback condition’ of
entire trees, e.g. as described by Landsberg (1989),
are more likely to reflect past rather than current
levels of defoliation and hence were of limited use
for annual monitoring as proposed here. Many
assessments of defoliation have been based on time-
consuming measurements of the volume of foliage
that has been removed from a tree. Visual estimates
of the proportion of leaf loss are more rapid and, as
reported by Landsberg (1989), can be estimated

101

INSECT DAMAGE TO WOODLAND EUCALYPTS

ISsNorS

Manildra

Figure 1. Location of the 17 paired eucalypt monitoring sites (A) in the Molong—Manildra area of

Central Western NSW.

(though often over-estimated) in a consistent manner
by a single observer but may vary between observers.
For the purposes of this survey, damage assessment
needed to be rapid and consistent, especially from year
to year, though not necessarily an absolute measure
of damage. Instead, an assessment of the proportion
of damaged leaves on each tree was undertaken.
This involved examining a clump of leaves on a tree
and estimating the proportion that showed obvious
damage, i.e. could be easily seen with the unaided eye
or with binoculars. This was repeated for a number of
clumps and the results averaged. Hence, a rating of 10
% meant that, on average, 10 % of the leaves on that
tree had obvious damage. ‘Damage’ included removal
of sections of leaf (either holes or along the margins),

102

presence of galls, “skeletonisation’ and other forms
of leaf stress. Removal of entire leaves could not be
estimated. In the first season (summer of 1989-90), a
three point scale was employed: “nil/minor’ damage
(<10 % of leaves were damaged), ‘moderate’ (>10—
50 %), ‘severe’ (>50 %). In subsequent years, the
proportion of damaged leaves was estimated to the
nearest 10%.

The procedure did not, therefore, distinguish
damage by Scarabeidae, which produce characteristic
sawtooth-like edges on leaves, from other forms of
insect damage; though whether damage was primarily
due to beetles was noted on field sheets. Another
consequence of this procedure was that removal of
a large part of an individual leaf was considered the

Proc. Linn. Soc. N.S.W., 128, 2007

W.S. SEMPLE AND T.B. KOEN

Figure 2. A relatively intact grassy E. albens woodland in a paddock at site Br3 during autumn
2000. This site has since been further degraded by tree removal and the planting of exotic trees.

same as removal of a small part.

It was also assumed that leaves being assessed
were different from those assessed the previous year,
i.e. that leaves were <1 year old. Although reports of
individual eucalypt leaves surviving for up to 3 years
exist (e.g. Lowman and Heatwole 1992), Jacobs
(1955) reported the average leaf-life of forest trees at
about 1.5 years, dependent on factors such as position
in the canopy, growth rate, wind changes and insect
attack. Jacobs noted that bursts of growth such as
occur after insect attack are associated with increased
leaf-fall such that the average life of remaining leaves
is <1 year. Bursts of growth that commonly occur on
woodland eucalypts in autumn and spring may also
be associated with increased leaf fall; and even if
some leaves live for >1 year, these flushes ensure that
many are <1 year old.

Site monitoring

Commencing in April 1990, the first 20 trees,
regardless of size, encountered along the roadside at
each site were assessed for insect damage. Species
and the main leaf type (adult, juvenile) were also
noted for each tree. This was repeated in the adjacent
paddock. The sites were reassessed each year — ideally
in March when Scarabeidae activity had ceased and

Proc. Linn. Soc. N.S.W., 128, 2007

before new autumn leaves were obvious — until 2004.
Trees were assessed by the same observer [WS] on
all occasions. Between 1991 and 1993, trees at five
of the sites (Table 2) were tagged so that damage to
individual trees could be monitored. When tree deaths
occurred or a tree could not be relocated, additional
trees were tagged to maintain the number of trees
monitored at 20.

The same groups of untagged trees were assessed
each year but this did not necessarily involve exactly
the same trees in each group. In most cases this was due
to tree deaths (natural and deliberate) and excessive
browsing by stock but in one case by the incorporation
of roadside trees into the adjacent paddock by new
fencing. These changes necessitated the assessment
of new trees, which were sometimes at a considerable
distance from the original site. Sites Mn1 (roadside)
and Mn2 (paddock) were abandoned in 2001 due to
insufficient numbers of trees being present within
reasonable walking distance. Monitoring of paddock
trees was progressively scaled back from 2001 (when
only tagged trees were monitored) to 2003 when only
roadside trees were monitored.

103

INSECT DAMAGE TO WOODLAND EUCALYPTS

Table 2. Brief descriptions of eucalypt sites monitored from 1990 to 2004

ROADSIDE TREES PADDOCK TREES

Siteno. § oud alb mell iror brid micr poly (juv) Ground alb mell iror brid micr poly (uv)
Brl * exotic 30 0 70 0 0 0 25 exotic 11 0 89 0 0 0 0
Br3 * native 66 0 34 0 0 0 OD) native 82 0 19 (0) 0 0 21
Br4 exotic 100 0 0 0 0 0) 52 native 92 0 8 0 0 () 44
Br5 exotic Del 44 30 0 () 0 29 exotic 7 79 14 0 0 0 2
Br6 exotic 0 51 49 0 0 0 16 native 0 0 100 0 0 0 49
Mean* 44 19 37 0 0 0 29 39 15 46 0 0 0 22
Cdl native 100 0 0 0 0 0 58 exotic 100 0 0 0) 0 0 0
Cd2 native 100 0 0 0 0 0 64 exotic 100 0 0 0 0 0 17
Cd3 native 100 0 0 0 0 0 4 exotic 100 0 0 0 0 0 0
Cd4 exotic 95 5 0 0 0 0 26 exotic 100 0 0 0 0 0 i
Cd5 exotic 1 83 16 0 0 0 2), exotic 0 95 6 0 0 0 0
Cd6 * exotic 0 50 0 50 (0) 0 20 native 25 5 70 0 0 0 0
Mean 66 23 3 8 0 0 29 69 17 15 0 0 0 3
Mnl exotic 51 45 0 0 4 0 VW exotic 63 1 0 0 37 0 9
Mn2 * native 80 | 0 0) 7 13 17 exotic 5 0 95 0 0 0 28
Mn3 exotic 0 Di 73 0 0 0 21 exotic 0 89 1 10 0 0 1
Mn4 * exotic 0 719 16 5 0 0 16 exotic 0 80 5 15 0 0 0
Mn5 native 65 35 0 0 0 0 25 native 56 44 0 0 0 0 5
Mn6 native 0 0 0 0 100 0 () native 0 0 0 0 100 0 1
Mean 32 31 15 1 19 2 24 20 38 15 5 23 0 7

#Means based on actual tree numbers over all observations and rounded to nearest integer

Proc. Linn. Soc. N.S.W., 128, 2007

104

W.S. SEMPLE AND T.B. KOEN

Data analysis

Over the 15 years of observation, 680 individual
trees were examined, yielding 8804 assessments of
leaf damage. Statistical analysis of this large data
set was complicated by the highly structured and
confounded nature of the classifying factors of: soil
landscape (Canowindra, Black Rock, Manildra),
location (paddock, roadside), pasture type (exotic,
native), species (E. albens, E. microcarpa, E.
melliodora, E. bridgesiana, E. blakelyi-dealbata),
leaf type (adult, juvenile) and time (1990, 1991 ...
2004). For example, E. microcarpa was only found
on the Manildra soil landscape and primarily on sites
with a native pasture understorey. Confounding of
this degree could result in the misinterpretation of the
cause of an apparent statistically significant outcome.
To minimise the chance of misinterpretation, a
number of separate linear mixed model analyses were
run. Initially all data were used to quantify the extent
of leaf damage over the full period of observation,
with focus only given to the main order effect of time.
Three further analyses were performed on subsetted
data formed by selecting the major species, viz. E.
albens, E. melliodora and E. blakelyi-dealbata, with
4033, 2106 and 1808 records respectively. Where
estimable, only main-order (e.g. soil landscape, time)
and interactions up to first order (e.g. soil landscape
by time) were included in these models.

The response variable for the analyses, viz.
proportion of leaves damaged on each tree, covered
the full range of 0 to 100%. Therefore to statistically
model the proportion, p, of severely damaged leaves
on a tree as a function of the above explanatory
factors, a logistic transformation (Cunningham et al.
2005) was applied within a generalised linear mixed
model framework using GenStat (2005). Significance
of interaction terms was tested by examining Wald
statistics formed by dropping, in turn, these individual
terms from the full fixed model. High degree splines
were fitted to these estimated means to indicate
general temporal trends.

Correlation coefficients were derived between
the estimated proportion of leaves damaged per tree
from the analysis of all data and various measures
of moisture status. One measure examined actual
rainfall in the form of either total annual, warm season
(September to February) or cool season (March to
August) rainfall lagged from the time of observation
by 1, 2, 3 or 4 years. Another measure re-expressed
each of these measures of rainfall as a standardised
index by creating a running mean of the three most
recent rainfall records, subtracting three times the
long term average and then dividing by the long term
average. This standardisation is similar to that used
by Foley (1957) and is referred to here as ‘three-year

Proc. Linn. Soc. N.S.W., 128, 2007

Foley’s index’.

To investigate the incidence of severe leaf
damage (defined as a tree having 240 % of leaves
damaged) in consecutive years, a contingency table
was formed from the annual damage levels of the
206 individually tagged trees. For trees that were
continuously monitored from 1991—93 to 2004, there
were 13-11 opportunities of severe leaf damage in
consecutive years.

RESULTS

Rainfall (Fig. 3a) during the 14 years of
monitoring varied considerably with occasional wet
years (e.g. 1992) and droughts (e.g. 1994-1995, 1997,
2002). The three-year Foley’s index (which provides
some indication of cumulative soil moisture over the
relevant period) for annual rainfall, peaked in 1992
reflecting a run of average to above-average seasons
but dipped in 1995 and 2003 reflecting runs of below-
average rainfall. A similar pattern was evident for
cool season rainfall (Fig. 3b) but not for warm season
rainfall where indices for the periods 1986—90 and
2002-03 were negative, and those for 1992—98 were
positive (Fig. 3c).

In Fig. 4b, it is evident that overall insect damage
was moderate when assessed in most autumns with
~20% of leaves on the ‘average tree’ damaged.
Significantly higher average damage levels of 33 %
and 37 % were evident in 1990 and 1994 respectively.
When damage to the three most common species, E.
albens, E. melliodora and E. blakelyi-dealbata, was
examined (Figs. 4c, d, e), time (year of observation)
was again the main significant explanatory variable
except for E. albens and E. melliodora, where soil
landscape interacted with time.

Much of the damage up to about 1995 was due to
Scarabeidae. Maximum leaf damage was recorded in
1990, anacknowledged ‘bad beetle year’ (Dick 1990a),
and again in 1994, though at this time damage by
other insects was also evident. Differences in damage
levels between tree species were suspected across the
period of observation but the statistical significance
was not determined due to the potential confounding
of effects. Nevertheless, when Scarabeidae were the
most obvious reason for damage, average damage to
species appeared to decrease in the following order:
E.albens or E. blakeleyi-dealbata, E. bridgesiana, E.
microcarpa and E. melliodora. Eucalyptus melliodora
was severely damaged in 1993-95 but leaf damage
(‘withered and dehydrated’) was consistent with the
occurrence of non-lerp-forming Psyllidae, rather than
Scarabeidae. From 1995 onwards, species ranking
with respect to damage changed, though E. albens

105

INSECT DAMAGE TO WOODLAND EUCALYPTS

(a)

Total Rainfall (mm)

(b)

Cool Season
Rainfall (mm)

(c)

Warm Season
Rainfall (mm)

Figure 3. (a) Annual, (b) Cool season (March to August) and (c) Warm
season (September to February) rainfall at Manildra P.O. in the years im-
mediately prior to and during the insect damage survey. Long term (1960-
90) averages shown by horizontal lines and three-year Foley’s indices by

undulating lines.

generally showed the most damage.

Apart from time of observation, very few of
the potential explanatory variables were statistically
significant in explaining leaf damage. Past rainfall,
however, correlated significantly with leaf damage.
Across all years of observation, highest correlations
were with Foley’s three-year moisture index of 4
years previous. There was a negative correlation
with the warm season index and a lower but positive
correlation with the cool season index (Table 3). The
relationship can be seen by comparing Fig. 4b (annual
leaf damage) with Fig. 4a, where the indices have
been staggered by 4 years. A deficit in warm season
rainfall was associated with higher leaf damage (at
least in 2 of the 15 years) and an increase by reduced
levels of damage 4 years later. An opposite trend was
evident for cool season rainfall.

However, during the period 1996-2004 (when
Scarabeidae were not particularly active), leaf damage
correlated negatively with the calendar year rainfall (r
= —0.86) and to a lesser extent, warm season rainfall (r

106

= —0.66) immediately prior
to the time of observation.
Repeated observations
of the same (206 tagged)
trees indicated that 24 %
(49) of the trees never
experienced severe leaf
damage by insects; a further
37 % (76 trees) experienced
severe leaf damage on up
to three occasions but not
in consecutive years. Most
trees (61 %), therefore,
were not subjected to severe
leaf damage in consecutive
years. One tree, however,
experienced ten severe
attacks over 14 _ years,
eight of which were in
consecutive years. Between
these two extremes, there
were low numbers of trees
that experienced severe leaf
damage in consecutive years

(Fig. 5).

Foley's 3 yr Foley's 3 yr

Foley's 3 yr

DISCUSSION

As it turned out, the
severe defoliation caused
by Scarabeidae in 1989-
90 was not a precursor of
widespread dieback in the Molong—Manildra area
of Central Western NSW. Nevertheless the number
of trees that died during the 15 years of survey was
surprising. In some cases, the cause was obvious
(e.g. deliberate removal associated with roadworks),
suggestive (e.g. severe browsing of saplings by
domestic stock or mature trees with a heavy mistletoe
burden) or not evident (e.g. rapid death of mature E.
melliodora) but, apart from the few cases of the latter,
it could not be described as ‘dieback’ (rapid and
widespread death; Nadolny 1995). However, further
south in the Boorowa—Young—Harden area, dieback
has been evident for many years and, for E. blakelyi
at least, appears to be associated with severe damage
by Psyllidae.

Damage by Scarabeidae was evident every
summer, particularly in 1989-90 and 1993-94, but
their role in overall damage declined after about
1994 and thereafter appeared to be confined mainly
to E. albens and E. blakelyi-dealbata. Reference

Proc. Linn. Soc. N.S.W., 128, 2007

W.S. SEMPLE AND T.B. KOEN

. Sigler a) Moisture
1.0 7 ars ( )

Moisture index
(@)
(@)

1986 1988 1990 1992 1994 1996 1998 2000

Year of Observation
1990 1992 1994 1996 1998 2000 2002 2004

(b) all data

~
™
—@— Canowindra

—©— Black Rock
see7--= Manildra

% leaves per tree exhibiting obvious leaf damage

(e) E. blakelyi - dealbata

1990 1992 1994 1996 1998 2000 2002 2004

Year of Observation

Figure 4. (a) Three-year Foley’s rainfall indices for cool season (C) and warm season (W) rainfall four
years prior to the average damage levels recorded for (b) all eucalypt trees, (c) E. albens trees, (d) E mel-
liodora trees and (e) E. blakelyi-dealbata trees when assessed each autumn from 1990 to 2004. Damage
is expressed as the estimated percentage of leaves with obvious insect damage on each tree and graphed
on the logit scale.

Proc. Linn. Soc. N.S.W., 128, 2007 107

INSECT DAMAGE TO WOODLAND EUCALYPTS

Table 3. Correlations between average leaf damage (“o) and previous (1 to 4 years) rainfall totals (cal-
endar year, warm season (September to February), cool season (March to August) and corresponding

three-year Foley’s indices.

ene: Seok ee! Dee yer
Total Foley’s Total Foley’s Total Foley’s
rainfall Index rainfall Index rainfall Index
4 years previously 0.26 0.28 0.52 0.51 —0.50 —0.76
3 years previously 0.20 0.19 0.43 O57 0.15 039
2 years previously 0.25 0.39 0.10 0.47 0.13 —0.15
| year previously —0.24 0.09 —0.06 0.20 —0.20 0.05

to Fig. 4 suggests that there was a low background
level of damage (i.e. ~20 % of leaves with obvious
damage), which was exceeded in occasional years

when Scarabeidae (or possibly

of E. melliodora around 1993—95) were active. These
results are consistent with the long-term observations
of Pook et al. (1998) in an E. maculata forest on the
south coast of NSW. No surveys were carried out in

2005 and 2006 but incidental
observations suggested
no major change in levels
of either leaf damage or
abundance of Scarabeidae.
This was corroborated by
a Molong landholder, who
observed that localised
explosions in the populations
of Anoploganthus spp. were
consistently associated with
severe damage to E. scoparia,
an introduced species in the
homestead garden.

The unbalanced nature
of the data, e.g. that some
species (E. bridgesiana and
E. microcarpa) did not occur

on all soil landscapes, and that species composition
at each roadside v. paddock ‘pair’ was not always
the same (Table 2) presented considerable difficulty

for analysis and interpretation.

made at the time the survey was designed in 1990

were subsequently considered

assumed that: (1) paddock sites were likely to be
more fertile (e.g. due to stocking and applications of
fertiliser) than those on roadsides, (2) that trees with
juvenile leaves were juveniles, i.e. young, (3) that sites
with an exotic-dominated groundstorey were more

108

fertile than those where natives dominated. As broad
generalisations they may be correct but observations
indicated that they were not consistent at the sites
surveyed. Although tests for significant associations
between the degree of leaf damage and some of the
site variables could not be carried out, it does not rule
them out as potential explanatory variables. However,
the survey results suggest that generalisations such as

Psyllidae in the case

1.00
0.75
0.50

0.25

Proportion of trees

0.00

1 2 3 4 5 6 7

Frequency of severe leaf damage in consecutive years

Figure 5. Proportion of trees (n = 206) that experienced varying frequen-
cies (potential maximum = 13) of severe leaf damage (>40 % of leaves
damaged) by insects in consecutive years.

“paddock trees are more likely to suffer severe insect
attack than those on roadsides’ or that ‘trees in intact
stands, e.g. those with native groundstorey, are less
likely to suffer severe insect attack’, are invalid.
Monitoring of tagged trees suggested that the
generalisation arising from the “nutrient enrichment’
explanation for chronic defoliation, i.e. that “the same
trees are attacked year after year’, is also invalid — at
least for damage by the totality of insects that feed
on eucalypts. Whether it is true when Scarabeidae
are abundant year after year cannot be determined

Further, assumptions

invalid: e.g. it was

Proc. Linn. Soc. N.S.W., 128, 2007

W.S. SEMPLE AND T.B. KOEN

from the data available, though it too is unlikely.
Scarabeidae were particularly active in the first 5 years
of the observation period, yet very few trees suffered
severe leaf damage in consecutive years on four or
more occasions (Fig. 5). A relatively small number
of trees suffered severe leaf damage in consecutive
years suggesting that some trees were particularly
susceptible to insect damage; though it was possible
that some (unhealthy?) trees did not produce frequent
flushes of new leaves and that the same leaves were
observed in consecutive years. Even so, 61 % of the
trees never experienced severe damage in consecutive
years and of these, some were never severely damaged
during the 15 years of observation.

In the case of leaf damage to the two most
common species, E. albens and E. melliodora, a
significant interaction was detected between time
and the soil landscape in which the trees occurred.
Whether this can be attributed to differing habitat
qualities (e.g. ‘fertility’) for build-up of insect
populations in average seasons (but of less relevance
when insects are more mobile in times of high
abundance) is conjectural. Of the three landscapes,
soils in Black Rock were the least fertile (Table 1) and
average levels of leaf damage were generally lower
on this soil landscape except during the high damage
years of 1990 and 1994 (Figs. 4c, d).

Considering the large number of insect species
that feed on eucalypts (e.g. Landsberg and Cork
1997) and the effects of seasonal conditions on their
abundance and on the insects and parasitoids that
feed on them (Heatwole and Lowman 1986), it was
surprising that correlations between leaf damage and
rainfall across the 15 years of survey, or even during
1996-2004 (when Scarabeidae were uncommon),
were evident at all. That the two correlations were
negative (r = —0.76 for the three-year Foley’s index
for warm season rainfall and leaf damage 4 years
later; and r = —0.86 during 1996-2004 for calendar
year rainfall and leaf damage the following autumn)
is difficult to explain. A correlation between rainfall
and insect abundance some years later would be
expected for Scarabeidae, which have long life
cycles, but not for those with short life cycles such
as Psyllidae. Hence, unless one or other type of insect
was responsible for the fluctuating levels of leaf
damage over time, correlations between leaf damage
and rainfall would not be expected. Since field
studies cannot afford the benefits of a fully controlled
laboratory study, these correlations may be indicative
of indirect associations with other non-measured
variables. Perhaps the correlations should be ignored
and other aspects of the rainfall record examined. For
example, it was assumed at the time that the decline

Proc. Linn. Soc. N.S.W., 128, 2007

in the abundance of Scarabeidae after 1994 was due
to drought, especially below-average cool season
rainfall, in 1994 and 1995 (Fig. 3).

CONCLUSIONS

Monitoring of insect damage to eucalypt leaves
over 15 years indicated significant change over time
with high damage levels evident in autumn 1990 and
1994. Though similarly high levels of damage were
recorded in particular situations at other times (e.g. E.
albens trees in the Manildra soil landscape in 2004),
overall background levels of proportions of leaves
damaged fluctuated between 10 and 25 %. At the
commencement of the monitoring period, Scarabeidae
were expected to be an on-going cause of leaf damage
but this did not eventuate and damage attributed to
them declined from 1994 to relatively low levels.
This was possibly in response to a prolonged deficit
in warm season rainfall up to 4 years earlier, though
the 1994-95 drought could not be ruled out. Damage
by other leaf-feeding insects, including Psyllidae,
was evident throughout and they were the main cause
of damage, which was generally low but high in some
species and/or situations, after about 1994.

Clear relationships between leaf damage and
site factors could not be determined but the results
provided little support for generalisations such as
‘trees in grazed paddocks suffer more insect damage
than those on roadsides’ or that ‘trees in disturbed
communities suffer more damage than those in
‘intact’ communities’. Relative damage to individual
species changed over time but during the period when
Scarabeidae were abundant, damage tended to be
higher on E. blakelyi-dealbata and E. albens than on
the other species. Studies of tagged trees offered little
support for the generalisation that once a tree suffers
severe damage it is likely to do so in consecutive
years. Though this occurred to a low number of
trees, many were not severely damaged at all or only
occasionally.

ACKNOWLEDGMENTS

Thanks to the former Regional Director of the Soil
Conservation Service in Orange, Tony McCarthy, who
initiated and supported this project though we must apologise
for taking 16 years to answer his questions of 1990 when they
probably could have been answered prior to his retirement
in 1999. Thanks to local landholders, Brian Evans and Bill
Marriott, who showed ongoing interest in the project, to
Penny Edwards, Suzanne Prober, Rod Fensham and Brian

109

INSECT DAMAGE TO WOODLAND EUCALYPTS

Murphy for providing useful information during the long
period when data were being collected and analysed, to
Jeff Bradley for preparing Figure 1, and to botanists at the
Royal Botanic Gardens, who helped with tree identification
at the start of the survey. Dr A. Raman kindly reviewed the
manuscript and provided some insights into the insects that
may have been responsible for eucalypt damage.

REFERENCES

Anon. (1990). CSIRO progress on insect-defiant gums.
The Land 1 November 1990, p. 25.

Cunningham, R.B., Lindenmayer, D.B., MacGregor,

C., Barry, S. and Welsh, A. (2005). Effects of trap
position, trap history, microhabitat and season on
capture probabilities of small mammals in a wet
eucalypt forest. Wildlife Research 32, 657-671.

Curtis, D. (1989). Eucalypt re-establishment on the
Northern Tablelands of New South Wales. MSc
Thesis, University of New England, Armidale.

Dick, A. (1990a). Beetles attack thousands of trees in
Central West. The Land 15 March 1990, p. 20.

Dick, A. (1990b). New threat faces billion trees plan. The
Land 14 June 1990, p. 18.

Edwards, P.B., Wanjura, W.J., Brown, W.V. and Dearn,
J.M. (1990). Mosaic resistance in plants. Nature 347,
434.

Edwards, P.B., Wanjura, W.J. and Brown, W.V. (1993).
Selective herbivory by Christmas beetles in response
to intraspecific variation in Eucalyptus terpenoids.
Oecologia 95, 551-557.

Foley, J.C. (1957). ‘Droughts in Australia. Review
of records from earliest years of settlement to
1955’. Bulletin No. 47, Bureau of Meteorology,
Commonwealth of Australia, Melbourne.

Fox, L.R. and Morrow, P.A. (1983). Estimates of
damage by herbivorous insects on Eucalyptus trees.
Australian Journal of Ecology 8, 139-147.

Floyd, R.B. and Farrow, R.A. (1995). The role of insects
in tree decline. In: “Redressing rural tree decline.
Proceedings of the ‘after dieback’ conference, May
1995, Orange, NSW’ (Ed A. Kater) pp. 25-33.
(Greening Australia, Sydney).

GenStat (2005). ‘GenStat for Windows.’ 8" edn. VSN
International Ltd., Hemel Hempstead, UK.

Heatwole, H. and Lowman, M. (1986). “Dieback, death
of an Australian landscape’. Reed Books, Frenchs
Forest.

Jacobs, M.R. (1955). “Growth habits of the eucalypts’.
Forestry and Timber Bureau, Canberra.

Kovac, M., Murphy, B.W. and Lawrie, J.W. (1990). ‘Soil
landscapes of the Bathurst 1:250 000 sheet’. Soil
Conservation Service, Sydney.

Landsberg, J. (1989). A comparison of methods for
assessing defoliation, tested on eucalypt trees.
Australian Journal of Ecology 14, 423-440.

110

Landsberg, J., Morse, J. and Khanna, P. (1990). Tree
dieback and insect dynamics in remnants of native
woodlands on farms. Proceedings of the Ecological
Society of Australia 16, 149-165.

Landsberg, J.J and Cork, S.J. (1997). Herbivory:
interactions between eucalypts and the vertebrates
and invertebrates that feed on them. In: “Eucalypt
ecology: individuals to ecosystems’ (Ed. J.E.
Williams and J.C.Z. Woinarsk1) pp. 342-372.
Cambridge University Press, Melbourne.

Lowman, M.D. and Heatwole, H. (1992). Spatial and
temporal variability in defoliation of Australian
eucalypts. Ecology 73, 129-142.

Nadolny, C. (1995). Causes of tree decline/dieback in
NSW. In: “Redressing rural tree decline. Proceedings
of the ‘after dieback’ conference, May 1995, Orange,
NSW’ (Ed A. Kater) pp. 11-18. (Greening Australia,
Sydney).

Pook, E.W., Gill, A.M. and Moore, P.H.R. (1998). Insect
herbivory in a Eucalyptus maculata forest on the
south coast of New South Wales. Australian Journal
of Botany 46, 735-742.

White, T.C.R. (1986). Weather, Eucalyptus dieback in

New England, and a general hypothesis on the cause
of dieback. Pacific Science 40, 58-78.

Wylie, F.R. and Landsberg, J. (1990). Rural dieback. In:
“Trees for rural Australia’ (Ed K.W. Cremer) pp. 243-
248. (CSIRO / Inkata Press, Melbourne).

Proc. Linn. Soc. N.S.W., 128, 2007

The Spatial Pattern of Invading Pinus radiata

Morera C. WILLIAMS AND GLENDA M. WARDLE

Institute of Wildlife Research, School of Biological Sciences, Heydon-Laurence Building, A08, The
University of Sydney, Sydney, NSW 2006 (mwilliam@bio.usyd.edu.au)

Williams, M.C. and Wardle, G.M. (2007). The spatial pattern of invading Pinus radiata. Proceedings of
the Linnean Society of New South Wales 128, 111-122.

The spatial pattern of invading populations can provide insight into mechanisms of invasion and
help establish the potential for further spread of a species. Pinus radiata has successfully invaded native
vegetation across southeastern Australia. The small scale spatial pattern of invading Pinus radiata was
investigated within two dry Eucalypt woodlands adjacent to commercial plantations in the upper Blue
Mountains, NSW Australia. This study aimed to identify the presence of a second generation of pines in
order to determine the sustainability of the invading population. We looked for evidence of 1) clustering of
pine seedlings; 2) positive associations between pine seedlings and reproductive pines. Spatial analysis of
20 m by 20 m plots using dispersion indices and Ripley’s K function revealed clustering of pine seedlings
at distances of up to 450 m from the plantation. Bivariate analysis found significant positive association
between seedlings and reproductive pines in two plots. Further evidence for self propagation was provided
by the correlation between seedling abundance and cone abundance. These results suggest that the invading

population is sustainable in the long term and is capable of spreading further into the native vegetation.

Manuscript received 29 May 2006, accepted for publication 13 December 2006.

Keywords: bivariate analysis, invasion, Ripley’s K, seedlings,

spatial distribution.

INTRODUCTION

The tree Pinus radiata has successfully spread
from commercial plantations into adjacent natural
areas across the southern hemisphere where it is widely
planted as a commercial timber crop (Richardson et al.
1994). Pine invasions are of particular concern in New
Zealand and South Africa where they are threatening
ecological and aesthetic values (Richardson et al.
1994, Ledgard 2001). Dense stands of self sown pines
can suppress understorey vegetation (Richardson and
Van Wilgen 1986, Richardson et al. 1989), and alter
a range of ecosystem properties including hydrology
(Van Wyk 1987), nutrient cycling and fire regimes
(Versfeld and Van Wilgen 1986).

The first phase of invasion by pines involves the
dispersal of seed, typically by wind, from plantation
trees and the subsequent establishment of wildlings
in the recipient habitat. The cones of Pinus radiata
are serotinous, opening and releasing seed during hot,
dry weather conditions and after fire (Fielding 1947,
McDonald and Laacke 2003). While the majority of
seeds are dispersed less than 100m from parent trees
(Van der Sommen 1978 cited in , Virtue and Melland

2003), initial recruits can establish at long distances
of up to 4km from the plantation (Williams and
Wardle 2005) and have been referred to as satellite
foci (Richardson et al. 1994). Yellow-tailed black
cockatoos (Calyptorhynchus funereus) may also be
responsible for long distance dispersal events as they
feed on cones and carry away seed (Attiwill 1970,
Buchanan 1989, Gill and Williams 1996). Rare long
distance dispersal events leads to a broken invasion
front and patchy progress across the landscape
(Hengeveld 1989).

The second phase of invasion occurs when first
generation wildlings mature and begin to reproduce,
creating a secondary seed source. Pines are known to
self-fertilise and perpetuate by establishing a colony
of seedlings (Bannister 1965). Self reproduction from
new loci located beyond the main invasion front
contributes greatly to the invading population (Moody
and Mack 1988) and increases invasive spread rates
(Clark et al. 1998). Seed dispersal processes are
central components of invasion dynamics and are
integral to models of pine spread which hope to gain
a predictive understanding of invasions (Higgins
and Richardson 1998). Determining the presence of

SPATIAL PATTERN OF INVADING PINUS RADIATA

a second generation of invaders and establishing the
time to their successful establishment will assist the
parametisation of models and the estimation of pine
spread rates.

Detecting the occurrence of pine wildling
reproduction requires an ability to distinguish between
seedlings sourced from the plantation and those
recruited from first generation colonisers. The spatial
relationship between adult and juvenile plants can
provide insight into dispersal patterns and seedling
origin. Limited seed dispersal can lead to small scale
aggregation of individuals (Prentice and Werger
1985). Positive associations between seedlings and
parent plants have been attributed to seed dispersal
mechanisms in a savanna palm tree population (Barot
et al. 1999). Similarly, the dispersal of seed by pine
wildlings would result in the aggregation of seedlings
and a positive association between these seedlings
and reproductive adults. Post-germination processes
and conditions including environmental heterogeneity
(Manabe and Yamamoto 1997), herbivore activity
(Janzen 1970) and competition with the mother plant
(Augspurger 1984) will influence seedling survival
and may result in a seedling distribution which
disguises the initial clustering of seeds around the
parent plant. Therefore, the detection of a clustered
spatial pattern after these processes have influenced
seedling distribution provides evidence for the natural
regeneration of pines.

Continuous dispersal of seed from the plantation
itself may also disguise natural regeneration patterns.
Seed in areas near to the plantation will arrive from
two sources; the plantation and first generation
wildlings. Constant seed dispersal from plantation
trees will result in a random pattern of seedlings in
the native vegetation, whereas recruitment from
already established wildlings will lead to a clustering
of seeds around reproductive adults. In areas where
first generation wildlings are the primary seed source,
i.e at distances away from the plantation, you would
expect to find a more detectable clustering pattern.

Invasions of P radiata have been reported
across southeastern Australia (Burdon and Chilvers
1977, Minko and Aeberli 1986, Lindenmayer and
McCarthy 2001, Williams and Wardle 2005), however
quantitative studies detailing the spatial pattern of the
invader are scarce. Evidence for self generation of
pines has been recorded within an invaded eucalypt
forest in the Australian Capital Territory (Chilvers
and Burdon 1983) where young seedlings were found
to be clustered around pines with mature cones.
The Blue Mountains region of New South Wales
is particularly at risk of invasion by P. radiata with
many plantations bordering large tracts of continuous

112

native vegetation. Two areas in particular, Newnes
and Lidsdale State Forests have suffered high levels
of invasion with pine densities exceeding 1,000 per
hectare in areas adjacent to the plantation (Williams
and Wardle 2005). Observations of high numbers of
pine seedlings clustered around reproductive adults at
both sites suggested the presence of self propagation
by wildling trees. This study aims to confirm the
occurrence of wildling reproduction by investigating
the small scale spatial pattern of invading Pinus
radiata within the two Eucalypt woodlands. It was
expected that spatial analysis of the survey area would
identify clustering of pine seedlings and a positive
association between pine seedlings and reproductive
pines.

METHODS

Study Sites

The study took place in two state forests
situated in the upper Blue Mountains in the central
tablelands of New South Wales. Newnes State Forest
(150°12’E, 33°24’S; altitude 1000-1170m) is located
approximately 7.5 km north east of Lithgow, NSW and
encompasses a 51 year old, 2000 hectare P. radiata
plantation. Lidsdale State Forest (150°3’E, 33°26’S;
altitude 900-1000 m) is situated approximately 7 km
north west of Lithgow and includes a 46 year old, 580
hectare P. radiata plantation. Field work took place
in February to August, 2003 within native vegetation
adjacent to the pine plantations.

Field Sampling

At each site, pine occurrence was investigated
within 20 m by 20 m plots located along six transects
placed perpendicular to the plantation boundary.
Transects were placed on multiple borders of the
plantation and ended when pines were no longer
present or terrain prevented further investigation.
Transects ranged in length from 150 m to 2.2 km at
Newnes and from 200 m to 750 m at Lidsdale. Plots
were placed in the nearest vegetation to the plantation
edge avoiding any forestry roads or fire breaks and
were eStablished at regular intervals within each
transect. The distance between plots varied from 100
m to 200 m to reflect changes in pine density. In total
28 plots at Newnes and 23 plots at Lidsdale were
sampled.

Six of the plots at each site in the area of Eucalypt
forest closest to the plantation were mapped. Mapped
plots were established in the first available vegetation
next to the plantation and located 50 m apart. Within
these plots the position of all dead and living P. radiata

Proc. Linn. Soc. N.S.W., 128, 2007

M.C. WILLIAMS AND G.M. WARDLE

individuals was recorded in Cartesian coordinates to
the nearest 0.1 m. For every pine inside the plot the
position of the center of the tree (x and y coordinates)
was recorded, the diameter at breast height, 1.4m
(Dbh) was measured with a tape, the height estimated
to the nearest 0.5 m and the number of cones and
branch whorls were counted. A whorl is the cluster of
branches arising from a node on the stem (Bannister
1962). Non mapped plots were divided into 16, five
metre by five metre quadrats. Within each quadrat,
the number of cones of P. radiata individuals with
whorls was estimated and the Dbh recorded. Pinus
radiata individuals without whorls were counted.

Pinus radiata individuals within all plots were
divided into five categories. 1) Seedlings; no whorls
present, 2) Saplings; whorls present and less than 1m
in height, 3) Juveniles; Greater than 1m in height and
a Dbh of less than 10cm, 4) Adults; trees with Dbhs
of 10cm or more; 5) Reproductive; trees of any size
with cones present.

Spatial Analysis
Dispersion Indices

Non mapped data were analysed using the index
of dispersion (ID) which estimates how a pattern
departs from spatial randomness. The index is
calculated as the ratio of mean to sample variance:

D = YWa, x) / 8
i=l
(Ludwig and Reynolds 1988)

A random arrangement of plants within the
sixteen quadrats will have a frequency distribution
similar to that of the Poisson distribution. Since the
variance and mean are equal in Poisson distributions,
a variance to mean ratio (ID) close to 1.0 is
indicative of a random distribution (Dale 1999).
Dispersion indices below 1.0 indicate a regular
distribution while those greater than 1.0 suggest a
clumped pattern.

Dispersion indices for pine seedlings were
calculated for all plots with more than 10 seedlings
present using the PASSAGE computer program
(Rosenberg 2001). Juveniles and saplings were
excluded from the analyses due to low numbers.
Results were compared to y? values to determine
statistical significance at the p = 0.025 significance
level.

Ripley’s K function
The use of dispersion indices to identify spatial

pattern is limited as it produces a single index of non

Proc. Linn. Soc. N.S.W., 128, 2007

randomness and fails to detect the scale at which this
pattern occurs. Ripley’s K function (Ripley 1977)
considers the variance in nearest-neighbour distances
(Haase 1995) and is favoured for its ability to detect
pattern across a range of spatial scales. A circle of
radius t is centred on each point and the number of
neighbours within the circle is counted. Density (A) is
estimated by dividing the number of individual points
present by the area sampled (A) (A= n/A). Ripley’s
K function, K(t), is defined as the expected number
of points within distance t of a point, as a proportion
of this estimated value for density. Under the null
hypothesis of Complete Spatial Randomness (CSR)
K(t) = mt. That is, the area of a circle of radius t and
a plot of VK(t) versus t should be linear. An estimator
of K(t) is calculated separately for every t (Appendix
1).

The distribution of pine seedlings within the
twelve mapped plots was investigated with Ripley’s
K function using the SPPA computer program (Haase
2002). The analysis began at a radius t of 0 m with
small 0.1 m increments up to 10 m (one half of the
plot length) to investigate small scale patterns. The
sample statistic was plotted as the derived variable
L(t),

L(t)= J(K())/p \-1=0

(Haase 2001)

as it has zero expectation for any value of t when the
pattern is random (Skarpe 1991). This analysis was
only performed for plots with 10 individuals or more,
as a lower number will not reveal consistent patterns
in the spatial distribution (Arevalo and Fernandez-
Palacios 2003).

Monte Carlo simulations produced 99%
confidence intervals. Positive values of L(t) above
the upper limit of the confidence interval signified
clumping at this scale. Significant negative deviation
specifies a regular pattern (Diggle 1983), while L(t)
values that remain within the confidence intervals
support the null hypothesis of Complete Spatial
Randomness (CSR).

Bivariate analysis

Bivariate analysis is an extension of Ripley’s K
function and allows an investigation of the nature of
a relationship between two different life stages of a
species (Couteron and Kokou 1997). Alternatives to
random bivariate patterns are clumped distributions
suggesting positive association or regular patterns
which suggest repulsion between two life stages.

The spatial relationship of two life stages, which

113

SPATIAL PATTERN OF INVADING PINUS RADIATA

may have different densities, 1,=n,/A and A, = n/A
is examined. The function 1,K,,(t) is defined as the
expected number of individuals of species two within
a radius t of an arbitrary individual of species one. The
function AK. (t) gives the expectation for the opposite
spatial relationship. The corresponding estimators are
then combined to a weighted mean single estimator
(Lotwick and Silverman 1982) (Appendix 2). The

derived K-statistic:

LiQ=j Imk,.O=7K1O) / (e™ =7,))-

(Kenkel 1994, Haase 2001)

is then plotted for values of t. If the two species are
independent of one another the expected value of
L,,(t) is 0. Negative values of L,,(t) indicate a
negative association while values of L,,(t) greater
than 0 suggest that the pattern is attractive (Mouer
1993) at that distance.

To determine the relationship between pine
seedlings and reproductive adults, bivariate analysis
was performed for all mapped plots using the SPPA
computer program (Haase 2002).

Analysis of non mapped data
The ability to detect spatial association

between reproductive adults and seedlings within non
mapped plots is restricted by the low resolution of the
quadrat data. This relationship was investigated on
a slightly larger scale by considering the correlation
between the number of cones on reproductive trees
and the number of seedlings in a plot. A square root
transformation of cone data was performed prior
to calculation of correlation coefficients in order to
account for the large variation in cone presence. Plots
located adjacent to the plantation boundary were
excluded from these calculations as it was assumed
that seedlings in these plots are more likely to have
been sourced from the plantation rather than the
reproductive trees present.

RESULTS

Distribution of seedlings

Establishment of seedlings was not equal between
plots with less than 10 seedlings counted in more
than half of the plots surveyed indicating low levels
of invasion in these areas. Eleven of the 28 plots at

Table 1. Clustering in P. radiata seedlings. Sample size (n), Index of dispersion (ID) and signifi-
cance level for seedlings at Newnes and Lidsdale. Asterices (*) indicate plots with significantly

(p<0.025) clustered seedlings.

NEWNES

Plot a eee an ID p-value
1B 50 65 3.27 0.0000*
1C 50 86 2.80 0.0000*
2D 50 A O83) OTD)
72183 50 50 6.08 0.0000*
2F 50 LEIS 1OL0341
3A 50 13. 1.68 0.0479
4A 50 16 1.73 0.0380
1D 150 30 6.18 0.0000*
dale) 250 12. 12.00 0.0000*
iB 350 60 4.53 0.0000*
1G 450 16 2.67 0.0005*

114

LIDSDALE

Plot sega ID p-value
1B 50 13 1.84 0.0241*
IC 50 14 1.41 0.1299
ie 50 77 4.27 0.0000*
5A 50 15 0.92 0.5407
2A 150 13 0.86 0.6142
4A 150 37 3.33 0.0380*
5B 250 22 1.83 0.0252
4B 350 20S Nessie On WAS
DE 550 12 1.16 0.2993

Proc. Linn. Soc. N.S.W., 128, 2007

M.C. WILLIAMS AND G.M. WARDLE

Newnes and 9 of the 23 plots surveyed at Lidsdale had
more than 10 seedlings present and were analysed to
determine the level of clustering. Dispersion indices
for seedling distributions were greater than 1.0 in
ten of the eleven plots at Newnes and in seven of the
nine plots at Lidsdale, indicating tendencies towards
clumping (Table 1). At Newnes, seven plots displayed
significant results with IDs ranging from 2.67 to 12.0.
Clustering was significant in three plots at Lidsdale.
All plots at both sites with more than 25 seedlings
were significantly clustered. There was some evidence
for a greater degree of clustering in plots further from
the plantation at Newnes with all four plots located

Index of Clustering (L (6)

9

more than 50m away returning significantly clustered
results (Table 1). However, clumping was also present
in plots close to the plantation and many plots were
excluded from analysis due to low seedling numbers
preventing a thorough examination on the effect of
distance.

Analysis of the twelve mapped plots using
Ripley’s K function revealed significant clustering of
seedlings across scales from 1 m to 9 m. All plots
at both sites with greater than 25 seedlings present
contained significantly clustered distributions (Fig.

1).

Distance (m)

Figure 1. Spatial analysis of the distribution of all P radiata seedlings for significantly clustered mapped
plots at Newnes (NIB, N1C, N2E) and Lidsdale (L1E). L(t) values > 0 indicate clustering. The dotted
lines give the 99% confidence intervals. Sample size (n) is in the top left corner and plot label is in the

top right corner of each graph.

Proc. Linn. Soc. N.S.W., 128, 2007

115

SPATIAL PATTERN OF INVADING PINUS RADIATA

Index of Association (L 12(t)) —

Distance (m)

Figure 2 a) Mapped locations of P. radiata individuals within 20 m by 20 m plots. Large open circles rep-
resent reproductive P. radiata. Small closed circles represent P. radiata seedlings. b) Positive association
between P. radiata seedlings and reproductive trees in two mapped plots at Newnes. L12(t) values > 0
indicate a positive association. The dotted lines give the 99 % confidence intervals. Plot label is in the top
right corner and sample size (n) is in the top left corner of each graph: seedlings/reproductive trees.

Association between seedlings and reproductive
adults

When seedling number was low, no significant
association between life stages was found at either
Newnes or Lidsdale. Of the twelve mapped plots,
only four had more than 25 seedlings present, two
of which revealed significant positive associations.
Maps indicating the position of P. radiata individuals
in the plots illustrate the clustering of seedlings

116

around reproductive adults (Fig. 2a). Bivariate
analysis revealed a significant positive association
at scales of 0.5 m to 7.5 m in plot 1C at Newnes
and at all distances in plot 2E at Newnes (Fig. 2b).
Seedlings were scattered across the remaining two
plots (Fig. 3a). Bivariate analysis confirmed spatial
independence between seedlings and reproductive
trees despite high seedling abundance (Fig. 3b).

Proc. Linn. Soc. N.S.W., 128, 2007

M.C. WILLIAMS AND G.M. WARDLE

=
——"

Index of Association (L j2(t))

Distance (im)

Figure 3. a) Mapped locations of P. radiata individuals within 20 m by 20 m plots. Large open circles
represent reproductive P. radiata. Small closed circles represent P. radiata seedlings. b) Spatial relation-
ship between P. radiata seedlings and reproductive trees in mapped plots at Lidsdale (LIE) and Newnes
(N1B). L12(t) values > 0 indicate a positive association. The dotted lines give the 99 % confidence inter-
vals. Sample size (n) is in the top left corner of each plot: seedlings/reproductive trees.

Association between seedlings and reproductive
adults in non mapped plots

Cone abundance per 20 m by 20 m plot
was extremely variable ranging from 0 to 131 at
Lidsdale and from 0 to 700 at Newnes. Significant
positive correlations were found between square
root cone abundance and seedling abundance for
far plots at both sites (Lidsdale: r = 0.77, p < 0.001.

Proc. Linn. Soc. N.S.W., 128, 2007

Newnes: r = 0.76, p < 0.01). This relationship was
most noticeable at Newnes where very large numbers
of cones in transect one produced large numbers of
highly clustered seedlings (Table 2). Plots within
transect one had an average of 615 (+ 13.96) cones
and 29.5 (+ 2.72) seedlings per plot. In comparison
transect two contained only three cones in total and
yielded no seedlings.

117

SPATIAL PATTERN OF INVADING PINUS RADIATA

Table 2. Relationship between cone production and seedling presence and dispersion within transects
1 (Plots 1D, 1E, 1F, 1G) and 2 (Plots 2G, 2H, 21) at Newnes. Plots placed adjacent to the plantation are

excluded.
Plot Distance from the Number of Total number Number of
plantation (m) reproductive trees of cones seedlings ID
NID 150 35 685 30 6.2
NIE 250 4 700 1) 12.0
NIF 350 6 620 60 4.5
NIG 450 4 456 16 21)
N2G 150 0 0 0 NA
N2H 250 1 3 0 NA
N2I 350 0 0 0 NA
DISCUSSION suggests that once seedling abundance is sufficiently

Many processes will influence the spatial
distribution of seedlings within a population. These
include seed dispersal mechanisms (Prentice and
Werger 1985, Hatton 1989, Barot et al. 1999), distance
from the mother plant and other seedlings (Janzen
1970, Augspurger 1984) and microsite conditions
such as light availability (Arevalo and Fernandez-
Palacios 2003) that will affect establishment. As
many of these processes will produce similar patterns,
care is needed in inferring causation. The process of
prime interest in this study was that of seed dispersal.
Generally, the result of initial seed dispersal is a
clustered distribution of seeds around the mother
plant (Bigwood and Inouye 1988). Recruitment from
parent plants has resulted in aggregation patterns
in a number of wind dispersed species (Westelaken
and Maun 1985, Hatton 1989). While the seeds of P.
radiata are adapted for long distance dispersal (Van
Wilgen and Siegfried 1986), experimental data from
the study sites indicates that the majority of seeds
released from pine wildlings in the Eucalypt habitat
will fall within 10 m of the parent tree with only rare
long distance dispersal events witnessed. This pattern
of seed dispersal leads to a clustering of seedlings
around the parent plant detectable by the methods
used in this study. Dispersion indices revealed
tendencies towards clumping among seedlings within
most plots. A general trend of greater clustering with
increased sample size was found. This relationship

118

high, a significant spatial pattern can be detected, and
a Clustered pattern is usually observed.

Significant positive associations between
seedlings and reproductive trees were found in two
plots at Newnes providing some evidence for natural
regeneration within areas adjacent to the plantation.
The spatial relationship between seedlings and adults
may change with time. Higher survival and growth
of seedlings close to the parent tree, 5-20 years after
fire, can strengthen the aggregation pattern (Ne’eman
et al. 1992). Spatial independence between seedlings
and reproductive trees in two plots adjacent to the
plantation with high seedling numbers also suggests
that seedlings have originated from the plantation
which therefore appears to provide an ongoing
contribution to recruitment.

Other process unrelated to regeneration may
also produce a pattern of clustered seedlings. Yellow-
tailed black Cockatoos, Calyptorhynchus funereus,
have been known to feed on cones and carry them
away (Buchanan 1989, Gill and Williams 1996) and
were observed feeding on plantation trees at both
study sites. The dropping of an entire pine cone by
these birds may result in an aggregation of seedlings
in the native vegetation, This event could occur at
any distance from the plantation and separating this
process from natural regeneration from wildlings is
difficult. Genetic analysis of the wilding population
is possible as microsatellite markers have already
been developed for the species (Devey et al. 2002).

Proc. Linn. Soc. N.S.W., 128, 2007

M.C. WILLIAMS AND G.M. WARDLE

However, while this may help confirm the patterns
observed in this study it would fail to distinguish
between wind and bird dispersed recruits. Determining
the relative contribution of the two vectors to overall
seed dispersal would be useful for modelling spread
patterns as long distance dispersal events have been
shown to be important for determining invasive
spread rates (Buckley et al 2005).

Stronger evidence for the production of a second
cohort of wildlings was provided by the relationship
between the number of cones within a plot and the
number of seedlings present. Seedlings were numerous
and highly clustered in plots containing highly
reproductive trees, suggesting that self regeneration
of pines is occurring at distances of at least 450 m
from the plantation. This observed pattern concurs
with another study which recorded the establishment
of seedlings around initial colonizers in an invasion
of Pinus radiata into South African fynbos vegetation
(Richardson and Brown 1986). The authors recorded
high pine densities of more than 2,700 individuals per
ha at distances of greater than one kilometre from the
plantation. Investigation of the invading population
found that initial colonisation occurred 13 years after
plantation establishment and that sufficient seed
was being produced by these recruits to generate a
second cohort of invaders within only 21 years of
afforestation. While first plantings of

P. radiata occurred 46 years ago at Lidsdale and
51 years ago at Newnes, the majority of planting took
place between the years of 1973 and 1981 at both
sites. Pinus radiata cone production peaks between
about 10 and 20 years of age (Lewis and Ferguson
1993) which means that a considerable amount of seed
has been available for over 20 years. The presence
of a second generation of wildlings at Newnes and
Lidsdale after this amount of time is not unexpected
considering the timescale of invasion observed by
Richardson and Brown (1986). We observed the
establishment of new seed sources at distances of up
to 450m from the plantation which may lead to further
spread of the population into the native vegetation.

Seed production within the plantation itself has
been shown to be a large determining factor in the
pattern of recruitment in areas surrounding plantations
(Dawson et al 1979). Peaks in the recruitment of first
generation wildlings may result in a wave of second
generation recruitment once wildlings enter peak cone
production. Current knowledge of cone production
within wildling populations is scarce and may assist
with the management of invasions by establishing a
time frame for the establishment of second generation
pine invaders.

While results of this study suggest that first

Proc. Linn. Soc. N.S.W., 128, 2007

generation wildlings are producing a second
cohort of pines, the capacity of these individuals to
contribute to the invading population and advance
the invasion front will depend on their capacity to
survive in their new environment. Environmental
variation including suitable conditions for seedling
establishment and survival will influence the rate of
invasion (Richardson and Bond 1991). Preliminary
investigations of seedling survival at Newnes suggest
that establishment rates and short term survival are
high following fire. However, this study also found
that the majority of plots had very low numbers of
seedlings indicating that while the pines have reached
the second stage of invasion, the establishment
process is slow. Further quantification of seedling
survival rates is required to determine the long term
sustainability of the invading pine population.

When inferring processes from pattern, it is
essential to look at temporal changes of spatial
distributions. Continued processes of recruitment
and seedling mortality will change the spatial
pattern of the invading pines. Monitoring of the pine
population within the study area over a longer time
period is desirable and may provide further evidence
for the generation of a second cohort of wildlings.
More importantly, the quantification of germination
success and survival rates of pines in the native
eucalypt vegetation will help determine spread rates
and evaluate the significance of pines as an invasive
species.

ACKNOWLEDGEMENTS

We acknowledge Joanne Ironside, Yvonne Davila,
Anja Divljan, Praveen Gopalan and Gayle Adams for
assistance with field work. Chris Banffy provided valuable
input and expertise on pine invasions in the Blue Mountains.
Permission to conduct this research was provided by
Department of Environment and Conservation (NSW) and
ForestsNSW and is gratefully acknowledged.

REFERENCES

Arevalo, J. R., and Fernandez-Palacios, M. (2003). Spatial
patterns of trees and juveniles in a laurel forest of
Tenerife, Canary Islands. Plant Ecology 165:1-10.

Attiwill, A. R. (1970). On the spread of pines and bridal
creeper by birds. South Australian Ornithologist
DPX,

Augspurger, C. K. (1984). Seedling survival of tropical
tree species: interactions of dispersal distance, light
- gaps and pathogens. Ecology 65:1705-1712.

19

SPATIAL PATTERN OF INVADING PINUS RADIATA

Bannister, M. H. (1962). Some variations in the growth
patterns of Pinus radiata in New Zealand. New
Zealand Journal of Science 5:342-370.

Bannister, M. H. (1965). Variation in the Breeding system
of Pinus radiata. Pages 353-372 in H. G. Baker and
G. L. Stebbins (editors). The Genetics of colonizing
species. Academic Press, New York, USA.

Barot, S., Gignoux, J and Menaut, J-C. (1999).
Demography of a savanna palm tree: predictions
from comprehensive spatial pattern analyses. Ecology
80:1987-2005.

Bigwood, D. W. and Inouye D. W. (1988). Spatial pattern
Analysis of Seed Banks: An improved method and
Optimised Sampling. Ecology 69:497-507.

Buchanan, R. A. (1989). Bush regeneration. Recovering
Australian landscapes. TAFE NSW, Sydney.

Burdon, J. J., and Chilver,s G. A. (1977). Preliminary
Studies on a Native Australian Eucalypt Forest
Invaded by Exotic Pines. Oecologia 31:1-12.

Chilvers, G. A. and Burdon J. J. (1983). Further studies on
a native Australian eucalypt forest invaded by exotic
pines. Oecologia 59:239-245.

Clark, J. S. Fastie, C., Hurtt, G., Jackson, S. T., Johnson,
C., King, G.A., Lewis, M., Lynch, J. Pacala, S.,
Prentice, C., Schupp, E. W., Webb, T. and Wyckoff,
P. (1998). Reid’s paradox of rapid plant migration:
dispersal theory and interpretation of paleoecological
records. Bioscience 48:13-24.

Couteron, P., and Kokou, K. (1997). Woody vegetation
spatial patterns in a semi-arid savanna of Burkina
Faso, West Africa. Plant Ecology 132:211-227.

Dale, M. R. T. (1999). Spatial Pattern Analysis in Plant
Ecology. Cambridge University Press, Cambridge.

Dawson, M. P., Florence, R. G., Foster, M.B. and
Olsthoorn, A. (1979). Temporal Variation in Pinus
radiata invasion of Eucalypt forest. Australian Forest
Research 9:153-161.

Diggle, P. J. (1983). Statistical analysis of spatial point
patterns. Academic Press, London.

Fielding, J. M. (1947). The seeding and natural
regeneration of Monterey Pine. Forest Timber Bureau
Australian Bulletin 29.

Gill, A. M., and Williams, J.E. (1996). Fire regimes
and biodiversity: the effect of fragmentation
of southeastern Australian eucalypt forests by
urbanisation, agriculture and pine plantations. Forest
Ecology and Management 85:261-278.

Haase, P. (1995). Spatial pattern analysis in ecology based
on Ripley’s K-function: Introduction and methods of
edge correction. Journal of Vegetation Science 6:575-
582.

Haase, P. (2001). Can isotropy vs. anisotropy in the spatial
association of plant species reveal physical vs. biotic
facilitation? Journal of Vegetation Science 12:127-
136.

Haase, P. (2002). SPPA. A program for Spatial Point
Pattern Analysis. 2.0

Hatton, T. J. (1989). Spatial analysis of a subalpine heath
woodland. Australian Journal of Ecology 14:65-75.

120

Hengeveld, R. (1989). Dynamics of Biological Invasions.
Chapman and Hall, London; New York.

Higgins, S. I., and Richardson, D. M. (1998). Pine
invasions in the southern hemisphere: modelling
interactions between organism, environment and
disturbance. Plant Ecology 135:79-93.

Janzen, D. H. (1970). Herbivores and the number of tree
species in tropical forests. The American Naturalist
104:501-528.

Kenkel, N. C. (1994). Bivariate pattern analysis of
jack pine - trembling aspen association. Abstracta
Botanica 18:49-55.

Ledgard, N. (2001). The spread of lodgepole pine (Pinus
contorta, Dougl.) in New Zealand. Forest Ecology
and Management 141:43-57.

Lewis, N. B. and Ferguson, I. S. (1993). Management of
Radiata Pine. Inkata Press, Australia.

Lindenmayer, D. B. and McCarthy, M. A. (2001). The
spatial distribution of non-native plant invaders in a
pine-eucalypt landscape in south-eastern Australia.
Biological Conservation 102:77-87.

Lotwick, H. W. and. Silverman, B. W. (1982). Methods
for analysing spatial processes of several types of
points. Journal of the Royal Statistical Society Series
B 44:406-413.

Ludwig, J. A. and Reynolds, J. F. (1988). Statistical
Ecology. Wiley, New York.

Manabe, T. and Yamamoto, S. (1997). Spatial distribution
of Eurya japonica in an old-growth evergreen broad-
leaved forest, SW Japan. Journal of Vegetation
Science 8:761-772.

McDonald, P. M., and Laacke, R. J. (2003). Pinus radiata
D. Don Monterey Pine. United States Department of
Agriculture Forest Service Northeastern Area. http://
www.na.fs.fed.us/spfo/pubs/silvics_manual/Volume_
1/pinus/radiata.htm Date Accessed: 11/06/2003.

Minko, G. and Aeberli, B. C. (1986). Spread of Radiata
pine into indigenous vegetation in North-eastern
Victoria. State Forests and land service conservation,
forests and lands 30:17-25.

Moody M. E. and Mack, R. N. (1988). Controlling the

spread of plant invasions. Journal of Applied Ecology

25:1009-1021.

Mouer, M. (1993). Characterizing Spatial Patterns of trees

Using Stem-Mapped Data. Forest Science 39:756-

VIS.

Ne’eman, G., Lahav, H and Izhaki, I. (1992).

Spatial pattern of seedlings 1 year after fire in a
Mediterranean pine forest. Oecologia 91:365-370.

Prentice, I. C. and. Werger, M. J. A. (1985). Clump
spacing in a desert dwarf shrub community. Vegetatio
63:133-139.

Richardson, D. M. and Bond, W. J. (1991). Determinants
of plant distribution: Evidence from pine invasions.
The American Naturalist 137:639-668.

Richardson, D. M. and Brown, P. J. (1986). Invasion of
mesic mountain fynbos by Pinus radiata. South
African Journal of Botany 52:529-536.

Proc. Linn. Soc. N.S.W., 128, 2007

M.C. WILLIAMS AND G.M. WARDLE

Richardson, D. M., Macdonald, I. A. W and Forsyth, G.
G. (1989). Reductions in plant species richness under
stands of alien trees and shrubs in the fynbos biome.
South African Forestry Journal 149:1-8.

Richardson, D. M. and Van Wilgen, B. W. (1986). Effects
of thirty-five years of afforestation with Pinus radiata
on the composition of mesic mountain fynbos near
Stellenbosch. South African Journal of Botany
52:309-315.

Richardson, D. M., Williams, P. A. and Hobbs, R. J.
(1994). Pine invasions in the Southern Hemisphere:
determinants of spread and variability. Journal of
Biogeography 21:511-527.

Ripley, B. D. (1976). The second order analysis of
stationary processes. Journal of Applied Probability
13:255-266.

Ripley, B. D. (1977). Modelling spatial patterns. Journal
of the Royal Statistical Society London Series B
41:368-374.

Ripley, B. D. (1981). Spatial Statistics. J. Wiley, New
York, NY.

Rosenberg, M. S. (2001). PASSAGE. Pattern Analysis,
Spatial Statistics, and Geographic Exegesis.1.1.1.3.
Department of Biology, Arizona State University,
Tempe, AZ.

Skarpe, C. (1991). Spatial patterns and dynamics of woody
vegetation in an arid savanna. Journal of Vegetation
Science 2:565-572.

Van der Sommen, F. J. (1978). The colonisation by Pinus
radiata D. Don of Eucalypt-dominated communities
in South Australia. Masters Thesis. University of
Adelaide.

Van Wilgen, B. W. and W. R. Siegfried. (1986). Seed
dispersal properties of three pine species as a
determinant of invasive potential. South African
Journal of Botany 52:546-548.

Van Wyk, D. B. (1987). Some effects of afforestation on
streamflow in the Western Cape Province, South
Africa. Water SA 13:31-36.

Versfeld, D. B. and Van Wilgen, B. W. (1986). Impact
of woody aliens on ecosystem properties. Pages
239-246 in Macdonald, I. A. W., Kruger, F. J. and
Ferrar, A. A. (editors). Ecology and Management
of Biological Invasions in Southern Africa. Oxford
University Press, Capetown.

Virtue, J. G. and Melland, R. L. (2003). The
Environmental Weed Risk of Revegetation and
Forestry Plants. South Australia. Department of
Water, Land and Biodiversity Conservation. Report,
2003/02.

Westelaken, I. L. and. Maun, M. A. (1985). Spatial pattern
and seed dispersal of Lithospermum caroliniense
on Lake Huron dunes. Canadian Journal of Botany
63:125-132.

Williams, M. C. and Wardle, G. M. (2005). The invasion
of two native Eucalypt forests by Pinus radiata in
the Blue Mountains, NSW, Australia. Biological
Conservation 125:55-64.

Proc. Linn. Soc. N.S.W., 128, 2007

WAL

SPATIAL PATTERN OF INVADING PINUS RADIATA

Appendix 1

Unbiased estimator of K(t):

a = i

KOsa Ay, vw, Ue)
I#J

(Ripley 1976, 1981)

A —area of the plot in m*

I - counter variable

OL distance between events 7 and j
w,.— weighting factor to correct for edge effects

Appendix 2

Estimators for bivariate analysis:

. * =|
Ke@)=(mn,)' A> > w, 1,u,)
(Lotwick and Silverman 1982)

Ki()=(4n)' AY > w, 1u,)

Combined estimator:

(m, +n,)—1 [n,K, ¢)+7,K, (2)

122 Proc. Linn. Soc. N.S.W., 128, 2007

Contemporary and Historical Descriptions of the Vegetation of
Brundee and Saltwater Swamps on the Lower Shoalhaven River
Floodplain, Southeastern Australia

Davip A. KEITH, CHRISTOPHER SIMPSON, MArK G. TOZER AND SUZETTE RODOREDA

Biodiversity Conservation Science, NSW Department of Environment and Conservation, PO Box 1967

Hurstville NSW 2220

Keith, D. A., Simpson, C., Tozer, M. G. and Rodoreda, S. (2007). Contemporary and historical
descriptions of the vegetation of Brundee and Saltwater Swamps on the lower Shoalhaven River
floodplain, southeastern Australia. Proceedings of the Linnean Society of New South Wales 128, 123-
153.

Coastal floodplains are functionally important and highly endangered ecosystems in southeastern Australia,
which have a long history of exploitation and environmental modification. In this study, we undertook a
systematic survey of contemporary vegetation in two recently established nature reserves on the south coast
of New South Wales and investigated historical records of the vegetation and environment to infer likely
changes since European settlement. An analysis of floristic samples showed that the present-day floodplain
vegetation includes a mosaic of woodlands, forests and saltmarsh/reedland (five communities) that contrast
markedly in species composition and structure to eucalypt forests that occupy the surrounding hills (two
communities). One hundred and forty-nine plant species were recorded in 24 0.04 ha samples within the
reserves, with Poacaeae and Cyperaceae represented by the most species on the floodplain. Some parts of
the floodplain contain substantial weed infestations, while other parts of the floodplain are largely free of
weeds. The vegetation underwent a series of changes since the first recorded observations in 1805. At that
time the floodplain included a mosaic of woodland, grassland and reedland. Native grassland now appears
to be extinct as a result of subsequent clearing, intensive cattle grazing, pasture improvement and changes
to drainage. A network of drains, initially constructed around 1900 and further developed in the 1960s,
resulted in soil oxidation. This may have made the floodplain soils more suitable for woody plant species,
but recruitment has been largely prevented by intensive cattle grazing. A recent expansion of Casuarina
and Melaleuca scrub and forest is evident within the nature reserves since their dedication and exclusion
of livestock in 2001, but not on adjoining properties where intensive cattle grazing continues. We conclude
that the reserves include important samples of remnant floodplain vegetation and that the vegetation is ina

continuing state of flux regulated by changing flood and tidal regimes and grazing regimes.

Manuscript received 1 November 2006, accepted for publication 15 January 2007.

KEYWORDS: coastal wetlands, Endangered Ecological Communities, European impact, floodplain,
Shoalhaven River, vegetation change, vegetation classification, vegetation history.

INTRODUCTION

Coastal wetlands are functionally important
ecosystems. They are significant carbon sinks,
resource-rich repositories of moisture and nutrients,
habitats for highly specialised plant life and important
breeding grounds for wetland-dependent fauna
including birds, frogs, fish, crustacea, molluscs and
other invertebrates.

Coastal floodplains are perhaps the most
endangered and heavily modified ecosystems in
southeastern Australia (Keith 2004, Keith and Scott
2005). The present status and parlous future of these
remnants is recognised in the listing of six Endangered
Ecological Communities under the Threatened
Species Conservation Act (1995) in New South Wales.
Complex interactions between multiple processes
has transformed temperate coastal floodplains
from natural systems to intensely managed human

landscapes. These processes include clearing of native
vegetation, intensive grazing by domestic cattle,
pasture improvement and cultivation, changes to water
regimes (water table depth, floods, tides), particularly
through construction of drains and tidal gates,
changes in soil chemistry and structure, and invasions
of alien plant and animal species. Collectively, these
processes have had a range of adverse impacts on
the ecological functions of floodplains. For example,
coastal floodplain soils often have high concentrations
of sulfur, which accumulated during the deposition
of marine sediments in the early stage of floodplain
development. Drainage works or other earthworks in
the swamps, expose their soils to oxidation, liberating
sulphuric acid into the soil solution (Johnston et al.
2003). As well as being toxic to plants and animals,
sulphuric acid leaches minerals that would otherwise
be fixed in the soil, including iron, aluminium and
magnesium. Thus, rapid drainage of swamps after

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

rainfall can cause high concentrations of acid and
dissolved metal ions in drains and estuaries, killing
fish and other animals dependent on the estuaries.
The decline of aquatic life in acid-affected swamps
also has impacts on the use of these areas by other
water-dependent fauna, including waterbirds.

Despite the intensity of past and present land
uses, many of the coastal floodplains still support
fragments of native vegetation, albeit in a highly
modified form (Keith 2004). These are maintained
in a state of flux in response to continuing changes
in land use, particularly water management, grazing
and cropping. Under the influences of global climate
change, the remaining native vegetation is likely to
become even less stable in the future. Vegetation
changes have profound effects on the ecological
functions of the floodplains. For example, projected
increases in atmospheric CO, concentrations, rising
temperatures and projected declines in precipitation
and flood frequency are predicted to accelerate the
release of carbon from wetland sediments (Gorham
1991, Freeman et al. 2004) and drive structural change
in habitats of specialised organisms (Hughes 2003).

At present, the mechanisms that regulate the
direction and rates of vegetation change on coastal
floodplains are poorly understood. The experimental
studies needed to elucidate these mechanisms
require baseline data as context for interpreting
future responses to environmental change. While
comprehensive surveys have been carried out in
wetlands on some subtropical latitudes (Pressey
1989a, b, Pressey and Griffith 1992), floodplains of
more temperate latitudes have been comparatively
neglected. On the Shoalhaven River floodplain
120 km south of Sydney, conservation reserves have
been established on parts of two floodplain wetlands,
which provide important reference areas for research
on wetland dynamics in response to changes in both
climate and land use. The aims of this study were
to describe and map present-day vegetation of the
reserves by gathering baseline data from the field, and
to describe past states of the vegetation by reviewing
historical information from the recent past and early
European times.

METHODS

Study area

Brundee Swamp and Saltwater Swamp are
located about 7 km south-east of Nowra (latitude
34°55'S, longitude 150°39'E) on the edge of the
broad, lowland floodplain of the Crookhaven River.
Brundee Swamp is on the upper floodplain of the
Crookhaven River, which subsequently runs through
Saltwater Swamp before entering the ocean close
to the mouth of the Shoalhaven River. The northern
parts of Brundee and Saltwater Swamps are freehold
land and have been used for grazing and associated
cropping for many years (Dalmazzo et al. 2000).
Nature reserves were established in southern parts of
the two swamps in January 2001 on former Crown
land that was subject to permissive occupancies for
cattle grazing by adjoining landholders. Brundee

124

Swamp covers an area of approximately 600 ha, of
which 230 ha are included within the nature reserve.
Saltwater Swamp covers approximately 480 ha, of
which 215 ha are included within the nature reserve.
The two nature reserves are separated by a distance
of about 1 km, comprising partially cleared freehold
land, which occupies a low rise above the floodplain.
To the south and west of the reserves, the floodplain
rises gradually into low forested hills of siltstone and
sandstone, which are partly freehold and partly within
Currambene State Forest.

Brundee and Saltwater Swamps are large, low-
lying, shallow, fresh to brackish wetlands mostly
at or below Mean Sea Level (+1.0m). Some small
areas are inundated semi-permanently, while large
areas of both swamps are inundated periodically
and are without surface water for most of the time
(Dalmazzo et al. 2000). Each swamp has a small
catchment of less than 2 000 — 3 000 ha. The swamps
apparently formed as a result of deposition of marine,
estuarine and fluvial sediments, which infilled old
coastal lagoons as sea levels rose at the end of the last
glaciation (Dalmazzo et al. 2000).

Field Sampling

Floristic composition and vegetation
structure were sampled at 21 sites within the Brundee
and Saltwater Swamp nature reserves. In addition,
four samples were obtained from a previous survey
project (Tindall et al. 2004), including three within
the reserves and one nearby. The sites were located
to cover a range of landforms, structural forms and
geographic locations within the study area including
forest, scrub and sedgelands on the floodplain, forests
on the margins of the floodplain, and forests on low
hills and slopes above the floodplain.

Fieldwork was carried out during 16-17 February
2006. Vegetation sampling methods were identical to
those used by Tindall et al. (2004) and Gellie (2005).
A global positioning system was used to record
the location and elevation of each survey site Tape
measures were used to mark out survey quadrats of
0.04 ha. These quadrats were square (20 x 20 m),
except where different dimensions were required
to ensure that landform and soils were reasonably
homogeneous within the plot (for example an 8 x 50
m quadrat was used along a drainage feature).

All vascular plant species rooted within or
overhanging the quadrat were recorded and assigned
a cover/abundance score using a modified Braun
- Blanquet scale (Poore 1955) as follows: 1- Rare,
one or few individuals present and cover < 5%; 2-
uncommon and cover < 5%; 3- common and cover <
5%; 4- very abundant and cover < 5% or 5% < cover
< 20%; 5- 20% < cover <50%; 6- 50% < cover <75%;
7- 75% < cover < 100%.

The height range and projected foliage cover
were estimated for all vegetation strata recognisable
at the site (e.g. tree, small tree, shrub, groundcover).
At the centre of the quadrat, a compass and clinometer
were used to measure the slope, aspect and horizon
elevations at compass bearings of 0, 45, 90, 135, 180,
225, 270 and 315°. Soils were examined by hand-
texturing and notes made on colour, texture, moisture

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

content and depth. Evidence of outcropping rock,
erosion, weed invasion, logging, soil disturbance or
recent fire was noted.

All native and exotic vascular plant
species were recorded. Plant species that could not
be identified in the field were collected for later
identification. Where necessary, collections were sent
for identification to the National Herbarium of NSW.
Nomenclature was standardised to follow Harden
(1990 - 2002) and Flora Online (http://plantnet.
rbgsyd.gov.au).

To allow for ongoing monitoring of the plots,
a steel star post marked with a fluorescent green
top and a uniquely numbered stainless steel tag was
located at the north eastern corner of each quadrat. A
photopoint was established at each site framing the
corner post from a bearing that was recorded on the
field data sheet. An indicative assessment of fuel loads
was made at each site based on the method for fine
fuels and total fuel load described in the Department
of Environment and Conservation’s Incident Field
Handbook (pp 24 - 25).

To assess spatial variation in growth stages
within each woody vegetation type, the angle count
(Bitterlich) method was used to estimate stand basal
area at each quadrat. The Bitterlich method uses an
angle gauge, a stick 1m in length with a 20mm
cross piece at one end, to assess all the trees around
a central sampling point (Mueller-Dombois and
Ellenberg 1974; Carron 1968). The operator stands at
the sampling point, sights along the stick to the cross
piece at the far end, and counts the number of trees
with diameter at breast height larger than, or equal to,
the angle indicated by the cross-piece. Counts were
tallied by assigning a full point to trees larger than
the cross piece and a half point score to trees equal
to the cross piece. With a 20 mm cross piece on a
stick of 1 m (ie a ratio of 1:50), the Bitterlich count
approximates the basal area of trees in the stand in
square metres per hectare.

Data analysis and description of Plant
Communities

A multivariate analysis of native species
composition data from the plots was carried out to
develop a classification of plant communities in the
two reserves. A data matrix was first assembled from
all available data, including the 21 plots recorded in
this survey and four plots recorded previously in and
around the reserves during the PSMA survey (Tindall
et al. 2004). All exotic species were excluded from
the data matrix so that the classification was based
on native species composition. Specimens that could
not be identified to species level were also omitted
from the analysis. Compositional dissimilarity among
samples was computed on unstandardised data using a
symmetric form of the Kulezynski coefficient (Belbin
1994). Hierarchical agglomerative clustering was
carried out using a flexible unweighted pair group
arithmetic averaging strategy with no adjacency
constraint and 8 = -0.1.

To assist interpretation of site groupings defined
in the cluster analysis, the 25 samples from Brundee
and Saltwater Swamps were added to a larger set of

Proc. Linn. Soc. N.S.W., 128, 2007

samples compiled by Tindall et al. (2004) from the
lower Shoalhaven district. This larger data set was
analysed using the same methods as those described
above. The correspondence between the site groups
for Brundee and Saltwater Swamps and existing
communities defined by Tindall et al. (2004) was
assessed by cross-referencing group membership for
the sites between the two classifications. This allowed
each new site from Brundee and Saltwater Swamps
to be assigned to one of the PSMA communities.
These interpretations were verified by assessing the
species list for each site against the list of diagnostic
species for the corresponding community described
by Tindall et al. (2004).

Map preparation

Existing vegetation boundaries on the PSMA
vegetation map (Tindall et al. 2004) were updated
by stereoscopic aerial photo interpretation (API) of
colour photography 1:15 000 scale flown in January
1996. Boundaries were further adjusted on-screen
using an orthorectified digital aerial photograph
flown in 2001. The photography was interpreted to
delineate all patches of native vegetation larger than
one hectare in size. In some cases, it was possible to
map additional patches that were smaller than one
hectare. Woody vegetation was mapped where crown
cover was = 5%. Interpretation of boundaries was
informed by location of sample sites and additional
field reconnaissance.

Vegetation and management history

We reviewed evidence of changes to the local
floodplain environment and its vegetation from
a number of different sources. These included
surveyors’s maps and reports for the Shoalhaven
floodplain from the Land Property Information
Service; historical articles; reports and journal articles
on local environmental studies; and aerial photographs
of the Brundee and Saltwater Swamps flown at
two recent dates (1996 and 2001). We interpreted
information from these sources to reconstruct likely
characteristics of historical water regimes, soils and
vegetation of the local floodplain.

RESULTS

Vegetation Classification

The 25 vegetation samples were classified into
seven groups on the basis of similarities in species
composition (Fig. 1). Each of these plant assemblages
was referrable to communities described previously
by Tindall et al. (2004). Eucalypt-dominated
communities were segregated in the dendrogram
from the remainder of the samples. Eucalypts were
generally confined to the margins of the floodplain
and the surrounding hills, while the floodplain itself
was characterised by a mosaic of forested wetlands
dominated by non-eucalypt tree genera and treeless
wetlands.

Description of Plant Communities
A local synopsis of the seven communities for

125

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

0.2610 0.4548 0.6486
| | |
pl0o7 Brun01 ( 1)
p107 Brun05 ( 5)
p07 NOWO11A ( 13)
pl0o5 Brun02 ( 2)
pl05 NOW004A ( 10) |
pl05 Salt06 ( 19) Vl
pl0os Salt01 ( 14)
pl05 Salt03 ( 16) |
pl0s5 Salt04 ( 17)
p106 Brun0s ( 3) ee
pl06é Salt08 ( Dik) ||
pl06 Brun09 ( 9)
p106 Salt07 ( 20) [pea FS
p509 Brun03 ( 3)
p509 Salt02 ( U5)
p509 Salt05 ( 18) |
ps09 NOW010A ( 12)
p85 Brun04 ( 4)
ps5 Salt10 ( ZS)
pss Salt15 ( 25)
p99 Brun07 ( 7)
p45 Brun06 ( 6)
p45 NOWOO5A ( iil)
p45 Salt11( 24) |
p45 Salt09 ( 22)
| | |
0.2610 0.4548 0.6486

0.8424 1.0362 1.2300
| | |

oneil l l
0.8424 1.0362 1.2300

Figure 1. Dendrogram showing the compositional relationships of 25 vegetation samples and the plant
community to which each sample was allocated (left-hand column). Scales at the top and bottom of the
dendrogram show Kulzcynski ultra-metric dissimilarity values.

Brundee and Saltwater Swamp Nature Reserves is
given below. Their mapped distributions are shown
in Fig. 2. The most extensive communities within the
reserves are Estuarine Fringe Forest and Floodplain
Swamp Forest, with Estuarine Creekflat Scrub and
Currambene Lowland Forest the next most abundant
communities (Table 1). The other three communities
are not common within the reserves.

Tozer et al. (2006) provide more general
descriptions for each of the seven mapped
communities across a larger region between Sydney
and the Victorian border. The codes for each map
unit follow those of Tozer et al. (2006), in which
capital letters represent abbreviations of structural
formations described by Keith (2004). These include:
FOW Forested Wetlands; SL Saline Wetlands, DSF
Dry Sclerophyll Forests, and WSF Wet Sclerophyll
Forests.

As of 31 December 2005, three Endangered
Ecological Communities (EECs) listed under the
NSW Threatened Species Conservation Act (TSC
Act 1995) occur in Brundee Swamp and Saltwater
Swamp NRs. The inferred relationships between each
of these EECs and the plant communities described
and mapped below are given in Table 2. A fourth
EEC, Freshwater Wetlands on Coastal Floodplains
of the NSW North Coast, Sydney Basin and
Southeast Corner bioregions, occurs nearby but is not
currently mapped within the reserves. No ecological
communities currently listed as Endangered under
the Commonwealth’s Environment Protection and

126

Biodiversity Conservation Act 1999 (EPBC Act 1999)
occur in the reserves.

Floodplain Swamp Forest (FOW p105)

Floodplain Swamp Forest (Fig. 3) within the
reserves is characterised by a typically dense to open
canopy dominated by Casuarina glauca with trees
or shrubs of Melaleuca ericifolia (or occasionally
M. styphelioides) in comparatively lower abundance.
The understorey generally lacks woody plants other
than sparse juvenile individuals of the canopy species.
Vines of Parsonsia straminea occur occasionally,
either as scramblers at ground level or ascending stems
of canopy trees. The groundcover comprises an open
cover of sedges, grasses and forbs, including Entolasia
marginata, Juncus kraussi subsp. australasicus,
Carex appressa and Cyperus polystachyos, with
Commelina cyanea, Lobelia anceps, Alternanthera
denticulata and Senecio hispidulus var. hispidulus,
which are dispersed within a dense cover of leaf litter
from the canopy. Patches of Phragmites australis
may occur in the understorey where the water table is
more frequently close to the surface (Fig. 4).

Floodplain Swamp Forest has been recorded
and mapped primarily around the eastern and
western margins of Brundee Swamp (from which
one site was recorded — Brun02) and a large patch
covering the central and eastern portions of Saltwater
Swamp (from which the remaining five sites were
recorded). These are raised areas of the floodplain
that are likely to experience lower levels of salinity in

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

nome ROA
_ DEC Estate

+ Sites
“77> DSF p85 Currambene Lowland Forest
> FOW p105 Floodplain Swamp Forest

FOW p106 Estuarine Fringe Forest
. FOW p106a Estuarine Fringe Forest

P24 FOW p107 Estuarine Creekflat Scrub
==— FOW p45 Coastal Sand Swamp Forest

E28 GW p3 South Coast Lowland Swamp
Woodland

YZ SL p109 Estuarine Mangrove Forest

2 SL p09 Estuarine Saltmarsh

WB WSF 199 Wawarra Gully Wet Forest
Cleared

AD 1 Kilometre
GES

Figure 2. Vegetation map of Brundee Swamp, Saltwater Swamp and surrounding areas.

Table 1. Estimated areas of plant communities within Brundee Swamp NR and Saltwater Swamp NR.

Estimated area in Estimated area in Total

Community Brundee Swamp NR Saltwater Swamp NR (ha)
(ha) (ha)

Floodplain Swamp Forest (FOW p105) 16.4 98.4 114.8
Estuarine Fringe Forest (FOW p106) 49.8 46.9 96.8
Pagoda Forest regenerating 119.0 119.0
Estuarine Creekflat Scrub (FOW p107) 329 il 41.9
Estuarine Saltmarsh (SL p509) Del 9.2 11.3
Coastal Sand Swamp Forest (FOW p45) 1.6 8.6 10.2
Currambene Lowland Forest (DSF p85) ee 41.5 42.8
Illawarra Gully Wet Forest (WSF p99) 4.4 4.4
Water 0.5 0.5

Total 227.3 214.2 441.5

their groundwater than stands of Estuarine Fringe depressions with more saline influence may support
Forest (FOW p106), which occur on lower parts (unmapped) patches of FOW p106.

of the floodplain and in local depressions. Much of

the eastern two-thirds of Saltwater Swamp NR has__ Estuarine Fringe Forest (FOW p106)

been mapped as FOW p105. Within this area, small Estuarine Fringe Forest (Fig. 5) within the

Proc. Linn. Soc. N.S.W., 128, 2007 D7

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

Table 2. Endangered Ecological Communities (Threatened Species Conservation Act 1995) found within
Brundee Swamp and Saltwater Swamp Nature Reserves and their relationship to map units defined in
this study. (The legal definitions of EECs are provided by the Final Determinations under the TSC Act.
Diagnoses as to whether any particular area of vegetation constitutes an EEC should be based on field

inspection and comparison with the relevant Final Determination.)

Endangered Ecological Communities listed
under the TSC Act 1995

Coastal saltmarsh in the NSW North Coast,
Sydney Basin and South East Corner bioregions

Swamp Oak Floodplain Forest of the NSW
North Coast, Sydney Basin and South East
Corner bioregions

Swamp Sclerophyll Forest on Coastal
Floodplains of the NSW North Coast, Sydney
Basin and South East Corner bioregions
(incorporating the formerly listed Sydney
Coastal Estuary Swamp Forest Complex in the
Sydney Basin Bioregion)

reserves is characterised by a typically dense canopy
dominated by Casuarina glauca with occasional trees
or shrubs of Melaleuca ericifolia. The understorey
generally lacks woody plants other than sparse
juvenile individuals of the canopy species, while
vines are generally absent. The groundcover typically
includes a prominent stratum of tussock sedges,
particularly Juncus kraussi subsp. australasicus, with
a groundcover of succulent forbs, Se/liera radicans
and Sarcocornia quinquefaria, and other forbs,
Commelina cyanea and Alternanthera denticulata,
which are typically dispersed amongst a dense cover
of leaf litter from the canopy. In the Brundee-Saltwater
area, Estuarine Fringe Forest has a very similar
tree canopy to Floodplain Swamp Forest, although
Melaleuca styphelioides does not occur in the former.
The principal differences are in the understorey,
with Estuarine Fringe Forest (FOW p106) generally
having lower overall diversity, no vines and a greater
abundance and diversity of succulent forbs than
Floodplain Swamp Forest (FOW p105).

Estuarine Fringe Forest has been recorded and
mapped primarily in low-lying areas in the northern
and central portions of Brundee Swamp NR (Brun08,
Brun09) and the western portion of Saltwater Swamp
NR (Salt08), although a patch also occurs within a
shallow depression in the east of this reserve (Salt07).
The lower areas of the floodplain that support
Estuarine Fringe Forest are likely to experience
higher levels of salinity in their groundwater (due to

128

Coastal Sand Swamp Forest

Corresponding Map Unit Relationship
Estuarine Saltmarsh SL p509 is included
(SL p509) within this broader EEC
Floodplain Swamp Forest FOW p105, FOW p106
(FOW p105) and FOW p107 are
+ included within the
Estuarine Fringe Forest broader EEC

(FOW p106)
+

Estuarine Creekflat Scrub

(FOW p107)

FOW p45 is included

(FOW p45) within the broader EEC

greater exposure to tidal inundation) than stands of
Floodplain Swamp Forest (FOW p105), which occur
on slightly more elevated parts of the floodplain.
Elsewhere in the Sydney-South Coast region, small
patches of Estuarine Fringe Forest are scattered along
the coast to both the north and south, “fringing the
high tide mark on the margins of tidal lakes, lagoons,
inlets and river estuaries” (Tindall et al. 2004). While
the mapped distribution in the Brundee-Saltwater
area appears atypical, it may reflect residual salinity
from former tidal flooding patterns, which have since
been modified by drainage works. It is also possible
that similar habitats to those supporting Estuarine
Fringe Forest in the Brundee-Saltwater area have
been destroyed elsewhere.

In Brundee Swamp NR, a large area of Estuarine
Fringe Forest (FOW p106) is currently an open
sedgeland with scattered individuals of Casuarina
glauca, which are usually immature (Fig. 6). This
area is mapped as p106a to distinguish the difference
in growth stage from the more typical form of the
community (mapped as p106), which has a much
denser and taller canopy of mature Casuarina
glauca. It appears that the woody component of the
community is re-establishing in this area following
the exclusion of livestock after the reserve was
declared in 2001. The difference in tree abundance is
reflected in the Stand Basal Area at Breast Height for
site Brun09 (1 m?/ha) compared to Brun08 (33 m7?/
ha). Sites within the area mapped as p106a also have

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

Figure 3. A typical stand of Floodplain Swamp Forest showing Casuarina glauca (background) with
Melaleuca ericifolia (foreground) and a sparse groundcover amongst a dense layer of compressed leaf
litter.

a higher abundance and proportion of exotic species
than sites within the area mapped as p106, reflecting
the association between weed invasion and livestock

grazing.

Estuarine Creekflat Scrub (FOW p107)

Estuarine Creekflat Scrub (Fig. 7) within the
reserves is characterised by a typically dense canopy
dominated by Melaleuca ericifolia with occasional
trees of Casuarina glauca. The understorey generally
lacks woody plants other than juvenile individuals
of the canopy species. Vines of Parsonsia straminea
occur occasionally, either as scramblers at ground
level or ascending stems of canopy trees or shrubs.
The groundcover is highly variable, with tall patches
of Phragmites australis and Gahnia clarkei, patches
of tussock sedges including Juncus spp., Carex
appressa and Cyperus lucidus, and a scattered cover
of forbs including Centella asiatica, Senecio minimus,
and Selliera radicans, interspersed with large patches
of bare ground covered by copious leaf litter from the
canopy.

Estuarine Creekflat Scrub has been recorded
and mapped primarily in a drainage depression at the
southern end of the Brundee Swamp floodplain (from
which two of the three sites were recorded), although
smaller stands are also mapped in the south-eastern
and south-western margins of Saltwater Swamp. The
third site (Brun05) was recorded in the mid-western

Proc. Linn. Soc. N.S.W., 128, 2007

portion of Brundee Swamp NR. This area has been
heavily affected by clearing, grazing and drainage
changes, and is currently mapped as a degraded area
of Estuarine Fringe Forest. However, the floristic
composition of Brun05 suggests that-parts of this
area could include some degraded stands of Estuarine
Creekflat Scrub.

Estuarine Saltmarsh (SL 509)

Estuarine Saltmarsh (Fig. 8) within the reserves
is essentially treeless, although it may have Casuarina
glauca, or rarely Melaleuca ericifolia, present as
scattered shrubs. Its most prominent feature is a
relatively dense, but variable cover of the tussock rush,
Juncus kraussi subsp. australasicus. In gaps between
dense patches of this rush, there is a more-or-less
continuous cover of succulent forbs, Se//iera radicans,
Sarcocornia quinquefaria and Lobelia anceps, and
the grass, Lachnagrostis filiformis. Patches of bare
ground are limited. Estuarine Saltmarsh (SL p509)
may be difficult to distinguish from regenerating
stands of Estuarine Fringe Forest (mapped as FOW
p106a), although saltmarsh typically occurs in local
depressions where soil conditions are more saline.
These two communities may be in a continuous
state of flux related to changes in water and salinity
regimes.

Estuarine Saltmarsh has been recorded and
mapped primarily in low-lying patches in the north-

129

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

Figure 4. A variant of Floodplain Swamp Forest at Salt03 (Saltwa-
ter Swamp NR) showing a patch of Phragmites australis beneath a
canopy of Casuarina glauca.

western section of Saltwater Swamp NR (Salt02,
Salt05). In Brundee Swamp, a restricted area of
saltmarsh has been recorded and mapped along a
brackish drainage channel in the eastern part of
the reserve (Brun03, NOWI10A). However, the
small occurrences of saltmarsh in Brundee Swamp
are uncertain, given the structural and floristic
resemblance to regenerating Estuarine Fringe Forest
(FOW p106a), which occurs adjacent to the putative
stands of saltmarsh and across large areas of Brundee
Swamp.

Elsewhere in the Sydney-South Coast region,
Estuarine Saltmarsh is scattered along the coast on
mudflats associated with estuaries in locations where
there is occasional tidal inundation. Collectively, these

130

saltmarshes encompass a diverse
group of assemblages (Adam et al.
1988). Some of these assemblages
may form fine-scale mosaics
within a single stand of saltmarsh,
possibly in response to local
variations in soil salinity and tidal
inundation regimes. The Estuarine
Saltmarsh in the Brundee-Saltwater
area 1S comparatively uniform in
composition and probably samples
the lower part of the range of
variation in soil salinity.

Coastal Sand Swamp Forest
(FOW p45)

Coastal Sand Swamp Forest
(Fig. 9) within the reserves is
characterised by a relatively
dense tree canopy dominated
by Eucalyptus robusta, with
occasional trees of E. botryoides
or E. longifolia at the margins of
stands. A diverse subcanopy is
dominated by Melaleuca ericifolia,
with M. styphelioides, M. decora,
M. lineariifolia and Casuarina
glauca occurring less frequently and
occasionally equalling the eucalypt
canopy in height. Cymbidium
suave occurs sporadically on the
tree branches. The understorey
includes varying densities of
juvenile individuals of the canopy
species, occasional clumps of the
large sedge, Gahnia clarkei, and
scattered shrubs of various species.
Occasional vines of Parsonsia
straminea, Marsdenia rostrata and
Kennedia rubicunda ascend tree
trunks, festoon shrubs or scramble
along the ground. Clumps of ferns,
Hypolepis muelleri and Pteridium
esculentum, punctuate a continuous
groundcover of grasses, including
Entolasia marginata, E. Stricta,
Imperata cylindrica, Microlaena
stipoides, Echinopogon ovatus
and Oplismenus imbecillus, and
forbs including Centella asiatica, Dichondra repens,
Opercularia diphylla, Pratia purpurascens, Senecio
hispidulus var. hispidulus and Veronica plebeia.
Other frequent groundcover species include Adiantum
aethiopicum and the twiners Cassytha pubescens and
Glycine spp. Coastal Sand Swamp Forest is the most
floristically rich and structurally complex vegetation
type within the reserves, and is therefore likely to be
important fauna habitat.

Coastal Sand Swamp Forest has been recorded
and mapped around the southern margins of both
Brundee (Brun06) and Saltwater (Salt09, Salt11)
Swamps. Small patches have also been mapped and
recorded to the west of Brundee Swamp (NOW05A),

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

Figure 5. Estuarine Fringe Forest at Salt08 (Saltwater Swamp NR) showing a tree canopy of Casuarina
glauca over scattered tussocks of Juncus kraussi subsp. australasicus, and patches of Selliera radicans
amongst copious Casuarina leaf litter.

outside the reserve. The stands in the reserves are
restricted to narrow ecotones with poorly drained,
humic sandy soils where the margins of the floodplain
receive freshwater runoff from the toeslopes of the
surrounding hills. These stands are only partly included
within the reserve boundaries. The stands associated
with Brundee Swamp are structurally simpler than
those associated with Saltwater Swamp, and this
probably reflects the influence of past management
regimes (Fig. 10). The mapped areas of Coastal Sand
Swamp Forest, outside the reserve, to the west of
Brundee Swamp are associated with shallow drainage
lines in slightly more elevated terrain.

Elsewhere in the Sydney-South Coast region,
Coastal Sand Swamp Forest occurs in small stands
associated with poorly drained swales and drainage
lines on coastal sandplains. The main areas of
occurrence are around Botany and Jervis Bays, though
little of the former remains (Tindall et al. 2004). Small
stands of the community are also associated with the
margins of Coomonderry Swamp. The stands in the
Brundee-Saltwater area are unusual in the sense that
they are associated with the margins of a floodplain
rather than a sandplain.

Proc. Linn. Soc. N.S.W., 128, 2007

Currambene Lowlands Forest (DSF p85)
Currambene Lowlands Forest (Fig. 11) within
the reserves is a relatively tall dry eucalypt forest
dominated by Corymbia maculata with Eucalyptus
globoidea and E. longifolia, occasionally with
E. paniculata, E. pilularis, E. punctata, Angophora
floribunda or Syncarpia glomulifera. Scattered trees
of Acacia irrorata and Allocasuarina littoralis make
up the subcanopy. An open shrub stratum comprises
Daviesia _ulicifolia, _ Leucopogon juniperinus,
Persoonia linearis and Pittosporum undulatum, in
addition to juveniles of the canopy species. Vines and
twiners festoon shrubs or scramble on the ground,
but rarely ascend trees. They include Billardiera
scandens, Eustrephus latifolius, Glycine clandestina,
Hardenbergia violacea, Parsonsia straminea and
Hibbertia scandens. An open groundcover is scattered
amongst a semi-continuous layer of eucalypt leaf litter
and occasional patches of bare ground. It comprises
graminoids, Entolasia — stricta, | Echinopogon
caespitosus, E. ovata, Imperata cylindrica, Lomadra
filiformis, L. longifolia, L. multiflora, Poa labillardieri
and Themeda australis and a range of forbs including
Brunoniella pumilio, Dianella caerulea, D. revoluta,
Dichondra repens, Lagenifera stipitata, Opercularia

Hi

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

aspera, Tricoryne elatior and Vernonia
cinerea.

Currambene Lowlands Forest
(DSF p85) occurs on hilly terrain with
well-drained yellow loams derived
from a mixture of siltstone, mudstone
and sandstone. The main occurrence is
in state forest to the south of Saltwater
Swamp. It also occurs on a low rise
that separates Saltwater Swamp from
Brundee Swamp, where it has been
fragmented by rural development,
although a small patch occurs within
the margin of Brundee Swamp NR.
The community occurs extensively in
the Nowra — Jervis Bay area (Tindall
et al. 2004).

Illawarra Gully Wet Forest (WSF
p99)

A moderate east-facing slope
on the western edge of the northern
section of Brundee Swamp NR
supports a eucalypt forest dominated
by Corymbia maculata with Eucalyptus
paniculata and E. globoidea (Fig. 12).
The understorey includes a number of
mesophyllous shrub species and a more
prominent groundcover of grasses
and forbs. The most abundant shrubs
include Bursaria spinosa and Notelaea
longifolia, with Clerodendrum
tomentosum and Olearia viscidulum.
Vines of Eustrephus _ latifolius,
Tylophora barabata and Pandorea
pandorana are prominent amongst
the shrubs, while the groundcover is
dominated by Oplismenus imbecillus,
Microlaena stipoides, Commelina
cyanea and Dichondra repens.

This forest has a canopy
composition similar to stands of
Currambene Lowland Forest in the
area. However, its understorey, with an
abundance of mesophyllous shrubs and
vines, anditsprominent groundcoverdominatedbysoft-
leaved grasses, distinguishes it from that community,
which typically has a sparse sclerophyllous shrub
stratum and an open groundcover of wiry graminoids
and forbs. The understorey features of the Brundee
stand more closely resemble those of two other
communities described for the south coast (Tindall
et al. 2004). The mesophyllous shrubs and vines and
soft-leaved grasses are characteristic of Illawarra
Gully Wet Forest (WSF p99), while Bursaria spinosa
and abundant grasses are characteristic of South
Coast Grassy Woodland (GW p34). Although the
forest on the western edge of Brundee Swamp NR has
characteristics of all three of these communities, its
species composition has closest overall resemblance
to Illawarra Gully Wet Forest (WSF p99), which is
mapped on several similar sheltered slopes in the
vicinity by Tindall et al. (2004).

Bz

Figure 6. Regenerating stand of Estuarine Fringe Forest at
Brun09 (Brundee Swamp NR) showing young Casuarina glauca
(background), tussocks of Juncus kraussi subsp. australasicus,
amidst dense growth of Aster subulatus, which overtops scat-
tered native forbs.

Other vegetation

Some small, permanently inundated areas
adjacent to Crookhaven River on the northern edge
of Saltwater Swamp NR support dense stands of
Phragmites australis that could not be sampled in
this survey and were too small to map. These small
patches represent an example of reedlands that may
previously have been more extensive on the lower
Shoalhaven floodplain. Along the northern boundary
of Saltwater Creek NR, there are also scattered
individuals of Avicennia marina, which occurs in
larger stands below the floodgates downstream on
Crookhaven River.

Vegetation structure and fuel characteristics
Coastal Sand Swamp Forest was the most
structurally complex vegetation in the study area,
with four vertical strata, each with a relatively high
cover of foliage (Table 3). Estuarine Saltmarsh had
the simplest structure, generally with only one or

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

Proc. Linn. Soc. N.S.W., 128, 2007

Figure 7. Estuarine
Creekflat Scrub at
Brun01 (Brundee
Swamp NR) showing
Melaleuca ericifolia
in the canopy and
as smaller shrubs
along a drainage line
beneath a gap in the
canopy.

Figure 8. Estua-
rine Saltmarsh at
site Salt05 (in Salt-
water Swamp NR)
showing tussocks
of Juncus kraussi
subsp. australasi-
cus with a mat of
Selliera radicans
and occasional
Sarcocornia quin-
quefaria.

133

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

Figure 9. Sand Swamp Sclerophyll Forest at site Saltl1 (Saltwater Swamp NR) showing a mixed tree
layer of Eucalyptus robusta (background right), Melaleuca spp. and Casuarina glauca (centre) with a
structurally complex groundcover of Hypolepis muelleri, numerous grasses and forbs.

two strata, although it consistently had the greatest
groundcover. Vegetation of the floodplain had shorter
trees than the eucalypt forests of the surrounding
hills. The basal area of trees on the floodplain and the
hills was generally similar (Table 3); however, the
floodplain tended to have greater densities of smaller
trees (Casuarina and Melaleuca spp.), while the hills
had fewer larger trees (Eucalyptus and Corymbia).
There was a clear inverse relationship between
the cover of tree canopies and groundcover where
Casuarina glauca was one of the dominant tree species
(Fig. 13). However, there was no clear relationship
where Melaleuca or Eucalyptus trees were dominant.

Litter fuels were generally greater than elevated
scrub fuels throughout the reserves (Table 4).
Floodplain Swamp Forest and Currambene Lowland
Forest generally supported the highest levels of
bushfire fuels, due to leaf litter contributions from
Casuarina glauca and Eucalyptus or Corymbia spp.,
respectively. Currambene Lowland Forest accounts
for a relatively small portion of the reserves (Table
1). The majority of vegetation on the floodplain
supports low to moderate fuel levels. For most of
the Floodplain Swamp Forest and Estuarine Fringe
Forest, litter fuels were composed primarily of
densely stacked Casuarina branchlets, which were

134

poorly aerated and therefore unlikely to support a
rapid rate of fire spread. The flammability of these
and other floodplain plant communities is also likely
to be reduced by the concentration of mineral salts
in foliage, which is likely to be higher than that in
eucalypt-dominated vegetation of the surrounding
hills.

Flora

One hundred and forty-nine native plant species
were recorded in the 24 quadrats located within the two
reserves (Appendix 1). The majority of native plant
species occurred in the eucalypt forest communities
around the margins of the floodplain, while the
floodplain vegetation was comparatively species-
poor (Table 5). Poaceae was the most prominent plant
family on both the floodplain and the surrounding
hills, with 25 species represented in total. Cyperaceae
(12 species), Asteraceae (5 species), Myrtaceae (5
species), Juncaceae (3 species) and Chenopodiaceae
(3 species) were also represented by numerous species
on the floodplain. The plant families, in addition to
Poaceae, that were represented by numerous species
on the hills included Fabaceae (12 species), Myrtaceae
(9 species), Asteraceae (4 species) and Lomandraceae

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

ND hi

plant diversity is retained in the ground layer.

(3 species).

Thirty-six exotic plant species were recorded
within the reserves (Appendix 1). The most frequently
recorded of these were Cirsium vulgare, Aster
subulatus, Chenopodium album, Phytolacca octandra
and Sonchus oleraceus. The majority of exotic species
recorded were short-lived disturbance opportunists
or introduced pasture grasses with agricultural
origins, reflecting recent land use in the reserves and
continuing land use in the surrounding area. Many of
these species are unlikely to persist in high abundance
in the absence of continuing disturbance, such as
livestock grazing; however, a number of exotic

Proc. Linn. Soc. N.S.W., 128, 2007

Figure 10. Coastal Sand Swamp Forest at site Brun06 (Brundee Swamp NR) _ the
showing Eucalyptus robusta with a sparse shrub stratum and continuous
ground layer dominated by Jmperata cylindrica and diverse herb layer. It is
likely that the structural complexity of this stand has been reduced (cf. site
Salt11, Fig. 8) as a result of past grazing and burning, although a substantial of

species are potentially
aggressive weeds
capable of further spread,
excluding native species
and reducing the diversity
of native vegetation. The
most problematic weed
species recorded include
Aster subulatus, Lantana
camara, Pennisetum
clandestinum, — Senecio
madagascariensis and
Xanthium occidentale.
Aster, Senecio and
Xanthium are relatively
abundant throughout
the treeless area in the
northern two-thirds of
Brundee Swamp NR,
while Lantana and
Pennisetum appear to be
largely restricted to the
area mapped as Illawarra

Wet Gully Forest on
the western edge of
the reserve. With the

exception of Aster, these
species are either declared
noxious weeds or listed as
a Key Threatening Process

under the Threatened
Species Conservation
Act.

Vegetation and
landscape history

The earliest
description of vegetation
on the Shoalhaven
River _ floodplain
is from the journal
account of Lieutenant

oS, B.Kent and James
; Meehan, who explored
area in February

1805. Weatherburn

(1960) reconstructed

and mapped the route

their exploration,

annotating a map with
their observations about a
number of locations. Kent
and Meehan apparently did not inspect Brundee and
Saltwater Swamps directly, but made observations
from very close-by. The following observation was
made from the southern bank of the Shoalhaven
River,

“This place is an extensive plain with no
trees on, is very low and apparently swampy,
is very thick grass intermixed with reeds.

The soil is a deep black mould.”
Weatherburn (1960) interprets the location of
this observation as close to Numbaa, opposite where
Broughton Creek joins the Shoalhaven River, about

135

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

* 4 é ce Gee

Bases

Figure 11. Currambene Lowlands Forest at site Salt15 (in Saltwater Swamp NR) show-
ing the dominant tree Corymbia maculata, with Eucalyptus globoidea, a sparse shrub

stratum and an open understorey of grasses, graminoids and forbs amongst copious
leaf litter.

*

: jy) ; ; \
j i bee es sos : \

Figure 12. Illawarra Gully Wet Forest at Brun07 (Brundee Swamp NR) dominated by
Corymbia maculata with an open shrub layer and prominent ground layer of forbs and
soft-leaved grasses, including Microlaena and Oplismenus. A clump of Lantana camara
dominates the middleground.

136 Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

3 km north of the northern edge of Brundee Swamp.
Terrara Swamp is very close to this location and
could have supported very similar vegetation to
Brundee Swamp at that time. Kent and Meehan made
similar observations about thick grass and reeds from
Broughton Creek near Jaspers Brush.

On the Crookhaven River, at a location interpreted
by Weatherburn (1960) as about 1 km downstream
from Saltwater Swamp, Kent and Meehan make the
following observation,

“This bank is low part of the brush, apparently
good soil.”
Looking further upstream, they remark,
“T suppose there is a good quantity of good
ground on the banks of this river.”
Property boundaries were surveyed in the area during
the period 1840-1885 and 1905 for the purpose of
drawing portion plans of land assigned to settlers
(Fig. 14). The surveyors frequently identified trees to
mark the ends of their survey lines and occasionally
recorded remarks about the land, vegetation and soils
on the plan. A parcel of land that separates Brundee
Swamp NR from Saltwater Creek NR (Parcel 1 on
Fig. 14) was surveyed in 1842 and was described as
follows,
“Swamp and poor forest land timbered with
swamp oak and spotted gum.”

One corer point of this parcel surveyed on
the northern boundary of Saltwater Creek NR was
marked on a tree identified as “Honeysuckle” [likely
to be Banksia integrifolia] and another on the eastern
boundary of Brundee Swamp on a tree identified
as “Apple” [likely to be Angophora floribunda or
A. subvelutina|. Immediately to the north of this
block, a land parcel surveyed in 1885 (Parcel 2 on
Fig. 14), notes the occurrence of “Oak” [Casuarina
glauca] and “Mangrove” [Avicennia marina] on a
tributary creek that emerges from the eastern edge
of Brundee Swamp. The survey plan for this portion
carries the annotation,

“Some swamp was cleared.”
Another block on the south-western boundary of
Brundee Swamp NR (Parcel 3 on Fig. 14), surveyed
in 1856 was described as,

“Rich swamp, fresh and brackish.”
None of the other portion plans in the vicinity are
annotated with descriptions. However, a number of
them identify “Oak” as a marker tree, particularly
in locations on the margins of the floodplain or
adjacent to streamlines. Many of the properties
adjoining Brundee Swamp NR that were surveyed
in 1908 identify their markers as “stakes” or “posts”,
suggesting either a lack of suitable trees or a change
in survey practice that preferred to install markers,
rather than use existing trees.

Two copies of the Numbaa Parish map obtained
from the NSW Land and Property Information Centre
(Fig. 15) show the general extent of “swamp” in the
area, but carry no other description of the vegetation.
The area immediately north of Brundee Swamp is
annotated “subject to tidal inundation,” and the small
tributary of the Crookhaven River emerging from
the eastern boundary of the present nature reserve is
marked “tidal”.

Proc. Linn. Soc. N.S.W., 128, 2007

Table 3. Structural characteristics of the vegetation. Data are means with standard errors.

Litter

Grass/forb

Max.

Height

Reed/rush

Shrub
Max. Max.

Tree

Stratum:

Cover

Cover Cover Cover Cover

Max.

Basal area

(%)

(%)

(%) Height (%) Height (%)

Height

(m?.ha"')

(m) (m) (m)

(m)

Community

92(5)
65(19)

11(4)
19(8)
5(4)

3(1.8) 29(11) 0.4(0.1)

3(3)
2(2)

6(-)
2.5(-)

39(3)
39(10)

11(0)
10(2)

20(3)
28(12)

Floodplain Swamp Forest (FOW p105)
Estuarine Fringe Forest (FOW p106)

0.2(0)
0.1(0)

19(6)

2(0)

2(0.8)

85(5)

AD) SACS =) O(-) —_ -(0.9)

14(14)

Estuarine Creekflat Scrub (FOW p107)

Estuarine Saltmarsh (SL p509)

18(3)
93(3)

60(23)

23(15) 0.5(0.1)

1(2)
1(0)
5G)

17(4) 48(14) 1.90.1) 2(1)
3(0.7)

22(2)
32(-)

22(3)
37(1)

Coastal Sand Swamp Forest (FOW p45)
Currambene Lowland Forest (DSF p85)

Illawarra Gully Wet Forest (WSF p99)

14(3)
90(-)

0.6(0.1)

0(-)

2(1)

38(3)
50(-)

10(-)

0.4(-)

0(-)

15(-)

5(-)

26(-)

137

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

100

3 80

S 60

o

8

e¢ 40

bom

°

O 20
a)

Canopy cover (%)

Figure 13. Relationship between tree canopy cover and groundcover on
the floodplain. Squares - sites with canopy dominated by Casuarina.
Triangles - sites with canopy dominated by Melaleuca.

i. "Stake ~ 2
Pale Post r

con
Stringybark 108k 'S55 Post pot

(Post :
Pag Posts Post
. “Kon A < -Oak
,

'

Post

a Fr

Oak sapling? \=2 Poe

vit

re}
Oak, sapling" oe Teatree t

(2

©) Survey Marker
|. 1 DEC Estate
omens Roel
Creeks
—— Crookhaven Ck
weer Cropkhaven R

A 1 K@ometre
EEE

Figure 14. Location of portion plan survey points showing identity of trees used as survey markers
1842 - 1905. Surveyors’ descriptions of land parcels 1, 2 and 3 are given in the Results text.

138 Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

Table 4. Estimated bushfire fuel loads. Means and standard errors in parentheses in tonnes per

hectare.

Community Ground fuel Elevated scrub fuel Total fine fuel
Floodplain Swamp Forest (FOW p105) 25 (5) 4 (2) 29 (4)
Estuarine Fringe Forest (FOW p106) 10 (3) 4 (2) 14 (4)
Estuarine Creekflat Scrub (FOW p107) 12 (7) 5 (2) 17 (5)
Estuarine Saltmarsh (SL p509) 0 (0) 5 (2) 5 (2)
Coastal Sand Swamp Forest (FOW p45) 3 (2) 6 (2) 9 (1)
Currambene Lowland Forest (DSF p85) 22 (6) 1 (0) 23 (6)
Illawarra Gully Wet Forest (WSF p99) 2 (-) 2 (-) 4(-)

Table 5. Mean (standard errors in parentheses) species richness and relative abundance of exotic spe-
cies in six plant communities of Brundee Swamp NR and Saltwater Swamp NR. (C/A — Braun-Blan-

quet Cover-Abundance).

Native species

Exotic species

Exotic species as
P Sum of C/A scores

Srimuntby richness richness P en outhe for exotic species
Floodplain Swamp Forest :

(FOW p105) 325) 0.8 (0.7) 5.7 (3.0)% 1.3 (0.8)
es Ee rE OW 7 A05) 2.0 (0.8) 21.3 (5.8)% 4.6 (2.2)
Estuarine Creekflat Scrub ‘

(FOW p107) 13.3 (4.3) By (2) 21.6 (5.5)% 6.3 (1.6)
Estuarine Saltmarsh (SL p509) 9.0 (0.7) 6.0 (0) 40.0 (1.8)% 12.3 (0.6)
Coastal Sand Swamp Forest ‘

(FOW p45) 28.8 (1.7) 3.3 (0.5) 10.2 (1.1)% 3.3 (0.5)
Currambene Lowland Forest ;

(DSF p85) 41.7 (3.9) 0.7 (0.5) 1.6 (0.1)% 0.7 (0.5)
Illawarra Gully Wet Forest A

(WSF p99) 49 (-) 15 (-) 23.4 (-)% 27 (-)

The Parish map shows the location of tracks
and streamlines, but does not show the location of
any artificial drains. One copy of the map carries
later annotations than the other. The latest date on
one copy of the map is 7" July 1903 (Fig. 15a),
while the other copy carries the date 13" November
1903. This second map (Fig. 15b) shows Brundee
Swamp divided into two or more portions, each of
which is annotated with an estimate of the area in

Proc. Linn. Soc. N.S.W., 128, 2007

acres “ex drains”. There are also annotations at the
base of this map referring to monetary assistance for
drainage work about Brundee Swamp, the formation
of a Union under the Drainage Act, and a warning to
settlers “re rate liabilities in Drainage Trust Districts”.
Portion plans from the floodplain of Broughton Creek
(a northern tributary of the lower Shoalhaven River),
approximately 8 km north of Brundee Swamp, show
drains in existence there during 1896.

139

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

_ 60x

Pea apfab he oe ae pO.
rant! line aerey i

CURR AM B

roveRothers applica fron for moneta
ance for drains age wark about Srunda>
p formation of Unien‘andertirs’ "ge Act
sted ise 26°3165 Dep vee i eee

ML 88

Settiors io t be SE te tae Sebilities in &
ie Tricts Digirets vide Ma 114204.

i
f -
i

Figure 15. Two copies of the Numbaa Parish map showing (a - upper) Brundee Swamp, which carries no
reference to drains and (b - lower) Saltwater Swamp and the southern part of Brundee Swamp, marked
with later annotations referring to drains and drainage works. Obtained from NSW Department of
Lands website (http://www.lands.nsw.gov.au/survey_mapping/parish_maps)

140 Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

A new phase of drain construction was initiated
in the area after the New South Wales Grant (Flood
Mitigation) Act (1964) was introduced (Dalmazzo
et al. 2000). The new drains were generally deeper
and more effective than the older drains. In addition,
floodgates were built on the Crookhaven River to
prevent tidal inundation of the swamps, and the
river was dredged to increase drainage and diverted
at Springbank Road via a new drain to Crookhaven
Creek. This significantly reduced the volume of water
flowing through Saltwater Swamp (Dalmazzo et al.
2000).

Recent aerial photography shows the change in
distribution of woody vegetation for an area of Brundee
Swamp NR over the period 1996-2002 (Figure 16).
Over that period, forest, woodland or scrub have
replaced treeless vegetation on some areas of the
floodplain. The most striking change has occurred
on a slightly elevated area along the eastern edge
of Brundee Swamp NR, where there has been a
substantial increase in the density of small trees,
mainly Casuarina glauca. Scattered shrubs, mainly
Melaleuca ericifolia, also appear to have become
more prominent on some lower lying parts of the
floodplain in the central part of the reserve.

Quadrat samples and field reconnaissance in
the regenerating area of Estuarine Fringe Forest in
Brundee Swamp (mapped as FOW p106a, Fig. 2)
during February 2006 also suggested that substantial
recruitment of Casuarina glauca and, to a lesser
extent Melaleuca ericifolia, had occurred in recent
years.

DISCUSSION

Vegetation Patterns

The vegetation of Brundee and Saltwater Swamps
is a mosaic of herbfields, scrubs and woodlands. The
distribution of these vegetation types is determined
to a large extent by the frequency, depth and duration
of inundation, the height of the water table and the
level of salinity in the water. Both of the reserves
include small areas of eucalypt forest on elevated
lithic substrates that adjoin the floodplain.

The influences of hydrology and salinity on
native vegetation of the floodplain may be inferred
from the distribution of plant communities within the
reserves and the surrounding landscapes. Eucalypt
forests (Currambene Lowland and Illawarra Gully
Wet Forest) are essentially excluded from floodplain
landscapes in this area. They are extensive on the
freely-draining, hilly terrain on lithic substrates that
surround the floodplains, but are restricted to small
marginal portions of the two nature reserves. Coastal
Sand Swamp Forest is restricted to a narrow zone
where the toeslopes of hills adjoin the floodplain.
Here, the soils have a somewhat sandy texture
and periodic inundation is by predominantly fresh
water, which accumulates in depressions around the
margins of the floodplain after it descends from the
surrounding hills. The mixed canopy composition
of Eucalyptus, Melaleuca and Casuarina reflects the

Proc. Linn. Soc. N.S.W., 128, 2007

transitional character of this habitat.

Estuarine Creekflat Scrub is associated with well-
developed humic soils on broad flats of the floodplain.
These habitats appear to receive greater quantities of
flowing water, which may be slightly more saline
than in habitats that support Coastal Sand Swamp
Forest. Floodplain Swamp Forest, Estuarine Fringe
Forest and Estuarine Saltmarsh form a replacement
sequence of communities with decreasing elevation
and increasing soil salinity on the open floodplain.
Subtle variations in relief of the floodplain appear to
influence the distributions of these three communities,
although their ecological relationships are likely to
have been obscured by disturbances to vegetation
and soils, and by alteration to drainage patterns.
Nevertheless, Floodplain Swamp Forest is apparently
associated with raised levees along streams and subtle
rises around the margins of the floodplain, whereas
Estuarine Saltmarsh is confined to low-lying sites
exposed to occasional (or past) tidal inundation.

Vegetation Change

There is little doubt that substantial changes have
occurred to floodplain vegetation since settlement as
a consequence of clearing, grazing and changes to
drainage and tidal flows. However, the precise causes
and mechanisms of change in the Brundee-Saltwater
area are not well understood, nor is the distribution
and make-up of native vegetation prior to agricultural
development of the floodplain. It seems likely that
these changes occurred in a series of episodes as
particular events took place, including initial clearing,
introduction of livestock, construction of drainage
channels, construction of deeper drains, installation of
tidal gates and successive changes to stocking rates.

The pre-settlement vegetation of the floodplain
apparently included extensive treeless areas with
thick grass intermixed with reeds. This vegetation
now appears to be locally extinct, as no native
grasslands were observed in or around the study area.
Based on remnants observed on other floodplains
(Keith 2004), Paspalum distichum (water couch)
was a likely dominant grass species. The community
is also likely to have included a number of other
grass genera, as well as sedges (notably Eleochaeris,
Cyperus and Schoenus) and forbs from families such
as Ranunculaceae, Apiaceae and Menyanthaceae. The
most abundant reed species are likely to have been
Phragmites australis, which persists in the vicinity
today in small patches and drainage lines inundated
by brackish water, or possibly Juncus kraussii. The
latter species is abundant in contemporary treeless
areas that are inundated with brackish water, but
less frequently than sites supporting P. australis.
Other possible ‘reeds’ include other species of
Juncus, Baumea and other cyperaceous genera. Early
accounts do not specifically mention saltmarsh,
although it seems likely to have been part of the pre-
European landscape, particularly as tides reached as
far upstream as the northern part of Brundee Swamp.

Woody vegetation was also part of the pre-
settlement floodplain landscape, although its
distribution was apparently patchy. The available
historical records and current ecological relationships

141

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

suggest that woodlands and forests were mainly
confined to the banks of streams and drainage lines
and to slightly elevated areas, including levees and
gentle slopes around margins of the floodplain.
Casuarina and Melaleuca are likely to have been the
major genera of trees, although historical descriptions
record honeysuckle (Banksia) and brush (rainforest
genera such as Acmena) along the banks of floodplain
streams.

After European settlers brought cattle onto the
floodplain, exotic pastures replaced the grasslands, and
most of the woodlands and forests. The surveys and
land grants dating from 1842 marked an intensification
of earlier grazing activities by squatters. The use of
stakes or posts, rather than marker trees, in the 1905
surveys may lend tacit support to the suggestion
that little native woody vegetation remained on the
floodplain at that time. However, this could also be
explained by a change in survey practice, and the fact
that some survey points were located in areas that
were originally treeless. While some of these survey
points are located in areas that now support forests
of Eucalyptus or Casuarina, few of the contemporary
trees are large enough to suggest pre-settlement
origin.

Co-ordinated drainage work to make the area
more suitable for agriculture probably began in the
Brundee-Saltwater area in the early twentieth century
(possibly 1903), although drains were in existence
elsewhere on the Shoalhaven a decade earlier.
Drains modified the hydrology of the floodplain by
lowering the water table and reducing the duration of
inundation. If appreciable areas of native grasslands
and reedlands persisted into the twentieth century, the
construction of drains would have accelerated their
replacement by exotic pasture.

Drainage works also had a profound effect on
soil and water chemistry, particularly after deep
drains and tidal gates were constructed in the second
half of the twentieth century. Widespread oxidation of
organic floodplain soils resulted in the release of acid
sulphates and discharge of strongly acidic water into
local streams, as was found behind the floodgates on
the Crookhaven River at Culburra Road during 1991-
92 (Lawrie 2005). The level of stored acidity appears
to be moderate in well-vegetated parts of the natures
reserves in comparison to other parts of the floodplain
(Lawrie 2005).

The reduced incursion of saline water after
installation of tidal gates on the Crookhaven River is
expected to have reduced the area of saltmarsh and
mangrove vegetation in and around Brundee and
Saltwater Swamps. Nevertheless, the persistence of
salt-tolerant plant species, including native species,
Juncus krausii, Casuarina glauca, Leptinella
longipes, Lobelia anceps, Sarcocornia quinqueflora
subsp. quinqueflora and Selliera radicans, as well as
exotics, Aster subulatus and Chenopodium album, in
both Brundee and Saltwater Swamps suggests that an
appreciable saline influence still exists in the reserves.
Recent analysis of a soil profile in Saltwater Swamp
NR confirmed this inference (Lawrie 2005). The
sample was taken from the floodplain, approximately
150m north of site Saltl5 (Fig. 2). The soil profile

142

was found to be moderately to strongly saline,
with salinity increasing with depth from c. 900 mg
chloride per kg dry soil near the soil surface to over
3000 mg.kg! below 1 m depth. Surface soil was too
saline for most pasture plants and no roots were found
below 55cm, probably due to the increased salinity at
this depth (Lawrie 2005). This salinity may be partly
residual, although some tidal incursion may occur, for
example, when large tides and floods breach the tidal
barriers.

The role of fire in vegetation dynamics on the
floodplain is uncertain. There was evidence of
charring on stems of some trees, suggesting that
parts of the reserves had burnt some time in the past
10-20 years. We observed no evidence of recent
subterranean (peat) fires during field reconnaissance.
Such fires may have a substantial impact on woody
vegetation in the wetlands. Typically, peat fires spread
slowly and are difficult to distinguish. Under most
fire conditions, however, the majority of the reserve
area is not expected to be highly flammable due to the
poorly aerated and/or low volume ground fuels and
high salts content of live and dead foliage.

The most recent episode of vegetation change
apparently involves the encroachment of woodland
and forest into treeless areas of the floodplain, through
the recruitment of Casuarina glauca and Melaleuca
ericifolia. The most likely cause of these changes
is the recent removal of livestock from parts of the
floodplain within the reserves. Evidence supporting
this interpretation includes the following:

¢ The encroachment of trees and shrubs since

1996 coincides with a synchronous reduction

and removal of livestock associated with public

acquisition of the grazing leases prior to dedication

of the reserves in January 2001;

¢ No significant drainage or tidal works have

occurred since 1996;

* The recent drought largely post-dates the

beginning of woody thickening as shown on the

aerial photographs flown in 2002 (Fig. 16); and

¢ There has been little if any expansion of woody

vegetation outside the reserves where stocking rates

have remained at similar levels.
The abundance of some weed species on the floodplain,
notably Aster subulatus, may also be a recent response
to the change in grazing regime. However, the inverse
relationship between tree canopy cover and ground
cover (Fig. 14) suggests that the abundance of both
native and exotic groundcover plant species will
decline as the densities of shrubs and small trees
increase further and cast more shade at ground level.

It is uncertain whether the expansion of forested
wetland vegetation into abandoned pastures represents
arecent return to pre-settlement vegetation. Anecdotal
information suggests that some areas of the floodplain
had been treeless prior to settlement (Weatherburn
1960), although location-specific information for
Brundee and Saltwater Swamps is extremely limited.
The nature of longer term historical changes depends
on the largely unknown effects that construction of
drainage channels had on native vegetation. It seems
likely that regrowth vegetation will differ from the
original vegetation because of lowered water tables,

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

Proc. Linn. Soc. N.S.W., 128, 2007

Figure 16. Aerial
photographs of
Brundee Swamp
flown in

(a - upper) 1996
and

(b - lower) 2002.

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

shorter periods of inundation and reduced salinity. For
example, soil drying and oxidation could have made
sites that previously supported open wetlands more
suitable for colonisation by native woody species,
which has been kept in check until recently released
from grazing. However, inferences about the nature
and causes of historical vegetation change remain
speculative in the absence of appropriate data.

Conservation of Floodplain Vegetation

Irrespective of the changes that may have taken
place in the past, Brundee and Saltwater Swamp
Nature Reserves both contain important examples of
Endangered Ecological Communities (EEC) that once
covered extensive areas on coastal floodplains of New
South Wales before these areas were developed for
agriculture. The endangered floodplain communities
make up the majority of the two reserves.

Floodplain Swamp Forest, Estuarine Fringe
Forest and Estuarine Creekflat Scrub represent three
variants of the Swamp Oak Floodplain Forest EEC.
None of these variants has appreciable representation
in conservation reserves in New South Wales. The
first two of these map units have less than 3% of their
estimated pre-settlement distribution within reserves
in southern NS W (Tozer et al. 2006), of which Brundee
and Saltwater Swamp Nature Reserves make up the
major portion. While estuaries and coastal lakes to
the south of Shoalhaven River retain many small
stands of Estuarine Creekflat Scrub, its occurrences
further north are generally small degraded remnants
of the original distribution. The reserves on the
Shoalhaven floodplain are therefore important for the
conservation of all three components of the Swamp
Oak Floodplain Forest EEC.

Coastal Sand Swamp Forest, part of the Swamp
Sclerophyll Forest on Floodplains EEC, is a naturally
restricted community on the south coast, of which
almost half has been cleared in the past. Its main
representation in conservation reserves is on the sandy
soils in Jervis Bay National Park. Although small,
the stands in Brundee and Saltwater Swamp Nature
Reserves are important because they are associated
with a major floodplain unlike those associated with
the Jervis Bay sandplain to the south. Furthermore,
they make important contributions to biodiversity,
harbouring a distinctive combination of plant species
and providing a structurally complex habitat that is
not replicated by other vegetation types in the area.
Larger areas of Coastal Sand Swamp Forest exist
on private property adjacent to, and nearby the
boundaries of both reserves. The condition of these
is unknown, but warrants investigation with a view to
fostering sympathetic management of the community
in the area.

Stands of Estuarine Saltmarsh within the reserves
represent a small part of the floristic and distributional
ranges of the highly variable Coastal Saltmarsh
EEC, which extends in small patches throughout the
coastline of New South Wales. In many parts of this
range, there are signs of mangrove transgression,
which potentially threatens the persistence and
diversity of saltmarsh vegetation (Saintilan and
Williams 1999). There is currently no evidence of

144

such changes in Brundee and Saltwater Swamp
Nature Reserves, perhaps because saltmarsh there
represents the terrestrial extreme of the variation in
the community and because minimal tidal influence is
maintained by current tidal regulation. Nevertheless,
the status of saltmarsh within the reserves is uncertain
and the possibility of encroachment by Casuarina
glauca or Melaleuca ericifolia warrants continued
monitoring.

Past and ongoing changes in native vegetation
demonstrate the sensitivity of floodplain vegetation to
environmental change. Insuch disequilibrium systems,
an important goal for contemporary management of
the reserves is conservation of a dynamic mosaic of
the endangered ecological communities that are under
threat throughout the broader region. The sensitivity
of the floodplain biota to environmental change also
demands an adaptive approach to management,
whereby actions are responsive to the direction
and magnitude of changes in the system (Burgman
and Lindenmeyer 1998). Adaptive strategies of this
kind rely on monitoring to diagnose contemporary
responses and inform decisions about future actions.

The permanently marked floristic sample
sites established in this study provide a crude
contemporary baseline for assessing changes in
composition and structure of the vegetation in the
future. However, it is likely that the sampling design
will require modification to provide answers to
specific management questions. To understand the
role of grazing in future vegetation changes on the
floodplain, for example, it would be necessary to
sample a number of sites where livestock continue
to graze, perhaps under a range of regimes (stocking
rates, frequency and duration of spelling, etc.).
This would require co-operation with neighbouring
landholders.

The sampling design could also be adapted
to examine the effects of any future changes in
drainage and tidal management regimes, through
the establishment of additional samples in suitable
control subcatchments. For example, the installation
of two-way floodgates has been suggested as a means
of increasing water quality by regular tidal flushing
(Lawrie 2005). A potential consequence of such
a change is the replacement of swamp oak forest
with saltmarsh in the low lying areas. Sustainable
conservation of biodiversity on coastal floodplains
depends on continuing evaluation of vegetation
responses to such changes in water management,
and on improved understanding of mechanisms that
influence vegetation dynamics in these landscapes.

ACKNOWLEDGEMENTS

We thank Les Mitchell and Phil Craven for initiating
this work, for their discussions on the design of the survey,
management and history of the reserve, and for their
comments on a draft manuscript. Les, Phil and Alex Deura
assisted with the field survey, which was funded by South
Coast Region of the NSW National Parks and Wildlife
Service.

Proc. Linn. Soc. N.S.W., 128, 2007

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

REFERENCES

Adam P., Wilson N.C. and Huntley, B. (1988) The
phytosociology of coastal saltmarsh vegetation in
New South Wales. Wetlands (Australia) 7: 35-85.

Belbin, L. (1994). PATN. Pattern analysis package. CSIRO,
Canberra.

Burgman, M.A. and Lindenmeyer, D.B. (1998).
“Conservation biology for the Australian environment’.
(Surrey Beatty and Sons, Chipping Norton).

Carron, L.T. (1968). “An Outline of Forest Mensuration’.
(Australian National University Press, Canberra).
Dalmazzo, P., Laing, J. and The Shoalhaven Remnant
Vegetation Committee (2000). Remnant Vegetation
Management Plan, Brundee Swamp. the City of

Shoalhaven, NSW. Unpublished report.

Freeman, C., Fenner, N., Ostle, N.J., Kang, N., Dorwick, D.
J., Reynolds, B., Lock, M.A., Hughes, S. and Hudson,
J. (2004). Export of dissolved organic carbon under
elevated carbon dioxide levels. Nature 430, 195-198.

Gellie, N.J.H. (2005) Native Vegetation of the Southern
Forests: Southeast Highlands, Australian Alps, South-
west Slopes, and SE Corner bioregions. Cunninghamia
9: 219-254.

Gorham, E. (1991). Northern peatlands: role in the carbon
cycle and probable responses to climatic warming.
Ecological Applications 1, 182-195.

Harden, G. J. (1990 - 2002). “Flora of New South Wales’.
Volumes 1-4. (University of New South Wales Press,
Sydney).

Johnston, S.G., Slavich, P.G. and Hirst, P. (2003) Alteration
of groundwater and sediment geochemistry in a
sulfidic backswamp due to Melaleuca quinquenervia
encroachment. Australian Journal of Soil Research
41: 1343-1367.

Lawrie R. (2005). Report on soil inspection and sampling,
Saltwater Swamp. NSW Department of Primary
Industries, Sydney. Unpublished report.

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 Scott, J. (2005). Native vegetation of coastal
floodplains—a diagnosis of the major plant communities
in New South Wales. Pacific Conservation Biology 11,
81-104.

Mueller-Dombois, D. and Ellenberg, H. (1974). “Aims and
Methods of Vegetation Ecology’. (J. Wiley and Sons,
London).

Poore, M.E.D. (1955). The use of phytosociological
methods in ecological investigations. I. The Braun-
Blanquet system. Journal of Ecology 43: 226-244

Pressey, R.L. (1989a). Wetlands of the lower Clarence
floodplain, northern coastal New South Wales.
Proceedings of the Linnean Society of NSW 111: 143-
155.

Pressey, R.L. (1989b). Wetlands of the lower Macleay
floodplain, northern coastal New South Wales.
Proceedings of the Linnean Society of NSW 111: 157-
168.

Proc. Linn. Soc. N.S.W., 128, 2007

Pressey R.L. and Griffth S.J. (1992). Vegetation of the
coastal lowlands of Tweed shire, northern New South
Wales, species and conservation. Proceedings of the
Linnean Society of NSW 113: 203-243.

Saintilan, N. and Williams, R. J. (1999). Mangrove
transgression into saltmarsh environments in south-
east Australia. Global Ecology and Biogeography 8:
117-124.

Tindall, D., Pennay, C, Tozer, M. Turner, K. and Keith,
D. (2004). Native vegetation map report series. No.
4. Department of Infrastruture Planning and Natural
Resources, Sydney.

Tozer, M. G., Turner, K., Simpson, C. Keith, D. A.,
Beukers, P., Mackenzie, B., Tindall, D. and Pennay,
C. (2006). Native vegetation of southeast NSW:
a revised classification and map for the coast and
eastern tablelands. Version 1.0. NSW Department of
Environment and Conservation and Department of
Natural Resources, Sydney.

Weatherburn, A. K. (1960). Exploration of the Jervis Bay,
Shoalhaven and Illawarra districts, 1792-1812. Journal
of the Royal Australian Historical Society 46: 83-97.

145

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

C2 Sale eS hwo es oe Se CS STS BSS SSN SS SaaS hehe AS = Ss EH oyiund pyjaiuounsg
Seta Se keep tat -F =DESGI5 SS Hee BPRS eevaes Ga S cg = SNIMADYIVI SNUOAT y
k= ee pyofijpsdjod piuosog

COR eae ACH See see ER OMI BS ee SS SSE o-e TER as © SES = SuapUuDds DAIIPADIFI_
Petes -e- be oa SESS S SES e= SESE ES CS es-ES Ss OSes psojid suapigy
S85 8-2 8 StS Gee See SSS 9S Bi PS Ea et S-BS-Be 2S Os ee & DIDINIYAD DauNnDg
DSOWBIDA

ea be -Se eo wa TERE eS See eS eee SAS ee aE PLE. 3 a e-6 & ‘IRA DSOWBIDA DIUOYJUDpO.Ssny
Cote enw Pap eg eer eke c SRE oe ee one fae re Ve oe HS es & ETE STICIS Vx.
erg te keane oN ea oe a OE Fe Se ae je = bee oS a q so1sods wmnipodosyjapy
Cage? CPS Se Se tG ERR GSE- Soe ee CS SES TAKES Zee ey - snuiu unipodosyjapy
-S26 =. > Sp RPaSeET BS es st See Sega eR M-e 2 aS KS E-8 = SUDBDA DpISiApy
S23 ck we Sia PP ERP BSCS SO E- £S SSS 84522 e E62 23 5-5 2 DAALII1AIS DINDAY
-SSSG See SISHSSSIER SES E- SFLSPSSSFR ORME FSS SES ppunquopf proydosuy
Ste Goes oe a ae ee (ee eee a SNIIU1BA1A UOBOdOApUP,
Seow oe ee oh ed = ee ae Poe ees er - PSU SSS rE Gg SISUBAAD SI]JDSDUP,
t2eae eee Clee cede geen ole oS Sie” Se SR SIe Le ge A ee LCE © LG AED EC EMA
Laan “Aeeco s + aces a ees o-- ~~ B= PS E- eB =) eee Sat els S1DAOYJ1] DULADNSDIO[p
fees o> a oe S E5 oy ESE pAoydouapn DUuljv1aB Py

C2 Gee ee a ee ee ee a ee en wnoido1yjap unjuvippy
bb 2- oo <ESce ==) FR ES = IISUADOUL DIDDIY

ee eS a ee ae Bo SE et SS og GE BS ey GE” pIofisuo] DIDDIy
I loo ele 2” 2 DIDAOAAL ‘ASQUS DIDAOAAL DIDDIP

oo oe ae ee Co ee eee So Nee on Sy ot AS ey Ee DIDAsaUIG DIDDIP

ST If OF 60 80 40 90 SO #0 £0 7O 10 ywiio voto ysoo vroo 60 80 140 90 SO +0 €0 ZO 10
WES HES WES WS eS WS Wes Wes Wes Wes WS WS MON MON MON MON UNG UNG Ug UNG UNG UNG UNG UNG UNG sus

sed spd ced spd gord 9014 sord 6osd cod sord 6osd sod cod 901d spd soid 901d 901d ped spd Lord sed 6osd sold zord Ayumnuu0D

‘soroods paonpoul sojouap , “(SpoujJoyl 99S) So100s souRpUNge-I9A09 Jonbur|g-uneIg ore sonteA
"SOAIOSOI SY} SPISINO SPM VSOOMON SUS “YN durems Joyemgyes pue YN duremsg sopunig 10j popiocel sotoods yueyd rejnoseA jo jst] *xipueddy

Proc. Linn. Soc. N.S.W., 128, 2007

146

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

a — Te i a Ee ee
Soe NA Bet Bale gS ge Se ees ie ie)
=) ot 8 ae ee as a ee ee ee Pe es:
G Sas IG Oo Oe ed te Reg mee ae
Siew SS basi eee ay SEAS: wee eri sk Se VE SA eh aie Gey
EHH eC eS GT PN a a Te | (iO ae a
FC Mei eele a MC asas wee ec Pa CnC Ee
Se a eee ee ee! iy (Ae aces ee et Meee) Pee
Sar en Re) Re Bech ee Re, re ae a
ss Sie en RS El ee an Oem! ee PR en eens ae
as Nt ot et oe ee (wg a =
seo AS ay ae Be Be A ee bake ter Se
= dy 25.40 80. 5 0 20 sy 3G as
$62 BoP loko Be irae By a ace ee
ae ee See ee S| ac el ee
CT ek eC” ee
ges Bye a fae ao ee ge Ee Se a Se
OS Bia. Se) SS) Bar eo Be a i ee Se ie Geo ee
Se ee = ie eS to ie fe ee eee en ee

St If OF 60 80 LO 90 SO ¥0 £0 CO 10 ywi19 Vo1o vsoo vro0 60
YES HES WES WES eS WES WES WES WES WES WES HBS MON MON MON MON UUg ug ung UMg UMg UNIg UNG UNg UNIg
sad spd ced std ord 901d sord 6osd soid soid 6osd soid zoid 901d spd sod ood 901d ped spd zoid sed eosd sord zoid

1
rei

sl
— Tt

i]
Sl

eae /

80 LO

90

49)

10

SYpizaqui sn4adXy
uoj{janp uopoudy
aapns wnipiquay

snjavifas uosodoqua)
DIoOfidouos0d DINO),
DIDINIDU DIquUic1oy

‘dds nzduo),
SISUaADUNS DZAUO) y
SisuaisDUuog 0ZA{U0)),,
pauvds DuYaUuUoy
uNnsSOJUuaUO] UNAPUaposa]D)
Saplou1ndjs ‘IeA Sapiou1ddjs syoualD
DIDISIUD SYDWUA]D

DADBINA UNIS) y
DAOYdups WNWUOWUDUUI),
unqp wnipodouay),
DavsYyJhsa WNLANDJUAD),
DINDISD DIJaJUayD

DIND]S DULAONSDD
suaasaqnd vyjAsspD
DIDIYIDAGABUG] XAADD
DSAAAUI XOADD)
DUDIPNDYIIPNDS XaIDD
SIUNIIABAG XAADD
pssadddp xai0)

pignp puanjyr0j0)
psouids piavsing

ants
Ayrumurvt09

147

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

NStN

cl

IT

Ol 60 80 LO 90 SO 0 €0 ZO
WES WES HES WES WES WES WES WeS WeS WES WES WES MON MON MON MON Unig unig unig UMg UNIg UNIg UNIg UMg uNIg
ssd spd sed spd 901d 901d soid 6osd sold sord 60sd sord zord so1d spd sord oo1d oo1d ved

I

C

€

I

I

I

10 yi1o volo vsoo wroo

€

60 80

LO

90

syd zoid sed 6osd sod zo1d

0)

140)

£0

4)

aN

N

10

‘dsqns njynjnoiund snjddjvongq
pyofisuo, sniddjpang
papiogojs snjdXjvany

sapiodajog snid«{jvanq
DID1AJS DISDIOWUT
DIDUIBADUL DISDIOJUT
4aqvIs “ADA Aaqvos snuid]q
DINID S1ADYI0I1q

SOUOS1A DIPOUIT

DIDISDY DIpDUIT

snjpao uosodou1yoq
snsojidsavo

IeA Snsojidsava uosodou1yoq
pAopfiadod D1ADISIG.

suadad DAPUOYIIgT
siunpsinbavui auyoojaysiq
DYJUDADIMA UYIDIQYIIG
DIN[OAAA “IEA DINJOAas DIJaUDIG
DYOfisuo] DjjauvIq.

Da[nsavo DIjaUuvIq

njasiiponb pixnakaq

SUDLADA UNIpoUsaq
unjjAydopydys unipouseaq
DYOfII1IN DISa1ADq

DIDAQUO]S S1]AJIDG y
sodyanjstjod snsada)
snpion] snsaday

sus
AyTUMUIWWOD

Proc. Linn. Soc. N.S.W., 128, 2007

148

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

N
'
'

E& 2
eee 7G

SI It Ot

60 80 LO 90 SO 0 £0 720
MES WES WES WES WES WeS WS eS WeS WeS WES WES MON MON MON AMON Bug uNIg UNG UMg UNG UMg UMg Ug UNG
sgd spd ced cpd go1d so1d cord 6osd cord soid 60sd soid zoid 901d spd soid 901d 901d ped spd zoid sgd 6osd sord oid

VIIO0 VOIO0 VSOO Vv00 60

LO

sisuaijoajsnp ‘dsqns lissnpsy snoune
ppiyjod paatkor

pyopunut sidajosy

Aofou “IeA vIpuiyao Dosaduy
ldajjanul sidajodAy]

DIDIIPDA S1daDYIOAAH]

DAGDIS SisaDYIOdAH ,
wunaulUDis und1deday]
SidDjnaunpad aj4joooapapy
suapuvas v1.4aqqiyy
pyofiajadua

‘dsqns oyofigjadua vy.iaqqiy
DAIDIOIA DIBAAQuapADET
snudsv4ja] Snd’vI0U0y)
DUIIDGD] aUuldAjH)

pypdydosniu auiadjH
DUIJSAPUDII aUu1IA]H)

LABPUDIOS “IeA 14apUuvjIOS WiNIUDAIH)
unsouds unisajdouojiay
unnbuidosd wunyoH
pUuv1AagalS DIUYDL

LoyADO DIU DD)

snyofijn] snydajsngq

pisngod snid{jvanq

pyojound snid«jpony

siapjniid snidjvang
pyojnaiupd

aS
Ayrunww0d

149

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

ee Gee Ze = = DN OSiBu0] evaI0; VIJOf{iBu0] vanjaj0N
eee Oe Ge te te ee Sapiodys ‘Tea sapiodys puavjos1y
sapioyaydijs ponavjay
Se de an a SS ee eh Ree ee a eS) A DYOf{MADUI] DINalDjIaW
Ww eS ee er AO ce. ae 9 pyofiaida panapvjayy
pAozap vDanalvjaN
eee at at ee ae ge a eS i Be ee Ee eS er Re DIDAJSOA DIUAPSADIN
ie ee a eS eS ee a a Ee ee eS Ce SS eS Ss SE poiuodol DAgIUOTy
DAOLfIjNU

= Bes ‘dsqns nsopjfiyjniu DAPUDULOT
DIJOfIsuo] DAPUDULOT
- - - smmsofiif ‘dsqns sus0fiif pApunuoT
eS SS este ET the Se 1 ed fs be eee PF aS sdaoup DyeqoT
pees ise ee te a eg eee ge es he ee ge Ge ie? ee SLADAUY] DADSPUIT

le aee we a a ea ae ee). ace ae? Se ee ee snuisadiun{ uosodoonaT
Be gS ky WC eee SS a eS as Oe ee Se Se Se Ze sadisuo] pjjauydaT
lee Gee eo) ee Se LS SS ee, a i) ee ee Sg ee Zee ee ajD4a]0] DuudadsopidaT
Pe te ake i eee a et ee eee en ee Es) ye See ca Ge DADUDI DUDJUDT »
ae a yee ee ees ct eho ee ai ce Ege ES Se pyopnidys péafiuasvT
ial a oS SS Se es eb Ee SS Ss = & 2 = = wm = = = SI[1DDAB DAafuasvT
Bh oe es BP ge ges Gas re ce ge ee ee ge ey Re a te SIUMAOLIIf SIISOABDUYIDT
EN CEO TEE)

oe ee oo a OR Oe ae se Se OE Ss ‘dsqns 1/21pA0]J1q S1IJSOABDUYIDT
eg ea ES SS PF oS Ee 2 2 S 2 = = = = eS 2 Z DINWAD SISOABDUYIDT
Ie Se ie Ee ae a ae Dpunaiqna pipauuay

Se ae ae ee a eae ee ee ee ee snyojisn snoune
SS ES a Sh ee I ee Re eee ee oe SIjou snoune

’
rN

'

'

’

I

'

I
=

'

I
rool

!

I
rN

1

1

I

I

'

t

I

I

I

I

1
i]
i]
1
1
i]
1
1
1
1
1
'
1
1
oO a4
1
1
1
1
i]
1
1
1
i]
1

|

t
NAN

—

t

t

1

1

1

t

I

I

1

i]

Loe)

1

1

1

i]

I

i]
oOwrn

1

1

i]

St It OF 60 80 40 90 $0 0 £0 CO 10 vito yo1o vsoo vroo 60 80 40 90 sO 0 €0 2 1
YES WES WES eS WLS Wes Wes WS Wes eS WES WLS MON MON MON MON UNG UNG UNG UNIg Ug UNG Ug Ung uNIg ous

sed spd sgd spd 901d 901d soid 60sd coid cord 6osd sold zo1d 901d spd sod so1d 901d ped spd zoid sed oosd cord oid AyunUIWO7Z

Proc. Linn. Soc. N.S.W., 128, 2007

150

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

ae

(4

cl

(4

II

(4

Ol

60 80 40 90 SO +0 £0 ZO
WES WES HES WES WES WES eS WES WeS NBS WES WS MON MON MON MON Ug ung UNIg UNIg UNIg UNIg UMg UNIg UNIg
sed cyd cad crd go1d 901d cord 6osd cord soid 6osd co1d zo1d 901d spd sod 901d 901d ped spd zoid sed 6ocd soid oid

10 vi10 volo vs00 wr00 60

80

I

LO

90

a 40)

snsopfiasod snyjuviaajd
DIDJOZIUD] OSDIUD] dx
unjojnpun wnsodsopig
unjnjodad unsodsojjig
pyofiuy ‘dsqns vyofiuy vajauig
DAPUDIIO DIDDIOIAY Ty
syuns snyjunyiayd
SI]DAJSND SaNWUsvAY I
AOUIUL SIADIDYT
payonbp Sis] DY dy
SIADAUY] DIUOOS1Ad
Suaidizap DIADIISAIg
uUNnUjsapuvjD uNnjasiuuad,
Dv vf vavjjad
unjojvj[1Ip unjodsvd,
DAUIWDAIS DISUOSADY
DUDIJISDAG DIYIKUOAD Gy
opus wunr1uDg
pun1opund palopund
snyofiusoip snuupyjozQ
supuuadad S10xXE

S1]1XO SIJDXO

Sijidaquir snuausijd¢C
snjnuap snuausyd¢C
ppidsiy v1Apjnz4adQ)
piadsp p1A0jnI4adQ)
DINPIISIA D1AD2IO

SUS
Ajromunw02d

151

Proc. Linn. Soc. N.S.W., 128, 2007

VEGETATION OF BRUNDEE AND SALTWATER SWAMPS

!
i]
i]
!
1
t
i]
t
i]
i]
re
1
!
i]
i]
i]
t
1
1
1
i]
t
1
1
1

(0661) 1Shoq nsuas g ‘ds pluDjnsiady
al Gt On eA: TA An nt bee ri cre ra AID AG AMS AMG OG BG Te oc te 6 fe Ot OT SN2IDABJO SNYIUOS,
suaasaonpjs ‘dsqns 4adsp snyouos,
OC Se Se Sth tari es Nore en ee oe ae Ve ota unyasund ununjo¢y
pene Be a a ae ee ee a oe ee = unaisdpzopnasd unuvjos
Bes Ba ee Pg gee ee a eT wunydydoutsd wnunjoy
Seg en Pe ree pe ee ee ee a es tl UNABIU UNUDIOS',
SoM eS wel So Ree ee Se ey ey ge a OT Se UNUDILIAUD WNnUDjOF
SLES OE ee aap J Pee ey td Pee Ree Meera (Cyn eae gee eae DIOfIquUoys OPIS,
ee i es ee saployjunuadd o12auag
SNUIUIU O1J2Uag
SISULADISDSDPDUL O199UAG,
SnjJassip ‘ea snynpidsiy o12auag
SUDIIPDA DABIYJI

psojfanbuinb

Peni Men RC Me Fos is gto ce Oe ey sey cee Hee iS gi) ee ees ee ‘dsqns vsopfanbuinb viusos0040¢
= = # = = =« snyofiaspd snqny
- = = = = 40109SIp SNQNY
Sept gee ee ee Oe Se ES Pe SnjDpunu snjnounuvoy
I eG = = = wunjuajnosa wnipidatd

24 = eG 1s Be at FT suaospandind o1joig
T= = eS DIJAYdOsIIM DABYIUDLOT

eee Wages Wee er ge ee, a ee ee St ee ee a DUIIA]DI DIABUAJOd
Sgie eee eee es ee SO Oe OE OE ES SO SSS Se UNAJSDUAAD UNUOSM]Od y
eee eer ogee ee ey ee ee Ree ee eer ae ene a Tee eee suapuvos wniqojopog
a 2 ee ig ries Ex eninge Bag)
COE tae a A eI I ee” BS NE ae ne we AR NE eS ESE Saj2auo1au DOd

’
'
1
1
a
'
1
N
l
!
N
'
1
!
1
'
1
'
'
'
'
'
'
'
'

!
N
’
=
'
!
1
1
-
col
’
'
1
1
'
co
1
1
1
'
'
1
1
ro
NAAN

St If Of 60 80 40 90 $0 ¥0 £0 72O 10 yi19 volo vsoo vroo 60 80 40 90 SO +0 £0 2 10
WES HES WHS WES WES WBS HES WES WES WeS eS WLS MON MON MON MON UWug ug UNG UNG UNIg UMg UNg Ung UNG SHS

sed spd sed spd 901d 9014 sord Gosd sord soid 60sd cord zo1d 901d spd cord oo1d 901d ved spd oid sad cosd sord zoid AyuMUuNW0Z)

Proc. Linn. Soc. N.S.W., 128, 2007

152

D.A. KEITH, C. SIMPSON, M.G. TOZER AND S. RODOREDA

SS SS. Ee ee ee 8 Se ES SS ES EE ee

a= Soo ef ee == = Che = peewee he Pes cS oe SE NYS D14217
ae SMe OE Se Gu 2 SSS SP EE aS Ae ES = are ajDjJUuap1990 UN1Y4JUDX ,
Siw ass £ Be ee = BY - = Be US esa 7s BS Se ee paavsapay DIOL
m Ao oe ea) SES 4 BOM eee BS ee Oe ee piaqajd va1uo01a/
Woes i S 6 ER- - EQ Besos Pereees = £ == = padauld “IeA DasUID DIUOUsAY,
OS 6 - = BR ee- | SRS She Peer ee Sts = & = 2 - = = DSIDU1 VINA)
PN sacl = BBS = EES 38 Fee a ees eS Re Me ee Sisuasuiumop pydA]
biome ows & 2S eM. .§ C2 2 Bo eb ee ee ee ce Se e- = e- ppgsvg véoydoj4],
Bus # . 2 6% 2. 2 SBR 2S eee EZ Cas - eS Sa 6 = s UNJDLAJS ULYIO] S14 J,
ae Ss 401019 QUAAOIIAT,

eS @.. = SS25 =—“2 2 S552: 65.5228 te “ac =. =F SIDAISND Dpawuay |,
Diafijnuojs

Aime ~ 2.8 S85 -S8F. Sia zo eee el> ek | Se ee ‘dsqns puafijnuojs vidivouag
S &. . 2 pees - 22, SES eS ever aas = Bee =) = AOUIM “IEA SNI1UIBAIA Snjogosody
Mise Of Le ee 6 eee S be gue eee a alee: Sawe 2 sna1pu1 snjoqo1ods,

ee ee ee

ST If Of 60 80 40 90 SO #0 £0 720 10 vwiio voto vsoo vroo 60 80 40 90 SO 0 £0 2 10
YES WES WS Wes Wes WS WS Wes Wes Nes NS WES MON MON MON MON UMg ug UNG UNG UNG UNG UNG UNIg Ung Sus

sed spd sed spd oid 901d cord 6osd sord cord 6osd sord zo1d g01d spd sord 901d 9014 ved spd Lord sed 6osd cord zoid AyUNUIOZ

a ee

153

Proc. Linn. Soc. N.S.W., 128, 2007

ee ees ee

ere

a at a a a i tae
TO 80 20 +0 co 10

‘
‘4
‘
‘
.
‘
‘
‘
‘
’
'
‘
‘
‘
'
‘
'
‘4
’
‘
‘
‘
—
'
‘

ts Es

'
‘
‘
'
‘
'
'
t
—
'
'
‘
‘
‘
‘
'
‘
‘
‘
t
'
-
'
'
'

_—
‘
~-
‘
'
’
‘
i
4
‘
‘
'
'
‘
‘
’
'
'
‘
'
'
‘
a
‘
’

‘

:

‘

:
w“
wT

'
vi

”y

’
—

.

>
rm

Ca
ft)
7
a
~ ra

~ ¥ suaaSTUMLING OITHAZ

. -_ == = = = = * [= + « on dydessne RATS
i > ~. DuiDAOD OLamet ag
r MIM LISOUD. LY ER ed,
SUAPHEN ML

pacer PaDryHi7O} IEA (ILIV

;
s
—_
“a
’
-
|
‘
—= =
i
i
erm

: : "= % +<¢
Aa ) = eS ahs we = <= crx ce 7 eh ace Ate « tes — « —— a S :

on @ ‘ g so4 — = g ° piers if")

~~

i34

The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales,
Australia. Part 6. Ginkgophyta

W.B.KeITtH HoLtmMes! AND Hetp1 M.ANDERSON?

‘46 Kurrajong Street, Dorrigo, NSW, 2453, Australia (Hon. Research Fellow, University of New

England, Armidale, NSW, 2351); 746 Kurrajong Street, Dorrigo, NSW. 2453, Australia (Hon. Palaeobotanist,

South African National Biodiversity Institute, Pretoria 0001 South Africa).

Holmes, W.B.K. and Anderson, H.M. (2007). The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales, Australia. Part 6. Ginkgophyta. Proceedings of
the Linnean Society of New South Wales 128, 155-200.

The Ginkgophyte Flora from two quarries in the Basin Creek Formation of the Middle Triassic
Nymboida Coal Measures of north eastern NSW Australia is described and illustrated. This includes the first
record from Australia of Hamshawvia, a female strobilus bearing a pair of megasporophylls. Hamshawvia
and the male strobilus Stachyopitys are regarded as the fructifications of the plants bearing Sphenobaiera
leaves. Two of the Hamshawvia specimens are placed in H. distichos sp. nov. and a third in H. sp. A. Several
specimens of Stachyopitys strobili are compared with S. matatilongus and S. lacrisporangia from the
Molteno Formation of South Africa. The Ginkgophyte leaves form c. 10% of the collected leaf fossils from
the Nymboida localities and are placed in the genera Ginkgoites (four morpho-species) and Sphenobaiera
(eight morpho-species). In the absence of preserved cuticle the morpho-species are differentiated on
characters of gross morphology. In five cases where sufficient specimens of a particular form are available
to indicate a natural range of variation they are placed in a ‘morpho-species complex’. New leaf taxa are
the morpho-species Ginkgoites nymboidensis sp. nov., Sphenobaiera paucinerva sp. nov., S. densinerva sp.

noy. and S. nymbolinea sp. nov.

Manuscript received 9 February 2006, accepted for publication 13 December 2006.

KEYWORDS: Ginkgoites, Ginkgophyta, Hamshawvia, Nymboida Coal Measures, Sphenobaiera

fructifications, Stachyopitys, Triassic Flora.

INTRODUCTION

In this sixth part of the series describing the early
Middle Triassic Nymboida Flora, leaves assigned to
the Ginkgophyte genera Ginkgoites and Sphenobaiera
and the affiliated fertile organs Stachyopitys and
Hamshawvia are illustrated and described. This
is the first description of the female Ginkgophyte
megasporophyll Hamshawvia from Australia.

Part 1 (Holmes 2000) of this series dealt with the
Bryophyta and Sphenophyta; Part 2 (Holmes 2001)
with the Filicophyta; Part 3 (Holmes 2003) with fern-
like foliage; Part 4 (Holmes and Anderson 2005a) with
the genus Dicroidium and its fertile organs Preruchus
and Umkomasia; Part 5 (Holmes and Anderson

2005b) with the genus Lepidopteris and its affiliated
fructifications Peltaspermum and Antevsia, and the
genera Kurtziana, Rochipteris and Walkomiopteris.

GINKGOPHYTA — A SHRINKING LINEAGE.

The Ginkgophytes have a long evolutionary
history ranging from the Carboniferous to the present
(Taylor and Taylor 1993; Holmes 1996). However
some early species are known only as isolated
specimens and their affinities are questionable
(Taylor and Taylor 1993; Rothwell and Holt 1997).
From the Late Palaeozoic to the early Tertiary the
Ginkgophyta had an almost global distribution with

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

a peak of diversity in the late Mesozoic and early
Tertiary (Zhou 1997; Rothwell and Holt 1997). The
last recorded occurrences in Gondwana are from the
Eocene of Argentina (Berry 1938) and the Palaeogene
of Tasmania (Hill and Carpenter 1999). It has been
suggested that the fleshy Ginkgo fruits were ingested
by herbivore dinosaurs that acted as dispersal agents
for the seeds. The extinction of the dinosaurs at the end
of the Cretaceous may have resulted in the contraction
in both range and diversity throughout the Tertiary
that brought Ginkgo to the brink of extinction (Tralau
1968). Today the last single extant representative is
Ginkgo biloba which may no longer exist in the wild
(Li 1956) but is now cultivated almost world-wide.
Previous Gondwana Triassic records of
Ginkgophyta include: from Australia — Tenison
Woods (1883), Ratte (1887, 1888), Johnston (1888),
Shirley (1887, 1898), Dun (1909), Walkom (1917,
1924, 1928), Retallack (1977), Retallack et al. (1977),
Holmes (1982, 1996); from South Africa — Seward
(1903, 1908), DuToit (1927, 1932), Baldoni (1980),
Anderson and Anderson (1983, 1985, 1989, 2003);
from South America — Frenguelli (1946), Menendez
(1951), Artabe (1985), Azcuy and Baldoni (1990),
Gnaedinger and Herbst (1999); from Malagasy —
Carpentier (1935) and from India — Pal (1984).

METHODS

The material described in this paper is based
mainly on collections made by the senior author
and his family from two Nymboida quarries over a
period of forty years and on limited material held
in the fossil collections of the Australian Museum,
Sydney, and also those in the Geology Department of
the University of New England, Armidale, as noted
in Retallack (1977) and Retallack et al. (1977). The
Nymboida material was collected mostly from blocks
fallen from the working quarry faces so the exact
horizon or source of most specimens is uncertain.

Details of the Coal Mine Quarry and Reserve
Quarry together with a summary of the geology of the
Basin Creek Formation, the Nymboida Coal Measures
and the Nymboida Sub-Basin were provided in
Holmes (2000). An earlier study by McElroy (1962)
of the Clarence-Moreton Basin included the older
geological units now regarded as formations within
the Nymboida Sub-Basin.

In the Holmes Nymboida collection, leaves
attributed to the Ginkgophyta comprise c. 10% of
the c. 2600 catalogued specimens with Ginkgoites
forming c. 2.5% and Sphenobaiera c. 7.5%. This

156

contrasts with that of the Benolong Flora in the
Napperby Formation near Dubbo, Australia,
(Holmes 1982) where Sphenobaiera leaves are the
most numerous of the preserved plant remains. In
their collections of fossil plants from the Molteno
Formation of South Africa, Anderson and Anderson
(1989) noted that leaves of Sphenobaiera were the
second most common taxon with a mean abundance
of 30% in 32 assemblages. Fertile organs attributed
to the Ginkgophyta are extremely rare in the fossil
record.

Leaves from a single plant of the extant Ginkgo
biloba may exhibit a great amount of variation which
points to the probability of a similar range of variation
in fossil Ginkgophyte leaves (Walkom 1917). Thus the
identification to species level of Ginkgophyte leaves
from large assemblages with a wide range of variation
and intergrading forms raises problems similar to those
experienced with the Dicroidium genus (Holmes and
Anderson 2005a) and the Lepidopteris and Kurtziana
genera (Holmes and Anderson 2005b). Anderson and
Anderson (1989) noted that at their Molteno localities
Mat111 and Birl11 three morphotaxa of Sphenobaiera
formed a complex and intergrading series. Harris
et al. (1974) also recognised that leaves of the
Ginkgoales in the Yorkshire Flora varied greatly in
form and that specific distinction was difficult. They
suggested that differences within a local assemblage
and between assemblages could arise from a slight
peculiarity of genetic balance between populations
of trees or by the environment. Zhou (1997) also
noted that almost all species that have been studied
in some detail exhibit a rather wide spectrum of leaf
variation. The fossil plants at Nymboida come from a
range of facies that represent different environments
so it is probable that some of the variation within the
Nymboida leaf complexes may have been influenced
by environmental conditions.

Harris (1935) believed that cuticular details were
essential for specific determination of Ginkgoites.
Anderson and Anderson (1989) noted that their
Sphenobaiera and Ginkgo (= Ginkgoites) cuticles
appeared to share a more or less equal number of
features in the epidermal structure and also with those
of the genera Lepidopteris and Dejerseya. At the
Nymboida localities all cuticle has been destroyed in
a heating event during the Cretaceous (Russel 1994).

Due to the absence of cuticle in any of our
otherwise well-preserved plant material the
identification of leaves is based on features of gross
morphology; e.g. size and form of lamina; angle of
divergence of the lamina from the base; form of the
petiole; nature of the segmentation and vein density in
distal portions of the lamina. In cases where numerous

Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

specimens of a somewhat similar form are available
this variation is encompassed in our identification of
the group as a “‘morpho-species complex’. Leaves
with a fan-shaped lamina (flabbelate) and a distinct
petiole are placed in Ginkgoites. Wedge-shaped
leaves (cuneate) contracting to a petiolate-like base
are placed in Sphenobaiera. This differentiation is
partly subjective and certain leaves could equally
well be placed in Ginkgoites or Sphenobaiera.

Based on the presence of associated or attached
fertile material in their large Molteno Formation
collections, Anderson and Anderson (2003) have
placed Ginkgoites and Sphenobaiera leaves in
separate Families and Classes. In the absence of any
leaves attached to fertile organs at Nymboida we have
not placed our morpho-taxa above generic rank.

The Ginkgophytes were widespread in the
Northern Hemisphere during the early Mesozoic
and some northern morpho-taxa are closely similar
in gross morphology to individuals in the Nymboida
flora. However, due to problems relating to time and
geographic separation we have made comparisons
only with Gondwana material.

The flabellate leaves of the Rochipteris genus
(Gnaedinger and Herbst 1998, Baronne-Nugent
et al. 2003, Holmes and Anderson 2005b) and the
leaves associated with Kannaskoppia (Anderson
and Anderson 2003) may be confused with the
ginkgophytes. They are differentiated by the usually
asymmetrical dissection of the lamina, the more
delicate venation with anastomoses and by their
distinct stomata.

The types and all illustrated material in this paper
have been allocated AMF numbers and are housed
in the Palaeontology Department of the Australian
Museum, Sydney.

SYSTEMATIC PALAEOBOTANY
Order Ginkgophyta
Genus Ginkgoites Seward 1919
Ginkgoites nymboidensis Holmes and Anderson
sp. nov.
Figures 1 A,B; 2 A,B.
Diagnosis
Leaf lamina semicircular, elegantly dissected
almost to base into six primary elongated narrow-

elliptic to spathulate segments, each segment again
deeply to shallowly incised, apices rounded.

Proc. Linn. Soc. N.S.W., 128, 2007

Description

A medium-sized Ginkgoites leaf with elegant
semicircular leaves, 75-100 mm long, 90-140
mm wide; basal angle 180°-200°; petiole slender,
3mm wide, length unknown; lamina symmetrically
dissected almost to the base into six primary segments;
each segment again incised from % to *%4 to the base;
ultimate lobes parallel-sided or elongate-spathulate,
6-12 mm in width, the inner segments longer than
the outside segments; apices rounded-obtuse or rarely
shallowly notched; veins forking in proximal portion
of primary segments then running straight and parallel
to the apical margin; vein density in the distal region
of the segments 12—14/10 mm.

Holotype
AMF129928. Australian Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, early Middle
Triassic.

Other material
AMF129926—7, AMF130170, Coal Mine Quarry;
AMF 129929, Reserve Quarry.

Discussion

This uncommon morpho-taxon bears some
features with the leaf illustrated by Tenison-Woods
(1883, pl. 4, fig. 3) as Jeanpaulia bidens. While that
specimen is somewhat similar in dissection pattern
and shape of the lobes and general proportions it is
smaller and too poorly preserved to make meaningful
comparisons. The leaves identified by Shirley (1897,
pl. 6, figs 1,2; 1898, pl.19, fig. 1, pl. 21) as Baiera
(Jeanpaulia) bidens and by Walkom (1917, pl. 3,
fig. 1) as Baiera bidens, differ from Tenison—Woods
specimen by the lamina being incised to the base into
linear segments, each of which is again incised almost
to the base to form elongated narrow elliptic segments
with acute apices. Walkom’s specimen was selected
as the type for Ginkgo denmarkensis by Anderson
and Anderson (1989). Ginkgoites waldeckensis
(Anderson and Anderson 1989), Gnaedinger and
Herbst (1999) and Ginkgoites koningensis Anderson
and Anderson (1989, 2003) are closely similar in
lamina shape to Ginkgoites nymboidensis but differ
by their much denser venation. G. matatiensis
Anderson and Anderson (1989) differs by the four
secondary segments on either side of the median
cleft. G. semirotunda Holmes (1982) is smaller and
with a slightly different dissection of the lamina.
Ginkgo palmata (Ratte) Anderson and Anderson

157

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

1989, which includes G. simmondsii Shirley (1898)
and Baiera simmondsii (Shirley) Walkom 1917,
differs from Ginkgoites nymboidensis by the greater
number of primary segments and denser venation.
G. nymboidensis is similar to some forms of G.
denmarkensis (Fig. 3B) but differs by its semi-circular
lamina and elegant partially dissected segments.

‘Ginkgoites denmarkensis Anderson and
Anderson 2003 complex’
Figures 3A,B; 4A; 5A,B; 6A; 7A; 8A.

Holotype
University of Queensland F1411, Denmark Hill,
Ipswich Coal Measures, Queensland.

Selected references

1917 Baiera bidens, Walkom, pl. 3, fig. (2).

1965 Ginkgo digitata, Hill et al. pl. T9, fig. 6.

1965 Ginkgoites simmondsii, Hill et al., pl. T
XI, fig. 3.

1965 Ginkgoites ginkgoides, Hill et al., pl. T XI,
fig.5.

1989 Ginkgo denmarkensis, Anderson and
Anderson p.227, t. fig. 2, pl. 322.

1999 ?Ginkgoites sp., Gnaedinger and Herbst,
p. 285, figs 4B, 5D-F.

2003 Ginkgoites denmarkensis, Anderson and
Anderson, p.197, t. fig. 3.

Description

Medium to large flabbelate leaves with a long
slender to strong petiole to 40 mm long, 2-4 mm
wide; lamina 100-170 mm long, 100-200 mm wide;
basal angle 140°-270°; divided into six primary
segments that range in form from those divided from
4 from apex to almost to the base to form elongated
almost linear or narrow spathulate segments that taper
distally to acute or rounded apices (Figs 3A,B, 4A),
to other forms where the primary segments become
broad-elliptical with entire, notched or shallowly
incised apices (Figs 5A,B, 6, 7, 8A). Veins fork near
the base then run straight and parallel to the apex of
the lobes; density in the distal portion 12—14/10 mm
and in rare specimens to 16—18/10 mm.

Material
AMF129930-4, 129936-—7, Coal Mine Quarry;
AMF 129935, Reserve Quarry.
Discussion
Leaves in the ‘G. denmarkensis complex’ are
common and variable. Some leaves from Nymboida
(Fig. 3A) are closely similar to the type specimen of

158

G. denmarkensis selected by Anderson and Anderson
(1989 p. 227) in which each of the six primary
lobes is dissected halfway or more to the base to
form elongated parallel-sided to slightly expanding
secondary lobes contracting distally to an acute apex.
Some forms e.g. Fig. 3B are close to G. nymboidensis
but are less elegantly symmetrical and with deeper
incisions of the secondary lobes. The very large leaf
(Fig. 4A) is close in form to the Molteno leaf G.
aviamnica that differs by the less dense venation (6/10
mm). Forms with less deep incisions and broadening
of the primary lobes (Figs 5A,B, 6A, 7A, 8A) that we
have included in this complex are close to the Molteno
G. matatiensis which differs by the dissection into
eight primary segments. The fragmentary leaves
figured by Gnaedinger and Herbst (1999, figs 4b,
5D-F) as Ginkgoites sp. from the Tranquilo Group of
Patagonia are closely similar in outline and venation
to G. denmarkensis.

Ginkgoites ginkgoides (Shirley 1898) Florin 1936.
Figures 9A,B

Type specimen
F104c Queensland Geological Survey, Denmark
Hill, Ipswich Coal Measures, Queensland.

Selected references
1898 Baiera ginkgoides Shirley, p. 13, pl. 3,
fig. 1.
1917 Baiera ginkgoides Walkom, p. 12, pl. 3,
figs 3,4.

Description

Medium-sized incomplete Ginkgoites leaves
divided into four to six widely separated primary
segments attenuated basally, expanding distally;
apices not known; petioles stout, to 6 mm wide,
length not known; veins prominent, density c. 10—
12/10 mm.

Material
AMF 1299412, Coal Mine Quarry.

Discussion

These rare specimens have close similarities
with those illustrated by Shirley (1898) and Walkom
(1917). At Nymboida they occur in allocthonous
deposits of mostly macerated plant fragments. These
specimens are probably decorticated or decaying
leaves and could possibly have been derived from G.
denmarkensis trees.

Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Ginkgoites sp. cf G. waldeckensis (Anderson
and Anderson 1989) Gnaedinger and Herbst 1999
Figures 8B—D.

Description

Small Ginkgoites leaves, lamina semicircular to
flabbelate, 25—35 mm long, 35-45 mm wide; petiole 10
mm long, 1—1.5 mm wide with an expanded leafbase:
lamina diverging from base at an angle of 130°-180°,
deeply incised to form six major segments, each
segment again less deeply incised to form 12 terminal
lobes with truncate or shallowly notched apices; each
lobe has from four to six veins giving a vein density
in apical portions of c. 16—24/10 mm.

Material
AMF 129938, 129940, Coal Mine Quarry;
AMF 129939, Reserve Quarry.

Discussion

The rare semi-circular, deeply segmented leaves
of this morpho-species differ from any previously
figured species from Australia. The few Nymboida
specimens are similar in lamina size and outline to
some forms of the variable Ginkgoites waldeckensis
from the Moltenon Formation of South Africa
(Anderson and Anderson 1989) but the long slender
petioles have not been observed. The leaves from
the El Tranquilo Group of Patagonia placed by
Gnaedinger and Herbst (1999) in G. waldeckensis
are larger in size and divided four or five times into
narrow linear segments with finer venation.

Genus Sphenobaiera Florin 1936

‘Sphenobaiera stormbergensis (Seward 1903)
Frenguelli 1948 complex’
Figures 10A; 11A,B; 12A,B; 13A

Holotype
Baiera stormbergensis Seward 1903, F11670
South African Museum

Selected references
1903 Baiera stormbergensis Seward fig.8(3).
1924 Biera bidens Walkom p1.21, fig.2.
1989 Sphenobaiera stormbergensis Anderson
and Anderson p.146, pl.91, pl.100, figs 9, 18.
1999 Sphenobaiera stormbergensis Gnaedinger
and Herbst fig.11 D,E.

Description
A medium-sized Sphenobaiera with wedge-

Proc. Linn. Soc. N.S.W., 128, 2007

shaped leaves contracted basally into a stout petiole to
6 mm wide and 50—60 mm long; angle of divergence
from 50°-80°; lamina from 110—150 mm long and to
140 mm wide; deeply incised to form 4 to 6 major
segments, each segment again shallowly incised to
form up to 12 ultimate, parallel-sided segments with
rounded apices; density of venation in distal portion
of segments c. 10/10 mm.

Material

AMF 129943 and counterpart AMF 129947,
AMF 129953, AMF121025—6, AMF130169, all Coal
Mine Quarry; AMF130168, Reserve Quarry.

Discussion

The Nymboida material assigned to this complex
is relatively common. The leaves are close to S.
stormbergensis from the Molteno Formation of South
Africa but sometimes differ by the presence of a
stout variously elongated petiolate base. S. coronata
Anderson and Anderson, also from the Molteno
Formation, has a similar density of venation and may
be stoutly petiolate but differs by a broader divergence
and more irregular incisions of the lamina. S.
stormbergensis differs from S. densinerva (below) by
the less dense venation. The woody interveinal striae
noted by Retallack et al. (1977) in a leaf fragment
referred to S. stormbergensis from Cloughers Creek
near Nymboida, have not been observed in material
from the Coal Mine and Reserve Quarries.

Sphenobaiera paucinerva Holmes and Anderson
Sp. nov.
Figures 10B, 14A—D

Diagnosis

A small Sphenobaiera leaf; lamina with deep
central incision, lateral segments again less deeply
incised to form four parallel-sided segments with
distal margins acute, rounded or shallowly cleft;
density of venation in distal portion of segments 6—
10/10 mm.

Description

Small wedge-shaped leaves; lamina 60-85 mm
long, 45-80 mm wide, contracting basally into a
stout petiole 3 mm wide, to 20 mm long. Angle of
divergence of lamina from base 60°—80°. Lamina with
deep medial incision; lateral pair of segments each
incised to a depth of c. 1/3 from distal margin to form
four equal or irregular segments with sub-parallel
margins and acute to rounded, notched or shallowly
incised tips. Venation coarse, bifurcating proximally

159

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

and then running parallel to distal margin, density in
distal portion c. 6—10/10 mm.

Holotype
AMF129954, Australian Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, early Middle
Triassic.

Other material
AMF129944, AMF129955—7, Coal Mine

Quarry.

Name derivation

paucinerva — referring to the less dense venation
than in the somewhat similar-shaped smaller forms of
leaves placed in S. densinerva below.

Discussion

S. paucinerva 1s an uncommon element in the
Nymboida collection. The leaves of S. insecta and S.
helvetica from the Molteno Formation of South Africa
(Anderson and Anderson 1989) are divided into four
segments with rounded apices. However both species
are significently larger than S. paucinervia and have
preserved and described cuticles. S. paucinervia
may be a small form of the much larger leaved S.
stormbergensis but there are no intermediate forms
in our collection.

‘Sphenobaiera densinerva Holmes and Anderson
sp. nov. complex’
Figures 10C; 15A—D; 16A,B; 17A—C; 18A—D; 19A;
20A,B.

Diagnosis

A small to large petiolate cuneate leaf divided
into four short primary truncate segments or with
longer parallel-sided or slightly expanding segments
sometimes again shallowly incised; apices truncate or
broadly obtuse; venation density in distal portion of
segments c. 16—24/10 mm.

Description

A very variable cuneate leaf ranging from small
to large; lamina 60-130 mm long, 40-100 mm wide;
diverging at 45°-80° from a well-defined slender
to stout petiole 10-30 mm long, 1.5—4 mm wide.
Lamina divided by deep incisions into four segments.
Some larger leaves have longer segments that may be
again incised to form further parallel-sided ultimate

160

lobes (Figs 17A, 18C), others with broad less incised
laminae (Fig. 20A,B). Apices of segments truncate
or broadly obtuse. Veins bifurcating only in basal
portion of the lamina and then running straight and
parallel to the apical margin; density of veins in distal
portion of the lobes from 16—24/10 mm.

Holotype
AMF 129945 and counterpart AMF 129949.

Type locality
Coal Mine Quarry, Nymboida. Basin Creek

Formation, Nymboida Coal Measures, early Middle
Triassic.

Name derivation

densinerva — referring to the more dense
venation when compared with most other Nymboida
ginkgophyte material.

Other material
AMF129959-68, 129971—2, Coal Mine Quarry;
AMF 129969, Reserve Quarry.

Discussion

Our illustrations represent what we consider
to be the range of variation of the abundant leaves
within this complex, from smallest to largest leaves.
The smaller leaves of S. densinerva (Figs 15A—D) are
distinguished from the less variable S. paucinerva
(Figs 14A—D) essentially by the denser venation. The
larger leaves of S. densinerva (Figs 18B—E) differ
from S. stormbergensis by the more parallel strap-
like segments and by the much denser venation. The
distal portion of the very large leaf in Figure 17A is
closely similar in form but very much larger than
the incomplete specimen of Baiera ipsviciensis of
Shirley (1898 pl.12, fig.2) and repeated in Walkom
(1917 pl.4, figs1,2)

‘Sphenobaiera schenckii (Feistmantel 1889) Florin
1936 complex’
Figures 21 A-E; 22A—D; 23A—D.

Lectotype
Baiera schencki Feistmantel 1889, pl. 3, fig. 2.
(see Anderson and Anderson 1989)

Selected references
1889 Baiera Schencki Feistmantel, pl. 3, fig 2.
1989 Sphenobaiera schenckii Anderson and
Anderson, p. 142, pls 57, 58, 79.
2003 Sphenobaiera schenckii Anderson and
Anderson, pp 211, 223.

Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Description

A small Sphenobaiera leaf bisected almost to the
base and each segment then symmetrically twice less
deeply dissected to form eight elongate, subparallel-
sided or slightly expanding ultimate segments. Leaf
lamina 40—100 mm long and wide, contracting basally
at a convergence angle of 45°-120° to a tapering
petiole 1-3 mm wide and to 25 mm long. Venation
forking in proximal portion of the lamina to form 4-8
veins running parallel to apex of ultimate lobes at a
density of 15—20/10 mm.

Material

AMF129958, 129973-4, 121027-30, 1210412,
121044, Coal Mine Quarry; AMF129970, 129975,
Reserve Quarry.

Discussion

The abundant Nymboida leaves that are placed
in this complex often form ‘autumnal banks’ on some
horizons at both Coal Mine and Reserve Quarries.
They are variable in size and manner of dissection and
are generally smaller than leaves of S. schenckii sensu
stricta from the Molteno of South Africa (Anderson
and Anderson 1989, 2003). The enlarged specimen
illustrated in Fig. 21E shows portions of three leaves
attached to a stem. The leaf assemblage on Fig. 22D
displays leaves showing a wide range of dissection.

Sphenobaiera leaves which form the most
commonly occurring element in the Middle Triassic
Benolong Flora of central western New South Wales
were described by Holmes (1982) as a new species
S. ugotheriensis. That taxon was synonymised with
S. schenckii by Anderson and Anderson 1989. The
Benolong leaves are generally longer than S. schenckii
and, in many cases, with much broader terminal
segments (Holmes 1982, Fig. 10C). We consider that
S. ugotheriensis should stand as a valid species.

Sphenobaiera sectina Anderson and Anderson
1989
Figures 24A; 25A,B

Holotype
Sphenobaiera sectina Anderson and Anderson
1989, Specimen BP/2/824

Selected reference
1989 Sphenobaiera sectina Anderson and
Anderson p.143, pl.64.

Description
Medium-sized Sphenobaiera leaves, 90-120 mm

Proc. Linn. Soc. N.S.W., 128, 2007

long, c. 40 mm wide, diverging at c. 30° from a short
stout petiolate base S—20 mm long, 0.3-4 mm wide.
Lamina with shallow to deep median incision to form
two long lanceolate segments with rounded apices;
veins forking in proximal third of the lamina and then
running straight and parallel to distal margin, venation
indistinct to well-defined at 8—12/10 mm across distal
portion of segments.

Material
AMF 126857, AMF125102, Coal Mine Quarry;
AMF 125101, Reserve Quarry.

Discussion

The rare Nymboida specimens are similar in
shape, size and venation to the once-divided leaves of
S. sectina from South Africa illustrated by Anderson
and Anderson 1989 P1.63 figs 1-7, P1.64 figs 3-
6,11,12. Cuticle of some of the South African material
has been preserved and described.

Sphenobaiera nymbolinea Holmes and Anderson
Sp. nov.
Figures 26A—C; 27A, 28A.

Diagnosis

Large Sphenobaiera leaf, lamina narrowly
diverging and bifurcating proximally into c. 16 long
linear segments.

Description

Leaf as preserved 145 mm long and probably to
200 mm when complete, 50 mm wide, bifurcating
four times in proximal third of leaf to form 16 linear
segments each 1—1.5 mm wide; angle of divergence
of the lamina from the stout sessile base c. 25°; veins
not visible but with possible midrib and longitudinal
striations.

Holotype
AMF 125104, Australian Museum, Sydney.

Type Locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMF125105, Coal Mine Quarry; AMF125106
and counterpart AMF125107, Reserve Quarry.

Name derivation

Contrived from type locality and the linear
ultimate leaf segments.

161

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Discussion

The only material of this taxon collected to date
is two individual leaves and a slab and its counterpart
bearing stems and leaves. S. nymbolinea is similar
in form but twice the size of a specimen from the
Ipswich Coal Measures referred to Czekanowskia
tenuifolia by Jones and deJersey (1947 Text fig.55)
and also illustrated by Hill et al. (1965 pl.T9 fig.2). In
form S. nymbolinea lies between the much smaller S.
pontifolia Anderson and Anderson (1989) with eight
segments and S. africana (Baldoni 1980) Anderson
and Anderson (1989) with 32 segments. The part
and counterpart surfaces of the slab illustrated in
Figures 26C, 27A and 28A show a mass of leaves
probably attached to the lateral axis of the main stem.
Unfortunately the crucial points of attachment are not
clear on either surface.

Sphenobaiera sp. cf. S. browniana Anderson and
Anderson 1985
Figures 29A, 30A,B

Description

Large incomplete acutely wedge-shaped
Sphenobaiera leaves; estimated length from 200 mm
to 350 mm; lamina deeply incised to form four or
more long parallel-sided segments from 15—20 mm
wide; apices not preserved; veins bifurcating near
the base then running straight and parallel towards
the apex, coarse, 5—8/10 mm across distal width of
segments.

Material
AMF125108, 125110, Coal
AMF 125109, Reserve Quarry.

Mine Quarry;

Discussion

By their large size and long strap-like segments
these uncommon Nymboida leaves are compared
with S. browniana, a species known only from three
incomplete leaf fragments from the Burgersdorp
Formation of South Africa (Anderson and Anderson
1983 p.156, 157, pl. 184; 1989 p.147). The South
African leaves are significantly larger being more
than 400 mm long, with assymetrical dissection and
unclear venation.

The surface of the slab illustrated in Figure 29A
shows an assemblage of several incomplete leaves.
The largest leaves of the Nymboida ‘S. stormbergensis
complex’ (Figure 12B, 13A) have long parallel-sided
segments but they are less than 200 mm long and have
denser venation than those of S. sp. cf. S. browniana.

162

Sphenobaiera sp. A
Figures 25C,D

Description

A very small incomplete leaf c. 20 mm long,
25 mm wide, base missing; angle of divergence c.
90°; lamina divided almost to the base into eight
straight-sided segments expanding distally. Veins are
conspicuous, bifurcating occasionally to run straight
and parallel to the segment margins; vein density near
segment apices c. 45/10 mm.

Material
AMF125103, Coal Mine Quarry.

Discussion

From its small size and very dense venation, this
leaf differs from all previously described ginkgophyte
leaves. However better preserved material is needed to
adequately describe this leaf form as a new species.

Fructifications associated with Ginkgophyte
leaves

From their large collections of Late Triassic
plants from the Molteno Formation of South Africa,
Anderson and Anderson (2003) have evidence with
varying degrees of certainty on the affiliation or
attachment of fertile organs with Ginkgophyte-like
leaves. The female strobilus Avatia and the male
strobilus Eosteria are regarded as the fertile organs
of plants bearing Ginkgoites leaves. Neither Avatia
nor Eosteria organs have, so far, been collected
at Nymboida. From the Molteno Formation both
the female (Hamshawvia) and male (Stachyopitys)
reproductive structures have been found in organic
attachment with Sphenobaiera leaves. On the basis of
their unique differences to other Gondwana Triassic
ovulate genera Anderson and Anderson (2003) have
placed the Ginkgoites and Sphenobaiera genera in
separate families and orders. While specimens of
both Hamshawvia and Stachyopitys are present in the
Nymboida collections, neither are found in organic
attachment but do occur in close association with
Sphenobaiera leaves (Figs 10D,E, 31A—C). The slab
from Nymboida bearing two Hamshawvia receptacles
was illustrated but not described in Holmes (1996).

Genus Hamshawvia Anderson and Anderson
2003

Type species

Hamshawvia baccata Anderson and
Anderson 2003 p.214

Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Hamshawvia distichos Holmes and Anderson sp.
nov.
Figures 10D,E; 31A—C

Reference
1996 ‘paired ovulate organ’ Holmes P1.1.1

Diagnosis

A large Hamshawvia strobilus with paired broad-
ovate fleshy megasporophylls bearing two rows of
embedded ovules on either side of the midrib.

Description

Peduncle stout 17 mm long, base missing, 2.5
mm wide, longitudinally striated, bifurcating into
short tapering pedicels c. 5 mm and 8 mm long, each
bearing a single terminal fleshy broad-ovate receptacle
c. 18 mm long, 15 mm wide, margins entire, apices
rounded; dorsal surface (Fig. 31A) showing a stout
median vein with c. six lateral veins on either side
departing at an acute angle, arching and dividing into
three proximally and two distally; ventral surface
(Fig. 31B,C) carbonaceous, revealing two rows of
embedded ovules on either side of midrib, the outer
row with c. eight ovules, the inner row with c. four;
ovules rounded, c. 1 mm in diameter.

Holotype
AMF129946 and counterpart AMF129950,
Australian Museum, Sydney.

Type Locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMF130176 and counterpart AMF130177 on
same slab as holotype and its counterpart.

Name derivation

distichos (Greek) two rows, referring to the
two rows of ovules on either side of the receptacle
midrib.

Discussion

Only the holotype (Fig. 31 A,C) and one other
strobilis, both on the same slab (Fig. 10A and 10E)
and their counterparts have been collected. They are
very similar and were possibly derived from the same
but unknown parent plant.

Hamshawvia distichos differs from all Molteno
Hamshawvia spp by its larger size, and by the greater

Proc. Linn. Soc. N.S.W., 128, 2007

number of ovules which are arranged in two rows
on either side of the midrib. Stiphorus, a genus of
ovulate organs from the Late Permian of Eurasia
(Gomankov and Meyen 1980; Meyen 1987) is similar
in gross morphology to Hamshawvia but differs by
the external attachment of the ovules. The strobili of
Hamshawvia distichos cannot be affiliated with any
specific leaf type as they are preserved on a bedding
plane of coarse grey shale in close association with
three distinct morpho-species of Sphenobaiera leaves
(Fig. 10).

Hamshawyia sp. A
Figures 32A,B

Description

A small Hamshawvia with an elongated slender
peduncle 18 mm long (as preserved), 1 mm wide,
bifurcating to form two pedicels (one missing); the
preserved pedicel arching and expanding into the
base of a reniform to semi-circular megasporophyll
6 mm long, 8 mm wide; margin entire; apex rounded;
surface verrucose; no venation or ovules visible.

Material
AMF125111, Coal Mine Quarry, Nymboida.

Discussion

This single small incomplete specimen is
obviously a Hamshavia but as the presence of ovules
in the receptacle cannot be determined it has not
been formally named. It is preserved in fine white
sandstone together with fragments of ferns. The shape
and surface texture of H. sp. A is somewhat similar
to H. longipedunculata from the Molteno Formation
(Anderson and Anderson 2003 p.215, pl.70, figs 1,4,5)
but lacks evidence of the embedded ovules that are
arranged radially about the midvein in that species.

Genus Stachyopitys Schenck 1867

Type species
Stachyopitys preslii Schenck from Bavaria,
Germany, Triassic.

A specimen with a strobilis of Stachyopitys
lacrisporanga and a leaf of Sphenobaiera africana
attached to a common bulbous base (Anderson
and Anderson 2003 pl.81(1-3)) from the Molteno
Formation of South Africa has confirmed the
affiliation of Stachyopitys as the male fructification of
the plant bearing Sphenobaiera leaves.

163

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Stachyopitys sp. cf. S. matatilongus Anderson and
Anderson 2003
Figures 33A—D, 34A, E

Description

The male strobilus (Fig. 33A) is c. 24 mm long,
base and apex missing, axis c. 1 mm wide, with
spirally attached pedicels 2—2.5 mm long, branching
and bearing terminal radial clusters of 5—8 elliptical
microsporangia each c. 1.5 mm long, 0.5 mm wide.

Material
AMF125112—5, Coal Mine Quarry.

Discussion

The rare specimens of S. sp. cf. S. matatilongus
from Nymboida are closely similar to the Molteno
specimens of S. matatilongus by the elongated
strobili bearing clusters of microsporangia borne on
branching pedicels, but differ by the straight elliptic
shape of the microsporangia. Shirley (1898) described
two species of Stachyopitys from the Ipswich Coal
Measures of Queensland as S. annularoides and S.
simmondsii. Anderson and Anderson (2003) regarded
S. annularoides of Shirley as a species of Pteruchus
but agreed that the dispersed microsporangia
described in Walkom (1917) as S. annularoides were
indeed Stachyopitys. The microsporangial clusters
of Walkom are similar to the Nymboida S. sp. ef. S.
matatilongus. S. simmondsii differs from S. sp. cf. S.
matatilongus by its much smaller size and from S.
sp. cf. S. lacrisporangia (below) by the shape of the
microsporangia.

Stachyopitys sp. cf. S. lacrisporangia Anderson
and Anderson 2003
Figures 34B—D

Description

A fragment of a strobilus (Fig. 34B) bearing four
microsporophylls and five detached microsporophylls
(Fig.34 C,D), each a radial cluster of c. 10 tear-shaped
microsporangia, c. 1 mm long.

Material
AMF126855-6, Coal Mine Quarry.
Discussion
This fragmentary Nymboida material has

microsporophylls with tear-shaped microsporangia
similar to, but larger than those of S. /acrisporangia
from the Molteno Formation and the complete
strobilus is not known.

164

CONCLUSION

We have endeavoured to illustrate the full range
of Nymboida Ginkgophyta to enable meaningfull
comparisons with future collections. Leaves
collected from the Basin Creek Formation of the
Nymboida Coal Measures have been placed, on the
basis of gross morphology, into the morpho-genera
Ginkgoites (with 4 spp.) and Sphenobaiera (with 8
spp). Where numerous specimens of a particular
form were available to show the range of variation
we have referred to that taxon as a ‘morpho-species
complex’ eg. ‘Ginkgoites denmarkensis complex’.
Seven of our species compare well with previously
described species from the Gondwana Triassic flora.
We have described one new species of Ginkgoites
and three new species of Sphenobaiera. Comparisons
have mainly been made with the recently reviewed
Ginkophyte floras from the El Tranquilo Flora of
Patagonia (Gnaedinger and Herbst 1999) and the
comprehensively collected and described Molteno
Formation of South Africa (Anderson and Anderson
1989, 2003). The presence of Hamshawvia distichos
and H. sp. A are the first descriptions of Ginkgophyte
ovulate structures from Australia. Two species of
Stachyopitys are compared with Molteno material.

ACKNOWLEDGMENTS

W.B.K.H. gratefully acknowledges the help of
his family over many years in collecting material from
Nymboida. A grant from the Betty Mayne Scientific
Research Fund provided financial assistance towards the
preparation of this paper. The Director and staff of the
National Herbarium, SANBI, Pretoria, South Africa are
thanked for the use of facilities and providing the Molteno
Fossil Plant Collection for examination.

REFERENCES

Anderson, J.M and Anderson, H.M. (1983). Palaeoflora of
southern Africa. Molteno Formation (Triassic) Vol.1:
Part 1, Introduction. Part 2, Dicroidium. Balkema,
Rotterdam.

Anderson, J.M and Anderson, H.M. (1985). Palaeofiora
of southern Africa. Prodomus of South African
megafloras, Devonian to Lower Cretaceous.
Balkema, Rotterdam.

Anderson, J.M and Anderson, H.M. (1989). Palaeoflora of
southern Africa. Molteno Formation (Triassic). Vol.2:
Gymnosperms (excluding Dicroidium). Balkema,
Rotterdam.

Anderson, J.M and Anderson, H.M. (2003). Heyday of the
gymnosperms: systematics and biodiversity of the
Late Triassic Molteno fructifications. Strelitzia 15,
1-398.

Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Artabe, A.E. (1985). Estudio systematico de la taphoflora
triasico de Los Menucos, Provincia Rio Negro,
Argentina. Part 2. Cycadophyta, Ginkgophyta y
Coniferophyta. Ameghiniana 22, 159-180.

Azcuy, C.L. and Baldoni, A.M. (1990). La flora Triasica
del Grupo el Tranquilo. Part 3 Ginkgoales.

5” Congresso Argentino de Paleontologia y
Biostratigraphie, Serie Correlacion Geologie 7,
109-115.

Baldoni, A.M. (1980). Baiera africana, una nueva especie
de ginkgoal del Triassico de Sud Africa. Ameghiniana
17, 156-162.

Barrone-Nugent, E.D., McLoughlan, S. and Drinnan, A.N.
(2003). New species of Rochipteris from the Upper
Triassic of Australia. Review of Palaeobotany and
Palynology 123, 273-287.

Berry, E.W. (1938). Tertiary flora from the Rio Pichileufu,
Argentina. Geological Society of America, Special
Papers Number 12.

Carpentier, A. (1935). Etudes palaéobotaniques sur le
Groupe de la Sakamena (Madagascar). Annales
Géologique Service Mines, Madagascar 5, 7-32.

Dun, W.S. (1909). Notes on fossil plants from lower
Mesozoic strata , Benolong, Dubbo District. Records
of the NSW Geological Survey 8, 311-317.

DuToit, A.L. (1927). The fossil flora of the Upper Karroo
Beds. Annals of the South African Museum 22,
289-420.

DuToit, A.L. (1932). Some fossil plant remains from the
Karroo System of South Africa. Annals of the South
African Museum 28, 369-393.

Feistmantel, O. (1889). Uberstichtliche Darstellung
der geologisch-palaeontologischen Verhaltnisse
Siid-Afrikas. Th 1: Die Karroo-Formation und die
dieselbe unterlagernden Schichten. Abhandlungen
Kiingliche Bohemische Geselschaft Wiessenschaft
Prague 7, 1-89.

Florin, R. (1936). Die Fossilen Ginkgophyten von Franz-
Joseph-Land nebst Erérterungen tiber vermeintliche
Cordaitales mesozoischen Alters. Palaeontographica
Band 81-82, 71-173.

Frenguelli, J. (1946). Contribuciones al conocimiento
de la flora del Gondwana superior en la Argentina.
33. Ginkgoales de los estratos de Potrerilles en la
Precordillera de Mendoza. Notas de Museo de La
Plata. Paleontologia 87, 11, 101-127.

Gnaedinger, S. and Herbst, R. (1998). La flora triasica
del Grupo El Tranquilo, provincia de Santa Cruz,
Patagonia. Parte V. Pteridophylla. Ameghiniana 35,
53-65.

Gnaedinger, S. and Herbst, R. (1999). La flora tridsica
del Grupo El Tranquilo, provincia de Santa Cruz,
Patagonia. Parte VI. Ginkgoales. Ameghiniana 36,
281-296.

Gomankov, A.V. and Meyen, S.V. (1986). Zatarina Flora
(composition and distribution in Late Permian of
Eurasia). Trudy Geologicheskgo Instituta Akademiya
Nauk. SSSR 401, 1-174 Gn Russian).

Harris, T.M. (1935). The fossil flora of Scoresby Sound,
East Greenland. Part 4. Ginkgoales, Coniferales,
Lycopodiales and isolated fructifications. Meddelelser
om Groenland 112, 1-176.

Harris, T.M. and Millington, W. (1974). The Yorkshire
Jurassic Flora. 4.1 Ginkgoales. pp.1-78. British
Museum of Natural History, London.

Proc. Linn. Soc. N.S.W., 128, 2007

Hill, R.H. and Carpenter, R.J. (1999). Gingko leaves
from Palaeogene sediments in Tasmania. Australian
Journal of Botany 47, 717-724.

Hill, D., Playford, G. and Woods, J.T. (1965).

Triassic Fossils of Queensland. Queensland
Palaeontographical Society, Brisbane. 1—32.

Holmes, W.B.K. (1982). The Middle Triassic flora from
Benolong, near Dubbo, central-western New South
Wales. Alcheringa 11, 165-173.

Holmes, W.B.K. (1996). Ginkgo biloba, the last of an
illustrious line: the fossil record of the Ginkgoales
with special reference to Gondwana occurrences.
International Dendrology Society Year Book 1995,
38-43.

Holmes, W.B.K. (2000). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 1. Bryophyta, Sphenophyta.
Proceedings of the Linnean Society of NSW 122,
43-68.

Holmes, W.B.K. (2001). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 2. Filicophyta. Proceedings of
the Linnean Society of NSW 123, 39-87.

Holmes, W.B.K. (2003). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 3. Fern-like foliage.
Proceedings of the Linnean Society of NSW 124,
53-108.

Holmes, W.B.K.and Anderson, H.M. (2005a). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 4.
Dicroidium. Proceedings of the Linnean Society of
NSW 126, 1-37.

Holmes, W.B.K. and Anderson, H.M. (2005b). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 5
The Genera Lepidopteris, Kurtziana, Rochipteris and
Walkomiopteris. Proceedings of the Linnean Society
of NSW 126, 39-79.

Johnston, R.M. (1888). The Geology of Tasmania.
Government Printer, Hobart. ‘

Jones, O.A. and de Jersey, N.J. (1947). The flora of the
Ipswich Coal Measures — morphology and floral
succession. Papers of the Department of Geology,
University of Queensland. New Series 3, 1-88.

Li, H.L. (1956). A horticultural and botanical history of
Ginkgo. Bulletin of the Morris Arboretum 7, 3-12.

McElroy, C.T. (1963). The geology of the Clarence
Moreton Basin. Geological Survey of NSW. Memoir
9, 1-172.

Menendez, C.A. (1951). Flora mesozoica de la Formacion
Llantenes, Provincia de Mendoza. Revista de
Museo Argentino de Ciencias Naturales ‘Bernadino
Rivadavia’ 2, 147-261.

Meyen, S.V. (1987). Fundamentals of Palaeobotany.
Chapman and Hall. London.

Pal, P.K. (1984). Triassic plant megafossils from the Tiki
Formation, South Rewa Gondwana Basin, India.
Palaeobotanist 32.3, 259-267.

Ratte, F. (1887). Note on two new plants from the
Wianamatta Shales. Proceedings of the Linnean
Society of NSW 1, 1078-1083.

Ratte, F. (1888) Additional evidence on fossil Salisburia
from Australia. Proceedings of the Linnean Society of
NSW 2, 159-162.

165

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Retallack, G.J. (1977). Reconstructing Triassic vegetation
of eastern Australia: a new approach for the
biostratigraphy of Gondwanaland. Alcheringa 1, 247-
278 and Alcheringa Fiche 1, G1—J16.

Retallack, G.J., Gould, R.E. and Runnegar, B. (1977).
Isotopic dating of a Middle Triassic megafossil
flora from near Nymboida, north-eastern N.S.W.
Proceedings of the Linnean Society of NSW 101,
77-113.

Rothwell, G.W. and Holt, B. (1997). Fossils and
phenology in the evolution of Ginkgo biloba. pp.
223-230. In T. Hore et al. (eds) ‘Ginkgo biloba—a
global treasure’. Springer, Tokyo.

Russel, N.J. (1994). A palaeothermal study of the
Clarence-Moreton Basin. Australian Geological
Survey Organisation Bulletin 241, 237-276.

Seward, A.C. (1903). Fossil Flora of the Cape Colony.
Annals of the South African Museum 4, \—122.

Seward, A.C. (1908). On a collection of fossil plants from
South Africa. Quarterly Journal of the Geological
Society 64, 83-108.

Shirley, J. (1897). On Baiera (or Jeanpaulia) bidens
Tenison- Woods. Proceedings of the Royal Society of
Queensland 12, 74-78.

Shirley, J. (1898). Additions to the fossil flora of
Queensland mainly from the Ipswich Formation.
Bulletin of the Queensland Geological Survey 7,
1-25.

Taylor, T.N. and Taylor, E.L. (1993). The Biology and
Evolution of Fossil Plants. Prentice Hall. New Jersey.

Tenison-Woods, J. (1883). On the fossil flora of the coal
deposits of Australia. Proceedings of the Linnean
Society of NSW 8, 37-180.

Tralau, H. (1968). Evolutionary trends in the genus
Ginkgo. Lethaia 1, 63-101.

Walkom, A.B. (1917). Mesozoic floras of Queensland.
Part 1. The flora of the Ipswich and Walloon Series.
(d) Ginkgoales, (e) Cycadophyta, (f) Coniferales.
Publications of the Geological Survey of Queensland
259, 1-49.

Walkom, A.B. (1924). On fossil plants from Bellevue, near
Esk. Memoirs of the Queensland Museum 8, 77-92.

Walkom, A.B. (1928). Fossil plants from the Esk district,
Queensland. Proceedings of the Linnean Society of
NSW 53, 458-468.

Zhou, Z. (1997). Mesozoic Ginkgoalean megafossils — a
systematic review. pp. 183—206. In T. Hore ef al
(eds). “Ginkgo biloba — a global treasure’. Springer,

Tokyo.

166 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 1. A,B. Ginkgoites nymboidensis Holmes and Anderson sp. nov. A. AMF 129926; B. AMF129927.
Both Coal Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 167

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 2. A,B. Ginkgoites nymboidensis Holmes and Anderson sp. nov. A. Holotype,
AMF 129928, Coal Mine Quarry; B. AMF129929, Reserve Quarry. Scale bar = 1 cm.

168 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 3. A,B. ‘Ginkgoites denmarkensis complex’. A. AMF129930; B. AMF129931. Both Coal Mine
Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 169

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 4. A. ‘Ginkgoites denmarkensis complex’. AMF 129932. Coal Mine Quarry. Scale bar = 1 cm.

170 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 5. A,B. ‘Ginkgoites denmarkensis complex’. A. AMF129933; B. AMF129934. Both
from Coal Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 171

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 6. A. ‘Ginkgoites denmarkensis complex’. AMF129935. Reserve Quarry.
Scale bar = 1 cm.

72 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 7. A. ‘Ginkgoites denmarkensis complex’. AMF 129936. Coal Mine Quarry.
Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 173

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 8. A. ‘Ginkgoites denmarkensis complex’. AMF 129937. Coal Mine Quarry; B—D. Ginkgoites sp.
cf. G. waldeckensis (Anderson and Anderson) Gnaedinger and Herbst. B. AMF129938. C. AMF129939.
D. AMF129940. B, D, Coal Mine Quarry; C, Reserve Quarry. Scale bar = 1 cm.

174 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 9. A,B. Ginkgoites ginkgoides (Shirley) Holmes and Anderson comb. nov. A.
AMF129941; B. AMF129942. Both Coal Mine Quarry. Scale bar = | cm.

Proc. Linn. Soc. N.S.W., 128, 2007 5)

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 10. Ginkophyte assemblage on one bedding plane from Coal Mine Quarry.
A. “Sphenobaiera stormbergensis complex’ AMF129943; B. Sphenobaiera paucin-
erva Holmes and Anderson sp. nov. AMF129944; C. ‘Sphenobaiera densinerva com-
plex’ Holmes and Anderson sp. nov. Holotype AMF129945; D. Hamshawvia disti-
chos Holmes and Anderson sp. nov. Holotype AMF129946. E. Hamshawvia distichos
AMF 130176. Scale bar = 1 cm.

176 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 11. A,B. ‘Sphenobaiera stormbergensis complex’. A. AMF 121025; B. AMF121026. Both Coal Mine
Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 177

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 12. A,B. “Sphenobaiera stormbergensis complex’. A. AMF130168. Reserve Quarry; B.
AMF 130169. Coal Mine Quarry. Scale bar = | cm.

178 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 13. A. ‘Sphenobaiera stormbergensis complex’. AMF 129953. Coal
Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 Wg)

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 14. A-D. Sphenobaiera paucinerva Holmes and Anderson sp. nov. A. Holotype. AMF129954;
B. AMF129955; C. AMF129956; D. AMF129957. All Coal Mine Quarry. Scale bar = 1 cm.

180 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 15. A-D. ‘Sphenobaiera densinerva Holmes and Anderson sp. nov. complex’ A. Holotype,
AMF 129945; B. AMF129959; C. AMF129960; D. AMF129961. All Coal Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 181

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 16. A,B. ‘Sphenobaiera densinerva Holmes and Anderson sp. nov. complex’. AMF 129962. B. Ar-
row showing insect damage. Coal Mine Quarry. Scale bar = 1 cm.

182 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 17. A-C. ‘Sphenobaiera densinerva Holmes and Anderson sp. nov. complex’. A. AMF 129963; B.
AMF 129964; C. AMF129965. All Coal Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 183

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 18. A—-D. ‘“Sphenobaiera densinerva Holmes and Anderson sp. nov. complex’.A. AMF 129966; B.
AMF 129967; C. AMF129968. All Coal Mine Quarry. D. AMF129969. Reserve Quarry. Scale bar = 1 cm.

184 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 19. A. ‘Sphenobaiera densinerva Holmes and Anderson sp. nov. complex’. AMF129968, venation
pattern. Coal Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 185

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 20. A,B. “Sphenobaiera densinerva Holmes and Anderson sp. nov. complex’. A. AMF129971; B.
AMF 129972. Both Coal Mine Quarry. Scale bar = 1 cm.

186 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 21. A-E. ‘Sphenobaiera schenckii (Feistmantel) Florin complex’. A. AMF129973; B.
AMF 129974; C. AMF129975; D. AMF129958; E. AMF129970. A,B,D Coal Mine Quarry; C, E Reserve
Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 187

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 22. A-D. Leaf assemblages of ‘Sphenobaiera schenckii (Feistmantel) Florin complex’.
A. AMF 121027; B. AMF121028; C. AMF121029; D. AMF121030. All Coal Mine Quarry.
Scale bar = 1 cm.

188 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 23. A-D. ‘Sphenobaiera schenckii (Feistmantel) Florin complex’. A. AMF121041; B. AMF121042;
C. AMF129975; D. AMF121044. A, B, D Coal Mine Quarry; C Reserve Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 189

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 24. A. Sphenobaiera sectina Anderson and Anderson. AMF 126857, leaf
assemblage. Coal Mine Quarry. Scale bar = 1 cm.

190 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 25. A,B. Sphenobaiera sectina Anderson and Anderson. A. AMF125101, Reserve Quarry.

B. AMF125102, Coal Mine Quarry. C, D. Sphenobaiera sp. A. AMF 125103, Coal Mine Quarry.
Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 191

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 26. A-C. Sphenobaiera nymbolinea Holmes and Anderson sp. nov. A. Holotype, AMF125104;
B. AMF125105, both Coal Mine Quarry. C. AMF125106, Reserve Quarry. Scale bar = 1 cm.

192 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 27. A. Sphenobaiera nymbolinea Holmes and Anderson sp. nov. Leaf assemblage. AMF125107.
Reserve Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 193

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 28. A. Sphenobaiera nymbolinea Holmes and Anderson sp. nov. Line drawing of stem
and leaf assemblage based on AMF 125106 and counterpart AMF125107. Scale bar = 1 cm.

194 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 29. A. Sphenobaiera sp. cf. S. browniana Anderson and Anderson. AMF 125108, Coal Mine
Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 195

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 30. A,B. Sphenobaiera sp. cf. S. browniana Anderson and Anderson. A. AMF125109, Reserve
Quarry. B. AMF125110, Coal Mine Quarry. Scale bar = 1 cm.

196 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 31. A-C. Hamshawvia distichos Holmes and Anderson sp. nov. A. Holotype, AMF 129946;
B,C. AMF129950, counterpart of holotype, Coal Mine Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 128, 2007 197

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 32.4,B. Hamshawvia sp. A. AMF125111, Coal Mine Quarry.
Scale bar = 1 cm.

198 Proc. Linn. Soc. N.S.W., 128, 2007

W.B.K. HOLMES AND H.M. ANDERSON

Figure 33. A-E. Stachyopitys sp. cf. S. matatilongus Anderson and Anderson.
A-C. AMF125112; D,E. AMF125113. Both Coal Mine Quarry. Scale bar = 1 cm.

\

Proc. Linn. Soc. N.S.W., 128, 2007 199

TRIASSIC FLORA FROM NYMBOIDA - GINKGOPHYTA

Figure 34. A,E. Stachyopitys sp. cf. S. matatilongus Anderson and Anderson. A. AMF125114;
E. AMF 125115; B—D. Stachyopitys sp. cf. Stachyopitys lacrisporangia Anderson and Anderson.
B. AMF126855; C, D. AMF126856. All Coal Mine Quarry. Scale bar = 1 cm.

200 Proc. Linn. Soc. N.S.W., 128, 2007

Revision of Microplasma parallelum Etheridge, 1899 (Cnidaria:
Rugosa) from the Middle Devonian Moore Creek Limestone of

New South Wales

YoncG YI ZHEN

Palaeontology Section, The Australian Museum, 6 College Street, Sydney NSW 2010, Australia
(yongyi.zhen@austmus.gov.au)

Zhen, Y.Y. (2007). Revision of Microplasma parallelum Etheridge, 1899 (Cnidaria: Rugosa) from the
Middle Devonian Moore Creek Limestone of New South Wales. Proceedings of the Linnean Society of
New South Wales 128, 201-208.

The holotype and sole known specimen of the rugosan coral Microplasma parallelum Etheridge, 1899 is
reassessed. This phaceloid species with only sporadic occurrence of isolated dissepiments or presepiments
is here selected as type species of the new subgenus Loyolophyllum (Fasciloyolophyllum), which is erected
to accommodate phaceloid species otherwise resembling Loyolophyllum (Loyolophyllum). Two other
species previously referred to Fasciphyllum, from the Devonian of China, are also ascribed to this new
subgenus. Review of the concept of Loyolophyllum sensu stricto leads to a reappraisal of those species

assigned to it.

Manuscript received 25 August 2006, accepted for publication 13 December 2006.

KEYWORDS: Devonian, Loyolophyllum, Moore Creek Limestone, Rugose corals.

INTRODUCTION

Microplasma parallelum is a poorly understood
tugose coral of Middle Devonian age, known
only from the holotype collected from the Moore
Creek Limestone (late Eifelian to early Givetian),
near Tamworth in the New England Fold Belt of
northeastern New South Wales (Fig. 1). The first (and
only) description of this specimen was made more
than one hundred years ago by Etheridge (1899).
Fletcher (1971), Hill (1978), and Pickett (2002) all
maintained Etheridge’s assignment of the species
to Microplasma, although Pedder (1967) listed it as
“Microplasma” parallelum”. Redescription of the
type material, including the original partially silicified
specimen and two thin sections illustrated by Etheridge
(1899), is here supplemented by eleven additional
sections which reveal new morphological details
supporting relocation of Microplasma parallelum to a
new subgenus Loyolophyllum (Fasciloyolophyllum).
Two other species referable to this subgenus are
recognised from the Early and Middle Devonian of
China. Emendation of the concept of Loyolophyllum
sensu stricto leads to a reappraisal of those species
attributed to it globally; four (possibly six) species

are included, but a further four previously assigned to
Loyolophyllum can now be placed in other genera.

TYPE AREA AND BIOSTRATIGRAPHY

Some 19 species of rugose and tabulate corals
described by Etheridge (1899) from several limestone
localities to the north of Tamworth (Fig. 1) were
mostly collected by Prof. T.W. Edgeworth David and
the New South Wales Government Geologist, Mr. E.F.
Pittman. The holotype of Microplasma parallelum was
collected by David from the Moore Creek Limestone
exposed immediately south of Moore Creek (Fig. 1;
see Etheridge 1899, p. 161). This area has been taken
as the type locality of the Moore Creek Limestone
Member of the Yarrimie Formation (Brown 1942,
Crook 1961, Mawson et al. 1997, Briihl and Pohler
1997). From this locality two other rugose coral
species, Disphyllum robustum (Etheridge, 1899)
and Australophyllum giganteum (Etheridge, 1899),
four species of tabulate corals, including Favosites
goldfussi d’Orbigny, 1850, Thamnopora crummeri
(Etheridge, 1899), Syringopora auloporoides
Etheridge, 1899, and Remesia porteri (Etheridge,

REVISION OF A DEVONIAN CORAL SPECIES

Tamworth
New South -
Wales

Canberra .

Fault
Road

Crown
Reserve _

co
oS
bei)
oO

Vegetation
and soil cover

Baldwin Fm.
(Late Devonian)

Yarrimie Formation
(Mid Devonian)

Moore Creek

10 KM
home!

&
Limestone Member ee

2
%s,.

Figure 1. Locality map showing outcrops of Moore Creek Limestone in the vicinity of Tamworth (type
Moore Creek occurrences are located about 18 km north of Tamworth city centre), northeast New
South Wales, and likely type locality of Loyolophyllum (Fasciloyolophyllum) parallelum (Etheridge,
1899) (modified after Benson 1915, Brith] and Pohler 1999).

1899), and a chaetetid species, Litophyllum konincki
Etheridge and Foord, 1884, were also recorded by
Etheridge (1899), who inferred the coral fauna to be
of Early Palaeozoic age.

Hill (1942) reported a much more diverse
rugose coral fauna from the Moore Creek Limestone
exposed to the south of Moore Creek, and also
included faunas from the “Woolomol Limestone”
to the west and the “Spring Creek Limestone”
further south (Fig. 1). Nine rugose coral species,
including Australophyllum giganteum (Etheridge,
1899), Carlinastraea halysitoides (Etheridge, 1918),
Campophyllum? sp. cf. lindstromi (Frech, 1886),
Disphyllum robustum (Etheridge, 1899), Mesophyllum
cornubovis (Etheridge, 1899), Phacellophyllum
porteri (Etheridge, 1890), Pseudomicroplasma
australe (Etheridge, 1892), Sanidophyllum davidis
Etheridge, 1899, and S. colligatum (Etheridge, 1920)
were recorded from the Moore Creek area, mainly

202

based on specimens collected by Ida Brown (Hill
1942). Hill suggested a Givetian age for the Moore
Creek fauna and correlated it with a fauna from the
Burdekin Formation of north Queensland.

In a recent revision of tabulate corals from
the Moore Creek Limestone, Briihl and Pohler (1999)
recorded seven species including Heliolites porosus
(Goldfuss, 1826), Thamnopora crummeri (Etheridge,
1899), Cladopora sp., Alveolites subordicularis
Lamarck, 1801, A. sp. aff. hemisphericus (Chernyshev,
1937), Syringopora auloporoides de Koninck, 1876
and Remesia porteri (Etheridge, 1899). Briihl and
Pohler (1999) indicated a Middle Devonian age for
the succession and demonstrated connections with
coeval faunas in Eurasia.

Conodonts from the Moore Creek Limestone
at Moore Creek suggested a late Eifelian (kockelianus-
ensensis zones) age (Philip 1966, Mawson and Talent
1994, Mawson et al. 1997). However, samples from

Proc. Linn. Soc. N.S.W., 128, 2007

Y.Y. ZHEN

nearby localities indicated that the top of the Moore
Creek Limestone might extend into hemiansatus
Zone and possibly early varcus Subzone of the early
Givetian (Mawson and Talent 1997).

SYSTEMATIC PALAEONTOLOGY

Phylum COELENTERATA Frey and Leuckart, 1847
Subphylum CNIDARIA Hatschek, 1888
Class ANTHOZA Ehrenberg, 1834
Family STAURIIDAE Milne-Edwards and Haime,
1850
Genus Loyolophyllum Chapman, 1914

Synonym

Columnaria (Loyolophyllum) Chapman, 1914, p.
306.

Loyolophyllum Chapman; Hill, 1939, pp. 239-242.

Loyolophyllum Chapman; Hill, 1981, p. F135.

Type species

Columnaria (Loyolophyllum) cresswelli Chapman,
1914, pp. 306-8, pl. li, figs 15—16, pl. lu, figs
17-18; Early Devonian (late Lochkovian-
Pragian), Loyola Limestone, Griffith’s Quarry,
Loyola, Victoria.

Diagnosis

Cerioid or phaceloid corallum; corallites small
with narrow peripheral stereozone; septa few, thin,
major septa unequal, some extending almost to axis;
minor septa short; tabulae complete, commonly
sagging, or horizontal; a few scattered dissepiments
or presepiments adhering to wall by both upper and
lower edges, or in some species an incomplete row of
dissepiments may be developed (modified after Hill
LOSE peels):

Remarks

The generic concept of Loyolophyllum is
amended herein to restrict L. (Loyolophyllum) to
cerioid forms and to establish a new subgenus,
L. (Fasciloyolophyllum) for phaceloid forms.
Loyolophyllum (Fasciloyolophyllum) differs from
Fasciphyllum Schiiter, 1885 by having only isolated,
rare occurrences of dissepiments or presepiments,
which never form a continuous series of dissepiments
as in Fasciphyllum. The type species of Battersbyia
Milne-Edwards and Maime, 1851 ispoorly understood.
So Battersbyia is better treated as a nomen dubium
and tentatively synonymised with Fasciphyllum (see
Hill 1981).

Proc. Linn. Soc. N.S.W., 128, 2007

The following four species are definitely assigned
to L. (Loyolophyllum):

Loyolophyllum cresswelli Chapman, 1914, p. 306;
from the Loyola Limestone, Early Devonian,
Victoria, Australia (cerioid form, corallites
1—1.13 mm [up to 2mm] in diameter, scattered
presepiments).

Loyolophyllum cerioides Soshkina, 1949, p. 109;
from Middle Devonian, Urals, Russia (cerioid
form, corallites 4-7 mm in diameter, short septa
in two orders 18—28x2, one discontinuous row of
dissepiments occurring in longitudinal section,
tabulae 7—11/10 mm).

Loyolophyllum urense Zhmaev in Khalfin, 1955,
p. 217, pl. 36, fig. 6; from Devonian, W.
Siberia, Russia (cerioid form, corallites 4.5
mm in diameter, septa in two orders 14—16x2,
dissepiments sporadic, tabulae 10—16/10 mm).

Loyolophyllum isolatum Cao in Cao et al., 1983, p.
136, pl. 45, fig. 4a—b; from Middle Devonian,
Lurie Formation, Gansu, northwest China
(cerioid form, corallites 2.5—-3 mm in diameter,
septa two orders 10x2, dissepiments rare, large
and isolated, tabulae complete 14—18/10 mm).

The following two species are only tentatively
included in L. (Loyolophyllum) pending further study
of the type material:

Loyolophyllum praesepimentosum Fligel and Saleh,
1970, p. 285, pl. 4, figs 7-8; from Silurian,
eastern Iran (cerioid form, corallite diameter
4.3—5.2 mm, tabulae 7—8/10 mm); and

Loyolophyllum savitskyi Wu, 1980, pp. 30-32, pl.
5, fig. 3a—b, from Lower Devonian, Uzbekistan
(cerioid form, corallites 2—2.5 mm in diameter).

Excluded from Loyolophyllum are the following
four species:

Loyolophyllum creviseptatum Bulvanker 1958,
p. 159; from Devonian, Russia (longitudinal
sections show mural pores on the wall; hence
this is a tabulate coral likely belonging to the
Syringoporidae).

Loyolophyllum crassispinosum Tchernchev in
Bulvanker et al., 1960, p. 244, and

Loyolophyllum originale Bulvanker in Bulvanker et
al., 1960, p. 243, both from Givetian, Middle
Devonian, Novaya Zemlya (both are cerioid
species of Disphyllidae, close to Spongonaria or
Zelolasma).

203

REVISION OF A DEVONIAN CORAL SPECIES

Loyolophyllum xizangense Yu and Liao, 1982, pp.
100, 101, pl., 2, figs 1-2, text-fig. 3; from Lower
Devonian, northern Xizang, China (cerioid
form, corallites 2.3-3 mm in diameter, tabulae
12-16/5 mm; 1-2 rows of steep dissepiments;
by showing continuous series of dissepiments,
it is here considered to represent a species
of Spongophyllidae, likely belonging to
Spongophyllum).

Loyolophyllum (Fasciloyolopyllum)
subgen. nov.

Type species

Microplasma parallelum Etheridge, 1899, Middle
Devonian (late Eifelian to early Givetian),
from Moore Creek Limestone, near Tamworth,
northeastern New South Wales.

Diagnosis
Like Loyolophyllum (Loyolophyllum), but
phaceloid.

Remarks

Apart from the type species M. parallelum, the
following two species are assigned to Loyolophyllum
(Fasciloyolophyllum):

Fasciphyllum guizhouense Li in Kong and Huang,
1978, p. 124, pl. 40, fig. 6; Givetian, Middle
Devonian, Dushan Formation, Guizhou, South
China (phaceloid form, corallites 6-9 mm in
diameter, septa in two orders 17—20x2, septal
stereozone 0.7—1.2 mm in thickness, dissepiments
elongated, in one discontinous row, tabulae
complete and concave).

Battersbyia qunlingensis Cao in Cao et al., 1983, p.
137, pl. 46, fig. 7a—b; from Early Devonian, Gala
Formation, Qinghai, northwest China (phaceloid
form, corallites 2.5—3.7 mm in diameter, septa in
two orders 12—14x2, septal stereozone 0.8—0.9
mm in thickness, nearly half of the corallite
radius, dissepiments rare and isolated, semi-
globose, tabulae concave).

Loyolophyllum (Fasciloyolopyllum) parallelum
(Etheridge, 1899)
(Figures 2, 3)

Synonymy

Microplasma parallelum Etheridge, 1899, p. 161, pl.
19, figs 1—2, pl. 30, figs 1-2; Fletcher, 1971, p.

204

31; Hill, 1978, p. 28.
Material

Holotype (monotypy): AM FT.3791(TS), AM
FT.4063 (LS), AM FT.14149-14159 (LSs and TSs),
all from AM F.35524 (original number: MMF843,
M568; transferred from Geological and Mining
Museum, Sydney in 1938), from Moore Creek
Limestone, Middle Devonian (late Eifelian to early
Givetian), near Tamworth, northeastern New South
Wales.

Description

Phaceloid corallum, dome-shaped with
dimensions of 15 cm in diameter and 13 cm in height;
corallites slender, regular in size (Fig. 2J), 2.2 mm in
average diameter, closely spaced and paralleling each
other (Fig. 2K), and lateral increasing with corallites
in contact or up to 5 mm apart at the peripheral part
of the corallum.

Septa well developed, in two orders, 11—15x2 for
adult corallites; peripherally dilated to form a narrow
peripheral stereozone up to 0.45 mm in thickness,
thin and weakly wavy in the tabularium, and with
weakly developed carinae (Fig. 2C, F, I); trabecular
structure obscured due to recrystallization; major
septa long, reaching or nearly reaching axis, nearly
radially extending (Fig. 2E); or unequal in length, in
some weakly bisymmetrically arranged (Fig. 2A, C,
F); minor septa variable in length, typically half to
one-third of the corallite radius.

In longitudinal sections, dissepiments (or
presepiments) sporadically developed, large and
elongated (0.5 mm wide and 1.5 mm high), and
vertically arranged with both ends attached to the
wall, occasionally two or three overlapping each
other (Fig. 3B, D). Tabulae complete, varying from
horizontal (Fig. 3F) to deeply concave (Fig. 3C), and
widely spaced, 4 to 5 per 2 mm vertically.

Discussion

The holotype of JL. (Fasciloyolophyllum)
parallelum, the sole known specimen, is partially
silicified with internal structures obscured in most
of the thin sections, and is heavily abraded without
preservation of the proximal tip and the calices. For
these reasons it has remained a poorly known species
in the Devonian coral literature. Etheridge (1899)
illustrated a longitudinal section showing concave,
widely spaced tabulae and lateral budding, and a
transverse section which lacks preservation of septal
structure due to recrystallization. However, the well
developed septa and complete tabulae as described
and illustrated here from some better preserved
corallites easily exclude this coral from the Order

Proc. Linn. Soc. N.S.W., 128, 2007

Y.Y. ZHEN

Figure 2. Loyolophyllum (Fasciloyolophyllum) parallelum (Etheridge, 1899). A, TS, a corallite from AM
FT.14149; B, TanS, a corallite from AM FT.14150; C, TS, a corallite from AM FT.14149; D, TS, a coral-
lite from AM FT.14149; E, TS, a corallite form AM FT.14151; F, TS, a corallite from AM FT.14149; G,
TS, a corallite from AM FT.14151; H, TS, a corallite from AM FT.14153; I, TanS, from AM FT.4063;

J, external upper view, AM F.35524; K, lateral external view, AM F.35524. A-B and D-I, x15 (see scale
bar in A); C, x20; J, x1.5; K, x1; Scale bars 1 mm, unless otherwise indicated.

Cystiphyllida. Its slender corallites, well developed
two orders of septa, complete tabulae, and in particular
the sporadic, elongated dissepiments are comparable
with those of L. (Loyolophyllum) cresswelli from the
Early Devonian Loyola Limestone of Victoria, except
that species is cerioid rather than phaceloid as in L.
(Fasciloyolophyllum) parallelum.

Proc. Linn. Soc. N.S.W., 128, 2007

L. (Fasciloyolophyllum) parallelum can
be distinguished from JL. (Fasciloyolophyllum)
guizhouense (Li in Kong and Huang, 1978) in having
smaller-sized corallites, fewer septa and a thinner
septal stereozone, and from L. (Fasciloyolophyllum)
qunlingensis (Cao in Cao et al., 1983) in having
strongly elongated dissepiments or presepiments.

205

REVISION OF A DEVONIAN CORAL SPECIES

Figure 3. Loyolophyllum (Fasciloyolophyllum) parallelum (Etheridge, 1899). A, LS, a corallite from AM
FT.4063; B, LS, a corallite from AM FT.4063; C, LS, a corallite from AM FT.14153; D, LS, a corallite
from AM FT.14150; E, LS, a corallite form AM FT.4063; F, LS, a corallite from AM FT.4063; all x20.
Scale bar Imm.

ACKNOWLEDGEMENTS

I thank Gary Dargan (Geological Survey of New
South Wales) for preparation of additional thin sections
of the holotype, Dr. Ian Graham (Mineralogy Section,
Australian Museum) for assisting with digital photography,
and Prof. Zhiyi Zhou for assistance in locating relevant
Chinese and Russian literature at the Nanjing Institute of
Geology and Palaeontology, Chinese Academy of Sciences.
Drs Ian Percival, Tony Wright and John Pickett provided

206

constructive comments on the manuscript.

REFERENCES

Benson, W. N. (1915). The geology and petrology of the
Great Serpentine Belt of New South Wales. Appendix
to Part V. The geology of the Tamworth district.
Proceedings of the Linnean Society of New South
Wales 40, 540-624.

Brown, I. A. (1942). The Tamworth Series (Lower and
Middle Devonian) near Attunga, NSW. Journal and

Proc. Linn. Soc. N.S.W., 128, 2007

DOYZEEN

Proceedings of the Royal Society of New South Wales
76, 166-176.

Briihl, D. and Pohler, M. L. (1999). Tabulate corals from
the Moore Creek Limestone (Middle Devonian: late
Eifelian-early Givetian) in the Tamworth Belt (New
South Wales, Australia). Jn R. Feist, J.A. Talent and
A. Daurer (eds), North Gondwana: Mid-Paleozoic
Terranes, Stratigraphy and Biota. Abhandlungen der
Geologischen Bundesanstalt 54, 275-293.

Bulvanker, E. Z. (1958). Devonskie chetyrekhluchevye
korally okrain Kuznetskogo basseyna: 2 vol., 212pp.,
93pls, Vses. Nauchno-issled. Geol. Inst. (Leningrad).
(in Russian)

Bulvanker, E. Z., Vasilyuk, N. P., Zheltonogova, V. A.,
Zhizhina, M. S., Nikolaeva, T. V., Spasskiy, N. Ya.
and Shchukina, V. Ya. (1960). Novye predstaviteli
chetyrekhluchevykh korallov SSSR. Jn B. P.
Markovskiy (ed.), Novye vidy drevnikh rasteniy 1
bespozvonochnykh SSR, 1 (1), 220-254. (in Russian)

Cao, X. D., Ouyang, X., Jin, T. A. and Cai, Z. J. (1983).
Rugosa. /n Xi’an Institute of Geology & Mineral
Resources, Xibei diqu gushengwu tuce: Shaan-Gan-
Ning fence [Palaeontological atlas of northwest
China. Shaanxi-Gansu-Ningxia Volume]. Part 2,
Upper Palaeozoic. 46-179, Geological Publishing
House, Beijing. (in Chinese)

Chapman, F. (1914). Newer Silurian fossils of eastern
Victoria, part 3. Records of Geological Survey of
Victoria 3, 301-316.

Chernyshev, B. B. (1937). Siluriyskie 1 devonskie Tabulata
Mongolii i Tuvy. Akademiya Nauk SSSR. Trudy
Mongolskii Komissii 30, 5-34.

Crook, K. A. W. (1961). Stratigraphy of the Tamworth
Group (Early and Middle Devonian). Tamworth-
Nundle District, N.S.W. Journal and Proceedings of
the Royal Society of New South Wales 94, 173-188.

Ehrenberg, C. G. (1834). Beitraége zur physiologischen
Kenntniss der Corallenthiere im allgemeinen
und besonders des Rothen Meeres, nebst einem
Versuche zur physiologischen Systematik derselben.
Physiologische Abhandlungen der KG6niglichen
Adademie der Wissenschaft, Berlin (1832), p. 225—
380.

Etheridge, R. Jr. (1890). On the occurrence of the genus
Tryplasma Lonsdale (Pholidophyllum Lindstré6m) and
another coral apparently referable to Diphyphyllum
Lonsdale, in the Upper Silurian and Devonian rocks
respectively of New South Wales. Records of the
Geological Survey of New South Wales 2, 15—21.

Etheridge, R. Jr. (1892). Class Actinozoa. Jn R. L. Jack
and R. Jr. Etheridge, Geology and palaeontology
of Queensland and New Guinea. Publications of
Geological Survey of Queensland 92, 50-64, 200—
201.

Etheridge, R. Jr. (1899). On the corals of the Tamworth
district, chiefly from the Moore Creek and Woolomol
Limestones. Records of the Geological Survey of New
South Wales 6, 151-182.

Etheridge, R. Jr. (1918). Two remarkable corals from
the Devonian of New South Wales (Spongophyllum

Proc. Linn. Soc. N.S.W., 128, 2007

halysitoides and Columnaria neminghensis). Records
of the Australian Museum 12, 49-51.

Etheridge, R. Jr. (1920). Further additions to the coral
fauna of the Devonian and Silurian of New South
Wales (Endophyllum schlueteri var. colligatum,
Columnopora (Gephuropora) duni, Vepresiphyllum
falciforme and Syringopora trupanonoides). Records
of the Geological Survey of New South Wales 9,
55-63.

Etheridge, R. Jr. and Foord, A. H. (1884). On two species
of Alveolites and one of Amplexopora from the
Devonian rocks of northern Queensland. Annals and
Magazine of Natural History (ser. 5) 14, 175-179.

Fletcher, H. O. (1971). Catalogue of type specimens
of fossils in the Australian Museum, Sydney. The
Australian Museum Memoir 13, 1-167.

Flugel, H. W. and Saleh, H. (1970). Die palaozoischen
Korallenfaunen Ost-Irans 1 - Rugose Korallen der
Niur-Formation (Silur). Jabrbuch der Geologischen
Bundesanstalt 113, 267-302.

Frey, H. and Leuckart, C. G. F. R. (1847). Beitrage
zor Kenntniss wirbelloser Thiere mit besonderer
Berticksichtigung der Fauna des Norddeutschen
Meeres. vilit170p., 2pls, Verlag von Friedrich
Vieweg und Sohn (Braunschweig).

Goldfuss, G. A. (1862). Petrefacta Germaniae. Erster
Theil. 12 S + 1-252, Taf. 1-71, Arnz, Diisseldorf
1826-1833, 2. Auflage iv + 1-134, Leipzig.

Hatschek, B. (1888-1891). Lehrbuch der Zoologie, eine
morphologische Ubersicht des Thierreiches zur
Einfuhrung in das Studium dieser Wissenschaft: Lief.
1-3, iv + 432p., 407 text-fig., Gustav Fischer (Jena).

Hill, D. (1939). The Devonian rugose corals of Lilydale
and Loyola, Victoria. Proceedings of the Royal
Society of Victoria 51, 219-256.

Hill, D. (1942). The Devonian rugose corals of the
Tamworth District, N.S.W. Journal and Proceedings
of the Royal Society of New South Wales 76, 142—
164.

Hill, D. (1978). Bibliography and index of Australian
Palaeozoic corals. Papers, Department of Geology,
University of Queensland 8, 1-38.

Hill, D. (1981). Rugosa and Tabulata. Jn C. Teichert
(ed.), Treatise on Invertebrate Paleontology, Part F,
Coelenterata, Suppl. 1. F1—762, Geological Society
of America and University of Kansas (New York and
Lawrence).

Jia, H. Z., Xu, S. Y., Kuang, G. D., Zhang, B. F., Zou, Z.
B. and Wu, J. S. (1977). Anthozoa. Jn Hubei Province
Institute of Geology (ed.), Zhongnan diqu gushengwu
tuce [Palaeontological atlas of central southern
China]. Volume 2, Wan gusheng dai bufen [Late
Palaeozoic Era]. 109-272, Geological Publishing
House (Beijing). (in Chinese)

Khalfin, L. L. (1955). Atlas of the leading forms of the
fossil fauna & flora of western Siberia. Moscow,
21-26, 153-154, 185-191, 212-213. (in Russian)

Kong, L. and Huang, Y. M. (1978). Tetracoralla. In
Guizhou Stratigraphy and Palaeontology Work Team
(ed.), Xinan diqu gushengwu tuce [Palaeontological

207

REVISION OF A DEVONIAN CORAL SPECIES

atlas of southwest China]: Guizhou, Volume 1,
Cambrian-Devonian. 35—161, Geological Publishing
House (Beijing). (an Chinese).

Koninck, L. G. de (1876-7). Recherches sur les fossiles
paléozoiques de la Nouvelle-Galles du Sud
(Australie). 373p., Atlas pls i-iv, 1876; pls v-xxiv,
1877, Bruxelles.

Lamarck, J. B. P. A. de M. (1801). Systeme des Animaux
sans Vertébrés, — vill + 432 p., published by the
author, Paris.

Mawson, R. and Talent, J. A. (1994). The Tamworth
Group (mid-Devonian) at Attunga, New South
Wales — conodont data and inferred ages. Courier
Forschungsinstitut Senckenberg 168, 37—58.

Mawson, R., Pang, D. and Talent, J. A. (1997). G.J.
Hinde’s (1899) Devonian radiolarians from
Tamworth, north-eastern New South Wales:
stratigraphic and chronologic context. Proceedings of
the Royal Society of Victoria 109, 233-256.

Milne-Edwards, H. and Haime, J. (1850-1855). A
monograph of the British fossil corals: p. 1-299,
pls 1-72. Palaeontographical Society Monograph
(London).

Orbigny, A. d’ (1850). Prodrome de paléontologie
stratigraphique universelle des animaux mollusques
et rayonnés. V. 1, lx + 349p., Victor Masson (Paris).

Pedder, A. (1967). The Devonian System of New England,
New South Wales, Australia. In D.H. Oswald (ed.),
International Symposium on the Devonian System,
Calgary, 1967. Volume 2. 135-142, Alberta Society
of Petroleum Geologists (Calgary, Alberta).

Philip, G. (1966). Middle Devonian conodonts from the
Moore Creek Limestone, northern New South Wales.
Journal and Proceedings of the Royal Society of New
South Wales 100, 151-161.

Pickett, J. (2002). Ozcorals: a bibliography and index of
fossil corals from Antarctica, Australia, New Guinea
and New Zealand. Version 2. http://www.es.mq.edu.
au/MUCEP/aap/downloads.htm

Schliiter, C. (1885). Diinnschliffe von Zoantharia rugosa,
Zoantharia tabulata und Stromatoporiden aus dem
palaontologischen Museum der Universitat Bonn,
Aussteller Professor Dr. C. Schliiter in Bonn. 52—56,
Catalogue de |’Exposition géologique, Congres géol.
Int. 3" sess. (Berlin).

Soshkina, E. D. (1949). Devonskie korally Rugosa Urala.
Trudy Paleontologicheskogo Instituta. Akademiya
Nauk SSSR 15 (4), 1-162. Gn Russian)

Soshkina, E. D. (1954). Devonski chetyrekhluchevye
korally Russkoy platformy [Devonian tetraradiate
corals of the Russian Platform]. Trudy
Paleontologicheskogo Instituta. Akademiya Nauk
SSSR 52, 1-76, pls 1-19. (an Russian)

Wu, D.L. (1980). New species of rugose corals from
the Upper Silurian and Devonian of northeastern
Fergana. Paleontologicheskiy Zhurnal 1980 (3),
28-33. (in Russian)

Yu, C. M. and Liao, W. H. (1982). Discovery of Early
Devonian tetracorals from Xainza, northern Xizang
(Tibet). Acta Palaeontologica Sinica 21 (1), 96-107.

208 Proc. Linn. Soc. N.S.W., 128, 2007

Ordovician Conodonts from the Watonga Formation, Port
Macquarie, Northeast New South Wales

Davip J. Ocu!”, IAN G. PERCIVAL” AND Evan C. Lettcu!

"Environmental Sciences, University of Technology, Sydney 2007, NSW; and ?Geological Survey of NSW,
Department of Primary Industries, 947-953 Londonderry Road, Londonderry 2753, NSW.

Och, D.J., Percival, I.G. and Leitch, E.C. (2007). Ordovician conodonts from the Watonga Formation,
Port Macquarie, northeast New South Wales. Proceedings of the Linnean Society of New South Wales

128, 209-216.

Conodonts of Middle to Late Ordovician age, obtained from cherts of the Watonga Formation exposed
in the Port Macquarie Block of the Mid North Coast region of New South Wales, establish this unit as
the oldest biostratigraphically-dated part of the southern New England Fold Belt subduction-accretion
complex. Correlation of the Watonga Formation with the Woolomin Formation, faunas from which are no
older than Pridoli, cannot be sustained. This revised age provides evidence of possible early Palaeozoic
subduction-accretion in this region at the same time as arc magmatism, volcaniclastic sedimentation and
exhumation of high-pressure metamorphic rocks were proceeding further west.

Manuscript received 16 November 2006, accepted for publication 15 January 2007.

KEYWORDS: Conodonts, New England Fold Belt, Ordovician, Silurian, Watonga Formation, Woolomin

Formation

INTRODUCTION

Port Macquarie is situated approximately 350
km north of Sydney, geographically located in the
Mid-North Coast region of New South Wales and
geologically in the southern New England Fold
Belt (Fig.la). South of the township, the Watonga
Formation (Leitch 1980), that has been invaded by
minor intrusions and serpentinite bodies, makes up the
eastern portion of the fault-bounded Port Macquarie
Block of Leitch (1974). The Watonga Formation
is well exposed along the coastline between Port
Macquarie and Tacking Point (Fig.lb) where it
comprises mostly broken formation inferred to result
from disruption of a once-stratified sequence of
basalt, chert, siliceous mudstone, siltstone, sandstone
and conglomerate. Several chert-dominated units can
be mapped west from the coast, and locally little-
disrupted basalt forms sections at least several tens
of metres thick. Previous descriptions of the rocks
of the Watonga Formation include those of Barron et
al. (1976), Leitch (1980) and Och et al. (2005), who
all concluded that the formation comprises a part of
the accretionary — subduction complex that is widely
exposed in the New England Fold Belt east of the
Peel-Manning Fault System (Fig. Ic).

The Watonga Formation is unconformably

overlain by early Triassic rocks of the Camden Haven
Group but is faulted against all earlier stratified rocks
(Leitch 1980). A Late Silurian to Late Devonian age
has been assigned to the Watonga Formation, based on
lithological correlation with the Woolomin Formation
of the Tamworth region (Leitch 1980), undescribed
palaeosceniid radiolarians from chert at Watonga
Rocks (Ishiga et al. 1988a) and conodonts reported
from chert at Tacking Point (Ishiga et al. 1988a, b).
In this paper we present new conodont data that more
accurately constrains the age of Watonga Formation
chert, and reassess previous microfossil identifications
from that formation and its possible correlatives in
order to reinterpret its tectonic significance.

CHERT IN THE WATONGA FORMATION

Chert occurs widely in the Watonga Formation, in
mappable chert-rich zones (Fig. 1b), in shear-bounded
blocks associated with basalt and/or mudstone,
and as discrete tectonically isolated blocks. Mass
flow deposits, although uncommon, also contain
chert clasts. Most exposures occur as ‘ribbon chert’
characterised by discontinuous stratification with
individual beds 5 mm to 200 mm thick intercalated

ORDOVICIAN CONODONTS FROM PORT MACQUARIE

2H OU ties, i 93 ‘ ox f 195 fe
“ Finkagt ee
‘Clarence - Town Beach St mee
Moreton e a A
{ wi iy Basin Eee =. Green Mound |
‘ Tablelands Oxleys Beach N

a Rocky Beach

Hupsey

Eis
| 22
Basin pnt ere Macquarie ’ Be Flynns Point

Nine Flynns Beach

om _— Ptmk436

B® Nobbys Beach

21

oO
pb)
“20% 2 beg
4)
=
©
-
Rosendahl
Reservior
19
Miners Beach
18
* _Ptmk432
Ptmk433
“= Tacking Point
remansy a Bt Ue ert (b)
| Watonga Rocks
17 17
89 ‘SORs) 94 $95 Fine
LITHOSTRATIGRAPHIC UNITS
Quarry Middle to Late Ordovician
——, ena A a Dominantly pillow and massive basalt ee
iver, creek and ponds F atonga
Cenozoic gy Dominantly chert Formation
hel Laterite, and alluvial, swamp, and dune complexes. ee Undifferentiated rocks
Permian to Triassic (2) Early Cambrian
BB Matic intrusives IB serpentinite

Figure 1. (a) Location of Port Macquarie in the southern part of the New England Fold Belt. (b) Geologi-
cal map of the northeast corner of the Port Macquarie Block showing localities sampled for conodonts.
Map grid is AMG-66. (Mapping by D. Och). (c) The Port Macquarie Block and adjacent tectonic ele-
ments of the southern New England Fold Belt. Pale grey (Tableland Complex) is mostly accretionary
— subduction complex terranes, grey (Manning Block and Nambucca Slate Belt) Early Permian overlap
sequences, and dark grey (Tamworth Belt) Palaeozoic arc and forearc deposits. Widespread latest Car-
boniferous-Triassic granite bodies omitted.

with thinner recessive dark mudstone units (Some common thicker beds (up to 1 m) are more laterally
mere films), and together forming irregular flat lenses continuous but change thickness along strike and
rarely exceeding 0.5 m in length (Figs 2a, b). Less mostly lens out or are terminated by faults and shears

210 Proc. Linn. Soc. N.S.W., 128, 2007

D.J. OCH, I.G. PERCIVAL AND E.C. LEITCH

Figure 2. (A) Ribbon chert showing discontinuous character of stratification, thickness variation in indi-
vidual layers, and irregular folds, all suggestive of soft sediment deformation. Tacking Point (Grid refer-
ence 493930 mE 6517660 mN, location of sample Ptmk 433). (B) Broken formation of prominent chert
lenses and recessive sheared mudstone matrix offset by a late fault, Town Beach (Grid reference 492600
mE 6522550 mN, location of sample Ptmk 434).

within individual exposures.

The chert is highly variable in colour, with
white and grey predominating but including red,
orange-brown, green and black varieties. Primary
sedimentary structures other than relic bedding are
absent. Widespread disharmonic mesoscopic folds,
lacking any signs of axial surface structure, of
variable orientation, and of a style ranging from box-
shaped to fluidal, are interpreted as pre-consolidation
slump structures. Ghosts of radiolarians are preserved
in some specimens but others consist solely of fine
anhedral quartz grains with a dusting of iron oxide
minerals. Samples collected from Tacking Point
(localities Ptmk 432 & 433) are strongly recrystallised
due to local thermal metamorphism, whereas those
from Flynns Beach and Town Beach (Ptmk 436 and
Ptmk 434, respectively), although highly veined, show
only the effects of the low grade regional alteration
that has affected all of the Watonga Formation.

PREVIOUS AGE-DATING OF THE WATONGA
FORMATION

Accurate dating of the rocks of the Port
Macquarie Block is important in understanding the
geological evolution of the region, better defining
the timing of accretion, determining relationships
of the block to other structural units of the southern
New England Fold Belt and establishing tectonic
relationships with other early Palaeozoic elements
in eastern Australia. Leitch (1980) first defined
Palaeozoic lithostratigraphic units in the area south

Proc. Linn. Soc. N.S.W., 128, 2007

and west of Port Macquarie, and recognised the
chert-rich Watonga Formation as having lithological
similarities to the Woolomin Formation exposed south
of Tamworth (Fig. 1c). On this basis he assigned a
generalised early Palaeozoic age to the Watonga
Formation although biostratigraphically useful fossils
had not been obtained from these rocks. Prior to the
present investigation, biostratigraphic data for the
Watonga Formation in the Port Macquarie Block
was limited to a record of the conodont Belodella
spp. (indicative of a generalised Late Silurian-Early
Devonian age) recovered from red cherts at Tacking
Point, and late Frasnian (Late Devonian) palaeosceniid
radiolaria including Palaeorubus hastingensis Ishiga,
1987 (in Ishiga et al. 1987) from chert at Watonga
Rocks (Ishiga et al. 1988a; b). Although this material
was never illustrated, these ages have been widely
accepted as dating the Watonga Formation.

BIOSTRATIGRAPHY OF THE WOOLOMIN
FORMATION

In the light of the inferred lithological correlation
of the Watonga Formation with the Woolomin
Formation of the Tablelands Complex to the west
near Tamworth, it is important to reassess available
data on ages from the latter unit. Major advances
in dating the Woolomin Formation resulted from a
collaborative project between geologists from the
University of Sydney and a consortium of Japanese
universities in the mid-1980s. Microfossils, including
radiolaria and conodonts, were extracted from cherts

211

ORDOVICIAN CONODONTS FROM PORT MACQUARIE

by dissolution in hydrofluoric acid; the results were
presented in a detailed report (Ishiga et al. 1988b)
with significant findings summarised in Ishiga et al.
(1988a). Conodonts identified by these authors as
Ozarkodina eosteinhornensis, Belodella cf. resima
and Walliserodus sp. were recovered from one locality
(WA-50) in red bedded chert north of Woolomin. This
assemblage, although limited, is clearly indicative
of a Late Silurian (Pridoli) age due to the presence
of the nominate species of the eosteinhornensis
zone, now referred to as Ozarkodina remscheidensis
eosteinhornensis. Additional specimens illustrated by
Ishiga et al. (1988b) confirm the identity of this species.
However, Belodella resima is a characteristic Early
Devonian form, not known to occur in the Silurian
and typically displaying strong marginal costae
not observed in the specimens from the Woolomin
Formation. The four specimens illustrated from
locality WA-50 by Ishiga et al. (1988b, pl. 3 figs 9-12)
are insufficiently well-preserved to show diagnostic
features of the Late Silurian species B. anomalis, so
they are probably best assigned to Belodella sp. The
specimen figured by Ishiga et al. (1988b, pl. 3 fig.
8) as Walliserodus sp. is more likely to be the non-
denticulate M element of this Belodella as it appears
to lack the pronounced costae on lateral faces that
characterises Walliserodus. \shiga et al. (1988a,
fig. 2p) illustrated a sharply-keeled element with
triangular cross section, planar to slightly concave
lateral and posterior faces and a deep basal cavity
that they also assigned to Walliserodus sp. from the
same locality; it is here identified as the long-ranging
genus Coelcerodontus sp. Four additional fragments
of S elements of ozarkodinids, depicted by Ishiga et
al. (1988b, pl. 3 figs. 13-16), are too incomplete for
identification but are consistent with Late Silurian
species. The Pridoli age inferred for cherts at locality
WA-50 contrasts with an early Carboniferous age
determined for similar cherts south of Woolomin
(locality 29 of Ishiga et al. 1988a; b) based on the
occurrence of the radiolarian A/baillella sp. Given the
highly imbricate structure of this area the possibility
of infaulting of younger units as occurs north of
Woolomin cannot be discounted (cf. Cawood 1982).
While the Pridoli age determined for the Woolomin
Formation is generally consistent with the previously
accepted age of the Watonga Formation, new data
reported here suggest that the Watonga Formation is
considerably older.

PW

Figure 3 (RIGHT). Conodonts from Watonga For-
mation chert, Port Macquarie Block

A — G from locality Ptmk 432, Tacking Point; A.
Panderodus recuryvatus?, MMMC 4348; B. eobe-
lodiniform element of Belodina sp., MMMC 4349;
C. Paroistodus sp.. MMMC 4350; D. drepanois-
todid element in oblique view, showing strongly
flared basal cavity (specimen intersected by V-
shaped cryptocrystalline vein), MMMC 4351; E.
indeterminate coniform element, MMMC 4352;
F. drepanoistodid element with expanded base,
MMMC 4353a; G. strongly reclined M element,
possibly related to Ansella sp., MMMC 4353b. H
— K from locality Ptmk 434, Town Beach; H. uni-
dentified platform? element with two prominent
peg-like denticles (remainder of specimen, in up-
per part of figure, is out of plane of focus), MMMC
4355; I. unidentified S element, possibly referrable
to Periodon, showing fine denticles but apparently
incomplete posteriorly, MMMC 4356; J. exten-
sively fragmented element of Belodina, anterior to
right, MMMC 4357; K. lateral view of unidentified
platform? element with discrete peg-like denticles
comparable with those shown in H, MMMC 4358;
L — W from locality 436, Flynns Beach. L. Pan-
derodus sp., MMMC 4359a; M. Panderodus recur-
vatus? showing curvature of cusp comparable with
specimen depicted in A (although expanded base
is more reminiscent of drepanoistodid), MMMC
4360a; N. Strachanognathus parvus, interpreted as
split along cusp and dextrally separated, MMMC
4361; O. unidentified element with two prominent
denticles on lateral process, MMMC 4359b; P.
Protopanderodus? sp. with posteriorly-extended
base, MMMC 4359c¢; Q. indeterminate coniform
element resembling Dapsilodus, MMMC 4362;
R. Periodon aculeatus, S element (note that cusp
is displaced into section below plane of focus),
MMMC 4360b; S. Phragmodus? sp., P element,
MMMC 4363a; T. unidentified Pb element (anteri-
or process unclear, but note three or more discrete
denticles on posterior process), MMMC 4364; U.
Pseudobelodina sp., MMMC 4360c; V. Pseudobe-
lodina sp.. MMMC 4363b; W. unidentified Sa?
(symmetrical) element, MMMC 4365. All ele-
ments illustrated in lateral view unless otherwise
indicated. Scale bar for each specimen represents
100 pm. Chert sections are housed in the microfos-
sil reference collection (MMMC) of the Geological
Survey of NSW, Londonderry Geoscience Centre,
Londonderry 2753.

Proc. Linn. Soc. N.S.W., 128, 2007

D.J. OCH, I.G. PERCIVAL AND E.C. LEITCH

PAB

Proc. Linn. Soc. N.S.W., 128, 2007

ORDOVICIAN CONODONTS FROM PORT MACQUARIE

NEW DATA ON THE AGE OF THE WATONGA
FORMATION

More than 25 conodont elements were observed
in cherts collected from outcrops of the Watonga
Formation on the coast immediately south of Port
Macquarie (see Fig. 1b for sample locations). Unlike
previous investigations, these microfossils were not
dissolved out of the cherts using HF. Instead, the
technique used involved preparing four or five large
(7.5 x 3.8 mm) thin sections from each sample, cut
parallel to bedding planes of the cherts, and ground
to a thickness of about 50-60 microns. Both sides
of the finished thin section were carefully examined
with a binocular microscope using transmitted light.
This method (first developed by Ian Stewart of
Monash University) has previously been employed
to investigate conodont biostratigraphy in cherts
exposed in the Narooma area on the NSW south coast
(Glen et al. 2004), and from the Tamworth Belt south
of Tamworth (Fig. lc) from where Stewart (1995)
determined a Middle to early Late Cambrian age for
conodonts in spiculitic chert of the Pipeclay Creek
Formation. Although conodonts found in this way can
only be observed in the plane of section in whatever
random orientation is presented to the viewer, the
thin section technique has the distinct advantage of
preserving extensively fractured elements that would
be destroyed by dissolution in HF.

Samples Ptmk 432 and Ptmk 433 were obtained
from grey-black chert outcrop at Tacking Point (Fig.
1b). Although the two cherts closely resemble each
other, sample Ptmk 433 was devoid of microfossils
whereas Ptmk 432 yielded eight elements. One of
these (Fig. 3B) is identified as the eobelodiniform
element of the Late Ordovician genus Belodina.
Associated Paroistodus (Fig. 3C) and unidentified
drepanoistodids are consistent with this age.

A further four conodont elements were found in
sections cut from sample Ptmk 434, a grey-black chert
at Town Beach. Although extensively fractured, the
belodiniform element of Belodina (Fig. 3J) is readily
identifiable, and provides support for recognition
of another element from the Belodina apparatus in
sample Ptmk 432. Elsewhere in New South Wales
several species of Belodina have been identified from
limestones of Gisbornian and Eastonian age (Late
Ordovician) in the central part of the state. Belodina
is also known from limestone olistoliths of Eastonian
age in the Wisemans Arm and Drik Drik formations
of the Tablelands Complex and Tamworth Belt,
respectively, in the New England Fold Belt (Furey-
Greig 1999; 2000). An unidentified element illustrated
in Fig. 3I most closely resembles a specimen identified

214

as Phragmodus sp. by Iwata et al. (1995: figure 2g)
from the Ballast Formation east of Cobar in central
NSW, though the Town Beach example is unlikely to
be that genus (R.S. Nicoll, pers. comm.. 2006). The
Ballast Formation fauna was tentatively assigned a
Late Ordovician age by Iwata et al. (1995).

Orange-brown chert from Flynns Beach (sample
Ptmk 436) was the most prolific of the samples
examined, yielding 14 conodont elements. Periodon
aculeatus (Fig. 3R) has a range of late Darriwilian
(late Middle Ordovician) to early Gisbornian (early
Late Ordovician). Pseudobelodina (Fig. 3U,V) and
Strachanognathus parvus (Fig. 3M) also occur over
this interval, which is consistent with ages of other
genera tentatively identified from this locality. Thus
locality 436 may be slightly older than both localities
432 and 434 which could be identical in age.

In summary, the maximum age range indicated
for cherts from the Watonga Formation appears to
extend from late Darriwilian to somewhere within
the Late Ordovician, possibly as young as the end
of the Eastonian (middle Late Ordovician). There is
no evidence for Silurian or Devonian ages in these
samples, thus conflicting with the previously reported
occurrence of Belodella spp. (Ishiga et al. 1988a; b).
Several Ordovician conodont elements, such as those
of Ansella and Belodina (the latter newly identified
from Tacking Point) resemble Belodella, though
without illustrations of specimens that were obtained
from the Watonga Formation by Ishiga it is impossible
to clarify whether or not this was misidentified.
Palaeosceniid radiolaria obtained from the Watonga
Formation at Watonga Rocks include Palaeorubus
hastingensis, otherwise known only from late
Frasnian rocks at Yarras in the western part of the
Hastings Block, to the west of the Port Macquarie
Block (Ishiga 1988). Until more information is
available concerning the age range of this form, the
Late Devonian age accorded the unit here cannot be
regarded as well established.

TECTONIC IMPLICATIONS

Although no coherent stratigraphic sequence
has been demonstrated in the Watonga Formation,
the association of chert and mudstone with basalt
mainly of mid-ocean ridge magmatic affinity (Och
et al. 2004), and interbedded graded volcaniclastic
sandstone and siltstone, is readily interpreted as the
product of sea-floor spreading, the accumulation of
pelagic oceanic rocks, and trench deposition on a plate
undergoing subduction. The early deformational style
of the Watonga Formation, involving progressive

Proc. Linn. Soc. N.S.W., 128, 2007

D.J. OCH, I.G. PERCIVAL AND E.C. LEITCH

disruption of the rocks in a continuum from soft
sediment deformation (Fig. 2a) to shearing and
stretching producing broken formation (Fig. 2b), is
consistent with accretionary — subduction tectonics.
Furthermore the Watonga rocks are of similar
structure to that of many units considered to be of
subduction — accretion origin, including pre-Permian
rocks from elsewhere east of the Peel-Manning Fault
System in New England.

The Middle—Late Ordovician chert samples Ptmk
432, 434 and 436 are the oldest biostratigraphically
dated rocks recorded from the New England
accretionary — subduction complex, and rule out
correlation of the Watonga Formation with the
Woolomin Formation chert from which only Late
Silurian and younger ages are known (see preceding
discussion). The Ordovician chert ages provide a
minimum age for the oceanic plate sediments that were
subducted, and raise the possibility that the Watonga
Formation was accreted in the early Palaeozoic,
at which time arc magmatism and volcaniclastic
sedimentation were proceeding further west in the
New England Fold Belt (Cawood 1983; Furey-Greig
et al. 2000; Offler and Shaw 2006) and in the eastern
Lachlan Fold Belt (Glen et al. 1998), and high pressure
metamorphic rocks now embedded within serpentinite
bodies at Port Macquarie were being exhumed under
blueschist facies conditions (Fukui et al. 1995; Och
et al. (2003). Scheibner (1998) and Glen (2005) have
suggested that the blueschists are accretionary rocks
that formed outboard of the Macquarie Arc in the
eastern Lachlan Fold Belt. Ordovician accretion of
the Watonga Formation rocks would be consistent
with them constituting a fragment of the Narooma
accretionary complex postulated to extend into New
England by these workers. The Ordovician age for the
Watonga Formation also provides a potential source
for chert blocks widespread in the (?) Early Devonian
Wisemans Arm Formation (Leitch and Cawood 1980;
Furey-Greig 1999).

The Watonga Formation is isolated from the
remainder of the New England accretionary complex
by the allochthonous Hastings Block (Fig. 1b) that
was moved north from along strike of the southern
Tamworth Belt in the Late Palaeozoic (Schmidt et al.
1994). It is likely that the Watonga Formation was
similarly displaced from a position much closer to
the Peel-Manning Fault System, in the vicinity of
which the Early Palaeozoic rocks of the New England
Fold Belt are concentrated. The central and eastern
parts of the New England accretionary — subduction
complex have yielded chert samples no older than
latest Devonian (e.g. Aitchison and Flood 1990; see
summary in Fergusson et al. 1993, fig. 12).

Proc. Linn. Soc. N.S.W., 128, 2007

CONCLUSIONS

Conodont faunas identified in bedding-parallel
thin sections from Watonga Formation cherts of
the Port Macquarie district are of Middle-Late
Ordovician age, thus making these cherts the oldest
biostratigraphically dated rocks from the extensive
accretionary — subduction complex of the New
England Fold Belt. The dates provide a minimum
age for the oceanic plate that was subducted, and
raise the possibility that the Watonga Formation was
accreted in the early Palaeozoic at the same time as
arc magmatism and volcaniclastic sedimentation were
preceding further west. A reassessment of the age of
a conodont fauna from the Woolomin Formation in
the southwestern part of the accretionary complex
indicates it is of Late Silurian (Pridoli) age, and hence
substantially younger than the Watonga Formation.
Thus the previously conjectured correlation between
the Woolomin and Watonga formations can no
longer be sustained. The Watonga Formation was
probably moved north in the late Palaeozoic during
displacement of the Hastings Block along strike from

the southern Tamworth Belt.

ACKNOWLEDGMENTS

We thank Anthony Och for his valuable assistance
in the field. Gary Dargan (Geological Survey of NSW)
prepared the chert thin sections. Conodont microfossils
were photographed by David Barnes (NSW Department
of Primary Industries). Bob Nicoll (Canberra) and John
Pickett (Londonderry) provided useful second opinions
on conodont identifications. Constructive reviews by Dick
Glen, Peter Cawood and Yong-yi Zhen were instrumental
in reorganising the manuscript for publication. David
Och and Ian Percival publish with permission of the
Deputy Director-General, NSW Department of Primary
Industries — Mineral Resources. The work of David Och
and Evan Leitch was supported in part by ARC Large Grant
A39601646. David Och acknowledges support from a
University of Technology, Sydney, Faculty of Science PhD
Research Scholarship. This paper is a contribution to IGCP
503: Ordovician Palaeogeography and Palaeoclimate.

REFERENCES

Aitchison, J.C. and Flood, P.G. (1990). Early
Carboniferous radiolarian ages constrain the timing
of sedimentation within the Anaiwan terrane, New
England Orogen, eastern Australia. Neues Jahrbuch
fiir Geologie und Paldontologie, Abhandlungen 180,
1-19.

215

ORDOVICIAN CONODONTS FROM PORT MACQUARIE

Barron, B.J., Scheibner, E. and Slansky, E. (1976). A
dismembered ophiolite suite at Port Macquarie, New
South Wales. Records of the Geological Survey of
New South Wales 18, 69-102.

Cawood, P.A. (1982). Structural relations in the
subduction complex of the Paleozoic New England
Fold Belt, Eastern Australia. Journal of Geology, 90,
381-392.

Cawood, P.A. (1983). Modal composition and detrital
clinopyroxene geochemistry of lithic sandstones
from the New England Fold Belt (eastern Australia):
a Paleozoic forearce terrane. Geological Society of
America Bulletin, 94, 1199-1214.

Fergusson, C.L., Henderson, R.A., Leitch, E.C. and
Ishiga, H. (1993). Lithology and structure of the
Wandilla terrane, Gladstone- Yeppoon district, central
Queensland, and an overview of the Palaeozoic
subduction complex of the New England Fold Belt.
Australian Journal of Earth Sciences 40, 403-414.

Fukui, S., Watanabe, T., Itaya, T. and Leitch, E.C. (1995).
Middle Ordovician high PT metamorphic rocks in
eastern Australia: evidence from K-Ar ages. Tectonics
14, 1014-1020.

Furey-Greig, T. (1999). Late Ordovician conodonts from
the olistostromal Wisemans Arm Formation (New
England Region, Australia). Abhandlungen der
Geologischen Bundesanstalt 54, 303-321.

Furey-Greig, T. (2000). Late Ordovician (Eastonian)
conodonts from the Early Devonian Drik Drik
Formation, Woolomin area, eastern Australia.
Records of the Western Australian Museum 58, 133-
143.

Furey-Greig, T., Leitch, E.C. and Cawood, P.A. (2000).
Early Palaeozoic arc-basin sequence in the Tamworth
Belt: constraints for East Gondwana tectonics.
Geological Society of Australia, Abstracts 59, 163.

Glen, R.A. (2005). The Tasmanides of eastern Australia.
In “Terrane Processes at the Margins of Gondwana’
(eds A.P.M. Vaughan, P.T. Leat and R.J. Pankhurst).
Special Publication of the Geological Society,
London 246, 23-96.

Glen, R.A., Stewart, I.R. and Percival, I.G. (2004). The
Narooma Terrane: implications for the construction of
the outboard part of the Lachlan Orogen. Australian
Journal of Earth Sciences 51, 859-884.

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.

Ishiga, H. (1988). Paleontological study of radiolarians
from the southern New England Fold Belt, Eastern
Australia. pp. 77-93 in “Preliminary Report on the
geology of the New England Fold Belt, Australia’.
Co-operative Research Group of Japan and Australia,
Department of Geology, Shimane University.

Ishiga, H., Leitch, E.C., Naka, T., Watanabe, T. and
Iwasaki, M. (1987). Late Devonian Palaeoscenidum
from the Hastings Block, New England Fold Belt,
Australia. Earth Science (Chikyu Kagaku) 41, 297-
302.

216

Ishiga, H., Leitch, E.C., Watanabe, T., Naka, T. and
Iwasaki, M. (1988a). Radiolarian and conodont
biostratigraphy of siliceous rocks from the New
England Fold Belt. Australian Journal of Earth
Sciences 35, 73-80.

Ishiga, H., Watanabe, T. and Leitch, E.C. (1988b).
Microfossil biostratigraphy and lithologic association
of bedded chert in the Macdonald Block of the New
England Fold Belt, Eastern Australia. pp. 47-59
in ‘Preliminary Report on the geology of the New
England Fold Belt, Australia’. Co-operative Research
Group of Japan and Australia, Department of
Geology, Shimane University.

Iwata, K., Schmidt, B.L., Leitch, E.C., Allan, A.D. and
Watanabe, T. (1995). Ordovician microfossils from
the Ballast Formation (Girilambone Group) of New
South Wales. Australian Journal of Earth Sciences
42, 371-376.

Leitch, E.C. (1974). The geological development of the
southern part of the New England Fold Belt. Journal
of the Geological Society of Australia 21, 133-156.

Leitch, E.C. (1980). Rock units, structure and
metamorphism of the Port Macquarie Block, eastern
New England Fold Belt. Proceedings of the Linnean
Society of New South Wales 104, 273-292.

Leitch, E.C. and Cawood, P.A. (1980). Olistoliths and
debris flow deposits at ancient consuming plate
margins: an eastern Australian example. Sedimentary
Geology 25, 5-25.

Och, D.J., Caprarelli, G. and Leitch, E.C. (2004).
Geochemical investigation of igneous rocks in
the Palaeozoic Port Macquarie Complex, NSW,
Australia. Geological Society of Australia, Abstracts
73, 178.

Och, D.J., Leitch, E.C., Caprarelli, G. and Watanabe, T.
(2003). Blueschist and eclogite in tectonic melange,
Port Macquarie, New South Wales, Australia.
Mineralogical Magazine, 67, 609-624.

Och, D.J., Leitch, E.C., Graham, I.T. and Caprarelli,

G. (2005). “A Fieldguide to the palaeosubduction
complex of Port Macquarie, NSW’. (Specialist Group
in Geochemistry, Mineralogy and Petrology Field
Guide, Geological Society of Australia: Sydney).

Offier, R. and Shaw, S. (2006). Hornblende Gabbro
Block in serpentinite melange, Peel-Manning Fault
System, New South Wales, Australia: Lu-Hf and
U-Pb isotopic evidence for mantle-derived, Late
Ordovician igneous activity. Journal of Geology 114,
211-230.

Scheibner, E. and Basden, H. ed. (1998). Geology of
New South Wales — Synthesis, Volume 2 Geological
Evolution. Geological Survey of New South Wales,
Memoir Geology 13(2), 666 pp.

Schmidt, P.W., Aubourg, C., Lennox, P.G. and Roberts,

J. (1994). Palaeomagnetism and tectonic rotation of
the Hastings Terrane, eastern Australia. Australian
Journal of Earth Sciences 41, 547-560.

Stewart, I. (1995). Cambrian age for the Pipeclay Creek
Formation, Tamworth Belt, northern New South
Wales. Courier Forschungsinstitut Senckenberg 182,
565-566.

Proc. Linn. Soc. N.S.W., 128, 2007

First Record of Thecostegites (Cnidaria: Tabulata) from Central
New South Wales

Gary DARGAN

Geological Survey of New South Wales, Department of Primary Industries, Londonderry Geoscience Centre,
943-957 Londonderry Rd, Londonderry NSW 2753

Dargan, G. (2007). First Record of Thecostegites (Cnidaria: Tabulata) from central New South Wales.
Proceedings of the Linnean Society of New South Wales 128, 217-221.

Thecostegites myolaensis, a new species of tabulate coral, is described from northwest of Parkes, New
South Wales. This is the only record of Thecostegites from the Australian mainland. Associated conodonts
establish a latest Ludlow (Late Silurian) age for this species, making this the oldest recorded occurrence
of the genus. Comparison with Thecostegites species from the Pridoli of Tadzhikistan and the Polar Urals
suggests that the genus originated in Australia and subsequently spread to these regions.

Manuscript received 16 November 2006, accepted for publication 15 January 2007.

KEYWORDS: Late Silurian, palaeobiogeography, tabulate coral, Thecostegites.

INTRODUCTION

The Australian record of the tabulate coral
genus Thecostegites is scant, with the only described
species being 7: ejuncidus Jell and Hill, 1969, from
the Point Hibbs Limestone (Pragian, Early Devonian)
of Tasmania. An older species, the earliest known
representative of the genus, is here described as the
new species 7: myolaensis. Its sole occurrence is in
Late Silurian limestone on the property ‘Myola’,
located 5 km southwest of the town of Trundle, 55
km northwest of the city of Parkes in central New
South Wales (Fig.1). The limestone contains abundant
tabulate and rugose corals and stromatoporoids, and
is the probable type locality for the stromatoporoid
Clathrodictyon __(Plexodictyon) —_ conophoroides
Etheridge, 1921 (Pickett and Ingpen 1990; Foldvary
2000). Sherwin (1996) mapped this limestone as
part of the Cookeys Plains Formation within the
Derriwong Group, assigning to it a Pridoli (latest
Silurian) to early Lochkovian (earliest Devonian)
age. He mentioned the occurrence of the conodont
Ozarkodina crispa, the nominate species of the crispa
zone, at a locality southeast of Trundle reported by
Pickett and Ingpen (1990, cover photos E and F)
and agreed with them that the age may be slightly
older than Pridoli. A latest Ludlow age is definitively
established for the limestone at the ‘Myola’ locality by

the presence of O. crispa (Pickett and Ingpen, cover
photo A). Simpson and Talent (1995) concur with a
latest Ludlow crispa zone age for these localities.

AGE AND BIOGEOGRAPHIC
CONSIDERATIONS

Thecostegites is well known from the Middle
Devonian of North America and the Middle and Late
Devonian of Europe and Asia. Nudds and Sepkoski
(1993) mentioned a Thecostegites from the Late
Silurian or Early Devonian of the Polar Urals as
the oldest known Thecostegites. This is most likely
Thecostegites tchernychevi Barskaya, 1965, from
the Greben Horizon in the Chernova Swell in the
Polar Urals (fide Chudinova 1986, Dubatolov et al.
1986). Both Chudinova and Dubatolov et al. gave
an “Upper Ludlow” age for this locality, but the
Greben Horizon is now known to be Pridoli (Talent
et al. 2001). Of comparable age is Thecostegites
isfaraensis Chekovich, 1960, from the Isfara Horizon
in Southern Fergana, Uzbekistan. Although in the
original description of this species the age is given
as “Upper Ludlow”, the upper part of the Isfara
Horizon is now also regarded as Pridoli (Talent et
al. 2001). Note that, in the original publication of T.
isfaraensis, Chekovich (1960) used the specific epithet
‘isfardensis’ in the text but labelled the illustrations

FIRST Thecostegites FROM NEW SOUTH WALES

NEW SOUTH WALES

TRUNDLE
a

Major roads
Tottenham Bogan Gate Railway

* Thecostegites myolaensis locality

Qutcrop of
Cookeys Plains Formation

33°00'S

Figure 1. A; Map of New South Wales showing general location of Trundle area, B; Map of Trundle
area showing outcrop of Cookeys Plains Formation and location of Thecostegites myolaensis (indicated

by star).

as T. isfaraensis. Subsequently Chudinova (1986)
used the name T. isfardensis when referring to this
species. This appears to be a case of Japsus calami
which I have corrected in accordance with ICZN
32.5.1 by using the spelling “isfaraensis’ to reflect the
stratigraphic occurrence of the type species.

Thecostegites myolaensis is the only species
known from mainland Australia and, with its age
established as late Ludlow, is also considerably older
than the only other Australian species T. ejuncidus,
known from the Early Devonian of Tasmania.

Although the geological record of Thecostegites
is patchy and undoubtedly incomplete, the available
data indicate that that the earliest known species is 7.
myolaensis. Species known from Uzbekistan and the
Polar Urals are of slightly younger, Pridoli age. This
suggests that the genus originated in Australia and
subsequently spread to these regions. As these two
areas were remote from Eastern Australia according
to Late Silurian reconstructions of global continental
distribution (Cocks and Torsvik 2002) it is difficult
to demonstrate a migration of Thecostegites without
occurrences (as yet undetected) in intervening
regions.

218

SYSTEMATIC PALAEONTOLOGY

Order AULOPORIDA Sokolov, 1947
Superfamily SYRINGOPORICAE de Fromentel,
1861
Family THECOSTEGITIDAE de Fromentel, 1861
Genus Thecostegites Milne-Edwards & Haime,
1849, p.261

Type species

Harmodites bouchardi Michelin, 1846, from the
Upper Devonian (Frasnian) at Ferques near Boulogne,
France, (by monotypy).

Diagnosis (Hill, 1981, p. 660)

Corallum massive and encrusting; corallites
slender, cylindrical, thick-walled, united by successive
irregular platform-like expansions of tabulate
tissue, each expansion in communication with the
tabularia through perforations arranged in verticils
in the walls of the corallites; the expansions may be
epithecate above and below; septal spines irregular
in development; tabulae in lateral expansions as well
as in the cylindrical corallites, irregular, horizontal,

Proc. Linn. Soc. N.S.W., 128, 2007

G. DARGAN

Figure 2. Thecostegites myolaensis sp. nov. holotype MMF 44854a-b. A, Transverse section showing
chains of interconnected corallites. B, Longitudinal section showing connecting tubes and a new corallite
branching off an adult with a deeply depressed tabula extending into the adult. C, Transverse section
showing merging of corallites to produce a platform-like area. D, Longitudinal section showing vertical
corallites arising from prostrate corallites in a part of the corallum where growth has been interrupted.
Scale bar equals 5 mm.

oblique, concave or with short axial tubes, which
may extend into the lateral expansions where they lie
horizontally, and may be crossed by small tabellae.

Thecostegites myolaensis sp. nov.
Fig. 2 A-D

Diagnosis

Thecostegites with long, closely spaced
corallites 1.1 to 1.5 mm. diameter, connected by short
verticillate tubular to platform-like lateral expansions;
septal spines absent; tabulae numerous, thin and
steeply inclined and incomplete or deeply depressed,
occasionally horizontal with a median depression.

Derivation of name

After the property “Myola” where the specimen
was found.

Proc. Linn. Soc. N.S.W., 128, 2007

Type Locality
Roadside paddock on ’’Myola” property (Trundle

1:50,000 map, grid reference 644549). Locality ‘D’ of
Pickett and Ingpen 1990 and locality ‘X’ of Foldvary
2000.

Holotype (and sole specimen)
Two pieces of a single corallum MMF 44854a-b
with one transverse and one longitudinal section.

Description

Corallites are long and closely spaced. Their
diameter ranges from 1.1 to 1.5 mm with an average
of 1.4 mm. Wall thickness ranges from 0.13 to 0.2
mm with an average thickness of 0.16 mm. Corallites
connected by numerous short verticillate tubes
spaced 5 to 7 per 5 mm. Tubes occur at the same
level on several adjoining corallites, resulting in long

INS)

FIRST Thecostegites FROM NEW SOUTH WALES

chains of connected corallites visible in transverse
section (Fig 2 A, C). Corallites increase in diameter
at the connecting tubes and sometimes merge to form
platform-like areas (Fig. 2 C). In the distal portion of
the colony corallites reach a maximum length of 18
mm.
Tabulae range from 11 to 17 per 5 mm. They are
usually steeply inclined and incomplete or deeply
depressed; rarely horizontal to slightly inclined with
a median depression and rarely forming a syrinx.
They pass through tubes into adjoining corallites and
are usually thin but sometimes thicken near corallite
walls. Horizontal tabulae are more common where
connecting tubes occur. Septal spines are absent.

In the proximal portion of the corallum some of
the vertical corallites arise from prostrate corallites.
This also occurs in portions of the corallum where
growth has been interrupted (Fig. 2 D). Increase is
lateral, non-parricidal and occurs via a connecting
tube in the parent corallite. A deeply depressed tabula
passes from the parent into the new corallite (Fig. 2
B).

Remarks

The corallum is approximately 120 mm across
and 45 mm high and has broken into two equal sized
pieces. It has grown on a corallum of Heliolites
daintreei.

The new species differs from 7’ ejuncidus in
possessing larger corallites, having more abundant
steeply inclined tabulae and lacking septal spines.

Thecostegites myolaensis has a similar corallite
diameter to T’'isfaraensis but horizontal tabulae are
more abundant in the latter. This also distinguishes
T-myolaensis from T. tchernychevi, which has a
similar corallite diameter to 7’ ejuncidus (Fig. 3).

ACKNOWLEDGEMENTS

Kathy Stait (University of Tasmania) kindly provided
access to type material of T: ejuncidus for examination. Jan
Percival (Geological Survey of NSW) and John Pickett
and an anonymous reviewer provided valuable advice
and criticism im preparing this manuscript. David Barnes
assisted with the photography and David Och with drafting
the map. Published with the permission of the Deputy
Director General, NSW Department of Primary Industries
— Mineral Resources Division.

REFERENCES

Chekovich, V.D. (1960). Novye vidi drevnikh rstenii
i bespozvonochnikh CCCP. (New species of
fossil plants and invertebrates of the USSR).
209-210, pl. 41, fig.1. Vsesoyuznii Nauchno-

mg 7.myolae
@ T.ejuncidus

A 7.isfaraensis

@ 7.tchemychevi

Horizontal Tabulae/TotalTabulae

0 0.5 1

1.5 7

Corallite Diameter mm.

Figure 3. Scatter plot of horizontal tabulae/total tabulae vs. corallite diameter for 7.
myolaensis, T. ejuncidus, T. isfaraensis and T. tchernychevi. Data for T. myolaensis and
T. ejuncidus were obtained from specimens. Data for T. isfaraensis and T. tchernychevi
were obtained from published measurements and illustrations.

220

Proc. Linn. Soc. N.S.W., 128, 2007

G. DARGAN

issledovatel’ski1 Geologicheskii Institut
Ministerstva Geologii 1 Okhrai Nedr CCCP. [in
Russian].

Chudinova, I.I. (1986). Sostav Sistema i Filogeniya
Iskopaemikh Koralloy Otryad Syringoporida.
(Systematics and Phylogeny of Fossil
Corals, Order Syringoporida). Trudy
Palaeontologicheskogo Instituta, Akademiya
Nauk SSSR, 216, 209 pp. [in Russian].

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.

Dubatolov, V.N., Chekovich, V.D. and Yanet, F.E.
(1968). Tabulyaty pogranichnikh sloev silura i
devona Altae-Sayanskoy gornoy oblasti i Urala.
(Tabulata of the boundary beds of the Silurian
and Devonian in the Altay-Sayan mountain
region and Urals). In “Korally pogranichnykh
sloev silura i devona Altae-Sayanskoy gornoy
oblasti i Urala’ (Ed. A.B. Ivanoskiy), 5-109.
(Nauka, Moscow). [in Russian].

Etheridge, R., Jr. (1921). Palaeontologia Novae
Cambriae Meridionalis — Occasional
descriptions of New South Wales fossils
—No.8. Records of the Geological Survey of
New South Wales 10(1), 1-11.

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.

Fromentel, E. de (1861). ‘Introduction a |’étude des
polypiers fossils.’ (F.Savy, Paris).

Hill, D. (1981). “Treatise on Invertebrate
Paleontology, Part F Coelenterata supplement
1, Rugosa and Tabulata, 2’. (Geological Society
of America, Boulder and The University of
Kansas, Lawrence).

Jell, J.S. and Hill, D. (1969). The Devonian coral
fauna of the Point Hibbs Limestone, Tasmania.
Papers and Proceedings of the Royal Society of
Tasmania 104, 1-15.

Michelin, J.L.H., (1846) ‘“Iconographie
Zoophytologique, description par localités et
terrains des polypiers fossils de France et pays
environnants’, 185-248. (P. Bertrand, Paris).

Milne-Edwards, H. and Haime, J. (1849). Mémoire
sur les polypiers appurtenant aux groupes
naturels des Zoanthaires perforés et des
Zoanthaires tabules. Comptes Rendus 29, 257-
263. Académie des Sciences, Paris.

Nudds, J.R. and Sepkoski Jr, J.J. (1993).
Coelenterata. In, ‘The Fossil Record 2’,

(Ed. M.J. Benton) p. 115. (Chapman & Hall,
London).

Pickett J.W. and Ingpen, I.A. (1990). Ordovician and
Silurian strata south of Trundle, New South
Wales. Geological Survey of New South Wales,
Quarterly Note 78, 1-14.

Proc. Linn. Soc. N.S.W., 128, 2007

Sherwin, L. (1996). Narromine 1:250,000
Geological Sheet SI/55-3: Explanatory Notes
104 pp. Geological Survey of New South
Wales, Sydney.

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. In,
Contributions to the First Australian Conodont
Symposium (AUSCOS) held in Sydney
Australia, 18-21 July 1995. (Eds. R. Mawson
and J.A. Talent). Courier Forschungsinstitut
Senckenberg 182, 79-216.

Sokolov, B.S. (1947). Novye Tabulata ordovi
Grenlandii. (New Ordovician Tabulata from
Greenland). Akademiya Nauk SSSR, Doklady
58(3), 467-472.

Talent J.A., Gratsianova, R.T. and Yolkin, E.A.
(2001). Latest Silurian (Pridoli) to middle
Devonian (Givetian) of the Asio-Australia
hemisphere: rationalization of brachiopod taxa
and faunal lists: stratigraphic correlation chart.
Courier Forschungsinstitut Senckenberg 236,
1-221.

2D

7,

FIRST ne ALES” |

AH pt 2 ei ornare
— nat Caer 2
tobe ss . a bn ein Rep.
oe vars 48 aos eine
th oF Ve
acwhie rp pare: if au’ fits eae

fant ody to aria’ oi) mot aangbonoo
inva ote 2% Weel bec Crise aga |? =
See ee ec eee) ae
ements nee PWS ot) ot contains ee) vr
Danae ct tod (2020 ApentigogmyZs ovritix.
ani abe) ORL che LM cattniA ison and
esahignec saith ncece A sepbaains Dd davabecdl aban eoranive
| aes SH WE when:
Te geno ile lindet eA We 4 AN +e) fi 4 ony
gut aludel najrivobr) wav) utbra ar) :
eae x We aA, ere “cesaahh Pehistaant) forte pi
ae ie ae: Te
ASTOmMISY Beni TR cA pater!
UW thier rohtety ew wh] aged {Rotperoie ww
/wilmuer Annie A delt berthatiowtiy netmovat}i(ertect (ny
) deen Legerainend 19 mes dteocee pthodigpsenatcs! jain)!
nado nolsleney acigagiele reeilanve Dame (Pie S
» OES eeltorkeend, Woieregdoci were)
0-]

ahaula

cere

rm

0 05 1

Coratite Diameter mm.

Ficure 3, Seatter plat of horizontal inbolae tains s mcngatine Way Ohta She}
mypolaensis,-T, ehuncidus, T. isfaraenels nen L tcherepchies. W¥ete for Fee rR

‘ Jus were obtained frota specimens. pees la AL ail
were ob ulin ¢ from peblished mepsurements an

222

aT UG eoinctDOS Bob elstiea”

1 ,

ns 4 _ Re . —

Cmax ot 1m .
aired {S008} gfe ye
Oye ar89y qroilticn ODD of 008 mut valgemong
\p lorsol .waiirat peoNyereeephiintambet 5
DeaRTED , me rekon esses
Hayton phar tlre She
ach arly spades se aula trench
phar a hese seat EERE BR
neeenyy ee
ond UNMAFENS,

be ASA

Minami ie a cinvaiitl (wos2oM ous
- eavot nigolowossled (1001) 2b, As

larbianaa cma aim)
eheeot este ak snte Ns 72 :

\o crew? inaiveionD otto eye =
ii “4 ft oe . ia ;

ward se ei ‘ikienaG

raninatqgae yesh Rete ani rachne
01-18 82 of
_ tsb abura'l & noneubotial’ 1081) ob. Aoieamort

nm snndsnaypl ao seitesiT” (12@1) Qi
_insmolqque eterotaafoo) FT net ogolotnoateT
yrsi202 Isvigolow) “f .stipludeT bas eeoguel | faa
Yo Ulerovia!s afTibas em itoraacs

teroo neimirvotl ov

(ea om 24 sab
sinderesT jsnoteon T addi in

Yo visined Veo wis ey ARETE

eitlid
gg -----aitlgesagnost” (ORSTY A. eae >
8 zation ton, nolyvorah swpigolonaqoos |
wa Hb sonar’ ob alt 1 acs Ste ah antsy
rine bossa % 8C-28!;"aemotee,

iaH bne .A zbrawbd-onliM ~~
PUI ART LIOG E91 THR eo

stout (O08 ee ae
esquotg Kon I
C29 BS etre mnqive’) actudat estiadingS :
2ivel apansind eob- sietrabsaA, aS aes
(E201) LU al ideodge? bas AL. S's

S inoos fideo oT” <n] atersinoiso)

e

Be ly

Early Ordovician Lingulate Brachiopods from New South
Wales

IAN G. PERCIVAL! AND MICHAEL J. ENGELBRETSEN7

‘Geological Survey of New South Wales, Department of Primary Industries, 947-953 Londonderry Road,

Londonderry, NSW 2753, Australia (ian.percival@dpi.nsw.gov.au); and

*Centre for Ecostratigraphy and Palaeobiology, Department of Earth and Planetary Sciences, Macquarie

University 2109. NSW. Australia.

Percival, I.G. and Engelbretsen, M.J. 2007. Early Ordovician Lingulate Brachiopods from New South
Wales. Proceedings of the Linnean Society of New South Wales 128, 223-241.

Lingulate brachiopods from the Lower Ordovician (lower Oepikodus evae conodont zone) Rowena
Formation in the far west of New South Wales are revised, and determined as Hyperobolus mootwingeensis
(Fletcher, 1964), Lingulobolus gnaltaensis (Fletcher, 1964), and the new genus Rowenaglossa, with
type species R. brunnschweileri (Fletcher, 1964). Specimens possibly conspecific with Lingulobolus
gnaltaensis are illustrated from Pine Gap, near Alice Springs, Northern Territory. In central western NSW,
Early Ordovician (Lancefieldian-Bendigonian) brachiopods are represented in the Yarrimbah Formation
of the Parkes region by the new species Palaeoglossa yarrimbahensis, a probable zhanatellid and an
indeterminate acrotretide. Allochthonous limestones in the Hensleigh Siltstone, south of Wellington,
of slightly younger (Bendigonian) age, yield Otariconulus sp. cf. O. intermedia and an unnamed new
ephippelasmatid. Although broadly contemporaneous, the lingulide brachiopods documented in this paper
lived in contrasting environmental settings. Those from the Koonenberry Belt in the far west inhabited
nearshore predominantly sandy substrates, whereas faunas from central western NSW lived in deeper water

outer shelf and slope to basinal environments flanking the Macquarie Arc.

Manuscript received 16 November 2006, accepted for publication 15 January 2007.

KEYWORDS: acrotretide, brachiopod, Koonenberry Belt, lmgulide, Macquarie Arc, Ordovician

INTRODUCTION

Taxonomic work on Early Ordovician lingulate
brachiopods from New South Wales has previously
been restricted to description of three lingulide
species from siliciclastic Lower Ordovician strata of
the Gnalta Shelf in the far west of the state (Fletcher
1964). Advances in our understanding of Ordovician
lingulides over the ensuing four decades, due chiefly
to a proliferation of taxonomic research on the part
of a small number of palaeontologists in Europe and
Russia specializing in this group, allows revision
of these species, which occur in the lower Rowena
Formation within the Mutawintji National Park,
northeast of Broken Hill (Fig. 1). Early Ordovician
lingulates are rarely encountered elsewhere in the
state, with the exception of the Yarrimbah Formation
near Parkes that is dominated by Palaeoglossa
yarrimbahensis sp. nov. Acrotretides are represented
by just a handful of specimens from this unit and
the slightly younger Hensleigh Siltstone, south of

Wellington (Fig. 1). Here we describe and illustrate
all known Early Ordovician lingulide and acrotretide
brachiopods from New South Wales. As such
brachiopods are becoming increasingly significant in
global biogeographic analysis, their documentation
is crucial even if based on limited material. This is
especially true for Australia, where contemporaneous
lingulate brachiopods remain very poorly known
(Percival 2000), with the exception of a largely
endemic fauna recently described from the Emanuel
Formation of the Canning Basin, Western Australia
(Brock and Holmer 2004). None of the latter species
is represented in the faunas documented herein.

STRATIGRAPHIC SETTING

Central Koonenberry Belt (Mootwingee - Mount
Wright area), far western NSW

The latest Cambrian to Early Ordovician
Mutawintji Group [previously the Mootwingee

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

30°00"

al
fult Arrowsmith

Fowlers
Gap

uth Wales!

4
5.

LEGEND

Main road
Minor road
Railway

River

Sample locality

Property name

Figure 1. Locality map showing sites (indicated by stars) in New South Wales yielding Early Ordovi-

cian lingulate brachiopods described in this paper.

Group; the change in spelling was requested by the
indigenous Mutawintji people to reflect its correct
pronunciation in the Parrkantyi language; T. Sharp,
pers. comm. 2003] includes (in ascending order) the
Nootumbulla Sandstone, Bynguano Quartzite and
Rowena Formation (Fig. 2). All three units consist
predominantly of coarse siliciclastic sediments
deposited in marginal marine to shallow shelf

224

conditions. The Rowena Formation also comprises
thin calcareous siltstone beds in the lower part of the
unit. From one such horizon, Zhen and Percival (2006)
reported a small but diverse conodont fauna dominated
by Erraticodon patu, a species characteristic of the
Early Ordovician evae Zone in the Tabita Formation
at Mount Arrowsmith in the northern Koonenberry
Belt (Zhen et al. 2003). Fletcher (1964) described

Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

KOONENBERRY “YARRIMBAH” BAKERS SWAMP
BELT NW OF PARKES SOUTH OF
CONODONT WESTERN NSW | CENTRAL NSW | WELLINGTON
ZONATION Rowena
Formation
; Hyperobolus
Reutterodus Oepikodus mootwingeensis,
andinus evae Lingulobolus
= gnaltaensis,
< < Rowenaglossa
(=) Py brunnschweilen
> = Hyperobolus ahs eig
o = mootwingeensis Siltstone
=] = Oepikodus | Prioniodus ¢ Olariconulus cf
a iu | communis | elegans I Palacogiossa O. intermedia
oO a yarrimbahensis
Yarrimbah Mitchell
>
audi Formation Formation
% Acodus deltatus - @ indet. acrotretide
= Oneotodus costatus ynguano

Macerodus dianae

LANCEFIELDIAN

Rossodus manitovensis

Cordylodus angulatus -
Chosonodina herfurthi
Cordylodus findstromi

WAREND
-IAN

Cordylodus prolindstromi
- Hirsudodontus simplex

Cordylodus proavus

L. CAMBRIAN
FURONGIAN

ae DATSONIAN

Quartzite

Nootumbulla
Sandstone

Nelungaloo
Volcanics

Figure 2. Stratigraphic levels at which Early Ordovician lingulate brachiopods occur in New South

Wales.

Obolus mootwingeensis, Lingulella (Leptembolon)
gnaltaensis, and Ectenoglossa brunnschweileri from
outcrops of the Rowena Formation (then known as
the upper part of the Gnalta stage of the Mootwingee
Series) in the vicinity of the disused Mootwingee-
White Cliffs mail coach road. The exact localities
from which the specimens described by Fletcher
were obtained are not known (they are noted in the
Australian Museum register as “8 miles along the
coach road from Mootwingee”). The most recent
geological map of Mutawintji National Park (Sharp
2004) shows that the Rowena Formation intersects
the old coach road along strike over a distance of
approximately 10 km. Most likely the type localities
are centred on GR 654505 mE 6270000 mN (Nuchea
7335 1:100,000 mapsheet, GDA94 coordinates) as
there are abundantly fossiliferous outcrops in the

Proc. Linn. Soc. N.S.W., 128, 2007

immediate vicinity. Since Fletcher’s pioneering
work no further research has been conducted into
these brachiopods. Revision of the type and topotype
material reveals new information about the internal
features of these brachiopods and allows reassessment
of their taxonomic status. All three taxa are reassigned
at genus level to Hyperobolus mootwingeensis,
Lingulobolus gnaltaensis, and the new genus
Rowenaglossa, with type species R. brunnschweileri.
Specimens of H. mootwingeensis and L. gnaltaensis
(collected in the mid-1960s, and curated in the
Geological Survey of NSW palaeontological
collection) also occur sporadically in the Bynguano
Quartzite at a locality three miles (approximately five
km) west of “Bilpa” homestead, ten miles (16 km)
south of Little Topar, in the southern extremity of the
Koonenberry Belt (Fig. 1).

DDS

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

Macquarie Arc, Lachlan Orogen, central NSW

The Yarrimbah Formation, exposed on
“Yarrimbah” property, 16 km west of Parkes (Fig.
1), has most recently been defined by Sherwin
(2000). Graptolites identified by Sherwin (1979,
1990; pers. comm. 1999) from laminated siliceous
mudstones forming the upper beds of this unit include
Didymograptus (Cymatograptus) sp., Tetragraptus
approximatus and Pendograptus fruticosus. These
indicate an age, in terms of the Victorian graptolite
zonation, of late Lancefieldian to early Bendigonian
(equivalent to the approximatus to fruticosus zones of
the Australasian and Chinese successions). Lingulide
brachiopods including prolific Palaeoglossa
yarrimbahensis sp. nov., and a possible zhanatellide
(the latter exceptionally rare) are the only other
fossils found in the upper Yarrimbah Formation.
One specimen of an indeterminate acrotretide with
a distinctive dorsal platform was recovered from
allochthonous limestone clasts that occur sporadically
in the lower part of the formation.

From the Hensleigh Siltstone in the Bakers
Swamp area, 26 km south of Wellington (Fig. 1B),
acrotretide brachiopods are represented by only
a couple of specimens. Two taxa, Otariconulus
sp. cf. O. intermedia and ephippelasmatidae gen.
et sp. nov. (insufficient material is available to
permit formal naming), are described herein from
allochthonous limestones in the lower part of this
formation. The age of the Hensleigh Siltstone, based
on graptolites, is middle to late Bendigonian (Be 2-3)
(Percival et al. 2001). Conodonts obtained from the
allochthonous limestones indicate an age equivalent
to the upper Prioniodus elegans conodont Biozone,
contemporaneous with the Oepikodus communis
conodont Biozone (Zhen et al. 2004).

COMPARATIVE PALAEOECOLOGY

Lower Ordovician rocks of the Mutwintji Group
were deposited in a nearshore shallow marine setting
on the Delamerian margin of Gondwana, as indicated
by trace fossils attributed to trilobites (Webby 1983)
and a diverse infauna (Droser et al. 1994) in the
Bynguano Formation (immediately underlying the
Rowena Formation). The lingulide brachiopods
redescribed herein from the lower Rowena Formation
present a further opportunity for palaeoecological
analysis, based on comparisons with studies of
similar faunas in contemporaneous clastic rocks of
central Europe.

In the Prague Basin (Czech Republic), the
Hyperobolus Community is interpreted as having
inhabited the most shallow water nearshore setting
(Havlicek 1982a), where Hyperobolus feistmantelli

226

(associated with other large and moderate to thick-
walled lingulates) lived in a semi-endobenthic habit,
i.e. being only partly or shallowly buried in an
unconsolidated sandy sea floor (Mergl 2002). Not
surprisingly, considering its close morphological
similarity, the Gnalta Shelf species H. mootwingeensis
seems to have occupied an identical ecological
niche.

Likewise, the presence of Lingulobolus and
Rowenaglossa in the Rowena Formation recalls the
co-occurrence of Lingulobolus? and the externally
homeomorphic Ectenoglossa in the Armorican
Sandstone of Brittany and the Montagne Noire
regions of France, in similar low diversity faunas
entirely dominated by large lingulate brachiopods
(Cocks and Lockley 1981).

However, there are subtle differences in the
distribution of these lingulates in the Rowena
Formation, compared with European nearshore
lingulate-dominated faunas. Firstly, the faunal
diversity of the Rowena Formation is somewhat lower
(three lingulate genera) than formations in the Prague
Basin (five to six genera in the Trenice Formation).
Secondly, the Rowena Formation lingulates are not
associated in the one horizon; rather, they occur as
monospecific shell beds, with Hyperobolus and
Lingulobolus confined to separate sandy quartzose
sediments, whereas Rowenaglossa generally occurs
in slightly finer lithofacies (silts rather than coarse
sands). Such variance in facies preference may
present an opportunity to more precisely define depth-
related communities in order to study transgressive-
regressive relationships in the Mutawinjti Group.

In contrast, Palaeoglossa in the Yarrimbah
Formation inhabited deep-water substrates. Although
no graptolites are found in direct association with
the brachiopod-bearing laminated siltstones, unlike
comparable occurrences in the Upper Ordovician
Malongulli and Gunningbland Formations (Percival
1978), graptolites are plentiful in directly underlying
beds, implying that water depth throughout deposition
of the Yarrimbah Formation was considerable.
Sponges, indicated by abundant spicules preserved
in the Malongulli Formation, may have provided a
substrate above the sediment-water interface for
some or all of the lingulate brachiopods found in that
unit; the apparent absence of sponge remains in the
Yarrimbah Formation suggests that P. yarrimbahensis
lay sessile on the sea floor or else may have lived
partly buried. Acrotretide brachiopods found in
allochthonous limestone clasts in both the Yarrimbah
and Hensleigh formations are interpreted as living
on outer carbonate shelves flanking volcanic islands
at relatively shallow depths, prior to displacement
downslope.

Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

SYSTEMATIC PALAEONTOLOGY

Type material, comprising specimens described
and illustrated or listed herein, is curated in the
palaeontological collections of the Geological
Survey of New South Wales (designated MMMC for
microfossil specimens, and MMF for macrofossils),
or in the type fossil collection of the Australian
Museum, Sydney (AM F). Responsibility for
taxonomic description and authorship of new taxa is
attributable to I.G. Percival for Lingulida, and M.J.
Engelbretsen for Acrotretida. For brevity, authorship
of taxonomic hierarchy above genus level is not
cited in the References; these bibliographic sources
are listed in the revised (2™ edition) Treatise of
Invertebrate Paleontology, Part H: Brachiopoda
(Williams et al. 2000).

Phylum Brachiopoda Duméril, 1806
Subphylum Linguliformea Williams, Carlson,
Brunton, Holmer & Popov, 1997
Class Lingulata Gorjansky & Popov, 1985
Order Lingulida Waagen, 1885
Superfamily Linguloidea Mencke, 1828
Family Obolidae King, 1846
Subfamily Obolinae King, 1846

Palaeoglossa Cockerell, 1911
Type species: Lingula attenuata J. de C. Sowerby,
1839

Palaeoglossa yarrimbahensis sp. nov.
Fig. 3A—M, 4D—N

Material

Holotype MMF 27463c; paratypes MMF
27621a, MMF 30329a, MMF 44823 — 44832
inclusive.

Diagnosis

A species of Palaeoglossa with a very short
ventral pseudointerarea; internally with scattered
pustules and fine pitting posteriorly; very fine radial
striations crowded around interior periphery of valves;
umbonal muscle scars confined to immediate vicinity
of pedicle groove; dorsal valve lacking median ridge
internally.

Description

Ventral valve elongately ovoid in outline with
broadly acute posterior end; dorsal valve tending
toward subcircular outline with rounded posterior;
lateral margins of both valves very gently curved

Proc. Linn. Soc. N.S.W., 128, 2007

to subparallel; maximum width in both valves
approximately coincident with midlength; shell
profile lenticular, weakly biconvex. Length:width
ratio ranges from 1.28-1.61 for ventral valves (average
1.33) and ranges from 1.08-1.22 (average 1.15)
for dorsal valves. Shell material moderately thin;
ornament confined to closely spaced hemiperipheral
growth lines (Fig. 4D), occasionally with accentuated
growth discontinuities (Fig. 3L), but lacking any
radial striations externally.

Ventral pseudointerarea reduced to pair of small
short triangular propareas flanking open pedicle
groove (Fig. 4G); flexure lines absent. Muscles
generally not deeply inserted, although paired
umbonal muscle scars are occasionally prominent in
the immediate vicinity of the pedicle groove (Figs.
3C, D, K, J), and paired anterior adductor scars with
relatively well-defined inner margins appear to be
present medially in one specimen (Fig. 3M). Isolated
small irregular pustules scattered on interior of valve
(Fig. 3C). Pallial canals not impressed.

Dorsal pseudointerarea extremely short but
apparently complete and undivided (Fig. 3)).
Although muscle scars are rarely impressed, an
umbonal scar (possibly paired: Fig. 4N) and some ill-
defined visceral markings may be visible posteriorly
(e.g. Figs 3H, I). Scattered pustules occasionally
present (Fig. 3L). One well-preserved internal mould
(Fig. 4M) shows evidence of very fine scattered pits
in visceral area. Pallial canals unknown. Margin of
dorsal valve displays very fine radial striations in
better-preserved specimens (Fig. 31).

Dimensions ;

Holotype MMF 27463c length 11.0 mm, width
8.2 mm. Smallest specimen is a subcircular juvenile
dorsal? valve (MMF 44832) 3.4 mm in length and
3.1 mm in width; subsequent hemiperipheral growth
accentuates elongation in adults, especially in ventral
valves, whereas dorsal valves may retain a subcircular
outline even in large adult specimens (e.g. Figs 4E,
F). Ventral valve length ranges from 5.1-11.1 mm,
dorsal valve length ranges from 7.9-12.7 mm; ventral
valve width ranges from 4.1-8.7 mm and dorsal valve
width ranges from 6.5-11.7 mm.

Discussion

The type species of Palaeoglossa, P. attenuata (J.
de C. Sowerby, 1839), from the Middle Ordovician of
the Shelve area, Shropshire, England (most recently
redescribed by Sutton et al. 1999), lacks flexure
lines in both pseudointerareas, and has internal
features such as muscle scars and pallial canals very
weakly impressed, with only sporadic pitting weakly

229,

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

Figure 3. Palaeoglossa yarrimbahensis sp. nov. from the upper Yarrimbah Formation, west of Parkes,
central west NSW. A — D: Two valves presumed to represent the one individual, MMF 44830; A, B, part
and counterpart and C, latex replica taken from A; A: internal mould of dorsal valve (on left) and ven-
tral valve (on right); B: partly exfoliated ventral valve (on left) and exterior of dorsal valve (on right); C:
latex replica clearly shows nearly circular juvenile shell on dorsal valve (right) and scattered pustules on
interior of ventral valve (left); note also deeply inserted umbonal muscle scars in this valve; D: Enlarge-
ment of posterior of ventral valve internal mould. E: ventral valve internal mould, MMF 44831.

F. ventral valve internal mould, MMF 44823. G. ventral valve internal mould, MMF 30329a, slightly dis-
torted. H — I: dorsal valve internal mould and latex replica, MMF 44827. J — K: Holotype, ventral valve
internal mould, MMF 27463c, J is an enlargement of posterior to show pedicle groove. L — M: dorsal
and ventral valves, possibly of the one individual, both are latex replicas taken from MME 44824; both
valves incomplete, with ventral valve vertically oriented in both photographs. Scale bar above H applies
to all specimens except D, F and J.

228 Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

Figure 4. Lingulate brachiopods from the upper Yarrimbah Formation, west of Parkes, central west
NSW. A — C: zhanatellide indet., internal mould (C) and latex replicas (B is enlargement of posterior
region) of ventral valve MMF 44843. D — N: Palaeoglossa yarrimbahensis sp. nov. D: ventral? valve exte-
rior view, displaying ornament of growth lines only, MMF 44828; E: dorsal valve internal mould, MMF
44829. F: dorsal valve internal mould with adherent shell, MMF 27621a. G —H. ventral valve internal
mould (H) and enlargement of posterior region (latex replica) to show pedicle groove, MMF 44825. I:
juvenile dorsal? valve, partly exfoliated internal mould, MMF 44832. J —N: dorsal valve internal mould
(L), and two latex replicas, (J) taken from that specimen and (K) from counterpart external mould,
(M) and (N) are enlargements of posterior regions of (K) and (L) respectively, MMF 44826. Scale bar
between F and J applies to all specimens except B, G, M and N.

developed internally in both valves. In all these
characteristics it resembles the Yarrimbah species.
The main features distinguishing P attenuata and
P. yarrimbahensis concern the complete absence in
the latter of any median ridge in the dorsal valve
(a weak median septum is occasionally developed
in P. attenuata), and the greatly reduced length of

Proc. Linn. Soc. N.S.W., 128, 2007

the ventral pseudointerarea in P yarrimbahensis,
resulting in a much less acuminate outline of the
ventral valve beak.

Lingulella, which appears to be closely related
to Palaeoglossa according to Sutton et al. (1999),
was also considered as a possible generic assignment.

229

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

Sutton et al. (2000) redescribed the type species of
Lingulella, L. davisii McCoy, 1851 from the Upper
Cambrian of North Wales, noting its diagnostic
characters as ventral propareas with well-defined
flexure lines, conspicuous deep pitting on interiors of
both valves, anterior lateral and central muscle scars
set close together, and vascula lateralia subperipheral
to peripheral, not sharply divergent proximally. Fine
sporadic pitting on the interior of P. yarrimbahensis
is neither as strongly developed nor as dense as that
which characterizes L. davisii. Furthermore, features
exhibited by the Yarrimbah species including very
reduced pseudointerareas lacking flexure lines, and
absence of any anterior extension of the dorsal visceral
field as a ridge or tongue, are atypical of Lingulella.

Distribution

Outcrop of upper Yarrimbah Formation adjacent
to “Yarrimbah” and “Wilga East” property boundary,
16 km west of Parkes (Fig. 1C); late Lancefieldian
(La3) to early Bendigonian (Bel).

Lingulobolus Matthew, 1895
Type species: Lingulella? affinis Billings, 1872

Lingulobolus gnaltaensis (Fletcher, 1964)
Fig. 5A—M

Synonymy
1964 Lingulella (Leptembolon) gnaltaensis
Fletcher, p. 287, pl. 31, figs 7, 9.

Figure 5. Lingulobolus gnaltaensis (Fletcher, 1964), from the lower Rowena Formation, Koonenberry
Belt, far western NSW. A: exterior of ventral? valve, holotype AM F49383; B — C: dorsal valve, internal
mould and latex replica, AM F50411; D — F: ventral valve, internal mould and latex replica, enlarged
in F to more clearly show muscle field, AM F50415; G: exterior of dorsal valve (latex replica from
external mould), AM F50401; H — I: interior (latex replica) and internal mould of dorsal valve, AM
F47492; J: exterior of ventral? valve (latex replica from external mould), AM F49384; K — M: latex
replica of dorsal internal mould shown in M, with enlargement of posterior half of valve shown in L,
AM F132242. Scale bar beneath D and E applies to all specimens except F and L.

230

Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

Material

Holotype AM F49383 and paratype AM F49384
(not 49344, a typographical error in Fletcher 1964,
p. 287), both internal moulds of ventral valves
(designated by Fletcher 1964); additional paratypes
selected herein are AM F47492, AM F50401, AM
F50411, AM F50415 and AM F132242 (illustrated in
Fig. 5).

Description

Shell broadly triangular in outline with
subacuminate beak in ventral valve, and with
maximum width attained about three-quarters length
from beak; length almost always greater than width
in ventral valves, but equal to or somewhat less than
width in dorsal valves; moderately to strongly biconvex
especially anteriorly; occasionally with fine growth
lamellae and stronger concentric rugae developed
anteriorly, most prominent in larger specimens (Fig.
5 A, C); no radial ornament. Ventral pseudointerarea
possibly undivided, very much reduced in length (Fig.
5 L, M). Ventral valve muscle field [well-displayed
on specimen AM F50415 (Fig. 5 F)] is confined to
posterior half of valve where it consists of a median
parallel track of anterior adductor scars(?) flanked by
pair of shorter but wider impressions (oblique muscle
scars?) that slightly diverge, and are bounded laterally
by traces of narrow linear pallial canals; fine closely-
spaced radial striae are impressed around anterior
margin of this specimen. Dorsal valve interior (Fig. 5
H) is essentially smooth and lacks impressed muscle
scars and vascular markings; pseudointerarea possibly
entire but very short and poorly preserved.

Dimensions

Holotype AM F49383: length 16.5 mm, width
14.1 mm. For other complete specimens, ventral
valve length ranges from 13.3-17.3 mm, dorsal valve
length ranges from 7.9-12.7 mm; ventral valve width
ranges from 11.8-17.6 mm and dorsal valve width
ranges from 11.3-12.5 mm.

Discussion

Although Fletcher (1964) stated in his description
of Lingulella (Leptembolon) gnaltaensis that only
brachial (i.e. dorsal) valves were figured, this was
corrected to pedicle valves in the caption to his
plate 31. Leptembolon Mickwitz, 1896, is close to
Lingulobolus as regards shell outline and profile, but
internally the two genera are readily distinguished
by the greatly thickened ventral visceral area of the
former and its slightly thickened dorsal visceral area
that extends as a low ridge nearly to the anterior valve
margin. As none of these features are apparent in

Proc. Linn. Soc. N.S.W., 128, 2007

Fletcher’s species its reassignment to Lingulobolus is
therefore justified.

The diagnosis given for Lingulobolus by Holmer
and Popov (2000) mentions a narrow subtriangular
ventral pseudointerarea divided by a broadly triangular
pedicle groove, with a vestigial undivided dorsal
pseudointerarea, and a weakly defined dorsal visceral
area extending beyond midlength with closely spaced
anterior-lateral and central muscle scars. The genus
is represented by L. affinis (Billings, 1872) and L.
spissus (Billings. 1872), both from Lower Ordovician
strata in Newfoundland, and possibly another two
species of Early Ordovician age, L.? hawkei (Rouault,
1850), found in southern Britain and France (Brittany)
(Cocks and Lockley 1981), and L.? brimonti (Rouault,
1850), co-occurring with L.? hawkei in these regions
as well as in the Montagne Noire of France (Havli¢éek
1980) and in Algeria, North Africa (Legrand 1971).
Lingulobolus gnaltaensis shows clear affinities with
the Newfoundland species L. spissus, most recently
illustrated by Holmer and Popov (2000, p. 47), in
shell outline, profile, rhomboidal shape of the ventral
visceral area, and disposition of fine radial striae in a
peripheral band at the anterior margin of the ventral
valve. However, in having a much reduced ventral
pseudointerarea without a distinct pedicle groove
(although this may be due to adverse preservation in
coarse clastic sediment) it differs from the Canadian
species. The external coarse rugae developed on L.
gnaltaensis, and lack of any exterior radial ornament,
are another distinguishing feature (compare with
detailed line drawings of both Newfoundland species
given by Walcott 1912, pl. XVI). Generally poor
preservation of L.? hawkei and L.? brimonti does not
permit meaningful comparisons with L. gnaltaensis.

Distribution

Lower Rowena Formation, “about 8 miles (13
km) from ‘Mootwingee’ homestead along old coach
road” according to Australian Museum register, just
east of Split Rock in Mutawintji National Park, far-
western NSW (Fig. 1A). Also occurs sporadically in
the Bynguano Quartzite in the southern Koonenberry
Belt, 16 km south of Little Topar (Fig. 1A) (MMF
15013, 15014, 15017, 15019, 15021).

Lingulobolus cf. L. gnaltaensis (Fletcher, 1964)
Fig.6 A—E

Discussion

Fletcher (1964) noted that his Lingulella
(Leptembolon) gnaltaensis was possibly congeneric
with specimens from the Pacoota Sandstone at a

y233I\

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

10mm

DBD

Figure 6. Lingulobolus cf. L. gnaltaensis (Fletcher,
1964), from the Pacoota Sandstone, Pine Gap
near Alice Springs, Northern Territory. All speci-
mens are located on one slab bearing the Austral-
ian Museum number AM F49586. A: latex replica
of interior of dorsal valve, AM F49586g; B: latex
replica of exterior of dorsal valve, AM F49586f;
C: latex replica of exterior of ventral valve, AM
F49586e; D: two ventral valve internal moulds,
one partly exfoliated, AM F49586a, b; E: dorsal
(left) and ventral (right) valve internal moulds,
possibly from the one individual, AM F49586c, d.

locality (Pine Gap), about 19 km southwest of Alice
Springs, Northern Territory. He neither described nor
illustrated these specimens, and to our knowledge they
have not subsequently been documented. Material
from the Pine Gap locality (now a restricted military
site) that was available to Fletcher was located in the
Australian Museum collection, and representative
specimens are here illustrated in Figure 6. Little
can be said about these shells other than their gross
morphology, as the coarse sandstone matrix has
prevented preservation of subtle internal features.
They seem to be slightly larger than Lingulobolus
gnaltaensis, but otherwise could well be conspecific.
Shell outline and profile appears near identical for
both the Rowena Formation and Pine Gap forms.
In the absence of internal features we tentatively
compare the latter with L. gnaltaensis.

Dimensions

AM F49586a (ventral valve) length 20 mm,
width 17.5 mm; AM F49586b (dorsal valve) length
16.4 mm, width 16.6 mm; AM F49586c (ventral
valve) length 18.8 mm, width 15.3 mm; AM F49586d
(ventral valve) length 16.7 mm, width 16.2 mm; AM
F49586e (dorsal valve) length 15.5 mm, width 15.3
mm.

Family Pseudolingulidae Holmer, 1991

Rowenaglossa gen. nov.
Type species: Ectenoglossa brunnschweileri
Fletcher, 1964

Diagnosis

Outline spatulate to elongate rectangular with
subparallel lateral margins; ventral pseudointerarea
extending about one-third valve length, divided by
narrow but deep pedicle groove; median septum
prominent in dorsal valve, much reduced or absent in

Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

ventral valve; muscle scars and pallial canals weakly
impressed.

Rowenaglossa brunnschweileri (Fletcher, 1964)
Fig. 7 A—P

Synonymy
1964 Ectenoglossa brunnschweileri Fletcher, p.
288, pl. 31, figs 4-6; pl. 32, fig 15.

Material

Holotype AM F49027 ventral valve, and
paratypes AM F48995 (ventral valve) and AM F49014
(dorsal valve), designated by Fletcher (1964); an
additional specimen AM F49044 was illustrated by
Fletcher but not included in the original type series.
It, together with specimens AM F48975, AM F49001,
AM F49011, AM F49034, AM F49050 and AM
F49843 are here designated as additional paratypes
(illustrated in Fig. 7).

Description

Shell outline spatulate, elongately acutely
triangular to subrectangular; ventral beak narrowly
acute, dorsal beak slightly less so; maximum width
reached between approximately two-fifths and half
valve length, after which lateral margins remain
subparallel or very slightly converge towards
anterolateral extremities; length to width ratio 1.6-
2.28 (ventral valves), 1.6-2.34 (dorsal valves); shell
profile weakly biconvex with maximum curvature
restricted to lateral periphery of valves. Ornament of
growth lines only (Fig. 7 E). Shell material moderately
thick.

Ventral pseudointerarea narrowly triangular,
extending to nearly one-third valve length (Fig. 7 A,
B), separated by narrow pedicle groove (Fig. 7 J, L)
that extends for several mm anteriorly, flexure lines
apparently lacking; in one specimen a short slightly
raised median ridge (Fig. 7 F, G) extends from front of
weakly impressed umbonal scar; median ridge absent
to very weakly developed medially; the anterior part
of this ridge may be flanked by pair of low furrows
that further accentuate it (Fig. 7 A, B); very weakly
impressed central muscle scars may be present
medially on either side of median ridge (Fig. 7 A) but
other muscle scars and pallial canals not observed.

Dorsal pseudointerarea short, very narrow,
entire but with very slight median depression to
accommodate pedicle (Fig. 7 A, C, D, H); paired large
median muscle scars (centrals) with subrectangular
outline are weakly impressed between about
midlength and two-thirds valve length (Fig. 7 H, J),
bisected by a thin but prominent median septum that

Proc. Linn. Soc. N.S.W., 128, 2007

commences in posterior third of some dorsal valves
and extends almost to anterior margin (Fig. 7 C, D, H,
I, M, N); other muscle scars not impressed; vascula
lateralia rarely visible (Fig. 7 1), slightly divergent
and mostly straight.

Dimensions

Holotype AM F49027 (ventral valve): length
27.2 mm, width 12.3 mm. For other complete
specimens, ventral valve length ranges from 15.2-
24.5 mm, dorsal valve length ranges from 14.5-28.1
mm; ventral valve width ranges from 8.0-15.3 mm
and dorsal valve width ranges from 9.0-13.3 mm.

Discussion

Fletcher (1964) referred this species to
Ectenoglossa Sinclair, 1945, which has a distinctive
spatulate to subrectangular outline near identical to
that of the new genus. Ectenoglossa is, according
to Holmer and Popov (2000), poorly known,
particularly as regards details of the dorsal interior
and the disposition of vascular markings in both
valves. Within the concept of the genus these authors
include only the type species E. /esueuri (Rouault,
1850), which occurs in Lower Ordovician rocks in
Brittany and the Montagne Noire in France (Havli¢ek
1980), and in pebbles eroded from these units and
redeposited in Devon, Britain (Cocks and Lockley
1981). Other species attributed to Ectenoglossa by
Goryansky (1969), including E. Jata and E. exunguis,
more closely resemble Pseudolingula in outline and
profile; E. /atain particular displays a prominent dorsal
median septum, and Holmer (1991) suggested that
it represented an undescribed new genus with close
affinity to Pseudolingula. Cooper (1956) tentatively
referred two species, E? rubra and E? sculpta, to
Ectenoglossa based on external details only; their
affinities remain unknown. A further species, E.
nymphoidea Cooper, 1956, from Middle Ordovician
strata of Tennessee and Virginia, is one of the largest
known lingulate brachiopods, with lengths estimated
up to 75 mm. Although other internal details are
unclear, the presence of a dorsal median ridge in E.
nymphoidea (Cooper 1956, pl. 2E, fig. 16) suggests
this species may be better placed in Pseudolingula.
A poorly preserved unnamed species provisionally
attributed to Ectenoglossa from Meiklejohn Peak,
Nevada (Krause and Rowell 1975) similarly displays
a median ridge in a presumed dorsal valve, and is also
more likely to be a pseudolingulide. Two additional
species assigned to Ectenoglossa have been described
from Upper Ordovician strata; EF. sorbulakensis Popov,
1980, from the Chu-Ili Range region of Kazakhstan
(see also Popov et al. 2002), and E. minor Zhan and

233

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

Figure 7. Rowenaglossa brunnschweileri (Fletcher, 1964), from the lower Rowena Formation, Koonen-
berry Belt, far western NSW. A — B: ventral valve, holotype AM F49027, A is latex replica and B is inter-
nal mould (note that adjacent dorsal valve on right-hand side of A is presumed to be the opposing valve
of this specimen); C — D: dorsal valve AM F49050, internal mould and latex replica; E: exterior of partly
exfoliated juvenile ventral? valve AM F48975; F—G: ventral valve AM F49001, internal mould and latex
replica; H — I: dorsal valve AM F48995, internal mould and latex replica; J: ventral valve AM F49843,
latex replica from internal mould; K — L: ventral valve AM F49011, internal mould and latex replica; M
—N: dorsal valve AM F49034, internal mould and latex replica; O: ventral valve internal mould, AM F
49840; P: dorsal valve internal mould, AM F49014.

234 Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

Cocks, 1998, from the Yushan area, Jiangxi Province
of South China. However, at least the former species is
apparently excluded from the concept of Ectenoglossa
outlined by Holmer and Popov (2000), which specifies
only an Early Ordovician range. Popov et al. (2002)
reported the presence of a weak dorsal median ridge
in E. sorbulakensis. No such feature is mentioned as
occurring in E. minor by Zhan and Cocks (1998) who
illustrated only ventral valves.

The most distinctive feature of Rowenaglossa
is the median septum prominently developed in the
dorsal valve (though also present in a much reduced
form in some ventral valves). This septum is not seen
in the type species of Ectenoglossa, whereas it is
characteristic of Pseudolingula and related genera,
thereby supporting inclusion of Rowenaglossa in
the Pseudolingulidae; absence of flexure lines is
another similarity. The “teeth” in the ventral valve
beak region, referred to by Fletcher (1964), seem to
be an artifact of preservation due to a fracture in one
specimen, and do not resemble the thin ridges flanking
the pedicle groove seen in some specimens of E.
lesueuri. Internal features such as the visceral field,
muscle scars, and vascular markings are generally
weakly impressed to absent in Rowenaglossa, and
serve to readily distinguish it from Pseudolingula
which displays massive thickening of the visceral
area and prominent muscle scars. The dorsal valve
of R. brunnschweileri illustrated in Fig. 7 H, I best
shows details of the weakly impressed muscle field
(central muscle scars) and pallial canals.

Distribution

Lower Rowena Formation, “about 8 miles (13
km) from ‘Mootwingee’ homestead along old coach
road” according to Australian Museum register, just
east of Split Rock in Mutawintji National Park, far-
western NSW (Fig. 1A).

Family Zhanatellidae Koneva, 1986

Hyperobolus Havliéek, 1982b
Type species: Lingula feistmanteli Barrande, 1879

Hyperobolus mootwingeensis (Fletcher, 1964)
Fig. 8A—P

Synonymy

1964 Obolus Mootwingeesis [nom. imperf. pro
mootwingeensis| Fletcher, p. 286, pl. 31, figs 1-3, 8;
pl. 32, figs 13-14.

2004 Leptembolon? gnaltaensis Fletcher, 1964;
Sharp, photograph 7.

Proc. Linn. Soc. N.S.W., 128, 2007

Material

Holotype AM F47427 (dorsal? valve), and
paratypes AM F47422 and AM F49056, designated
by Fletcher (1964), who also illustrated (but did
not designate as types) AM F47428, 47439, 48963,
and 48974. Of these, AM F47439 is a steinkern of
the holotype specimen, and clearly therefore must
be regarded as forming part of the holotype. Also
mentioned, but neither illustrated nor designated as a
type by Fletcher (1964), is AM F47424, which is the
counterpart of paratype AM F47422. All specimens
cited above are now included in the type series,
and are supplemented by the following additional
paratypes AM F47425, AM F48985 and AM F49056
(illustrated in Fig. 8).

Description

Shell large, distinctly triangular with straight
posterolateral margins diverging at approximately
50-75 degrees from beak; maximum width attained
at about four-fifths length of ventral valve and
approximately three-quarters length of dorsal valve;
anterolateral corners well rounded; anterior margin
straight. Profile biconvex, ventral valve flattened
medially, most convex in anterior quarter; dorsal
valve more evenly convex. Shell moderately thick,
multilayered, with concentric growth lamellae flaking
anteriorly to reveal fine radial ornament and dendritic
vascular markings peripherally (Fig. 8 L-N).

Details of ventral pseudointerarea uncertain,
but likely to be very narrow and elongate, extending
about one-third length of lateral margin of valve
(poorly preserved on specimen in Fig. 8 B). Muscle
field not clearly defined; there is a suggestion of scars
developed medially, flanked by pair of narrower
slightly longer scars in posterior quarter of valve
(Fig. 8 I) but any further comment on these would
be conjecture. Pallial canals also not well impressed
except in posterolateral third of valve where they
separate into finely dendritic canals peripherally — this
is seen in several specimens e.g. Fig. 8 L-N.

Dorsal pseudointerarea entire, very short and
narrow (Fig. 8 O, P). Muscle field is poorly defined,
possibly with undivided umbonal scar and small
subcircular posterolateral (oblique muscle?) scars
(Fig. 8 O, P). Vascular markings unknown except for
probable dendritic canals posterolaterally.

Dimensions

Holotype AM F47427/47439 (dorsal? valve):
length 27.5 mm (slightly incomplete), width 21.6 mm.
For other complete specimens, ventral valve length
ranges from 16.3-32.0 mm, dorsal valve length ranges
from 13.5-17.0 mm (2 specimens only); ventral valve

235

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

Figure 8. Hyperobolus mootwingeensis (Fletcher, 1964), from the Koonenberry Belt, far western NSW.
All specimens except K (from Bynguano Quartzite on “Bilpa” in southern Koonenberry Belt) are from
the lower Rowena Formation in Mutawintji National Park. A — D: ventral valve, AM F47422 and AM
F 47423, (A) latex replica of partly exfoliated external mould, (B) latex replica of internal mould shown
in (C), (D) counterpart external of ventral valve. E — G: dorsal? valve external mould, holotype AM
F47427, and counterpart dorsal? valve internal mould AM F474339, (G) is a latex replica taken from
external mould AM F47427. H: ventral valve exterior, AM F47428. I. ventral valve internal mould, AM
F49056. J: ventral? valve exterior, AM F48985. K: ventral valve internal mould, MMF 15012. L — M:
partial valve exterior with exfoliated shell, showing detail of shell structure and dendritic canals, AM
F 48963. N: dorsal valve exterior, AM F48974. O — P: fragment of posterior of dorsal valve interior, show-
ing pseudointerarea and weakly impressed muscle scars on internal mould (O) and latex replica (P), AM
F 47425. Scale bar beneath I applies to specimens A — I and K; scale bar above L applies to specimens J,
Land O—- P; M and N have individual scale bars.

236 Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

width ranges from 14.0-29.2 mm and dorsal valve
width ranges from 13.1-13.8 mm (2 specimens).

Discussion

Fletcher (1964) compared this species to H.
feistmanteli(Barrande), at the time attributed to Obolus
(as was originally H. mootwingeensis). Apart from a
general agreement in characters (including presence of
a marginal rim), Fletcher commented on a similarity in
the appearance of marginal sinuses crossing the inner
surface of the valves near the anterior-lateral margins
in both species. He distinguished “O”. mootwingeensis
from “O”. feistmanteli by the generally larger size,
and more acuminate subtriangular outline, of the
NSW species. Fletcher also remarked on the presence
of a “posterior median sinus” which was found only
in “O”. mootwingeensis, referring in his description
of this species to “a fine median ridge” extending
from the beak in the pedicle valve. In the caption to
the illustrated internal mould (Fletcher 1964, pl. 31,
fig. 2) the same feature is termed a “median internal
sinus”. Neither is correct; the apparent fine groove is
an artefact of the lighting of the specimen, and in fact
represents a longitudinal fracture in the shell. Despite
this slight misinterpretation, Fletcher’s overall close
comparison with H. feistmanteli (Barrande, 1879),
type species of Hyperobolus Havliéek, 1982b remains
valid, and supports the reassignment of the NSW
species to that genus. Mergl (2002), who augmented
the detailed description of Havliéek (1982b) with
further illustrations of specimens from Bohemia,
figured (Pl. 20, fig. 14) a dorsal internal mould of
H. feistmanteli which is identical to Figure 8 M, N
herein.

Of the three previously known species of
Hyperobolus, the type H. feistmanteli comes from the
Upper Tremadocian Trenice Formation of Bohemia.
H. fragilis Holmer, Koneva and Popov, 1996 is known
from the Middle Ordovician Zhyrykaus Formation
of the Malyi Karatau Range, southern Kazakhstan,
and H. andreevae Popov and Holmer, 1994 was
described from the Akbulaksai Formation (late Early
Ordovician), South Urals.

It has proven difficult to differentiate dorsal from
ventral valves in H. mootwingeensis as both valves
appear to be relatively acuminate in outline, and few
internal features are known. However, even in poorly
preserved material, fully grown specimens of H.
mootwingeensis are readily distinguished in the field
by their considerably larger size, much flatter profile,
and more acuminate ventral valve outline compared
to Lingulobolus? gnaltaensis. These criteria allow
recognition of probable H. mootwingeensis from
fragmentary remains of external moulds in the

Proc. Linn. Soc. N.S.W., 128, 2007

Yandaminta Quartzite, on the western flank of
Mount Arrowsmith in the northwestern part of
the Koonenberry Belt, at a stratigraphic level
comparable with the lower Rowena Formation
(J.R. Paterson, Macquarie University, pers. comm.
2006).

Distribution

Lower Rowena Formation, “about 8 miles (13
km) from ‘Mootwingee’ homestead along old coach
road” according to Australian Museum register,
just east of Split Rock in Mutawintji National Park,
far-western NSW; probably also in Yandaminta
Quartzite, western flank of Mount Arrowsmith (Fig.
1A). Also sporadically occurs in the Bynguano
Quartzite in the southern Koonenberry Belt, 16 km
south of Little Topar (MMF 15012, 15015, 21080).

Zhanatellide? indet
Fig. 4 A—C

Material
One partial ventral valve represented by
internal and external moulds, MMF 44843.

Description

Valve ovoid or elliptical with slightly pointed
beak, maximum width at midlength or beyond;
lateral margins subparallel; anterior margin
probably broadly curved (based on extrapolation
from prominent growth lines which disrupt entire
shell thickness); profile very low in convexity.
Pseudointerarea narrow, with prominent v-shaped
emarginature for pedicle emergence; a short broad
ridge extends anteriorly from emarginature; small
umbonal muscle scars are separated by posterior
end of prominent pedicle nerve impression which
splays anteriorly as two deeply incised grooves.
Valve interior smooth to extremely finely pustulose
medially with sporadic small pits scattered
randomly.

Dimensions
Specimen length 9.1 mm, width 7.7 mm; both
dimensions incomplete.

Discussion

This specimen is attributed to the Zhanatellidae
on the basis of the prominent pedicle nerve
impression and the v-shaped emaginature. There is
insufficient material to compare with known taxa.

237

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

Distribution

Outcrop of upper Yarrimbah Formation adjacent
to “Yarrimbah” and “Wilga East” property boundary,
16 km west of Parkes; late Lancefieldian (La3) to
early Bendigonian (Bel).

Order Acrotretida Kuhn, 1949
Superfamily Acrotretoidea Schuchert, 1893
Family Acrotretidae Schuchert, 1893

Acrotretide gen. et sp. indet.
Fig.9 A, B

Material
One fragmentary dorsal valve, MMMC 4190.

Description

Valve thick-shelled, robust, with external
ornament of faint widely-spaced fila; larval shell
relatively small; posterolateral margins converging to
medial point. Pseudointerarea well developed, narrow,

triangular, with straight anterior margin; median
plate slightly depressed, flanked by pair of triangular
propareas. Valve interior bears high, blade-like
median septum supported by robust median buttress
sporting prominent node on each anterolateral corner;
median septum bears thickened, concave ventral
edge. Muscle scars ovate, raised, relatively small but
distinct.

Discussion

Although poorly preserved and lacking a
corresponding ventral valve, assignation of the
specimen to the Acrotretidae is supported by the
presence of a well-defined pseudointerarea, a high
median septum supported by a robust median buttress
and thickened, raised cardinal muscle scars.

Distribution

Allochthonous limestone clast in lower
Yarrimbah Formation, “Wilga East” property, 16 km
west of Parkes; Early Ordovician, late Lancefieldian
(La3) to early Bendigonian (Bel) age.

Figure 9. Acrotretide brachiopods from Early Ordovician limestones of central west NSW. A — B: Ac-
rotretide gen. et sp. indet. from allochthonous limestone clast in lower Yarrimbah Formation, interior
and oblique lateral views of dorsal valve MMMC 4190. C — D: Otariconulus sp. cf. O. intermedia (Popov
and Holmer, 1994) from allochthonous limestone pod in lower Hensleigh Formation, exterior and inte-
rior views of ventral valve MMMC 4203. E — H: Ephippelasmatidae gen. et sp. nov. from allochthonous
limestone pod in lower Hensleigh Formation, interior and oblique lateral views of dorsal valve MMMC
4193, and enlargements (G, H) of surmounting plate on median septum and spinose projections. Scale
bar equals 100 pm.

238 Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

Family Eoconulidae Rowell, 1965

Otariconulus Holmer and Popov, 2000
Type species: Ofariella prisca Popov and Holmer,
1994, by subsequent designation of Holmer and
Popov (2000, p.134).

Otariconulus sp. cf. O. intermedia (Popov and
Holmer, 1994)
Fig. 9 C, D

Material
One complete dorsal valve, MMMC 4203.

Description

Valve outline subcircular, 84% as long as wide
(0.57 mm long, 0.68 mm wide), moderately convex;
subcircular larval shell, 81% as long as wide (0.15
mm long, 0.19 mm wide), with ornament comprising
circular pits of one size; beak marginal, swollen with
postlarval shell bearing rugellae superimposed on
closely spaced growth lamellae; valve interior with
pair of large, poorly defined cardinal muscle scars
extending just short of midlength; median ridge
lacking.

Discussion

This valve agrees in most respects with the
description given by Popov and Holmer (1994, p.
143) of Otariconulus intermedia from the Lower
Ordovician (Tremadoc) Satpak Formation of north-
central Kazakhstan. However, in the absence of
the pedicle valve, only a tentative identification is
proposed.

Distribution
Allochthonous limestone in lower Hensleigh
Siltstone; mid-Bendigonian age (?Be2-Be3).

Family Ephippelasmatidae Rowell, 1965

Ephippelasmatidae gen. et sp. nov.
Fig. 9 E-H

Material
One dorsal valve, MMMC 4193.

Description

Valve outline subcircular with narrow, straight
posterior margin interrupted by gently protuberant
beak; external ornament comprises fine, regular,

Proc. Linn. Soc. N.S.W., 128, 2007

closely spaced fila; valve broadly sulcate in anterior
view and posteriorly with slightly inflated umbo.
Pseudointerarea relatively narrow, poorly divided,
propareas separated by median depression. Median
septum high, arising abruptly from near midlength
of valve at angle of approximately 45°, unsupported
by median buttress, bearing concave surmounting
platform; surmounting platform abruptly expands
anteriorly to form ovate plate in plan view with deep,
narrow concavity continuing medially to anterior edge
of plate; lateral edges of plate folded dorsomedially
beneath upper surface, appearing to continue
posteriorly but diminishing in width; undersurface
of plate bears single, prominent, robust, anteriorly
directed spine, possibly with a second much shorter
spine beneath it.

Discussion

Principal distinguishing characteristics of the
single known dorsal valve include absence of a
median buttress, median septum supporting ovate
surmounting plate with dorsomedially folded lateral
margins and a single prominent, anteriorly directed
spine on its undersurface. Unfortunately, intractable
matrix obscuring parts of the valve and lack of a
corresponding ventral valve preclude naming of a
new taxon at this time, although observable features
appear to be generically distinctive and unique.

This dorsal valve is comparable with that of
Lurgiticoma Popov (in Nazarov and Popov, 1980)
in configuration of the median septum and size
and shape of the pseudointerarea. In plan view the
median septum of both Lurgiticoma and the new
genus abruptly expands anteriorly and bears spines
on the undersurface; however, in the latter genus the
septum is not supported by a median buttress and
lacks the numerous spines of Lurgiticoma. Although
the pseudointerarea of the new genus is also similar
to that of Lurgiticoma in being relatively long with
propareas separated by a relatively large median
depression, it differs in bearing a median concave
indentation along the anterior margin.

Distribution
Allochthonous limestone in lower Hensleigh
Siltstone; mid-Bendigonian age (?Be2-Be3).

ACKNOWLEDGMENTS
David Barnes (NSW Department of Primary Industries)
expertly prepared the photographic illustrations, and

Dean Oliver drafted Figure 1. Reviews by Glenn Brock
(MUCEP, Macquarie University) and Lawrence Sherwin

239

EARLY ORDOVICIAN LINGULATE BRACHIOPODS

(Geological Survey of NSW) greatly facilitated polishing
of the manuscript for publication. Ian Percival publishes
with permission of the Deputy Director-General, NSW
Department of Primary Industries — Mineral Resources.
This paper is a contribution to IGCP Project No. 503:
Ordovician Palaeogeography and Palaeoclimate.

REFERENCES

Barrande, J. (1879). ‘Systéme silurien du centre de la
Bohéme. I** partie. Recherches paléontologique. Vol.
5. Classe des Mollusques. Ordre des Brachiopodes’.
(Published by the author: Prague and Paris). 226 pp.

Billings, E. (1872). On some fossils from the primordial
rocks of Newfoundland. Canadian Naturalist and
Geologist (new series) 6, 465-479.

Brock, G.A. and Holmer, L. (2004). Early Ordovician
lingulate brachiopods from the Emanuel Formation,
Canning Basin, Western Australia. Memoirs of the
Association of Australasian Palaeontologists 30, 113-
132.

Cockerell, T.D.A. (1911). The name Glossina. Nautilus
25, 96.

Cocks, L.R.M. and Lockley, M.G. (1981). Reassessment
of the Ordovician brachiopods from the Budleigh
Salterton Pebble Bed, Devon. Bulletin Natural
History Museum London (Geology) 35, 111-124.

Cooper, G.A. (1956). Chazyan and related brachiopods.
Smithsonian Miscellaneous Collections 127, 1245 pp.
+ 269 pl.

Droser, M.L., Hughes, N.C. and Jell, P.A. (1994).

Infaunal communities and tiering in Early Palaeozoic
nearshore clastic environments: trace-fossil evidence
from the Cambro-Ordovician of New South Wales.
Lethaia 27, 273-283.

Fletcher, H.O. (1964). New linguloid shells from Lower
Ordovician and Middle Palaeozoic rocks of New
South Wales. Records of the Australian Museum, 26:
283-294.

Goryansky, V. Yu. (1969). Bezzamkovye brakhiopody
kembrijskikh I ordovikskikh otlozhenij severo-zapada
Russkoj platformy. [Inarticulate brachiopods of the
Cambrian and Ordovician of the northwest Russian
Platform.] Ministerstvo Geologii RSFSR, Severo-
Zapadnoe Territorial ’noe Geologicheskoe Upravlenie
6, 1-173. Nedra, Leningrad. [in Russian]

Havliéek, V. (1980). Inarticulate brachiopods in the Lower
Ordovician of the Montagne Noire (South France).
Mémoire de la Société d’Etudes Scientifiques de
l’Aude 1, 1-11.

Havlicek, V. (1982a). Ordovician in Bohemia:
development of the Prague Basin and its benthic
communities. Sbornik geologickych véd, Geologie
37, 103-136.

Havlicek, V. (1982b). Lingulacea, Paterinacea, and
Siphonotretacea (Brachiopoda) in the Lower
Ordovician sequence of Bohemia. Sbornik
geologickych véd, paleontologie 25, 9-82.

240

Holmer, L.E. (1991). The systematic postition of
Pseudolingula Mickwitz and related lingulacean
brachiopods. In “Brachiopods through time.
Proceedings of the 2nd International Brachiopod
Congress’ (Eds D.I. MacKinnon, D.E. Lee and J.D.
Campbell) pp. 15-21. (Balkema: Rotterdam).

Holmer, L.E., Koneva, S.P., Popov, L.E. and Zhylkaidarov,
A.M. (1996). Middle Ordovician (Llanvirn) lingulate
brachiopods and conodonts from the Malyi Karatau
Range, Kazakhstan. Paldontologische Zeitschrift 70,
481-495.

Holmer, L.E. and Popoy, L.E. (2000). Class Lingulata.

In “Treatise on Invertebrate Paleontology, Part H.
Brachiopoda (Revised), Part 2’ (Eds A. Williams,
C.H.C. Brunton and S.J. Carlson) pp. 20-146.
(Geological Society of America: Boulder, and
University of Kansas Press: Lawrence).

Krause, F.F. and Rowell, A.J. (1975). Distribution and
systematics of the inarticulate brachiopods of the
Ordovician carbonate mud mound of Meiklejohn
Peak, Nevada. The University of Kansas,
Paleontological Contributions 61, 1-74.

Legrand, P. (1971). A propos de la présence de Dinobolus
(?) aff. Brimonti (M. Rouault) au Sahara Algérien.
Mémoire B.R.G.M. 73, 79-91.

Matthew, G.F. (1895). Traces of the Ordovician System
on the Atlantic coast. Royal Society of Canada,
Transactions (series 1, section 4) 1, 253-279.

McCoy, F. (1851). On some new Cambro-Silurian fossils.
Annals and Magazine of Natural History (series 2) 8,
387-409.

Mergl, M. (2002). Linguliformean and craniiformean
brachiopods of the Ordovician (Tfenice to Dobrotiva
Formations) of the Barrandian, Bohemia. Acta Musei
Nationalis Pragae, series B — Historia Naturalis 58,
1-82.

Mickwitz, A. (1896). Uber die Brachiopodengattung
Obolus Eichwald. Memoires de l’Académie Impériale
des sciences de St. Petersbourg 4, 275 pp.

Nazarov, B.B. and Popov, L.E. (1980). Stratigrafiia 1
fauna kremnisto-karbonatnykh toltshch ordovika
Kazakhstana (Radioliarii i Bezzamkovye
Brakhiopody). [Stratigraphy and fauna of Ordovician
siliceous-carbonate deposits of Kazakhstan
(Radiolarians and inarticulate brachiopods)]. Trudy
Geologicheskogo Instituta Akedemii Nauk SSSR 331,
1-192. [in Russian].

Percival, I.G. (1978). Inarticulate brachiopods from the
Late Ordovician of New South Wales, and their
palaeoecological significance. Alcheringa 2, 117-141.

Percival, I.G. (2000). Brachiopods, pp.73-75 in Webby,
B.D., Percival, I.G., Edgecombe, G.D., Cooper,
R.A., VandenBerg, A.H.M., Pickett, J.W., Pojeta,

J., Playford, G., Winchester-Seeto, T., Young, G.C.,
Zhen, Y-y., Nicoll, R.S., Ross, J.R.P. and Schallreuter,
R. Ordovician palaeobiogeography of Australasia.
Memoir of the Association of Australasian
Palaeontologists 23, 63-126.

Proc. Linn. Soc. N.S.W., 128, 2007

I.G. PERCIVAL AND M.J. ENGELBRETSEN

Percival, I.G., Webby, B.D. and Pickett, J.W. (2001).
Ordovician (Bendigonian, Darriwilian to Gisbornian)
faunas from the northern Molong Volcanic Belt of
central New South Wales. Alcheringa 25, 211-250.

Popov, L.E. (1980). [New brachiopod species from the
Middle Ordovician of the Chu-Ili Range]. Ezhegodnik
Vsesoyuznovo Palaeontologicheskovo Obshshestva
23, 139-158. [in Russian].

Popov, L.E., Cocks, L.R.M. and Nikitin, IF. (2002).
Upper Ordovician brachiopods from the Anderken
Formation, Kazakhstan: their ecology and
systematics. Bulletin Natural History Museum
London (Geology) 58, 13-79.

Popov, L.E. and Holmer, L.E. (1994). Cambro-
Ordovician lingulate brachiopods from Scandinavia,
Kazakhstan, and South Ural Mountains. Fossils &
Strata 35, 1-156.

Rouault, M. (1850). Note préliminaire sur une nouvelle
formation découverte dans le terrain silurien inférieur
de la Bretagne. Bulletin de la Société Geologique de
France (series 2) 7, 724-744.

Sharp, T.R. (2004). “Geological history of Mutawintji
National Park’. (Geological Survey of New South
Wales: Sydney).

Sherwin, L. (1979). Age of the Nelungaloo Volcanics, near
Parkes. Quarterly Notes of the Geological Survey of
New South Wales 35, 15-18.

Sherwin, L. (1990). Early Ordovician graptolite from the
Peak Hill District. Quarterly Notes of the Geological
Survey of New South Wales 90, 1-4.

Sherwin, L. (2000). Nelungaloo Volcanics; Yarrimbah
Formation. In ‘Forbes 1:250 000 Geological Sheet
SI/55-7, 2™4 edition. Explanatory Notes’ (compilers
and editors P. Lyons, O.L. Raymond and M.B.
Duggan). AGSO Record 2000/20, 29-30.

Sinclair, G.W. (1945). Some Ordovician lingulid
brachiopods. Transactions of the Royal Society of
Canada, series 3, section 4, 39, 55-82.

Sowerby, J. de C. (1839). Shells. In R.I. Murchison, ‘The
Silurian System’ part 2, pp. 579-712. (J. Murray:
London).

Sutton, M.D., Bassett, M.G. and Cherns, L. (1999).
Lingulate brachiopods from the Lower Ordovician of
the Anglo-Welsh Basin. Part 1. Palaeontographical
Society Monograph 153, 1-60.

Sutton, M.D., Bassett, M.G. and Cherns, L. (2000). The
type species of Lingulella (Cambrian Brachiopoda).
Journal of Paleontology 74, 426-438.

Walcott, C.D. (1912). Cambrian Brachiopoda. United
States Geological Survey Monograph 51, part 1 872
pp., part 2 363 pp. + 104 pl.

Webby, B.D. (1983). Lower Ordovician arthropod trace
fossils from western New South Wales. Proceedings
of the Linnean Society of New South Wales 107, 61-
76.

Proc. Linn. Soc. N.S.W., 128, 2007

Williams, A., Brunton, C.H.C. and Carlson, S.J. (eds).
(2000). ‘Treatise on Invertebrate Paleontology, Part
H. Brachiopoda (Revised), Part 3’. (Geological
Society of America: Boulder, and University of
Kansas Press: Lawrence).

Zhan, R.-b. and Cocks, L.R.M. (1998). Late Ordovician
brachiopods from the South China Plate and their
palaeogeographical significance. Special Papers in
Palaeontology 59, 70 pp.

Zhen, Y.-y. and Percival, I.G. (2006). Late Cambrian-Early
Ordovician conodont faunas from the Koonenberry
Belt of western New South Wales. Memoir of the
Association of Australasian Palaeontologists 32,
267-285.

Zhen, Y.-y., Percival, I.G. and Webby, B.D. (2003). Early
Ordovician conodonts from far western New South
Wales, Australia. Records of the Australian Museum
55, 169-220.

Zhen, Y.-y., Percival, I.G. and Webby, B.D. (2004).

Early Ordovician (Bendigonian) conodonts from
central New South Wales, Australia. Courier
Forschunginstitut Senckenberg 245, 39-73.

241

oyEeramaamoa

tees) L& yrocing hak . hb? yricnamed peelllay
nef vgvletaaple-oteslatseva rtecretten th? (UA) alees
jac yoluond) |,’ Etre D (bse 6 elon Gadel 465 Wi
is ime vinit- be pohiuold asian be vibiagBno
oN ed omen ‘Peeiastde te
natavabs ote.) (G0). Ma ch ladoot thee - d-4 oul
vise) bre atelT ape dived act mot ehoqoidond
v4. rap’ loon? sommuimgie lsoiiqnigosgosstag

othe), aes eens Re LEM ES

wVOV Se ‘cacercenlod aay srnin yr erro AB ae teeny!
vlw-neticene) ond (WS) DA Jeviowt bos we ¥ fel ; moons
mmedisnood ad) ara l-esmal), inebanoe.noterain? rr e rhe elentid nda Coenal
als Yo onivh vole? dtantemamgteow bales | haviees snihaaiionntt lotsa =
SE wakgobosreep ho cam want Wu ere niabopinede- Sp “Vee ont wena meget ,
‘ nd PotBietahe, ' ee
ixgal 4 S005). 6 edie. bine, Rob day tof tiadS. svensk ment abn onl ashe 2
hve? wert mew sa aol aaebeedo natoinebsty SnieGirabinmtenitin’.
voucwenl,. mew atk y AeA Tells A 2aldW ulepramp af Escees borat Lanadibtep, 22m
PF ) OER OA a2 wale vibeeaitire origin! 51G: eee ,
{P0009} OLAd yediioW baw OirdevubRop eds camaniaaaneds “sols ON eee yO He
rece. athe > fosioenibesthl apitrobat) (bok Jo spigot aiined nb ro sna 2
shar) nilatiead, asd Giuod to ier} | 3 red n ype, YY 9
FT-OF PL yevlustoned witeanlgnadseot
cong JTS ‘ 4
moe fre w iM, Reiman iF 1) ts Segoe
| Reaseeserneee wo ni lag MED sy LN)
the Mandhergeds y macpcarcay yeni: Area
herd = Albottix c | Rive Reashwil # wR
12a oath nami otboatg cel ach DOE 2 w
panda seat piaenainains rineeny Sell
1244 me be dead Ned 3 eddietanedit Aid
ites’ .conevlov opisgauioM beans a 1M
Acrictad2 intiegeiin so AO D8 So fomsdrath* am areo"L
rraiqinoy} vei oem
ott bas bicurted tA ened Abana
DERG 02 00D Winwi O208) term
biliyel! aatcivolaO same — wick
yo wisSod, Heel UAV to sqctcaneoplym :
| OD SE kb tio Leetonia
af Tf! eek a4 pre peter a t
neni meatal one a
seeeayen fenieiin aM sag aie of
‘to nawolst) weyrod vat mod sbogvidaedt 2
Vera aerssstoanieth Line artigo A
ale ain’ dd EEL be : uy A
oth (QOL conch’) ke SIM. neeestl AC eet
( nba tweiiedenne’>) eAlakagns To easoqe oq
pipers”
sun sharpener S440 1 !
oe pop 6 eR ay

a ae iB ond DEe
© eo aes; plea snaapy of arta,
tame «0 te cor qenah of Aatiratostan ots
ee ee

Morphometric Relationships and Catch Composition
of Wobbegong Sharks (Chondrichthyes: Orectolobus)
Commercially Fished in New South Wales, Australia

CHARLIE HUVENEERS!, NICHOLAS M. Otway? AND Rosert G. Harcourt!.

‘Marine Mammal Research Group, Graduate School of the Environment, Macquarie University, Sydney
NSW 2109 (charlie.huveneers@gse.mq.edu.au); and * NSW Department of Primary Industries, Port Stephens
Fisheries Centre, Taylors Beach Road, Taylors Beach NSW 2316

Huveneers, C., Otway, N.M. and Harcourt, R.G. (2007). Morphometric relationships and catch
composition of wobbegong sharks (Chondrichthyes: Orectolobus) commercially fished in New South
Wales, Australia. Proceedings of the Linnean Society of New South Wales 128, 243-249.

Wobbegongs (Orectolobiformes) are commercially targeted in New South Wales, Australia. Catches have
declined approximately 60% in a decade, leading to concerns over the fishery’s sustainability. However,
length and weight composition of the catch is unknown as carcasses are trunked (i.e. beheaded and
eviscerated) before landing. We provide parameters for length—length, weight—-weight and weight—length
relationships to convert carcass length and carcass weight measurements to total lengths and total weights
used in fisheries assessments. Neonates and small juveniles were conspicuously absent in the length-
frequency distributions of all three species, suggesting the potential existence of nursery areas not available

to the commercial fishery.

Manuscript received 5 January 2007, accepted for publication 24 January 2007.

KEYWORDS: commercial fishery, morphometric relationship, Orectolobus, wobbegong.

INTRODUCTION

Three species of wobbegong shark: the spotted
wobbegong, Orectolobus maculatus, the dwarf
ornate wobbegong, O. ornatus, and the large ornate
wobbegong, O. halei (Huveneers 2006) occur in
coastal waters off New South Wales (NSW), Australia
and are commercially targeted by the Ocean Trap and
Line fishery. Wobbegongs have been sold as ‘boneless
fillets’ or ‘flake’ and their catch has declined from
~150 tonnes in 1990/91 to ~70 tonnes in 1999/00,
a decrease of > 50% in less than a decade (Pease
and Grinberg 1995; NSW Department of Primary
Industries, unpublished data). This decline led to
wobbegongs being listed as ‘Vulnerable’ (in NSW)
and ‘Near Threatened’ (globally) under the World
Conservation Union (IUCN) Red List assessment
(Cavanagh et al. 2003) and to concerns over the
sustainability of the fishery.

Given that many shark species, including
wobbegongs, are trunked prior to landing, partial
length and carcass weight are usually the only
measurements that can be recorded (FAO 2000).
Relationships between partial length and carcass

weight and their respective total length and total weight
are a fundamental requirement for an assessment of
the catch composition, and towards the ecologically
sustainable management of the fishery.

This study presents length—length, weight—
weight, and weight—length relationships for each
of the three species caught in the NSW commercial
fishery. Catch composition and length-frequency
distributions recorded during the study are also
presented.

MATERIALS AND METHODS

Wobbegongs were collected from commercial
fishers at six locations in NSW (Nambucca Heads,
Port Stephens, Newcastle, Terrigal, Sydney and
Eden) (Fig. 1). Wobbegongs were caught on setlines
with O’Shaughnessy style hooks size 10/O or 12/0,
with a 50-100 cm long wire or nylon trace attached to
the bottom line by a stainless sharkclip. Hooks were
baited with black fish (Girella tricuspidata), mullet
(Mugil cephalus) or Australian salmon (Arripis
trutta). Lines were set before sunset and hauled at

WOBBEGONG MORPHOMETRICS IN NEW SOUTH WALES

Foo ot Socal it me ole | =. SES ae) JURE RE ESS Sb om Piette tt Leeeters
a 4
: :
2 :
i ar
rage Nambucca Heads |
bm i
Pe :
é ' 4
ri 2} 33"
} :
' 4
Newcastle g Port Stephens
errigal '
: Sydney, Lsase
‘ 1
|
; io
40 km a
st 4
Sian en ae no an cs ce a oe GE Re I 1 40000 an an A OR eg TS em Ow A a oy a a ’
148 1st 132° 14 14s

Figure 1. Sampling locations for collection of wobbegongs in New South Wales, Australia

sunrise on the following day.

The species, gender and a series of length
measurements were recorded (to the nearest mm)
for each shark caught. The length measurements
included: total length (TL), snout to anal-fin insertion
length (SAL), and partial length from the pectoral-fin
origin to the caudal-fin origin (PL). SAL was taken
instead of fork length as upper and lower caudal fin
lobes of wobbegongs are not discernible. Total weight
(TW) and carcass weight (CW) were recorded using
spring balances (scale: 100 + 0.2 kg, 20 + 0.2 kg, 5
+ 0.1 kg).

Linear regressions of TL on SAL, TL on PL,
and TW on CW were determined for each of the
three species using data pooled across all sites. Log-
transformed data were used for the regressions of TW
on TL and CW on PL and corrected for biases caused
by natural logarithmic transformation (Beauchamp
and Olson 1973). Analyses of covariance (ANCOVA)
were used to test for differences between sexes in
all regressions. When the slopes and intercepts did
not differ significantly between sexes the data were
pooled and a common regression determined.

RESULTS

A total of 904 wobbegongs (435 males and 469
females) was collected comprising: 183 male and 202
female O. ornatus (combined range 471-1,017 mm
TL), 97 male and 88 female O. maculatus, (combined
range 870-1,575 mm TL), and 155 male and 179
female 334 O. halei (combined range 869-2,065
mm TL). Most O. ornatus (86.5%) were collected
off Nambucca Heads with none caught south of

244

Port Stephens. Orectolobus maculatus catches were
distributed among Nambucca Heads (26.5%), Port
Stephens (30.8%) and Sydney (37.8%), with none
caught in Eden. Orectolobus halei were caught at
all locations, with the majority caught off Sydney
(62.6%), and sporadic captures at the remaining
locations (Table 1). Neonates (born at ~21 cm for O.
ornatus and O. maculatus and ~30 cm for O. halei)
and small juveniles were absent in the catches of all
three species (Fig. 2).

The conversion parameters estimated are
applicable to the size range analysed (Table 1) which
covers most of the population size range, with the
exceptions of neonates and small juveniles (not
caught by the commercial fishery). All regressions
were significant with 19 correlation coefficients out
of 22 over 0.84 (Table 2 and 3).

The slopes of the regressions of TL on SAL (Table
2) did not differ significantly between the sexes for O.
ornatus and O. maculatus (ANCOVA: Baie. =2.17 and
0.62 respectively, P> 0.05), but the intercepts differed
significantly between males and females (ANCOVA:

nuttt ag 5.29 and 11.06 respectively, both P < 0.05).
The adjusted means showed that male O. ornatus
and O. maculatus had a significantly greater TL for a
given SAL compared to females. Similarly, the slopes
of the regressions of TL on PL (Table 2) did not differ
significantly between the males and females of O.
ornatus and O. maculatus (ANCOVA: De ae 3.06
and 0.17 respectively, P > 0.05). Again, the intercepts
of the regressions of TL on PL (Table 2) differed
significantly between the sexes (ANCOVA: 1 see
9.24 and 2.44, P< 0.001 and P < 0.05, respectively).
The adjusted means showed that the male O. ornatus
and O. maculatus had a significantly greater TL for

Proc. Linn. Soc. N.S.W., 128, 2007

C. HUVENEERS, N.M. OTWAY AND R.G. HARCOURT

S

Frequency (%)
oso8 88& SES

0
xs x

Total length (rm)

RL JS SF. FJ SJ Sf
|, ee lame ll
CEEOL CO ES S

intercepts were significantly different
between the sexes (ANCOVA:

seeps ~ 20:20,and's:49) 2 0'001
and P < 0.05, respectively). The
adjusted means showed that females
of O. maculatus and O. halei had a
significantly greater TW for a given
TL when compared to males.

Neither the slopes nor intercepts
of the regressions of CW on PL (Table
3) differed significantly between the
sexes for O. ornatus, O. maculatus
and O. halei (ANCOVA: F =

slopes

IES 2 alorandeletS sb: = 0.01

intercepts i ?

0.04 and 0.60; all P > 0.05 for O.
ornatus, O. maculatus and O. halei ,

respectively).
DISCUSSION
The spatial distribution of

wobbegong catches provides an
indication of their distribution within
NSW waters. Port Stephens was the
southern-most location where O.
ornatus was caught. Although O.
ornatus have been recorded as far
south as Sydney (207 km south of Port
Stephens), no O. ornatus was caught
around Sydney. Museum registered
specimens have been collected as far
north as the Whitsunday Islands (20°
20'S 148° 54’E, Australian Museum
specimen IA 3831), restricting the
distribution of O. ornatus from

Figure 2. Length-frequency distribution of wobbegongs caught tropical to warm temperate waters

during sampling period for (a) O. ornatus, (b) O. maculatus, and (c) Of eastern Australia.
O. halei for males (solid bar) and females (open bar).

a given PL when compared to females. Neither the
slopes nor intercepts of the regressions of TL on SAL
and TL on PL (Table 2) differed significantly between
the sexes for O. halei (ANCOVA: TL on SAL: F
= 2.18 and F.

intercepts

= 1.57, both P > 0.05; TL on PL:
Bee 7 Oo and Fae. — 040, both P0105):

The slopes of the regressions of TW on TL (Fig.
3 and Table 3) differed significantly between male
and female O. ornatus (ANCOVA: Bee Bie 6.62, P
< 0.05) with weight increasing at a faster rate than in
females. In contrast, slopes of the regressions of TW
on TL (Table 3) for male and female O. maculatus and
O. halei did not differ significantly (ANCOVA: eeu
= 0.32 and 0.04 respectively, both P > 0.05), but the

slopes

Proc. Linn. Soc. N.S.W., 128, 2007

Orectolobus
maculatus is abundant in central NSW,
around Port Stephens and Sydney.
Orectolobus maculatus is caught in

larger numbers in northern NSW than O. halei and

has been recorded as far north as Gladstone (Kyne
et al. 2005). In contrast to O. halei, O. maculatus

was rarely caught around Merimbula and Eden (S.

Fantham, pers comm.), restricting its distribution in

eastern Australia from tropical to temperate waters.

Orectolobus halei catches were low in northern NSW

and higher around Sydney and Eden, where it was

the only species caught during this study. In NSW,

O. halei is more abundant in temperate waters with

abundance decreasing in warm temperate waters.

There is apparently a similar trend for O. halei

collected in Western Australia (WA) (J. Chidlow, pers

comm.).

245

WOBBEGONG MORPHOMETRICS IN NEW SOUTH WALES

Table 1. Number (with TL size range in mm) of wobbegong caught during June 2003—May 2006

Location O. ornatus O. maculatus O. halei Total

Nambucca Heads 333 (471-994) 49 (1,160—1,485) 31 (1,175-1,972) 411
Port Stephens 52 (577-1,017) 57 (870-1,440) 10 (1,280—1,875) 119
Newcastle 7 (1,265-1,435) 3 (1,444-1,755) 10
Terrigal 2 (unknown) 8 (1,860—1,930) 10
Sydney 70 (1,055—1,575) 209 (869—2,065) 278
Eden 73 (1,190-1,870) 64
Total 385 (471-1,017) 185 (870—1,575) 334 (869-2,065) 904

Table 2. Relationships between length—length and weight—weight. Estimated parameters (and standard
error) from the linear regression analysis to derive the equation Y = a+bX; a and b are parameters;

n is sample size; r’ is square of correlation coefficient; rmse is root mean square error; and P is prob-
ability of statistical significance between sex with ns representing P > 0.05, * P< 0.05, ** P< 0.01, ***
P< 0.001. TL is total length; SAL is snout to anal-fin insertion length; PL is partial length; TW is total
weight; CW is carcass weight.

P
Y-X Species Sex n a (S.e.) b (s.e.) i rmse slope intercept
TL-SAL O. ornatus Male 161 44.80 (15.52) 1.16(0.02) 0.94 19.66 ns rs
Female 164 71.79 (15.51) 1.12(0.02) 0.94 21.54
O. maculatus Male 93 26.98 (24.33) 1.22(0.02) 0.97 25.32 ns ai
Female 77 ~—-41.52 (19.03) 1.200.002) 0.98 16.52
O. halei Combined 236 10.34 (14.17) 1.23(0.01) 0.98 33.38 ns ns
TL-PL O. ornatus Male 113 164.26 (26.42) 1.28(0.05) 0.86 34.73 ns Bos
Female 124 96.00 (18.76) 1.38(0.03) 0.93 25.60
O. maculatus Male 63 159.61 (51.08) 1.40(0.06) 0.90 43.4 ns z
Female 60 184.39 (45.98) 1.34(0.05) 0.91 39.32
O. halei Combined 174 103.97 (23.34) 1.49(0.02) 0.96 54.63 ns ns
TW-CW OO. ornatus Combined 73 1.33 (00.14) 1.33(0.06) 0.87 0.31 ns ns
O. maculatus _ Combined 93 3.95 (00.75) 1.01(0.08) 0.61 1.83 ns ns
O. halei Combined 148 1.67 (00.77) 1.53(0.05) 0.87 3.90 ns ns

Neonates and small juveniles were rarely caught
by commercial wobbegong fishers at any location.
Several reasons may account for their absence.
Neonates and small juveniles might occupy crevices
to avoid predation and forage on small prey living in
the crevices. This may provide a physical partitioning
of the habitat within a given location. Gear selectivity
could also decrease neonate catch because hooks and

246

baits used in the commercial wobbegong fishery are
too large. However, gear selectivity is unlikely to
explain the absence of larger juveniles because O.
ornatus of 700-1000 mm TL are commonly caught
using the same gear and in the same areas where only
a few O. halei smaller than 1300 mm TL are caught.
It seems more likely that small wobbegongs are not
available to the fishery and occur within different

Proc. Linn. Soc. N.S.W., 128, 2007

C. HUVENEERS, N.M. OTWAY AND R.G. HARCOURT

—s

Total mass (kg) &
Oe MiWwRiUnm~) 0010

600 800 1000

(b) 30

s (kg)

Total mas
S

Qo

400 800 81200 1600

700
Total length (mm)

1400

2100

OreMwWwLUnDd~ 0010

200 400 600

400 800 §=1200 81600

0

700 1400
Total length (mm)

2100

Figure 3. Relationships between total weight and total length of wobbegongs in NSW. Plots of mean total
weight against TL (—), with 95% confidence limits (- — —) and 95% prediction intervals (---), for males
(left), and females (right) for (a) O. ornatus, (b) O. maculatus, and (c) O. halei. Values for parameters and
statistical quantities from regression analysis are given in Table 3.

habitats. Furthermore, a similar study in WA yielded
no O. maculatus smaller than 900 mm TL and only
one O. halei (synonym O. ornatus) smaller than 1200
mm TL (Chidlow 2003). Size segregation might
therefore occur with neonates and small juveniles
living in primary and/or secondary nursery areas.
Size segregation in habitat use is commonly found
in chondrichthyans (e.g. Simpfendorfer 1992), with
neonates living in nursery areas for the first weeks,
months or years (Heupel and Hueter 2002). Nursery
areas are thought to provide neonates and small sharks
with increased food availability and/or protection
against predators (Heupel and Hueter 2002).

Proc. Linn. Soc. N.S.W., 128, 2007

The regression parameters in Tables 2 and 3
are provided for scientists and fisheries managers
as an aid to determining size when TL and TW are
required but cannot be measured, but where SAL, PL
or CW are available. The absence of sex differences
in the CW-PL relationships although correlation
coefficients are high suggested that somatic growth
was similar between males and females (Braccini
et al. 2006). However, the regressions of TW on TL
differed significantly between males and females with
greater body weight in females. Sex-based differences
in body weight are often due to discrepancies in
the weights of internal organs and are common in

247

WOBBEGONG MORPHOMETRICS IN NEW SOUTH WALES

Table 3. Relationships between total weight (TW)-total length (TL) and carcass weight (CW)-partial
length (PL). Estimated parameters (and standard error) for the relationships for males and females
derived from the equation TW=acTL” and CW=acPL’; a and b are parameters; c is the Beauchamp and
Olson (1973) correction factor; other parameters and statistical quantities as in Table 2.

P
Shark category n a (s.e. range) x 10° b (s.e.) c ie rmse slope intercept
TW-TL
O. ornatus ez ee
Males 129 21.1 (10.1-44.1) 2.82(0.11) 1.008 0.84 3.28
Females 159 1.81 (0.95-3.46) 3.20 (0.10) 1.010 0.88 4.62
O. maculatus ns ee
Males 86 57.4 (26.3-125) 2.69 (0.11) 1.008 0.88 2.88
Females 73 31.7 (12.8-78.3) 2.78 (0.13) 1.007 0.87 2.64
O. halei ns =
Males 86 73.6 (39.2—138) 2.69(0.11) 1.008 0.88 2.88
Females 106 6.52 (3.88-11.0) 3.01 (0.070 1.008 0.95 5.21
CW-PL
O. ornatus 26 47 (3.12-709) 2.83 (0.43) 1.008 0.9 0.16 ns ns
O. maculatus 94 1,090 (405-2,920) 2, SOxl)pmeOl, Ors, ‘Ox ns ns
O. halei 149 69.9 (40.8-120) 2.80 (0.08) 1.013 0.64 0.13 ns ns

chondrichthyans (e.g. Walker 2005). Differences
occur due to the inclusion of pregnant females, and
the heavier reproductive organs and liver in females
(Stevens and Wiley 1986). In contrast, male O.
ornatus and O. maculatus had significantly greater
TL for a given SAL and PL compared to females. The
reason for this sex difference is unknown.

Most life history parameters used in fisheries
assessments are determined as a function of total
length or weight. Wobbegongs landed in the NSW
Ocean Trap and Line Fishery are, however, beheaded
and eviscerated preventing the measurement of total
length and total weight. The regression relationships
documented in this study allow estimates of total
length and total weight to be obtained from landed
carcasses enabling future assessments of the
ecological sustainability of the fishery through a
more accurate knowledge of the catch composition
of this fishery. Although many studies provide
relationships between total length and total weight
(e.g. Stevens and McLoughlin 1991), we concur with
recommendations of the International Plan of Action
for the Conservation and Management of Sharks
(IPOA-Sharks) (FAO 2000) that future studies should

248

also incorporate the measurement of partial lengths
and carcass weight. Only when this is done routinely,
will it be possible to estimate, with accuracy, total
length and total weight and provide much needed
information on the length/weight composition of the
catch of shark fisheries.

ACKNOWLEDGEMENTS

The authors thank Reala Brislane, Jason Moyce, Ian
Puckeridge, Mark Phelps and Shannon Fantham for
assistance aboard their fishing vessels, several interns
and volunteers for help with sampling, Simon Allen for
comments on an earlier version of the manuscript, and
Matias Braccini for giving the authors the idea of this
manuscript. Terry Walker is also thanked for his help with
data analysis. Charlie Huveneers was supported by an
international Macquarie University Research Scholarship.
Financial support was provided by the Graduate School of
the Environment, NSW Department of Primary Industries
and the Australian Geographic Society.

Proc. Linn. Soc. N.S.W., 128, 2007

C. HUVENEERS, N.M. OTWAY AND R.G. HARCOURT

REFERENCES

Beauchamp, J.J., and Olson, J.S. (1973). Corrections
for bias in regression estimates after logarithmic
transformation. Ecology 54, 1403-1407.

Braccini, J.M., Gillanders, B.M., and Walker, T.I. (2006).
Total and partial length—length, mass—mass and
mass—length relationships for the piked spurdog
(Squalus megalops) in south-eastern Australia.
Fisheries Research 78, 385-389.

Cavanagh, R., Kyne, P., Fowler, S.L., Musick, J.A., and
Bennett, M.B. (2003) ‘The conservation status of
Australasian chondrichthyans. Report of the IUCN
Shark Specialist Group Australia and Oceania
regional red list workshop. Queensland, Australia,
7-9 March 2003.’ (The University of Queensland:
Brisbane).

Chidlow, J. (2003) The biology of wobbegong sharks
(family: Orectolobidae) from south-western
Australian waters. Masters thesis, James Cook
University, Townsville, Australia.

FAO (2000) ‘Fisheries management. 1. Conservation and
management of sharks.’ (FAO: Rome, Italy).

Heupel, M.R., and Hueter, R.E. (2002). Importance of
prey density in relation to the movements patterns
of juvenile blacktip sharks (Carcharhinus limbatus)
within a coastal nursery area. Marine and Freshwater
Research 53, 543-550.

Huveneers, C. (2006). Redescription of two species of
wobbegongs (Chondrichthyes: Orectolobidae) with
elevation of Orectolobus halei Whitley 1940 to
species level. Zootaxa 1284, 29-51.

Kyne, P., Johnson, J.W., Courtney, A.J., and Bennett,
M.B. (2005). New Biogeographical information
on Queensland Chondrichthyans. Memoirs of the
Queensland Museum 50, 321-327.

Pease, B.C., and Grinberg, A. (1995) “New South Wales
Commercial Fisheries Statistics 1940 to 1992.’ (NSW
Fisheries: Sydney, NSW, Australia).

Simpfendorfer, C.A. (1992). Reproductive strategy of the
Australian sharpnose shark, Rhizoprionodon taylori
(Elasmobranchii: Carcharhinidae), from Cleveland
Bay, northern Queensland. Australian Journal of
Marine and Freshwater Research 43, 67-76.

Stevens, J.D., and McLoughlin, K.L. (1991). Distribution,
Size and Sex Composition, Reproductive Biology and
Diet of Sharks from Northern Australia. Australian
Journal of Marine and Freshwater Research 42,
151-199.

Stevens, J.D., and Wiley, P.D. (1986). Biology of two
commercially important carcharhinid sharks from
northern Australia. Australian Journal of Marine and
Freshwater Research 37, 671-688.

Walker, T.I. (2005) Reproduction in fisheries science.

In ‘Reproductive biology and phylogeny of
Chondrichthyes: sharks, rays and chimaeras’. (Ed. W.
C. Hamlett) pp. 81-127. (Science Publishers Inc.:
Enfield, USA).

Proc. Linn. Soc. N.S.W., 128, 2007

249

THODSAH 2

~ | pernveemers (aad cur a weap fr ae bina
a

» het=eon tutel e« thn Ci eae ale iA er 4 (CW)- parva

Wes TL* ened [W008 *

| & “inh ome TERE LO

vy ortew | veimenes sali eoeeiedaaataiaaaaeiaae a:
TOb1-2084 2 qgoosA mois
(ONS oh sti tam; cA persbestfi =
tae eeman-ezers ae pach obo ee
' oo Robmag-beohq ade wot sqisiauotselay gignsl
- Lntlevien A arstene-dies ry Cre

OhE-208 MY craves creat =
bas WAAL Jointly 21% Jalwot A omyA A
‘to eutute noliavrscnas sl (2000) 8.M pre
rd). | 444] ) 7.2) 16 wu vfs iio ahi mney dtiiarodts a

1 Peer singe A) aw silence

ISSR 20 (0 1 ihatedA talon 4 Scolar ew :
cbantensau() bo Ciasavin'd onT) * FORK dans ( a aa
A
09 10. daypaih Ghtpeddea fo: ells 8 (£005) Lo =
reat then sere, (abidqlatz0 tee rf
doo’) zone reeds avn tole sulk
aie A stivatwOF vtiervi

_ a gonerogne) | psreysqagn evirorlait’ (QO0S) O;
(laa so 3075)” aftnds Yo tosmegense ;
LO sonkrinaiet (2005) AStyeasel bre , AM,
nmoting aaomsvom of o) aoielsy ai yesh Ys
Cowvodwwll enniteeoteens) adds qattonld ahaa 0
ier’) Weenies 2ote ecregue (opapoo A niet va :
Oe2-ES £2 does
Ky anishqe dvr ts wongieseasbs st (8) Dee

dtiy (oubidgigized) gevdidpishaod > 2 ve

a ODO | vobid ve ioled aatiobeynt

12-22 ,.A8S1 pzpooS ova asiage |

ws
—
r

drooetl bas obé wana hd chi Tee
Mang) at bectslomesenieh Wi 1) Mode bet
a eon simedidanaot 3. yee) ms 0 “a

Tete BE yatyest, : s

slew Ulin? Gays” (Beet) po — a
WHAT SOOT aa 01 jeassse Sea dat ePSAAOS
Wate W2A veaby?e eaitadeit -

wile ygetet: svitanbewp a (L001) AD ashobosiqntn
winkna eohoroernorhs herds oeoeperlz anteeuA -
boulovel) nadtysabiaaow) wiisiatdomacta), *
Wie owes «otucvat barslaaseeQ mednon ant 5
Va, bho stan seamen y bin on AVES. tg

AO we {27} pilthgoods mae, ig Ai. ee

one tanks svibeubegge Sa aotip er 2g
wallnasyy aileueuh, inadno A myo Rings ~
SE tropes oxsautivas Ty ehede', Mae

cert 'hy ehnlordl ABBOT G4 sl ai bone Lt amelie
renga} whale byushadaned 3 leiaoes ge
\oywa « a ashes ely i Se

ool wradetidat sons) TS 8 qq (Hotel pee
an dloitns

es

Linnaeus’ Philosophia Botanica
translated by Stephen Freer
Oxford University Press
Paperback edition 2005 (ISBN 0 19 856934-3)
(translation first published in hardback in 2003 — the paperback edition incorporates a few
minor corrections)

Carl Linnaeus was one of the towering figures of
eighteenth century science, renowned as the father
of binomial nomenclature and commemorated in the
several Linnean Societies, including our own.

Apart from acknowledging his historical significance
why would anyone today read Linnaeus in translation?
I would argue that there is much to learn from such an
exercise, not least because it should inspire humility
—in many respects Linnaeus was the very model of a
modern academic — and when it comes to pedagogy
there has really been little change over the last two
hundred and fifty years.

Although remembered today as a taxonomist,
Linnaeus was a long standing teacher at the University
of Uppsala where he attracted record audiences to
his lectures. Students and former students remained
important to Linnaeus’ work — in this he was in
marked contrast to Darwin who remained outside
academia and worked alone. Daniel Solander, who
accompanied Joseph Banks to Botany Bay, and who
is remembered in Cape Solander and a memorial
garden in the Royal Botanic Gardens, was a student
of Linnaeus, but unlike his teacher was a reluctant
publisher and did not himself describe the many
Australian plants he collected.

Linnaeus had broad interests in what today we would
call biodiversity, and was a pioneer in zoological
systematics as well as in botany, but it is clear that his
main fields of interest were botanical.

In 1736 Linnaeus had written Fundamenta
Botanica, consisting of 365 aphorisms on matters
botanical. Philosophia Botanica was published, in
Stockholm and Amsterdam, in 1751. It consists of
the 365 aphorisms of the Fundamenta, arranged in
12 chapters, but each aphorism is now followed by
explanatory text.

In this translation ‘Philosophia’ is rendered as
‘Science’, as the ‘Science of Botany’ is the best
explanation to a modern audience of the nature of

the book. (The modern concepts of science, and
scientist, had yet to be developed by William Whewell
— ‘Scientia’ translates as ‘knowledge’ which would
not completely encompass the content of Philosophia
Botanica).

The explanatory text which the Philosophia adds
to the Fundamenta are essentially lecture notes
— material which today, along with the illustrations,
would be made available to students via the web. As
lecture notes, they are in brief, almost staccato, point
form, and provide opportunity for scathing attacks on
the errors Linnaeus perceived in the work of others.
This is the sort of thing that can be done to spice up
lectures but would normally be absent from “serious”
scientific writing. Indeed such flamboyance is absent
in the much more serious Species Plantarum, the
commencement of modern botanical nomenclature,
published only two years later in 1753. There are
also numerous references to, and examples from,
Linnaeus’s other publications. This frequent self-
citation has been viewed as self-aggrandisement — not
quite in good form — but if the Philosophia is seen
as a set of lecture notes it is more understandable as
being Linnaeus showing his students that he had runs
on the board — his publication record showing that he
was at what we would now call the cutting edge of
research so that you could take what he said as being
right. The self-citation was a means of attracting the
interest of students rather than representing an ego
trip by the author.

The Philosophia also includes memoranda — notes
of practical instruction on matters such as preparing
herbarium specimens and making notes on collections.
These also show that, long before his time, Linnaeus
included as advice to his students the appropriate
Occupational Health and Safety warnings. (“Botanical
outings are arranged differently by different people:
with us, the following [arrangements] are usual.

Very light and very loose clothing, proper to
botanists, (where circumstances permit) and the most
appropriate for the business.................

BOOK REVIEW

The clothing of the herborisant, beside linen, should
be a short coat, very thin breeches extending from the
hypochondria to the heels; smooth shoes, a hat with a
very large brim, or else a sunshade, so that he turned
by the way, the warmth, heat or sweat”.

When one looks at photographs of late nineteenth
century botanists in Australia, dressed in heavy
tweeds, it is clear that Linnaeus’ eminently sensible
advice took a long time to become acceptable!)

Even today, the basic structure of the Philosophia
would provide a very good framework for an
introductory botany course, starting with a historical
review and introduction to the literature, before
exploring a number of topics in detail.

Linnaeus accepted that species had been created,
but he had a very good understanding of variation
within species and was at great pains to stress that
variants should not be elevated to the rank of species.
A whole chapter (IX) is devoted to varieties, and
the topic also arises elsewhere in the Philosophia.
Linnaeus recognized (section 306) the practical need
to recognize varieties.

“The use of varieties in gardening, cookery, and
medicine makes it necessary to recognize them in
ordinary life; otherwise, varieties do not concern
botanists, except in so far as the botanists bother
about them, so that the several species shall not be
multiplied or confused”’.

However, he contrasts the different taxonomic
treatments in zoology and botany (section 259)

“In the animal kingdom, no sensible person would
readily say that varieties are distinct species.

White, black, red, grey, and variegated cows; small
and large, thin and fat, smooth and hairy cows; no
one has said that there are so many distinct species.

Exresences, crowns of the head, and sutures of the
skull have demonstrated that dogs, whether Melitean,
spaniels, mastiffs, Greek, poodles, etc. are all of the
one species”

and suggests that one of the reasons for the proliferation
of species names by botanists was “Contagious

madness among lovers of flowers”.

“Definitions that pass off varieties as species are
erroneous” and as an “horrendous example” of this

jis)

bad practice Linneaus conducts a demolition of
Micheli’s treatment of Trifolium (the clovers).

Linneaus had a surprisingly detailed understanding of
the causes of variation within species, including light
(sun versus shade), drainage (water logging versus
dry), soil type and both disease and attack by insects.
He advocated an experimental approach to studying
variation (section 316. “Cultivation is the mother of
very many varieties and is the best means of testing
varieties”). It was a long time before such an approach
became common place in what developed as a very
observation based science.

Section 334 provides a remarkably succinct
introduction to ecology and biogeography. In the
discussion of variations in flora and vegetation in
relation to latitude there are indications of the ideas
subsequently developed by von Humbolt. The
lengthy discussion of the relationship between species
and habitats concludes with the observation that “So,
by mere inspection of the plants, the earth and soil
beneath can be discerned”, a concept which still
underlies a great deal of ecological survey. The next
section (335) provides an overview of phenology and
demonstrates an understanding of the role of factors
such as temperature and day length in determining
features such as germination and flowering, although
it was to be many years before physiologists elucidated
the mechanisms involved.

Students today are always anxious that their courses
contain material of practical value; it was obviously
the same in the eighteenth century, and Linneaus
obliged, witha final chapter (XII), entitled “Potencies’,
dealing with economic botany. Much of this material
is still relevant, and with the emphasis on natural
medicines would have renewed appeal today even
though some of the claims still need to be rigorously
tested. Nevertheless if Linneaus was correct in his
observation (section 341) that both tomatoes and
eggplant were “Maddening and narcotic with our
people” it could explain a great deal!

What does Philosophia tell us about the development
of Linnaeus’ taxonomic ideas?

The binomial system of nomenclature is essentially
complete and a great deal of the Philosophia involves
laying down nomenclatural rules and guidance,
although it has to be admitted that many of these
rules were subsequently ignored or bent. (Section
236. “Generic names should not be misused to gain
the favour, or preserve the memory, of saints, or of

Proc. Linn. Soc. N.S.W., 128, 2007

BOOK REVIEW

men famous in some other art. It is the only prize
available to botanists; therefore it should not be
misused” — to which one could add that it is also a
prize to zoologists and palaeontologists).

The binomial system is one of Linnaeus’ greatest
legacies. There are those who argue that it should
be abandoned as the old hierarchical system
of classification does not accord with modern
understanding of the relationships between organisms
derived from molecular studies, but for sheer
practicality it is unlikely to be replaced (Defences of
the binomial system are provided by, Wheeler (2004)
and Knapp et al (2004)).

The Philosophia clearly explains Linnaeus’ belief
that the basis of taxonomy should a Natural System
and illustrates his attempts to develop such a system,
based on appropriate invariant characters (and
rejecting classifications based on phenotypic variation)
and in particular on floral characters (relevant to
Linnaeus’ sexual system). The importance of natural
systems of classification (which, it would now be
understood as reflecting as far as possible phylogeny)
was soon accepted by most biologists. Although
Linnaeus’ approach to developing a natural system
was subsequently overtaken by newer versions, it
is remarkable how many of the taxa recognized by
Linnaeus have stood the test of time.

Linnaeus’ chapters on floral structures and breeding
systems in plants were major contributions to
biological science (and the lectures on sex, with their
colourful use of analogies, no doubt went down well
with his student audience — again, some things never
change). The chapter on sex (V) contains, among
numerous other details, probably the first published
data on the annual seed production of individual
plants, and this could be said to be the pioneering
work in plant demography — a field which did not
develop for another two hundred years. I was struck
by Linnaeus pointing out the occurrence of arils in a
number of species. This feature of seeds of so many
tropical species is found in a few European plants, but
is not mentioned at all in many subsequent northern
hemisphere textbooks.

Linnaeus explains clearly the definitions of many
features — leaf shapes, floral structures etc — thus
providing a consistent framework for all subsequent
descriptive studies. The definitions, and illustrations,
provide evidence of Linnaeus’ keen and careful eye
for detail. This attention to detail is also seen in the
distinction drawn between right and left handed

Proc. Linn. Soc. N.S.W., 128, 2007

climbers, long before Flanders and Swann drew
attention to the same phenomenon.

Although Linnaeus provides a broad overview of
botany for his students it is clear that he expected
that many of the examples he presents would have
been familiar (for those more exotic species from
foreign lands he provides a bit more detail). In
this regard, Linnaeus, if put before a 21% century
undergraduate audience, would find life a little harder.
Although the modern student would be familiar with
many topics unknown in the eighteenth century
(biochemistry, genetics, computing) the broad natural
history knowledge, and the ability to make detailed
observations, is perhaps much less well developed
(despite marvellous television documentaries, natural
history has become more a spectator sport than a
participatory one — as noted by Marren 2002).

The Philosophia concludes with the statement

“Tn natural science the elements of truth ought to be
confirmed by observation”.

This is as true today as it was two hundred and fifty
years ago.

REFERENCES

Knapp, S. Lamas, G. Nic Lughadha, E. and Novarino,
G. (2004). Stability or stasis in the names of
organisms: the evolving codes of nomenclature.
Philosophical Transactions of the. Royal Society
of London B. 359, 611-622.

Marren, P., (2002). ‘Nature Conservation: a review
of the conservation of wildlife in Britain 1950-
2001’. London, HarperCollins.

Wheeler, Q.D. (2004). Taxonomic triage and the
poverty of phylogeny Philosophical Transactions
of the Royal Society of London B. 359, 571-583.

Paul Adam

Tre)

repos et
Morr e in

awed), gaol .<«medimeks
cont ote alt ob Musi

gvoinies 4

Lent 9 «veg 2sscomd dager

cx» of, sudtceclo eidt-eebusr cid vol vasied

| blucw aimsesrg of eviqenexd oft Yo. yor terit

" VE a ow). 200) oa itiers caged

m bd: zebivorg,ehcsbeul fgpiasst

el f rider -1u) th eoSatie A (ora ee)

| fected tog itt seachuectmdarrobas

f w ttelana oshom ait dguodily
; od a ewreniny soho yam
ONC Geom eo eemae tna asd )

hae oybwwrons.roiatd

\othcad hat a nella iedo
! ! ighwueliovinm stiqesh)
“uit om ancl yroteil

Marr betidry Bite YEN Et BN

it OV web

: LuHOUG's 5
f ek lent nee eat
: bowen
: co :
i ae 8
al
. we
’ 7 « ja
bet ; ‘ ) )
f od LE :
‘ yawn
,
t » 4
\
Ae ‘
OT
+
”
=e 4
7 4 »
:
g

aA ion

egeyhae add wtutacreisconee Girembmeh |
od eer bheode da: ee Totet ogahhoatnsd aot olde
6 cele wm h dar? bbs bluos ono doidw ot — “besth
nPess jel samergolainy abn deri ate
at oa so? variation within Spec $25, inching eta
tates. eemgeni jl You tine vet comets ieieapnidraatT
Ublooske ti sd ere pieces asarmed By oko :
Hostéyeteaidarsrait-n bien stp) liesobaads i ed
jae (sUitobyidaos “Catiriesahe avutesitiedalo
airy nag tected aqud zentslircsltitoge) ipciatin
wee: 70) | duck. wth atuislemncknaity
te asensiots :baosigo: +d aingisdiinekaperitilso
COS tolesd Wiad babey sg ate enoleyz imo
44-002) \o 19 qqsnid ba
"7 eomarkably succinet
Joideth.aeaseni de sarigicuase rere givkeyiolith ts
init bluada ycionexnsd eipetkods tm
aawayns douaolivel of wqnugite auke Ber
‘bise)oumnibetatio. Lape vin serqorad to
(notary op opulqe bsneshwer tie biieanha gene
an! is&velos) ticle Lani ste relacocinsapat
Sune inovoetreuri sl : (gmake yedbeasd ‘exis
dostive} eas Leradiaesel 2 to:oien
Take aa wl ee gresaiios es, bObmereky
gaboid: tages vid bedtpbos cheer ital

Oat EUS y

les ise

TH itys
P bad OSL

et wen chiniw
‘ut moldy: rit

(tiger “TP

(elmuten eogmeedeaboos (doaorgts Jidasam
i MOAI, ZPVrAN: 6 bg data y Linsupe ciriasi
b Ot exe ort: le ieumitiwod: oidnalutonie
nnd Lo tan cel howe avad
nbosd Lae eswisitte tinal mo eriqars *
Ohlenoiwdiniogs: Wye sxwaeraly heb gum

oiutsaloiltiees) goaomadsvigols
on.ewgaiegs, larudul lirneeiga
eSrothais iebhasen and idle
oe Wigaivodt» riage
4hetql seal guano

yet lola
& risky cow trigseyy
ron cpl cones SOA
jnoordW t Bee

} i i $ * «gl ei
b-seleid che ee ob dado

wee: Oo nowiehwsg bow lage edltegoe
waseooeg ofl! Sct obese ot biped wartie Damea 38
teare--Dib ath char bkoed iqeroub Iasi nt
bi aa lr 200 bonito bwediose- ret qolay

eat ales tA sort 94; Tutn-g seeing eased
any aietsoqx to toda
J wot ext baud al eaissq? leoiged
rrediinss sdeupsedua grain 2: dhe dy Bonsnasaptoe :

amoodiae —e

Vusas te emonngab afl winasds nal axes seated
aul — i avionete din 2stypides bedl/odiseniredt
easdi ie nt Hevworsh rotteizned egritherenn
anotitea!} bom 2norietiieersd T)20ihista ovatitasab
w¢e ldtsino dw: imal euageeid forsbesbrys Shion] n
ott ci (ew olaidestahebnormbeninicdieishsant ——
bolasd fol bastigr-nemetedr asthe mestoated —

FTTH

Ss - - pages
Voen Or lmeiow to.cmisal

ied eg NAO

OOS VAG Wi 5d SB} 28 ni

ae oat

Lampreys: Life without jaws
Martin W. Hardisty (2006)
Forrest Text, Cardigan UK

Although my research interests have always been
with mammals, for many years I taught a course
“Vertebrate Zoology”. My very favourite non-
furry animals within this subject were not actually
vertebrates but more correctly Chordates - the hagfish
and lamprey. I devoted far more time to this small
group than was proportionate, probably because they
are so delightfully bizarre. The Agnathans (animal
without jaws) would no doubt be ignored by a
creationist or proponent of intelligent design, because
one must either accept a creator with a bizarre sense
of humour or a totally mad designer. Perhaps it was
simply the challenge of designing a predator without
jaws or limbs that preys on much larger fish, or an
animal that changes from a fresh-water herbivore to a
marine predator and back again.

I wish this book had been available when I was
teaching at university, but it might have extended my
hagfish/lamprey lectures even longer. Martin Hardisty
has gathered together in this one work everything that
is known or suspected about lampreys. Dr Hardisty
died, at the age of 94, before the book was finished
and it seems some of his colleagues, and perhaps
some of his many students, finished the work. In so
doing, they appear to have made minimal changes,
except for adding missing references, in order to
remain true to Hardisty’s original text. Unfortunately
they did not add an index, and the lack of an index in
a reference crammed with facts is close to infuriating,
especially these days when word processing makes
the production of an index relatively simple.

Lampreys do not occur in the tropics and their
distribution is strongly biased towards the northern
hemisphere, where there are 34 species as opposed to
four in the southern hemisphere. Interestingly, there
are none in Africa except along the Mediterranean.
The Australian fauna is typical of the diversity within
the group as a whole. Two species, Geotria australis
(the pouched lamprey) and Mordacia mordax (the

short-headed lamprey) are anadromous, with the
ammocoete larva undergoing metamorphosis to a
marine parasite, which returns to freshwater to breed.
Geotria australis is also found as an ammocoete
in New Zealand, Chile and Argentina. Mordacia
mordax is found in SE Australia only and is a true
parasite, feeding on the blood of the fish to which it
has attached with its oral sucker and its horny teeth.
Geotria australis is more of a true predator, feeding
off muscle of the attacked fish. The third Australian
species, Mordacia praecox (the precocious lamprey),
remains in the drainages of SE Australia throughout
its life cycle.

Most information in this book of course relates to
northern species, such as the one that finished off
King Henry I after overindulgence in lamprey stew.
The story of the invasion of lampreys into the Great
Lakes of North America is well documented here and
is a woeful tale of ecological disaster. There are many
useful lessons in this book and I recommend it for
general readers as well as specialists. And I will leave
it to each reader to decide if lampreys are fish or not.

M.L. Augee
Sydney 2006

awe) tuodtiv slit :ayorqemd
(200°) yiaitaeH .W nine he
AU negibiaD jzsT teen

ii bane om (vorqinel bobsled:

8 O 2eolentmiom umemebas eal sisseqmuhm
orth OF izoft of 2iuin ibulw tinea oariasn
(ent iz 2n boot olen 4 24s antesD
Haare l ain basins wom m
ing adleues 14 nt bowel et eolam

ibe taft orf To boald or) ao gribed) .cifemaq

2 lune at diw bedosite and

goibee? 1 sn) e to aoe af eiletiews efittesD
gs T .dei) bodoste of to oben fio
Pear vl: } obi aioshaoM 2ersoqe
POU : ; fc 1iOP uariond orl) fl Pratt Tad
siovy old ati

go 1 seurod to Aood ell m@ eoteendim tol
fis dei it oe ; | srsege modnor
pete i santeyluboheve tafe | gush gaia
lev ht e¥orgs of £ i to Whale aif
A ifrralt To etobe.l
(fi 7 9 Io slat hifsaw avi

tel a ae i Voeeel tutes

wes! rw 11 afl b 27ebast letensg
F Hons ond

+ fs 5 if

More 2p

oued 2yewle sve enon dowsseet yen devotiil
sewoo # tigue! | naesy vase wil Alemene
non sihuovel yov YM “ygelooS “etadars!
yilsutos ton stow tosidue aidt widiitw elamine (rt
Hengad od) - zstabrorl > yOosiwes ston ted estat
Heene 2irlt of smc sor wl beloved I power
vod? seusced yidsdory sreansyoqorq aw nedi ¢
lemins) znelisngA efT -onasid yilihiigifeb 02%
6 Yd betongi od idveb on Gitow (avn
a2uBoed .1gizsb Hrogitistal 16 Insnadqor 20 tek
ese onExid & iw Iotestd #999908 irlis
zew ti eqartsT ssaglesh here sgliaiot's 30 toma
tuoritiw olebow s unidgiesb to ogrsilels ort vi ;
ns to .lzit wsgie! dodo mo eyony tart admit 4

wae:
UOT

it F

8 of s1Ovidiall Tstaw-destt-« mowtt eognads teds Tene
<é
miuge toad bao 1olsbery sh
aaw | nadw sldalieve aosd bed dood eid di

wm babasixe oved tigint ni ind vinetovims te anifed
(ibeH rinaM sognel seve eowios yore! Heiae
niovs sow ono euls a retltegat berating f
mH 170 .eynqmal mods bepsoqzre to nwo
ted +? To Spe orld ts

eouusoliod 2if Fo seme ansoe IF B
st bedsiant sessile yea 2st to. Sine
Zopmato farniuion ovad oF tmotfqn ‘york
odin anibie wi tse
2 visit ef suv nb

* ; — wis ] aw +
psiignan zsw dood alt aw

bits

es al aH , W

MIS 1II90
“Shar fi

Visiaaunoicl? ase lene

nt gob BE to dost srt be J nt hbe ton bib x
grsurwasiil at aeots ai aioe? dtr beeitens a }
vlan orieesog bnow csi meab geen? vileinsqes

viov leisy sabai ae to nolfoubeng’
ban eoique off of wed fon eb eysiqet
mainen ot ebiewel foebtd yirnowe a soltudintdll
on bazo¢pypo ex esrioqe FE ote oreil) aro ae atl
sen) Visanevicin! srorqetrmd oradtene off nf
neotensibel sit gnois iqovxzo KOrRA ai snbir sé
aiiw yiiersvil ot To levi) 2! enwd) nsiiedewA ss
alone nnioel) 2amooqe of stontw aa query *
ab) xetrom fsbroM bra (yinqmel bedoueq el

tear?

SF aed

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 (M.L. 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, ‘and’ is spelled out (not &), and et al. is not in italics.

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, W.K. (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 1s 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 scanned at 600 dpi. Jpeg is not
acceptable and figures of any kind set in WORD are unacceptable. 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 WORD 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
from the website or from the Secretary.

11. It is helpful if authors suggest a running head of less than 40 characters.

258

{ pur tustien, scientific nomencleture, em @e-e @ Reeed te ee ageeglegs
or fawn the Secretary

ehertytul if authors sugges? © roy Gee? Gee Gee

ves coat Oy eeparate from the teat fle Sheet coend gee oreables ane

oe ag seg’ oF Green grmenated EXCEL or WORE Head I

cm theme Wty oO Meee al Sineecat ketene ity figures: noite: ling
. —— Ce cathe a required, @ aie Oe presented a5 a bur within the fign
f=! 9 « Oe odie) Qeteegpetien weoenions et onletge fig<wres a4 nece ands
, 4° @ ow epee eee si
eee ee SOF Ake scenes a .
(on ime at es WUE ge mapequete biesdnutngs inst Veale is stied
> weed we Re ae Cine Camas eeen tee te tate. i;
wt Aas © ad pay pee See Se amma paliery 00 have the
io amt kgmeee emia at Se Geeie de Of Gaels the legend, Migere
——- oe np Ce eal we Gee ’

a ir oe tm). aii ta—geein are nos et ie
=o tho Figen gee 4qumelle dheiype Gaetitte qi (geet ard cian

r dv ‘epend 6 meget & i @& pee, Peppy SeGeceeyy ne
iw vate! puller ip maa Ade GP o¢ aw Gee Gi one page aie

vin at tome on bles unless atentinty name oth dem cole
‘ace pe a Denier erouml hice
UN CEL tables ere acceptable, tut WORD @ eration

PROCEEDINGS OF THE LINNEAN SOCIETY OF N.S.W.
VOLUME 128

MITHSONIAN IN

INH A

3 9088 0132

Issued 23 February 2007
CONTENTS

1 Todarello, P. and Chalmers, A.
The characteristics of five species of hollow-bearing trees on the New South Wales
central coast.
15 Rose, S. and Martin, H.
The vegetation history of the Holocene at Dry Lake, Thirlmere, New South Wales.
57 Robbie, A. and Martin, H.
The history of the vegetation from the last glacial maximum at Mountain Lagoon, Blue
Mountains, New South Wales.
81 Kellermann, J. and Udovicic, F.
A revision of the Cryptandra propinqua complex (Rhamnacea: Pomaderreae).
99 Semple, W.S. and Koen, T.B.
Observations of insect damage to leaves of woodland eucalypts on the central
western slopes of New South Wales: 1990-2004.
111 Williams, M.C. and Wardle, G.M.
The spatial patterns of invading Pinus radiata.
3) 2583 Keith, D.A., Simpson, C., Tozer, M.G. and Rodoreda, S.
Contemporary and historical descriptions of the vegetation of Brundee and Saltwater
Swamps on the lower Shoalhaven River floodplain, southeastern Australia.
155 Holmes, W.B.K. and Anderson, H.M.
The middle Triassic megafossil flora of the Basin Creek Formation, Nymboida Coal
Measures, New South Wales, Australia. Part 6. Ginkgophyta.
201 Zhen, Y.Y.
Revision of Microplasma parallelum Etheridge, 1899 (Cnidaria: Rugosa) from the Mid
Devonian Moore Creek limestone of New South Wales.
209 Och, D.J., Percival, |.G. and Leitch, E.C.
Ordovician conodonts from the Watonga Formation, Port Macquarie, northeast New
South Wales.
217 Dargan, G.
First record of Thecostegites (Cnidaria: Tabulata) from central New South
Wales.
223 Percival, I.G. and Engelbretsen, M.J.
Early Ordovician lingulate brachiopods from New South Wales.
243 Huveneers, C., Otway, N.M. and Harcourt, R.G.
Morphometeric relationships and catch composition of wobbegong sharks
(Chondrichthyes: Orectolobus) commercially fished in New South Wales, Australia.
251 Book review: Linnaeus’ Philosophia Botanica translated by Stephen Freer.
255 Book review: Lampreys — Life without jaws
257 Instructions for authors.

Printed by Southwood Press Pty Ltd,
76-82 Chapel Street, Marrickville 2204

PROCEEDINGS

of
NEW SOUTH WALES

VOLUME 129

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 election
to the Society must be recommended by two members. The present annual subscription is $456.00.

The current subscription rate to the Proceedings is set at A$110.00 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 2006/2007

President: D.R. Murray

Vice-presidents: M.L. Augee, I.G. Percival, J.P. Barkas

Treasurer: 1.G. Percival

Secretary: J-C. Herremans

Council: A.E.J. Andrews, M.L. Augee, J.P. Barkas, M. Cotton, G. Edgecombe (to March 2007),
M.R. Gray, J-Cl. Herremans, D. Keith, R.J. King, H.A. Martin, E. May, D.R. Murray, P.J. Myerscough,
LG. Percival, J. Pickett (from April 2007), S. Rose, H.M. Smith (from April 2007) and K.L. Wilson
Editor: M.L. Augee

Assistant Editor: Elizabeth May

Auditors: Phil Williams Carbonara

The postal address of the Society is: P.O. 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: a diagram of a portion of the abdomen of the ironic fly Ironomyia nigromaculata from the
paper by David McAlpine, page 28 this volume.

PROCEEDINGS
of the

LINNEAN
SOCIETY

of
NEW SOUTH WALES

For information about the Linnean Society of New South Wales, its publications and
activities, see the Society’s homepage

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

VOLUME 129
March 2008

bith 2olteoi hud ett ze huis# wait Yo visi2d2 npeantt set

NEW SO UTR Ww ae 4 :
iS ote: O47X <0 hy

Rae
rae

| eta wo = et greet 10

pocan Shartcis Pel) thin Boies artes

tie een af the Lonpram. Matlesy Lac
Microtiolugy af tine: lL alveeity ai Sy dite
cnibslishany Ch Poneiogediogs. 14 iaahale fabian

atihie moetungs, ineludiiie te Meantal Sit i
aes Memanal Lectiw deliverdd tie WOE,
Nien! (f sore pr «al Que) eeseeieae ? ;

: Hon ine 12S eine Toe Dae " ofeeciaind 9
ry Wh cei. The crema @ehnas ST ad 7
crates

. “sgaqoinolt Abi 32 Mol She lest eon ume

crcitt CHOI SOQnNti ~\UG INDI. 8I6 WW ioe
(> pus | ie Proven ere symdabte v1 4 poche ue

hit ie aia t= vu i , NWihen rhe Scul Tiary

bp aee® vl othe G. Pagecvenie dies
ito, & Maw, ti vet i ;
La ei CT Oe: Ay preg? A es ».. #7 wit

by LA \ iad Cy rive
‘mt se
aie a! ~Litiensar itn, tra ot
oe
\ ait! yy ¥ ne 7 ' j
pea.
rthors OF th ot we tinic. fly dees
‘ - -
rate o> il _ 7 7
- Ge 7 a
os nas a

The Ecology of Episodic Saline Lakes of Inland Eastern

Australia, as Exemplified by a Ten Year Study of the Rockwell-

Wombah Lakes of the Paroo.

BRIAN V. TIMMS

School of Environmental and Life Sciences, University of Newcastle, Callaghan, NSW, 2308, Australia.

Email: brian.timms@newcastle.edu.au

Timms, B.V. (2008). The ecology of episodic saline lakes of inland eastern Australia, as exemplified by a
ten year study of the Rockwell-Wombah Lakes of the Paroo. Proceedings of the Linnean Society of New
South Wales 129, 1-16.

Studies on salt lakes are mostly snapshots of their unique characteristics and relationships. Longer term
studies provide different perspectives on variability in hydrology, salinity and biological communities.
Such a study on five lakes near the Paroo River in the northwestern Murray-Darling Basin showed most
hold water episodically for about 80% of the time, but each fluctuate over a characteristic salinity range
: unnamed lake 0.6 — 19 gL1, Wombah 1.2 — 30 gL7, North Blue 0.3 — 31 gL", Mid Blue 0.7 — 103 gL",
and Bulla 1.8 — 262 gL. Generally, instantaneous biodiversity is low and not necessarily correlated with
salinity, but unlike southern seasonal salt lakes, species accumulation lists are long, approaching 80 species
of invertebrates, 50 bird species and a few fish species per lake. Diversity is promoted by salinity fluctuation
and habitat heterogeneity. Most species reach peak abundance in any season as long as conditions are within
their physiological salinity tolerances. Invertebrate fauna is of inland affinities, but with some localized
substractions and additions explained by hydrology and/or salinity; waterbird numbers are influenced by
events elsewhere in Australia as well as by local conditions. Like most naturally salinised lakes, production
can be high, especially at low to moderate salinities and algal blooms occur naturally from time to time.

Manuscript received 18 April 2007, accepted for publication 19 September 2007.

KEYWORDS: benthos, biodiversity, episodic lakes, fish, littoral, salinity variability, waterbirds,

zooplankton.

INTRODUCTION

Over the last few decades much has been learnt
about the numerous salt lakes in Australia. Limno-
logical summaries are available for Victoria (Williams
1981) and the eastern inland (i.e. Queensland, New
South Wales, adjacent South Australia) (Timms 2007)
and southwestern Western Australia (Pinder et al.
2004a), Tasmania (De Deckker and Williams 1982),
and southern South Australian lakes (De Deckker
and Geddes 1982; Williams 1984. In essence, most
Australian salt lakes are shallow, intermittent (either
seasonal or episodic), chemically dominated by NaCl,
markedly alkaline, and have a crustacean-dominated
fauna that decreases in diversity with increasing
salinity (see above references). Additionally, unlike
salt lakes in the northern. hemisphere, there is
considerable regionalisation of the fauna (Williams
1984; Timms 2007).

Australian saline lakes are changing due to

various adverse environmental pressures, as reviewed
by Williams (2002) and Timms (2005), In Australia
the greatest problem is secondary salinisation and it
is manifest mainly in southwestern Western Australia
(Davis et al. 2003; Halse et al. 2003) and to a lesser
degree in some areas of southern Victoria and South
Australia (e.g. Lake Baird, Pellana Lagoon, Lake
Wangary on Eyre Peninsula (author, unpublished
data))(Fig 1 in Timms, 2005). However in the remote
inland, saline systems have remained unaffected by
anthropogenic secondary salinisation.

One such inland area is the middle and lower
Paroo River catchment in northwestern New South
Wales and adjacent southwestern Queensland (Fig.
1 in Timms, 2006). Here there are many freshwater
wetlands but a few are naturally salinised. Although
surrounding terrestrial environments are degraded
after 120 years of grazing, aquatic habitats are almost
pristine (Timms 2001a). Study of them provides a
‘control’ for investigations into secondary salinised

ECOLOGY OF EPISODIC SALINE LAKES

8 7

Nites a

North Blue Lake

Lake Bulla

| Z} Unnamed Lake

7

|
Channel from Pay6o River

Rockwell
‘Station

Queenslan

|
Channel to Cuttaburra Creek

Figure 1. Map showing the five lakes on Rockwell and Wombah stations, southwest Queensland.

systems. However, they have accumulated their salt
over millennia, so the rate of salinisation is different
to that in anthropogenically affected systems. Also,
the latter are affected by multiple degrading factors
rather than just by salt increase (Davis et al. 2003).
Much of the existing data on saline lakes were
collected on single field trips, or perhaps at best,
a few over a year, so that long-term fluctuations
associated with variable weather patterns have gone
unrecognised. Yet these fluctuations could be a major
difference between primary and secondary salinised
lakes (Hudson et al. 2003). So understanding of inland
lakes over a longer time frame than provided by most
studies will provide better understanding of some
of the impacts in secondary salinised systems. The
present study aims to do this in a context of other
saline lakes in the Paroo and of Australia in general.

THE LAKES

The five lakes lie on a creek line (called Number
10 creek upstream of North Blue Lake) that flows
southwestward to join an anabranch of the Paroo
River on the New South Wales — Queensland border
40-50 km east of Hungerford (Fig. 1). The creek is
channelled between the upper three lakes (North Blue,
Mid Blue and Bulla) which coalesce when full, but
is obliterated by dunes downstream of Lake Bulla. It
reappears as a small stream near Lake Wombah (Fig.
1). Paroo floodwater often enters Lake Wombah and
leaves by the same route, but rarely reaches upstream
to Lake Bulla. The last times this occurred were in
1974 and 1990 (P. Tuite, pers com.). The upper three
lakes fill from Number 10 Creek, together with local
streams from the east. The unnamed lake fills entirely

Proc. Linn. Soc. N.S.W., 129, 2008

B.V. TIMMS

Table 1. Geomorphic and physicochemical features of the five lakes

Lake

Wombah Lake

0.2 (1.8)

Unnamed

Mid Blue
Lake

Nth Blue
Lake

2.3 (3.6)

Lake Bulla

4.4 (4.9) 3.3 (3.9)

Mean water
Temperature (° C)

Mean Turbidity + SE
(FTU)

34+ 10.9 Bass 27) 14+3.5 W325i)
Mean pH + SE 3) 22 (71 8.7+0.16 8.8 + 0.19 8.9 + 0.12 8.8 + 0.23

* Figures in brackets are heights of highest water marks, but were not recorded during study.

from local streams and Lake Wombah receives most
of its water from the Paroo River.

METHODS

Areas of the lakes were established from aerial
photographs and depths from gauges in the lakes
and the height of stranded beaches/wave cut notches
determined with a dumpy level (Table 1). The later
relate to intense local rainfall and large Paroo floods
such as occurred in 1974 and 1990 (P. Tuite, pers.
com.).

The lakes were visited at approximately three-
monthly intervals from July 1995 to June 2004 for
the purposes of determining some physicochemical
parameters and sampling zooplankton and littoral
organisms On each visit a surface water sample from
about 50m offshore was taken for the immediate
measurement of temperature with a mercury
thermometer, pH with a HANNA HI8924 meter and
later measurement of total dissolved solids (TDS) by
gravimetry, with turbid waters being allowed to settle
for many months in a sealed container. Turbidity was
measured spectrophotometrically in the laboratory at
450 nm with the results recorded in Fittou’s Turbidity

Proc. Linn. Soc. N.S.W., 129, 2008

Units. Five times during 1997 and 1998 nitrate and
phosphate was determined in the field on the water
samples from Wombah, Bulla, Mid Blue and North
Blue using a HACH DR/2000 spectrophotometer
and method 8171 for dissolved nitrate and 8048 for
dissolved phosphate. These measurements were made
during a hyposaline and a meso/hypersaline stage.
Nutrients were measured once only (July 1998) in the
unnamed lake.

Zooplankton and littoral organisms were
collected with nets of mesh size 159 um and 1mm
respectively, identified and counted as outlined in
Timms and McDougall (2004). Benthos was sampled
with a Birge-Ekman grab just once in the four larger
lakes— Wombah in April 2004 and Bulla, Mid Blue
and North Blue in December 2001. Five grabs were
taken from each lake near the deepest part, sieved
onshore through a 0.4 mm mesh and sorted while
organisms were alive. Fish were caught sometimes in
the littoral net and often in yabbie traps set for an hour
or two to catch Cherax crayfish. These traps were not
employed regularly or in any pattern.

Waterbirds were counted from January 1998
onwards. Shore based total counts were made with
binoculars and/or a spotting scope on each visit.
Counts were always made from the same vantage

ECOLOGY OF EPISODIC SALINE LAKES

points each time, and covered the full surface area
of all lakes except Lake Wombah. In this lake only
the northern half was surveyed, for if the southern
part was included it would take several hours to
count the birds and impossible to account for bird
movements because of the terrain. Small waders
were underestimated in Lakes Wombah and Bulla
when full as distances were too great for accurate
determination.

RESULTS

Physicochemical Conditions

Average annual rainfall for the 10-year period
was 366 mm, but yearly totals varied widely and
included unusually wet years of 1998 (618 mm) and
2000 (622 mm) and the particularly dry year of 2002
(92 mm)(Fig. 2). Major local inflows to the top three
lakes from intense local rain occurred in April 1998
and March-May 2000, and Paroo floods filled Lake
Wombah in July 1998, March 2000 and February
2004. Between major fillings lake levels fell slowly
with evaporation or rose slightly with minor inputs,
so that four of the lakes held water for ca 80% of the
time (Table 1). High salinities were experienced only
briefly during March to December 1997 (to April
1998 in Bulla) and again variously during much of
2001-2002 in Wombah, Bulla, Mid Blue and North
Blue (Fig. 3). The small unnamed lake held water for

a few months in 1998-99 and again in 2000, both as
a consequence of unusually heavy rainfall in those
years (see above).

All five lakes varied widely in salinity, from
fresh in all cases to hyposaline (i.e. to ca 20gL") in
the unnamed lake, to mesosaline (i.e. 20-50 gL) in
Wombah and North Blue, to hypersaline (i.e. >50 gL
') in Mid Blue and very hypersaline (> 200 gL" in
Bulla (Table 1). Mean salinities are deceptive due to
the short time the lakes spend at higher salinities (Fig
3) so that median salinities are more representative
than mean salinities of typical conditions in the lakes
(Table 1); in this respect all lakes are often hyposaline,
with Bulla the most saline and the unnamed lake the
least.

Mean water temperature for the ten years was
21.6 °C, with a minimum of 8 °C and maximum
of 36 °C. Lake waters were generally clear, except
following major inflows, or when shallow and wind-
stirred, or in Lake Wombah’s river-derived water
(Table 1). Variability was greatest in the two lakes
(Wombah and North Blue) receiving mostly flood
water and least in Bulla where abundant salt helped
to settle colloidal clay (Table 1). All five lakes had
well buffered, markedly alkaline water, except again
for some variation in the two lakes receiving most
floodwater (Table 1).

Nutrients were high, around 1-2 mgL" nitrate
and 0.2-0.4 mgL! phosphate, with highest values in
Lake Bulla (Table 1).

pn | Lal 102 gl-'*) 158 gL!
' J ) 193 gL i
il
jibe [ot
== | Bulla} =| ‘i
tial! "
Poteet ‘
= Rvqabl
a | |
— 504 f
= { | i'
az / . | Mid Blue ne
/ _ 1
OMB ; North Blue a
25 4 ’ layers
unnamed ae
=, a“ e :
+ -”. ¥ 1
a. . Sa
0 T T T T = T pe ees en Sas LE oT; T Wo, Tt beg he boro crrerai aise
95 1996 1997 1998 1999 2000 2001 2002 2003 04
Figure 3. Fluctuations in TDS in the five lakes.
4 Proc. Linn. Soc. N.S.W., 129, 2008

B.V. TIMMS

Monthly rainfall (mms)

200

140

100

50

i

2000

il)

a a boll
1997

1998 1999 2001

Figure 2. Monthly rainfall at Rockwell station homestead 1996-2004.

]

2002

| nt
Oe
mh | iat
EL

2003

Proc. Linn. Soc. N.S.W., 129, 2008

ECOLOGY OF EPISODIC SALINE LAKES

Aquatic Plants

Aquatic macrophytes were most common
and persistent in the clear waters of Lake Bulla,
where Myriophyllum verrucosum Lindl, Lepilaena
bilocularis Kirk, and Chara spp. were common at
lower salinities (<30 g/L). At higher salinities (30-60
g/L) the dominants were Chara sp. and Ruppia sp.
These species occurred in the other lakes, but were
less abundant, except occasionally in North Blue. A
new species of Chara occurs in Bulla and Mid Blue
(A. Garcia, pers. comm..) In summer and autumn,
filamentous algae often shrouded plants, and appeared
to be associated with their demise for the season.

Zooplankton

At least 37 taxa occur in the five
lakes, with 12-28 species per lake (Table
2). The dominant freshwater species were
Boeckella triarticulata, Daphnia angulata
and D. lumholtzi, while the most common
saline species were Apocyclops dengizicus,
Daphniopsis queenslandensis, Cyprinotus
sp. and Diacypris spp. (Table 2). There
are broad similarities between average
percentage composition between the lakes
(Table 2), though the most intermittently filled
unnamed lake had the highest percentage of
clam shrimps (14.3%) and the most saline
lake (Bulla) had the highest percentage of
saline species (69.9%), followed in order by
Mid Blue (56.0%), North Blue (44.4%) and Wombah
(14.7%). Freshwater eulimnetic species with limited
salinity tolerance (e.g. Calamoecia lucasi, Daphnia
lumholtzi, Daphnia angulata, Ceriodaphnia spp.,
Diaphanosoma spp., Moina australiensis) comprised
<22% in all lakes and were lowest in Bulla (10.2%)
and the unnamed lake (1.1%). While salinity
influenced a few species, momentary species richness
was not related significantly to salinity in any of the
lakes (Wombah r = -0.2148, n=27; unnamed lake r =
0.0832, n=7; Bulla m-0.4885, n=31; Mid Blue r = -
0.1425, n=35; North Blue r = -0.2505, n=29).

Typically, composition and dominance varied
markedly between sampling dates in each lake and
there was little seasonal repeatability over the ten
years. Exceptions to this were Daphnia lumholtzi
which often peaked in autumn in the fresher lakes,
and Daphniopsis queenslandensis which was
most common in winter-spring and Moina baylyi
in summer-autumn, when lakes were saline. The
most predictable occurrences were of saline species
when lake salinity was elevated (e.g. Daphnia n.sp,
Cyprinotus sp. at lower salinities, and Apocyclops
dengizicus and Diacypris spp. at higher salinities).

Table 2 (RIGHT) . Zooplankton of the five lakes,
showing mean percentage composition and maxi-
mum salinity for each species.

The amount of zooplankton in each sample
varied greatly, with means ranging from 22.2 (Bulla)
to 4.1 (North Blue) mls per minute of standardised
collecting (Table 3). Variablility was greatest in the
more saline lakes (Table 3). Dense populations were
relatively uncommon but sometimes occurred in
winter, soon after filling, or when the lakes increased
in salinity as they dried rapidly (as a consequence of
long periods of warm dry weather), but this was not
significant statistically.

Table 3. Measure of standing crops

Zooplankton Benthos

Lake

a oe gm-2 + SE
hanno

15.5 £5.5

45401
49-004

* densities based on lakes being on average half full

3)3)-5) ae 0).S,

Littoral Invertebrates

At least 78 species occur in the five lakes, with
an accumulated species richness in each lake between
36-58 species (Table 4). Mean momentary species
richness in each lake ranged from 6.4 to 10.9 (Table
4). The dominant species in all lakes was Micronecta
sp., with the two ostracods Mytilocypris splendida
and Trigonocypris globulosus and Anisops gratus
and other backswimmers and boatmen common in
most lakes, except in the unnamed lake. All common
species had wide salinity tolerances (Table 4). Though
Tanytarsus barbitarsus had by far the highest salinity
tolerance, it was common only in mesosaline waters
and above.

Beetles, although diverse, were not common in the
lakes, and larvae were uncommon too. Odonatans were
most abundant in lakes with extensive macrophytes
(Bulla, Mid Blue). Ephemeropterans were most
abundant in the lower salinity lakes (unnamed, Mid
Blue and North Blue). River dominants such as large
crustaceans Macrobrachium australiense and Cherax
destructor were hardly encountered and then only in

Proc. Linn. Soc. N.S.W., 129, 2008

B.V. TIMMS

SPECIES max. Lake unnamed} Bulla Mid Blue | Nth Blue
salinity | Wombah} Lake Lake Lake Lake

ae a
Lae
<01
istislag

Boeckella triarticulata Thomson 54.5 18.5 16.9

Daphnia n. sp. : 5 :
| Daphniopsis queenslandensis Sergeev | 59 | 19 [| | 95 | 95 | 88
~aseipieinac Srl Cl
Fev Nl
Sorcerers 7 | NE ae
Emme ae | es) ail oa |
ee oe AD OE TA EE
oe | | 3 |_ ot) fe) | aaron foooan |
eee PN |NSSL OO. is | 02 | olen ok |
ere el ee | oma noniannh
ie A TOS ETT
| 02
ha a
eo 1 |

N i)
— i)

3.0

I= || S) S)
oo | 4 ies)

Trigonocypris globulosa De Deckker

0.2
4

E

Proc. Linn. Soc. N.S.W., 129, 2008 7

ECOLOGY OF EPISODIC SALINE LAKES

Table 4. Littoral invertebrates in the five lakes. All numbers expressed as mean log abundance
(Log 1 = 10) individuals per 15 minute collection).

Nth Blue

| Take
| Wombah_
Bem
(cca
o> Fee
fie Te
a_i
eae
| 0.04

Ti
N
a

Anostraca
Branchinella australiensis (Richters)
Branchinella buchananensis Geddes 3
Spinicaudata
Caenestheria sp. 3

ard
ae
ie
ied At
eS
Caenestheriella packardi Spencer & Hall 0.04 0.03
mse al
fell
mete
iret

Eocyzicus parooensis Richter & Timms 3 Cre ws.
Eocyzicus 0 sp 1 res hall oe eed
Limnadopsis birchii (Baird) 3 oo :
Ostracoda

Mytilocypris splendida (Chapman) men

Trigonocypris globulosa De Deckker 68-05 dt |e OO | Sesigene- | memo ee 1.16
Decaopda tH]

Cherax destructor Clark ered

Macrobranchium australiense Riek
Ephemeroptera

Cloeon sp.

4
Tasmanocoenis tillyardi (Lestage)

Odonata
4
30
12
12
12

Austrolestes annulosus (Selys)
Ischnura heterostricta (Burmeister)

Xanthoagrion erythroneurum Selys

Diplacoides bipunctata (Brauer)
Hemicordulia tau (Selys)
Hemianax papuensis (Burmeister)

Trapezostigma loweii (Kaup)

Anisops calcaratus Hale

[006
[<r
[006
pre
—_
atl
P01
[035
oe ee mr
=
aaa
[008
a
maerras
[or
[000

26
13
12
94
17
17
17

Anisops gratus Hale
Anisops stahi Kirkaldy
Anisops thienemanni Lundbald

Micronecta sp.

19a

Agraptocorixa eurynome Kirkaldy
Agraptocorixa hirtifrons Hale
Agraptocorixa parvipunctata Hale
Sigara sp.

Trichoptera
Notolina sp.

5)
5
3
4
3
5
Oecetis sp. a 5
5

6444
jh]
su
[5 7)
eral
ea
[= 20]
[2a
[2]
[es
itis 246)
[am
Hemiptera [alas iad
[me ol
ean
Le]
[2s
[84]
Le
ns
are]
ena
4]
Sh
fame al
[ae

Oéecetis sp. b

8 Proc. Linn. Soc. N.S.W., 129, 2008

Triplectides australicus Banks

Cybister tripuncatus Olivier
Eretes australis (Erichson)
Hydroglyphus leai (Guignot)
Hyphydrus elegans (Montrouzier)

Sternopriscus multimaculatus (Sharp)
Haliplus sp.

Hydraena sp.

Hydrochus sp.

Berosus approximans Fairmaire

B.V. TIMMS

ey Sp aba Fe
eee are

iT
a |
sib fe

0.04
.08

0.0 0

Berosus australiae Mulsant

[oie]
EAT 9

Berosus macumbensis Blackburn
[BerosusnumsMackesy | | id os |
| BrocirusandersoniBuckoum | 7 | 006 | oa [ on | om |
| iicchnus macutcers Wakes) | es
aviebeiparimae | 7 |__| an [an [| an
meio Gon | 8 | | an [om | oe
imam | | an | [om [an [|
ree aor po
rr ee
F Bectialarae Berosussp. | | oo |_| oo | 00+ | oor
imma pat | oa |). | oo
Re Chan tiv pal) pe |
~ each, a a aa ATEN
ee es | el oe
isan | 30 | or | ome [om | 009 | 0
ppm tf 026 || 025) oan | 035

Tanytarsus barbitarsus Freeman 0.12

Berosus debillipennis Blackburn

unidentified ceratopogonid larvae
Aedes sp.
Anopheles sp.

unidentified pyralidae larvae

Proc. Linn. Soc. N.S.W., 129, 2008

ECOLOGY OF EPISODIC SALINE LAKES

Table 4 continued

ESOS salinity

Hydrocarina
Arrenurus sp.
Eylais sp.
Hydrachna sp.
Limnesia sp.
Poina sp.
unidentified water mite

Gastropoda

ies)

Coxiella gilesi (Angas)
Glyptophysa sp.
momentary species richness (MSR) 6
MSR SE

number of species

Lake Wombah. Coxiella gilesi was abundant only in
Lake Bulla and Mid Blue Lake. Chironomids were
present in most lakes and were no doubt more diverse
than indicated in Table 4. The unnamed lake was
the most distinctive of the five lakes, with a fauna
dominated by large branchiopods.

Momentary species richness correlated negatively
with salinity in three of the five lakes: Wombah r = -
0.3700, n=27, ns; unnamed lake r = -0.6135, n=7 ns;
Bulla r = -0.6245, n=31 significant at P< 0.01; Mid

max. Lake
Wombah

<0.01
0.22
0.13
0.17

4

Nn

leinkoniers| Pe akg
unnamed Bulla Mid Blue | Nth Blue
lake Lake Lake
[nae gy earress
co

nr
ET
Pian

0.79

0.21
0.03
0.30
9.5
58

oS
aN
NO

0.31

lc
36

Blue r= -0.5072, n=35, significant at P < 0.01; North
Blue r= -0.4853, n=29, significant at P< 0.01.

to
i
2
NQ

?
—

Benthos

Based on limited sampling, benthos of
unvegetated offshore zone of the lakes was abundant
but communities were simply structured (Table 5).
Chironomids dominated, with ceratopogonids and
large ostracods present, and an oligochaete in one
lake. Biomass was by far the greatest in Lake Bulla.

Table 5. Benthic invertebrates of four of the lakes (numbers per m7’).

Lake Wombah Bulla Mid Blue Nth Blue
TDS(gL") 15 17.1 6 30.9
Depth (m) 5 2-—2.4 1—1.55 0.2
Dero digitata (Muller) Mp) 38)
Geen splendida 177 + 166
ae ie globulosa De 1320 £219
Chironomus sp. 6600 + 895 100 + 62
Procladius sp. 310492 1364 + 182 BPS NEE 28 3720 + 267
Tanytarsus sp. 25655 + 944
Ceratopogonid larva 280 + 55 250 + 108 324 + 83
Total numbers (m7) 6910 + 768 30796 + 712 3622 + 544 4044 + 248
Biomass (gm?) Oise ile 33.5+0.9 4.5+0.7 4.9+0.4

10

Proc. Linn. Soc. N.S.W., 129, 2008

B.V. TIMMS

Fish

A variety of fish occurred intermittently in
the lakes. Bony Herring (Nematalosa erebi) and
Spangled Perch (Leiopotherapon unicolor) were
present in Lakes Bulla, Mid Blue and North Blue
from at least mid 1998 to 2001 and Mosquito fish
(Gambusia holbrooki) and Carp (Cyprinus carpio)
were common in Lake Wombah over the same period.
Other fish could have been present in these four lakes
as the sampling technique was not designed to catch
all species. No fish were seen in the unnamed lake.

There was a major fish kill in Lake Bulla in
March 2001 at the same time as there was a bluegreen
algal bloom. Carp in Lake Wombah died en masse in
December 2001 as the salinity reached 30 gL". Fish
in the Blue lakes seemed to disappear slowly over
2001, after which piscivorous birds were rarely seen
on these lakes.

Waterbirds

Forty-eight species of waterbirds were seen on
the lakes; Wombah had 34 species, the unnamed
lake 15, Bulla 43, Mid Blue 41 and North Blue 38.
The most common species were Grey Teal (Anas
gracilis)(particularly in Wombah), Pink-eared Duck
(Malacorhynchus membranaceus)(particularly in
Bulla) and Eurasian Coot (Fulica atra)(mainly in Bulla
and North Blue). Other species present in appreciable
numbers included: Hardhead (Aythya australis), the
Australasian (Zachybaptus novaehollandiae) and
Hoary- headed (Poliocephalus poliocephalus) grebes,
Black Swan (Cygnus atratus), Australasian Pelican
(Pelecanus conspicillatus), Little Black Cormorant
(Phalacrocorax sulcirostris), Little Pied Cormorant
(Phalacrocorax melanoleucos), Silver Gull (Larus
novaehollandiae), Black-winged Stilt (Himantopus
himantopus) and Red necked Avocet (Recurvirostra
novaehollandiae). Only Black Swans bred on the
lakes during 1995-2004, mainly on the islands in
Bulla.

Numbers fluctuated greatly between lakes and
observations (Fig. 4 and Table 3). Mean numbers of
birds on the lakes varied between 18.6 ha! (North
Blue) and 7.3 ha! (Wombah) (assuming the lakes were
on average three-quarters full)(Table 3). Variability
was greatest in Bulla which had the greatest salinity
range, and there was no correlation between species
richness or bird numbers with salinity (r = 0.1364 and
r = -0.0839 respectively).

DISCUSSION

Naturally salinised lakes in the eastern inland of

Proc. Linn. Soc. N.S.W., 129, 2008

Australia have many distinctive features as itemised
below, and with particular reference to the Rockwell-
Wombah lakes.

1. Hydrological regime and geomorphic factors
determine the presence of salt lakes and their salinity
fluctuations (Hammer 1986; Williams 1998a; Timms
2006). When hydrologically closed, 1.e. without
overflow, bodies of water tend to accumulate salt.
This explains the different salinity regimes in the
Rockwell-Wombah lakes. Lake Wombah is the
only one to be flushed by river flood water and then
evaporation concentrates the isolated waters. Turbidity
is greatest in Lake Wombah, due to the colloidal clays
of Paroo River water (Timms 1999). Lake Bulla is
generally the evaporative terminus of a blocked
stream system flowing onto the Paroo floodplain, so
it has accumulated the most salt and, associated with
this, its waters are the clearest. Every few decades
it overflows and presumedly salt is lost so that it is
only hypersaline as it nears dryness (similar to Lake
Wyara on the other side of the Paroo floodplain —
Timms 1998). Mid Blue and North Blue Lakes are
intermediate settling basins; differences between
them largely reflect the relative shallowness of North
Blue Lake. The unnamed lake is the most intermittent
of the five due to its relatively small catchment; salt
accumulation consequently is quite modest.

2. For episodic saline lakes even 10 years is
not long enough to encounter all the variations in
environmental conditions. Judged from comparative
rainfall averages at Rockwell during 1995-04 (366mm)
compared to long term averages for the area (325 mm
over 105 years at Boorara nearby — A. McGarth, pers.
com; 302 mm over 75 years at Warroo nearby — M.
Dunk, pers. com.), the Rockwell-Wombah lakes were
probably full for longer than a long-term average
during 1995-2004. On the other hand, if the lakes had
been studied during the much drier 1910s to 1930s
they would have held only a little water occasionally
and would have been more saline on average. An even
longer time frame would encompass climate change
and quite different hydrological conditions (Bowler
1983; Pearson et al. 2003).

3. While there is a classic negative relationship
between diversity and salinity in salt lakes (Hammer
1986; Timms 1993), other factors may mask it
(Williams 1998b); in the case of these lakes only the
littoral assemblage in Lake Bulla and the two Blue
Lakes were significantly correlated with salinity.
None of the other influencing factors considered by
Williams (1998b) seems be important in these lakes.

11

"€007-8661 SOL] 94} JO AMOy uy sxoquinu p.arq UY sUONLIALA “p 2.1NSTY

ECOLOGY OF EPISODIC SALINE LAKES

Number of birds per lake

OO00L
OO0SL

un
=
=
oS

LO 10 00 00 00 O00 00 66 86 66 86 86
r Jey ueraonN das inr Ae el 99q Inrudy das uer

= a |

Ss © 2m

| =

=
=
o

=| =

=35 2
O

o

o oo

ho = =

or

Ns aw
=

= = terete teeta @

= wm

o Oc] El

i) Qo fs)
=

=) = yl a8

ie & =
wm

= = To ooo

Proc. Linn. Soc. N.S.W., 129, 2008

B.V. TIMMS

Indeed, the wide salinity tolerance of many species
is in accord with that in the remainder of Australia
so that it is only where lakes exceed 50 gL" that
salinity has a significant impact as a determinant of
community structure (Williams et al. 1990; Williams
1998b).

4. Salt lake fauna is regionalised in Australia
(Williams 1984; Timms 2007), though all areas share
a dominance of crustaceans, particularly Parartemia
spp., Apocyclops dengizicus, Daphniopsis spp., and a
host of ostracods (Diacypris, Heterocypris, Reticypris,
Platycypris) including a few mytilicyprinid genera
(Australocypris, Mytilocypris, Trigonocypris)
(Williams 1981; Pinder et al. 2002, 2004a, 2004b;
Halse & McRae 2004; Timms 2007;). Lakes in more
propititious climates have higher crustaceans such
as amphipods (Austrochiltonia spp) and the isopod
Haloniscus searlei which have no resistant stage
and need at least dampness to survive dry periods,
and calanoid copepods (Calamoecia spp.) and more
ostracod genera (Williams 1984; Pinder et al. 2002).
The eastern inland has a distinct regionalised fauna,
though as more is being learnt about inland Western
Australian fauna (Pinder et al. 2002, 2004a; Timms
et al. 2006) many of the distinctive species of the
east are being found in the west as well. Examples
include recent discoveries of Eocyzicus parooensis.,
Daphniopsis queenslandensis, Moina _ baylyi,
Celsinotum spp., and Trigonocypris globulosa in
inland Western Australia (Halse et al. 2000; Pinder
et al. 2002, 2004; Timms et al. 2006). The eastern
inland is still characterised by the unique presence of
Parartemia minuta, a new Daphnia sp. in hyposaline
waters, many halotolerant insects, and many
monotypic genera whereas there has been multiple
Speciation in other areas, particularly in Western
Australia (Geddes et al. 1981; Pinder et al. 2002,
2004a), and a greater relative importance of a variety
of insects (mainly odonatans, corixids, notonectids,
and coleopterans).

5. Not all lakes have a full complement of
the regional fauna probably due to a mixture of
local environmental factors, stochastic events and
inadequate sampling. In the case of the Rockwell-
Wombah lakes, species composition is typical for
lakes in the middle Paroo with fluctuating salinity
(Timms, 1993; 1999; Timms and Boulton, 2001).
Notable absences include Calamoecia canberra
(because waters are not turbid enough when fresh —
Timms 2001b), and Parartemia minuta (because the
lakes often have fish). Notable unusual occurrences
include the snail Coxiella gilesi (absent in most other

Proc. Linn. Soc. N.S.W., 129, 2008

Paroo lakes investigated), Daphnia lumholtzi (more
common in these lakes than elsewhere in Paroo), and
Branchinella buchananensis (this is the only known
site in southwest Queensland). Some differences
between the lakes can be explained in terms of various
environmental factors. The presence of shrimps and
yabbies in Lake Wombah when it is fresh-subsaline
is explained by its episodic river connection. The
importance of large branchiopods in the unnamed lake
can be associated with the permanent absence of fish
in it. Finally the longer species list of coleopterans
in Mid Blue Lake (23 species cf 13 -17 in the other
lakes) could be a reflection of this lake’s greater
littoral heterogeneity, among other factors.

6. Lakes that fill predictably each season have
a shorter list of species than those fill episodically
(areas being of the same magnitude). The Rockwell-
Wombah lakes, together with other Paroo lakes and
Lake Gregory in northwest Western Australia with
episodically fluctuating salinities, have relatively
long species lists (ca 70-100 species) (Table 2 and 3;
Timms 1998; Halse et al. 1998; Timms & Boulton,
2001; Timms & McDougall 2004). By contrast,
seasonally-filled salinas in southern Victoria (Geddes
1976) and southeast South Australia (Geddes & Brock
1978; De Deckker & Geddes 1980) have less than
half this number (ca 25 -40 species). The explanation
lies mainly in the significant periods of hypsosaline
conditions (these lakes have lower median salinities
than most southern seasonally filled lakes), and also
in the unpredictable and fluctuating conditions in the
episodic lakes suiting various species at different
times, compared with predictable and generally
muted environmental conditions of the seasonal
lakes allowing just one suite of tolerant species
to persist throughout much of the season. Lack of
salinity variation to maintain diversity may well be
a major factor contributing to low species richness in
secondarily salinised lakes (Hudson et al. 2003).

7. Habitat heterogeneity is another major factor
influencing biodiversity in saline lakes (Timms 1998,
2001c; Williams1998b), and needs to be considered in
conjunction with salinity variation. Large salinas such
as Lake Eyre are far more homogeneous than small
salt lakes and have fewer species (Timms 1998). A
Paroo example is a comparison of Lake Wyara (3400
ha) and nearby Lake Bulla (420 ha), both studied for
10 years, both exhibiting similar salinity fluctuations
over this time and both studied by similar methods
— with 34 species recorded in Lake Wyara, but 73
in Bulla (Timms 1998). This role for environmental
heterogeneity as a species richness driver is confirmed

13

ECOLOGY OF EPISODIC SALINE LAKES

by the presence of 84 species in the Werewilka Inlet
of Lake Wyara (Timms 2001c).

8. Distinctions between descrete zooplankton,
littoral and benthic communities are blurred in most
saline lakes, because most are shallow and many
have rich macrophyte growth throughout the lake.
The situation is not helped by the giant ostracods
(Australocypris, Mytilocypris, Trigonocypris species
) and the large branchiopods (Parartemia spp..,
Eocyzicus parooensis., Triops ‘australiensis’ and
halobiont species of Branchinella) which live and
fed variously in all three habitats (Marchant and
Williams 1977 on Parartemia zietziana; Timms
1981 on giant ostracods in L.Gnotuk; others from
unpublished data, author). Many studies (e.g. Geddes
1976; Geddes & De Deckker 1980) therefore make
little distinction between habitats, while others try,
but with overlapping species lists (e.g. present work).
Most undersample the benthos either by failing to dig
the littoral net sufficiently into the sediments, or no
or few specialised studies with quanitative devices
such as Birge-Ekman grabs, or both. Thus while
many benthic species seem to be recorded in so called
littoral samples, indications of abundance are far too
low (cf chironomids in Tables 4 and 5).

9. Hyposaline and mesosaline lakes attract
waterbirds because of their rich resources (Kingsford
et al. 1994; Kingsford & Porter 1994; Timms 1997).
Even hypersaline lakes can support vast numbers
of a limited diversity of birds, mainly large waders
(Chapman & Lane 1997) and can be important
breeding sites for some species (e.g. Banded Stilts—
Burbridge & Fuller 1982). The Rockwell-Wombah
lakes supported many thousands of birds belonging
to 48 species. Numbers fluctuated widely (Fig. 4)
due in part to varying food resources (Kingsford &
Potter 1994; McDougall & Timms 2002), but these
were not studied in these lakes. Mobile waterbirds are
also influenced by events elsewhere in the inland. For
example, the low numbers in the Rockwell-Wombah
Lakes and also in nearby Lake Yumberarra during
2000 (Timms & McDougall 2004) probably marks
their movement to the Lake Eyre Basin in response to
even better conditions there (Roshier et al. 2002).

10. Saline lakes are usually productive,
especially when hyposaline or mesosaline, but not
when euhypersaline (Williams 1972; Hammer 198 1a,
1981b; Timms 1983). In the Rockwell-Wombah lakes
there are indications that Lake Bulla, the most saline
lake is also the most productive (Table 3). However at
any instant, some of the present lakes are unproductive

14

due to either the aftermath of a bluegreen algal bloom
(as at Lake Bulla) or by filamentous algae overgrowing
and killing macrophytes as seen in many summers at
the Blue lakes.

ACKNOWLEDGEMENTS

This paper was first presented orally at a LIMPACS
‘Salinity, Climate and Salinisation’ conference at Mildura,
Victoria in October, 2004. I sincerely thank John and
Heather Buster, Robin and Rhonda Davis and family, James
and Cheryl Hatch and Paul and Geraldine Tuite, variously
present and former owners and managers of Rockwell
and Wombah, for permission to work on their properties,
for extraordinary generous hospitality, for assistance on
occasions when I got hopelessly bogged, and to James
and John for bulldozing tracks to make my passage easier
between lakes. I appreciate the assistance over the years
of many volunteer field assistants, including John Vosper
and Sarah Wythes who came many times. I thank Olivier
Rey-Lescure for drawing Figure 1, and Jane McRae,
Adrian Pinder and Rus Shiel for identifications. Finally
I am grateful to Jenny Davis and Lien Sim of Murdoch
University, Perth, contributed with their helpful comments
on the manuscript.

REFERENCES

Bowler, J.M. (1983). Lunettes as indices of hydrologic
change: a review of Australian evidence. Proceedings
of the Royal Society of Victoria 93: 147-168.

Burbridge, A.A. and Fuller, P.J. (1982). Banded Stilts
breeding at lake Barlee, Western Australia. Emu 97:
212-216.

Chapman, A. and Lane, J.A.K. (1997). Waterfowl usage
of wetlands in the south-east arid interior of Western
Australia 1992-93, Emu 97: 51-59.

Davis, J.A., McGuire, M, Halse, S.A., Hamilton, D.,
Horwitz, P., McComb, A.J., Froend, R.H., Lyons, M
and and Sim, L. (2003). What happens when you add
salt: predicting impacts of secondary salinisation on
shallow aquatic ecosystems by using an alternative-
states model. Australian Journal of Botany 51: 715-
724.

De Deckker, P. and Geddes, M.C. (1980). Seasonal fauna
of ephemeral saline lakes near the Coorong Lagoon,
South Australia. Australian Journal of Marine and
Freshwatar Research 31: 677-699.

De Deckker, P. and Williams, W.D. (1982). Chemical
and biological features of Tasmanian salt lakes.
Australian Journal of Marine and Freshwater
Research 33: 1127-1132.

Geddes, M.C. (1976). Seasonal fauna of some ephemeral
saline waters in western Victoria with particular
reference to Parartemia zietziana Sayce (Crustacea:
Anostraca). Australian Journal of Marine and
Freshwater Research 27: 1-22.

Proc. Linn. Soc. N.S.W., 129, 2008

B.V. TIMMS

Geddes, M.C. and Brock, M. (1978). Limnology of
some lagoons in the southern Coorong. In (eds)

“The Southern Coorong and Lower Younghusband
Peninsula of South Australia’ (Eds. D.D.Gilbertson,
and M.R. Foale) pp E47-8 (The Nature Conservation
Society of South Australia Inc, Adelaide)

Geddes, M.C., De Deckker, P., Williams, W.D., Morton,
D.W. Topping, M. (1981). On the chemistry and
biota of some saline lakes in Western Australia.
Hydrobiologia 81/82: 201-222.

Halse, S.A., Shiel, R.J., Storey, A.W., Edward, D.H.D.,
Lansbury, I, Cale, D.J. and Harvey, M.S. (2000).
Aquatic invertebrates and waterbirds of wetlands
and rivers of the southern Carnarvon Basin, Western
Australia. Records of theWestern Australian Museum
Supplement. 61: 217-265.

Halse, S.A. and McRae, J. (2004). New genera and species
of ‘giant’ ostracods (Crustacea: Cyprididae) from
Australia. Hydrobiologia 524: 1-52.

Halse, S.A., Ruprecht, J.K. and Pinder, A.M. (2003).
Salinisation and prospects for biodiversity in rivers
and wetlands of south-west Western Australia.
Australian Journal of Botany 51: 673-688.

Hammer, U.T. (1981a). A comparative study of primary
production and related factors in four saline lakes in
Victoria, Australia. Internationale Revue Gesamten
Hydrobiologie 66: 701-743.

Hammer, U.T., (1981b). Primary production in saline
lakes. A review. Hydrobiologia 81: 47-57.

Hammer, U.T. (1986). “Saline lake ecosystems of the
world.’ (Junk, Dordrecht).

Hudson, P., Sheldon, F. and Costelloe, J. (2003). Aquatic
macroinvertebrate biodiversity in the Western Lake
Eyre Basin: The role of naturally fluctuating salinity.
Records of the South Australian Museum Monograph
Series No. 7: 135-144.

Kingsford, R.T. and Porter, J.L. (1994). Waterbirds on
an adjacent freshwater lake and salt lake in arid
Australia. Biological Conservation 69: 219-228.

Kingsford, R.T., Bedward, M. and Porter, J.L. (1994).
Waterbirds and Wetlands in northwestern New South
Wales. Occasional Paper No. 19, NSW National
Parks & Wildlife Service, Hurstville.

Marchant, R., and Williams, W.D. (1977). Population
dynamics and production of a brine shrimp
Parartemia zietziana Sayce (Crustacea: Anostraca) in
two salt lakes in western Victoria. Australian Journal
of Marine and Freshwater Research 28: 417-438.

Pearson, S., Gayler, S, Hartig, K. and Timms, B.

(2003). Ecosystem health in the Paroo: an arid
frontier? In “Proceedings of The Air Waters Places
Transdisciplinaey Conference on Ecosystem Health
in Australia’ (Ed, G. Albrecht) pp 252-264. (School
of Environmental and Health Sciences, University of
Newcastle)

Pinder, A.M., Halse, S.A., Shiel, R.J., Cale, D.C. and
McRae, J.M. (2002). Halophile aquatic invertebrates
in the wheatbelt region of south-western Australia.
Verhandlungen Internationale Vereinigung Limnologie
28: 1687-1693.

Proc. Linn. Soc. N.S.W., 129, 2008

Pinder, A.M., Halse, S.A., McRae, J.M. and Shiel, R.J.
(2004a). Occurrence of aquatic invertebrates of the
Wheatbelt region of Western Australia in relation to
salinity. Hydrobiologia 543: 1-24. PROBLEM

Pinder, A.M., Halse, S.A., McRae, J.M. and Shiel,

R.J. (2004b). Aquatic invertebrate assemblages

of wetlands and rivers in the Wheatbelt region of
Western Australia. Records of the Western Australian
Museum Supplement 67: 7-37.

Roshier, D.A., Robertson, A.J. and Kingsford, R.T. (2002).
Responses of waterbirds to flooding in an arid region of
Australia and implications for conservation. Biological
Conservation 106: 399-411.

Timms, B.V. (1981). Animal communities in three
Victorian lakes of differing salinity. Hydrobiologia
81: 181-193.

Timms, B.V. (1983). A study of benthic communities
in some shallow saline lakes of western Victoria,
Australia. Hydrobiologia 105: 165-177.

Timms, B.V. (1993). Saline lakes of the Paroo, inland New
South Wales, Australia. Hydrobiologia 267: 269-289.

Timms, B.V. (1997). 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: 287-313.

Timms, B.V. (1998). A study of Lake Wyara, an
episodically filled saline lake in southwest
Queensland, Australia. International Journal of Salt
Lake Research 7: 113-132.

Timms, B.V. (1999). Local Runoff, Paroo Floods and Water
Extraction Impacts on the Wetlands of Currawinya
National Park. 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). Wetlands of Currawinya National
Park: Conservation and Management. In “Research
needs for managing a Changed Landscape 1n the
Hungerford/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. (2001b). A new species of Calamoecia
(Copepoda: Calanoida) from arid Australia, with
comments on the calanoid copepods of the Paroo,
northwestern Murray-Darling Basin. Memoirs of the
Queensland Museum 46: 783-790.

Timms, B.V. (2001c). A study of the Werewilka Inlet
of the saline Lake Wyara, Australia — a harbour of
biodiversity for a sea of simplicity. Hydrobiologia
466: 245-254.

Timms, B.V. (2005). Salt lakes in Australia: present
problems and prognosis for the future. Hydrobiologia
552: 1-15.

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.

Timms, B.V. (2007). The limnology of the saline lakes of
central and eastern inland of Australia: A review with

15

ECOLOGY OF EPISODIC SALINE LAKES

special reference to their biogeographical affinities.
Hydrobiologia.576: 27-37.

Timms, B.V. and Boulton, A. (2001). Typology of arid-
zone floodplain wetlands of the Paroo River, inland
Australia and the influence of water regime, turbidity,
and salinity on their aquatic invertebrate assemblages.
Archiv fiir Hydrobiologie 153: 1-27.

Timms, B. V., Datson, B. and Coleman, M. (2006). The
Wetlands of the Lake Carey Catchment, Northeast
Goldfields of Western Australia, with special
reference to large branchiopods. Journal of the Royal
Society of Western Australia 89: 175-183

Timms, B.V. and McDougall, A. (2004). The limnology of
Lake Yumberarra, an episodic arid zone wetland, with
special reference to its use by waterbirds. Wetlands
(Australia) 22: 11-28.

Williams, W.D. (1972). The uniqueness of salt lake
ecosystems. In “Productivity Problems of
Freshwaters’ Eds S. Kajak and A. Hillbricht-
Illowska) pp. 349-361. (Polish Academy of Sciences,
Warsaw).

Williams, W.D. (1981). The limnology of saline lakes in
Western Victoria. Hydrobiologia 82: 233-259.

Williams, W.D. (1984). Chemical and biological features
of salt lakes on the Eyre Peninsula, South Australia,
and an explanation of regional differences in the
fauna of Australian salt lakes. Verhandlungen
Internationale Vereinigung Limnologie 22: 1208-
IDWS,

Williams, W.D. (1998a). ‘Management of Inland Saline
Waters. Guidelines of Lake Management Vol
6’.(International Lake Environment Committee
United Nations Environment Programme. Kusatsu,
Japan).

Williams, W.D. (1998b). Salinity as a determinant of the
structure of biological communities in salt lakes.
Hydrobiologia 381: 191-201.

Williams, W.D. (2002). Environmental threats to salt lakes
and the likely status of inland saline ecosystems in
2025. Environmental Conservation 29: 154-167.

Williams, W.D., Boulton, A.J. and Taafe, R.G. (1990).
Salinity as a determinant of salt lake fauna: a
question of scale. Hydrobiologia 197: 257-266.

16 Proc. Linn. Soc. N.S.W., 129, 2008

New Extant Species of Ironic Flies (Diptera: Ironomyiidae) with
Notes on Ironomyiid Morphology and Relationships

Davip K. McALPINE

Australian Museum, 6 College Street, Sydney 2010

McAlpine, D.K. (2008). New extant species of ironic flies (Diptera: Ironomyiidae) with notes on
ironomyiid morphology and relationships. Proceedings of the Linnean Society of New South Wales 129,

17-38.

The Ironomyiidae or ironic flies (a family of lower Cyclorrhapha) are previously known from one
Holocene Australian species and allegedly several Cretaceous or even Late Jurassic fossil species (Northern
Hemisphere countries). Aspects of morphology are discussed, particularly that of the antenna and prelabrum
(“clypeus” in error), and several alternatives as to possible phylogenetic relationships are mentioned. The
Cretaceous genus Lebambromyia Grimaldi and Cumming is removed from the Ironomyiidae to incertae
sedis (though possibly cyclorrhaphous), but the Jurassic-Cretaceous subfamily Sinolestinae is perhaps
related to Ironomyiidae. A key to species of Jronomyia White is given. Ironomyia francisi sp. nov. and J.
whitei sp. nov. are described from temperate eastern Australia.

Manuscript received 17 April 2007, accepted for publication 19 September 2007.

KEY WORDS:, antennal sacculi, Australian endemic family, comparative morphology, living fossil,
lower Cyclorrhapha, pedicellar button, phylogenetic relationships.

INTRODUCTION

The family Ironomyiidae was established
by J. McAlpine and Martin (1966) for the extant
Australian genus Jronomyia, which had previously
been placed in the Empididae-Hybotinae (now
Hybotidae), and later in the Platypezidae. Some
Jurassic-Cretaceous fossil genera of the northern
hemisphere have since been placed in the family
(as discussed below), but, on the basis of Holocene
(Recent) fauna, the Ironomyiidae remain one of the
few families of Diptera endemic to Australia.

I class the Ironomyiidae as a living fossil
taxon, because of their present limited diversity and
distribution compared with those suggested by the
fossil record, and because they show a degree of
morphological stasis since their latest Cretaceous
record (more than 70 m years ago), compared with
the great majority of cyclorrhaphous families.
Ironomyiids resemble monotremes (egg-laying
mammals) in these respects, and also in that both
groups are now restricted to the Australasian Region
where each is represented by three living species.
Whether the parallel between the two groups can be
taken an additional step, and the Ironomyiidae can
also be classed as the sister group to a taxon of major
Holocene diversity, is a question for future research,
as indicated below.

I include in the Appendix all genera
mentioned in the text with their authors’ names. I
arrange these in a provisional classification which
is based on a number of recent publications listed
in References (including Stuckenberg 2001). This
does not necessarily mean that I am convinced of the
accuracy of every step in this classification. Wiegmann
et al. (2003) presented an outline phylogeny of the
lower (non-eremoneuran) Heterodactyla (which
approximates to Muscomorpha of Woodley 1989, not
of J. McAlpine 1989). These steps are omitted from my
Appendix, as I have not referred to any included taxa
in the present text. J. McAlpine (1989) hypothesised
a monophyletic superfamily Platypezoidea including
Ironomyiidae and the phoroid families, and Collins
& Wiegmann (2002) found only limited support for
such a clade. Further morphological considerations
(some discussed below) suggest the possibility of
alternative associations, and fossil studies (e.g.
Grimaldi & Cumming 1999; Mostovski 1995) suggest
very early origins for some lineages. Monophyly of
the Aschiza (= Cyclorrhapha minus Schizophora),
though supported by J. McAlpine (1989), is refuted
by virtually all later studies. The taxon is therefore
omitted from my classification, and I leave open
several hypothesised associations among the lower
cyclorrhaphans, avoiding use of formal names of new
status. Classification above family level within the

NEW SPECIES OF IRONIC FLIES

Schizophora is omitted as irrelevant to this present
study, except that the acalyptrates are listed first. The
spelling Homoeodactyla as used by Hennig (1973)
and Sabrosky (1999) is considered most appropriate.
I use the spellings Asiliformae and Empidiformae in
accord with basic Latin grammar for taxa of above
superfamily status but corresponding in content to
the old superfamilies Asiloidea (some usages) and
Empidoidea. These were originally termed Asiliformia
and Empidiformia by Hennig (1948).

I use the term ironic flies (preferred German
version Ironiefliegen) as a family-level common
name for ironomyiids. This is simply a translation
of the name of the type genus. Such common names
have proved useful in communicating with non-
specialists.

In my morphological study I have generally
used a stereo light microscope (SLM), but for some
work a compound light microscope (CLM) or
scanning electron microscope (SEM) was used.

Inlistingmaterial, the followingabbreviations
refer to institutions housing specimens:

AM Australian Museum, Sydney

ANIC Australian National Insect Collection,
CSIRO, Canberra

BM The Natural History Museum, London

CAS California Academy of Sciences, San
Francisco

CNC Canadian National Collection, Agriculture

Canada, Ottawa
UQ University of Queensland Insect
Collection, Brisbane
The following collectors’ names are abbreviated to the
initials: D.R. Britton, C.J. Burwell, J.M. Cumming,
A. Daniels, G. Daniels, B.J. Day, E.D. Edwards, D.K.
McAlpine, I.D. Naumann, E.S. Nielsen, N. Power,
J.H. Skevington, E. Tasker, A. White, D. White, S.
Winterton.

NOTES ON IRONOMYIID MORPHOLOGY

The following observations supplement
the detailed study by J. McAlpine (1967). Other
brachyceran taxa are mentioned for comparative
purposes.

The Antenna

The antenna of Jronomyia (Figs 1, 2, 4, 6, 7)
has an unusual structure for the lower Cyclorrhapha.
The broadly rounded segment 3 (postpedicel) is
compressed in an oblique plane so as to have broad
dorsomedial and ventroexternal surfaces (state a), and
has a transverse basal slot at right-angles to this plane,

18

dividing the basal part into two strongly gibbous
projections. Each of these projections contains a
complex sensory sacculus opening by a separate pore
on its ventroexternal surface (state b). Segment 2 (or
pedicel) has a stout, distally swollen conus (sensu
Disney 1988b), fitted to the basal slot of segment 3,
and also has an angular distal exposure or projection
(“lobe”) on each of the two surfaces (dorsomedial
and ventroexternal). The conus is bridged on each
side to one of these distal projections, so that, when
exposed by removal of segment 3 (as in Fig. 6), it is
not free distally (in contrast to that of Lonchopteridae,
Sciadoceridae, and probably most taxa of Phoridae)
(state c). The dorsoexternal and ventromedial
surfaces of segment 2 each have a separate distally
facing concavity receiving the corresponding basal
projection of segment 3 (state d). The almost terminal
three-segmented arista has a covering of numerous
microtrichia extending as minute pubescence to the
apex.

This antennal structure includes four
apomorphic character states (a, b, c, d above) or one
very complex apomorphic state, not known in other
extant cyclorrhaphous taxa of pre-syrphoid grade.

The distal surface of the conus bears, on the
outer side of the distal foramen, a pedicellar button
(shown in Fig. 7). This is a new term for the smooth
subcircular cuticular structure, having slightly sunken
margins and surrounded by a smooth, convex cuticular
ring, located on either the conus or the distal articular
surface of segment 2. It is probably connected with a
chordotonal organ, which is contained within segment
2. I have found the button to be present in all taxa of
Cyclorrhapha, Empidiformae, and Homoeodactyla
which I have examined adequately with SEM ina very
preliminary survey (e.g. the platypezid Lindneromyia,
Figs 8, 9), but, like the conus, it is only visible after
removal of segment 3 from the more basal segments.
As studies of the button and conus in the Cyclorrhapha
are continuing, further details are reserved for a future
publication.

I here use the term sacculus, following
Lowne (1895: 586-589, pl. 41), for a deep, sac-like
invagination of the cuticle of antennal segment 3,
containing a number of sensilla and opening to the
exterior by a relatively small pore. This structure is
distinct from the often numerous simple pits in the
cuticle, each of which is probably associated with a
single sensillum, or various saucer-like pits that may
contain several sensilla. In Drosophila melanogaster
Meigen the sensilla in the sacculus are of several
structural kinds. Some are olfactory and some
have a “thermo-/hygrosensitive” function (Stocker
2001). Because one or more sacculi are present in

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Figures 1-3. 1, Jronomyia nigromaculata segment 3 and base of arista of left antenna, outer view. 2, the
same, segments 1 and 2 dorsal view, setulae omitted. 3, Hormopeza sp., Yukon Territory, segment 3 and
stylus of right antenna. s, sacculi.

Figures 4, 5. 4, antennae of Ironomyia nigromaculata, male. 5, antennae of Lindneromyia sp., male.

Proc. Linn. Soc. N.S.W., 129, 2008 19

NEW SPECIES OF IRONIC FLIES

many cyclorrhaphous taxa and have so rarely been antennal segment 3 of Jronomyia with such rhagionid
mentioned in the taxonomic literature, I brieflyreview or athericid genera as Symphoromyia, Suragina,
the examples that have come to my attention. and Atherix. My examination of the superficially

J. McAlpine (1967) compared the ‘reniform’ cyclorrhaphan-like antenna of the athericid

Figures 6-9. 6, Ironomyia nigromaculata, left antennal segment 2 after removal of segment 3, anterior
view. b, bridges between distal projections and conus. bu, pedicellar button. c, conus. 7, same, detail of
part of conus showing pedicellar button. 8, Lindneromyia sp., distal articular surface of right antennal
segment 2, after removal of segment 3, pedicellar button indicated. bu, pedicellar button. 9, detail of part
of same showing pedicellar button.

20 Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Suraginella shows (with CLM) that it lacks a sacculus
in the ventrally gibbous segment 3, and that the long
dorsal tubular arista, though unsegmented, has a
minute attenuated solid apical section.

I have studied the antenna of the one
available example of Hormopeza sp. (Empididae;
Q, Canada: Yukon Territory, CAS, see Fig. 3).
Segment 3 contains two large, structurally different
sacculi with remarkably wide external pores. One
of these opens on the dorsal surface and one on the
outer lateral surface of the segment. This is the only
example of possession of sacculi known to me in
the Empidiformae (or Empidoidea), though I have
studied the antennae of only a meagre cross-section
of empidiform taxa. Sinclair (1995) interpreted the
style of Hormopeza as two-segmented with an apical
bristle. This interpretation of the apical part as a bristle
or enlarged macrotrichium appears to be justified by
its solid structure and lack of pubescence, in contrast
to the two preceding sections.

The antenna of Sciadocera lacks the sacculi
in segment 3. Segment 2 is short in its externally
visible surfaces, but has an elongate central knob, the
conus, arising from the centre of its distal surface.
The conus fits into a deep, rounded central basal
cavity in segment 3 (Disney 2001: fig. 7) and bears
the foramen of articulation on its terminal surface.
In S. rufomaculata White segment 3 is so securely
anchored to segment 2 by means of the conus that
it cannot readily be disarticulated, even after the
intersegmental connection is snapped by rotation.

In typical taxa of the Phoridae sacculi are
absent (as far as I can determine without a major
study), and relations between segments 2 and 3 are
very like those of Sciadocera, but segment 2 tends to
become more reduced so as to be often represented
by little more than the concealed conus (Disney 1994:
fig. 1.2). The antenna of Lonchoptera furcata (Fallén)
(Lonchopteridae) also has a large conus and no
sacculi, whereas that of Melanderomyia kahli Kessel,
Lindneromyia spp., and probably other platypezids
(Platypezidae) has no sacculus in segment 3 and an
almost planate distal articular surface of segment 2
with only slight indication of a conus (Fig. 8). The
antenna of the Lonchopteridae is thus much more
like that of the sciadocerid-phorid alliance than
that of any other lower cyclorrhaphous (aschizan)
taxon. J. McAlpine (1989) is in error in stating that
Lonchopteridae share with Platypezidae the absence
of the conus (“apex of pedicel never deeply inserted
into base of first flagellomere.”’)

A single large sacculus is probably usually
present on the outer surface of segment 3 in the
Syrphidae, though minor sensory pits may also be

Proc. Linn. Soc. N.S.W., 129, 2008

evident [examples studied with CLM: Microdon
variegatus (Walker), Eristalis tenax (Linné),
Melangyna sp.|; but Deineches sp. apparently has
numerous sacculi. Eudorylas sp. (Pipunculidae) also
has one sacculus.

In the Schizophora one or more sacculi
are usually present. The acalyptrate taxa generally
have one sacculus [no exceptions yet confirmed;
examples studied by me: Liriomyza chenopodii
(Watt) (Agromyzidae), Asteia sp. (Astetidae),
Aulacigaster sp. (Aulacigastridae), Zalea major
(McAlpine) (Canacidae s.l.), Clisa australis
(McAlpine) (Cypselosomatidae), Scaptomyza
australis Malloch (Drosophilidae), numerous taxa of
Ephydridae, Gobrya cyanea (Enderlein) (Gobryidae),
Tapeigaster spp., Borboroides spp., and Heleomicra
sp. (Heteromyzidae s.l.), Huttonina furcata Tonnoir
& Malloch (Huttoninidae), Poecilohetaerus aquilus
Schneider, 7rigonometopsis sp. (Lauxantidae), Badisis
ambulans McAl\pine, Compsobatafemoralis (Meigen),
Cothornobata aczeli McAlpine, Metopochetus
spp., and Mimegralla spp. (Micropezidae), Nemo
centriseta McAlpine (Neminidae), Neurochaeta
capilo McAlpine, Neurocytta prisca (McAlpine),
Neurotexis freidbergi McAlpine, and Nothoasteia
clausa McAlpine (Neurochaetidae), Nothybus
decorus de Metjere (Nothybidae), Teloneria sp.
(Nertidae), Maorina sp. (Pallopteridae), Cyamops
sp. (Periscelididae), Euprosopia armipes McAlpine,
Lenophila coerulea (Macquart), and Rhytidortalis
averni McAlpine (Platystomatidae), Lasionemopoda
hirsuta (de Meyere) (Sepsidae), Strongylophthalmyia
sp. (Tanypezidae), Somatia aestiva (Fabricius)
(Somatiidae), Teratomyza undulata McAlpine
(Teratomyzidae)]. °

In the calyptrate Schizophora the number of
sacculi in segment 3 is variable, but I have had time
to examine very few taxa, especially as the usually
darkly pigmented cuticle makes study more difficult.
Fannia canicularis (Linné) (Muscidae or Fanniidae)
has one major sacculus near the middle of the outer
surface, but there are also numerous smaller, shallow
pit-like structures, some of which contain several
sensilla. Scathophaga sp. (Scathophagidae) has a
sacculus on the outer surface and another on the
medial surface nearer the base. Musca vetustissima
Walker has one major subbasal sacculus only.
Calliphora augur (Fabricius) (Calliphoridae) has
apparently c. nine major sacculi. The statement
by Lowne (1895: 586), that C. vicina Robineau-
Desvoidy (as C. erythrocephala) has “about eighty
large sacculi”, apparently includes the smaller pits on
the medial surface. Chetogaster sp. (Tachinidae) has
one large sacculus on the outer surface and several

21

NEW SPECIES OF IRONIC FLIES

smaller pits on the inner (medial) surface. In the
Axiniidae (Colless 1994) there is commonly one
sacculus (“sensory pore”), but it is multiple or absent
in various taxa.

The presence of more or less separate
dorsoexternal and ventromedial concavities (or,
for comparative purposes, simply upper and lower
concavities) on the distal articular surface of
segment 2 of /ronomyia is interesting. Such separate
concavities are absent in most cyclorrhaphous taxa
of pre-syrphoid grade, including the platypezids
Lindneromyia (Fig. 5) and Melanderomyia (the latter
with a strongly bilaterally compressed segment 3),
but are distinguishable in many syrphids, pipunculids,
and schizophorans.

The arista of Jronomyia is three-segmented,
as is probably also the case in the Cretaceous
ironomyiid genus Cretonomyia. This is the most
frequent condition in the Cyclorrhapha and _ is
probably the groundplan condition, in contrast to
that of most Empidiformae. However, fewer than
three aristal segments are present in the Opetiidae,
apparently in some lonchopterid-like and platypezid-
like fossil taxa, in numerous taxa of Syrphidae,

and in a sprinkling of taxa in numerous families of
Schizophora (D. McAlpine 2002). Most or all of
these examples represent derived character states and
are of multiple origin.

The Prelabrum

The prelabrum of the Cyclorrhapha is often,
with inadequate justification, homologised with the
clypeus (J. McAlpine 1981) or sometimes with the
(usually fused) tormae (e.g. J. McAlpine 1967: figs
1, 2). See D. McAlpine (2007) for discussion of this
problem. In Jronomyia (Fig. 10) it is remarkably
prominent and resembles that of many schizophorans
in appearance. However, it differs from the latter in
being very weakly sclerotised on its median section. I
note that in Sciadocera and at least some platypezids
(Fig. 11), the prelabrum is divided in two or almost
so, a condition which suggests (but does not prove)
its origin from paired sclerites, such as the tormae. In
these taxa and in Lonchoptera the prelabrum is more
or less flattened against the ventral surface of the head,
in contrast to that of Ironomyia. In the Syrphidae and
the Schizophora the prelabrum is generally undivided
and broadly sclerotised across its median part.

Figures 10, 11. 10, Ironomyia nigromaculata, subcranial region of male. 11, Lindneromyia sp., subcra-
nial region of male. Medially desclerotised prelabrum indicated for both taxa.

DD

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Figures 12-15. Prosternal region of cyclorrhaphous flies. 12, Ironomyia nigromaculata, subcoxal scler-
ite indicated. 13, Lindneromyia sp. 14, Sciadocera rufomaculata, precoxal bridge indicated. 15, Eristalis
tenax (Linné), subcoxal sclerite indicated.

The prelabrum of all species of Ironomyia is sexually
dimorphic, being larger in females. This condition
may be correlated with dimorphism of the head
capsule in taxa having holoptic tendency in the males.
However, in some cyclorrhaphous taxa without male
tendency to holopticism, the prelabrum of the female
is Significantly larger than that of the male — e.g.
Borboroides spp. and Heleomicra spp. (Heleomyzidae
s.1. or Heteromyzidae), Rivelliaspp.(Platystomatidae),
Gluma_ spp. (Coelopidae). Strongylophthalmyia
spp. (Tanypezidae), Acartophthalmus nigrinus
(Zetterstedt) (Acartophthalmidae), A/lometopon spp.
and Tetrameringia ustulata McAlpine (Clusiidae),
Traginops sp. (Odiniidae), Dasyrhicnoessa spp.
(Tethinidae or Canacidae); Stenomicra sp. (sp. B in
AM, Periscelididae). Cyamops spp. (Periscelididae)
also have the prelabrum much smaller in the male,
but in this genus the facial region of the head capsule,
not the postfrons, is narrowed by encroachment of the
eyes in the male.

The Prosternum
In Jronomyia the prosternum (Fig. 12) is a

Proc. Linn. Soc. N.S.W., 129, 2008

very broad sclerite covering most of the ventral surface
of the thorax in the space between the fore coxae. Its
lateral margin forms a raised flange on each side, and
next to the anterior part of the lateral.margin there is
a separate, often rather weakly defined, subtriangular
sclerite — the subcoxal sclerite. Anterolaterally the
prosternum is separated from the propleuron on each
side by a membranous zone, i.e. there is no precoxal
bridge. In the platypezids Lindneromyia (Fig. 13)
and Microsania the prosternum is slightly narrower,
with neither raised lateral margin nor precoxal
bridge, and the subcoxal sclerite is at most vestigial.
However, in the platypezid Agathomyia, I find a large
triangular subcoxal sclerite. In Sciadocera (Fig. 14)
the prosternum is broadly triangular, with slightly
raised lateral margin and distinct precoxal bridge; no
subcoxal sclerite is distinguishable in the position it
occupies in /ronomyia, but a minute sclerotised spot
near the anterior angle of the coxal base perhaps
represents the subcoxal sclerite. Typical phorids
generally have the prosternum very like that of
Sciadocera but more narrowed posteriorly and without
raised lateral margin. Lonchoptera furcata also has a

23

NEW SPECIES OF IRONIC FLIES

similar prosternum, without visible subcoxal sclerite
or raised lateral margin.

The Syrphidae show a range of shapes in
the prosternum, but the following features are nearly
always present (Fig. 15): the prosternum is broad with
margin strongly raised and produced posterolaterally
into a lobe on each side; the prosternum is isolated
from the propleuron on each side, there being no
precoxal bridge; the subcoxal:sclerite is large and
approximated to the lateral margin of the prosternum.
This suite of characters is remarkably similar to that
of Jronomyia, differing mainly in that the raised
lateral margin forms a prominent posterolateral lobe
on each side. The prosternum of the Pipunculidae
and Conopidae, so far as I have observed it, looks
like a reduced version of that of the Syrphidae, being
generally more narrowed anteriorly, with the subcoxal
sclerite little developed. In the Schizophora prosternal
structure is very diverse-(see Speight 1969), but,
as the diversity is scarcely relevant to this study of
Ironomytidae, it is not treated here.

The Tarsus

Sawlines are present on the mid and hind
tarsi of Ironomyia spp., except on the terminal
segment of each. I have described these structures
for the schizophoran families Syringogastridae and
Diopsidae and mentioned their presence in a few other
cyclorrhaphous families including Sciadoceridae,
Phoridae, Platypezidae, and several families of
Schizophora (D. McAlpine 1997: 172). The modified
setulae comprising ironomytid sawlines (Fig. 16)
appear to lack the double structure seen in diopsoid
flies (op. cit. figs 40-43).

In all three Jronomyia spp. segment 4 of the
hind tarsus is very asymmetrical, on account of an
elongate, subconical distally directed process on the

posterior side. This process bears a sawline and several :
other setulae, of which a subapical one is enlarged. It
does not appear to be sexually dimorphic.

In I. nigromaculata the empodium consists
of a pubescent, broadly tubercle-like basal plate,
apically bearing a large, simple, upwardly curved
setiform process with smooth surface, in contrast to
the longitudinally sculptured setulae on segment 5
(Figs 16, 17). °F

In Sciadocera rufomaculata the empodium
has the same basic structure as in Jronomyia, but
is much smaller and the setiform*process is not
curved upwards. The hind tarsus shows strong
sexual dimorphism. In the female, segment 4 is
approximately symmetrical. In the male segment 4
is very asymmetrical, but of different form from that
of Ironomyia, and it is the anterior, not the posterior,
side that is strongly extended distally. Schmitz (1929)
described the empodium of Sciadocera as ‘pad-
shaped and not bristle-like.’ This is certainly not true
for Sciadocera s.str., but he may have referred to
the condition in Archiphora, which-was then treated
as a subgenus of Sciadocera, and which I have not
examined. However, Schmitz admitted that he had a
pair of Sciadocera (Sciadocera) nigromaculata for
study.

The Wing

The wing venation of /ronomyia is probably
quite diagnostic within the whole field of extant
brachyceran flies (Fig. 23). The subcosta is free
basally, but becomes fused with vein 1 (R,) for
more than one third of the wing-length, and diverges
from it distally to terminate separately in the costa;
vein 3 (R,,.) is unbranched; the discal cell is well
developed and distally separately emits three veins,
each of which extends to the wing margin; the anal
cell (CuP) is enclosed, moderately short and emits

Figures 16, 17. Ironomyia nigromaculata, parts of left hind leg. 16, distal part of tarsus ventral view,
sawlines indicated. 17, terminal view of tarsus.

24

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

distally a long vein 6 (CuA, + A,) which is distinctly
sclerotised approximately to the margin. Also, the
costa has two or three almost longitudinally aligned
anteroventral costagial bristles. Jronomyia agrees
with Sciadoceridae and Phoridae in that the costa
terminates at the end of vein 3, whereas in typical
platypezids it extends to vein 4. Disney (1988a)
pointed out a close resemblance in costal chaetotaxy
between Ironomyiidae and Sciadoceridae. Though
he initially regarded the relevant character state as
a synapomorphy, he later (Disney 2001) described
it as ‘probably a plesiomorphic feature’, in order to
support an alternative cladistic hypothesis.

In all specimens of Jronomyia that I have
examined in detail, the anal region is broadened so
that that part of its margin just beyond the alula is
almost at right angles to the longitudinal axis of the
wing (Fig. 23). However, in previously published
illustrations of the wing of 1. nigromaculata the anal
region is erroneously shown as less prominent and
more evenly rounded, presumably because it was
furled or partly concealed in the available specimens.
The part of the anal margin nearest the alula naturally
possesses a series of elongate setulae, which vary
considerably in number. These setulae are so fragile
that they are often lost in preserved specimens.

The alula of Jronomyia, though not large,
forms a better developed lobe than in Sciadoceridae
and Phoridae. A distinct, pigmented ambient vein is
present on its margin, and bears numerous setulae
(socket-based macrotrichia), but has very limited
extension on to the anal lobe.

The hair-fringe on the posterior margin of
the wing beyond the anal lobe consists of microtrichia
only, which, like those of most lower cyclorrhaphans
other than Lonchopteridae, lack basal sockets.

The Preabdomen

In Jronomyia abdominal tergites 1 and 2
are separated by a narrow intersegmental membrane
across the dorsal part of the abdomen but this
membrane is discontinued for about one fifth of the
total width of the tergite on each side, so that the
two tergites become sclerotically continuous on this
marginal region. A groove in the apparently uniformly
sclerotised cuticle is all that defines the two tergites
in this region, as there is no well defined marginal
incision. By contrast, tergites 2 to 6 are all separated
by complete intersegmental membranes.

My studies, which are far from a complete
coverage of taxa, indicate that partial to complete
fusion of tergites 1 and 2 is a general rule in the
Cyclorrhapha. The only included groups in which
these tergites are completely separated by a strip of

Proc. Linn. Soc. N.S.W., 129, 2008

intersegmental membrane are, to my knowledge, the
Opetiidae (Chandler 1998), Sciadoceridae (author’s
observation on Sciadocera), and Phoridae (author’s
limited observations and examples figured by Disney
1994). The Empidiformae usually have tergites 1 and 2
separate, and thus, if Empidiformae and Cyclorrhapha
are sister groups, the separate condition is likely to
be the groundplan condition for the Eremoneura. The
resemblance between Ironomytidae and the rest of the
Cyclorrhapha (apparently including the Platypezidae)
is interesting, but cannot at this stage be affirmed as a
synapomorphy.

Chandler (2001) points out that the
Platypezidae show so many plesiomorphic conditions
that it has not yet been conclusively demonstrated how
they differ from the groundplan of the Cyclorrhapha.
At least partial fusion of tergites 1 and 2 occurs in the
genera Agathomyia, Lindneromyia, Melanderomyia,
and Microsania (my observations); so, if the condition
proves to be uniformly present in the Platypezidae, this
would be an apomorphic character state of the family
relative to the groundplan of the Cyclorrhapha.

The Male Postabdomen

The male postabdomen of Jronomyia is
essentially symmetrical; segment 6 has the tergite and
sternite occupying their primitive, respectively dorsal
and ventral positions; segment 7 is not represented
by any sclerite, and sternite 8 is large, approximately
symmetrical, dorsally located, and connate on its
posterior margin with the epandrium. J. McAlpine
(1967) terms sternite 8 “sternite 7+8”, but there is no
evidence for inclusion of any part of segment 7 in this
sclerite, either in the groundplan of the Cyclorrhapha
or in Jronomyia. In the Schizophora many taxa possess
an identifiable, usually asymmetrical sternite 7, and a
few taxa possess a small sclerite associated with right
spiracle 7 which could be a vestige of tergite 7. The
absence in Jronomyia of any sclerite representing
segment 7 agrees with Sciadocera but differs from
the Platypezidae. According to Chandler (1998: fig.
8), Opetia has the postabdominal segmentation even
further reduced. J. McAlpine (1967) identifies the
two median sclerites between the aedeagus and the
cerci in Jronomyia as ‘sternum 10?’ and ‘sternum 11?’
While I am also doubtful of the homologies of these
sclerites, I provisionally use those designations.

FOSSILS
The fossil record (entirely Mesozoic)

of apparent or possible ironomyiid flies has been
reviewed by Mostovsky (1995) who provided a key

XS

NEW SPECIES OF IRONIC FLIES

(in Russian); and Chandler (2001) made some general
comments on the fossil taxa. Grimaldi and Cumming
(1999) described an additional fossil taxon.

The only one of these fossil taxa which
I consider to be unambiguously ironomyiid is
Cretonomyia pristina J. McAlpine 1973, in Upper
Cretaceous Canadian amber. The wing venation
(including partial fusion of subcosta and vein 1) and
visible detail of antennal segments 2 and 3 confirm
the impression of a fairly general agreement with
Ironomyia in other characters, though numerous
morphological features of the unique fossil are not
visible. Ironomytidae s. str. (=subfamily Ironomyiinae
of Mostovsky) includes only the genera Jronomyia
and Cretonomyia.

Lebambromyia acrai Grimaldi & Cumming,
1999, was based on two specimens in lower Cretaceous
Lebanese amber. The published details suggest
cyclorrhaphous status for Lebambromyia, but in my
view any synapomorphy with Ironomyiidae is at best
doubtful. Ifthe prelabrum is absent, then perhaps some
doubt would be thrown on the cyclorrhaphous affinities
of Lebambromyia (but the condition of the prelabrum
is apparently also unrecorded for Opetia, which I
have not examined). Grimaldi and Cumming mention
the similarity between Lebambromyia, Cretonomyia,
and /ronomyia in the sclerotised pterostigma confined
to the subcostal cell. Agathomyia spp. (England, AM;
New York state, AM) have a lightly sclerotised zone
in the apex of the subcostal cell, and this appears
to be present also in some other platypezids figured
by Kessel (1987). I do not consider the presence of
a pterostigma restricted to the subcostal cell to be a
reliable diagnostic indicator for the Ironomyiidae.

Lebambromyia \acks what I consider to be
diagnostic apomorphies for the Ironomyiidae, viz.
the partial fusion of the subcosta with vein 1, and
the highly specialised articulation between antennal
segments 2 and 3. I find no acceptable evidence for
inclusion of Lebambromyia in the Ironomytidae and
formally remove it from the family. Those making
future studies of the genus should decide whether a
new family is required for it.

The remaining fossil taxa previously
referred to the Ironomyiidae are placed in the
subfamily Sinolestinae by Mostovski (1995). These
are impression fossils from the Upper Jurassic to
Lower Cretaceous of northern Asia, and are thus
among the very earliest putative cyclorrhaphans,
if correctly identified as such. Included genera are:
Eridomyia, Hermaeomyia, and Palaeopetia (syn.
Sinolesta). Twenty-one nominal species are included.
These flies have, except in the region of the subcosta,
venation typical of the Platypezidae, with the fork

26

of vein 4 (M,_,) located beyond the discal cell. The
subcosta, beyond the base, becomes fused or connate
for a considerable distance with vein 1, then becomes
free distally so as to delimit basally, in at least some
species, a short, apparently sclerotised or pigmented
pterostigma. This condition of the subcosta in, for
example, Palaeopetia gemina Mostovski (1995:
fig. 1) is a precise replica of that in the Holocene
Ironomyia (Fig. 23), but shows less resemblance to
the Upper Cretaceous Cretonomyia, which has a much
shorter extent of fusion between the subcosta and
vein 1. It has been doubted if actual fusion between
the subcosta and vein 1 can be demonstrated in the
impression fossils, but the sinolestine specimens
are fairly numerous, and the compression to which
they have been subjected should emphasise any gap
between veins in a percentage of specimens.

It is improbable that details of antennal
structure would be preserved in these impression
fossils, but the drawings by Mostovski sometimes
indicate amore rotund segment 3 than is generally seen
in the less specialised empidoids and the Platypezidae
(other than Melanderomyia). That of Eridomyia
captiosa Mostovski shows antennae reminiscent of
true ironomyiids, but I am unsure how much of the
visible outline represents actual structure.

Without seeing any of these sinolestine
fossils, I remain impressed by their resemblance to
later ironomyiids, and suggest that the Sinolestinae
remain provisionally in or near the Ironomyiidae
until further evidence tends to confirm or negate this
position.

PHYLOGENETIC CONSIDERATIONS

The IJronomyiidae share three possible
synapomorphies with the probably monophyletic
group Sciadoceridae + Phoridae (treated as one family
by Tonnoir 1926, Disney 2001): (1) the subcosta
becomes fused with vein 1 a short distance beyond
base; (2) the costa extends to and is discontinued
near vein 3 (in contrast to less reduced forms of
Platypezidae, and probably also to Lonchopteridae
and Opetiidae where it is continued as an ambient
vein); (3) segment 7 of the male postabdomen has no
distinguishable sclerite. In the Cretaceous tronomyiid
Cretonomyia fusion between the subcosta and vein 1|
commences much further from the base and continues
for a much shorter distance than in Jronomyia; i.e. the
subcosta is somewhat less phoroid in appearance.
The case for synapomorphy between Ironomytidae
and the more typical phoroids is thus weakened, and
homoplasy between, for example, Jronomyia and

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Sciadocera becomes less unlikely. Also characters (2)
and (3) can show instability in many cyclorrhaphous
lineages, and are subject to frequent convergence.

Griffiths (1972) considered the Ironomytidae
+ Sciadoceridae + Phoridae to constitute the clade
Hypocera (Phoroidea is the name now preferred), and
included two further autapomorphies for the group:
“apex of second antennal article deeply inserted into
base of third;” and “anal cell shortened.” I reject
these from the apomorphic evidence for inclusion
of the Ironomyiidae in this alliance, because (1) the
articulation between antennal segments 2 and 3 in
Ironomyia is fundamentally different from that of
the other two families as indicated above, (2) a well
developed conus inserted into segment 3 is present in
many cyclorrhaphous taxa, and (3) the anal cell (cup)
of Jronomyia and Cretonomyia is only marginally
shorter than that of numerous platypezid taxa and
probably not as short as that of Opetia.

Tronomyia also differs from more typical taxa
of the phoroid alliance in the absence of pubescence on
the marginal setulae of the alula, in the partial fusion
of abdominal tergites 1 and 2, and in the absence of
prothoracic precoxal bridges. When on tree trunks or
in glass containers, ironomyiids are inactive or walk
slowly, in contrast to typical sciadocerids, phorids,
and platypezids, which usually run actively.

The nature of the articulation between
antennal segments 2 and 3, and the presence of
well developed sensory sacculi in segment 3,
appear to separate the Ironomytidae from all other
cyclorrhaphans of pre-syrphoid grade, so far as they
have been investigated for these structures, but a few
approaches to these ironomyiid-like conditions have
been noted in certain taxa in the Empidiformae.

As an alternative hypothesis, two shared
apparent apomorphies (?synapomorphies) suggest that
the Ironomyiidae are closely related (? the sister group)
to the Eumuscomorpha (= Syrphidae + Pipunculidae
+ Schizophora). These are: (1) antennal segment 3
containing one or more sensory sacculi; (2) antennal
segment 2 with more or less separate upper and lower
concavities on distal articular surface, which receive
the upper and lower basal gibbosities of segment 3. Of
these conjectured synapomorphies, I have particular
reservation concerning (2). In such taxa as the
platypezid Lindneromyia (Fig. 5) antennal segments
2 and 3 are only slightly bilaterally compressed, and
this probably plesiomorphic approximation to radial
symmetry round the central longitudinal axis results
in the distal concavity of segment 2 being annular.
In many syrphids, platypezids, and schizophorans
these segments, particularly segment 3, have become

Proc. Linn. Soc. N.S.W., 129, 2008

dilated and compressed and such annular concavity is
consequently squeezed into upper and lower elements.
But bilateral compression of segment 3 and the
adjoining part of segment 2 has evolved several times
in diverse non-cyclorrhaphous brachycerans (e.g. in
Athericidae and Dolichopodidae). Such examples
must throw a degree of doubt on any hypothesis of
a single permissible origin of such compression in
Cyclorrhapha. Nevertheless, I mention condition
(2) because it is probably a groundplan condition
of the three abovementioned main groups of
Eumuscomorpha, as well as of Ironomyia; also because
the platypezid Melanderomyia, with its strongly
compressed segment 3, still lacks the distal concavities
of segment 2. As explained above, the prominent,
sexually dimorphic prelabrum of Jronomyia is, apart
from its probably primitive bipartite structure, more
like that of certain eumuscomorphous taxa than that of
any other cyclorrhaphous taxon of presyrphoid grade.
But much variation in the Eumuscomorpha creates
difficulty in determining the groundplan condition of
the prelabrum in this group.

The partial fusion of abdominal tergites 1
and 2 in Jronomyia is more in agreement with the
Eumuscomorpha than with the Phoroidea, though it
occurs also in some other lower cyclorrhaphans.

According to the study by Wada (1991), the
Eumuscomorpha differ from all other cyclorrhaphans
investigated in the nature of the sensory epithelium of
the compound eyes. Unfortunately he was not able to
examine /ronomyia for this condition.

If the fusion between the subcosta and vein
1 is a homologous condition through the Sinolestinae
and Ironomytidae s.str., then this apomorphy must
have evolved in the lineage by the end of the Jurassic
Period. The ancestral eumuscomorphan (on the
assumption of monophyly for this group), having a
more plesiomorphic subcosta, could not have been
subsequently derived from such ironomytiid lineage.

It is now desirable that Jronomyia be
incorporated into DNA phylogenetic studies, such
as that of Collins & Wiegmann (2002), to test
support for one of three conceivable alternatives:
(1) Ironomyiidae are closest to the typical phoroid
families Sciadoceridae and Phoridae; (2) Ironomyiidae
are close to the Syrphidae and Pipunculidae or to
the possibly monophyletic Eumuscomorpha; (3)
Ironomylidae are a very isolated group of lower
cyclorrhaphans. If either alternative (2) or (3) is
favoured, a separate superfamily would be necessary
for the ironomyiids.

Di

NEW SPECIES OF IRONIC FLIES

Genus Ironomyia White

Ironomyia White 1916: 216-217, fig. 39. Type species
(monotypy) I. nigromaculata White.

Description
See J. McAlpine (1967).

Distribution

Queensland — as far north as Atherton
Tableland. New South Wales — Coast districts to
Western Slopes. Victoria — few records. Tasmania
— probably widely distributed. In New South Wales
and Queensland they extend from wet coastal
districts to drier inland districts such as Mendooran
and Millmerran. With further collecting, they will
probably be found to have a wider distribution in
Victoria, and it would not be surprising if they were
discovered in South Australia and Western Australia.

Notes

This, the only extant genus of Ironomytidae,
shows very little structural diversity, and has been
well described by J. McAlpine (1967). I give some
further morphological data above. The wing venation
(Fig. 23) is unique among Holocene Brachycera, and
the structure of antennal segments 2 and 3 (Figs 1, 2)
is also distinctive.

The flies generally live in forested country,
where they are sometimes found on tree trunks (e.g.
Acacia s.1.). They are also taken by light-trapping and
occasionally by sweeping vegetation. They are most
often found in late spring or early summer, but are
generally uncommon.

Key to species of Ironomyia

1 Scutellum with dorsal setulae or mollisetae (in
addition to marginal bristles, Fig. 19), or, if
these (rarely) absent, then numerous marginal
setulae located among marginal bristles; apex of
scutellum without pale spot; abdominal tergites
2 to 6 each with black median zone; male:
surstylus compressed, plate-like, with setulae
little developed on anterior surface ...............00
eS Ae ee nigromaculata White

- Scutellum without either dorsal setulae or dorsal
mollisetae, nor with setulae among marginal
bristles, with yellowish apical spot (Fig. 21);
median zone of tergites 2 to 6 pale grey to
yellowish, flanked by black zone on each side
(sometimes brown on tergite 6); male: surstylus
not thus compressed, with very numerous long or
rather short setulae on anterior surface ............ Z

2 Wing with blackish apical spot covering end of

28

submarginal cell and adjacent parts of marginal
and first posterior cells (Fig. 24); median pale
zones on tergites 2-5 relatively narrow (Figs 27,
28); male: surstylus stoutly ovoid, with mixed
large and small anterior setulae (Fig. 26)... ...... :
LEELA whitei n.sp.
- Wing without distinct apical spot (apices of veins
2 and 3 often darkened); median pale zones on
tergites 2-5 relatively broad (Fig. 22, broader
in female); male: surstylus in profile curved,
obliquely truncate, with numerous short setulae
only on anterior surface (Fig. 25) ...francisi n.sp.

Figure 18. Ironomyia nigromaculata. Abdominal
tergites 1-6 of male, diagrammatic, spread flat to
show pattern.

Ironomyia nigromaculata White
Figs 1, 2, 4, 10, 12, 16-20

Ironomyia nigromaculata White 1916: 217-218,
fig. 39; J. McAlpine 1967: 226-227, figs 1-15
(redescription).

Material examined (localities only)

Queensland: Hugh Nelson Range, S of
Atherton (only known specimen from tropics,
ANIC); Mount Moffatt vicinity, Carnarvon National
Park (AM, UQ); Rainbow Beach, Tin Can Bay

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

district (AM); near Mount Glorious (UQ); Brisbane
district — several localities (AM, UQ); 43 km WSW
of Millmerran (AM); Tamborine Mountain (UQ);
Amiens State Forest, W of Stanthorpe (AM). New
South Wales: Dorrigo (ANIC); New England National
Park (ANIC); Wollomombi Falls, Armidale district
(AM); Warrumbungle National Park (AM, UQ);
Carrai State Forest, W of Kempsey (AM); Goonoo
State Forest, near Mendooran (AM); Kurrajong (AM);
Mount Boyce, Blue Mountains (AM); Ku-ring-gai
Chase National Park (AM); Royal National Park, S
of Sydney (AM); Otford (AM); Macquarie Pass, near
Albion Park (ANIC); Minnamurra Falls, W of Kiama
(AM); Clyde Mountain, E of Braidwood (ANIC);
Bawley Point, Ulladulla district (AM, ANIC); Depot
Beach, near Bateman’s Bay (ANIC). Victoria: 26 mi
(c. 42 km) NNE of Orbost (ANIC); Young’s Creek, 12
km N of Orbost (ANIC). Tasmania: Mount Barrow,
near Launceston (AM); Cradle Valley (ANIC); Lake
Saint Clair (ANIC); Bronte Park (ANIC); Franklin-
Gordon Wild River Park (UQ); Hobart (holotype
BM; ANIC); Arve River, Geeveston district (ANIC).

I have not examined the holotype, but J.
Chainey has kindly checked its diagnostic characters

19

and confirmed (in litt.) the identification made here
and by J. McAlpine (1967).

Description

See J. McAlpine (1967). Inote some variation
and some diagnostic characters in the relatively large
series now available.

Coloration. Antennal segment 3 generally
dark grey or grey-brown in southern populations, but
specimens from Queensland usually with segment
3 yellow, as in J. francisi and I. whitei. Mesoscutal
coloration sexually dimorphic as in J. francisi, but that
of female more variable in width and extent of dark
bands; scutellum always without pale apical zone.
Dorsal abdominal pattern variable, but black zones
generally more extensive in males than in females,
(Fig. 18; J. McAlpine 1967: figs 4, 5) and often less
extensive in southern Queensland specimens, but
male from Hugh Nelson Range (northern known limit
of range) as dark as any southern specimens; tergites
2 to 6 always with black median zone (J. McAlpine’s
fig 4 is in error regarding tergite 6, being based on a
damaged specimens which I have checked).

Thorax. Scutellum with few to numerous

20

Figures 19, 20. Ironomyia nigromaculata. 19, scutellum, dorsal view. 20, epandrium and associated

parts, left lateral view, scale = 0.1 mm.

Proc. Linn. Soc. N.S.W., 129, 2008

29

NEW SPECIES OF IRONIC FLIES

dorsal setulae, often fewer in northern populations,
specimen from Hugh Nelson Range with setulae
interspersed with marginal bristles, but apparently no
dorsal setulae.

Male _postabdomen. Epandrium relatively
stout; surstylus more compressed and plate-like than
in other species, with very oblique base and broadly
rounded apex, setulosity on anterior surface little
developed; hypandrium broadly rounded anteriorly;
cerci and sternite 11 located more terminally than
dorsally on epandrium; sternite 11 broader than in
other species.

Dimensions. Total length, 4 3.7-6.0 mm,
Q 2.9-6.2 mm; length of thorax, d 1.5-2.5 mm, 2
1.3-2.5 mm; length of wing, 3 4.2-6.4 mm, ° 3.6-6.9
mm.

all eps

Brisbane Forest Park, near The Gap, 27°25’41”S
152°50°18”E, 28.1x.-15.x.2002, JLHS. & J.M.C.,
Malaise trap (J.H.S. #13323, AM). Glued to small
card point.

Paratypes. Queensland: 2 3, 4 2, same data
as holotype (AM,ANIC): 1 4, Scrub Road, Brisbane
Forest Park, 3-10.x.1997, S.W., N.P., D.W. (UQ);
1 4, 1 9, W of Highvale, near Samford, 27°23’S,
152°47’°E, 19.ix.1986, G.D. (UQ); 2 3, 3.5 and 4
km WSW of Point Lookout, North Stradbroke I., 3-
5.iv.1992, G.D., C.J.B. (UQ); 1 4, North Stradbroke
1., 4.v1.1987, C.J.B. (UQ).

Other material. New South Wales: 1 6,24km
W of South Grafton, 29°37’S, 152°44’E, 1.xii.1990,
A.D., G.D. (AM); 4 2, Mooney Mooney Creek,
near Gosford, 20-29.xi1.1975, 1.xii.1989, BJ.D.,

Figures 21, 22. Ironomyia francisi. 21, dorsal view of thorax of female to show pattern. 22, tergites 1-6

of male, diagrammatic, showing pattern.

Distribution

Eastern Australia: Atherton Tableland
(Queensland) to southern Tasmania. Records cover
almost the whole known range of the genus.

Tronomyia francisi sp. nov.
Figs 21-23, 25

Material examined
Holotype. 3, Queensland: Scrub Creek,

30

G.D., D.K.M. (AM, CNC); 1 9, Sydney, no other
original data (ANIC); 1 3, 3 km E of Wedderburn,
Campbelltown district, 34°08’S 150°49’E, 19.x.2003,
D.R.B. (AM, in alcohol).

Description (3, 2)
With the general morphology of the genus.
Coloration. Head grey, largely densely
pruinescent; most setulae black; subgenal region
largely shining brown. Antenna yellow, with brown
arista. Sclerotised lateral part of prelaburm with

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

shining brown zone which is larger in female; median
desclerotised part of prelabrum grey-pruinescent;
palpus largely black, with variable yellow to
brownish apical zone. Thorax largely grey to greyish-
brown pruinescent; mesoscutum of male dark brown,
becoming grey posteriorly, intensity and extent of
colour-zones changing with angle of view, goldish
zones on and near notopleural areas visible from
some angles; mesoscutum of female (Fig. 21) grey
with complex pattern consisting essentially of five
longitudinal dark brown stripes and a dark blotch in
front of each postalar callus; scutellum brown-black,

with grey to yellowish anterolateral zone on each side
and yellowish apical spot which is larger in female.
Legs largely grey-brown; femora apically yellowish;
tibiae yellowish, each with diffuse brown sub-basal
ring; mid and hind tibiae with, in addition, brown
subapical ring, that on hind tibia much larger and
darker; tarsi yellowish, each usually with terminal
segment variably browned. Wing clear; pterostigma
buff to pale brown; part of costa adjoining end of
veins 2 and 3 slightly darkened, with very little brown
pigment often visible on membrane at ends of these
veins. Halter yellowish, with capitellum dark brown.

Figures 23-26. 23, Ironomyia francisi, wing. 24, I. whitei, wing tip. 25, I. francisi, epandrium and
associated parts, left lateral view. 26, I. whitei, the same. Scale for Figs 25 and 26 = 0.1 mm.

Proc. Linn. Soc. N.S.W., 129, 2008

31

NEW SPECIES OF IRONIC FLIES

Abdominal tergites 1 to 6 variegated black and pale
yellowish grey, approximately as in Fig. 22 in male,
with median pale zones broader in female, but extent
of zones changing with angle of view so that median
pale zones appear much darker in anterodorsal view;
tergite 6 lacking median black zone.

Head structurally similar to that of J.
nigromaculata.

Thorax. Scutellum quite without dorsal
setulae, with three or four pairs of marginal bristles
often irregular and asymmetrical.

Male postabdomen. Epandrium more slender
than in 1. nigromaculata, slightly narrowed basally,
setulose, most strongly so laterally on posterior half,
with well sclerotised anteroventral bridge in front of
hypandrium, with posterodorsal bight for insertion of
proctiger longer than in other species; surstylus stout
but somewhat curved, distally obliquely truncate so
that anterodistal angle is more acute than posterodistal
angle, with relatively few large setulae posteriorly,
and with numerous rather dense short setulae on
anterior surface; hypandrium relatively slender,
consisting of pair of narrow, anteriorly converging
and shortly fused plates with mostly small setulae;
aedeagus moderately elongate, sclerotised, curved,
relatively slender beyond base, with slender basal
apodeme; cerci well separated, setulose, markedly
narrower than in J. nigromaculata and I. whitei;
proctiger glabrous; apparent sternite 10 with pair of
rounded finely setulose prominences between bases
of surstyli; apparent sternite 11 narrower than in J.
nigromaculata, with pair of large setulae and several
small setulae.

Dimensions. Total length, ¢ 2.9-3.8 mm,
Q 2.7-3.4 mm; length of thorax, ¢ 1.2-1.7 mm, 9
1.3-1.6 mm; length of wing, 3 3.3-4.1 mm, 9 3.4-4.1
mm.

Distribution

Queensland: Brisbane district and North
Stradbroke Island. New South Wales: Grafton district
to Sydney district.

Notes

I. francisi is readily distinguished from
I. nigromaculata by having a pale median zone on
tergites 2 to 6 and no dorsal or seriate marginal setulae
on the scutellum. It also differs in details of the male
postabdomen as in above description and Fig. 25. For
comparison with /. whitei see under that species.

The specimens that I collected at Mooney
Mooney Creek were found on trunks of Acacia sp.
Most of the Queensland specimens were taken in
Malaise traps or at mercury vapour light.

By

The specific epithet refers to James Francis
(Frank) McAlpine, who established the family
Ironomyiidae (with J.E.H. Martin) and contributed
much to knowledge of its morphology and
relationships.

Ironomyia whitei sp. nov.
Figs 24, 26-28

Material examined

Holotype. 3, Tasmania: Pieman River, near
Rosebery, 15.1.1960, D.K.M. (AM).

Paratypes. Tasmania: 1 2, Pelion (Mount
Pelion vicinity, Cradle Mountain-Lake Saint Clair
National Park), 111.1990, LD.N. (ANIC); 1 &,
Clayton’s, near Melaleuca, Bathurst Harbour district,
43°23’S 146°08’E, 1.1991, E.S.N., E.D.E. (ANIC).
New South Wales: | 9, Carrai State Forest, W of
Kempsey, 30°54’33”S 152°16’28”E, 3-8.x11.1997,
E.T. (AM).

Description

Resembling /. francisi in most characters
and agreeing with description of that species, except
as indicated below.

Coloration. Head and antenna largely as
given for J. francisi. Palpus almost entirely blackish in
male, with tawny-brown apex in female. Mesoscutum
and scutellum with sexually dimorphic pattern
resembling that of I. francisi. Legs with markings
possibly resembling those of J. francisi but all
specimens somewhat faded. Wing with pterostigma
brown (darker in less faded holotype); apical brown
zone covering veins 2 and 3 and intervening area.
Abdominal pattern of male (Fig. 28): tergite 1 broadly
yellow-grey medially, with black paramedian zones,
yellowish brown lateral parts, and posterior part of
lateral margin black; posterior margin largely pale,
with transverse silvery-pruinescent zone on each side;
tergite 2 with moderately small, rounded pale median
zone, which appears yellowish in anterodorsal view,
silvery and slightly larger in posterodorsal view, with
large black zone on each side of median zone, with
lateral parts pale yellowish grey, and extreme lateral
margin apparently black; central part of posterior
margin black; sublateral part with transverse silvery
pruinescent zone; tergites 3 and 5 resembling tergite 2,
but median pale zone progressively narrower (lateral
margins not visible); tergite 4 also generally similar
to above, but narrow median pale zone consisting of
silvery pruinescence on a largely brown to blackish
ground-colour, so that this zone largely disappears in
anterodorsal view; tergite 6 tawny-yellow with two

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Figures 27, 28. Ironomyia whitei. 27, female tergites 1-6, diagrammatic. 28, male tergites 1-6.

pairs of dark brown zones connected on each side
by zone of brown suffusion, and with entire median
zone and posterior margin tawny-yellow. Abdominal
pattern of female: somewhat resembling that of male;
tergites 2 to 5 with narrow median stripe of silvery
pruinescence (most distinct in posterodorsal view)
either located on broader zone of yellowish ground
colour which grades into darker paramedian zone,
or delimited laterally by such dark zone; tergite 6
variable, with only the posterior pair of blackish
brown zones (Carrai specimen, Fig. 27) or more
extensively darkened (Tasmanian specimens).

Thorax. Scutellum with two or three pairs of
marginal bristles, without dorsal setulae.

Male postabdomen., Epandrium as described
for I. francisi, but much more elongate, slightly
gibbous basally; surstylus very stout, with medially
inclined subacute to narrowly obtuse (depending on
angle of view) apex, with numerous small and large
setulae on anterior surface and few large setulae
on and near posterior surface; hypandrium much
tapered anteriorly, not divided except at posterior
end around base of aedeagus; aedeagus larger than
in I. nigromaculata, stouter than in J. francisi and not
tapered; cercus broader than in /. francisi; sternite 10
relatively broad, with slight convexity on each side;
sternite 11 as described for J. francisi.

Proc. Linn. Soc. N.S.W., 129, 2008

Dimensions. Total length, 3 3.1 mm, @ 3.3-
3.7 mm; length of thorax, J 1.6 mm, 9 1.3-1.6 mm;
length of wing, ¢ c. 4.5 mm, @ c. 3.9-5.0 mm.

Distribution
Tasmania: western parts of state. New South
Wales: eastern edge of Northern Tablelands district.

Notes

I. whitei is readily distinguished from other
species of Ironomyia by the obvious dark brown to
blackish apical wing spot (represented at most by a
trace in other species), also by the distinctive colour
pattern of the abdominal tergites (Figs 27, 28) and
details of the male genitalia (Fig. 26).

In the holotype (the only known male) I
have not been able to confirm the presence of a black
lateral marginal zone on most abdominal tergites and
have omitted them from Fig. 28, though they may be
present. Females from Tasmania have the black zones
on the tergites more extensive than that from Carrai
Forest, New South Wales (Fig. 27), but observed
variation is no greater than that in . nigromaculata.

This species has been rarely collected. I
obtained many flies by sweeping vegetation during
an exceptionally hot day in a relatively cool rainforest
gully leading to the Pieman River, near Rosebery,

33

NEW SPECIES OF IRONIC FLIES

Tasmania, in Jan. 1960. These included heleomyzids
(genera Austroleria McAlpine, Diplogeomyza
Hendel, Trixoleria McAlpine), lauxaniids (genera
Ceratolauxania Hendel, Jncurviseta Malloch,
Sapromyza s.l), and other flies including the holotype
of Ironomyia whitei.

The specific epithet refers to Arthur White,
pioneer student of Tasmanian Diptera, who named
the genera Jronomyia and Sciadocera.

ACKNOWLEDGEMENTS

J.M. Cumming, G. Daniels, W.N. Mathis, K. Ribardo,
J.H. Skevington, F.C. Thompson, and D.K. Yeates provided
significant study material. J. Chainey and E.D. Edwards
gave information on material in their care. D.J. Bickel, S.F.
McEvey, and G. Theischinger made useful comments on the
manuscript. M.S. Moulds and D.J. Bickel also gave support
to this study. S. Lindsay carried out electron microscopy.

S. Cowan prepared electronic copy.

REFERENCES

Chandler, P.J. (1998). 3.2. Family Opetiidae. In
“Contributions to a Manual of Palaearctic Diptera’
(Ed. L. Papp & B. Darvas) 3, 17-25. (Budapest,
Science Herald).

Chandler, P.J. (2001). The flat-footed flies (Diptera:
Opettidae and Platypezidae) of Europe. Fauna
entomologica scandinavica 36, 280 pp.

Colless, D.H. (1994). A new family of muscoid Diptera
from Australasia, with sixteen new species in four
new genera (Diptera: Axiniidae). Invertebrate
Taxonomy 8, 471-534.

Collins, K.P. & Wiegmann, B.M. (2002). Phylogenetic
relationships of the lower Cyclorrhapha (Diptera:
Brachycera) based on 28S rDNA sequences. Insect
Systematics and Evolution 33, 445-456.

Disney, R.H.L. (1988a). Unusual costal chaetotaxy in
the phylogenetically interesting Ironomytidae and
Sciadoceridae (Diptera). Annales entomologici
fennici 54, 19-20.

Disney, R.H.L. (1988b). The form of articulation between
the pedicel and first flagellar segment of the antenna
in flies (Diptera). The Entomologist 107, 99-103.

Disney, R.H.L. (1994). ‘Scuttle flies: the Phoridae’, 467
pp. (Chapman and Hall, London).

Disney, R.H.L. (2001). Sciadoceridae (Diptera)
reconsidered. Fragmenta faunistica 44, 309-317.

Ferrar, P. (1988). A guide to the breeding habits and
immature stages of Diptera Cyclorrhapha, 907 pp.
(E.J. Brill, Leiden).

Griffiths, G.C.D. (1972). The phylogenetic classification
of the Diptera Schizophora with special reference to
the structure of the male postabdomen, 340, pp. (W.
Junk, The Hague).

34

Grimaldi, D.A. and Cumming, J. (1999). Brachyceran
Diptera in Cretaceous amber and Mesozoic
diversification of the Eremoneura. Bulletin of the
American Museum of Natural History 239, 124 pp.

Hennig, W. (1948). Die Larvenformen der Dipteren 1, 185
pp. (Akademie-Verlag, Berlin.)

Hennig, W. (1973). 31. Diptera (Zweifltigler). Handbuch
der Zoologie 4(2), 2, 337 + 4 unnumbered pp.

Kessel, E.L. (1987). 50. Platypezidae. In J.F. McAlpine
(editor): Manual of Nearctic Diptera 2, 681-688.
(Canadian Government Publishing Centre, Hull,
Quebec).

Lowne, B.T. (1895). ‘The anatomy, physiology,
morphology, and development of the blow-fly
(Calliphora erythrocephala)’, 778 pp. (R.H. Porter,
London.)

McAlpine, D.K. (1997). Gobryidae, a new family of
acalyptrate flies (Diptera: Diopsoidea), and a
discussion of relationships of the diopsoid families.
Records of the Australian Museum 49, 167-194.

McAlpine, D.K. (2002). Some examples of reduced
segmentation of the arista in Diptera-Cyclorrhapha,
and some phylogenetic implications. Studia
dipterologica 9, 3-19.

McAlpine, D.K. (2007). The surge flies (Diptera,
Canacidae, Zaleinae) of Australasia and notes on
tethinid-canacid morphology and relationships.
Records of the Australian Museum 59, 27-64.

McAlpine, J.F. (1967). A detailed study of Ironomyiidae

(Diptera: Phoroidea). Canadian Entomologist 99,

225-236.

McAlpine, J.F. (1973). A fossil ironomyiid fly from

Canadian amber (Diptera: Ironomyiidae). Canadian

Entomologist 105, 105-111.

McAlpine, J.F. (1981). 2. Morphology and terminology —
adults. In J.F. McAlpine (editor): Manual of Nearctic
Diptera 1, 9-63. (Canadian Government Publishing
Centre, Hull, Quebec).

McAlpine, J.F. (1989). 116. Phylogeny and classification
of the Muscomorpha. In J.F. McAlpine (editor):
Manual of Nearctic Diptera 3, 1397-1518. (Canadian
Government Publishing Centre, Hull, Quebec).

McAlpine, J.F. and Martin, J.E.H. (1966). Systematics of
Sciadoceridae and relatives with descriptions of two
new genera and species from Canadian amber and
erection of family Ironomyiidae (Diptera: Phoroidea).
Canadian Entomologist 98, 527-544.

Mostovski, M.B. (1995). New taxa of Ironomytidae
(Diptera: Phoromorpha) from the Cretaceous of
Siberia and Mongolia. Paleontologicheskii Zhurnal
29, 318-331. In Russian.

Peterson, B.V. (1987). 49. Lonchopteridae. In J.F.
McAlpine (editor): Manual of Nearctic Diptera 2,
675-680. (Canadian Government Publishing Centre,
Hull, Quebec).

Sabrosky, C.W. (1999). Family-group names in Diptera.
Myia 10, 1-360.

Schmitz, H. (1929). Fascicle 1. -Sciadoceridae and
Phoridae. Diptera of Patagonia and South Chile 6,
1-42.

Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Sinclair, B.J. (1995). New species of Hormopeza
Zetterstedt from South Africa and Tasmania (Diptera:
Empididae). Annals of the Natal Museum 36, 203-
208.

Speight, M.C.D. (1969). The prothoracic morphology
of acalyptrates (Diptera) and its use in systematics.
Transactions of the Royal Entomological Society 121,
325-421.

Stocker, R.F. (2001). Drosophila as a focus in olfactory
research: mapping of olfactory sensilla by fine
structure, odor specificity, odorant receptor
expression, and central connectivity. Microscopy
Research and Technique 55, 284-296.

Stuckenberg, B.R. (2001). Pruning the tree: a critical
review of classifications of the Homeodactyla
(Diptera, Brachycera), with new perspectives and
an alternative classification. Studia dipterologica 8,
3-41.

Tonnoir, A.L. (1926). A new and primitive sub-family
of the Phoridae (Dipt.). Records of the Canterbury
Museum [Christchurch] 3, 31-38, pl. 4.

Wada, S. (1991). Morphologische Indizien fiir das
unmittelbare Schwestergruppenverhaltnis der
Schizophora mit den Syrphoidea (“Aschiza’) in der
phylogenetischen Systematik der Cyclorrhapha
(Diptera: Brachycera). Journal of Natural History 25,
1531-1570.

White, A. (1916). The Diptera-Brachycera of Tasmania.
Part III. Families Asilidae, Bombylidae, Empidae,
Dolichopodidae & Phoridae. Royal Society of
Tasmania Papers and Proceedings 1916, 148-266.

Wiegmann, B.M., Yeates, D.K., Thorne, J.L., and Kishino,
H. (2003). Time flies, a new molecular time scale for
brachyceran fly evolution without a clock. Systematic
Biology 52, 745-756.

Woodley, N.E. (1989). 115. Phylogeny and classification
of the ‘Orthorrhaphous’ Brachycera. In J.F. McAlpine
(editor): Manual of Nearctic Diptera 3, 1371-1396.
(Canadian Government Publishing Centre, Hull,
Quebec).

Proc. Linn. Soc. N.S.W., 129, 2008

35

NEW SPECIES OF IRONIC FLIES

APPENDIX
Classification of genera mentioned in text.
Suborder Brachycera
Infraorder Homoeodactyla
Athericidae
Atherix Meigen
Suragina Walker
Suraginella Stuckenberg
Rhagionidae

Symphoromyia Frauenfeld
Infraorder Heterodactyla s.1.
Division Asiliformae
Division Eremoneura
Subdivision Empidiformae
Empididae
Hormopeza Zetterstedt
Subdivision Cyclorrhapha
Informal grade ‘lower Cyclorrhapha’ s.str
Opetiid group
Opetiidae
Opetia Meigen
Platypezoidea
Platypezidae
Agathomyia Verrall
Lindneromyia Kessel
Melanderomyia Kessel
Microsania Zetterstedt
Lonchopteroidea
Lonchopteridae
Lonchoptera Meigen
Phoroidea
Sciadoceridae
Archiphora Schmitz
Sciadocera White
Phoridae
Sinolestine group
Sinolestinae
Eridomyia Mostovski
Hermaeomyia Mostovski
Palaeopetia Zhang (= Sinolesta Hong & Wang)
Ironomyiid group
Ironomyiidae
Cretonomyia J. McAlpine
Ironomyia White
Uncertain group
Lebambromyia Grimaldi & Cumming
Cohort Eumuscomorpha
Syrphid group
Syrphidae
Deineches Walker
Eristalis Latreille
Melangyna Verrall
Microdon Meigen

36 Proc. Linn. Soc. N.S.W., 129, 2008

D.K. McALPINE

Pipunculid group
Pipunculidae
Eudorylas Aczél
Group Schizophora
Acartophthalmidae
Acartophthalmus Czermny
Agromyzidae
Liriomyza Mik
Astetidae
Asteia Meigen
Aulacigastridae
Aulacigaster Macquart
Canacidae (including Tethinidae)
Dasyrhicnoessa Hendel
Zalea D. McAlpine
Clusiidae
Allometopon Kertész
Tetrameringia D. McAlpine
Coelopidae
Gluma D. McAlpine
Cypselosomatidae
Clisa D. McAlpine
Drosophilidae
Drosophila Fallén
Scaptomyza Hardy
Ephydridae
Gobryidae
Gobrya Walker
Heteromyzidae (Heleomyzidae) s.l.
Austroleria D. McAlpine
Borboroides Malloch
Diplogeomyza Hendel
Heleomicra D. McAlpine
Tapeigaster Macquart
Trixoleria D. McAlpine
Huttoninidae
Huttonina Tonnoir & Malloch
Lauxaniidae
Ceratolauxania Hendel
Incurviseta Malloch
Sapromyza Fallén
Micropezidae
Badisis D. McAlpine
Compsobata Czerny
Cothornobata Czerny
Metopochetus Enderlein
Mimegralla Rondani
Neminidae
Nemo D. McAlpine
Neurochaetidae
Neurochaeta D. McAlpine
Neurocytta D. McAlpine
Neurotexis D. McAlpine
Nothoasteia Malloch

Proc. Linn. Soc. N.S.W., 129, 2008

NEW SPECIES OF IRONIC FLIES

Nothybidae

Nothybus Rondani
Neriidae

Telonerius Aczél
Odintidae

Traginops Coquillett
Pallopteridae

Maorina Malloch
Periscelididae

Cyamops Melander

Stenomicra Coquillett
Platystomatidae

Euprosopia Macquart

Lenophila Guérin-Méneville

Rivellia Robineau-Desvoidy

Rhytidortalis Hendel
Sepsidae

Lasionemopoda Duda
Somatiidae

Somatia Schiner
Tanypezidae

Strongylophthalmyia Heller
Teratomyzidae

Teratomyza Malloch
Calliphoridae

Calliphora Robineau-Desvoidy
Muscidae s.1.

Fannia Robineau-Desvoidy

Musca Linné
Scathophagidae

Scathophaga Meigen
Tachinidae

Chetogaster Macquart

Proc. Linn. Soc. N.S.W., 129, 2008

Early Natural History of the Greater Glider, Petauroides volans

(Kerr, 1792)

K. SHANE MALoney! AND JAMIE M. Harris?

' School of Biological Sciences, University of Wollongong, NSW 2522 (ksm99@uow.edu.au);

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

Gharril 1 @scu.edu.au)

Maloney, K.S. and Harris, J.M. (2008). Early natural history of the greater glider, Petauroides volans (Kerr,
1792). Proceedings of the Linnean Society of New South Wales 129, 39-55.

Early accounts of the greater glider Petauroides volans (Marsupialia: Pseudocheiridae) are reviewed,
starting with Arthur Phillips’ 1789 account in The Voyage of Governor Phillip to Botany Bay and proceeding
to the latest taxonomic works. This species has a quite complicated and confusing taxonomic history. It
has been listed as a member of no fewer than 10 genera with about 23 different binomial names since its
discovery. In this paper, we review some of this taxonomic complexity and early descriptions of the species’
morphology, dentition, behaviour, distribution and abundance. We found that taxonomic descriptions of P.
volans have been frequently confused with those of a number of other gliding possums, particularly the
yellow-bellied glider Petaurus australis. Early descriptions of the morphology of P. volans were given only
in broad general terms. More value can be placed on the early behavioural observations, and on the earliest
records of its occurrence. This paper examines some of the oldest accounts of P. volans and assesses their

significance.

Manuscript received 14 Februry 2007, accepted for publication 12 December 2007.

KEYWORDS: natural history, nomenclature, Petauroides volans, Pseudocheiridae.

INTRODUCTION

The greater glider, Petauroides volans
(Marsupialia: Pseudocheiridae), is the largest gliding
marsupial and is endemic to eastern mainland
Australia (McKay 1995). Currently, there are two
recognised sub-species: P. volans volans, which
occurs in south eastern Australia (from Victoria in
the south, through mainly coastal New South Wales
(NSW) to the Rockhampton district in north-east
Queensland (Qld)); and P. v. minor, which occurs in
very far north-east Qld (from the Dawson River to the
Barron River) (Flannery 1994). It is around the size of
a domestic cat, with females being larger than males
(Flannery 1994; Kavanagh and Wheeler 2004). Most
individuals are jet black on the dorsum and creamy
white on the ventrum, but pure white forms are not
uncommon and intermediate colours are also found
(Flannery 1994; McKay 1995; Lindenmayer 2002).
This species is nocturnal, arboreal and folivorous
and is dependent on tree hollows for its nesting
requirements.

Petauroides volans is relatively conspicuous and
was quickly noticed by the early colonists (Phillip
1789). Subsequently, descriptions of this species were

included in many of the earliest zoological accounts
of the Australian fauna. However, few modern
zoologists are aware of the historical significance
and value of this old literature as it relates to this and
other species (see also Harris 2006). Whilst some
of this literature on P. volans has been reviewed by
McKay (1982), this was limited to aspects of the
nomenclature of the genus name Petauroides (and
also Petaurus). In this contribution, we have sought
to provide a comprehensive survey of the early
natural history literature pertaining to P volans,
including information on discovery, taxonomy,
dentition, morphology, distribution, abundance, diet
and behaviour.

TAXONOMY AND NOMENCLATURE

Governor Arthur Phillip reported ‘black flying
opossum’ from NSW (Phillip 1789). A male specimen
owned by Henry Constantine Nowell was illustrated
(Figure 1), although no details on the precise
collection locality were published. Presumably it was
found in the vicinity of Port Jackson. Phillip (1789,
1790) recognised that it represented a new species
and suggested taxonomic affinity with American

EARLY HISTORY OF THE GREATER GLIDER

Figure 1: Black flying opossum (=Petauroides volans) drawn by P. Mazell and published in Phillip (1789).
Note the opposable clawless hallux and syndactylous digits on each of the hind feet.

40 Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

Didelphis, although a specific name was not offered.
A few years later, Kerr (1792) named Phillip’s
specimen Didelphis volans, derived from the Latin
word ‘volare’ meaning ‘to fly’ (Strahan 1981). A
year later, Meyer (1793) named Phillip’s specimen
D. voluccella, and a year later still, Shaw (1794)

proposed the name D. macroura. In Shaw’s work,
The Zoology of New Holland, a juvenile specimen
drawn by James Sowerby was illustrated (Figure 2).
Shaw (1800) explained that it was sent to him by
John White, who was the first Surgeon-General for
the colony of NSW.

Figure 2: Long-tailed opossum Didelphis macroura (=Petauroides volans) from Zoology of New Holland
(1794) by George Shaw. The figure was drawn by James Sowerby. This illustration was also reproduced

in Shaw (1800) and Desmarest (1820).

Proc. Linn. Soc. N.S.W., 129, 2008

41

EARLY HISTORY OF THE GREATER GLIDER

Cuvier (1798) followed use of the name D.
volans (Kerr 1792) but questioned the affiliation
with the genus Didelphis. Nevertheless, Shaw (1800)
continued the use of D. macroura. Bechstein (1800)
elevated the name Voluccella, used by Meyer (1793),
to generic level and proposed V. nigra for the subject
species, but he evidently confused the greater glider
and the yellow-bellied glider Petaurus australis in
synonymy. His proposed V. nigra incorporated D.
voluccella Meyer, 1793 (=Petauroides volans) and
“Hepoona Roo” White, 1790 (=Petaurus australis).
It is understood that Hepoona Roo is P. australis
and not Petauroides volans (McKay 1982, 1988).
Bechstein (1800) also advanced V. macroura as a
separate species that incorporated D. volans Kerr,
1792 and D. macroura Shaw, 1794. Thus, V. nigra
and V. macroura are both synonyms of P. volans.
Voluccella Bechstein, 1800 was discontinued for the
subject species because this genus name had already
been advanced by Fabricius (1794) for a species of fly
(Diptera: Bombyliidae) (Thomas 1888; McKay 1988;
Evenhuis 1991). Hence, Voluccella Bechstein, 1800
is a Junior generic synonym for Petauroides but not
Voluccella Fabricius, 1794.

Phalanger volans was used by Lacépéde (1801),
whilst Desmarest (1803) and Tiedemann (1808)
placed it under Phalangista (see also Schinz 1821;
Thomas 1888). Turton (1806) mistakenly thought
that the descriptions by Kerr (1792: D. volans) and
Shaw (1794: D. macroura) represented two separate
species. Oken (1816) made a similar mistake, but
also erroneously included Petaurus australis in the
synonymy for one of his proposed species. This was
Petaurus niger, and the epithet was a gender change
of Bechstein’s (1800) nigra (see also Iredale and
Troughton 1934). Oken’s (1816) second species was
Petaurus macroura.

Desmarest (1817) listed three species (Petaurus
macrourus, P. peronii and P. taguanoides). Petaurus
macrourus included a slight change in the epithet to
standardise the gender of the binomial. Desmarest’s
explanation thatthe membrane of P. peronii “terminates
at the elbow” is good evidence that this specimen
was also P. volans. For P. taguanoides however, the
synonymy was confused with the yellow-bellied
Glider [i.e. Didelphis petaurus of Shaw (1791) and
“Hepoona Roo” of White (1790)]| and the descriptions
about the patagium ending at the wrist suggested to us
that this specimen was not the greater glider. However,
according to the publications of the Muséum National
d’ Histoire Naturelle (MNHN) the type specimen of P.
taguanoides as described by Desmarest is indeed P.
volans (de Beaufort 1966; Julien-Laferriére 1994). To
confirm this identification we contacted the MNHN

directly, and obtained a photograph of the specimen
(number CG1990-408) and although no patagium
was evident in the photograph, it looks like a greater
glider because of its substantially long tail and hairy
ears. The arrangement of Desmarest’s (1817) was
later followed by Cuvier (1826), Lesson (1827, 1828,
1830, 1838), and Fischer (1829). Bennett (1837) also
used Desmarest’s (1817) terminology, although he
appears to have used P. peronii in reference to the
sugar glider Petaurus breviceps.

Desmarest (1820) applied Petaurista to supercede
Petaurus, and maintained Petaurista taguanoides, P.
macroura and P. peronii as separate species (later
followed by Cuvier 1827, 1829). However, this was
flawed as Petaurista had been advanced for the giant
flying squirrels (Rodentia) by Link (1795) (see also
Fischer 1814; Thomas 1888; Sherborn 1902; Palmer
1904). Waterhouse (1838b), Gloger (1842), Gould
(1863) and Thomas (1885), persisted with this invalid
generic name for the greater glider.

Frédéric Cuvier (1825) mentioned Petaurus
didelphoides Geoffroy, an apparent new name for the
subject species (Thomas 1888; Iredale and Troughton
1934; de Beaufort 1966). However, later works by
F. Cuvier and also his brother Georges, made no
reference to P didelphoides (Cuvier 1826, 1827,
1829). de Beaufort (1966) noted that Cuvier (1825)
offered no specific descriptions, and stated that he
was unable to find any reference to Geoffroy as the
authority for the name. It is uncertain whether Cuvier
intended this name for the greater glider. Iredale and
Troughton (1934) considered it a vernacular name.

Lesson (1828, 1830, 1838) listed the “Black
Flying Opossum” of Phillip (1789) (=the greater
glider) as a junior synonym of Petaurus taguanoides.
This was subsequently repeated by Fischer (1829),
Wagner (1843), Schinz (1844) and Giebel (1859).
Waterhouse (1838a) then stated that two specimens of
Petaurista taguanoides were held in the Museum of
the Zoological Society of London (ZSL), one of which
was a ‘white variety’. Waterhouse’s (1841) included
an illustration of a greater glider (Figure 3) and stated
that “Specimens which are totally white, and others
which are white and irregularly variegated with grey,
are not rare”. Waterhouse (1841) was wrong when
he suggested that P. macrourus is P. flaviventer (=P.
australis) (see also Wagner 1855; Giebel 1859; Gould
1863). Descriptions of taguanoides specimens in many
19" century publications subsequent to Waterhouse
(1841) appear to represent the greater glider (e.g.
Owen 1841, 1845; Gloger 1842; Waterhouse 1846;
Gervais 1855: Gerrard 1862; Brehms 1880; Flower
1884; Forbes-Leith and Lucas 1884; Krefft 1864;
Haswell 1886; Jentink 1886; Lucas 1890).

Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

Figure 3: A print from Waterhouse (1841) that is clearly Petauroides
volans because of the length of the tail and the hairy ears. This
image was also reproduced in Waterhouse (1843) and Lydekker
(1896).

Major T.L. Mitchell collected a presumed new
species of glider “from the banks of the Murray”,
named it Petaurus leucogaster and “deposited [it]
in the Australian Museum (AM)” (Bennett 1837;
Mitchell 1838). Gray (1841) suggested that it
“may only be a variety of [the] P taguanoides” of
Waterhouse (=the greater glider) (see also mention
of P. leucogaster in Gray 1842, 1843; Krefft 1864).
Several authors considered J/eucogaster to be
synonymous with P volans (Gould 1863; Thomas

Proc. Linn. Soc. N.S.W., 129, 2008

1888; Iredale and Troughton 1934;
McKay 1982). However, McKay
(1988) stated that P Jeucogaster
was ‘Incertae sedis’ (of uncertain
position) because the specimen
could no longer be found at the
AM. He suggested that the locality
for Mitchell’s specimen was outside
the range of P. volans and may have
been Petaurus norfolcensis.

M.R. Oldfield Thomas, of the
British Museum of Natural History
(BMNH), revised the taxonomy of
the subject species several times
during the period 1879-1923. Thomas
(1879) noted that the specific name
volans Kerr antedated taguanoides
Desmarest, and maintained that
the correct binomial was Petaurus
volans. A few years later, however, he
listed it as Petaurista volans (Thomas
1885). After finding that Petaurista
was unavailable, Thomas (1888)
advanced Petauroides to replace the
previous generic names. He listed
two subspecies: Petauroides volans
typicus as the southern form; and P. v.
minor as the northern form (following
Collett 1887). Later, Thomas (1923)
received further examples from Qld
and considered that there were two
additional subspecies: Pv. incanus
and P. vy. armillatus.

Thomas (1923) mentioned
that Ogilby (1892) referred to “Dr
Ramsay’s P. cinereus” and that it
“seems never to have been described”.
However, Ramsay (1890) did indeed
publish a description of a supposed
new species, which he named
Petaurides cinereus. This was based
on two specimens obtained from
the Bellinden-Ker Range, north-
east Qld. The name Petaurides is a
definite misspelling of Petauroides Thomas 1888 (see
Ramsay 1890). It is also noted that these specimens
had earlier been exhibited at a meeting of The Linnean
Society of NSW under the name of Belideus cinereus
(Anon 1890).

The next taxonomic contribution was by Iredale
and Troughton (1934). They argued that the generic
name Schoinobates Lesson 1842 had been published
before Petauroides Thomas 1888, and advanced the
name S. volans with four subspecies: S. v. volans;

43

EARLY HISTORY OF THE GREATER GLIDER

S. v. incanus; S. v. armillatus and S. v. minor.
Subsequently, S. volans was in use for around 50 years
(Fleay 1947, 1968; Tate 1945; Anon 1946; Troughton
1935, 1941; Marlow 1958, 1962; de Beaufort 1966;
Ride 1970; Strahan 1980, 1981). However, the
nomenclatural change by Iredale and Troughton
(1934) was groundless. McKay (1982) pointed out
that Schoinobates was first used by Lesson (1842)
to supersede Petaurista leucogenys Temminck, 1838
(=Pteromys leucogenys; the Japanese flying squirrel).
In fact, this was an error on Lesson’s part because
there are no marsupials in Japan (Palmer 1904).
Nevertheless, it was highly irregular for Iredale and
Troughton to amend the type locality of P. Jeucogenys
from “Japan” to “Sydney”. Probably, Iredale and
Troughton (1934) did not view the original account
and illustration of P. Jeucogenys in Fauna Japonica
(Temminck 1838), which clearly depicts a sciurid.
Schoinobates Lesson, 1842, is therefore properly
placed as a junior synonym of Petaurista Link, 1795.
Thus, McKay’s (1982) assessment that the name
Schoinobates was unavailable and that Petauroides
must stand was justified.

Iredale and Troughton (1934) also nominated
Petaurus maximus as a synonym for the subject
species, listing Partington (1837) as the authority. This
was accepted by McKay (1982) and Flannery (1994).
However, McKay later attempted unsuccessfully
to track down the original reference and stated that
the relevant page in the book he examined “contains
no reference to this or any other mammal” (McKay
1988). We note that McKay (1988) misread Iredale
and Troughton’s (1934) reference to Partington (1837:
424) because P. maximus is indeed described in The
British Cyclopedia of Natural History, but not in the
The British Cyclopedia of Arts and Sciences, which
was read by McKay (1988). After reading Partington
(1837) with its reference to some “almost white”
specimens, we accept P. maximus as synonymous with
the greater glider (following Iredale and Troughton
1934). The preceding literature review of taxonomy
of the Greater Glider is presented in Table 1.

Common names for the subject species
have included ‘black flying opossum’ (Phillip
1789), ‘flying opossum’ (Kerr 1792; Turton 1806;
Waterhouse 1841), ‘long-tailed opossum’ (Shaw
1794, 1800; Turton 1806; Waterhouse 1841), ‘large-
tailed Petaurista’, ‘Peron’s Petaurista’ (Cuvier 1827),
‘white-bellied flying squirrel’ (Bennett 1837), ‘grey
flying squirrel’ (Bennett 1837; Waterhouse 1841)
‘large-tailed flying squirrel’ (Bennett 1837), ‘taguan
flying opossum’ (Waterhouse 1838b), ‘taguan flying
phalanger’ (Waterhouse 1846; Thomas 1888, 1923;
Fleay 1933), ‘greater flying phalanger’ (Gould 1863;

44

LeSouef and Burrell 1926; Fleay 1933), ‘the brill’ (De
Vis 1886), “flying phalanger’ (Haswell 1886), ‘great
flying oposssum’, ‘flying squirrel’ (Lucas 1890),
‘dusky glider’ (Fleay 1933; Ride 1970), ‘greater
glider-possum’ (Iredale and Troughton 1934; Anon
1946), and ‘greater glider’ (Marlow 1958). Stability
in the vernacular name was achieved in 1980 when
a committee of the Australian Mammal Society
formalised it as the ‘greater glider’ (Strahan 1980).

MORPHOLOGY

The morphology was first described by Phillip
(1789). He stated that the “tip of the nose to root of tail
[was] 20 inches [=508 mm], tail 22 inches [=559 mm],
loins 16 inches [=406 mm].” The ears were described
as “large and erect”, the fur “glossy black” on top,
“mixed with grey”, and “the under parts...white”. It
was noted that the fur “continued to the claws”, and
that the membrane “expanded on each side of the
body”. Phillip (1789) also described and illustrated
the foot (Figure 1). He observed that the “fore legs
have five toes on each foot, with a claw on each; the
hinder ones four toes, with claws, (the three outside
ones without any separation) and a thumb without a
claw”. Following Phillip (1789), similar descriptions
were also published by subsequent authors based on
his original account and from the illustration provided
(1.e. Kerr 1792; Meyer 1793; Bechstein 1800). Shaw
(1794) provided morphological descriptions based on
the illustration reproduced in Figure 2.

One diagnostic feature of P. volans is the flying
membrane that runs from the elbow to the knee, and
this was noted by several early zoologists (Kerr 1792;
Turton 1806; Desmarest 1817; Waterhouse 1841,
1846). Thomas (1888) added that the membrane is
“very narrow along the sides of the forearm and lower
leg”. Ramsay (1890) stated that the “parachute” or
“wing membrane” commences a little in front of the
elbow-joint, and extends to about half-way below the
knee-joint. Numerous early authors also noted the
syndactylous hind feet (Kerr 1792; Shaw 1794, 1800;
Bechstein 1800; Lacepede 1801; Tiedemann 1808;
Desmarest 1820; Partington 1837).

Some authors have compared the size of this
Species to animals known from Europe. For example,
it has been suggested to be about the size of a “black
rat’ (Shaw 1800), “flying squirrel” (Desmarest
1803; Tiedemann 1808), “surmulot” (Cuvier 1817;
Desmarest 1820; Lesson 1827), “squirrel of Europe”
(Desmarest 1820; Lesson 1827), and “brown
rat’ (Partington 1837). More recently, it has been
suggested to be about the size of a domestic cat

Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

Table 1: New synonymy based on the current review.

Petauroides Thomas, 1888
Petaurus Shaw, 1791
Didelphis Kerr, 1792
Voluccella Bechstein, 1800
Phalanger Lacepede, 1801
Phalangista Desmarest, 1803
Petaurista Desmarest, 1820
Petauroides Thomas, 1888
Petaurides Ramsay, 1890
Belideus Anon, 1890
Schoinobates Iredale and Troughton, 1934
Petauroides volans (Kerr, 1792)

Petauroides volans volans (Kerr, 1792)
Didelphis volans Kerr, 1792
Didelphis voluccella Meyer, 1793
Didelphis macroura Shaw, 1794
Voluccella nigra Bechstein, 1800
Voluccella macroura Bechstein, 1800
Phalanger volans Lacepede, 1801
Phalangista volans Desmarest, 1803
Petaurus macroura Oken, 1816
Petaurus niger Oken, 1816

Petaurus taguanoides Desmarest, 1817
Petaurus macrourus Desmarest, 1817
Petaurus peronii Desmarest, 1817
Petaurista taguanoides Desmarest, 1820
Petaurista macroura Desmarest, 1820
Petaurista peronii Desmarest, 1820
Phalangista macroura Schinz, 1821
Petaurus didelphoides Cuvier, 1825
Petaurus maximus Partington, 1837
Petaurus volans Thomas 1879

Petaurista volans Thomas 1885
Petauroides volans typicus Thomas, 1888
Petauroides volans incanus Thomas, 1923
Petauroides volans armillatus Thomas, 1923

Schoinobates volans volans Iredale and Troughton 1934

Schoinobates volans incanus Iredale and Troughton 1934

Schoinobates volans armillatus \redale and Troughton 1934
Petauroides volans minor (Collett, 1887)

Petaurista volans minor Collett, 1887

Belideus cinereus Anon, 1890

Petaurides cinereus Ramsay, 1890

Schoinobates volans minor Iredale and Troughton 1934

(Flannery 1994).

Colouration was also frequently commented on.
For example, Cuvier (1817) reported that the fur exists
in different tones of brown, with many varieties, and
others are whitish. Fully white specimens were also
noted (Lesson 1827; Waterhouse 1841; Krefft 1864;

Proc. Linn. Soc. N-S:W., 129, 2008

Le Souef and Burrell 1926). Gould (1863) stated that
“it is subject to very great variety in the colouring of
its fur, some specimens being entirely blackish brown
[see Figure 4], grey to cream and others quite white”.
Krefft (1871) reported that the species “varies much
from creamy-white to spotted black and white and

45

EARLY HISTORY OF THE GREATER GLIDER

Figure 4: Petaurista taguanoides from Gould (1863) (= P. volans). Note: The front arms of the back-
ground glider are shown in the wrong position as P. volans tucks them under the chin when gliding
(Fleay 1933; Grzimek 1967; McKay 1989).

46 Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

Table 2: Dental formulas provided in the early natural history literature for Petauroides volans. Abbre-
viations: I = Incisors; C = Canines; M = Molars; P = Premolars. For Thomas (1885, 1888) an asterisk
indicates that the tooth is sometimes or commonly absent.

Source Dental formula
Desmarest 1820 6 Tee ey 6 6 7
= 3C = Op —— WI Op —— = Si) oye
2 00 22 66 66
Cuvier 1825, 1826 6 00 88
J Meee Ws OS
2 00 68
Lesson 1827, 1830, 1838; 6 00 8 8
Fischer 1829 lipoma mca OS
2 00 V7
Waterhouse 1838b 23 Pine 44
= Cy ple a lp Sh
11 Oey! 44
Krefft 1871; Collett 1887 B13 einige 44
mind Cone eam — = 40
11 ly owdhe
Thomas 1885 HOD 3 a. I 23 aged
I———C— P———M -——— x2 = 34 or 40
HOO 42435 1 2.3)4
Thomas 1888 12 3 1 1 034 1234
mn Oca i mm (Cl) ae ee eee et (at most) 3 (On lo-4)x 2— 40
I 2BEOo Oso 03*4 1234

perfect black, beneath the fur is always white.” Le
Souef and Burrell (1926) stated that “as a rule [the]
colour [is] darker in winter than in summer.” They
also stated that “animals from Gippsland (Victoria)
[were] dead black above and on tail; pure white on
undersides”, whereas Qld and NSW specimens were
“usually smoky grey” and “white specimens [were
reportedly] common.”

Other notable morphological features described
in the early literature include the ears, tail and size
differences between the sexes. Waterhouse (1838a,
1841, 1846) stated that “the ears are entirely covered
externally with long and dense fur, flesh-coloured and
almost bare within” (see also Krefft 1864, Thomas
1888, Ramsay 1890). The tail was reported as not
being prehensile (Lacepede 1801; Tiedemann 1808;
Partington 1837), and longer than the body (Shaw
1800; Turton 1806; Cuvier 1817). Thomas (1888)
described and illustrated the naked tip of the tail.
Gould (1863) stated the “sexes offer no external
difference, except that the female is somewhat
smaller than the male” (see Flannery 1994, as this is
erroneous). Various other aspects of the morphology
of this species are discussed in the literature, but
lack of space precludes a detailed discussion here.
However, these aspects include skull structure
(Waterhouse 1846; Collett 1887; Thomas 1888) and
myology (Haswell 1886).

Proc. Linn. Soc. N.S.W., 129, 2008

DENTITION

Phillip (1789) stated that in “the upper jaw
forwards are four small cutting teeth, then two canine
ones, and backwards five grinders: the under jaw
has two long large cutting teeth, five grinders, with
no intermediate canine ones, the space being quite
vacant’. Similarly worded descriptions were provided
by Kerr (1792) and Turton (1806).

A dental formula for the species was first provided
by Desmarest (1820) (see Table 2). He counted six
upper and two lower incisors, but was uncertain about
the number of canines and premolars. This uncertainty
led him to indicate a total of 32 or 34 teeth. Cuvier
(1825) and Lesson (1827) counted a total of 38 teeth.
Cuvier (1825) reported that the space between the
incisors and molars is occupied by two rudimentary
teeth. Waterhouse (1838b, 1841) and Owen (1841,
1845) mentioned they had never observed any of
these diminutive teeth in the specimens they had
examined. Waterhouse (1841) suggested that Cuvier
(1825) may have inadvertently described the dentition
of Phalangista cookii (=Pseudocheirus peregrinus;
common ringtail possum). These two species do
have great similarity in their dental characteristics, as
noted by early zoologists (Owen 1841, 1845; Giebel
1853, 1855; Thomas 1885; Collett 1887) and more
modern authors (Tate 1945; Triggs 1996). Waterhouse
(1838b) provided a dental formula indicating a total
of 34 teeth. Subsequent authors concurred with this

47

EARLY HISTORY OF THE GREATER GLIDER

observation (Waterhouse 1841, 1846; Wagner 1843;
Collett 1887; Ramsay 1890). Early illustrations of the
dentition in Cuvier (1825, 1827), Waterhouse (1846)
and Giebel (1853, 1855) support the dental formula
of Waterhouse (1838b).

Krefft’s (1871) dental formula (Table 2) was for
a total of 40 teeth (see also Collett 1887). Thomas’
(1885) assessment was that the number of teeth
varied from 34 to 40, dependent on the presence or
absence of a small canine and two premolars in the
lower jaw. Thomas (1888) attempted to improve his
earlier dental formula by changing the position of
the lower canine to the incisor position (Table 2),
and remarked that the “presence or absence of the
minute teeth is not of any systematic importance”.
Thomas (1888) provided illustrations of the upper
and lower jaw of P. v. volans and P. v. minor, although
these are not consistent with his dental formula. Later
reviewers have alluded to a socket in the lower jaw
where a small incisor would be present (i.e. Archer
1984; Triggs 1996). Twenty-one P. volans specimens
in the AM were recently examined by us, and four
(19%) were noted to have minute teeth between the
incisors and pre-molars.

HABITAT AND DIET

Some information on the habitat and diet of
P. volans is available in the early literature. Gould
(1863) stated that the species seeks “blossoms of the
Eucalypti...together with the tender buds and shoots
of the same trees”. Similarly, Le Souef and Burrell
(1926) stated that the “food consists of the leaves and
buds of eucalyptus-trees”. They also added that:

‘careful examination of the contents of
several stomachs of animals taken from
the forests has not revealed anything
else, but in the Myall Lakes district
[NSW]... we have observed this
species on the casuarina-trees; in one
such case the contents of the stomach,
although much masticated, seemed to
be the casuarina-leaves. Mr. Ralph C.
Blackett, forest ranger at Queanbeyan
[NSW]..., states that they chiefly
feed on E. regnans, and to a lesser
extent on E. viminalis, E. fastigata, E.
australasiana, and other narrow-leaved
peppermints.’

In captivity, P- volans has been observed to eat
E. sieberiana readily, “being especially fond of the

flowers, and preferring the bark of the branches to the
leaves” (Le Souef and Burrell 1926). Fleay (1933)
stated “one of the chief difficulties in captivity is
the maintenance of an abundant supply of the tender
leaves of acceptable species of eucalypts” and reported
on collecting trips to obtain sufficient amounts of
leaf from E. elaeophora and E. australiana. He also
reported that “captive specimens could be persuaded
to acquire an additional taste for bread and milk spread
with a sweet jam, but only as an adjunct to the diet
of eucalypt leaves.” Grzimek (1967) stated “because
[P. volans] are exclusive in their diet, like koalas, no
specimen has ever reached a European zoo alive.”
Menkhorst and Knight (2004) stated that it “eats
only eucalypt leaves and buds.” However, Maloney
and Harris (2006) report feeding observations from
several non-eucalypts.

In terms of habitat, Gould (1863) wrote that it “is
strictly an inhabitant of the extensive brushes which
stretch along the south-eastern and eastern portions of
New South Wales”. It has also been reported to occur
in Eucalyptus forests (Le Souef and Burrell 1926,
Anon 1946). Fleay (1933) stated that the species was
found “favouring the taller timber areas and generally
inhabiting dead trees in the gullies of mountainous
country”. Marlow (1958) reported that P volans was
more abundant in dry than wet sclerophyll forests
and less common in open woodland. Ride (1970)
stated that “the habitat is sclerophyll forest and tall
woodland”.

DISTRIBUTION AND ABUNDANCE

The earliest statements on the distribution of the
subject species was that it inhabits NS W (Phillip 1789;
Kerr 1792) or “New Holland” (=Australia) (Meyer
1793; Shaw 1794; Cuvier 1798; Bechstein 1800). The
earliest specific localities mentioned were for places
in NSW, i.e. Botany Bay, Port Jackson, Sydney, Blue
Mountains, Port Macquarie, Bathurst, Maitland,
Clarence River and Goulburn Plains (Cuvier 1826,
1827; Lesson 1830; Bennett 1837; Waterhouse 1841;
Gray 1841; Krefft 1864). Other early distributional
records for NSW include Sutherland (1908, AM
M2003), Helensburgh (1909, AM M2051), Bowral
(1918, AM M2724), Myall Lakes (1922, AM
M33762), Gerringong and Milton (Troughton 1935,
1941), Geehi Gorge (Mt Kosciuszko area) (Anon
1946), Armidale and Tidbinbilla Nature Reserve
(1974), (see Maloney and Harris 2006).

Early literature records from Queensland are
north of the Herbert River (de Vis 1886), Herbert
Vale, Coomooboolaroo, Calliungal (Collett 1887),

Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

Bellenden-ker Range (Ramsay 1890), Eidsvold, Gin-
Gin (Thomas 1923), Atherton Tablelands, Evelyn
Station, Dimboola and Mount Spurgeon Stations
(Tate 1945).

In Victoria, they have been reported from
Templestowe around 1865; east and north-east
of Melbourne (Lucas 1890), and also from the
south, south-west and questionably north-west
areas of the State (Forbes-Leith and Lucas 1884;
Lucas 1897). Other Victorian distributional records
include Allambee East, Newham, Bullengarook
(1905), Dandenong (1923), Mitta Mitta (1931),
Upper Beaconsfield, Traralgon, Daylesford, Bendoc
(1933); Buchan (1960), Matlock (1961), Healesville,
Yellingbo, Powelltown (1963), Woori Yallock,
Darlimurla (1966), Upper Thompson Valley (1968),
Marysville (1969), Porepunkah, Mount Buffalo and
Upper Lerderderg Valley (1970) (see Maloney and
Harris 2006).

Gould (1863) believed that its range was from
“Port Phillip to Moreton Bay”. Krefft’s (1864)
assessment was that it occurred in the “mountainous
coast districts of the Australian continent”, from
Victoria to Qld; also that it was “not found upon the
plains of the interior”. Thomas (1888) and Lydekker
(1896) reported that its range was from Qld to Victoria.
Fleay (1933) stated that the range “extends down the
highlands of eastern Australia from southern Qld. to
Victoria’, and that he had “never observed the species
further west than the Ballarat-Daylesford forest” in
Victoria. Marlow (1958) found that the western limits
of its distribution in NSW were Barraba, Orange
and Tumut. Ride (1970) reported the distribution
to be from the Dandenong Ranges (Victoria) to
Rockhampton, Qld.

In terms of abundance, the species has been
described as the “most abundant of the arboreal
marsupials in the forests to the east and north-east
of Melbourne” (Lucas 1890), “very plentiful in the
heavy eucalypt forests” of eastern Australia (Le Souef
and Burrell 1926); and “among the most numerous
of arboreal marsupials” in East Gippsland (Fleay
1933). Marlow (1958) reported that P volans was
“abundant” in NSW (see also Calaby 1966; Flannery
1994; McKay 1995). Currently, P. volans is not listed
as threatened in the three states that it occurs, and
recent distribution maps are provided by Eyre (2004)
and Winter et al (2004) for Qld, Kavanagh (2004) for
NSW and van der Ree (2004) for Victoria.

BEHAVIOUR

The gliding ability of P volans was first reported

Proc. Linn. Soc. N.S.W., 129, 2008

by Phillip (1789) and then by Shaw (1794, 1800),
Cuvier (1798) and Turton (1806). Later authors
remarked that it moves with a gliding motion, but
this was not true flying (Desmarest 1817; Lesson
1827; Owen 1841, 1845; Lydekker 1896). Le Souef
and Burrell (1926) record a “flight by one of these
animals from the top of one tall eucalypt to the base
of another was 80 yards [=73 m]; another flight, of
55 feet [=17 m], occupied 1 % seconds.” Troughton
(1935, 1941) stated that it is “the record glider of the
possum world” and reported that one individual was
observed at Milton NSW, covering a distance of 590
yards [=540 m] in six successive glides. Two of these
glides were 120 yards [=110 m], and one of 70 yards
[=64 m] from a tree 100 feet [=30 m] high. Wakefield
(1970) stated “that some long glides, attributed in
the literature to P. volans, belong in fact to Petaurus
australis”. He discussed the report by Troughton
(1935, 1941) and stated:

‘The 70 yard [=64 mJ] glide from a
100-foot [=30 m] tree indicates an
angle of descent of 26 degrees to the
horizontal, and, even allowing for
sloping ground and a margin of error
in the measurements, this performance,
though well within the capabilities
of Petaurus, is quite outside that of
Petauroides. Also, for the 120-yard
[=110 m] glides P. volans would require
for its 40 degree descent, a take-off
point approximately 300 feet [=90 m]
high, while Petaurus would need a 200-
foot [=60 m] tree. Other features of the
Milton resident’s report — that during
the performance the animal “lost no
time in ascending three more trees” and
that “it uttered its peculiar squealing
call” — leave no doubt that the “record
glider” was, in fact, Petaurus australis
and not Petauroides volans.’

The voice and gliding accomplishments of
Petaurus australis have been credited erroneously to
P. volans, which is, in fact, a sedentary, slow-moving,
silent animal of minor gliding ability (Wakefield
1970; McKay 1989). Many authors have mistakenly
accredited P volans with the vocalisations of P.
australis: for instance Lydekker (1896) was the first
to erroneously report “when disturbed, or in flight,
they utter a loud piercing scream, audible for a
long distance” (see also Le Souef and Burrell 1926;
Troughton 1935, 1941; Fleay 1933, 1947; Calaby
1966 for similar reports).

49

EARLY HISTORY OF THE GREATER GLIDER

It was also recognised quite early that this species
was nocturnally active and utilised tree hollows as den
sites during the day (Oken 1816; Desmarest 1817;
Lesson 1827; Partington 1837; Waterhouse 1846;
Thomas 1885; Collett 1887; Aflalo 1896). Gould
(1863) stated that “on the approach of evening [it]
emerges from its retreat.”” Lydekker (1896) reported
that they “spend the day in some hollow branch or the
stem itself, whence they issue forth for their nocturnal
flight’.

Le Souef and Burrell (1926) suggested that the
only predators of P. volans are the powerful owl
Ninox strenua and the introduced fox Vulpes vulpes;
‘the latter occasionally catches them on the ground’
(see also Fleay 1933, 1947, 1968). However, Maloney
and Harris (2006) reported P. volans falling prey to a
range of other predators such as the cat Felis catus,
dog Canis familiaris, fox V. vulpes, wedge-tailed
eagle Aquila audax, quoll Dasyurus maculatus and
sooty owl Tyto tenebricosa. Other recorded predators
of the greater glider include the dingo C. f dingo
(Robertshaw and Harden 1985), lace monitor Varanus
varius (Weavers 1989) and carpet python Morelia
spilota (Lindenmayer 2002).

Fleay (1933) reported:

“Wandering under the trees on a still
night, when the dusky gliders [P
volans| are feeding overhead, rarely
leads to their discovery without resort
to intent listening. Perhaps the faint
sound of a leaf being pulled from a
stalk, or a sudden rustle as the animal
plunges its weight from one slender
limb to another, betrays its position
to a searching torch beam held so that
the observer’s eyes look straight along
the path of light. Then the blazing
orbs of the animal, certainly the most
brilliant light reflectors that I know of
among the marsupial family, regard the
intruder with some curiosity’.

In terms of its reproduction and breeding
behaviour, Desmarest (1817) reported “females have
a pouch under the belly, where the young spend the
first part of their existence”. Fleay (1933) made the
following observations on captive specimens: “only
two mammae are found in the pouch” and “only
one embryo is reared at a time.” He also reported as
follows:

‘In Vic. this minute naked creature
seems to appear usually in July or

50

August, and it is difficult to realize
that such a mite, no larger than the
head of a drawing-pin, may indulge
some day in graceful aerial “flights”.
Gradually as the youngster increases
in bulk, it is noted that the limbs and
tail are extraordinarily long, the loose
volplaning membrane from fore limb
to hind limbs is plainly visible, and
the colour of the furless embryo is
pink with very dark ears. The little
fellow becomes free of its inseparable
attachment to the mamma when some
six weeks of age. Later the eyes open
and a covering of short fur indicates
plainly the contrast between the black
and white of the upper and lower
surfaces respectively. It then spends the
daylight hours out of the pouch, and by
night is carried around as a large bulge
in it. At four months it has become too
bulky to. be contained in the pouch any
longer. Between the growing of fur and
the forsaking of the mother’s “pocket
nursery” the young Taguan Phalanger
[P. volans] is one of the most curious
and pathetic babes that one can imagine
with its lanky legs, very long tail and
thin weedy body. Having outgrown the
pouch, though still being nourished
from it, the little phalanger clings to
its mother’s back during her nocturnal
wanderings, though perhaps the gliding
leaps are out of the question unless the
youngster remains in the home tree or
sleeping hollow’.

CONCLUSION

Petauroides volans has had a long and sometimes
confusing taxonomic history. It has been listed as a
member of 10 genera (Belideus, Didelphis, Petaurista,
Petaurides, Petauroides, Petaurus, Phalanger,
Phalangista, Schinobates, and Voluccella) and there
have been at least 23 different binomial names
used for it since its discovery. This geographically
widespread species was sent to different museums
throughout Europe by collectors, and given different
designations by 19" Century zoologists. These early
zoologists were often rivals, each of whom was more
anxious to discover and name species, than to find out
the habits of the species already known (Partington
1937). Consequently errors were made, and some of

Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

these have persisted into the modern literature. For
example, Flannery (1994) mistakenly lists Hepoona
Roo (© Petaurus australis) as synonymous with P
volans.

Early descriptions of the morphology of P
volans such as its colouration, size and the presence
of a gliding membrane, are given in broad general
terms but nevertheless they do have value from a
historical viewpoint. Dental descriptions in the early
literature vary, and some confusion with the similarly
structured dentition of Pseudocheirus peregrinus
is evident. Early behavioural observations include
the ability to glide, and that it is nocturnally active
using tree hollows as den sites. The earliest records
of occurrence were centred about the Sydney district.
As the colony expanded so did its recorded range.
Gould (1863) reported that its distribution was from
Port Phillip (Victoria) to Moreton Bay (Qld), and
this is reasonably accurate when compared to our
understanding of its current range.

ACKNOWLEDGEMENTS

We wish to thank the library staff at the AM, Sydney;
Natural History Museum, London; and the University of
Wollongong. For translating French articles, we are grateful
to Betty Hassen, and for translating German we acknowledge
Peter Simmons. We also thank Sandy Ingleby for access to
the mammal collection in the AM, and Cécile Callou for
correspondence and a photograph regarding a specimen
from the MNHN. Finally, we owe a debt of gratitude to
Ross Goldingay and Rob Whelan for helpful discussion and
comments on earlier versions of the manuscript, and also to
two anonymous referees for their advice.

REFERENCES

Aflalo, F.G. (1896). A sketch of the natural history of
Australia with some notes on sport. (MacMillan and
Co: London). pp. 48-58.

Anon. (1890). Notes and exhibits. Proceedings of the
Linnean Society of New South Wales 4 (2), 1030.

Anon. (1946). Report to the trustees of Kosciusko State
Park by the joint scientific committee of the Linnean
Society of N.S.W and the Royal Zoological Society
of N.S.W on a reconnaissance natural history survey
of the park. p. 51.

Archer, M. (1984). The Australian marsupial radiation.
In “Vertebrate zoogeography and evolution in
Australasia. (Animals in space and time)’ (Eds. M.
Archer and G. Clayton) pp. 633-808. (Hesperian
Press: Carlisle, Western Australia).

Bechstein, J.M. (1800). Thomas Pennant’s allgemeine
Uebersicht der vierfiissigenn Thiere. Aus dem

Proc. Linn. Soc. N.S.W., 129, 2008

englischen tibersetzt und mit Anmerkungen and
Zusatzen versehen von J.M. Bechstein. Weimer:
Industrie-Comptoir’s 2, 323-768. [351-353, 686]

Bennett, G. (1837). A catalogue of the specimens of
natural history and miscellaneous curiosities
deposited in the Australian Museum. (James Tegg
and Co: Sydney). pp. 3-4.

Brehms, A.E. (1880). Thierleben, allgemeine kunde des
thierreichs. Grosse Ausg. 2, 573.

Calaby, J.H. (1966). Mammals of the Upper Richmond
and Clarence Rivers, New South Wales. Division of
Wildlife Research Technical Paper No. 10, CSIRO,
Australia.

Collett, R. (1887). On a collection of mammals from
central and northern Queensland. Zoologische
Jahrbuecher 2, 829-940. [926]

Cuvier, F. (1826). Dictionnaire des sciences naturelles,
dans lequel on traite methodiquement des differens
étres de la nature 39, 418-419.

Cuvier, G. (1798). Tableau elementaire de l’histoire
naturelle des animaux. (Baudouin: Paris). p. 126.

Cuvier, G. (1817). Phalangers. In “Regne animal distribué
d’aprés son organisation’. (Deterville: Paris). 1, 178-
180.

Cuvier, G.L.C.F.D. (1825). Petaurus. In “Des dent
des Mammiferes considérés comme characteres
zoologiques’. (F.-G. Levrault: Paris). pp. 128-130,
253.

Cuvier, G. (1827). Synopsis of the species of the class
Mammalia as arranged with reference to their
organization by Cuvier, and other naturalists, with
specific characters, synonyma etc etc. In ‘The
animal kingdom, arranged in conformity with its
organization, by Baron Cuvier, with additional
descriptions of all the species hitherto named, and
of many not before noticed’, by Griffith, E. (G.B.
Whittaker: London). 5, 198-205.

Cuvier, G. (1829). Phalangers. In “Régne animal distribué
d’aprés son organisation’. 2" edition. (Deterville:
Paris). 1, 181-184.

de Beaufort, F. (1966). Catalogue des types des
mammiferes du Muséum National d’ Histoire
Naturelle, Paris. VI Monotremata. VII Marsupialia.
Bulletin du Museum National d’Histoire Naturelle
38, 509-553.

Desmarest, A.G. (1803). Nouveau dictionairie d’histoire
naturelle, appliquée aux arts, a l’agriculture, a
l’économie, rurale et domestique. (Deterville: Paris).
17, 381.

Desmarest, A.G. (1817). Nouveau dictionairie d’histoire
naturelle, appliquée aux arts, a l’agriculture, a
l’économie, rurale et domestique, a la medicine etc.
(Deterville: Paris). 25, 400-404.

Desmarest, A.G. (1820). Mammalogie ou description des
espéces de mammiferes. Encyclopédie méthodique
histoire naturelle. (Mme Veuve Agasse: Paris). 1,
268-270. (suppl. 8. fig. 4).

de Vis, C.W. (1887). On new or rare vertebrates from the
Herbert River, north Queensland. Proceedings of the

51

EARLY HISTORY OF THE GREATER GLIDER

Linnean Society of New South Wales 1 (2), 1129-
1137.

Evenhuis, N.L. (1991). World catalog of genus group
names of bee flies (Diptera: Bombylidae). Bishop
Museum Bulletins in Entomology 5, 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: Chipping Norton).

Fabricius, J.C. (1794). Entomologia systematica emendata
et aucta. Secundum classes, ordines, genera,
species adjectis synonimis, locis, observationibus,
descriptionibus. (C.G. Proft: Hafniae: Copenhagen).
4, 412-413.

Fischer, V.W.G. (1814). Zoognosia tabulis synopticus
illustrata: In usum praelectionum academiae
medico-chirugicae mosquensis edita. Editio tertia
(Vsevolozsky: Moscow). 3, 498-501.

Fischer, J.B. (1829). Synopsis mammalium. (Sumtibus
J.G. Cottae: Stuttgardtiae). pp. 278-279.

Flannery, T.F. (1994). Pseudocheiridae. In ‘Possums of
the World: a monograph of the Phalangeroidea’.
(GEO Publications in association with the Australian
Museum: Sydney). pp. 102-151.

Fleay, D. (1933). The greater flying phalanger . The
Victorian Naturalist 50, 135-142.

Fleay, D. (1947). Gliders of the gum trees. (Bread and
cheese club: Melbourne).

Fleay, D. (1968). Nightwatchmen of bush and plain.
(Jacaranda Press: Melbourne).

Flower, W.H. (1884). Catalogue of the specimens
illustrating the osteology and dentition of vertebrated
animals, recent and extinct, contained in the museum
of the Royal College of Surgeons of England.
(Printed for the College: London). 2, 705.

Forbes-Leith, T.A. and Lucas, A.H. (1884). Catalogue of
the fauna of Victoria. Vertebrates: Mammalia. The
Victorian Naturalist 1, 4-6.

Gerrard, E. (1862). Catalogue of the bones of Mammalia
in the collection of the British Museum, London.
(Printed by order of the Trustees: London) p. 120.

Gervais, P. (1855). Tribu des Phalangistins. In “Histoire
naturelle des mammifeéres avec |’indication de
leurs moeurs, et de leurs rapports avec les arts, le
commerce et |’agriculture’. (L. Curmer : Paris). 2,
276.

Giebel, C. (1853). Odontographie; vergleichende
darstellung des zahnsystems der lebenden und
fossilen wirbelthiere. (A. Abel: Leipzig). p. 42 (plate
18, figure 3).

Giebel, C. (1855). Odontographie; vergleichende
darstellung des zahnsystems der lebenden und
fossilen wirbelthiere. (A. Abel: Leipzig). p. 42 (plate
18, figure 3).

Giebel, C. (1859). Die Saugethiere in zoologischer,
anatomischer und palzontologischer Beziehung
umfassend dargestellt. (A. Abel: Leipzig). p. 701.

Gloger, C.W.L. (1842). Gemeinniitziges hand-und
hilfsbuch der naturgeschichte: fiir gebildete leser

CA
NR

aller staénde, besonders fiir die reifere jugend und ihre
lehrer. (Verlag von Aug. Schulz: Breslau). p. 85.

Gould, J. (1863). The mammals of Australia: incorporating
the 3 original volumes with modern notes by Joan M.
Dixon. (Macmillan 1977: South Melbourne).

Gray, J.E. (1841). Contributions towards the geographical
description of the Mammalia of Australia, with notes
on some recently discovered species. Appendix
inGrey, G. ‘Journal of two expeditions of discovery
in northwest and Western Australia during the years
1837, 38 and 39’. (1. and W. Boone: London). pp.
397-414.

Gray, J.E. (1842). Catalogue of Australian Mammalia.
Tasmanian Journal of Natural Science, Agriculture,
Statistics, &c. 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). pp. 83-84.

Grzimek, B. (1967). Marsupials learnt how to fly three
times. In ‘Four-legged Australians, adventures with
animals and men in Australia’. (Collins: Sydney). pp.
68-75.

Harris, J.M. (2006). The discovery and early natural
history of the eastern pygmy-possum, Cercartetus
nanus (Geoffroy and Desmarest, 1817). Proceedings
of the Linnean Society of New South Wales 127, 107-
124.

Haswell, W.A. (1886). Jottings from the biological
laboratory of Sydney University. Proceedings of the
Linnean Society of New South Wales 1 (2), 176-182.

Iredale, T. and Troughton, E.L.G. (1934). A check-list of
the mammals recorded from Australia. Australian
Museum Memoir 6, 1-122. [28-30].

Jentink, F.A. (1886). Muséum d’histoire naturelle des
pays-bas. Catalogue ostéologique des Mammiferes.
(E.J. Brill: Leiden). 9, 316.

Julien-Laferriere, D. (1994). Catalogue des types de
Mammiféres du Muséum National d’ Histoire
Naturelle. Order des Marsupiaux. Extrait de
Mammalia. 58, 19-20.

Kavanagh, R.P. (2004). Distribution and conservation
status of possums and gliders in New South Wales. In
“The biology of Australian possums and gliders’ (Eds
R. L. Goldingay and S. M. Jackson) pp. 130-148.
(Surrey Beatty and Sons: Chipping Norton).

Kavanagh, R.P. and Wheeler, R.J. (2004). Home-range of
the greater glider Petauroides volans in tall montane
forest of southeastern New South Wales, and changes
following logging. In ‘The biology of Australian
possums and gliders’ (Eds R. L. Goldingay and S.

M. Jackson) pp. 413-425. (Surrey Beatty and Sons:
Chipping Norton).

Kerr, R. (1792). The animal kingdom, or zoological
system, of the celebrated Sir Charles Linnaeus: Class
1. Mammalia: containing a complete systematic
description, arrangement, and nomenclature, of all
the known species and varieties of the Mammalia,
or animals which give suck to their young: being
a translation of that part of the Systema Nature,
as lately published, with great improvements, by

Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

Professor Gmelin of Goettingen. Together with
numerous additions from more recent zoological
writers, and illustrated with copperplates. (J. Murray
and R. Faulder: London). 1, 199.

Krefft, G. (1864). Catalogue of the Mammalia in the
collection of the Australian Museum. (Government
Printer: Sydney). pp. 38-39.

Krefft, G. (1871). The mammals of Australia, with a
short account of all the species hitherto described.
(Government Printer: Sydney).

Lacépéde, R. P. (1801). Memoires de |’institut national
des sciences et arts (sciences mathématiques et
physiques). Tableau des divisions, sous-divisions,
ordres et genres des mammiferes. Deuxieme sous-
division. Les pieds de derriere en forme de mains.
Pedimanes. Deuxieme ordre. p. 491.

Le Souef, A.S. and Burrell, H. (1926). Greater Flying
Phalangers. In ‘The wild animals of Australasia’.
(George Harrap and Company: Sydney). pp. 259-261.

Lesson, R.P. (1827). Les Petauristes in “Manuel de
Mammalogie, ou histoire naturelle des mammiferes’.
(Roret: Paris). pp. 223-224.

Lesson, R.P. (1828). Petaurus. In ‘Dictionaire classique
d’histoire naturelle par Messieurs Audouin, Isid.
Bourdon, Ad. Brongniart, de Candolle... et Bory de
Saint —Vincent’. 13, 286-289.

Lesson, R.P. (1830). Histoire naturelle générale et
particuliére des mammiferes et des oiseaux
découverts depuis 1788 jusqu’a nos jours. Suite des
mammiféres. (Baudouin Freres: Paris). 4, 440-442.

Lesson. R.P. (1838). Les Petauristes in “Complément de
Buffon races humaine et mammiféres’, 2 ™ Edition,
(P. Pourrat Freres: Paris). pp. 450-453.

Lesson, R.P. (1842). Famille Petaurusideae. In ‘Nouveau
tableau du régne animal. Mammiferes’. (A. Bertrand:
Paris). pp. 189-190.

Lindenmayer D.B. (2002). Gliders of Australia: a natural
history. (UNSW Press: Kensington).

Link, H.F. (1795). Beytrage zur naturgeschichte. Bd. 1,
(2). p. 52.

Lucas, A.H.S. (1890). Zoology. Vertebrata. In “Handbook
of Melbourne’. (Ed. W.B. Spencer) (Spectator
Publishing Co: Melbourne). p. 62.

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.

Lydekker, R. (1896). The Taguan Phalangers, Genus
Petauroides. In “Lloyd’s Natural History, A hand-
book of the Marsupialia and Monotremata’. (Wyman
and Sons: London). pp. 100-102 (plate xiv).

Maloney, K.S. and Harris, J.M. (2006). Annotated records
of the greater glider Petauroides volans from The
Victorian Naturalist. The Victorian Naturalist 123(6),
230-236.

Marlow, B.J. (1958). A survey of the marsupials of New
South Wales. CSIRO Wildlife Research 3, 71-114.

Marlow, B.J. (1962). Greater glider Schoinobates volans.
In “Marsupials of Australia’. (Jacaranda Press:
Brisbane). pp. 92-93.

Proc. Linn. Soc. N.S.W., 129, 2008

McKay, G.M. (1982). Nomenclature of the gliding possum
genera Pefaurus and Petauroides (Marsupialia:
Petauridae). Australian Mammalogy 5, 37-39.

McKay, G.M. (1988). Petauridae. 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. 87-97. (Australian
Government Publishing Service: Canberra).

McKay, G.M. (1989). Petauridae. In ‘Fauna of Australia.
Vol. 1B Mammalia’. (Eds. D.W. Walton and B.J.
Richardson) pp. 665-679. (Australian Government
Publishing Service: Canberra).

McKay, G.M. (1995). Greater Glider Petauroides
volans. In “The mammals of Australia. The national
photographic index of Australian wildlife’. (Ed. by R.
Strahan) pp. 240-241. (Reed New Holland: Sydney).
Menkhorst, P.W and Knight, F. (2004). Greater glider
Petauroides volans. \n ‘A field guide to the mammals
of Australia’. pp. 98-99. (Oxford University Press:
Melbourne).

Meyer, F.A.A. (1793). Systematisch-summarische
Uebersicht der neuesten zoologischen Entdeckungen
in Neuholland and Afrika. Nebst zwey andern
zoologischen Abhandlungen. (Dykischen
Buchhandlung: Leipzig) 184 pp. [26]

Mitchell, T.L. (1838). Three expeditions into the interior
of eastern Australia, with descriptions of the recently
explored regions of Australia felix, and of the present
colony of New South Wales. (T&W Boone: London).
p. XVil.

Ogilby, D. J. (1892). Petauroides. In ‘Catalogue of
Australian Mammals with introductory notes on
general mammalogy’. Australian Museum Catalogue
16, 31-32.

Oken, L. (1816). Lehrbuch der Naturgeschichte. Dritter
Theil Zoologie. (A. Schmid und Comp: Jena). 3,
1117-1120. ‘

Owen, R. (1841). Marsupialia (from the cyclopedia of
anatomy and physiology). (Martin, Singer and Smith,
London). pp. 7-9.

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. 1 (text), 384-387.

Palmer, T.S. (1904). Index generum mammalium. North
American fauna, U.S Department of Agriculture.
(Government Printing Office: Washington). 23, 1-
987. [624]

Partington, C.F. (1837). The British cyclopzedia of natural
history: combining a scientific classification of
animals, plants and minerals; with a popular view of
their habits, economy, and structure. (Orr and Smith:
London). 3, 424-425.

Phillip, A. (1789). The voyage of Governor Phillip to
Botany Bay, with an account of the establishment
of the colonies of Port Jackson & Norfolk Island,
compiled from authentic papers, which have been
obtained from the several departments, to which
are added the journals of Lieuts. Shortland, Watts,

3/3)

EARLY HISTORY OF THE GREATER GLIDER

Ball, & Capt. Marshall; with an account of their new
discoveries, embellished with LV copper plates, the
maps and charts taken from actual surveys, & the
plans and views drawn on the spot, by Capt. Hunter,
Lieuts. Shortland, Watts, Dawes, Bradley, Capt.
Marshall, &c. (1st edition). (J. Stockdale: London ).
pp. 296-298.

Phillip, A. (1790). The voyage of Governor Phillip to
Botany Bay, with an account of the establishment
of the colonies of Port Jackson & Norfolk Island,
compiled from authentic papers, which have been
obtained from the several departments, to which
are added the journals of Lieuts. Shortland, Watts,
Ball, & Capt. Marshall; with an account of their new
discoveries, embellished with LV copper plates, the
maps and charts taken from actual surveys, & the
plans and views drawn on the spot, by Capt. Hunter,
Lieuts. Shortland, Watts, Dawes, Bradley, Capt.
Marshall, &c. (2nd edition). (J. Stockdale: London).
pp. 135-137.

Ramsay, E.P. (1890). On a new species of Petaurides from
the Bellenden-Ker Range, N.E. Queensland. Records
of the Australian Museum 1, 77-78.

Ride, W.D.L. (1970). Petauroides volans In ‘A guide to
the native mammals of Australia’. (Oxford University
Press: Melbourne).

Robertshaw, J.D. and Harden, R.H. (1985). The ecology of
the dingo in north-eastern New South Wales II. Diet.
Australian Wildlife Research 12, 39-50.

Schinz, H.R. (1821). Phalangista macroura. In “Das
thierreich eingetheilt nach dem bau der thiere
als grundlage ihrer naturgeschichte und der
vergleichenden anatomie, von dem herrn ritter von
Cuvier’. 1, 259-261.

Schinz, H.R. (1844). Systematisches verzeichniss aller
bis jetzt bekannten saéugethiere; oder, synopsis
mammalium nach dem Cuvier’schen system. (Jent
und Gassmann, Solothurn). 1, 530-533.

Shaw, G. (1791). Petaurus. In ‘The Naturalist’s
Miscellany: or coloured figures of natural objects;
drawn and described immediately from nature’.
(Fredericus Nodder: London). 2, [text to pl. 60].

Shaw, G. (1794). Zoology of New Holland (the figures by
James Sowerby). (Published by J. Sowerby: London).
1, 33 (also plate 12).

Shaw, G. (1800). Petaurine opossum and long-tailed
opossum. In “General zoology, or, Systematic natural
history’. (Fredericus Nodder: London). 1 (2), 496-
501 (ancluding plates 112, 113).

Sherborn, C.D. (1902). Index animalium, sive index
nominum quae ab A.D. 1763 generibus et speciebus
animalium imposta sunt, societatibus eruditorum
adiuvantibus. (Cambridge University Press: London).
[738].

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 eae ya piohepraphiedl notes. (Angus and

54

Tate, G.H.H. (1945). Notes on the squirrel-like and mouse-
like possums (Marsupialia). American Museum
Novitates 1305, 1-12.

Temminck, C. (1838). Fauna Japonica. I. Mammalia.

Paris. p. 46 (Plate 13).

Thomas, O. (1879). On Robert Kerr’s translation of the
“Systema Naturae’ of Linnaeus. Annals and Magazine
of Natural History 4 (5), 396-397.

Thomas, O. (1885). Phalanger. In “The Encyclopzedia
Britannica: a dictionary of arts, sciences, and general
literature’. (Adam and Charles Black: Edinburgh).
(9th Edition) 18, 727-729.

Thomas, O. (1888). Petauroides. In ‘Catalogue of the
Marsupialia and Monotremata in the collections of
the British Museum’ (Natural History), London.
pp. 163-166. (Plate 17 figures 2 and 3, and Plate 18
figures 1, 2 and 3).

Thomas, O. (1923). On some Queensland Phalangeridae.
Annals and Magazine of Natural History 11 (9), 246-
250.

Tiedemann, F. (1808). Zoologie. Zu seinen vorlesungen
entworfen. Allgemeine Zoologie, mensch
und saugthiere. (Landshut, in der Weberschen
Buchhandlung: Heidelberg). 1, 432-433.

Triggs, B. (1996). Tracks, scats and other traces: a field
guide to Australian mammals. (Oxford University
Press). pp. 304-305.

Troughton, E. (1935). The largest gliders or “flying
possums”. Australian Museum Magazine 5 (9), 314-
319.

Troughton, E. (1941). Greater glider-possum. In ‘“Furred
animals of Australia’. (Angus and Robertson:
Sydney). pp. 101-105.

Turton, W. (1806). A general system of nature, through
the three grand kingdoms of animals, vegetables, and
minerals, systematically divided into their several
classes, orders, genera, species and varieties with
their habitations, manners, economy, structure and
peculiarities by Sir Charles Linne; translated from
Gmelin, Fabricius, Willdenow, &c.; with a life of
Linne and a dictionary of the terms of natural history.
1, 68-69.

Van Der Ree, R. Ward, S.J. and Handasyde, K.A. (2004).
Distribution and conservation status of possums and
gliders in Victoria. In ‘The biology of Australian
possums and gliders’ (Eds R. L. Goldingay and S.

M. Jackson) pp. 91-110. (Surrey Beatty and Sons:
Chipping Norton).

Wagner, J.A. (1843). ‘Die Saugethiere, in Abbildungen
nach der Natur, mit Beschreibungen. Fortgesetzt
von A. Goldfuss. (Ed. J.C.D. von Schreber).
Supplementband 3’. (Erlangen: Voss). pp. 84-88.

Wagner, J.A. (1855). ‘Die Saugethiere, in Addildungen
nach der Natur, mit Beschreibungen. Fortgesetzt.

(Ed. J.C.D. von Schreber). Supplementband von J.A.
Wagner’. (Erlangen: Voss) Suppl. 5, 278-279.

Wakefield, N.A. (1970). Notes on the glider-possum,
Petaurus australis (Phalangeridae, Marsupialia). The
Victorian Naturalist 87, 221-236.

Waterhouse, G.R. (1838a). Taguan Flying Opossum.

In ‘Catalogue of the Mammalia preserved in the

Proc. Linn. Soc. N.S.W., 129, 2008

K.S. MALONEY AND J.M. HARRIS

museum of the Zoological Society of London’.
(Richard and John E. Taylor: London). (2nd edition)
p. 68.

Waterhouse, G.R. (1838b). Observations on certain
modifications observed in the dentition of the flying
opossums. Proceedings of the Zoological Society 6,
149-156.

Waterhouse, G.R. (1841). Genus Petaurus. In ‘The
Naturalist’s Library. Mammalia’. (Ed. W. Jardine).
(W.H. Lizars & H.G. Bohn: Edinburgh & London).
11, 282-289. (including plate 27).

Waterhouse, G.R. (1843). Genus Petaurus. In ‘The
Naturalist’s Library. Mammalia’. (Ed. W. Jardine).
(W.H. Lizars & H.G. Bohn: Edinburgh & London). 8,
282-289. (including plate 27).

Waterhouse, G.R. (1846). Genus Petaurus. In ‘A natural
history of the Mammalia containing the order
Marsupiata, or pouched animals’. (Hippolyte
Baillere: London) 1, p.318-341 (plate 19, fig. 4).

Weavers, B.W. (1989). Diet of the lace monitor (Varanus
varius) in south-eastern Australia. Australian Zoologist
25 (3), 83-85.

Winter, J.W., Dillewaard, H.A., Williams, S.E. and
Bolitho, E.E. (2004). Possums and gliders of north
Queensland: distribution and status. In ‘The Biology
of Australian Possums and Gliders’ (Eds. R. L.
Goldingay and S. M. Jackson) pp. 26-50. (Surrey
Beatty and Sons: Chipping Norton).

White, J. (1790). Journal of a voyage to New South Wales,
with sixty five plates of non descript animals, birds,
lizards, serpents, curious cones of trees, and other
natural productions. (J. Debrett: London) p. 299
(plate 60).

Proc. Linn. Soc. N.S.W., 129, 2008

35

56

oo 1 Suing cfALIRce <cibw mip, 4h linge
i

eae ae
ab ts) teen am i

~

i" thi tie aa hah
a ater pci iye ae ie ef
‘yA NOI TEAMS AA “en

wiles: Nenwae it Livres Aas te

- Roetthak ANGE TRattoemete 3 3 ;
(lem bs yer pile “an m
ee (hee bert get Ris I
SES ote path eine tan (C8)
. amb: VDE 0S therein fine onbe :
BA dolbserad oly apa tebe: <A acts
“ad byte” MieeconntlS ata! st)
monn ithiveithury'S Atmel Dit YS 2
inne ity Seb Ta abe milan ait W
why agg aK ecm: hackansqne Baul
(hai) RS akin ies et Beet (hho #
vanett®) sof sost alt his Posh aera W ae
spades iie-Aie : .
niwirton, Aikewteines Pediat: shee LES fj -
ia Rm lle pecead A dal, iteeesialsit AL Sins
than table, othe ainileey 4) sQ0l> 73 a. ultile
vgn of lof ae
ude cba wrebrtieinn senior
sscingiten eds Zhan
they, © Mei ieee ie
seceded hoa netpatrtagesh
son beh, herrsdetes ovals serves iy yamine whl dt

‘ TL A pp iow Cee OT heli

“he pete Leora (Ff apis, ogee hieny
aA, Oye 2helly divigued: iia. fice 5

ce w@ite, éfist & Telos wel Vareeehy

y

7

us ‘ ana em Oe Ay tindies | eee, ert
of mired) of rele, (ioe Willen Ae, aoe ne
Luho gull «i eaary ob The tegis al Lathe hen

ak fe , : Ther Poe © lie fe adh hiailite NA fee
(Lewis (ee Oh ef nmeiretitce olgtie gif pee

1 Vicootw th bie tiolepy if AGaiplian
ai MGS Chile dl (ladies
ar new sf “ees Bs. Wi 11. | ope Ties weet Sores
‘ : 1m ¢ Noetem 7 _ ia
teacd) eo alr, he Teaioetnemmn Kortemehyy «
f My Bus } A \ ae [eheber)}” : 54
bbe Ir Hbear ome Tagg ee Oe os.
cine uated, (0509 ‘Lit Qin, In cs &
fuichiher Shate, deal ahaa chine ‘Poy ey i) i Al
A rayen Chal hn FN eeeat Rites q oy, 5 ay lou rewind ary tA oe
Vila bed yi yarns /2 nea, Pee ; Sonny F & vA rr) a
Wakcic! cA Pe Noe a he cle ponent, ;
‘ tr edt yucomaln (aspects whiranydjies
iT, My pari Ne RE S91 aM. «~
mw et Weatcrieryee, (5, CRI Leswciy Peg Cyne

"Uwtahugiw inf the Monenatia weer ise

BONE [ODE M62

er

Ordovician (Early Darriwilian) Conodonts and Sponges from
West of Parkes, Central New South Wales

YonG Yt ZHEN! AND JOHN PICKETT?”

1. Palaeontology Section, The Australian Museum, 6 College Street, Sydney, NSW 2010, Australia (yongyi.
zhen@austmus.gov.au);
2. Geological Survey of New South Wales, NSW Department of Primary Industries, State Geoscience Centre,
947-953 Londonderry Road, Londonderry, NSW 2753; Research Associate, Australian Museum, Sydney.

Zhen, Y.Y. and Pickett, J.W. (2008). Ordovician (Early Darriwilian) conodonts and sponges from west of Parkes,
central New South Wales. Proceedings of the Linnean Society of New South Wales 129, 57-82.

A well preserved conodont fauna and an associated small sponge assemblage recovered from a limestone lens
exposed on Kirkup Station, 15 km west of Parkes, New South Wales are described and illustrated. The conodont
fauna is exceptionally rich by Australian standards, represented by nearly 4,000 specimens, but low in diversity
including only six species: Erraticodon balticus Dzik, 1978, Kirkupodus tricostatus gen. et sp. nov., Protopanderodus
cf. varicostatus (Sweet and Bergstr6m, 1962), Protopanderodus? nogamii (Lee, 1975), Juanognathus serpaglii
Stouge, 1984, and Pseudooneotodus mitratus (Moskalenko, 1973). The species definition of E. balticus is revised
based on the current collection of over 1,700 specimens. Co-occurrence of E. balticus, J. serpaglii and P. cf.
varicostatus suggests an early Darriwilian (Da2) age for the fauna, which is correlated with that from the basal
Weemalla Formation exposed further east near Orange. Two anthaspidellid sponges occur in the assemblage. The
stromatoporoid Janilamina kirkupensis gen. et sp. nov. is the oldest stromatoporoid reported from Australia, and
among the oldest known. A shallow-water, near-shore setting for the fauna is supported by the abundant occurrence

of algal oncolites and certain sedimentary features in the limestone lens.

Manuscript received 30 June 2007, accepted for publication 6 February 2008.

KEYWORDS. a stromatoporoid, Conodonts, Darriwilian, Goonumbla Volcanics, Middle Ordovician, New South

Wales, new taxa, sponges.

INTRODUCTION

The western flank of the Forbes Anticline, west of
Parkes, New South Wales, is made up of a generally
conformable succession of andesitic volcanics and
diverse, mostly shallow-water sediments of Middle
and Late Ordovician age (Fig. 1). An analysis of Late
Ordovician coral and conodont faunas in this area
was provided by Pickett and Percival (2001). Their
assemblages were derived from a series of limestones,
mostly not continuous along strike for any great
distance, occurring in a series of formations called by
them the Goonumbla Volcanics (oldest), the Billabong
Creek Limestone and the Gunningbland Formation
(youngest). In that paper, the oldest conodont
assemblages reported were from a level about 270
m below the top of the Goonumbla Volcanics. The
small faunas were not finely age-diagnostic, the
only identifiable species being Periodon aculeatus
Hadding, Panderodus cf. gracilis (Branson and Mehl)

and Drepanodus arcuatus Pander (sample C874, Fig.
2). Some 100 m higher a cluster of samples yielded
Pygodus anserinus Lamont and Lindstrém, Pygodus
anitae Bergstr6m and Eoplacognathus spp. A sample
from the base of the Billabong Creek Limestone
(C828) yielded both Pygodus anserinus and P. serra
(Hadding), indicating that the base of that formation
lies within the kielcensis Subzone, placing the base
of the Billabong Creek Limestone just below the
top of the Darriwilian (Da4). Re-examination of the
specimen from C828 referred to P. serra (Pickett
and Percival 2001, fig. 4C) suggests that it is better
placed in P. protoanserinus Zhang, 1998b, since the
distance between the inner and central denticle rows
is greater than that between the outer and inner rows.
Zhang’s fig. 2 suggests that the ranges of P. anserinus
and P. protoanserinus do not overlap, so their co-
occurrence implies an age right on the boundary
between the serra and anserinus Zones, and the
age can be refined to Da4b. Her detailed analysis

ORDOVICIAN CONODONTS AND SPONGES

QA
ena Z
PINS Sa
GUNNINGBLAND \
“Sunnyside”

S-Dd)
Opic
’ “New Durra

ey
6330000mN
7 f
Opb’

i

YY

1

600000mE

GEOLOGY OF THE GUNNINGBLAND AREA
West of Parkes, Central New South Wales

| | @ Alluvium and colluvium
Mz Mesozoic sediments

s-Dd Derriwong Group

I. Omz Monzonite

vv] Opw Wombin Volcanics

[SSN Oph Goonumbla Volcanics

[SS Opbe Goonumbla Voleanics - breccia
ps Gunningbland Formation

formation (PEC fauna)

Opa! Undiff limestone in Gunningbland Formation
FHP Limestone in Gunningbland Formation
ALL Allochthonous limestone in Gunningbland

ee

PEC Billabong Creek Limestone

HTR Billabong Creek Limestone

FBS Billabong Creek Limestone

Opi Pre- FBS Billabong Creek Limestone

Opie Billabong Creek Limestone, volcaniclastcs
Cy Yarrimbah Formation

Onv Nelungaloo Volcanics

q Quartz N

Grid Projection AMG Zone 55

“—— _ River

— Fault

—— Main Road
—-— Secondary Road
—-— Track

Railway

“Kirkup” = Homestead
nN

A Measured Section
~~ Kirkup unconformity

2km

Figure 1. Map of the study area west of Parkes, showing Kirkup locality (arrowed) and location of sec-
tion A — A’. After Pickett and Percival (2001).

58

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

of Pygodus species (Zhang 1998b) also indicates
no overlap between the ranges of P anitae and P
anserinus. The specimen from sample C900 referred
to P. cf. anserinus (Pickett and Percival 2001, fig. 4G)
is a tertiopedate element, and, although more strongly
denticulate than that figured by Zhang (1998b, pl.
2, fig. 17), it is possible that 1t belongs to P. anitae,
thus removing the anomalous co-occurrence of P
anitae and P. anserinus. Nonetheless, the presence of
P. anitae, possibly a later form of the species, since
the morphology of the specimen figured by Pickett
and Percival (2001, fig. 4D) lies between the two
forms figured by Zhang (1998b, fig. 2), indicates the
presence of strata at least as old as the top of the E.
suecicus Zone or basal P. serra Zone (upper Da3).

The type locality of the Nelungaloo Volcanics lies
in an excavation not more than 100 m stratigraphically
below the level of the oldest conodont assemblages
in the Nelungaloo section (Fig.1, A-A’) . Glen et al.
(2007, fig. 3) have indicated the probable presence of
Yarrimbah Formation strata between the Goonumbla
Volcanics and Nelungaloo Volcanics in this section,
as Simpson et al. (2005) have done for the Kirkup
area. However, it has not proved possible, at
least at Nelungaloo, to date the oldest strata of the
Goonumbla Volcanics, and, consequently, the onset
of post-unconformity sedimentation, as other than
Da3. The interest of the present locality is that it lies
but a short interval (a few metres) above the Kirkup
Unconformity, and gives the opportunity to constrain
the age of the base of the Goonumbla Volcanics (=
top of the Kirkup Unconformity, see below) more
precisely.

Recently, Simpson et al. (2005) have mapped a
widespread unconformity (here named the Kirkup
Unconformity) between the Yarrimbah Formation
and Goonumbla Volcanics, with the Nelungaloo
Volcanics occupying the core of the Forbes Anticline.
Their figures 1 and 2 incorporate age determinations
purportedly derived from information in Pickett and
Percival (2001). There appear, however, to have been
some errors in transcribing the data to their figures,
since the localities bearing the ages Da2 and Da4,
and lying north and slightly west of “Nelungaloo”
homestead, are those referred to above as Da3 and
Da4, respectively. Additionally, the locality with
age Da4, just southwest of “Kirkup” homestead, is
that of the locality forming the focus of the present
report, and for which no age has been given either by
Pickett and Percival (2001) or any other authors to
date. No assemblages as old as Da2 were reported by
Pickett and Percival. Consequently, the data shown
by Simpson et al (2005) as indicating the age of the
youngest post-unconformity strata are misleading.

Proc. Linn. Soc. N.S.W., 129, 2008

REFERENCE 0
E-] Siltstone
t- 50
3 Limestone <
rs fe)
f ® | 100 2
fv] Volcanics o ©
E =
ro}
150 e
50}
<
200 o
To)
D
&
<
N =
Orthograptus oO
Ea3 Climacograptus
Dicellograptus
827, C893, C894
Ea2 aint
no
outcrop
o
c
le}
se)
(77)
ea ES :
C839 = = _ C897 £
Ea 1 C836 agg CO 5
cS Sa ~<
oO
®
—
C836 O
833 : D
Gi C832 C913 5
C830 et 6902 ®
C901 =
a
Da 4 eae Pygodus anserinus
n
L
=
o
2
(e)
Da3 csgs,c905 Pygodus anitae >
C899,C907 Is}
Te}
=
=}
=
fe)
fe}
Oo
C910 Car
C906
Da 2 =e A

Figure 2. Stratigraphic section A — A’ from Fig. 1,
showing location of samples discussed in the text,
the level of the sampled limestone lens at Kirkup
correlated back to this section (arrowed), and ages
in terms of the Australian Ordovician stages (Da =
Darriwilian, Gi = Gisbornian, Ea = Eastonian).

59

ORDOVICIAN CONODONTS AND SPONGES

BIOSTRATIGRAPHY AND BIOFACIES

The association of Juanognathus serpaglii,
although very rare, with abundant Erraticodon
balticus and Protopanderodus cf. varicostatus in
the Kirkup fauna indicates an early Darriwilian age
(Da2, upper variabilis Zone) for the fauna. This age
determination is also supported by the occurrence of
E. balticus in the basal Weemalla Formation exposed
in the Panuara district, southwest of Orange in central
New South Wales. In the basal Weemalla Formation,
E. balticus (referred to as E. sp.) was found co-
occurring with Ansella jemtlandica?, Periodon
macrodentatus, Drepanodus? bellburnensis, Par-
oistodus originalis?, Protopanderodus cooperi, P.
robustus, P. varicostatus, and Dzikodus hunanensis,
which also suggested an early Darriwilian (Da2) age
(Zhen and Percival 2004b). This age for the base of
the Weemalla Formation is consistent with the Da3
graptolite occurrence ata higher level inthe unit (Smith
1966; Zhen and Percival 2004b). In the Table Head
Formation of western Newfoundland, Juanognathus
serpaglii and Erraticodon balticus occur together
in the upper part of the Histiodella tableheadensis
Assemblage Zone, which was correlated with the
upper variabilis Zone (Stouge 1984).

In his study of the conodont faunas from the
Table Head Formation Stouge (1984) recognized a
Parapanderodus-Scalpellodus biofaces, which was
further subdivided into an inner shelf sub-facies
dominated by the occurrence of Erraticodon balticus,
and an outer shelf sub-facies characterized by the

dominant occurrence of Ansella. The Kirkup fauna
with its dominance of E. balticus and Kirkupodus
tricostatus gen. et sp. nov. is most similar to the fauna
of the inner shelf sub-facies of the Parapanderodus-
Scalpellodus biofacies.

SAMPLING LOCALITIES

The limestone yielding the assemblage reported
here lies on Kirkup station, 15 km west of the town of
Parkes in the central west of New South Wales (Fig.
1). The outcrop extends for a few hundred metres,
from approximately GR 594700 6237900 to GR
595000 6238300 (m, AMG; Parkes 1:50,000 sheet,
8531 I & IV). Its thickness is about 1.5 m, though
the tumbled nature of the outcrop hinders accurate
measurement.

In addition to the extensive conodont assemblage
which affords the basis of the age determination,
there is a small fauna of anthaspidellid sponges, all
of which are completely desilicified and broken,
and the stromatolite-like stromatoporoid Janilamina
which is quite common and reaches considerable
size. A few, generally damaged brachiopods and
rare gastropods are the only other macrofossils. The
macrofossils occur in the more terrigenous parts of
the unit. Algal oncolites with abundant Girvanella are
not uncommon (Fig. 3A, B), oolites occur frequently
although a true oolitic limestone is never developed,
and there are patches of a coquina of small shells
most of which lie in the concave-up position (Fig.

Figure 3. Sedimentary features of the limestone lens at Kirkup. A, thin section of oncolites, MMF 44877,
x 3.7. B, detail of the smaller oncolite, showing tubes of the alga Girvanella sp., x 40. C, vertical thin sec-
tion of coquina of small shells, younging upward, MMF 44874, x 4.6.

60

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

©liplao|?e ©
Je}OL Nl] 0 | 2) eo ©
R/U s/s/° ("13
(oe) (co) [<o) (=)
oveco |8]e//$| | [8
veo felsfele| |
= =
weco |elele|a| | [8
= lS (90) N
evezo |8/8/8/8/>] [8
N wT N
eveco |8/8/8/8/-| |S
ww N w
we (Sla/al8] | |g
LO bse =
ovezo |8/e/3/8/ | |e
ceca [8[-lo[~| | |s
zoeeo [~| | [| | [e
oe fal -||2| | Ie
owco [=1-| [>| | a
wo [51 | [ol |
eezeo [-| | | | | [-
wsezo || | [>| | |@
ered |S [afolo| |
>
S)
eS
See
in} Y
allt
SQ} S| 5 <
L} S] oo | 8
S} do] 2/ S| &
=| S|] S] 0] 8
Q on £/ 8S] 81s
= S| 3] So
S/S} 8/8) 8/8
SSIs] LQ] 2) 3
S|} 5] S18] BIS
§|3|3} |S] 8
S/S) S/S8] Sis
=/§/ 8/8 S18
S/S) S131 3/3] _
S/S ]L] S$] s o
Ss) 2) S| SESS hs3
Ry} AL} AY Stale

Proc. Linn. Soc. N.S.W., 129, 2008

Table 1. Distribution of conodont species in samples from the limestone lens exposed near Kirkup Station, Gunningbland, New South Wales

(samples C2340 - 2347 collected from calcarenite).

3C). Sedimentary features are thus in accord with a
shallow-water situation, well within the photic zone
and above wave-base, and tally with its position
immediately above the Kirkup Unconformity.

There are two major lithologies in the limestones
recognized in the outcrop, calcarenite and packstone.
Conodonts were rare in the former, but very abundant
in the greyish fine grained packstone (see Table 1).

SYSTEMATIC PALAEONTOLOGY

Illustrations in Figures 3-4 are optical microscopic
photographs of thin sections in transmitted light.
These specimens bear the prefix MMF, and are
housed at the Londonderry Geoscience Centre of the
Geological Survey of New South Wales. Many of the
sections used in this study were made well prior to
identification of the taxa, and some of the originally
numbered specimens contain more than one species.
Consequently, parts of and thin sections from a single
block are differentiated by a lower case letter appended
to the specimen number. All photographic illustrations
shown in Figures 5 to 11 are SEM photomicrographs
of conodonts captured digitally (numbers with the
prefix IY are the file names of the digital images) and
are held in the Palaeontology Section of the Australian
Museum. Figured specimens bear the prefix AM
F. and are deposited in the type collections of the
Palaeontology Section at the Australian Museum in
Sydney. Conodont samples with the prefix C form part
of the collections of the New South Wales Geological
Survey at Londonderry.

Phylum PORIFERA Grant, 1836
Class DEMOSPONGIAE Sollas, 1885
Informal taxon LITHISTIDA Schmidt, 1870
Family ANTHASPIDELLIDAE Ulrich, 1889
Genus Patellispongia Bassler, 1927

Type species
Patellispongia oculata Bassler, 1927

?Patellispongia sp.
Figures 4A — 4E

Material

Several large fragmentary sponges (MMF29060,
29069, 29978, 29979, 35561b, 44871-3), with five
thin sections.

Description
In hand specimen the material presents the

61

ORDOVICIAN CONODONTS AND SPONGES

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

appearance of a bladed or occasionally possibly
patellate sponge. The thickness of the soma varies
from 4.5 to 9.5 mm, with a lateral dimension which
may exceed 12 cm. The fragments are all firmly
embedded in matrix, so that the precise shape of
the sponge cannot be determined, as is the case for
its surface, though this appears to be undulose and
somewhat uneven.

The thin sections are more or less random, but
those that have an orientation nearly parallel to the
growth direction show an axis of divergence of the
trabs which is much closer to one side of the soma
than the other, interpreted as the inner, or, possibly,
the excurrent surface. The trabs are 0.2 — 0.25 mm
wide and separated by a distance of 0.3 —0.5 mm. The
shafts of the dendroclones are c. 0.07 mm in diameter,
and clearly show the typical ladder-like anthaspidellid
arrangement. No details of coring or accessory
spicules could be observed due to calcification.

Remarks

Although the specimens are all fragmentary,
the manner in which the trabs diverge is similar to
that of Patellispongia australis Rigby and Webby, as
figured by them (1988, pl. 13, fig. 3). In that species
the axis of divergence of the trabs lies much nearer to
the upper, or concave surface, and this is the basis for
the interpretation of the present material. P. australis
has coring monaxons and, to judge by pl. 13, fig. 9 of
Rigby and Webby (1988), also oxeas which lie free in
the spicule network. Calcification precludes obtaining
any further confirming evidence from our material.

Anthaspidellid gen et sp. unident.
Figure 4F

Material

A single specimen with one good transverse
section, MMF44878, a probable second specimen,
MMEF3556la, and a third specimen too small for
sectioning, MMF29070.

Description
Body of sponge cylindrical or possibly obconical,
reaching 2.5 cm in diameter, and probably exceeding
this in length. The axial area is occupied by a bundle
of rounded excurrent canals 1.3 —2.0 mm in diameter,

the group itself about 1 cm across, and comprising
roughly 20 canals, separated by a screen made of up
of a single layer of spicules. Details of the exterior
are unknown, but it is apparently fairly smooth. The
offcut from the transverse section suggests that the
excurrent canals end in an apical depression, but if
so it was probably shallow. There is no indication
of a dermal layer of differentiated spicules, but the
sponges were probably somewhat eroded.

The skeleton is typically anthaspidellid, the
trabs made up of fused spicule rays and reaching
a maximum diameter of 0.3 mm. The trabs are
near vertical at the axis, but diverge and are nearly
horizontal at the periphery. Between the trabs the
spicule shafts are 0.25 — 0.3 mm apart. The material is
entirely desilicified, and it is not possible to determine
if the trabs include coring spicules.

Affinities

The scant material and its preservation make a
generic assignment hazardous. Of the more or less
cylindrical anthaspidellids described by Rigby and
Webby (1988) only those ascribed to Aulocopium and
Hudsonospongia can be compared with the present
specimen, although it appears to lack the deep apical
spongocoel of those forms. The former genus has
since been transferred to the family Streptosolenidae
(Finks et al. 2004), but in the present material
the dendroclones lie parallel to the surface, as is
characteristic of Anthaspidellidae.

Class STROMATOPOROIDEA Nicholson and
Murie, 1878 k
Order 7?CLATHRODICTYIDA Bogoyavlenskaya,
1969
Family unassigned
Tanilamina Pickett and Zhen gen. nov.

Type species
Tanilamina kirkupensis Pickett and Zhen sp. nov.

Remarks

The genus is named for our friend and colleague
Dr Ian Percival, in recognition of his contribution to
knowledge of the Ordovician System in New South
Wales.

Figure 4 (LEFT). Anthaspidellid sponges from Kirkup. A — E, ?Patellispongia sp. A, longitudinal (left)
and near transverse (right) sections of two specimens, MMF 44872, x 2.8. B, detail of left specimen
from A, showing locus of axis of divergence of trabs close to right side of skeleton, x 7.1. C, section of
blade , MMF 35561b, x 4.2. D, section of curved blade, MMF 29060a, x 1.7. F, anthaspidellid gen. et sp.

unident., MMF 44878, x 2.4.

Proc. Linn. Soc. N.S.W., 129, 2008

63

ORDOVICIAN CONODONTS AND SPONGES

Diagnosis

A stromatoporoid whose skeleton consists of thin,
extensive, densely porose laminae, with a thread-like
tissue occupying some latilaminae.

Tanilamina kirkupensis Pickett and Zhen sp. nov.
Figure 5

Material

MMEF29887 (holotype), paratypes MMF35560,
MMF44870, 44875, 44876, 44879; eight thin sec-
tions.

Description

The organism forms stromatolite-like bodies,
initially broadly encrusting, but rapidly developing a
domical shape, sometimes expanding upwards. The
margins are smoothrather than ragged, the major bursts
of growth in macroscopic appearance being more
or less enveloping. These bodies reach dimensions
greater than 12.5 cm wide and 9 cm high. The largest
specimen (MMF44879) is an irregularly laminate
body with undulose laminae and many inclusions
of sediment, and spreading to a width of at least 20
cm, while not more than 7 cm high; the holotype is
domical, 20 cm x 14 cm and 11 cm high. The skeleton
is comprised of latilaminae ( = incremental units of
Stearn and Pickett, 1994) which range from 0.1 mm
to 1.3 mm in thickness, are discontinuous laterally,
and present varying appearance in longitudinal
section, due in the main to diagenetic features. The
upper surfaces of the latilaminae are defined by the
thin laminae, which are remarkably smooth, appear as
a very thin, discontinuous, dark line, which, in areas
where there has been development of sparry calcite,
usually simply vanish, though this can be seen to
be a progressive degradation of the structure during
diagenesis. The laminae frequently are turned down
to terminate on the upper surface of the previous
lamina (Fig. 5E).

Some latilaminae present a brown and rather
flocculent appearance between the laminae. Others

are light in colour, demarcated by the dark line
(lamina) on their upper surface, and show internally
a vague network of thread-like calcified tissue,
whose structure is not clearly delineated and is
never as strongly calcified as the laminae (Fig. 5F).
The appearance of these layers intergrades with that
of the brownish, flocculent layers, so it is probable
that the latter are layers which have undergone more
diagenetic alteration.

In tangential section the thin, dark laminae can
be seen to be minutely, irregularly porous (Figs 5F,
5J). The pores are subangular to subrounded, have an
internal diameter of 0.05 — 0.1 mm, and are separated
by a delicate meshwork of calcified tissue about
0.025 mm wide between the pores. No structures
approximating to astrorhizae have been identified.

Associated features

The vertical succession of latilaminae is
occasionally interrupted. This is most commonly the
result of accumulation of sediment on the surface,
which is then covered by the next incremental
unit (Fig. 5B). Interruptions may also be caused
by algal mats in which tubes of Girvanella may
clearly be seen, or by overgrowth by an unidentified
anthaspidellid sponge (Fig. 5D). Finally, there are
small encrustations of what are probably the early
stages of bryozoan colonies (Fig. 5G) or possibly
algae, but the small size of these (c. 1 mm, with tubes
0.075 — 0.1 mm in diameter) suggests that they were
rapidly overgrown by the stromatoporoid.

Remarks

The most similar form described so far is the
marginally older Zondarella Keller and Fligel,
1996, from the late Arenig (= earliest Darriwillian
of Argentina. The type species, Z. communis, forms
large, stromatolite-like masses which even construct
reefs, quite different from the scale of the present
occurrence. The poorly developed vertical elements
of lanilamina resemble to some extent those of
Zondarella, but the well developed pores in the
laminae of the former have not been described from

Figure 5 (RIGHT). lanilamina kirkupensis gen. et sp. nov., all from a limestone lens at the base of the
Goonumbla Volcanics, Kirkup station, Gunningbland, NSW. A, appearance in hand specimen, MMF
44870b, x 0.6. B — F, longitudinal sections. B, MMF 44876, encrusting on anthaspidellid sponge (lower
centre), x 1.2. C, MMF 44875a, also encrusting an anthaspidellid, x 1.2. D, section MMF 35560a, speci-
men with rather flat laminae, and enclosing a small anthaspidellid (top right), x 1.4. E, detail of B, show-
ing terminations of laminae, x 6.8. F, somewhat oblique section through one incremental unit, showing
porous laminae and thread-like internal structure, MMF 44875b, x 13.2. G, detail of C, showing small
encrusting ?bryozoans, overgrown by later incremental units, x 10.5. H, tangential section of holotype
MME 29887b, section has traversed a rather flat lamina (dark area, top centre), x 1.7. J, detail of central
part of H, showing detail of the pores in the lamina, x 49.

64 Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

Proc. Linn. Soc. N.S.W., 129, 2008 65

ORDOVICIAN CONODONTS AND SPONGES

Zondarella. On the other hand, breaks in the laminae of
Zondarella are quite common (e.g. Keller and Fliigel
1996, pl. 47, fig.6), and it may be that these equate
to the pores of Janilamina. The epizoic ?bryozoans
illustrated by Keller and Fligel (pl. 47, fig. 8) are
almost identical with those observed on Janilamina.
Photographs of thin sections of topotype specimens of
Zondarella communis, kindly supplied by Dr Marcelo
Carrera, do not show laminae with pores similar to
those of Ianilamina. Zondarella was assigned to the
family Pulchrilaminidae Webby 1993, which was
included questionably as the last family in the order
Labechiida by Webby, in Stearn et al. (1999). The soft
tissue of members of that order appears to have been
external to the skeleton, whereas the anatomy of most
other stromatoporoid groups suggests, by analogy
with Vaceletia, e.g. Stearn and Pickett (1994), that
most of the soft tissue was internal. In spite of other
similarities with Zondarella the absence of pores
in that genus is particularly significant, pointing to
labechiid affinities. The pores of Janilamina however
imply relationship to non-labechiid stromatoporoids,
and for this reason the genus is tentatively included
amongst the clathrodictyids.

A corollary of the virtual absence of pillars or
other well calcified vertical structures between the
laminae would have been a skeleton which was
relatively weak structurally. In spite of this, there
is very little evidence of internal damage, although
a number of the specimens have clearly suffered
external damage.

There has been much discussion with colleagues
over the affinities of this material. The preliminary
field identifications cast it as stromatolitic, chiefly
perhaps due to the rather straggly outline of the
individual masses; and indeed, many layers within
them are clearly sedimentary. It is possible that the
structures here interpreted as porous laminae may
be residual features of algal structures otherwise
destroyed by diagenetic processes. However, the
following features have swayed the interpretation
as a stromatoporoid: 1) in addition to the layers of
sediment, sponges and bryozoans mentioned above,
there are occasional layers of undoubted algal mats,
clearly differing in preservation and structure from the
tissue immediately surrounding them; 2) the porous
laminae are extremely thin, while the true algal layers
have a substantial vertical dimension; 3) the laminae
turn down rather abruptly to terminate cleanly on the
surface of the lamina beneath, suggesting incremental
growth rather than surficial accumulation of sediment.
The problems faced in the interpretation of Janilamina
are in effect the same as those experienced by Keller
and Fliigel (1996) when describing Zondarella, and

66

efficiently summarised by them (p. 186). Many
of their comments also are relevant to the present
material, noting particularly the association with a
high-energy environment, provision of a hard surface
for encrusting organisms, and layers of different
composition or structure.

Phylum CHORDATA Balfour, 1880
Class CONODONTA Pander, 1856
Order PRIONIODINIDA Sweet, 1988
Family CHIROGNATHIDAE Branson and Mehl,
1944
Genus Erraticodon Dzik, 1978

Type species
Erraticodon balticus Dzik, 1978.

Diagnosis

Septimembrate or octomembrate apparatus with
a ramiform-ramiform structure including makellate
M, alate Sa, bipennate Sb and Sc, tertiopedate or
tripennate Sd, digyrate Pa, trigyrate Pb, and often
tripennate Pc, hyaline elements with a prominent
cusp, discrete peg-like denticles on the processes, and
a shallow basal cavity.

Discussion

Represented by at least 10 species with high
morphological variation, Erraticodon consisting
of large, hyaline ramiform elements is an easily
recognizable common element in shallow-water,
inner-shelf conodont faunas. A number of Erraticodon
species with a septimembrate or octomembrate
apparatus have been recently documented from late
Early and Middle Ordovician strata in Australia
(Zhen et al. 2003, Zhen and Percival 2004a, Zhen
and Percival 2004b, Zhen and Percival 2006; Nicoll
and Kelman 2004) and in South China (Zhen ef
al. 2007). Known as the oldest genus in the family
Chirognathidae (Sweet 1988), Erraticodon was
geographically widely distributed with a stratigraphic
range from the late Early Ordovician (evae Zone)
to the late Mid Ordovician (Darriwilian) (Zhen
et al. 2007). The following species are assigned to
Erraticodon:

Erraticodon balticus Dzik, 1978; defined herein
as having a septimembrate apparatus; type material
from an erratic boulder of Middle Ordovician (possibly
early Darriwilian) age, found near Kartuzy, Poland. It
was also recorded from the lower Darriwilian (Da2)
of Newfoundland (Stouge 1984), and central New
South Wales (this study).

Erraticodon bellevuensis Zhen and Percival

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

2004a; reported as having a septimembrate apparatus
(Zhen and Percival 2004a, p. 94-96, figs. 12, 13),
from allochthonous limestone (Middle Ordovician,
Darriwilian, ?Da3) in the Oakdale Formation of
central New South Wales.

Erraticodon fenxiangensis Ni, in Ni and Li
1987; not adequately documented, four morphotypes
referrable to M, Sa, Sb and Sc elements recognised
(see Ni, in Ni and Li 1987, p. 408-409, pl. 60, figs.
6=Sc, 7=Sa, 14=Sb, 15=M), from the lower and
middle parts of the Guniutan Formation (Middle
Ordovician, Darriwilian) of Yichang, Hubei, South
China.

Erraticodon . gratus (Moskalenko, 1977)
(see Moskalenko 1989); recorded as having a
septimembrate apparatus, from the Ordovician of
Russia.

Erraticodon hexianensis An and Ding, 1985;
revised as having an octomembrate apparatus (An and
Ding 1985; Zhen ef al. 2007), from the late Dawanian
to early Darriwilian of South China.

Erraticodon patu Cooper, 1981; recorded as
having a septimembrate or possibly octomembrate
apparatus (see Zhen et al. 2003; Nicoll and Kelman
2004) from the Horn Valley Siltstone (Early
Ordovician, evae Zone) of the Amadeus Basin of
central Australia, and also reported from the Tabita
Formation (and various other units of the same age)
in western New South Wales (Zhen ef al. 2003, Zhen
and Percival 2006).

Erraticodon tangshanensis Yang and Xu, in An et
al. 1983; reported having a septimembrate apparatus
(see An et al. 1983, p. 95-97), from the Majiagou
Formation and Beianzhuang Formation of Middle
Ordovician age in North China.

Erraticodon tarimensis Zhao et al. 2005; reported
as having a septimembrate apparatus (Zhao ef al.
2005, p. 30, pl. 2, figs 1-13), from the middle part of
the Upper Quulitag Formation (late Early Ordovician)
in the Tarim Basin of Northwest China.

Erraticodon sp. Lofgren, 1985 (p. 124, fig.
4AT—AY); with M, Sc, Pa, and Pb? (or Sa?) elements
illustrated from core samples of Middle Ordovician
age (upper para to lower variabilis zones) at
Finngrundet, south Bothnian Bay, Sweden.

Erraticodon balticus Dzik, 1978
Figures 6-8

Synonymy

Erraticodon balticus Dzik, 1978, p. 66-67, text-
fig. 6a-e, pl. 15, figs 1-3, 5-6 (text-fig. 6d and pl. 15,
fig. 5 = M element; text-fig. 6e and pl. 15, fig. 6 = Sa

Proc. Linn. Soc. N.S.W., 129, 2008

element; text-fig. 6c and pl. 15, fig. 3 = Sb element;
text-fig. 6a and pl. 15, fig. 1 = Sc element; text-fig. 6b
and pl. 15, fig. 2 = Sd element); Stouge 1984, p.84-
85, pl. 17, figs 9-19 (fig. 11 = M element; figs 17-18
= Sa element; figs 9-10 = Sb element; fig. 19 = Sc
element; figs 13-15 = Sd element; figs 12, 16 = Pb
element).

Erraticodon sp. Zhen and Percival 2004b, p. 167-
168, figs 8A-H, 9A-I (9A-B=M element; 9C-G=Sa
element; 9H=Sb element; 9I=Sc element; 8A-B and
8H=Pa element; 8C-G=Pb element).

Material

1726 specimens recovered from 16 samples (see
Table 1) collected along the strike of the 1.5 m thick
limestone lens.

Diagnosis

A species of Erraticodon with a septimembrate
ramiform-ramiform apparatus, including makellate
M, alate Sa, bipennate Sb and Sc, tripennate (modified
bipennate) Sd, digyrate Pa, and trigyrate (modified
digyrate) Pb elements; all elements hyaline with a
prominent cusp, and discrete, peg-like denticles on
the processes, and a shallow open basal cavity often
with basal cone attached; Sa element with short lateral
processes each bearing a single denticle and with a
long posterior process; Sa, Sb and Sc elements with
an excessively enlarged denticle (location varying
from the first to fourth away from the cusp) developed
on the posterior process.

Description

M element makellate (Fig. 6A-D) with a long
outer lateral process bearing four to seven pointed
denticles, and a sharp costa along the inner lateral face
with neither denticles nor anti-cusp (Fig. 6A); cusp
robust, antero-posteriorly compressed, and distally
curved posteriorly, with broad anterior and posterior
faces, and sharp costate lateral margins (Fig. 6A-C);
basal buttress weakly developed on the posterior face
(Fig. 6C); basal cavity shallow, tapering into narrow
groove towards the outer lateral end of the base (Fig.
6A-B) with gently arched basal margin in anterior
(Fig. 6D) or posterior view (Fig. 6A).

Sa element alate, symmetrical, bearing a long
denticulate posterior process, and a short lateral
process on each side with a single denticle (Fig.
6E-K); cusp triangular in cross section with a broad
anterior face, and a sharp costa along its posterior
margin and on each antero-lateral side; three sharp
costae extending basally and merging with the upper
margin of the posterior and the lateral processes;
posterior process long, in most specimens broken

67

ORDOVICIAN CONODONTS AND SPONGES

Figure 6. Erraticodon balticus Dzik, 1978 A-D, M element; A, AM F.133041, C2340, posterior view
(1Y85004); B, AM F.133042, C2340, basal view ([Y85003); C, AM F.133043, C2340, upper view (LY85005);
D, AM F.133044, C2340, anterior view (1Y85001). E-K, Sa element; E-G, AM F.133045, C2340, E-F, an-
terior views ([TY85022); G, upper view (IY85023); H, AM F.133046, C2343, lateral view (LY85010); I-J,
AM F.133047, C2340, I, posterior view ([Y85020), J, basal view (TY85021); K, AM F.133048, C2342,
lateral view (1Y88017). L-P, Sb element; L, AM F.133049, C2340, inner lateral view (1Y85018); M-N,
AM F.133050, C2342, M, basal view (1Y87033), N, inner lateral view ([Y87034); O, AM F.133051, C2340,
upper view (1Y85019); P, AM F.133052, C2343, outer lateral view (1TY85010). Scale bars 100 pm.

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

before or after the excessively larger denticle, which
is typically the first or second denticle away from the
cusp, and typically over twice as wide as the cusp
on lateral view (Fig. 6K); single denticle on lateral
processes antero-posteriorly compressed with a broad
anterior and posterior face and a sharp edged lateral
side; basal cavity shallow and small with strongly
arched basal margin in lateral view (Fig. 6H, K).

Sb element bipennate, asymmetrical with a sub-
erect cusp, a long posterior process, and a downward
extending and strongly inner laterally curved anterior
process (Fig. 6L-P); cusp moderately compressed
laterally with a sharp anterior costa and a posterior
costa that extends downward to form the upper
margin of the anterior and posterior processes (Fig.
60); posterior process long, bearing more than five
denticles, with the third or fourth denticle away from
the cusp excessively larger than other denticles,
typically as large as the cusp (Fig. 6L) or even larger
(Fig. 6N-O); denticles on posterior process moderately
compressed laterally with a sharp costa along the
anterior and posterior margins; long anterior process
strongly curved inward forming an angle of about
90° or less with posterior process (Fig. 6M, O), and
bearing five to eight or even more denticles, which
are more closely spaced than those on the posterior
process, and laterally compressed with sharp edges;
basal cavity shallow, often with basal cone attached,
showing a sickle-like outline in basal view (Fig.
6M).

Sc element bipennate (Fig. 7A-E) with a
prominent cusp, a downwardly extended, slightly
inner laterally curved anterior process bearing three or
more denticles (Fig. 7C), and a long posterior process
bearing four or more long denticles (Fig. 7C-D); cusp
laterally strongly compressed with sharply costate
anterior and posterior margins and smooth lateral
faces; one denticle on posterior process (typically
the third away from the cusp) excessively larger than
other denticles; basal cavity shallow with an arched
basal margin in lateral view (Fig. 7D).

Sd element tripennate (modified bipennate) with
a prominent cusp, a short anterior process bearing
two or three denticles, a long posterior process
bearing five or more denticles, and a long, outer
lateral process bearing up to seven denticles (Fig. 7F-
M); cusp tricostate bearing sharp anterior, posterior
and outer lateral costae, which extend downward to
merge with the upper margin of the three processes;
cusp rounded in cross section (Fig. 7H) with outer
lateral costa more towards posterior, and curved inner
laterally and posteriorly with broad, convex inner
lateral face and antero-outer lateral face, but with
slightly concave postero-outer lateral face (Fig. 7H);

Proc. Linn. Soc. N.S.W., 129, 2008

anterior process inner laterally curved and extending
downward with a straight basal margin nearly normal
to the basal margin of the posterior process, denticles
closely spaced with smallest at the distal end and the
largest, which is only slightly shorter than the cusp,
next to the cusp; posterior process long, but broken
in most of the specimens recovered; denticles on both
anterior and posterior processes compressed laterally,
oval in cross section with a sharp costa along their
anterior and posterior margins; outer lateral process
long, posteriorly curved varying from a 70° angle with
the posterior process (Fig. 7G) to nearly parallel with
the posterior process (Fig. 7J), denticles moderately
compressed antero-posteriorly, oval in cross section
with sharply costate lateral margins, and curved
posteriorly.

Pa element digyrate with a less prominent cusp,
a long sinuously curved inner lateral process, and a
long, anteriorly twisted outer lateral process (Fig. 8A-
G); cusp rounded in cross section with a sharp lateral
costa on each side, and broad anterior and posterior
faces; denticles on both processes also rounded or
weakly antero-posteriorly compressed; basal cavity
shallow and inverted with a shallow pit underneath
the cusp (Fig. 8D).

Pb element trigyrate (modified digyrate) with a
lateral process on each side, and an anterior process
bearing three or more denticles (Fig. 8H-N); cusp
less prominent than that of the M and S elements,
with a broad posterior face and with a sharp costa
along anterior margin, and on each side; outer lateral
process long, bearing seven or more denticles; inner
lateral process shorter bearing typically two or three
denticles, anterior process outer laterally curved;
denticles antero-laterally compressed on the lateral
processes and laterally compressed on the anterior
process; basal cavity shallow with a pit underneath
the cusp (Fig. 8N).

Discussion

The type material of Erraticodon balticus was
recovered from an erratic boulder found near Kartuzy,
Pomerania, Poland, but believed to be transported
from the Baltic region. Dzik (1978) originally
defined the species as consisting of a seximembrate
apparatus, but only five elements including makellate
M, alate Sa, bipennate Sb and Sc, and tripennate
(modified bipennate) Sd were represented in the
type material. However he suggested an additional
spathognathiform element (Dzik 1978, Fig. 2 =
digyrate Pb element) represented by a specimen
illustrated as “Chirognathus” sp. by Viira (1974, pl.
11, fig. 22). Subsequently, Dzik (1991) indicated that
the species had a septimembrate composition, but

69

ORDOVICIAN CONODONTS AND SPONGES

Figure 7. Erraticodon balticus Dzik, 1978. A-E, Sc element; A-B, AM F.133053, C2343, A, basal-posterior
view (LY85012), B, basal view (1Y85013); C, AM F.133054, C2343, inner lateral view (1Y85017); D-E,
AM F.133055, C2342, D, outer lateral view (1Y87031), E, basal view (LY87032). F-M, Sd element; F, AM
F.133056, C2340, outer lateral view (LY85009); G-H, AM F.133057, C2343, G, upper view (1Y85015),
H, close up of upper view showing cross section of tricostate cusp (1Y85016); I, AM F.133058, C2343,
outer lateral view (LY92029); J-M, AM F.133059, C2343, J, outer lateral view ([Y92026), K, upper view
(LY92025), L, posterior view ([Y92027), M, inner lateral view (LY92028). Scale bars 100 um.

unfortunately provided neither description nor further
details of the revised apparatus. Based on material
from the Table Head Formation (Darriwilian, Da2)
of western Newfoundland, Stouge (1984) suggested a
septimembrate apparatus for E. balticus, its elements
being comparable with the M, Sa, Sb, Sc, Sd and Pb

elements defined herein from the Kirkup fauna. No
Pa element was reported by Stouge, (1984) and his
zygognathiform and sannemanulliform elements are
interpreted herein as variants of the Sd element with
the outer lateral process curved posteriorly in varying
degree.

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

Figure 8. Erraticodon balticus Dzik, 1978. A-G, Pa element; A-B, AM F.133060, C2340, A, upper view
(TY85008); B, anterior view (1Y85006); C-E, AM F.133061, C2342, C, basal-posterior view (1Y88011), D,
basal view (IY88014), E, anterior view (1Y88012); F-G, AM F.133062, C2342, F, upper view (LY88008),
G, anterior view (1Y88009). H-N, Pb element; H-J, AM F.133063, C2342, H, outer lateral view (1Y88003);
I, anterior view (IY88002), J, posterior view (IY88004); K-L, AM F.133064, C2300, K, upper view
(TY86007), L, inner-anterior view ([Y86006); M-N, AM F.133065, C2347, M, basal view (IY88005), N,
posterior view (1Y88006). Scale bars 100 nm.

Proc. Linn. Soc. N.S.W., 129, 2008

ORDOVICIAN CONODONTS AND SPONGES

E. balticus is characterized by having an
accentuated denticle on the posterior process of the
Sa, Sb and Sc elements. The holotype (Dzik 1978.
pl. i5, fig. 6) exhibits this accentuated denticle as the
third denticle away from the cusp on the posterior
process, which is slightly wider (thicker) near the base
than the cusp in lateral view. An accentuated denticle
similar to that of the holotype was also developed on
the posterior process of the Sb and Sc elements in the
type material (Dzik 1978. text-fig. 6). In our abundant
material from Kirkup, this accentuated denticle on
the posterior process of the Sa, Sb and Sc elements
is observed as a consistent character, but its position
may vary from the first to fourth away from the cusp,
and the size can also vary from equal to that of the
cusp to over twice its size (Fig. 6K).

Specimens have previously been referred to this
species by various authors (e.g. Watson 1988; Lehnert
1995; Albanesi, in Albanesi et al. 1998; Zhang 1998a),
but their illustrated specimens apparently lack this
character and should be excluded from £. balticus.
For instance, none of the Sa, Sb and Sc elements
described by Watson (1988) from the the Goldwyer
Formation of the Canning Basin of Western Australia
illustrated this diagnostic character of the species.
None of the illustrated Sa, Sb and Sc elements referred
to as E. balticus by either Zhang (1998a, pl. 9, figs 9-
10, 13) from the Guniutan Formation in Hubei and

Hunan provinces or by Ding ef al. (in Wang 1993)
from the same stratigraphic unit near Nanjing in South
China shows an accentuated denticle on the posterior
process. Zhang’s (1998a, pl. 9, fig. 9) illustrated Sa
element bears more than one denticle on the lateral
processes. The material illustrated by Ding et al. (in
Wang 1993, pl. 37, figs 18-28) as this species includes
elements belonging to different genera, and the
material illustrated by Zhang (1998a) may be more
comparable to E. hexianensis (see Zhen et al. 2007).
That species, recorded from the upper Dawanian to
early Darriwilian in South China (Zhen et al. 2007),
closely resembles EF. balticus, but its Sa, Sb, and Sc
elements lack an excessively enlarged denticle on the
posterior process, and its Sd element is tertiopedate
rather than tripennate, as in E. balticus.

Order PRIONIODONTIDA Dzik, 1976
Family OISTODONTIDAE Lindstrém, 1970
Genus Juanognathus Serpagli, 1974

Type species
Juanognathus variabilis Serpagli, 1974.

Juanognathus serpaglii Stouge, 1984
Figure 9C-F

Figure 9. A-B, Pseudooneotodus mitratus (Moskalenko, 1973) AM ¥.133066, C2347, A, upper view
(TY92012), B, outer lateral view ([Y92013). C-F, Juanognathus serpaglii Stouge, 1984. C-D, asymmetrical
element, AM F.133067, C2343, C, posterior view (1Y92021), D, upper view (LY92022); E-F, symmetrical
element, AM F.133068, C2343, E, posterior view (1Y92020), F, basal view ([Y92019). Scale bars 100 um.

72

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

Synonymy
Juanognathus serpaglii Stouge, 1984, p. 58-59,
pl. 5, figs 10-20 (cum syn.).

Material
Three specimens, two (both illustrated) from
sample C2343, and one from C2342.

Discussion

Stouge (1984) originally recognised a bimembrate
(symmetrical and asymmetrical) apparatus for the
species. Both elements are represented by only one
specimen each in our material from Kirkup station.
They are identical with the type material from the
Table Head Formation of western Newfoundland.
Both elements have a cusp with a blade-like costa on
each side, broadly convex anterior and posterior faces
and a prominent basal surface defined by a ledge-like
costa parallel to and slightly above the basal margin,
which is similar to that described in Cooperignathus
Zhen, in Zhen ef al. 2003 and Protoprioniodus
Cooper, 1981 (see Zhen ef al. 2003). This ledge-like
costa and basal surface are also well developed in the
type material of J. serpaglii (Stouge 1984, pl. 5, figs
10, 12, 16), but have not been recognised in the other
species of Juanognathus including the type species.

Order PROTOPANDERODONTIDA Sweet, 1988
Family PROTOPANDERODONTIDAE Lindstrém,
1970
Genus Protopanderodus Lindstrém, 1971

Type species
Acontiodus rectus Lindstrém, 1955.

Protopanderodus cf. varicostatus (Sweet and
Bergstrém, 1962)
Figure 10A-P

Synonymy

Protopanderodus cf. varicostatus (Sweet and
Bergstrém); Lofgren, 1978, p. 191-193, pl. 3, fig.
26-31.

Material

295 specimens recovered from 12 samples (see
Table 1) collected along the strike of the 1.5 m thick
limestone lens.

Description

Five morphotypes of this species are recognized
and assigned to Sa, Sb, Sc, Pa and Pb elements. Sa
element symmetrical, bearing a proclined to suberect

Proc. Linn. Soc. N.S.W., 129, 2008

cusp, with a broad anterior face, an antero-lateral
costa and a lateral costa on each side, and a costa
along its posterior margin (Fig. 10A-E); a broader,
shallow groove developed between the two costae
and a deep and narrow furrow positioned to the
posterior side of the lateral costa. Sb element like
Sa, but asymmetrical, with a concave inner face and
a convex outer face, and weaker development of the
antero-lateral costa on each side, and with a deep
furrow developed between the two costae on each
lateral face. Sc element strongly asymmetrical and
laterally compressed, with a convex outer face bearing
a median costa and a furrow to its posterior side, a
concave inner face bearing two costae and separated
by two furrows (Fig. 10J-K), and with a sharp costa
along its anterior and posterior margins. Pa element
with a proclined cusp and a shorter base, and with a
sharp costa along its anterior and posterior margins;
outer lateral face convex and smooth or often with
a weak and short postero-lateral costa developed
around the curvature of the cusp (Fig. 10M); inner
lateral face with two costae and a furrow in between.
Pb element like Pa element, but with a suberect cusp
and a longer base.

Discussion

The current material resembles that described as
Protopanderodus cf. varicostatus from the Middle
Ordovician of northern Sweden (Léfgren 1978),
particularly the S elements. The scandodiform
Pb elements from the Kirkup fauna generally
have a longer base, which was not recognized in
P. varicostatus, and some of the symmetrical Sa
elements show a broader anterior face (Fig. 10A)
in comparison with the corresponding elements of
Protopanderodus variabilis. Some scandodiform P
elements of those referred to Protopanderodus cf.
varicostatus by Léfgren (1978, pl. 3, fig. 30) exhibit
intermediate features between the current material,
which has a longer base (Pb), and typical P. variabilis
with a very short base.

Protopanderodus? nogamii (Lee, 1975)
Figure 11A-S

Synonymy

Scolopodus cf. bassleri \go and Koike 1967, p. 23, pl.
3, figs. 7, 8, text-fig. 6B.

Scolopodus sp. A Hill et al. 1969, p. 0.14, pl. OVI,
fig. 13.

Scolopodus sp. C Hill et al. 1969, p. 0.14, pl. OVII,
fig. 15.

“Panderodus” sp. Serpagli 1974, p. 59, pl. 24, figs.

73

ORDOVICIAN CONODONTS AND SPONGES

Figure 10. Protopanderodus cf. varicostatus (Sweet and Bergstrém, 1962). A-E, Sa element; A-B, AM
F.133069, C2343, A, basal-posterior view (1Y87005), B, lateral view (LY87004); C-E, AM F.133070,
C2342, C-D, lateral views (LY87025, [Y87023), E, basal view (IY87026). F-I, Sb element; F-H, AM
F.133071, C2343, F, basal view ([Y87013), G, inner lateral view ([Y87011), H, outer lateral view
(TY87014); I, AM F.133072, C2343, inner lateral view (LY87016). J-L, Sc element, AM F.133073, C2343,
J, inner lateral view ([Y92038), K, upper view (IY92036), L, outer lateral view ([Y92037). M-N, Pa
element; M, AM F.133074, C2343, outer lateral view (1Y87007); N, AM F.133075, C2340, inner lat-
eral view (1Y¥91040). O-P, Pb element; O, AM F.133076, C2340, outer lateral view (LY91038); P, AM
F.133077, C2340, inner lateral view ([Y91039). Scale bars 100 pm.

12 e IB ple sOy ttesaeatS: 1983, p. 140, pl. 13, fig. 27, pl. 14, figs. 1-8 (cum
Scolopodus nogamii Lee 1975, p. 179, pl. 2, fig. 13. syn.); An and Zheng 1990, p. 173, pl. 2, figs. 7-
Protopanderodus primitus Druce (MS), in Cooper 11, 13, 14, 16.

1981 (nomen nudum), p. 174, pl. 27, figs. 3, 4 Protopanderodus nogamii (Lee); Watson 1988: p.

(cum syn.); Stait and Druce 1993, p. 307, figs. 124, pl. 3, figs. 1, 6; Zhen et al. 2003, p. 207-209,

13A-C, 18D, E, G-K (cum syn.). fig. 23A-P, ?Q (cum syn.); Zhen and Percival
Scolopodus euspinus Jiang and Zhang, in An ef al. 2004a, p. 104-105, fig. 18A-K.

74 Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

Figure 11. Protopanderodus? nogamii (Lee, 1975). A-F, Sa element; A, AM F.133078, C2343, lateral view
(1Y88032); B-D, AM F.133079, C2343, B, basal view (LY88030), C-D, lateral views (LY88029, TY88031);
E-F, AM F.133080, C2340, E, basal view (1Y88044), F, lateral view (1Y88045). G-H, Sb element, AM
F.133081, C2343, G, basal view (1Y88035), H, inner lateral view (1Y88034). I-L, Sd element; I-J, AM
F.133082, C2343, I, inner lateral view (1Y88041), J, basal view (1Y88039); K-L, AM F.133083, C2343,
K, close up showing furrow does not cut through the basal margin and surface striation (1Y88038), L,
outer lateral view (1Y88036). M-O, Pb element; M, AM F.133084, C2343, basal view(1Y88018); N-O,
AM F.133085, C2343, N, inner lateral view (1Y88024), O, outer lateral view ([Y88023). P-S, Pa element;
P-R, AM F.133086, C2343, P, inner lateral view (1Y88027), Q, close up showing furrow does not cut
through the basal margin ([Y88028), R, basal view ([Y88026); S, AM F.1330087, C2343, outer lateral

view (1Y88021). Scale bars 100 pm, unless otherwise indicated.

Proc. Linn. Soc. N.S.W., 129, 2008 75

ORDOVICIAN CONODONTS AND SPONGES

Protopanderodus? nogamii (Lee); Zhen and Percival
2004b, p. 170-172, fig. 11P, Q.

Parapanderodus paracornuformis Ethington and
Clark; Albanesi, in Albanesi ef al. 1998, p. 116,
partim pl. 12, fig. 13, 8?-10?, non 11, 12.

?Panderodus nogamii (Lee); Cantrill and Burrett
2004, p. 410, pl. 1, figs 1-16.

Material

458 specimens recovered from 11 samples (see
Table 1) collected along the strike of the 1.5 m thick
limestone lens.

Discussion

The concept of P.? nogamii and its constituent
elements has been reviewed extensively in several
recent publications (Zhen et al. 2003; Zhen and
Percival 2004a, b; Cantrill and Burrett 2004). It
consists of a seximembrate apparatus including
short-based, bi-furrowed Pa and Pb, long-based,
bi-furrowed Sa, Sb and Sd, and long-based, uni-
furrowed Sc elements; with furrows and coarse striae
disappearing just before reaching the basal margin.
This species is excluded from Panderodus as it lacks
a true panderodontid furrow, which cuts deep into
basal margin. Typical species of Panderodus have
only one furrow on the outer lateral face (except
for a rare bi-furrowed symmetrical element). Most
of the P? nogamii elements have a furrow on each
lateral side, and the furrows disappear just before
reaching the basal margin (Fig. 11K, Q). Stait and
Druce (1993) recognized a uni-furrowed element in
the P.? nogamii species apparatus. Zhen et al. (2003)
found this element (referred as the Sc element) was
extremely rare in the Early Ordovician material from
Mt. Arrowsmith in far western New South Wales. In
the Kirkup fauna, P ? nogamii is one of the dominant
species, but no uni-furrowed element has been
identified.

Family ACANTHODONTIDAE Lindstrém, 1970
Genus Kirkupodus Zhen and Pickett gen. nov.

Derivation of name

After the property name, Kirkup Station, where
the type species (and only species assigned to the
genus) was recovered.

Type species
Kirkupodus tricostatus Zhen and Pickett gen. et
Sp nov.

Diagnosis
A septimembrate coniform-coniform apparatus

including nongeniculate bicostate M, bicostate Sb
and Sc, tricostate Sa and Sd, and drepanodiform Pa
and Pb elements; all elements hyaline, with sharp
blade-like costae, and ommamented with fine striae; S
elements forming a symmetry transitional series; M
element with non-expanded base, S and P elements
with posteriorly or laterally extended base.

Discussion

Kirkupodus is defined herein as having a
septimembrate apparatus. The blade-like costae of
the S elements and the general morphology of the
P elements resemble the corresponding elements
of Triangulodus van Wamel, 1974, but Kirkupodus
lacks a geniculate M element. The P elements and
asymmetrical, bicostate Sb and Sc elements of
Kirkupodus tricostatus can be closely compared
with the P and S elements of Scalpellodus Dzik,
1976. However, tricostate Sa and Sd elements were
not recognised in the species apparatus of the type
species, S. Jatus, nor from the apparatuses of the
other two species (S. gracilis and S. viruensis) of
Scalpellodus, from the Baltic, although descriptions
of all these species were based on a large number of
specimens (Lofgren 1978). These differences support
the establishment of a new genus, Kirkupodus, to
accommodate the current species from central New
South Wales.

As Lofgren (1978, p. 98) pointed out, Dzik’s
original definition (1976, p. 421) of Scalpellodus
included elements belonging to both Scalpellodus
and Cornuodus. Lofgren (1978) revised the genus
as having a trimembrate apparatus including a
scandodiform, a long-based drepanodiform and a
short-based drepanodiform elements. She recognised
three species from the upper Middle Ordovician
(Darriwilian) of Jamtland, northern Sweden, and
noted that in the younger species, S. viruensis Lofgren,
1978, the long-based drepanodiform element was
apparently missing.

The type species, S. J/atus, was originally
described as having a trimembrate apparatus including
symmetrical short-based, symmetrical long-based,
and asymmetrical scandodiform elements (van
Wamel 1974), all with prominent surface striation.
The holotype (van Wamel 1974, pl. 4, fig. 2a-b)
was defined as a symmetrical short-based element,
which Léfgren (1978) referred to as the short-based
drepanodiform element (= Pb element of our current
notation). The symmetrical long-based element (van
Wamel 1974, pl. 4, fig. la-b) was called the long-
based drepanodiform element by Lofgren (1978) (=Pa
element of our notation). Only one type of S element
was recognized for S. /atus by both van Wamel (1974)
and Léfgren (1978) as the asymmetrical scandodiform

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

element (= ?Sc element herein).

Sweet (1988) interpreted the genus as consisting
of a bimembrate apparatus. More recently, Lofgren
(2000, 2003) considered the scandodiform elements
to represent the ?S element and the drepanodiform
elements to be the ?P elements. Lofgren (2000)
illustrated three specimens of S. /atus from the lower
Middle Ordovician (7. quadrangulum Subzone) of
northern Oland, Sweden, and assigned them to the S
element (L6fgren 2000, fig. 4S), M? element (Lofgren
2000, fig. 4T), and P? element (L6fgren 2000, fig. 4U).
The doubtful M element of S. /atus is a nongeniculate
element with a short base, and shows some similarity
to the M element of Kirkupodus tricostatus from
central New South Wales (Fig. 12A-C). Later Lofgren
(2003) also illustrated three specimens of a slightly
younger species, S. gracilis, from the upper Middle
Ordovician (L. variabilis Zone, early Darriwilian) of
southern Sweden, as the Sb? element (Léfgren 2003,
fig. 7V), the Sd? element (Lofgren 2003, fig. 7X) and
the P? element (Lé6fgren 2003, fig. 7W). Although
L6éfgren’s revised notation for the species apparatus
of Scalpellodus has not been formally published, it
seems clear that Baltic species of Scalpellodus do not
have tricostate Sa and Sd elements similar to those of
Kirkupodus.

Stouge (1984) endorsed the definition of
Scalpellodus given by Lofgren (1978) as consisting
of a trimembrate apparatus and recognised a tricostate
element in both Scalpellodus pointensis Stouge, 1984
and Scalpellodus biconvexus (Bradshaw, 1969). It
exhibits sharp, edge-like posterior, anterior and outer-
lateral costae, and is comparable with the Sd element
of Kirkupodus tricostatus from central New South
Wales. However, neither a symmetrical tricostate
Sa element nor a nongeniculate M element was
recorded in the apparatus of these two species from
Newfoundland. They are considered here as doubtful
species of Kirkupodus, pending further study.

Kirkupodus tricostatus Zhen and Pickett gen. et
sp. nov.
Figure 12

Derivation of name

Tricostate, referring to the distinctive tricostate
Sa (symmetrical) and Sd (asymmetrical) elements of
the species.

Material

1480 specimens recovered from 14 samples
(see Table 1) collected along the strike of the 1.5
m thick limestone lens; including figured holotype

Proc. Linn. Soc. N.S.W., 129, 2008

AM F.133090 (Fig. 12D-F), and 18 paratypes, AM
F.133088-89, AM F.133091-133106 (Fig. 12A-C, G-
AA).

Diagnosis

A species of Kirkupodus consisting of a
septimembrate apparatus including bicostate M,
bicostate Sb and Sc, tricostate Sa and Sd, and
drepanodiform Pa and Pb elements; M weakly
asymmetrical with an antero-posteriorly compressed
cusp, Sa symmetrical, Sb asymmetrical with a
posteriorly flared base, Sc strongly asymmetrical
with base flared more postero-inner laterally, Sd
asymmetrical, Pa with a proclined cusp and a
unexpanded and inner laterally flared base, and Pb
with a sub-erect cusp and a posteriorly-expanded
base; all elements ornamented with fine striae.

Description

M element weakly asymmetrical; cusp curved
posteriorly and slightly outer laterally, antero-
posteriorly compressed with a broadly convex anterior
face, a less convex posterior face and a sharp blade-
like costa on each side forming the lateral edges; base
slightly extended posteriorly, clam-shaped in outline
in basal view with a broadly arched antero-basal
margin (Fig. 12A-C).

Sa element symmetrical and tricostate (Fig.
12G); cusp triangular in cross section (Fig. 12G),
proclined with a broad anterior face, a sharp blade-
like antero-lateral costa on each side, and a costa
along the posterior margin; base posteriorly extended
(Fig. 12D-G).

Sb element asymmetrical with a broad anterior
face, an antero-laterally located costa on the inner
side (Fig. 12L), a postero-laterally located costa on
the outer lateral side (Fig. 12J), and a posteriorly
extended base, which is oval in outline in basal view
(Fig. 12H-I, K).

Sc element strongly asymmetrical with a laterally
compressed cusp, and a base which is flared both
posterolaterally and on the inner side (Fig. 12M-
P), and widest at the posterior end (Fig. 12P); cusp
suberect to weakly proclined with sharp costa along
its anterior and posterior margins and with smooth
lateral faces.

Sd element asymmetrical with a broad anterior
face, and with a sharp, blade-like costa on each
lateral side and along the posterior margin (Fig. 12Q-
T); cusp proclined, triangular in cross section (Fig.
12S), with inner lateral costa more anteriorly located,
and varying from weakly asymmetrical (Fig. 12T) to
strongly asymmetrical with a twisted cusp (Fig. 12Q);
base extended posteriorly, oval to triangular in basal

77

ORDOVICIAN CONODONTS AND SPONGES

Proc. Linn. Soc. N.S.W., 129

Y.Y ZHEN AND J. PICKETT

view (Fig. 12Q, T).

Pa element with a proclined cusp and an inner
laterally flared base (Fig. 12X-AA); cusp with sharp
anterior and posterior margins, and broadly convex
lateral faces, outer lateral face bearing a broad carina
located antero-laterally; sharp anterior and posterior
edges disappearing before reaching the basal margin;
base flared inward, long and posteriorly without
prominent expansion (Fig. 12X, AA), flared inward,
widest at the mid point (Fig. 12Y-Z) and the outer
side of the basal margin more or less straight in basal
view (Fig. 12Z).

Pb element similar to Pa, but weakly asymmetrical
with a suberect cusp (Fig. 12U-W) and a shorter,
less inner laterally flared, but posteriorly, strongly-
expanded base (Fig. 12U-V).

Discussion

Kirkupodus tricostatus can be differentiated
from the known species of Scalpellodus in having a
symmetrical tricostate Sa element and an asymmetrical
tricostate Sd element, and by lacking the long-based
elements in the apparatus. In comparison with S. /atus
and other two species from the Middle Ordovician of
northern Sweden (Lofgren 1978), the current species
has strongly developed blade-like costae in the M and
S elements, and finer surface striae, although the P
elements are comparable with the Baltic species of
Scalpellodus.

Order and Family Uncertain
Genus Pseudooneotodus Drygant, 1974

Type species
Oneotodus? beckmanni Bischoff and
Sannemann, 1958.

Pseudooneotodus mitratus (Moskalenko, 1973)
Figure 9A-B

Synonymy

Ambalodus mitratus mitratus Moskalenko, 1973, p.
86, pl. 17, figs 9-11.

Pseudooneotodus mitratus (Moskalenko); Nowlan
and McCracken in Nowlan ef al. 1988, p. 34, pl.
16, figs 2-6 (cum syn.); Pohler and Orchard 1990,
pl. 6, fig. 12; Trotter and Webby 1995, pl. 4, figs
21-22 (cum syn.); Zhen and Webby 1995, p. 285,
pl. 4, figs 16-17; Zhen et al. 1999, p.92-94, fig.
9.14-9.15; Zhen et al. 2003, fig. 6Q.

Material
A single specimen from sample C2346.

Discussion

Occurrence of this species in the Kirkup fauna
is very rare, only represented by one element among
a huge collection. It is rather common in the Upper
Ordovician in central New South Wales, previously
reported from the Fossil Hill Limestone (Zhen and
Webby 1995), the Ballingoole Limestone of the Bowan
Park Group (Zhen et al. 1999) and from the Late
Ordovician allochthonous limestones in the Barnby
Hill Shale (Zhen et a/. 2003). Morphologically, the
current specimen is identical with the morphotype
showing a smooth surface without nodes on the flanks

Figure 12 (LEFT). Kirkupodus tricostatus gen. et sp nov. A-C, M element; A-B, AM F.133088, paratype,
C2343, A, posterior view (I1Y86017), B, basal view ([Y86018); C, AM F.133089, paratype, C2343, up-
per view (1Y86019). D-G, Sa element; D-F, AM F.133090, holotype, C2340, D, lateral view ([Y91034),
E, basal view (IY91035), F, posterior view ([Y91036); G, AM F.133091, paratype, C2347, upper view
(1Y92001). H-L, Sb element; H, AM F.133092, paratype, C2344, upper view ([Y92007); I, AM F.133093,
paratype, C2344, basal view ([Y91027); J, AM F.133094, paratype, C2343, basal view (TY86020); K, AM
F.133095, paratype, C2343, outer lateral view (1Y86022); L, AM F.133096, paratype, C2343, inner lateral
view (1Y86024). M-P, Sc element; M-N, AM F.133097, paratype, C2343, M, inner lateral view (TY86029),
N, closing up showing surface striation ([TY86030); O-P, AM F.133098, paratype, C2343, O, outer lateral
view (1Y86027), P, basal view (1Y86028). Q-T Sd element; Q, AM F.133099, paratype, C2344, postero-
basal view (1Y91019); R, AM F.133100, paratype, C2343, inner lateral view (1Y86040); S, AM F.133101,
paratype, C2343, upper view (1Y86033); T, AM F.133102, paratype, C2343, basal-outer lateral view
(LY 86036). U-W, Pb element; U, AM F.133103, paratype, C2343, inner lateral view ([Y86009); V-W, AM
F.133104, paratype, C2343, V, outer lateral view ([Y86011), W, basal view ([Y86014). X-AA, Pa element;
X-Y, AM F.133105, paratype, C2343, X, inner lateral view ([Y86012), Y, basal view ([Y86013); Z-AA,
AM F.133106, paratype, C2344, Z, basal view (1Y91028), AA, outer lateral view (LY91029). Scale bars

100 pm, unless otherwise indicated.

Proc. Linn. Soc. N.S.W., 129, 2008 72)

ORDOVICIAN CONODONTS AND SPONGES

of the posterior and lateral ridges described from
various stratigraphic units of the Late Ordovician
in central New South Wales. The occurrence of P.
mitratus in the Kirkup fauna of early Darriwilian age
represents the earliest record of this species.

ACKNOWLEDGEMENTS

Field work was supported by a grant (Betty Mayne
Scientific Research Fund) to ZYY from the Linnean Society
of New South Wales. Gary Dargan (Geological Survey of
New South Wales) assisted with acid leaching and residue
separation. Scanning electron microscope photographs
were prepared in the Electron Microscope Unit of the
Australian Museum. We are grateful to Dr Marcelo Carrera
for supplying photographs of Zondarella communis. JWP
is an Honorary Research Associate of both the Australian
Museum and the Geological Survey of New South Wales,
and is grateful to these organisations for the provision of
facilities. The authors appreciate the careful commentary
of two anonymous reviewers. This is a contribution to
IGCP Project 503: Ordovician Palaeogeography and
Palaeoclimate.

REFERENCES

Albanesi, G.L., Hiinicken, M.A. and Barnes, C.R.
(1998). Bioestratigrafia, biofacies y taxonomia
de conodontes de las secuencias ordovicicas del
Cerro Porterillo, Precordillera central de San Juan,
R. Argentina. Actas de la Academia Nacional de
Ciencias 12, 1-249.

An, T.X. and Ding, L.S. (1985). Ordovician conodont
biostratigraphy in Hexian, Anhui Province.
Geological Review 31 (1), 11—20 (Gn Chinese).

An, T.X., Zhang, F., Xiang, W.D., Zhang, Y.Q., Xu, W.H.,
Zhang, H.J., Jiang, D.B., Yang, C.S., Lin, L.D.,
Cui, Z.T. and Yang, X.C. (1983). “The Conodonts
of North China and the Adjacent Regions’. 1—223,
Science Press, Beijing (in Chinese with English
abstract).

An, T.X., and Zheng, S.C. (1990). “The conodonts of the
marginal areas around the Ordos Basin, northern
China’. 1-201, Science Press, Beijing (Chinese with
English abstract),

Bassler, R.S. (1927). A new Early Ordovician sponge
fauna. Journal of the Washington Academy of
Science 17, 390-394.

Bischoff, G. and Sannemann, D. (1958). Unterdevonische
Conodonten aus dem Frankenwald. Notizblatt des
hessischen Landesamtes fiir Bodenforschung 86,
87-110.

Bradshaw, L.E. (1969). Conodonts from the Fort Pefia
Formation (Middle Ordovician), Marathon Basin,
Texas. Journal of Paleontology 43, 1137-1168.

80

Branson, E.R. and Mehl, M.G. (1944). Conodonts. In
‘Index fossils of North America’ (Eds. H.W. Shimer
and R.R. Shrock) p. 235-246. (Wiley, New York).

Cantrill, R.C. & Burrett, C.F. (2004). The Greater
Gondwana distribution of the Ordovician conodont
Panderodus nogamii (Lee) 1975. Courier
Forschungsinstitut Senckenberg 245, 407-419.

Cooper, B.J. (1981). Early Ordovician conodonts from
the Horn Valley Siltstone, central Australia.
Palaeontology 24, 147-183.

Drygant, D.M. (1974). Prostye Konodonty Silura iz
nizhego Devona Volyno-Podolya [Simple conodonts
from the Silurian and lower Devonian of Volhynia-
Podolia]. Paleontologicheskiy Sbornik 10, 64—69 (in
Russian).

Dzik, J. (1976). Remarks on the evolution of Ordovician
conodonts. Acta Palaeontologica Polonica 21,
395-455.

Dzik, J. (1978). Conodont biostratigraphy and
paleogeographical relations of the Ordovician
Mojcza Limestone (Holy Cross Mts, Poland). Acta
Palaeontologica Polonica 23, 51—72.

Dzik, J. (1991). Evolution of oral apparatuses in the
conodont chordates. Acta Palaeontologica Polonica
36, 265-323.

Finks, R.M., Reid, R.E.H., & Rigby, J.K. (2004). “Treatise
on Invertebrate Paleontology. Part E, Porifera
(revised)’, vol. 3 (Demospongea, Hexactinellida,
Heteractinida, Calcarea). 1-872. (Geological Society
of America and University of Kansas).

Glen, R.A., Crawford, A.J., Percival, I.G., and Barron,
L.M. (2007). Early Ordovician development of the
Macquarie Arc, Lachlan Orogen, New South Wales.
Australian Journal of Earth Sciences 54 (2-3), 167-
179.

Hill, D., Playford, G. and Woods, J.T. eds. (1969).
‘Ordovician and Silurian fossils of Queensland’.
0.2-0.15, and S.2-S.18. (Queensland
Palaeontographical Society, Brisbane).

Igo, H. and Koike, T. (1967). Ordovician and Silurian
conodonts from the Langkawi Islands, Malaya, Part
I. Geology and Palaeontology of Southeast Asia 3,
1-29.

Lee, H.Y. (1975). Conodonten aus dem unteren
und mittleren Ordovizium von Nordkorea.
Palaeontographica Abteilung A 150, 161-186;
Stuttgart.

Lehnert, O. (1995). Ordovizische Conodonten aus der
Prakordillere Westargentiniens: Ihre Bedeutung
fiir Stratigraphie und Palaogeographie. Erlanger
Geologische Abhandlungen 125, 1-193.

Lindstrém, M. (1955). Conodonts from the lowermost
Ordovician strata of south-central Sweden.
Geologiska Féreningens i Stockholm Férhandlingar
76, 517-604.

Lindstrém, M. (1970). A suprageneric taxonomy of the
conodonts. Lethaia 3, 427-445.

Lindstrém, M. (1971). Lower Ordovician conodonts
of Europe. In “Symposium on conodont

Proc. Linn. Soc. N.S.W., 129, 2008

Y.Y ZHEN AND J. PICKETT

biostratigraphy’ (Eds. W.C. Sweet and S.M.
Bergstr6m). Geological Society of America, Memoir
127, 21-61.

Lofgren, A. (1978). Arenigian and Llanvirnian conodonts
from Jamtland, northern Sweden. Fossils and Strata
13, 1-129.

Léfgren, A. (1985). Early Ordovician conodont
biozonation at Finngrundet, south Bothnian Bay,
Sweden (Geology of the southern Bothnian Sea, Part
Ill). Bulletin of the Geological Institutions of the
University of Uppsala (N.S.) 10, 115-128.

Léfgren, A. (2000). Early to early Middle Ordovician
conodont biostratigraphy of the Gillberga quarry,
northern Oland, Sweden. GFF 122, 321-338.

Léfgren, A. (2003). Conodont faunas with Lenodus
variabilis in the upper Arenigian to lower
Llanvirnian of Sweden. Acta Palaeontologica
Polonica 48 (3), 417-436.

Moskalenko, T.A. (1973). General survey of Ordovician
conodonts of the Siberian Platform: Akademiya
Nauk USSR, Siberian Branch, Trudy Instituta
Geologii i Geofiziki 47, 87—135 (in Russian).

Moskalenko, T.A. (1989). Konodonty verkhney chasti
nizhnego i srednego Ordovika. Trudy Instituta
Geologii i Geofiziki 751, 125-138.

Ni, S. and Li, Z. (1987). Conodonts. Jn Wang X.F., Ni,
S.Z., Zeng, Q.L., Xu, G.H., Zhou, T.M., Li, Z.H.,
Xiang, L.W. and Lai, C.G., ‘Biostratigraphy of
the Yangtze Gorge area 2: Early Palaeozoic Era’.
386-447, 549-555, 619-632. (Geological Publishing
House, Beijing) (in Chinese with English abstract).

Nicoll, R.S. and Kelman, A. (2004). Arrangement of
elements in the Early Ordovician Likmas type
ramiform-ramiform conodont Erraticodon patu
Cooper, 1981: interpretation and implications.
Memoirs of the Association of Australasian
Palaeontologists 30, 207-220.

Nowlan, G.S., McCracken, A.D. and Chatterton, B.D.E.
(1988). Conodonts from Ordovician-Silurian
boundary strata, Whittaker Formation, MacKenzie
Mountains, Northwest Territories. Geological Survey
of Canada, Bulletin 373, 1-99.

Pohler, S.M.L., and Orchard, M.J. (1990). Ordovician
conodont biostratigraphy, western Canadian
Cordillera. Geological Survey of Canada Paper 90
= 115, 1-357.

Pickett, J.W., & Percival, I.G. (2001). Ordovician faunas
and biostratigraphy in the Gunningbland area,
central New South Wales, Australia. Alcheringa 25,
9-52.

Rigby, J.K., & Webby, B.D. (1988). Late Ordovician
sponges from the Malongulli Formation of central
New South Wales, Australia. Palaeontographica
Americana 56, 1-147.

Sergeeva, S.P. (1974). Nekotorye novye konodonty iz
ordovikskikh otlozhenii leningradskoy oblasti [Some
new Ordovician conodonts from the Leningrad
region]. Paleontologicheskiy Sbornik 11 (2), 79-84.

Proc. Linn. Soc. N.S.W., 129, 2008

Serpagli, E. (1974). Lower Ordovician conodonts from
Precordilleran Argentina (Province of San Juan).
Bollettino della Societa Paleontologica Italiana 13,
17-98.

Simpson, C.J., Cas, R.A.F., & Arundell, M.C. (2005).
Volcanic evolution of a long-lived Ordovician
island-are province in the Parkes region of the
Lachlan Fold Belt, southeastern Australia. Australian
Journal of Earth Sciences 52, 863-886.

Smith, R.E. (1966). The geology of Mandurama-Panuara.
Journal and Proceedings of the Royal Society of
New South Wales 98, 239-262.

Stait, K. and Druce, E.C. (1993). Conodonts from the
Lower Ordovician Coolibah Formation, Georgina
Basin, central Australia. BMR Journal of Australian
Geology and Geophysics 13, 293-322.

Stearn, C.W., & Pickett, J.W., 1994. The stromatoporoid
animal revisited: building the skeleton. Lethaia 27,
1-10.

Stearn, C.W., Webby, B.D., Nestor, H., and Stock, C.W.
(1999). Revised classification and terminology of
Palaeozoic stromatoporoids. Acta Palaeontologica
Polonica 44 (1), 1-70.

Stouge, S. (1984). Conodonts of the Middle Ordovician
Table Head Formation, western Newfoundland.
Fossils and Strata 16, 1-145.

Sweet, W.C. (1988). “The Conodonta: Morphology,
Taxonomy, Paleoecology, and Evolutionary
History of a Long-Extinct Animal Phylum’. 212pp.
(Clarendon Press, Oxford).

Trotter, J.A., and Webby, B.D. (1995). Upper Ordovician
conodonts from the Malongulli Formation, Cliefden
Caves area, central New South Wales. AGSO
Journal of Geology and Geophysics 15 (4), 475-499.

van Wamel, W.A. (1974). Conodont biostratigraphy of
the Upper Cambrian and Lower Ordovician of
north-western Oland, south-eastern Sweden. Utrecht
Micropalaeontological Bulletins 10, 1-125.

Viira, V. (1974). “Konodonty Ordovika Pribaltiki
[Ordovician conodonts of the east Baltic]’. 142pp.
(Valgus, Tallinn).

Wang, C.Y., ed. (1993). ‘Conodonts of the Lower Yangtze
Valley - an index to biostratigraphy and organic
metamorphic maturity’. 326pp. (Science Press,
Beijing) (in Chinese with English summary).

Watson, S.T. (1988). Ordovician conodonts from
the Canning Basin (Western Australia)
Palaeontographica, Abteilung A, 203 (4-6), 91-147.

Webby, B.D., 1993. Evolutionary history of Palaeozoic
Labechiida (Stromatoporoidea). Memoir of the
Association of Australasian Palaeontologists 15,
57-67.

Zhang, J.H. (1998a). Conodonts from the Guniutan
Formation (Llanvirnian) in Hubei and Hunan
provinces, south-central China. Stockholm
Contributions in Geology 46, 1-161.

Zhang, J.H. (1998b). The Ordovician conodont genus
Pygodus. In “Proceedings of the Sixth European
Conodont Symposium (ECOS VI)’. (Ed. H.
Szaniawski). Palaeontologia Polonica 58, 87-105.

81

ORDOVICIAN CONODONTS AND SPONGES

Zhao, Z.X., Huang, Z.B., Du, P.D., Zhang, G.Z., Xiao,
J.N. and Tan, Z.J. (2005). New species of the Lower-
Middle Ordovician conodonts from the Tarim Basin
in Xinjiang. Acta Micropalaeontologica Sinica 22,
29-38. (in Chinese with English abstract)

Zhen, Y.Y. and Percival, I.G. (2004a). Middle Ordovician
(Darriwilian) conodonts from allochthonous
limestones in the Oakdale Formation of central New
South Wales, Australia. Alcheringa 28, 77-111.

Zhen, Y. Y. and Percival, I. G. (2004b). Middle Ordovician
(Darriwilian) conodonts from the Weemalla
Formation, south of Orange, New South Wales.
Memoirs of the Association of Australasian
Palaeontologists 30, 153-178.

Zhen, Y.Y. and Percival, I.G. (2006). Late Cambrian-Early
Ordovician conodont faunas from the Koonenberry
Belt of western New South Wales. Memoir of the
Association of Australasian Palaeontologists 32,
267-285.

Zhen, Y.Y. and Webby, B.D. (1995). Upper Ordovician
conodonts from the Cliefden Caves Limestone
Group, central New South Wales, Australia. Courier
Forschungsinstitut Senckenberg 182, 265-305.

Zhen, Y.Y., Liu, J.B. and Percival, I.G. (2007). Revision
of conodont species Erraticodon hexianensis
from the upper part of the Meitan Formation
(Middle Ordovician) of Guizhou, South China.
Palaeontological Research 11 (2), 143-160.

Zhen, Y.Y., Percival, I.G. and Webby, B.D. (2003). Early
Ordovician conodonts from far western New South
Wales, Australia. Records of the Australian Museum
55, 169-220.

Zhen, Y.Y., Webby, B.D. and Barnes, C.R. (1999).

Upper Ordovician conodonts from the Bowan Park
succession, central New South Wales, Australia.
Géobios 32 (1), 73-104.

82

Proc. Linn. Soc. N.S.W., 129, 2008

Emsian (Early Devonian) Tetracorals (Cnidaria) from Grattai
Creek, New South Wales

A.J. WRIGHT

School of Earth and Environmental Sciences, University of Wollongong, Wollongong N.S.W 2522
(tony_wright@uow.edu.au)

Wright, A.J. (2007). Emsian (Early Devonian) tetracorals (Cnidaria) from Grattai Creek, New South
Wales. Proceedings of the Linnean Society of New South Wales 128, 83-96.

The tetracoral species Phillipsastrea scotti sp. nov. and Trapezophyllum grattaiensis sp. nov. are described
from strata assigned to the middle Emsian (nothoperbonus to inversus conodont zones: Early Devonian)
part of the Cunningham Formation at Grattai Creek, west of Mudgee, N.S.W. For comparison with the
former, Phillipsastrea oculoides, from the Early Devonian (late Pragian or early Emsian) Garra Formation
in the Wellington area of N.S.W., is revised on the basis of the type material; new longitudinal thin sections
show indisputable horseshoe dissepiments and trabecular fans in this species.

Manuscript received 8 August 2007, accepted for publication 6 February 2008.

KEYWORDS: Early Devonian, Emsian, Grattai Creek, Phillipsastrea, tetracorals, Trapezophyllum.

INTRODUCTION

The two new tetracoral species described here
were collected by Martin Scott from the Cunningham
Formation on Grattai Creek near Mudgee, N.S.W.
during remapping of the Dubbo 1:250 000
geological sheet (Meakin and Morgan 1999). The
tetracorals and associated tabulate corals, bryozoans,
stromatoporoids, pelmatozoans and comminuted
shelly debris are scattered through a mass flow unit
at about the middle of the Cunningham Formation;
pebbles and cobbles of calcareous, volcanic and
metasedimentary rock types occur at this locality in
the formation, clearly transported from a shallow
water zone into the deepwater environment of the
Hill End Trough (HET). The original source of these
fossils was probably on the Capertee High to the
east, although the direction of transportation of this
fossiliferous debris has not been established.

The Grattai Creek fauna described here is
important as age-diagnostic macrofossils are rare
in strata of the Hill End Trough; biostratigraphic
calibration of the HET sequence has been hampered
by the lack of such fossils, so any new occurrences

are noteworthy. To put this in perspective, Table 1
shows the stratigraphic positions of important faunas
from HET strata. This fauna was first discussed by
Percival (1998); further details of the occurrence were
given by Meakin and Morgan (1999) and Packham
et al. (2001) who provided a full biostratigraphic
discussion. Conodonts identified by Percival in
Packham et al. (2001) from limestone clasts from
the Grattai Creek locality indicated a maximum age
within the nothoperbonus to inversus conodont zones
(middle Emsian, late Early Devonian).

As stated above, this fossiliferous horizon lies
within the Cunningham Formation, the highest
formation of the Hill End Trough sequence. The main
locality is on the southern bank of Grattai Creek;
material has been collected loose or as clasts in the
outcrops. The fossiliferous bed extends for some
hundreds of metres along strike, mainly north but also
south of the creek, with the maximum development of
carbonate clasts in exposures on the southern bank of
the creek. The largest coral specimen measured about
10 cm in maximum dimension. It is likely that much
loose material was washed away by severe floods in
2000, and little coral material is now available.

EARLY DEVONIAN TETRACORALS

Table 1. Early Devonian rock units of the eastern HET and the Limekilns area. Full details were given
by Packham et al. (2001, table 1 and accompanying text). Numbers indicate occurrence of age-diagnostic
fossils as follows: 1, this paper; 2, Packham et al. (2001); 3, Talent and Mawson (1999), Rickards and
Wright (2001), Wright and Haas (1990); 4, Garratt and Wright (1988); 5, Wright (in prep.).

Age

latest Lochkovian?-late
Emsian

HET sequence

Cunningham Fm!” Limekilns Fm?

Limekilns sequence

Merrions Tuff Merrions Tuff

Crudine Group

Lochkovian-Pragian
Lochkovian

Crudine Group
Waterbeach Fm

Waterbeach Fm
Turondale Fm*>

latest Silurian Chesleigh Fm

GEOLOGICAL SETTING OF FAUNAS
DISCUSSED HERE

Palaeogeographic units for the Early Devonian
of this part of central-western N.S.W. were defined by
Packham (1960, 1968). Two shallow water structures,
the Capertee High to the east and the Molong High
to the west were separated by the deepwater Hill
End Trough (HET). Biostratigraphic control of
Devonian strata on the two highs is provided by
the abundant fossils that have been collected from
shallow water strata there, but Devonian fossils are
rare in HET strata. This, combined with the difficulty
of establishing lithostratigraphic correlation between
strata of the highs and the trough, makes fossils in
HET strata most important in establishing correlations
and deciphering the evolution of the region. Therefore
the Grattai Creek locality, in yielding conodonts and
corals, is important as one of the rare fossil localities
in the region.

The fossiliferous sequence at Limekilns, N of
Bathurst, has been crucial in providing biostratigraphic
control for dating the upper part of the HET sequence
(Table 1). The oldest fossils from the Limekilns
area are the Pridoli (Late Silurian) graptolites from
the Chesleigh Formation described in Packham et
al. (2001). The oldest Devonian shelly fauna from
the HET in the Limekilns area is the brachiopod
fauna from the Turondale Formation at the Paling
Yards locality just N of Limekilns (see discussion
by Packham et al. 2001); study of this fauna is in
progress, but Garratt and Wright (1988) asserted that
this fauna is earliest Devonian (Lochkovian). In their
paper describing the Limekilns Formation trilobite
Paciphacops crawfordae (assigned to Echidnops by

[@.2)

-.

Turondale Fm?
Cookman Fm

Cookman Fm
Chesleigh Fm?

Sandford [2002]), Wright and Haas (1990) concluded
the occurrence was Pragian, largely on the basis of
the occurrence of the tentaculite Nowakia sulcata.
However, graptolites from probably low in the
Limekilns Formation were identified by Rickards
and Wright (2001) as the Lochkovian Monograptus
uniformis; they pointed out that this determination was
anomalous in view of the inferred ages of the Paling
Yards shelly fauna from the Turondale Formation
(below) and the Limekilns Formation fauna (above).
The Limekilns Formation, therefore of presumed
Pragian age, is overlain (apparently abruptly) by
limestone (‘the Jesse Limestone’) with a serotinus
conodont fauna (late Emsian), as well as rich coral
(Etheridge 1892; Pedder et al. 1970; Pickett 1972;
Wright, unpublished data), stromatoporoid (Webby
and Zhen 1993), brachiopod (Wright, unpublished
data), trilobite (Chatterton and Wright 1988) and
conodont (Talent and Mawson 1999) faunas.

EMSIAN CORAL FAUNAS IN THE REGION

On the Capertee High, Emsian coral faunas
are known from several localities, especially near
Mudgee. Prolific, largely undescribed coral faunas
occur near Mudgee at Mount Frome (Wright 1966,
1981) and in the Sutchers Creek Formation in
the Queens Pinch belt (Wright 1966, 1979); both
formations have yielded serotinus age conodont
faunas (Pickett 1978; Wright 1981; McCracken
1990; Talent et al. 2000), although the upper beds
of the Mount Frome Limestone are probably Middle
Devonian (Pickett 1978). From the Early Devonian
Garra Formation (Joplin and Culey 1938) on the

Proc. Linn. Soc. N.S.W., 129, 2008

A.J. WRIGHT

Molong High to the west of the Hill End Trough,
Hill (1942b) described the new species Phillipsastrea
oculoides, from probably Pragian or early Emsian
strata (see below). Phillipsastrea also occurs in the
Mount Frome Limestone, and Trapezophyllum occurs
with Phillipsastrea in the Sutchers Creek Formation;
neither genus is yet known from Limekilns.

Elsewhere in the Lachlan’ Fold Belt,
Trapezophyllum occurs with Phillipsastrea in
serotinus age beds of the Taemas Formation (Pedder
et al. 1970) at Wee Jasper, N.S.W, as Phillipsastrea
currani iaspiculensis Pedder, 1970 and Sulcorphyllum
pavimentum Pedder, 1970. The two genera have been
documented from the New England Fold Belt by
various authors including Hill (1942a) and Pedder
(1968).

SYSTEMATIC PALAEONTOLOGY

In all cases, comments are based on examination
of type material as well as original descriptions.
Material studied is from the Australian Museum
(AM F for rock specimens, AM FT for thin sections),
the Mining Museum (MMF) and Museum Victoria.
Terminology is that customarily used for fossil corals
after Hill (1981); in corals with a well-defined pipe of
horseshoes, the internal diameter of the pipe equals
the diameter of the tabularium.

Phylum CNIDARIA Hatschek, 1888
Order TETRACORALLIA Haeckel, 1866
Family PHILLIPSASTRAEIDAE Hill, 1954 (pro
Phillipsastreidae C.F. Roemer, 1883)

Phillipsastrea ad’ Orbigny, 1849
Type species
Astraea (Siderastrea) hennahi Lonsdale, 1840, p. 697;
subsequently designated by Edwards and Haime
1850, p. Ixxi (see Lang, Smith and Thomas 1940,
p. 99).

Diagnosis

“Astreoid, thammnasteriod, pseudocerioid to
partially aphroid coralla. Septa extend variably into
tabularium, showing fusiform tabularial boundary,
while in dissepimentarium they range from being fully
continuous to breaking down into isolated spines.
Horseshoe dissepiments vary from intermittent, as
in the type species, to a continuous pipe completely
surrounding the tabularium’ (after McLean 1994a;
modified after McLean 1989, p. 239).

Proc. Linn. Soc. N.S.W., 129, 2008

Remarks

Asnotedbymany authors, therelationship between
Phillipsastrea and genera such as Medusaephyllum
F.A. Roemer, 1855 and Pachyphyllum Edwards and
Haime, 1850 has been uncertain. Two syntypes of
Medusaephyllum ibergense F.A. Roemer, 1855 (the
type species of Medusaephyllum) were located and
sectioned by Dr Alan Pedder (McLean 1986, p. 445;
1994a, p. 53), but the species is not yet redescribed
and evaluated. Pachyphyllum was discussed by
McLean (1986, 1989) who recognised it as a separate
genus.

The general consensus is that the type species
of Phillipsastrea probably has good development
of the horseshoe pipe but that the horseshoes are
largely obscured by coarse trabeculae (McLean 1986;
McLean 1994a; Sorauf 1998). McLean’s (1994a)
generic diagnosis (see above) asserted that the genus
could include species with a variable development of
horseshoes from intermittent to a continuous pipe.
Sorauf (1998) commented extensively on the matter
of diagnostic characters of this genus, and specifically
stated that the presence of trabecular fans is a more
important generic character than horseshoes.

McLean (1994a, pp. 55-6) also indicated that
some N.S.W Early Devonian taxa described by Hill
and others and originally referred to Phillipsastrea
(currani currani Etheridge, 1892; currani
iaspiculensis Pedder, 1970; maculosa Hill, 1942a;
linearis Hill, 1942a; oculoides Hill, 1942b; speciosa
Chapman, 1914) appeared to be congeneric and could
constitute a new genus.

Phillipsastrea scotti sp. nov.
(Figs la-d, 2a-d)

Material
Holotype MMF 44852a-b-c, paratypes MMF
34136a-d, MMF 44853a-b.

Locality

Grattai Creek, west of Mudgee, N.S.W. Grid
reference 726100E 6382100N, Burrendong 1:50 000
topographic sheet.

Etymology

The species is named for Martin Scott (formerly
of the Geological Survey of N.S.W.), collector of the
original material.

Diagnosis

An astraeoid species of Phillipsastrea with 14-
16 major septa almost reaching to axis; corallites

85

EARLY DEVONIAN TETRACORALS

it wh
% i

Figure la-d. Phillipsastrea scotti sp. nov. a-b, transverse views of holotype MMF 44852a. c, longitudinal
views, holotype MMF 44852b. d, transverse view, paratype MMF 34136c. Bar scales = 5 mm.

86 Proc. Linn. Soc. N.S.W., 129, 2008

A.J. WRIGHT

Figure 2a-d. Phillipsastrea scotti sp. nov. a-c, longitudinal views, holotype MMF 44852b, holotype MMF
44852c, holotype MMF 448852a, respectively; d, longitudinal view, paratype MMF 44853. Bar scales = 5
mm.

separated by weakly depressed zone of flattish
dissepiments outside the horseshoes, seen in transverse
section flanking septa. Horseshoe dissepiments
form a vertical array of small horseshoes flanked on
both sides by small, inclined globose dissepiments;
trabecular fans prominent; pipe diameter 4.5 - 5 mm.
Tabular floors convex, consisting of weakly convex
tabulae, with a gutter where upwardly convex tabellae
and globose dissepiments flank the tabulae.

Proc. Linn. Soc. N.S.W., 129, 2008

Description

Original colony dimensions unknown, but
available material suggests a bun-shaped colony
about 75 mm in size. Growth form astraeoid, with
little continuity of septa between corallites. Corallite
axes from 7.5-13 mm apart, separated by weakly
depressed dissepimental area. Pipe diameter about
4.5-5 mm. 14-16 major septa in mature corallites,
being equally but slightly withdrawn from axis; minor

87

EARLY DEVONIAN TETRACORALS

septa very short, just penetrating into tabularium.
Both orders of septa dilated in typical spindle-form,
but thinner inside and outside tabularium. Tabular
floors elevated, with convex tabulae flanked by axially
inclined tabellae. Pipe well defined in longitudinal
and transverse sections; width of horseshoes about 0.5
mim in longitudinal section; globose dissepiments rare
to present inside horseshoes, but abundant flattened
small sub-horizontal dissepiments occur between
corallites forming a weakly depressed coenosteum,
and are clear in longitudinal view flanking septa.

Remarks

Of the several N.S.W. Early Devonian species
that have been assigned to Phillipsastrea, only P.
oculoides and P. currani iaspiculensis (see below
for comments) should be compared with the Grattai
species. The late Emsian serotinus CZ P. currani
(from the Jesse Limestone Member at Limekilns,
N.S.W.) lacks a pipe of horseshoes, and has been
discussed by Pedder in Pedder et al. (1970) and
McLean (1986, 1989, 1994a, 2005). Two undescribed
late Emsian new species of this genus occur in the
Mount Frome Limestone near Mudgee, and a further
late Emsian new species occurs in the Sutchers Creek
Formation at Queens Pinch near Mudgee (Wright,
unpublished data). The undescribed species from the
Sutchers Creek Formation has a larger pipe and more
major septa than oculoides, and the Mount Frome
Limestone species are distinguished on the basis
of the number and detailed nature of septa, and the
diameter of the pipe. The Mount Frome, Sutchers
Creek and Limekilns occurrences are all from the
serotinus CZ.

The late Emsian Phillipsastrea currani
iaspiculensis Pedder, 1970 from the Taemas Fotmation
at Wee Jasper (N.S.W) resembles oculoides in
having up to 16 short major septa and discrete septal
fragments, but has better developed horseshoes (albeit
irregular), flat tabulae, a smaller tabularium and 17-
21 vepreculate septa of each order.

Phillipsastrea oculoides Hill, 1942b (see Figure
4 herein) from the Garra Formation near Wellington
differs markedly from P. scofti in having: 16-19 septa
of each order, of which the major septa are strongly
dilated and extend only halfway to the axis; septal
fragments in the tabularium, and between corallites
where septa tend to break into longitudinal fragments;
pipe diameter of 4.5-5 mm; corallite axes 9.5-12
mm apart; quite irregular, depressed tabular floors
consisting of incomplete globose tabulae flanked
by some dissepiments just inside the pipe; strongly
depressed dissepiments between tabularia.

The probably Eifelian Bensonastreaea praetor

88

Pedder, 1966 from the Timor Limestone in the New
England Fold Belt is a close relative of Phillipsastrea,
but can be distinguished from P. scoftti on the basis
of several attributes, including its larger tabularium,
complex dissepimentarium, and 17-21 vepreculate
septa of both orders.

Species of this genus from China listed by McLean
(1994a) include: the Middle Devonian ganxiensis
He, 1978 and yanbianiense He, 1978; and the Early
Devonian primitiva Jin and He, 1981; and producta
Jin and He, 1981. Of these, none closely resembles
P. scotti: P. primitiva is aphroid; P. primitiva and P.
yanbianense have 18-21 x 2 septa; and P. ganxiensis
appears to have globose dissepiments inside the
horseshoes.

Phillipsastrea oculoides Hill, 1942b
(Figs 3a-c)

Synonymy

1942b Phillipsastrea oculoides Hill, p. 186, pl. VI,
fig. 9. :

1946 Phillipsastrea oculoides Hill; Basnett and
Colditz, table 1, p. 42.

1965b Phillipsastrea oculoides Hill1942; Strusz, p.
565, pl. 73, fig. 7; text-fig. 22.

1968 Phillipsastrea oculoides Hill; Philip and
Pedder, p. 1033.

1994a Phillipsastrea oculoides Hill 1942c; McLean,
p. 56.

Type material

Only the holotype (Figs 3a-c herein) is known
(formerly SUGD 5281, now AM F 98541); it consists
of three rock pieces and 4 thin sections as follows:
AM FT 7896 (oblique LS: Strusz 1965, fig. 22); AM
FT 12085 (oblique LS not figured); AM FT12696
(oblique TS not figured); 12791 (good TS, figured
by Hill and Strusz: Fig. 3a herein); and three new
longitudinal sections, AM FT 14511-14513 have been
prepared (Figs 3b-c herein). Hill’s original transverse
section was re-figured by Strusz (1965, pl. 73, fig. 7)
so is not illustrated here at that scale, but magnified
views of several corallites are given; these show the
horseshoe dissepiments and trabecular fans (Figs 3b-

Cc).

Locality data

The specimen is from the Garra Formation
(Joplin and Culey 1938, Strusz 1965a); locality
details given by Hill (1942b, p. 186) are: ‘portion
247, parish of Mickety Mulga (sic.), about 6 miles
N.W. of Wellington’. The collectors of this specimen,
Basnett and Colditz (1946, table 1) showed this

Proc. Linn. Soc. N.S.W., 129, 2008

AM FT

9

transverse view

holotype AM F98541. a,

9

=5

89

A.J. WRIGHT

SRS
cae

ATS

SSSA ON
KOC Src (

Figure 3a-c. Phillipsastrea oculoides Hill, 1942b;

longitudinal section of a corallite showing good horseshoes and trabecular

b,

.
>)

12791. b-c, AM FT 14511
fans and c, longitudinal v

mm; bar scale for (c)

ing) of two corallites and coenosteum. Bar scales for (a) and (b)

(draw

iew
5 mm.

°

=)

Proc. Linn. Soc. N.S.W., 129, 2008

EARLY DEVONIAN TETRACORALS

locality (their locality IV) also as portion 247. Strusz
(1965) gave further details as follows: ‘locality MM-
10; (parish of Mickety Mulga, county Gordon [sic.]);
boundary of portions 60 and 247, c. 500 yds. west
of portion 208; grid ref. 1810.9863 (Dubbo); outcrop
extends south from fence (portion 60), 200 yds.
towards road.’ Note that the parish of Micketymulga
lies in County Lincoln, not County Gordon. General
maps of outcrops of limestones of the Devonian
Garra Formation have been provided by Carne and
Jones (1919), Basnett and Colditz (1946) and Strusz
(1965a, 1965b). Furey-Greig (1995) gave 3 locality
citations for P. oculoides (grid references 674781
6408371, 675705 6408611, and 675350 6408 750)
but clearly one locality yielded the unique specimen
of P. oculoides.

From whatis probably the locality near Wellington
that produced the P. oculoides coral fauna, Philip
and Pedder (1968, p. 1033) gave a faunal list which
included the corals Embolophyllum, Tipheophyllum,
Zelolasma gemmiforme and P. oculoides, and
conodonts including Spathognathodus exiguus and
Sp. linearis which, together with Z. gemmiforme,
suggested to them a late Siegenian-early Emsian
age. Philip and Pedder (1968, p. 1033) stated that
this was the youngest Garra fauna known to them,
but probably older than the Cavan Bluff Limestone
at Taemas; by implication, it is younger than the
dehiscens age conodont fauna from the Garra which
they also listed. In current terminology this would
make the Garra occurrence of oculoides late Pragian
or early Emsian (Mawson et al. 1992).

Diagnosis

An astraeoid species of Phillipsastrea with 16-19
short major septa which extend about halfway to axis,
with minor septa just entering tabularium. Trabecular
fans well developed. All septa dilated adjacent to
pipe. Septa tend to break into longitudinal fragments
both between and within tabularia, and generally
do not extend far outside the pipe, but are rarely
continuous between corallites. Pipe diameter of 4.5
- 5 mm. Tabular floors depressed, as are dissepiments
between corallites, which are separated by 9.5-12
mm. Convex to slightly concave, incomplete tabulae
are flanked by tabellae and dissepiments. Horseshoes
continuous, width 0.5 — 1.2 mm in longitudinal
section. Increase apparently non-parricidal, occurring
within coenosteum.

Remarks

This species was described by Hill (1942b) who
illustrated only a transverse section of the holotype and
Strusz (1965b) who reillustrated the transverse section

90

with a sketch of an oblique longitudinal section. In
order to clarify the true nature of the horseshoes and
the tabularium, three new longitudinal sections have
been prepared. The species is not redescribed in full
here.

Breakdown of septa as seen in the tabularium
of P. oculoides is variable in Phillipsastrea species
described by McLean (1994) and Sorauf (1999) but
is not known from other Australian Early Devonian
Phillipsastrea.

Trapezophyllum Etheridge, 1899

Synonymy

1899 Cyathophyllum (Trapezophyllum) Etheridge, p.
31, pl. B, figs 2-4.

1963 Sulcorphyllum Pedder, p. 366, text-figs 2a-b.

1968 Stellatophyllum Spassky in Bulvanker et al., p.
30.

21977 Cystitrapezophyllum Jia in Jia et al., p. 148.

21977 Parasulcorphyllum Jia in Jia et al., p. 149.

Type species
Cyathophyllum elegantulum Dun, 1897, pp. 83-87,
plate III, figs 5-6.

Type material

The type material of this species was incorrectly
stated by Fletcher (1971) to be hand specimen GSV
41717 (Sweet collection specimen number 107,
formerly held by the Geological Survey of Victoria,
now housed in the National Museum, Melbourme,
Victoria), from which thin sections AM.2 and
AM.3805 (from Sweet collection number 101) were
cut. Australian Museum material includes no hand
specimens for these thin sections above, but only 3 thin
sections: AM.3805 (Dun’s LS, the illustrated holotype),
and 2 numbered AM.2 (Etheridge’s longitudinal and
transverse sections). AM.2 (Etheridge’s transverse
section) and longitudinal section (AM.FT14479:
Etheridge’s re-numbered longitudinal section) appear
to be cut from different rocks. It can be stated that
the thin section AM.FT14479 was cut from the rock
NMV 41717. Hill (1981, figs 183a-b-c) figured new
material of elegantulum (UQF31114, UQF54725:
neither could be located on July 2, 2007, pers comm.
Dr AG. Cook) showing globose dissepiments outside
the pipe.

Occurrence

Dun’s type species of the genus was based
on material from Loyola, Victoria (late Pragian:
Cooper1973; Mawson et al. 1992), and was also

Proc. Linn. Soc. N.S.W., 129, 2008

A.J. WRIGHT

described from the Coopers Creek Limestone by
Philip (1965); according to Talent et al. (2000) the
Coopers Creek Limestone spans the sulcatus and
dehiscens conodont zones, and is thus late Pragian to
early Emsian.

Remarks

Hill (1942a) described two new species from
the Sulcor Limestone in the New England Fold Belt
of N.S.W. (late Emsian or younger: Mawson et al.
1985), Trapezophyllum coulteri and Prismatophyllum
brownae. The latter species was chosen as type
species of Sulcorphyllum by Pedder (1963). Pedder
(1968) noted that the former species is from the
lowest and middle of the three faunas he recognised
in the Sulcor Limestone, whereas brownae is from the
middle fauna.

Pedder in Pedder et al. (1970) described a
further species, Sulcorphyllum pavimentum Pedder,
1970, from Wee Jasper, where it occurs with P.
currani iaspiculensis Pedder, 1970; S. pavimentum
is also known from rather highly deformed Emsian
limestone beds in the Capertee Valley, N.S.W (Wright,
unpublished data). Undescribed N.S.W species occur
in the Sutchers Creek Formation in the Mudgee
district and in the Wellington district (Bunny 1962:
probably from the Cunningham Formation).

Sulcorphyllum pavimentum was selected as
type species of Parasulcorphyllum by Jia (1977)
but this genus has been regarded as a junior
synonym of Trapezophyllum by most subsequent
authors. Other new genera established by Chinese
workers, including Neotrapezophyllum Jia & Wang
in Jia, 1977; Cystotrapezophyllum Cai, 1983; and
Heterotrapezophyllum Cai, 1983 are in need of further
evaluation beyond the scope of this paper, but appear
to be junior synonyms of Trapezophyllum. At least 16
species, mostly Late Devonian, have been described
from China, Germany, Russia and Belgium, but none
closely resembles the new Grattai Creek species.

Sulcorphyllum was erected by Pedder (1963)
on the premise that it possessed abundant globose
dissepiments that occurred outside the pipe of
horseshoes; at that time he stated that these are
not developed in Trapezophyllum. Study by the
present author of type, topotype and other material
of 7. elegantulum material has shown that such
globose dissepiments do occur outside the pipe in 7.
elegantulum (see also Hill 1981, p. 284, fig. 183-2c),
so the use of the occurrence of globose dissepiments
outside the pipe in Sulcorphyllum cannot be used
to distinguish the two genera, which are regarded
as synonyms. The same view was expressed by
McLean (1989, p. 242) and Sorauf (1998). However,

Proc. Linn. Soc. N.S.W., 129, 2008

Pedder (2006, p. 52) maintained that the type species
of Trapezophyllum, T. elegantulum, lacks septal
trabeculae and continued to recognise Sulcorphyllum
as a separate genus. I have studied all available
topotype material of this species which certainly
does show mostly very fine trabecular structure but
trabecular fans are definitely seen in material (Wright,
unpublished data) where septal dilation is developed,
so I conclude that Su/corphyllum is a subjective junior
synonym of Trapezophyllum.

I have examined the type material of
Stellatophyllum lateratum Spasskiy, 1968 which
is the type species of this Russian genus. Detailed
remarks will be presented elsewhere, but I can assure
the reader that this material should be referred to
Trapezophyllum.

Trapezophyllum grattaiensis sp. nov.
(Fig. 4a-b)

Material
MME 34138a-c, holotype.

Locality

Grattai Creek, west of Mudgee, N.S.W. Grid
reference 726100E 6382100N, Burrendong 1:50 000
topographic sheet.

Etymology
The species is named after the locality on
Grattai Creek.

Diagnosis

Trapezophyllum with 12-15 long major septa,
reaching to or almost to corallite axis; minor septa
just extend into tabulartum. Pipe of well-formed,
continuous horseshoe dissepiments about 2.5-3 mm in
internal diameter; maximum corallite diameter 6 mm;
outer rank of dissepiments flat, accessory globose
dissepiments outside these absent. Tabulae often
complete and slightly convex upwards, supplemented
by some globose peripheral tabellae.

Description

Corallum cerioid, originally about 100 mm in
size; corallites with 5-7 sides. Wall poorly preserved
and thin. Mode of increase unknown. Mature
corallites up to 6 mm in diagonal dimension. 12-
15 major septa, somewhat wavy and thorny in
tabularium, long and thin, with variable dilation over
horseshoes; extend almost to corallite axis where they
may be in contact. Minor septa about half as long
as majors, and just reach into tabulartum. Dilation

91

EARLY DEVONIAN TETRACORALS

iM
é'

RR paper
* oe

Figure 4a-b. Trapezophyllum grattaiensis sp. nov., holotype. a, MMF 34186b, transverse view ; b, MMF

34186a, longitudinal view. Bar scales = 5 mm.

of septa weakly spindle-form, lesser outside pipe.
Pipe about 2.5-3 mm in diameter, well-defined in
transverse section; horseshoes continuous and well-
formed in longitudinal view, reaching about 0.4 mm
in width in longitudinal view; flat dissepiments about
0.5 mm in longitudinal view, about 16 in 5 mm of
corallite length. Tabulae mostly complete, and gently
convex upwards, supplemented by rare, gently axially
inclined, globose tabellae; no dissepiments inside
pipe. Only flat dissepiments outside the pipe, with no
globose dissepiments.

Remarks

The very distinctive type species, 7. elegantulum,
has been recorded only from Victoria, at the type
locality at Loyola (Dun 1897; Etheridge 1899; Hill
1939) and in Gippsland (Philip 1965). It has sporadic
globose dissepiments outside the pipe; distant, often
complete and sagging to horizontal tabulae; rarely
visible minor septa; and short major septa barely
entering the tabularium. Hill (1939) stated that there
are 10-12 major septa alternating with minor septa;
in material I have studied, the range is normally 12-

13 in mature corallites. Minor septa are very slightly
shorter than major septa, but are generally discernible,
especially where there is skeletal dilation. Tertiary
septa as were reported in the type species by Hill
(1939, p. 235), but these appear to be limited to very
occasional spikes projecting from the wall between
minor septa. The type species differs profoundly from
the Grattai and all other Zrapezophyllum species in
septal length, the nature of the tabulae, possession of
globose dissepiments outside the horseshoe pipe, and
in size. Another diminutive and previously overlooked
species is T. terecktense Spasskty, 1971; this species
will be discussed in detail elsewhere, but it clearly
differs markedly from our new species.

Other Australian species of Trapezophyllum
occur in the Tamworth district (7. coulteri Hill,
1942a and 7. brownae (Hill, 1942a), both from the
Sulcor Limestone (Emsian: Pedder 1968, p. 139), and
T. pavimentum from the Taemas area (Pedder et al.
1970)

Trapezophyllum coulteri Hill, 1942a is
characterised by having: a corallite diameter of
about 4-6 mm and 13-15 septa of each order; pipe

Proc. Linn. Soc. N.S.W., 129, 2008

A.J. WRIGHT

diameter of 2.5-3.5 mm; and an axial space of about
1 mm. The holotype has major septa that are clearly
withdrawn from the axis, and are highly dilated over
the horseshoes; the tabulae consist of a set of convex
axial plates, flanked by convex tabellae; globose
dissepiments are rare. Trapezophyllum grattaiensis
differs in having longer major septa, lesser septal
dilation and less regularly convex tabulae.

The holotype of 7. brownae (SUP 8152, now AM
F 133004) was figured by Hill (1981, figs 1823a-b-c)
for the first time. It differs clearly from T. grattaiensis
in having larger dimensions (maximum corallite
diameter of 6-7 mm, pipe diameter 2.5-3.5) and 15-
18 major septa. The species is further characterised
by the wide zone of globose dissepiments outside
the horseshoes, being markedly sloping downwards.
Hill (1942a, pl. Ill, figs 4a-b) figured the paratype
(formerly SUGD 7246, not 7236 as stated by Hill
1942a, p. 152).

ACKNOWLEDGEMENTS

I am grateful to: Dr Jan Percival and Gary Dargan
(NSWGS) for making the original Gratta1 Creek material
available for study; Martin Scott (formerly NSWGS) for
guiding me to the locality; Professor Barrie Rickards and Dr
Lucy Muir for assisting with field work; Dr Yong Yi Zhen
who facilitated access to Australian Museum material and
allowed preparation of further longitudinal sections of the
holotype of P. oculoides; Dr Rolf Schmidt and Dr David
Holloway who allowed me to borrow Trapezophyllum
material in their keeping at Museum Victoria; and Dr Ross
McLean and Dr Alan Pedder who assisted with literature.
Dr Olga Kossovaya (VSEGEI, St Petersburg) very kindly
assisted with copies of Spasskiy (1968), and Dr Valentina
Stolbova (Geological Museum, Mining Academy, St
Petersburg) facilitated my study of Nikolai Spasskiy’s
type material of Stellatophyllum in the VSEGI Museum in
St Petersburg. David Barnes, Penny Williamson, Dudley
Simon and Yong Yi Zhen gave valuable technical advice
for production of illustrations. Two anonymous reviewers
provided comments that improved the paper. This study has
been undertaken during my tenure of the Linnean Macleay
Fellowship, which I gratefully acknowledge.

REFERENCES

Basnett, E.M. and Colditz, M.J. (1946). General geology
of the Wellington district, N.S.W. Journal and
Proceedings of the Royal Society of New South Wales
79, 37-47.

Besprozvannykh, N.I., Dubatolov, V.N., Kravtsov,

A.G., Latypov, Yu. Ya. and Spassky, N. Ya. (1975).
Devonskie rugozy Taymyro-Kolymskoy provintsii.

Proc. Linn. Soc. N.S.W., 129, 2008

Trudy Instituta Geologii i Geofiziki, Akademiya Nauk
SSSR, Sibirskoe Otdelenie 228, 172 pp. (Russian).

Birenheide, R. (1978). Rugose Korallen des Devon.
Leitfossilien. Begriinder von GEORG GURICH. 2.
vollig neu bearbitete Auflage herausgegeben von
Prof. Dr. KARL KROMMELBEIN, Kiel. Gebriider
Borntraeger, Berlin, Stuttgart. 265 pp. (German).

Bulvanker, E.Z., Goryanov, V.B., Ivanovskiy, A.B.,
Spasskiy, N.Ya., Shchukina, V. Ya. (1968). Novye
predstavilteli chetyrekhluchevykh korallovykh
polipov SSSR. (New representatives of tetraradiate
coral polyps of the USSR) in Markovskiy, B.P. (ed.)
New taxa of fossil plants and invertebrates of the
USSR, volume 2, part 2, pp. 14-45, Nedra, Moscow
(Russian).

Bunny, M. (1962). The geology of the parishes of Namina,
Wellington and Galwadgere, Wellington, N.S.W.
Unpublished BSc (Hons) thesis, University of
Sydney.

Cai Tu-ci and Zeng Cai-lin (1983). Atlas of palaeontology
of NW China, Xinjiang (Sinkiang) volume, Part 2, p.
113-216, 684-704.

Carne, J.E. and Jones, L.J. (1919). The limestone
resources of New South Wales. New South Wales
Geological Survey, Mineral Resources 25, 411 pp.

Chapman, F. (1914). Newer Silurian fossils of eastern
Victoria. Records of the Geological Survey of Victoria
3, 301-316.

Chatterton, B.D.E. and Wright, A.J. (1988). Early
Devonian trilobites from the Jesse Limestone, New
South Wales. Journal of Paleontology 62, 93-103.

Coen-Aubert, M. (1994). Stratigraphie et systématique
des Rugueux de la partie moyenne du Frasnien de
Frasnes-lez-Couvin (Belgique). Bulletin de l'Institut
des Sciences Naturelles de Belgique (Sciences de la
Terre) 64, 21-56. (French).

Cooper, B.J. (1973). Lower Devonian conodonts from
Loyola, Victoria. Proceedings of the Royal Society of
Victoria 86, 77-84.

D’Orbigny, A. (1849). Notes sur des polypiers fossiles. 12
pp. Victor Masson, Paris.

Dun, W-.S. (1897). Contributions to the Palaeontology
of the Upper Silurian rocks of Victoria, based on
specimens in the collections of Mr. George Sweet.
Part 1. Proceedings of the Royal Society of Victoria
10, 79-90.

Edwards, H.M. and Haime, J. (1850). A monograph of
the British fossil corals. Palaeontographical Society
Monograph, i-\xxxy, 1-71.

Etheridge, R. Jr (1892). Descriptions of four Madreporaria
Rugosa — species of the genera Phillipsastraea,
Heliophyllum, and Cyathophyllum — from the
Palaeozoic rocks of N. S. Wales. Geological Survey
of New South Wales, Records 2, 165-174.

Etheridge, R. Jr (1899). Descriptions of new or little-
known Victorian Palaeozoic and Mesozoic fossils,
No. 1. Progress Report, Department of Mines,
Geological Survey of Victoria XI, 30-36.

93

EARLY DEVONIAN TETRACORALS

Fletcher, H.O. (1971). Catalogue of Type Specimens of
Fossils in the Australian Museum, Sydney. Memoir of
the Australian Museum 13, 1-167.

Furey-Greig, T. (1995). Checklist of the known fossil
localities on the Dubbo 1:250,000 sheet area.
Geological Survey of New South Wales, Unpublished
Palaeontological Report 95/01, 174 pp.

Garratt, M.J. and Wright, A.J. (1988). Late Silurian to Early
Devonian biostratigraphy of southeastern Australia.
vol. Ill, 647-662. In N.J. M*‘Millan, A.F. Embry and
D.J. Glass, (eds). Devonian of the World. Memoir of the
Canadian Society of Petroleum Geologists 14.

Haeckel, E. (1866). Generelle Morphologie der
Organismen. 2, Allgemeine Entwickelungsgeschichte
der Organismen. G. Reimer, Berlin.

Hatschek, B. (1888-1891). Lehrbuch der Zoologie, eine
morphologische Ubersicht des Thierreiches zur
Einftihrung in das Stadium dieser Wissenschaft.
Gustav Fischer, Jena, Lief 1-3, 1v + 432 pp.

He Yuan-xiang (1978). Subclass Rugosa, pp. 98-179. In
Chengdu Institute of Geology and Mineral Resources
(ed.) Atlas of fossils of southwest China. Sichuan
volume, Part 1. From Sinian to Devonian. Geological
Publishing House, Beijing, China (Chinese).

He Jin-yan (1988). Rugosa, pp. 178-195. Jn Devonian
stratigraphy, palaeontology and sedimentary facies of
Longmenshan, Sichuan. Chengdu institute of Geology
and Mineral Resources and Institute of Geology,
Chinese Academy of Geological Sciences (editor),
Geological Publishing House, Beijing. 487 pp.

Hill, D. (1939). The Devonian corals of Lilydale and
Loyola, Victoria. Proceedings of the Royal Society of
Victoria 51, 219-256.

Hill, D. (1942a). Devonian rugose corals from the
Tamworth district, N.S.W. Journal and Proceedings
of the Royal Society of New South Wales 76, 142-164.

Hill, D. (1942b). Middle Palaeozoic rugose corals from the
Wellington district, N.S.W. Journal and Proceedings
of the Royal Society of New South Wales 76, 182-189.

Hill, D. (1981). Treatise on Invertebrate Paleontology,
Part F, Coelenterata, Supplement I Rugosa,
Tabulata). Geological Society of America and
University of Kansas, Boulder Colorado and
Lawrence Kansas. i-xl, + pages F1-378 (volume 1)
and pages F379-761 (volume 2).

Ivanovski, A.V. and Shurygina, M.V. (1980). Reviziya
devonskikh rugoz Urala. (Revision of the
Devonian rugose corals of the Urals). Trudy
Paleontologicheskogo Instituta, Akademiya Nauk
SSSR, 186, 64 pp. (Russian).

Jia Hui-zhen (1977). Devonian rugose corals, pp. 112-168.
In Geological Scientific Research Institute of Hubei
(eds) Palaeontological Atlas of central and southern
China, part 2, Late Palaeozoic. Geological Publishing
House, Beijing (Chinese).

94

Jin Shan-yu and He Jin-han (1981). The Devonian rugose
corals of Guangxi, their sequence and systematic
descriptions. In Bai, S.L., Jin S.Y. and Ning, Z.S.
(eds), The Devonian biostratigraphy of Guangxi
and adjacent area, pp. 109-148, 160-165. Peking
University Press, Beijing, China (Chinese).

Joplin, G.A. and Culey, A.G. (1938). The geological
structure and stratigraphy of the Molong-Manildra

district. Journal and Proceedings of the Royal
Society of New South Wales 71, 267-281.

Lakhov, G.V. (1981). Novye vidy kolonial’nykh
devonskikh rugoz Novoy Zemli. (New genera of
colonial rugose corals from the Devonian of Novaya
Zelyia). Zapiski Leningradskogo Gornogo Instituta
85, 65-74. (Russian).

Lang, W.D., Smith, S. and Thomas, H.D. (1940). Index of
Palaeozoic coral genera. 231 pp. British Museum
(Natural History), London.

Lin Bao-yu and others (1995. Monograph of Palaeozoic
corals. Rugosa and Heterocorallia. Geological
Publishing House, Beijing, China. 778 pp (English
summary, pp. 718-756).

Lonsdale, W. (1840). In Sedgwick, A. and Murchison,
R.I., On the physical structure of Devonshire, and on
the subdivisions and geological relations of its older
stratified deposits, etc. Geological Society of London,
Transactions, series 2, 5, 697 pp.

Mawson, R., Jell, J.S. and Talent, J.-A. (1985). Stage
boundaries within the Devonian: implications
for application to Australian boundaries. Courier
Forschungsinstitut Senckenberg 75, 1-16.

Mawson, R. and Talent, J.A. (1994). The Tamworth
Group (mid-Devonian) at Attunga, New South
Wales: conodont data and inferred ages. Courier
Forschungsinstitut Senckenberg 168, 37-59.

Talent, J.A., Mawson, R. and Simpson, A.J. (2000).
Silurian to Early Carboniferous (Tournaisian)
platform-slope sequences in eastern Australia: recent
advances in stratigraphic alignments. Geological
Society of Australia Abstracts 61, 114-120.

Mawson, R., Talent, J.A., Brock, G.A. and Engelbretsen,
M.J. (1992). Conodont data in relation to sequences
about the Pragian-Emsian boundary (Early Devonian)
in south-eastern Australia. Proceedings of the Royal
Society of Victoria 104, 23-56.

McCracken, A. (1990). Stratigraphy, chronology and
regional significance of Devonian allochthonous
blocks and debris flows in the vicinity of Queens
Pinch, near Mudgee, N.S.W. Unpublished BSc
Honours thesis, Macquarie University, Sydney.

McLean, R.A. (1986). The rugose coral Pachyphyllum
Edwards and Haime in the Frasnian of Western
Canada. Current Research, Part B, Geological Survey
of Canada, Paper 86-1B, 443-455.

Proc. Linn. Soc. N.S.W., 129, 2008

A.J. WRIGHT

McLean, R.A. (1989). Phillipsastreidae (Rugosa) in
the Frasnian of western Canada. Association of
Australasian Palaeontologists, Memoir 8, 239-249.

McLean, R.A. (1994a). Frasnian rugose corals of western
Canada. Part 3A: the massive Phillipsastreidae —
Phillipsastrea, Chuanbeiphyllum. Palaeontographica
A, 320, 39-76.

McLean, R.A. (1994b). Frasnian rugose corals of western
Canada. Part 3B: the massive Phillipsastreidae
— Pachyphyllum, Smithicyathus, Frechastrea.
Palaeontographica A, 320, 77-96.

McLean, R.A. (2005). Phillipsastreid corals from the
Frasnian (Upper Devonian) of western Canada:
taxonomy and biostratigraphic significance. NRC
Research Press, Ottawa, Ontario. 109 pp.

McLean, R.A. and Sorauf, J.E. (1989). The distribution of
rugose corals in Frasnian outcrop sequences of North
America, pp. 379-396. In N.J. M°Millan, A.F. Embry
and D.J. Glass, (eds). Devonian of the World. Memoir
of the Canadian Society of Petroleum Geologists 14,
vol. III.

Meakin, S. and Morgan, E.A. (compilers) (1999).
Explanatory Notes, Dubbo Geological Sheet, 1:250
000 S1/55-4. Geological Survey of New South
Wales, Sydney and Australian Geological Survey
Organisation, Canberra. 504 pp.

Packham, G.H. (1960). Sedimentary history of part of
the Tasman Geosyncline in South Eastern Australia.
Reports of the XXI International Geological
Congress, 2, 74-83.

Packham, G.H. (1968). The geology and sedimentary
tectonics of the Hill End-Euchareena district, New
South Wales. Proceedings of the Linnean Society of
New South Wales 93, 111-163.

Packham, G.H., Percival, I.G., Rickards, R.B. and Wright,
A.J. (2001). Late Silurian and Early Devonian
biostratigraphy in the Hill End Trough and the
Limekilns area, New South Wales. Alcheringa 25,
251-261.

Pedder, A.E.H. (1963). Two new genera of Devonian
tetracorals from Australia. Proceedings of the
Linnean Society of New South Wales 88, 364-367.

Pedder, A.E.H. (1966). The Devonian tetracoral
Haplothecia and new Australian phacellophyllids.
Proceedings of the Linnean Society of New South
Wales 90, 181-189.

Pedder, A.E.H. (1968). The Devonian System of New
England, New South Wales, Australia. D.H. Oswald
ed., International Symposium on the Devonian
System, Calgary, 1968, volume 2, 135-142.

Pedder, A.E.H. (2006). Zoogeographic data from studies
of Paleozoic corals from the Alexander terrane,
southeastern Alaska and British Columbia. In
J.W. Haggart, R.J. Enkin and J.W.H. Monger,
eds, Palaeogeography of the North American
Cordillera: Evidence For and Against Large-Scale
Displacements. Geological Association of Canada,
Special Paper 46, pp. 1-48; Appendix A, Systematic

Proc. Linn. Soc. N.S.W., 129, 2008

Paleontology, pp. 48-57; Appendix C, Zoogeographic
data from studies of Paleozoic corals from the
Alexander terrane, southeastern Alaska and British
Columbia: Biogeographic Database, pages numbered
1-45.

Pedder, A.E.H., Jackson, J.H. and Philip, G.M. (1970).
Lower Devonian biostratigraphy in the Wee Jasper
region of New South Wales. Journal of Paleontology
44, 206-251.

Percival, I.G. (1998). Early Devonian (Pragian-

Emsian) faunas from the Cunningham Formation.
Unpublished Palaeontological Report 98/02,
Geological Survey of New South Wales, 5 pp.

Philip, G.M. (1965). The palaeontology and stratigraphy
of the Siluro-Devonian sediments of the Tyers area,
Gippsland, Victoria. Proceedings of the Royal Society
of Victoria 75, 123-246.

Philip, G.M. and Pedder, A.E.H. (1968). Stratigraphical
correlation of the principal Devonian limestone
sequences of Eastern Australia. D.H. Oswald ed.,
International Symposium on the Devonian System,
Calgary, 1968, volume 2, 1025-1041.

Pickett, J.W. (1967). Untersuchungen der Familie
Philltpsastreidae (Zoantharia rugosa).
Senckenbergiana lethaea 48, 1-89.

Pickett, J.W. (1972). Correlation of the Middle Devonian
formations of Australia. Bulletin of the Geological
Society of Australia 18, 457-66.

Pickett, J.W. (1978). Conodont faunas from the Mount
Frome Limestone (Emsian-Eifelian), New South
Wales. Bulletin of the Bureau of Mineral Resources,
Geology and Geophysics (Australia) 192, 97-107.

Rickards, R.B. and Wright, A.J. (2001). Early Devonian
graptolites from Limekilns, New South Wales.
Records of the Western Australian Museum,
Supplement, 58, 122-131.

Roemer, C.F. (1883). Lethaea geognostica, Theil 1:
Lethaea palaeozoica: Leif. 2, pp. 113-544, E.
Scweizerbart’sche, Stuttgart.

Roemer, F.A. (1855). Beitrage zur geologischen Kenntnis
des nordwestlichen Harzegebirges, Dritte abtheilung.
Palaeontographica 5, i-1v, 44 pp.

Sandford, A.C. (2002). Systematics, biostratigraphy and
palaeoenvironments of Echidnops, a new genus of
trilobite from the Late Silurian — Early Devonian of
south-eastern Australia. Memoirs of the Association
of Australasian Palaeontologists 27, 1-31.

Sorauf, J.E. (1972). Middle Devonian coral faunas from
Washington and Oregon. Journal of Paleontology 46,
426-439.

Sorauf, J.E. (1998). Frasnian (Upper Devonian)
rugose corals from the Lime Creek and Shell
Rock Formations of Iowa. Bulletins of American
Paleontology 113 (355), 159 pp.

Soshkina, E.D. (1949). Devonskie korally Rugosa Urala.
(Late Devonian rugose corals from the Urals).
Akademiya Nauk SSSR, Trudy Palaeontologiceskogo
Instituta 15 (4), 162 pp.

25

EARLY DEVONIAN TETRACORALS

Soshkina, E.D. (1951). Pozdnedevonskie korally Rugosa
Urala, ikh sistematika i evolutsitya. (Late Devonian
rugose corals of the Urals, their systematics
and evolution). Akademiya Nauk SSSR, Trudy
Paleontologicheskogo Instituta 34, 124 pp. (Russian).

Spasskiy, N. Ya (1968). Zakonomemosti prostranstvenno-
vremennogo rasprostranenija rodov i vodov.

(Na primere cetrechlucevych korallov Devona).
(Regularity in the space-time distribution of genera
and species [exemplified by Devonian tetraradiate
corals]). Ezeghodnik vses paleontologiceskii
Obschestvo 18, 3-14 (Russian).

Spasskiy, N. Ya in Dubatolov, V.N. and Spasskiy,

N. Ya (1971). Devonian corals of the Dzhungaro-
Balshaschkoi province. Trudy Instituta Geologii i
Geofizikii, Siberskoi Otdelenie, Akademiya Nauk SSR
74, 5-40, 74-109 (Russian).

Spasskiy, N. Ya (1977). Devonskie rugozy SSR. (Devonian
rugosa of the USSR). Isdatelstvo Leningradskogo
Universiteta, Leningrad. 344 pp. (Russian).

Strusz, D.L. (1965a). A note on the stratigraphy of the
Devonian Garra Beds of New South Wales. Journal
and Proceedings of the Royal Society of New South
Wales 98, 85-90.

Strusz, D.L. (1965b). Disphyllidae and Phacellophyllidae
from the Devonian Garra Formation of New South
Wales. Palaeontology 8, 518-571.

Talent, J.A. and Mawson, R. (1999). North-Eastern
Molong Arch and Adjacent Hill End Trough
(Eastern Australia): Mid-Palaeozoic conodont data
and implications. Abhandlungen der Geologischen
Bundesanstalt 54, 49-105.

Webby, B.D. and Zhen, Yong Yi (1993). Lower Devonian
stromatoproids from the Jesse Limestone of the
Limekilns area, New South Wales. Alcheringa 17,
327-352.

Wright, A.J. (1966). Studies in the Devonian of the
Mudgee district, N.S.W. Unpublished PhD thesis,
University of Sydney.

Wright, A.J. (1979). A new Early Devonian solitary
“cystimorph’. Alcheringa 3, 135-40.

Wright, A.J. (1981). A new phillipsastraeimid tetracoral from
the Devonian of New South Wales. Palaeontology 24,
589-608.

Wright, A.J. and Chatterton B.D.E. (1988). Early Devonian
trilobites from the Jesse Limestone, New South Wales.
Journal of Paleontology 62, 93-103.

Wright, A.J. and Haas, W. (1990). A new Early Devonian
spinose phacopid trilobite from Limekilns, New
South Wales: morphology, affinities, taphonomy and
palaeoenvironment. Records of the Australian Museum
42, 137-47.

Yoh. S.S. (1937). Die Korallenfauna des Mitteldevons aus
der Provinz Kuangsi, Stid-China. Palaeontographica
A 87, 45-76.

Yu Changmin and Liao Weihua (1978). Middle Devonian
rugose corals of Longdongshui Member, Houershan
Formation from Dushan district, Guizhou. Memoirs
of the Nanjing Institute of Geology and Palaeontology
12, 107-145 (English summary 145-146).

96 Proc. Linn. Soc. N.S.W., 129, 2008

A 38,000 year History of the Vegetation at Penrith Lakes, New
South Wales

JANE M. CHALSON! AND HELENE A. Martin?

'46 Kilmarnoch St. Engadine, N.S.W. 2233

? School of Biological, Environmental and Earth Sciences, University of New South Wales, Sydney Australia

2052 (h.martin@unsw.edu.au)

Chalson, J.M. and Martin, H.A. (2008). A 38,000 year history of the vegetation at Penrith Lakes, New
South Wales. Proceedings of the Linnean Society of New South Wales 129, 97-111.

Sediments in an abandoned river channel on the flood plain of the Nepean River at Penrith record about
38,000 calibrated years (38 k cal. yr BP) of deposition. Sections of sediments of a 860 cm core proved
barren of pollen, but sufficient pollen was recovered from three sections aged about (1) 38-36 k cal. yr BP,
middle glacial period, (2) 27-16 k cal. yr BP, middle-late glacial period, including the last glacial maximum
and (3) 6 k cal. yr BP to present, late Holocene.

During the 38-36 k cal. yr BP period, the vegetation was an open sclerophyll forest with Eucalyptus
viminalis and Leptospermum polygalifolium prominent. A ‘spineless Asteraceae’, thought to be Cassinia
arcuata was prominent in the understorey. E. viminalis was the most common eucalypt and it is the most
cold-tolerant of the suite of possible eucalypts. During the 27-16 k cal. yr BP period, a shrubland of Cassinia
arcuata with some grasses was present. The lack of eucalypts during the height of the last glacial period
suggests a cold, arid climate and agrees with estimates that the rainfall was about half that of today. In the
period 6 k cal. yr BP to present, a Eucalyptus tereticornis and Leptospermum juniperinum woodland with a
grassey understorey occupied the site.

When compared with other records in the Sydney Basin, the vegetation through the last glacial maximum

at Penrith Lakes is the only one with a shrubland/grassland community.

Manuscript received 26 February 2007, accepted for publication 24 October 2007

Key words: Climate change, History of the vegetation, Holocene, Last glacial maximum, Palynology,

Penrith.

INTRODUCTION

Penrith, situated on the Cumberland Plain, is
just east of the Lapstone Monocline which defines the
eastern edge of the Blue Mountains Plateau (Bembrick
et al, 1980). At Penrith (Figs 1, 2), the Nepean River
flows from a confining Triassic sandstone gorge onto
shale lowlands where sediments have accumulated
since Tertiary times. The Nepean and the Wollondilly
Rivers together drain much of the southern part of the
Sydney Basin. In the late Pleistocene, it transported
abundant gravels over a braided plain. Nanson and
Young (1988) use this evidence of exceptional fluvial
activity to argue for a pluvial period which ended
about 40-45,000 years ago. The river quickly became
confined to two stable channels, but the easternmost
channel was abandoned about 34-37,000 years ago,
leaving only the western channel, the present course
of the Nepean River (Nanson and Young 1988).

The gravels have been extracted for building
aggregate and the excavations converted into
the Penrith Lakes for recreation (Penrith Lakes
Development Corporation, 1983/84). A core through
the sediments filling the abandoned channel has
been used for this study, the base dating from 38 k
cal. yr BP. This time span includes the last glacial
maximum at about 20-18 k yr ago. There are three
other histories of the vegetation extending back to the
last glacial maximum in the Sydney Basin: (1) Lake
Baraba, one of the Thirlmere Lakes in a confined
sandstone gorge (Black et al., 2006), (2) Mountain
Lagoon, at 500 m altitude in the Blue Mountains
(Robbie and Martin, 2007) and (3) Redhead Lagoon,
a now coastal location south of Newcastle (Williams
et al., 2006). These studies come from very different
environments to that of Penrith Lakes, hence this
study will add significantly to our understanding of
the history of the vegetation of the Sydney Basin.

VEGETATION HISTORY OF PENRITH LAKES, NSW

450° 30° 451° 151° 30°

Redhead Lagoon}:

+ Mountain Lagoon

e. Richmond
Fig. 2

= Penrith

e@ Camden

+ Lake Baraba

Wollongong:

151° 30

Figure 1. Locality map showing place names men-
tioned in the text.

THE ENVIRONMENT

The Penrith Lakes site is located on a river
terrace of the Nepean River, approximately 4.5 km
north of Penrith and 1 km west of Cranebrook Village
(Fig. 2), at 33° 42’ S and 150° 41’ E, and an altitude
of 17-19 m asl. The site was a swamp overlying the
black clay of the channel fill. The channel cut into
the Cranebrook Terrace sediments, which overlie
the Ashfield Shale. To the east, is another older and
higher terrace of Tertiary origin (Fig. 2).

Evidence of alluvial deposition along the Nepean
River extends well back into Tertiary time. Following
the Tertiary deposition, the river excavated a broad
trench running parallel to the Lapstone monocline,
where the Quaternary alluvium of the Cranebrook
Terrace is inset (Fig. 3). The thick basal gravels were
deposited almost contemporaneously with a sandy
clay overburden until the river became confined to
two stable channels. Since the easternmost channel
was abandoned, the Nepean River appears to have
only occupied the western channel and the abandoned
channel filled with fine sediments (Nanson et al.
1987). A bedrock bar at the Castlereagh Neck (Fig.
2) isolated the river from eustatic changes and the

98

Nepean River has left no other significant alluvial
deposits since the last glacial maximum. However,
stripping and replacement of overburden appears to
have occurred over the western part of the terrace
about 14 k yr BP (Fig. 3).

Before gravel extraction commenced, the swamp
collected runoff from the west and southwest, draining
the entire region between the levees of the Nepean
River. Water gradually passed through the swamp and
eventually entered a small tributary of Cranebrook
Creek to the north of the swamp. The swamp acted as
a sump during times of low runoff but as a drainage
channel during periods of higher runoff (Chalson
1991).

MOUNTAINS

4
0)
2
jae
S
o
a1
®
rz

Reworked overburden: Late Pleistocene-
Holocene. c10,000-13,000 yr BP

Original overburden: Pleistocene.
c40,000-45,000 yr. BP

Channel infill: Pleistocene-Recent :
€36,000 yr BP

Fig. 2. The Cranebrook Terrace, showing the al-
Juvial formations and the location of the core site.
The cross section A-B is shown in Fig. 3. Modified
from Nanson et al. (1987)

Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

West

Nepean River

a
13,000 yr BP
=L Yee yt BF

Metres (AHD)

EZ | Sandy clay: reworked overburden

NI fe 33
Sandy-clay: original overburden

| aia Gravels
== Channel infill

Figure 3. Cross section of the Cranebrook Terrace, showing the alluvial units. For location of cross sec-

tion, see Fig. 2. Modified from Nanson et al. (1987)

The soils of the Cranebrook Terrace are weakly
differentiated on the alluvium of the western side
near the river, where the sediments were deposited
9,000-12,000 years BP. On the eastern side, where the
sediments are some 38,000 years old, deep weathering
has produced strongly differentiated profiles. The soils
are red and yellow podzolics with complex variability
(Young et al. 1987).

The closest currently operating meteorological
station is at Richmond on the University of Western
Sydney, Hawkesbury campus, some 20 km NNE
of Penrith and at a similar altitude. Here, the mean
annual rainfall is 800 mm pa, with January to
February the wettest months with an average of 89-
96 mm per month and July to September the driest
months, with an average of 43-47 mm per month. The
mean daily maximum annual temperature is 23.9°C,
with a mean daily maximum of 28.9-29.4°C for the
hottest months of January-February. The mean daily
minimum annual temperature is 10.5°C, with a mean
daily minimum of 3.2-4.4°C for the coldest months of
July-August (BoM 2006).

A meteorological station at Penrith, not operating
now, recorded a long term mean annual rainfall of 685
mm pa (Bureau of Meterology 1966) and a short term
record at Penrith Lakes, 687 mm pa (Chalson 1991).

A survey of the vegetation in the Penrith area
(Benson, 1992) found that very little of the natural
vegetation remains because of the suitability of the
soils for agriculture. The vegetation patterns relate
strongly to the underlying geology with major
groups of communities being restricted to either
the Wianamatta Shale, Tertiary alluvium, Holocene
(and other Quaternary) alluvium or Hawkesbury
Sandstone (Benson, 1992). The swamp surface had
the most significant natural vegetation remaining in
the area (Chalson 1991).

Proc. Linn. Soc. N.S.W., 129, 2008

Melaleuca linariifolia tall shrubland with
Eleocharis _sphacelata, Typha orientalis and
Philydrum lanuginosum covered the swamp, with
Triglochin procerum in shallow standing water.
Juncus usitatus and Persicaria spp. were common in
waterlogged areas (Benson 1992). The southern end
of the swamp was almost completely covered with
Carex appressa and occasional M. Jinariifolia, with
Melaleuca styphelioides in deeper water. Midway
along the swamp, ™. linariifolia was associated with
Paspalum distichum, C appressa, T. orientalis and T.
procerum. Toward the northern part of the swamp,
P. distichum and E. sphacelata were dominant, with
C. appressa and J. usitatus in the marginal areas. M.
linariifolia is still found at the northern extreme of the
swamp (Chalson 1991).

Only small patches of dryland vegetation
remained on the river flats, and from the flora of
these patches, together with early botanists’ reports,
some idea of the original vegetation may be achieved
(Benson 1992). Casuarina cunninghamiana was
found in these remnant sites, and would have
fringed the river. Acacia spp., Bossaiea rhombifolia,
Pultenea flexilis and Kennedia rubicunda may have
been present in the understorey. The floodplain once
supported a river flat open forest with Eucalyptus
amplifolia and Angophora subvelutina dominant.
A remnant of open forest east of the swamp had E.
amplifolia, E. baueriana, E. eugenioides and E.
moluccana (Chalson 1991). Eucalyptus tereticornis
was predominant downstream around Richmond. The
Tertiary alluvium supported remnants of E. fibrosa
open forest and the Ashfield Shale supported E. crebra
and Syncarpia glomulifera. Understorey species found
in these remnants were Bursaria spinosa, Themeda
australis, Aristida ramosa, Daviesia ulicifolia and
Grevillea juniperina (Benson, 1992).

99

VEGETATION HISTORY OF PENRITH LAKES, NSW

For thousands of years, the Penrith Lakes area
was extensively used by the original inhabitants,
the Darug people (Penrith Lakes Development
Corporation 2006). Most of the preserved sites are
surface middens on the river terraces and are probably
less than 3,000 years old. Stockton and Holland
(1974) found stone implements in the gravels at 9 m
depth and obtained radiocarbon dates of 27,000 BP,
but Nanson et al. (1987) regard this as contamination
since all of the gravels were deposited by 47,000-
45,000 BP. However, stone artifacts have been found
in the tumble at the foot of the quarry, and since all
of the overburden was removed before quarrying the
gravels, the possibility of contamination is remote.
The wear on the artifacts suggest that they were
dropped close to the site where they were found
(Nanson et al. 1987).

Excavation of the Shaws Creek KII rockshelter
(Kohen et al. 1984) has revealed more than 13,000
years of occupation. This site, close to both mountane,
riverine and plain environments would have enabled
access to an abundance and variety of plants and
animals. The gravels and boulders in the bed of the
river would have supplied a variety of rock types for
stone implements. There is an older phase of relatively
Sparse occupation and a younger phase of seemingly
more intense occupation associated with a change in
stone tool technology about 4,000 BP (Kohen et al.
1984).

Europeans settled in the region shortly before
1800 AD and settlement accelerated between 1801
and 1806. Initially, settlement focused on timber
getting and by 1810, the cedar and rose mahogany
had been cleared. Subsequently, the settlers began
to grow wheat from around 1801 until the 1820’s
when this was replaced by grazing, probably due to
falling fertility. In the late 1800’s, market gardens and
dairying developed to cater for the Sydney market
(Chalson 1991). The entire site has been extensively
altered since 1991 by the extraction of building
aggregate and subsequent construction of the Sydney
Olympic Games rowing course. Very little evidence
of the original clay channel remains today.

METHODS

The core site was located where it was thought
that the sediments of the abandoned channel would
be the deepest. The top 20 cm was a dense fibrous
mat that had to be trenched. From 20 cm to 50 cm,
the sediments were hand-cored, using a Russian D-
corer (Birks and Birks 1980). From 75 cm to 860 cm,
a continuous core of 50 mm diameter was obtained

100

using a drill truck kindly provided by the Penrith
Lakes Development Corporation. Samples for pollen
analysis were taken at 10 cm intervals. The uppermost
sample for radiocarbon dating was taken from the
trench and the other samples were taken from the core
at postulated zone boundaries and at the base of the
core. A kilogram or more of the clay was required for
each radiocarbon sample.

Very little of the natural vegetation remains
around Penrith (Benson 1992), hence it was not
surveyed. An initial survey of surface samples from
degraded remnant stands showed that the pollen was
poorly preserved and contained over 50 % Poaceae,
reflecting the disturbed nature of the vegetation.
Hence the present day pollen deposition was unlikely
to assist in interpretation of the core pollen spectra,
however a study of pollen deposition in natural
vegetation across the Blue Mountains (Chalson 1991)
may give some insight for interpretation, although the
environments of these sites are somewhat different to
the Penrith site.

Pollen preparations extracted from the core
sediments were spiked with of a known concentration
of Alnus, then treated with hydrochloric and
hydrofluoric acids to remove siliceous material (Birks
and Birks 1980), oxidised with Schultz solution (a
saturated solution of potassium perchlorate in nitric
acid), cleared in 10% potassium carbonate and the
residue was mounted in glycerine jelly (Brown,
1960). Reference pollen was treated with standard
acetolysis (Moore et al. 1991).

Pollen was identified by comparing grains from
the core with reference pollen. Special attention was
paid to pollen of the Myrtaceae which was identified
following the method in Chalson and Martin
(1995). Poaceae was extremely abundant in some
of the samples and several different types could be
recognised (Appendix 1), although they could not be
identified with any taxon within the family.

Grains were counted along transects across the
slides and tests showed that a count of more than 140
grains adequately sampled the residues. The counts
were presented as percentages of the total count and
pollen concentrations were calculated for the most
important taxa.

The abundance of charcoal retained after
sieving was estimated subjectively on scale of | to
8. Counts of microscopic charcoal for a swamp at
Kings Tableland showed that the two methods gave
similar results, although the latter was more variable
(Chalson 1991).

Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

Table 1 The sediments. The Troels-Smith method of description (Birks and Birks 1980) has been

Olive grey, greyish brown, yellowish brown

Dark grey, greyish brown, dark brown, pale brown

Grey, dark grey, greyish brown, yellowish brown, with a few

Mottled grey, light grey, brown, pale brown, yellowish brown
Grey, light grey, brown, pale brown, yellowish brown, olive

brown, with a few mottled bands

followed.
Depth (cm) Sediment type Colour
Trench
0-15 Rooty peat
16-20 Peaty clay Brown/dark brown
Hand core
21-50 Clay
51-74 No core recovery
Drill core
75-90 Clay As for 21-50 cm
91-280 Clay
mottled bands
281-570 Clay
571-660 Clay
661-760 Clay
761-860 Silt, silty clay

RESULTS

The core revealed some 20 cm of rooty peat at
the top, then clay down to 760 cm, and finally silt and
silty clay down to the base of the core at 860 cm (Table
1). The colour of the clay is predominantly grey and
greyish brown, with some dark grey and yellowish
brown colours. There are minor bands of mottling
below 91 cm and consistent mottling between 281
cm and 570 cm. The radiocarbon dates are presented
in Table 2 and show that the record extends back
approximately 38,000 calibrated years BP (38 k cal.
yr BP).

Clay usually has a lower pollen content than
peat, having been deposited in a lake where pollen

Brown, yellowish brown, olive grey.

Greyish brown, dark grey, grey, dark yellowish brown

must be transported to the site, whereas with peat, the
plants growing on site contribute pollen directly into
the sediments. Mottling indicates a fluctuating water
table which is destructive to pollen. There are several
sections in the core which failed to yield workable
pollen spectra (Fig. 4). Nevertheless, sufficient
pollen has been recovered to provide a history of the
vegetation for certain periods and to illustrate changes
in the vegetation over time. ‘

In sediments such as these, the possibility of
differential pollen destruction must be addressed.
Cyperaceae and Poaceae are thin-walled and fragile
grains, and may be expected to be destroyed first. The
pollen spectra from the clays have a proportion of
these fragile grains and are thus are no different from

Table 2. Radiocarbon dates (standard C14 technique) on bulk samples (see Methods). Calibrated years
have been calculated according to the Radiocarbon Calibrated Program Calib Rev 5.0.2 (Stuiver and

Reimer, 1985-2005).

Laboratory Age (radiocarbon years) Calibrated years BP
Depth (cm)
number (yr BP) (cal. yr BP)
40-45 SUA-2489 280 + 50 1,650
410-440 SUA-2349 11,140 + 200 11,150
795-830 SUA-2490 33,500 + 700 39,100
830-857 SUA-2350 32,000 + 500 37,600

Proc. Linn. Soc. N.S.W., 129, 2008

101

VEGETATION HISTORY OF PENRITH LAKES, NSW

©
)
Bg ns
§ eh = S & #
& —~ @ @
S & 9 © § 2 y % S
Sy A
as 2 §F gFE§ €e Se" Se
So & GF ~€ SS ¥ e & y s IS
SS & ££ Ses g PS G @ 8 es N
x oO @ NES >
So 4 & ESLL ESS eee se se
0 TK
SOXSK
20 \ A1
280 +50 Fe 4 F A2
60
NO CORE RECOVERY
80 - =
a ks é [ :
120 e
140 |v vv |
B
160 BARREN OF POLLEN
480 —
200 - | | | | | | |
410-440 cm
41,1404200 BARREN OF POLLEN

Total pollen sum = Total pollen and spores counted

i-—t a}
SCALE Pollen percentages 20% SCALE Pollen concentrations 10 * grains/ec

Figure 4 (above and opposite). Pollen diagram. The upper solid line represents percentages and the lower
broken line the pollen concentrations. For species included in the pollen type name, see Appendix 3.
‘Subjective scale for macroscopic charcoal, the higher the number the more charcoal: see Methods.

102 Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

ee Vv % 6
® v ©
e Y g © <4 @
CS ES Ses aS
e e sa isa w & w Macroscopic Total pollen concentration x10* grains/cc
s g S g ge < ge charcoal
= Bo ge Le ao oe SE EON La ACRE Gree oer 6 ee
i ee ee eee Eee Ee ee eee
ees = ee a fe ==
OMe gigas on ey Sea Al
4h) -== SSE aes
i= F A2

NO CORE RECOVERY

80

100
REE
Bessoesd
Besse

120 sSeecens Peat

BARREN OF POLLEN

res

BARREN OF POLLEN nen Silt, silty clay

Lie | Mottling

BARREN OF POLLEN

Total pollen sum = Total pollen and spores counted

3: t_J
SCALE Pollen percentages 20% SCALE Pollen concentrations 10 * grains/cc

Proc. Linn. Soc. N.S.W., 129, 2008 103

VEGETATION HISTORY OF PENRITH LAKES, NSW

the spectra from well preserved sediments.

The pollen spectra are presented in Fig. 4.
and the taxa represented by the name on the pollen
diagram are found in Appendix 3. Percentages for
all taxa identified and pollen concentrations for the
most abundant taxa are shown. The total pollen
concentration is high at the top, then low through
the clay with some high concentrations in the basal
silt and silty clay. When total pollen concentrations
are high, concentrations of individual taxa generally
parallel percentages, but when the total is low,
individual concentrations remain low, even though
percentages may be high, a reflection of the relative
nature of the percentages.

The pollen spectra have been zoned as follows
(Fig. 4), and an age model has been deduced from Fig.
5, assuming a uniform rate of sediment deposition.
The sediments are relatively uniform throughout the
core and accumulation has probably been similar to
today, where the abandoned channel acts as a drainage
sump.

Depth (cm)
0

Zone A

Zone B

Zone D

Open sclerophyil forest E. viminalis

Zone D, 860-790 cm, c.38 to 36 cal. yr BP (see Fig.
S):

Total pollen concentration is low but increases
towards the top of the zone. Eucalyptus viminalis
and Leptospermum polygalifolium are prominent
here and the Casuarina content is low. There are
moderate amounts of Asteraceae/Tubuliflorae and
the form species Tubuliflorites pleistocenicus, which
probably represents the shrubby Cassinia arcuata
(see Appendix 2). Other sclerophyllous shrubs, e.g.
Acacia, Haloragaceae and Monotoca are present
also. Poaceae types 3 and 6 are present, with minor
amounts of Cyperaceae towards the top of the zone.
There is also a large amount of charcoal.

790-650 cm, c. 36.5 to 27 k cal. yr BP, barren of
pollen.

Zone C, 650-500 cm, c. 27 tol6 k cal. yr BP.
Total pollen concentration is low throughout the
zone. Myrtaceae pollen is too degraded for specific

VEG ET, ATION Sediments

Woodland/grassiand. E. teriticornis
7

Woodland/grassland. E. teriticornis

wese
Oran

Figure 5. Summary diagram of the history at Penrith Lakes. This model assumes continuous deposition
(see text). For legend of sedimentary symbols, see Fig. 4. Radiocarbon dates are shown here, and for

calibrated dates, (crosses), see Table 2.

104

Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

identification, and there is far less of it than in the
zone below. The shrubby 7. pleistocenicus is the
most prominent pollen type, there are some Poaceae,
especially type 6, and Cyperaceae is more abundant
than in the zone below. Podocarpus is usually
present and has the most consistent representation
for the profile. The vegetation of this zone, which
includes the last glacial maximum, would have been
predominantly shrublands with few Eucalyptus spp.
Charcoal content is low, with the exception of one
higher value.

500-220 cm, c. 16 to 6 k cal. yr BP, barren of pollen.
Zone B, 220-75 cm, c. 6 to 2.2 k cal. yr BP.

Total pollen concentration is low in the lower part
of the zone, but increases towards the top. Eucalyptus
tereticornis and Leptospermum juniperinum are
present here and there is an increase in Casuarina,
although it is not large. The shrubby T. pleistocenicus
is completely lacking, Poaceae types 2, 3 and 5 are
more abundant, but type 6 is not present. Cyperaceae
and trilete fern spores increase towards the top of the
zone. The ‘other tricolporate grains’ group, probably
representing herbs and shrubs, is consistently present
in the upper part or the zone. Most of the charcoal
values are low.

75-55cm, ¢. 2.2 to 2 k cal. yr BP, no core recovery.
Zone A, 55 cm to surface, c. 2 to 0 k cal. yr BP.

Zone A has high pollen concentrations, especially
in the peat, and is subdivided into two sub-zones, A2
and Al. The major contributors to A2 sub-zone are
Myrtaceae, Poaceae, Cyperaceae and the trilete spore
group. Unfortunately, the poor preservation does not
allow identification of the Eucalyptus species in the
older Subzone A2.

In Subzone Al, Eucalyptus tereticornis and
Leptospermum juniperinum are prominant. The
other tricolporate grains group (shrubs and herbs)
and Asteraceae/Tubuliflorae type have increased.
Asteraceae/Liguliflorae is consistently present through
the sub-zone. Poaceae has increased, especially types
1 and 2 and the Cyperaceae content is maintained.
The introduced Pinus occurs here and the charcoal
content is high.

HISTORY OF THE VEGETATION

The abandoned river channel would have been
a lake or pond for almost the entire time, becoming

Proc. Linn. Soc. N.S.W., 129, 2008

a swamp supporting rooted vegetation for probably
only a few hundred years (see Fig. 5) prior to the
present. The lake was probably quite shallow, subject
to drying out on occasions of long dry spells. The
fluctuating water table would not have favoured
pollen preservation, hence there are long sections of
the profile with no pollen and thus no record of the
vegetation.

he periods where there is arecord of the vegetation
(Fig. 5) show that from about 38 to 37 k cal. yr BP (zone
D), Eucalyptus viminalis was dominant, with minor
Casuarina and some Leptospermum polygalifolium.
If the total Myrtaceae pollen of this zone is compared
with the top part of the profile, which is assumed to
represent the modern vegetation, then Eucalyptus
species, presumably the tree cover, would have been
greater than in the modern vegetation. Sclerophyllous
shrubs were present in the understorey, particularly
Cassinia arcuata which colonised disturbed sites (see
Appendix 2). There would have been some grasses
and a few Cyperaceae, the latter probably fringing
the lake. The vegetation was probably an open
sclerophyllous forest.

During the period 26 to 16 k cal. yr BP (zone
C), which includes the last glacial maximum, the
vegetation was predominantly a shrubland/grassland,
with the shrub Cassinia arcuata dominant, and
minimal trees. Some grasses were present and they
are the same types as found in the older zone below.
Cyperaceae was more abundant than in the preceding
zone.

By the period 6 to 2.2 k cal. yr BP (zone B), some
trees had returned, with Eucalyptus tereticornis and
Casuarina suggesting an open woodland, particularly
in the upper part of the zone. The shrub Cassinia
arcuata had disappeared entirely. Grasses were more
common, but the types found in the Holocene are
mostly different to those on older zones, showing that
the grass flora had changed. Trilete fern spores are far
more abundant, suggesting a moister environment.

After 2 k cal. yr BP, the increased Myrtaceae
pollen content suggests that the tree cover may have
increased, but preservation is too poor to identify
the species in Zone A2. In the younger Zone Al
Eucalyptus tereticornis has been identified, but it
decrease towards the top of the profile, most likely
due to European wood cutting. Grasses are more
abundant and the uppermost levels would have
included agricultural and introduced grasses and
herbs. Trilete spores decrease, especially in the upper
Zone Al.

The charcoal content is consistently higher in the
basal Zone D and the top Zone A. The higher content
may indicate more fires, but alternatively, it may
indicate more fuel to burn. Zone D has the greatest

105

VEGETATION HISTORY OF PENRITH LAKES, NSW

tree cover, hence more fuel is likely. The tree cover is
less in Zone A, but this period encompasses European
settlement which may have been the cause of greater
burning. There are high charcoal values at a few
other levels in the core, but overall the content is low
between these two zones.

CLIMATIC HISTORY

The climatic parameters for the tree species in
the region today and identified in the sediments (Table
3) give some basis for deducing climatic changes at
Penrith. Eucalyptus viminalis has the lowest of the
mean minimum temperatures for the coldest month,
hence is the most cold tolerant, and is predominant
in the period 38-36 k cal. yr BP, suggesting that
temperatures were lower than today. The tree cover
was probably greater than the present-day, suggesting
better effective soil moisture. Even if rainfall was the
same, the cooler temperatures would ensure more
effective moisture. This study concurs with previous
studies (reviewed by Allan and Lindesay, 1998; Pickett
et al., 2004) which indicate cool and moist climatic
conditions about 32 k yr BP, with temperatures some
2.5 °C lower than today.

The period 26-16 k cal. yr includes the last
glacial maximum. The vegetation was shrubland,
suggesting that temperatures and rainfall was less
than the minimum required by the tree species (Table
3). Other episodic events, for example, extreme
frosts or drought, may have contributed to keeping
the river flats treeless (Hope, 1989). Previous studies
indicate that the period 25 to 20 k yr BP was colder

and drier, with temperatures some 3-5 °C lower than
today. During the last glacial maximum (c. 18 k yr
BP), rainfall was up to half of present day values
and air temperatures were as low as 7-8 °C below
present, and winds were some 20 % stronger (Allan
and Lindesay, 1998). This study is thus in accord with
previous studies.

The transition period from glacial maximum
to the Holocene, when temperatures and rainfall
increased to more like the present, is missing from
this record. In the early Holocene (c. 9-6 k yr BP), the
climate was wetter than today and the late Holocene
appears to have been drier than the early Holocene
with less dramatic changes (Allan and Lindesay 1998,
Pickett et al. 2004).

Trees had returned to the river flats by 6 k cal. yr

BP, but their density was not as great when compared
with the period 38-36 k cal yr BP. The tree species
in this younger period are different as well.. Grasses
increased, but it was a different suite of species. There
was a much increased fern component, suggesting
wetter conditions, at least around the abandoned
channel. Grasses increased further and trees decreased
towards the present, probably due to the influence of
European settlement.
These climatic interpretations follow the overall
trends expected from previous studies, but moisture
relationships would have been the result of both
rainfall and river activity which at times may have
augmented or subtracted moisture from the site. It is
unknown how much influence river activity would
have had on the moisture relationships, but climatic
change would have had an effect on river discharge
also.

Table 3. Climatic parameters of some of the tree species found in the Penrith area (Benson, 1992), from
Boland et al (2002). P, found in the area today. F, found as a fossil in the core. Present day climatic
averages are included for Richmond, the nearest operating meteorological station and the mean annual

rainfall for Penrith (see text).

Mean max. Mean min. NE non cee Mean annual

Species temp, Hoes: temp., OU eE ders jaar ene rainfall, mm
month, °C month, °C

Penrith n/a n/a n/a 686
Richmond 28.9-29.4 3.2-4.4 n/a 800
Casuarina cunninghamii ' P, F 25-40 0-15 >50 500-1500 !
Eucalyptus eugenioides, P 25-33 0-6 >50 700-1100
E. moluccana P 26-32 0-10 >50 700-1200
E. tereticornis P, F 24-36 1-19 30 630-3000
E. viminalis F 20-32 -4-8 0-100 500-2000

‘A riverine species, hence rainfall alone is no indication of available moisture

i=)
Or

Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

DISCUSSION

The Penrith Lakes history may be compared with
other records in the Sydney Basin. At Lake Baraba
near Thirlmere (Fig. 1) (Black et al. 2006), lacustrine
clays were being deposited from >43 k yr BP until the
early Holocene, when peat deposition started about
8 k yr BP. Bands of oxidised sediments suggest lake
level fluctuations, hence the depositional environment
was generally similar to Penrith Lakes, except for the
different dates of the onset of peat deposition. Unlike
Penrith Lakes, however, the vegetation at Lake Baraba
was a Casuarinaceae woodland/shrubland which
remained relatively stable from >43 k yr BP, through
the glacial maximum until the early Holocene, when
Myrtaceae expanded at the expense of Casuarinaceae
(Black et al. 2006). Lake Baraba is set in a sandstone
gorge and it may have been sufficiently protected to
function as a refugium for woodland during the last
glacial maximum.

Mountain Lagoon, at about 500 m altitude in the
Blue Mountains (Robbie and Martin 2007) has a 23
k cal. yr BP record. The lagoon was a lake initially,
but peat formation started about 20 k cal. yr BP.
Both Casuarinaceae and Myrtaceae were prominent
in the vegetation throughout the whole time. While
the species of Eucalyptus changed with time, a few
Species were present the whole time, so some tree
species survived the glacial period. The vegetation was
thus remarkably stable through the climatic changes
of the glacial period. The reason for this stability may
lie in the favourable moisture relationships. If, during
the last glacial maximum, the rainfall of Mountain
Lagoon was just 50% of the current rainfall, it would
have been about the lower limits required by some of
the Eucalyptus species, hence they could remain at the
site. The location of Mountain Lagoon is sheltered,
hence moisture relationships would have been further
enhanced (Robbie and Martin 2007). Half the current
rainfall at Penrith Lakes (Table 3), however, would
have been less than the lower limits required by all
of the Eucalyptus species identified. Also, moisture
relationships of the open floodplain site of Penrith
Lakes would not be so favourable.

At Redhead Lagoon, south of Newcastle (Fig. 1),
the sedimentary record goes back some 75 k cal. yr
BP (Williams et al. 2006). At 40 k cal. yr BP, there
was an increase in Poaceae and a decline in woody
taxa, reflecting drier times. At the height of the last
glacial period, Casuarinaceae was prominent and
Angophora/Corymbia and Eucalyptus were present
also, hence the glacial maximum was not treeless at
this site. There was an increase in the shrubby taxa
Monotoca, Proteaceae and Asteraceae, including

Proc. Linn. Soc. N.S.W., 129, 2008

the spineless Asteraceae T. pleistocenicus (Williams
2005) also found in this study. Overall, the vegetation
communities were less complex during the last glacial
maximum when compared with those of today,
indicative of a harsh dry environment (Williams et al.
2006)

Each of these four sites in the Sydney Basin
represents a different environment and each has its
own distinctive vegetation and pattern of change
through the last glacial maximum. The Penrith Lakes
site is the only one that would have been a treeless
shrubland.

CONCLUSIONS

This study reveals periods of three different kinds
of vegetation on the river flats at Penrith in successive
times:

1) During 40-36 k cal. yr BP, a Eucalyptus
viminalis, Leptospermum polygalifolium open
sclerophyllous forest with a shrubby understorey of
predominantley Cassinia arcuata and a few grasses.

2) During 27-16 k cal. yr BP, a Cassinia arcuata
shrubland, with Cyperaceae, probably fringing the
abandoned channel, and few grasses.

3) During 6 k cal. yr BP to present, a Eucalyptus
tereticornus, Leptospermum juniperinum grassy
woodland with Cyperaceae and ferns, the latter
probably closer to the abandoned channel. The grass
flora here would have been substantially different to
that in the older periods recording the vegetation.

The pollen frequencies suggest that the tree
cover of the oldest E. viminalis vegetation unit was
greater than the youngest E. tereticornis unit. The
higher percentages of Poaceae in the younger unit
would depress the percentages of E. tereticornis, but
pollen concentrations of the E. tereticornis are low
also, in accord with the percentage evidence and a
lesser tree cover.

The climatic interpretations follow the general
trends of other studies: colder and drier during 40-36
k cal. yr BP, colder and drier still during 27-16 k cal.
yr BP, then warmer and wetter from 6 k cal. yr BP
to the present. The peak in fern spores infer that 3-2
k cal. yr BP was the wettest period hydrologically,
which may have been climatic, but river activity and
altered drainage may have affected the local moisture
relationships.

The vegetation through the last glacial maximum
at other sites in the Sydney Basin was specific to each
site. Penrith Lakes is the only one which would have
been a treeless shrubland.

107

VEGETATION HISTORY OF PENRITH LAKES, NSW

ACKNOWLEDGEMENTS

We are indebted to the Joyce W. Vickery Research Fund
of the Linnean Society of NSW, the River Group Fund of
the Federation of University Women, and the Penrith Lakes
Development Corporation for financial assistance with
this project. Our thanks go to Dr. John Turner, Dr. Mike
Barbetti, the National Parks and Wildlife Service of NSW
and the Forestry Commission of NSW for assistance. To the
many friends, relatives and colleagues who gave unstinting
help and encouragement, our heartfelt gratitude.

REFERENCES

Allen, R. and Lindesay, J. (1998). Past climates of
Australasia. In: “Climates of the Southern Continents’
(Eds J.E. Hobbs, J.A. Lindesay, H.A. Bridgman) pp.
208-247. (Wiley & Sons, Chichester).

Bembrick, C., Herbert, C., Scheibner, E., Stuntz, J. (1980).
Structural subdivision of the Sydney Basin. In ‘A
Guide to the Sydney Basin’ (Eds C. Herbert and R.
Helby) pp 3-9. (Department of Mineral Resources,
Geological Survey of New South Wales Bulletin 26,
Sydney).

Benson, D.H. (1992). The natural vegetation of the Penrith
1:100 000 map sheet. Cunninghamia 2, 503-662.

Black, M.P., Mooney, S.D. and Martin, H.A. (2006). A>
43,000-year vegetation and fire history from Lake
Baraba, New South Wales, Australia. Quaternary
Science Reviews 25, 3003-3016.

Birks, H.J.B. and Birks, H.H. (1980) ‘Quaternary
Palaeoecology’. (Edward Arnold, London).

Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall,
N. H., Hyland, B.P.M., et al. (2002). “Forest Trees of
Australia’ 4" Edition (Thomas Nelson: Australia).

BoM, (2006). Commonwealth Bureau of Meteorology
Website (http://www.bom.gov.au). Accessed August
2006.

Brown, C.A. (1960). “Palynological Techniques’ (C.A.
Brown, Baton Rouge) 188 pp

Bureau of Meteorology (1966). Rainfall Statistics in
Australia. (Commonwealth of Australia, Melbourne)

Chalson, J.M. (1991). The late Quaternary vegetation
and climatic history of the Blue Mountains, NSW,
Australia. PhD Thesis, University of New South
Wales (unpubl.)

Chalson, J.M. and Martin, H.A. (1995). The pollen
morphology of some co-occurring species of the
family Myrtaceae in the Sydney region. Proceedings
of the Linnean Society of New South Wales 115, 163-
191.

Harrison, S.P. (1993). Late Quaternary lake-level changes
and climates of Australia. Quaternary Science
Reviews 12, 211-231.

Hope, G. (1989). Climatic implications of timberline
changes in Australasia from 38 000 yr BP to
present. In “CLIMANZ 3, Proceedings of the
Third Symposium on the Late Quaternary Climatic
History of Australasia’ (Melbourne University 28-

108

29 November 1987) pp 91-99. (CSIRO, Institute of
Natural Resources and Environment).

Kohen, J.L., Stockton, E.D. and Williams, M.A.J.

(1984). Shaws Creek KII Rockshelter: a prehistoric
occupation site in the Blue Mountains Piedmont.
Archaeology in Oceania 18 (2) 19 (1/2), 57-73.

Macphail, M. and Martin, T. (1991). “Spineless”
Asteraceae (episode 2). PPAA Newsletter 23, 1-2.
(Palynological and Palaeobotanical Association of
Australasia: Melbourne).

Martin, H.A. (1973). The palynology of some Tertiary
Pleistocene deposits, Lachlan River Valley,

New South Wales. Australian Journal of Botany
Supplement 6, 1-57.

Moore, P.D., Webb, J.A. and Collison, M.E. (1991).
“Pollen Analysis’. (Blackwell Scientific Publications,
Oxford).

Nanson, G.C. and Young, R.W. (1988). Fluviatile evidence
for a period of late Quaternary pluvial climate in
coastal southeastern Australia. Palaeogeography,
Palaeoclimate, Palaeoecology 66, 45-61.

Nanson, G.C., Young, R.W. and Stockton, E.D. (1987).
Chronology and palaeoenvironment of the Cranbrook
Terrace (near Sydney) containing artefacts more than
40 000 years old. Archaeology in Oceania 22, 72-78

Penrith Lakes Development Corporation, (1983/84).
Penrith Lakes Scheme Regional Environmental
Study. Department of Environment and Planning,
Sydney.

Penrith Lakes Development Corporation, (2006). http://
www.penrithlakes.com.au Accesed June 2006.

Pickett, E.J., Harrison, S.P., Hope, G., Harle, K. et al.
(2004). Pollen-based reconstructions of biome
distributions from Australia, Southeast Asia and the
Pacific (SEAPAC region) at 0, 6000 and 18,000 “C
yt BP. Journal of Biogeography 31, 1381-1444.

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.

Stockton, E.D. and Holland, W.N. (1974). Cultural
sites and their environment in the Blue Mountains.
Archaeology and Physical Anthropology in Oceania
9 (1), 36-65.

Stuiver, M and Reimer,P.J. (1986-2005). Radiocarbon
calibration program Calib. Rev 5.0.2. http://calib.qub.
ac.uk/calib/calib.html (accessed November 2007)

Williams, N.J. (2005). The environmental reconstruction
of the last glacial cycle at Redhead Lagoon in coastal
eastern Australia. PhD Thesis, University of Sydney
(unpubl.).

Williams, N.J., Harle, K.J., Gale, S.J. and Heynis, H.
(2006). The vegetation history of last glacial-
interglacial cycle in eastern New South Wales,
Australia. Journal of Quaternary Science 21, 735-
750.

Young, R.W., Nanson, G.C. and Jones, B.G. (1987).
Weathering of late Pleistocene alluvium under a
humid temperate climate: Cranebrook Terrace,
southeastern Australia. Catena 14, 469-484.

Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

APPENDIX 1. POACEAE POLLEN TYPES.

Poaceae Diameter Exine thickness Gmercadeieeisties
type pm pum
Faintly granular surface pattern, grains often
1 ~35 5
crumpled
2 ~30 ~1.5 Smooth surface, grains retain spherical shape
3 25x35 Si Faintly granular surface, grains retain ovate shape,
with the broad end with pore depressed
Faintly granular surface, grains retain spherical
4 ~35 ~1
shape
5 ~30 ~1.5 Surface smooth, grains retain spherical shape
6 25-30 <0.5 Surface granular, grains usually crumpled

Proc. Linn. Soc. N.S.W., 129, 2008 109

VEGETATION HISTORY OF PENRITH LAKES, NSW

APPENDIX 2. THE IDENTITY OF TUBULIFLORITES PLEISTOCENICUS, ASTERACEAE

The form species Tubuliflorites pleistocenicus Martin 1973 was described to accommodate pollen
which was identifiable with the family Asteraceae but which lacked the spines usually seen on these grains.
The ‘spineless’ Asteraceae may be found in considerable abundance in the last glacial period pollen spectra
and is less common in younger sediments. This pollen type is found in Calomeria and several other genera
(see Chalson 1991). Calomeria is a shrub of high rainfall areas of Gippsland and southern New South Wales,
and is found in wet sclerophyll and the margins of rainforest (Chalson 1991). Descriptions by early botanists
of the species in ‘scrubby brushwood’ along the Nepean River, some 50 km upstream from Penrith, include
Calomeria amaranthoides and other wet sclerophyll species (Benson 1992). However, C. amaranthoides is
thought unlikely during the last glacial period (Zone C) which is expected to be much drier than the present.

This ‘spineless’ Asteraceae pollen is also found in Cassinia arcuata which also closely resembles
T. pleistocenicus. C. arcuata is widespread across the drought and frost-prone western slopes of NSW and
extends on the Central and Southern Tablelands. Moreover, it readily colonised bare and disturbed ground
(Macphail and Martin 1991), a habitat which could be created by river activity on a regular basis. It is thought
that C. arcuata was more likely during the glacial period, but Calomeria cannot be ruled out for Zone D (c.
35-31 ka), where there was a good forest cover, although no other wet sclerophyll species have been recorded
in the pollen spectra.

110 Proc. Linn. Soc. N.S.W., 129, 2008

J.M. CHALSON AND H.A. MARTIN

APPENDIX 3. POLLEN TYPE NAME ON THE POLLEN DIAGRAM (FIG. 4) AND THE PROBABLE

Name on the pollen diagram

Podocarpus

Pinus

Eucalyptus tereticornis
E. mannnifera

E. viminalis

Leptospermum polygalifolium

L. juniperinum
Unidentified Myrtaceae
Casuarina

Other tricolporate grains
Chenopodiacese
Acacia

Haloragaceae
Asteraceae/Liguliflorae
Asteraceae/Tubuliflorae
T: pleistocenicus
Monotoca
Caryophyllaceae
Cyperaceae
Myriophyllym

Trilete spores
Monolete spores
Poaceae type |

Poaceae type 2
Poaceae type 3

Poaceae type 4
Poaceae type 5
Poaceae type 6

SOURCE IN THE VEGETATION.

Probable source in the vegetation

Probably Podocarpus spinulosus, shrub to small tree
Pinus sp(p)., Introduced

Eucalyptus tereticornis

E. mannnifera

E. viminalis

Leptospermum polygalifolium

L. juniperinum

All other pollen types in the family

Probably Casuarina cunninghamii, possibly Allocasuarina sp(p)
Unidentified tricolporate grains, probably herbs and shrubs
All taxa in the family

All species in the genus

Haloragis/Gonocarpus

Faiesits taxa in the subfamily Liguliflorae
Echinate taxa in the subfamily Tubuliflorae

Most likely Cassinia arcuata, see Appendix 2

All species in the genus

Family Caryophyllaceae

All species in the family

all species in the genus

Ferns

Ferns

Poaceae, see Appendix |

Poaceae, see Appendix 1

Poaceae, see Appendix 1

Poaceae, see Appendix 1

Poaceae, see Appendix |

Poaceae, see Appendix |

Proc. Linn. Soc. N.S.W., 129, 2008 iLUL

IRCA SACPRETOR I

COMAt DIY MASOAIE mcrae :
et 1 EC BE

‘dow tall aE TGR KG Z | om Wien a att - a
palseen jody We pale PLE woh AOVOR AINE EES sae 4,
jatar < eh Aan ¥ tc iin a "ath EMER ay
ey eyruns : rental wear ae al tie acted Hi
nyt ‘foe ar be eet nl SPAN levis Kemriptlons

J ie i +N

; ser Wl: Rone arti Chheg “tianeren i. ied
ty fl period IPI se no deer ban
fi i ah Wnxal wich as ne ai aut
£0 ’ waning aanago phan ed hare pe ith be
22 oe MER ed oor vin nn a sop Ubi bast TF
rarer ig erate lesan 2 0 esa isi
7100 fe ll worerud: Me lenny h vil gates res Have b
(eye satirical} vidiaaoe Keafiedenoes 4 wcrc ive Adudes ‘7
etliride foe echt yideder) enter simiqloonm bathitebtag
viinnat ond ob een HLA,
wtttny cel? mnt voice HLA
sans wrt WAGEMOW
seating! yitrodoe cal nr ezal olmmesng'
ondtilwdel yugtidee wit ay axa sinned
C 2ibnogr A se@ ion elaine. yioihizoM_ «
sunay ot) nf esisogk ILA
susoullyiqovieD ‘vent
vitere! wit 1) eoiaoge ILA
aioeg orl ni esinoge Le
aryl
amet

| zibliogy A: san Ssooand

SihosyqgA soa ,ocsw0F

vibatsqnA ost costo
rituroqnA sea ekdoROF
| xibroqy A so sescdn
| xibnaqyA soz casoe01

a ee

112 ) |

The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales,
Australia. Part 7. Cycadophyta

W.B. KeitH HoLMes! AND Herpi M. ANDERSON?

146 Kurrajong Street, Dorrigo, NSW, 2453, Australia (Hon. Research Fellow, University of New England,
Armidale, NSW, 2351); 746 Kurrajong Street, Dorrigo, NSW. 2453, Australia (Hon. Palaeobotanist, South
African National Biodiversity Institute, Pretoria 0001 South Africa).

Holmes, W.B.K. and Anderson, H.M. (2008). The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales, Australia. Part 7. Cycadophyta. Proceedings
of the Linnean Society of New South Wales 129, 113-149.

Cycadophyte fronds comprise c. 4% of the catalogued specimens in the Holmes’ collections from two
quarries in the middle Triassic Nymboida Coal Measures of the Nymboida sub-Basin in north-eastern New
South Wales. The fronds are placed in fifteen taxa in the Cycadales and one in the Bennettitales. Eight new
species are described; Pseudoctenis nymboidensis, P. rigbyi, P. prolongata, P. cursanervia, P. grandis, P.
nettiana, Moltenia sparsispinosa and Ctenis marniana. Halleyoctenis megapinnata is nominated as a new

genotype for Halleyoctenis.

Manuscript received 24 June 2007, accepted for publication 6 February 2008.

KEYWORDS: Cycadophyta, Middle Triassic flora, Nymboida Coal Measures, palaeobotany.

INTRODUCTION

This is the seventh part of a series describing
the early-middle Triassic Nymboida flora Part 1
(Holmes 2000) of this series described the Bryophyta
and Sphenophyta, Part 2 (Holmes 2001) the
Filicophyta, Part 3 (Holmes 2003) fern-like foliage,
Part 4 (Holmes and Anderson 2005a) the genus
Dicroidium and its fertile organs Umkomasia and
Pteruchus, Part 5 (Holmes and Anderson 2005b)
the genera Lepidopteris, Kurtziana, Rochipteris and
Walkomiopteris and Part 6 (Holmes and Anderson
2007) the Ginkgophyta.

A description of the Coal Mine and Reserve
Quarries, the source localities of our described
material, together with a summary of the geology
of the Basin Creek Formation, the Nymboida Coal
Measures and the Nymboida Sub-Basin were provided
in Holmes (2000).

In this paper, leaves with cycadophyte affinities
are described and illustrated. No fertile material has
been found. The cycadophytes include both true
cycads in the Order Cycadales, and the bennettitites
in the extinct Order Bennettitales (~Cycadioidales).
The origins of the cycadophytes date back to the

Upper Carboniferous (Taylor and Taylor 1993).
Cycad survivors of the End-Permian Extinction
diversified and reached their maximum development
and world-wide distribution during the Mesozoic
Era. Reconstructions of the Mesozoic landscape
often portray dinosaurs in close association with
cycadophytes (White 1990). Cycadophytes have been
in a decline over the last 100 ma. and today true cycads
are a relatively small group of plants distributed
through tropical and warm temperate regions of the
northern and southern hemispheres. Extant cycads
comprise c. 190 species in c. 11 genera (Jones 1993)
and new species continue to be discovered and
described (Singh and Radha P 2006). Two species of
cycads, Lepidozamia peroffskyana and Macrozamia
johnsonii, still survive in the Nymboida region of
northern NSW (Hill and Osborne 2002).

METHODS

The material described in this paper is based
mainly on collections made by the senior author and
his family from two Nymboida quarries (Coal Mine
Quarry and Reserve Quarry) over a period of forty
years and on limited specimens in old collections of
the Australian Museum, Sydney, and in the Geology

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Department of the University of New England,
Armidale, as noted in Retallack (1977) and Retallack
et al. (1977).

The exact horizon or source of most specimens
is uncertain as the plant fossil material was collected
mostly from blocks fallen from the working quarry
faces. The Coal Mine Quarry has not been exploited
for many years but its weathering high-wall exposure
provides an excellent cross-section of beds that
demonstrate the palaeo-environmental conditions
at the time of sedimentation (Retallack 1977). The
Reserve Quarry with excellent exposures was active
until recent times. During 2006 the whole quarry was
bulldozed for “restoration” purposes into a featureless
bowl — a great scientific loss!

In the Holmes collections from Nymboida,
leaves attributed to the Cycadophyta comprise c.
4% of the 2600 selectively catalogued specimens of
the preserved floodplain flora. However, in life, the
Cycadophyta may have been more common in upland
areas where their remains would have had very little
chance of becoming fossilized.

The Nymboida specimens are preserved in
mudstones, siltstones and sandstones as carbonaceous
compressions or impressions in which the gross
morphology is usually well-preserved. However,
spores and cuticles have been destroyed by a tectonic
heating event during the Cretaceous Period (Russel
1994). Therefore our identification of taxa is based
only on characters of gross morphology.

Cycadophyte leaves have been recorded in
the Triassic of Gondwana from South America by
Frenguelli (1950), Menendez (1951), Stipanicic and
Bonetti (1965), Bonnetti (1968, 1972), Artabe (1985),
Herbst and Troncoso (2000), Troncoso and Herbst
(2000), Ottone (2006); from India by Lele (1956);
from South Africa by Du Toit (1927), Anderson
and Anderson (1983, 1989, 2003); from Australia
by Johnston (1888), Shirley (1897, 1898), Walkom
(1917, 1924, 1925, 1928), Jones and De Jersey (1948),
De Jersey (1958), Hill et al. (1965), Flint and Gould
(1975), Retallack (1977), Rigby (1977), Webb (1980),
Holmes (1982) and from Antarctica, a cycad stem
(Smoot et al. 1985), a cycad pollen cone (Klavens et
al. 2003) and cycad cataphylls (Hermsen et al. 2006).
Where possible, identification and comparisons of the
Nymboida material have been made from descriptions
and illustrations in the above publications. Due to
time and geographical separation our material has not
been compared with northern hemisphere taxa, i.e.
non-Gondwana species.

Type and illustrated material is housed in the
Australian Museum, Sydney. Some additional
specimens are in the collections of the Geology

114

Department, University of New England, Armidale
and the University of Queensland.

SYSTEMATIC PALAEOBOTANY
CYCADOPHYTA

During the Mesozoic Era the cycadophytes
comprised two orders, the Cycadales (cycads) and
the Bennettitales (cycadioids). They are distinguished
essentially by their reproductive organs and frond
cuticle structure (Anderson and Anderson 1989, pp
276-279; Taylor and Taylor 1993). In the absence
of cuticles and/or fertile structures, as is the case at
Nymboida, the correct assignment of cycadophyte
fronds is difficult. In this paper we follow Anderson
and Anderson (1989, 2003) who, from their large
Molteno collections, albeit with little preserved
cuticle, classified all their cycadophyte foliage in
the Cycadales except for the genus Halleyoctenis
that was placed in the Bennettitales. The simple
leaves belonging to the form genus 7aeniopteris have
been placed in the Bennettitopsida by Anderson and
Anderson (2003). Taeniopterid and other simple leaves
from Nymboida will be described in a forthcoming
paper of this series.

Anderson and Anderson (1989) and Herbst and
Troncoso (2000) noted the polymorphic character of
their form species especially when large collections
were available. They noted the presence of
intergrading forms and Herbst and Troncoso (2000)
also questioned the erection of some new form
species. We acknowledge this as a constant problem
in palaeobotanical taxonomy. Despite the sometimes
limited material, we have described and separated
our Nymboida cycadophyte fronds into form species
based on all available characters of frond and pinna
gross morphology and especially on venation
architecture and density.

Order Cycadales
Family incertae sedis

Genus Pseudoctenis Seward 1911
Type species
Pseudoctenis eathiensis (Richards) Seward

1911.

Pseudoctenis fissa DuToit 1927
Figures 1A, B; 2A; 3A—F

Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Selected references

1927 DuToit Pseudoctenis fissa, p. 386, fig. 22 (3).

1968 Bonetti Pseudoctenis anomozamoides, P|. 2,
figs 1, 2.

1968 Bonetti Pseudoctenis cf. falconeriana, PI. 2,
figs 3, 4; Pl. 3, figs 1, 4.

1989 Anderson and Anderson Pseudoctenis fissa, p.
286, t. figs 1-11; Pls 155—157, 162 (1-5), 167
(7-18), 323 (1-2).

Description

A very variable small to medium-sized frond,
narrow-elliptic to elongate-spathulate, 150-200 mm
long, width at mid-lamina 25—90 mm; rachis to 3 mm
wide at base tapering distally. Pinnae attached laterally
at high angle in lower half, becoming slightly acute
apically, opposite to subopposite, mostly separate to
the base, or more widely separated and confluent, or
with strongly decurrent acroscopic bases; pinna shape
from narrow to broad oblong, adjacent pinnae often of
irregular width, basal pinnae short and broad becoming
more elongated to 2/3 to apex then reducing in length,
apical pinnae sometimes conjoined, apices truncate
to broadly obtuse or shallowly cleft. Veins departing
from rachis at high angle, sometimes forking close to
base or in mid-lamina, running parallel to each other
to apex; vein density in mid-lamina 14—18/10 mm.

Material

AMEF126860, 133960, 133961, 133962, 133963,
133964, 133965, 133968, Coal Mine Quarry;
AMF133966, 133967, Reserve Quarry.

Discussion

The Nymboida fossils placed in P. fissa reflect the
range in size and form of the material from the Upper
Umkomaas and Hlatimbe Valley localities in the
Molteno Formation of South Africa (Anderson and
Anderson 1989) and from Argentina (Bonetti 1968)
as listed in the selected references. However, some
Nymboida specimens are larger than the Umkomaas
material and exceed in size even those from the
Hlatimbe Valley locality (Anderson and Anderson
1989, Pl. 167, figs 12-15).

One slab, AMF133963 (Fig.1B), shows two
virtually complete leaves aligned probably from a
common point of attachment. There is some woody
tissue (? stem) close to the base of the fronds but no
clear connection.

Pseudoctenis nymboidensis Holmes and Anderson

sp. nov.
Figures 4A; 5A, B; 6A—E; 7A; 8A

Proc. Linn. Soc. N.S.W., 129, 2008

Diagnosis

Medium to large variable Pseudoctenis frond;
pinnae closely spaced, spathulate to broad-elliptic to
broad-linear, base straight or slightly contracted, apex
broadly rounded, vein density 12—16/10 mm, once-
forked proximally or medially.

Description

Frond medium to large, probably ovate to broad-
elliptic but as no complete specimens are available
the total length and shape is unknown, to >300 mm
long and >200 mm wide, the leafbase is not known;
at mid-frond the rachis is up to 10 mm wide. Pinnae
semi-dorsally attached, closely spaced with confluent
bases; ranging in form from spathulate to elongate-
elliptic to broad linear, adjacent pinnae often differing
in width, from 8-20 mm wide and to 105 mm long;
attached at high angle to rachis, 75°— 90° on lower and
mid-portions of frond, becoming more acute apically,
slightly contracted proximal to base, expanding to
mid-pinna then contracting slightly to broad obtuse
apex or broad-linear with parallel margins. Veins
attached straight to rachis or basiscopically decurrent;
some veins forking close to rachis and occasionally
medially; vein density across mid lamina 12—16/10
mm.

Holotype
AMF 133969, Australian Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material

AMEF133970, 133971, 133972, 133973, 133974,
133975, 133976, 133977, 133978, 133979, Coal
Mine Quarry; AMF133980, 133981, Reserve

Quarry.

Name derivation
nymboidensis — from the type locality —
Nymboida Coal Measures.

Discussion

Although complete fronds are not known,
Pseudoctenis nymboidensis appears to be one of
the larger of the Nymboida cycadophytes. The
holotype (Fig. 4A) is of two incomplete fronds lying
sub-parallel to each and suggesting autocthonous
preservation of fronds abscissed from a nearby
parent plant. Pseudoctenis nymboidensis is relatively
common and the specimens include a continuum of

115

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

intergrading pinna forms ranging in outline from
spathulate (Fig. 5A) when they are somewhat similar
to specimens of the Molteno P. spathulata Du Toit
(Anderson and Anderson 1989, pls 175-178), to
broad linear with unconstricted bases (Fig. 6C). The
forms with unconstricted bases and longer pinnae
(Fig. 8A))are similar in shape to P. longipinnata
Anderson and Anderson and the much larger P.
brownii from the Burgersdorp Formation of South
Africa (Anderson and Anderson 1989) but differ
by the less dense venation. Large fronds from the
Ipswich Coal Measures, similar in size and outline
to the largest of P. nymboidensis specimens, were
compared with P. brownii (as Nilssonia cf browni Du
Toit) by Jones and DeJersey (1947). These Ipswich
leaves differ from P. nymboidensis by the denser
venation. Pseudoctenis megaspatulata Herbst and
Troncoso (2000 p. 286) from Chile is a very much
larger spathulate frond also with denser venation. The
largest specimens of P. nymboidensis approach in size
P. grandis (described below) but differ by the finer
denser venation. Pseudoctenis multilineata (Shirley)
Herbst and Troncoso 2000 differs by the significantly
denser venation (see comments under Halleyoctenis
below).

Pseudoctenis rigbyi Holmes and Anderson sp. nov.
Figures 9A; 10A

Selected references

1917 Pseudoctenis eathiensis, Walkom, p. 19, Pl. 7,
figs 1,2.

1965 Pseudoctenis eathiensis, Hill et al., Pl. T7,
fig 5.

1975 Pseudoctenis eathiensis, Flint and Gould, P1.
2, ihe, B

Diagnosis

Small to medium-sized Pseudoctenis frond;
rachis stout; pinnae well-separated, decurrent,
elongate elliptic to broad linear, apices acute, adjacent
pinnae often of irregular width; venation once-forked
proximally then straight and parallel to apex; vein
density 16—20/10 mm.

Description

A small to medium-sized Pseudoctenis frond,
length >300 mm long, to 160 mm wide, ovate to broad-
elliptic, rachis stout, to 8 mm wide at base, tapering
apically. Pinnae well-separated, semi-dorsally
attached at 90° near base, more closely-spaced in
mid-frond at c. 75° and more acute apically, slightly
contracted proximally but base expanded at point
of attachment, decurrent to occasionally confluent,

116

broad-linear to elongate elliptic, in mid-frond c.
60-80 mm long, adjacent pinnae often of irregular
width, from 3—8 mm, apices acutely rounded but
rarely preserved, length to breadth ratio c. 10 to 1 but
variable due to irregular pinna widths. Veins forking
once close to base then running straight and parallel
to apex; vein density in mid-lamina 16—20/10 mm.

Holotype
AMF 133982, Australian Museum, Sydney.

Type Locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMEF133983, 133984, 133985, 133986,
UNEF13451, Coal Mine Quarry.

Name derivation
rigbyi, for Dr J.F. Rigby, a long-time researcher
of Australian fossil plants.

Discussion

This is an uncommon cycadophyte at Nymboida.
The fronds display a range of variation in pinna size
and spacing along the frond rachis. The holotype
specimen shows three fronds aligned parallel to
each other (Fig. 9A), which suggests they may have
abscissed from a nearby parent plant. Pseudoctenis
rigbyi differs from P prolongata (below) by the
shorter pinna length to width ratio and to all other
Nymboida cycadophytes by the broad linear to elliptic
pinnae with variously contracted bases. Pseudoctenis
rigbyi is close in vein density and pinna shape to some
Molteno specimens of P. gracipinnata (Anderson and
Anderson 1989, pls 159, 160, 168) but is generally
a very much larger frond. Pseudoctenis longipinnata
and P. harringtonia from the Molteno Formation
(Anderson and Anderson 1989) are similar in venation
density to P rigbyi. Pseudoctenis longipinnata differs
by the larger frond size and by the longer, closely-
spaced confluent pinnae; P harringtonia differs by
the basally uncontracted and shorter pinnae. In gross
morphology the Nymboida specimens are closely
similar to fronds referred to Pseudoctenis eathiensis
(Richards) Seward by Walkom (1917), Hill et al.
(1965) and Flint and Gould (1975). We believe the
epithet eathiensis is inappropriate as it is based on
Jurassic material from Scotland. Specimen UQF158
from the Esk Beds of Queensland and listed by
Walkom (1917) under P. eathiensis has preserved
cuticle and was redescribed by Joshi et al. (2004) as

Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

P. pantii. Under the International Code of Botanical
Nomenclature (2001), P. pantii is a form taxon based
only on leaves with cuticle preserved.

Pseudoctenis sanipassiensis Anderson and
Anderson 1989
Figures 11A; 12A—C

Reference
1989 Pseudoctenis sanipassiensis, Anderson
and Anderson p. 289, figs 1, 2.

Description

At Nymboida, known only from incomplete
specimens; original length and width not known.
Rachis to 5 mm wide. Pinnae semi-dorsally attached
from 60°—90° to the rachis; linear-lanceolate, variable
in width, from 4-12 mm, length >90 mm, apices
not known, slightly contracted near the base with
attachment decurrent to confluent. Veins decurrent on
rachis, forking once close to the base then running
straight and parallel to the apex; vein density c. 16/10
mm.

Material
AMF133987, 133988, 133989, 133990, 133991,
133992, 133993, 133994, 133995, Coal Mine

Quarry.

Discussion

Although incomplete, the above material, except
for the semi-dorsal attachment of the pinnae, agrees
well with fronds of P. sanipassiensis from the Molteno
Formation as described and illustrated by Anderson
and Anderson (1989 pls 185, 186). It differs from
other Pseudoctenis spp. with similar venation density
by its longer, narrower pinnae.

Pseudoctenis prolongata Holmes and Anderson
sp. nov.
Figures 13A; 14A—C

Diagnosis

Small to medium Pseudoctenis frond; pinnae
well-spaced, very narrow elongate elliptic, length to
width ratio 16—23 to 1; vein density 16—20/10 mm.

Description

A small to medium-sized frond, broad elliptic to
> 300 mm long and to 200 mm wide; rachis stout, to 5
mm wide at base, tapering apically. Pinnae with semi-
dorsal attachment, well-separated, bases expanded,
decurrent to barely confluent, attached at high angle
but becoming slightly acute towards frond apex, very

Proc. Linn. Soc. N.S.W., 129, 2008

narrow elongate-elliptic, to 100 mm long in mid
frond, 3—5 mm wide; length to breadth ratio of 16—23
to 1, apex acutely rounded. Veins decurrent or straight
from rachis, proximally forking once and then running
straight and parallel to the acutely-rounded apex; vein
density in mid-pinna 16—20/10 mm.

Holotype
AMF 133996, Australian Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMF 133997, 133998, 133999, Coal Mine
Quarry; AMF134000, 134001, Reserve Quarry.

Name derivation
prolongata — Latin, lengthened, referring to the
extremely elongate-elliptic pinnae.

Discussion

Pseudoctenis prolongata differs from all
described Gondwana cycadophytes by the long narrow
elliptic pinnae. Pseudoctenis gracipinnata Anderson
and Anderson (1989 p. 240) is somewhat similar but
differs by its smaller overall size and shorter pinna
length to breadth ratio.

Pseudoctenis nettiana Holmes and Anderson sp.
nov.
Figures 15A—G

Diagnosis

A very small Pseudoctenis frond; pinnae well-
separated, slightly confluent, linear, width irregular,
apices obtuse or rarely lobate to lacerate; vein density
2430/10 mm.

Description

Frond variable from very small to small,
lanceolate to broad elliptic, to 100 mm long, c. 40—
50 mm wide; rachis c. 2 mm wide at base, tapering
distally. Pinnae semi-dorsally attached at a high
angle, becoming more acute apically, near frond
base well-separated, in mid-frond and apically more
closely spaced, slightly confluent, broad to narrow
linear, opposite to alternate, adjacent pinnae often of
irregular width, 1.5—2.5 mm wide, 15—25 mm long,
apices obtuse but on specimen AMF134003 (Fig.
15A) slightly lobed or lacerate. Venation fine, forking
close to the base then running straight and parallel to
apex; vein density 24—30/10 mm.

117

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Holotype
AMEF134003, Australian Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material

AMF134002, 134004, 134005, 134006, 134007,
134008, 134009, Coal Mine Quarry; AMF134010,
134011, Reserve Quarry.

Name derivation

nettiania, for Netta Holmes-Lee, daughter of
the senior author, who, for many years, assisted on
family collecting trips.

Discussion

The fronds placed in P. nettiana are close to
the fragmentary type specimen of P. harringtonia
Bonetti 1968 (Pl. 3, fig 4) and refigured by Anderson
and Anderson (1989, Pl. 323, fig. 3). The parameters
of P. harringtonia were expanded by Anderson and
Anderson (1989) to include leaves from the Molteno
Formation of South Africa and by Herbst and
Troncoso (2000) for leaves from Chile. Pseudoctenis
harringtonia, as defined by those authors differs from
P. nettiana by the larger size, coarser venation and
the pinnae tapering to an acute apex. Pterophyllum
parvum Shirley (1898, Pl. 17, fig. 4), an apical portion
of a small frond from Queensland, differs from P.
nettiana by the irregular length and arrangement of
the pinnae and by the coarser venation.

Pseudoctenis grandis Holmes and Anderson sp.
nov.
Figures 16A, B

Diagnosis

A large Pseudoctenis frond, length not known,
>260 mm wide, pinnae long, broad-linear; attached
semi-dorsally, slightly contracted basally; vein
density 22—24/10 mm.

Description

Known from only four incomplete fragments,
frond length not known, width >260 mm. Pinnae
closely spaced, broad linear, semi-dorsally attached,
to >130 mm long but complete pinnae not preserved,
14-17 mm wide, slightly contracted close to the
base, apex not known. Veins sometimes forking once
proximally then running straight and parallel distally;
vein density 22—24/10 mm.

Holotype
AMF 134012, Australian Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMEF134013, 134014, 134015, Coal Mine
Quarry, Nymboida.

Name derivation
grandis — Latin, /arge, referring to the apparent
large frond size.

Discussion

This is amongst the largest of known Gondwana
cycadophyte fronds. Pseudoctenis grandis differs
from the large leaf P cursanervia (below) by the
veins being more than twice as fine and dense.

Pseudoctenis cursanervia Holmes and Anderson
sp. nov.
Figures 17A, B

Diagnosis

A large Pseudoctenis frond with broad-linear
pinnae, almost parallel-sided, closely spaced but
separate to the base, to >175 mm long; veins very
coarse, density 10/10 mm.

Description

Based on the rachis and pinna dimensions of
the two incomplete specimens with basal and apical
portions missing; the complete fronds were very
large; rachis to 10 mm wide. Pinnae broad-linear to
>175 mm long and 15-20 mm wide, aligned at 90°
to the rachis with semi-dorsal attachment, separate to
the base, slightly constricted proximally, basiscopic
and acroscopic attachment slightly decurrent, apex
obtuse. Veins attached straight to the rachis or around
the basiscopic and acroscopic margin following the
line of the expanded base; some veins forking once
at or near the pinna base then running straight and
parallel to the apex; veins coarse, to 0.3 mm in width,
some veins with two or three longitudinal striations;
vein density 10/10 mm.

Holotype
AMF134016, Australian Museum, Sydney.

Type locality
Coal Mine Quarry, Nymboida. Basin Creek

Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMF 134017 Coal Mine Quarry, Nymboida.

Name derivation
cursa — Latin, coarse; — nervia —Latin, vein,
referring to the coarse venation.

Discussion

Pseudoctenis cursanervia and P. grandis
(described above) differ by their large broad-linear
pinnae from all described Gondwana cycadophyte
fronds. Pseudoctenis cursanervia differs from P.
grandis by the much coarser veins. Pseudoctenis
megaspathulata Herbst and Troncoso (2000), a very
large frond from Chile, differs by the spathulate
pinnae. Pseudoctenis brownii (DuToit 1927) Anderson
and Anderson is a poorly-known large leaf with veins
apparently closely spaced and differs from P grandis
by the moderately contracted pinna bases.

Pseudoctenis azcaratei (Herbst and Troncoso)
Holmes and Anderson comb. nov.
Figures 18A—D; 19A
References
2000 Pterophyllum azcaratei Herbst and Troncoso,
p. 29, figs 2F, 5A, B.
2000 Pterophyllum azcaratei Troncoso and Herbst,
p.140.

Description

No complete fronds from Nymboida have been
collected; mid-frond fragments are up to 100 mm
long and to c. 140 mm wide but complete pinnae are
rarely present. Rachis to 4 mm wide. Pinnae broad-
linear, semi-dorsally attached at c. 90° but becoming
inclined apically; adjoining pinnae often irregular in
width, closely spaced, 2-6 mm wide, to c. 70 mm
long, margins straight and parallel, lamina is not
proximally contracted but at point of attachment
is slightly decurrent or sometimes confluent with
adjacent pinnae. Veins arise straight from the rachis,
some forking close to the base or sometimes more
distally, then running straight and parallel to the apex;
vein density ranges from 26—28/10 mm.

Material
AMF 134018, 134019, 134020, 134021, 134022,
Coal Mine Quarry.

Discussion
This Nymboida material is closely similar to

Proc. Linn. Soc. N.S.W., 129, 2008

fronds described as Pterophyllum azcaratei from the
La Ternera Formation of Chile (Herbst and Troncoso
2000). It differs from all other Nymboida taxa by the
broad linear pinnae with very dense venation. We
are hesitant in following the Pterophyllum generic
assignation by Herbst and Troncoso as Pterophyllum
is essentially a Northern Hemisphere genus and
based on cuticle morphology is bennettitalean.
From the studies of Anderson and Anderson on the
Cycadophyta of the Molteno Formation, it appears
that most Gondwana cycad genera are probably
cycadalean. As Pterophyllum azcaratei is closely
comparable in gross morphology with many other
Pseudoctenis species we believe it is better placed
in the latter genus. Future study of specimens with
well-preserved cuticle will determine the correct
taxonomic assignment.

Pseudoctenis sp. cf Pseudoctenis strahanii
(Johnston) Anderson and Anderson 1989
Figure 19B

Description

Frond probably medium-sized; fragment
preserved 80 mm long, c. 110 mm wide base and apex
missing; rachis 3 mm wide. Pinnae well-separated,
dorsally attached at high angle to rachis, 5-7 mm
wide, to 56 mm long, basiscopic base decurrent,
margins parallel, apices irregularly cleft into 2—3
irregular lobes. Basiscopic veins following decurrent
base, acroscopically departing straight from rachis,
forking once close to the rachis then running straight
and parallel into apical lobes. Vein density in mid-
lamina 18/10 mm.

Material
AMF134023, Coal Mine Quarry.

Discussion

The above description is based on a single
specimen. It is compared with Pseudoctenis strahanii,
the only previously described Pseudoctenis with
deeply cleft and lobed pinna apices. The lectotype
from Tasmania and illustrated in Anderson and
Anderson (1989, pl. 323, fig. 4) differs from the
Nymboida specimen by the pinnae with deeper
apical clefts and by the less dense venation (12/10
mm). A single specimen of the assemblage placed in
P. nettiana (see above) has lobed or lacerated pinna
apices but differs from P. sp. cf. P. strahanii by the
much smaller size and more dense venation.

Pseudoctenis sp. A
Figures 19C, D

119

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Description

This form is known from two apical fragments.
Frond small, probably broad-elliptic, length not
known, width to c. 60 mm; rachis slender, 1.5 mm
wide. Pinnae well-separated to the rachis, bases
slightly expanded, semi-dorsally attached at high
angle in mid-frond, becoming more acute apically;
narrow-oblong with truncate apices, pinnae to 40 mm
long, adjacent pinnae irregular in width, 3—6 mm wide.
Basiscopic veins decurrent on rachis, others attached
straight, most forking once then running straight and
parallel to apex. Venation density 20—26/10 mm.

Material
AMF134024, 134025, Coal Mine Quarry.

Discussion

By the elongate-oblong pinnae Pseudoctenis
sp. A is similar to P. multilineata (= Pterophyllum
multilineata of Shirley 1897, fig. 7a and see below)
but differs by the much smaller size and less dense
venation.

Pseudoctenis sp. B
Figure 24A

Description

A Pseudoctenis probably of medium size but
known from only a single apical fragment. Pinnae well-
spaced, attached semi-dorsally to rachis, decurrent to
confluent, at high angle then becoming more acute
apically on frond; elongate-elliptic, expanding from
a narrow base to mid-pinna then contracting slightly
distally, > 60 mm long and to 6 mm wide, apices not
known. Veins decurrent on rachis then decurving
and running parallel to margin to apex, forking once
proximally to medially. Vein density across mid-pinna
12-14/10 mm.

Material
AMF 134026, Coal Mine Quarry.

Discussion

This is a rare frond form known only from an
apical fragment. Pseudoctenis sp B differs from all
other Nymboida cycadophytes by the pinna shape and
the coarse venation pattern. In outline P. gracipinnata
of Anderson and Anderson (1989 p. 290. figs 2—4) is
similar to P. sp B but differs by its larger size and less
dense venation.

Genus Moltenia Du Toit 1927

The genus Moltenia was erected to include

cycadophyte foliage with pinnae exhibiting variously
serrate margins and lacerated or lobed pinna apices.

Type species
Moltenia dentata Du Toit 1927

Moltenia sparsispinosa Holmes and Anderson sp.
nov.
Figures 20A; 21A—D
Diagnosis
Medium-sized Moltenia frond; pinnae elliptic
with margins very sparsely spinulate; apices truncate
or serrate; vein density c. 12-16 / 10 mm.

Description

Frond medium-sized although complete fronds
not preserved; probably broad-elliptic in outline;
incomplete specimens >185 mm long and to 120 mm
wide. Rachis ribbed, slender, 3 mm wide basally and
tapering apically. Pinnae broad-elliptic, to 60 mm long
and 8-15 mm wide, attached semi-dorsally to rachis
at a high angle but becoming more acute apically;
variously contracted proximally, bases confluent,
apices rarely preserved, cuneate, sometimes serrate;
margins entire or with rare and widely-separated
short conical spines. Vein attachment at basiscopic
base strongly decurrent, midveins straight and
acroscopic veins decurrent upwards; veins forking
once in an irregular manner — one pinna may have
all veins forking close to the rachis while an opposite
pinna has veins forking away from the rachis, all then
running sub-parallel to the apex; the outermost vein
terminates in a spine when present; vein density in
mid-lamina c. 12—16/10 mm.

Holotype
AMF 134028, Australiam Museum, Sydney.

Type locality

Coal Mine Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Other material
AMF 134033 paratype; 134027, 134029, Coal
Mine Quarry; 134030, Reserve Quarry.

Name derivation
sparsus — Latin, few, rare; spinosa — Latin, spiny

Discussion

The slab bearing the holotype specimen shows
three incomplete sub-parallel fronds. Occasional
pinnae bear rare small conical marginal spines and

Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

one pinna (Fig. 21A) shows a lacerate apex. On the
basis of similar shape and venation, a fragmentary
specimen without observable preserved spines,
AMF 134030 (Fig. 21B) is included in this taxon.
Moltenia sparsispinosa is closest to M. paucidentata
Anderson and Anderson (1989, p. 350) but differs
by the relatively fewer spines and by the wider but
shorter elliptical pinnae. Herbst and Troncoso (2000,
fig. 4E) illustrated a frond from Chile as Pseudoctenis
longipinnata that showed a few distinct serrations
and should perhaps be placed in Moltenia dentata Du
Toit.

Genus Crenis Lindley and Hutton 1834

Type species
Ctenis falcata Lindley and Hutton 1834.

The genus Cfenis includes cycadophyte leaves
characterized by the laterally attached entire pinnae
with veins more or less parallel, anastomosing and
meeting the lamina margin (Anderson and Anderson
1989). It is a rare element in Gondwana Triassic
floras. Poorly preserved material has been referred to
Ctenis from Queensland (Jones and DeJersey 1948)
and from Argentina (Menendez 1951). Anderson
and Anderson (1989) described two species from the
Molteno Formation of South Africa.

Two incomplete fronds and their counterparts,
described below as Ctenis marniana and Ctenis sp. A,
were collected at the Reserve Quarry from the same
horizon of grey siltstone. The pinna shape of each
specimen and the anastomosing pattern are somewhat
different but due to their close proximity they may
possibly be fronds from a single variable species. We
place them in the genus Cfenis on the basis of their
clearly preserved anastomosing venation but with
reservations as the pinnae are semi-dorsally attached
and the outer veins continue parallel to the lamina
margin to terminate at the pinna apex.

Ctenis marniana Holmes and Anderson sp. nov.
Figures 22A—C; 23A

Diagnosis

A small pinnate elliptic frond; pinnae at midfrond
broad-elliptic, distally conjoining to form entire apex;
venation anastomosing; areoles irregular elongate
rhomboidal, becoming narrower distally; vein density
across mid-pinna c. 12/10 mm.

Description

A small pinnate frond, base missing; rachis at
broken base 2mm wide. Pinnae semi-dorsally attached,

Proc. Linn. Soc. N.S.W., 129, 2008

basally at c. 100°, in mid-frond at 90° and becoming
slightly acute apically; confluent; succeeding pinnae
increasing in length from short ovate to elongate-
oblong or broad-elliptic; pinnae at mid-frond 22 mm
long, variable in width from 11—15 mm, tapering to
broad obtuse apices; distally the pinnae conjoin to
form an entire obtuse Gontriglossa-like apex to the
frond. Veins attached straight to rachis, forking then
anastomosing with adjacent veins to form irregular
elongate rhomboidal areoles from 4-10 mm long
and 0.6—1 mm wide, becoming shorter and narrower
towards the pinna apex, outer veins running parallel
to margin and terminating around the pinna apex;
vein density across mid-lamina c. 12/10 mm.

Holotype
AMF 134031 and counterpart AMF134032,
Australian Museum, Sydney.

Type locality

Reserve Quarry, Nymboida. Basin Creek
Formation, Nymboida Coal Measures, Middle
Triassic.

Name derivation

marniana — for Marnie Holmes-Kaner, daughter
of the senior author, a keen-eyed helper on collecting
trips.

Discussion

The entire apex of AMF134031 (Fig. 22A) is
reminiscent of the apices of Triassic Glossopteris-
like leaves (Holmes 1992) now placed in the genera
Gontriglossa and Cetiglossa by Anderson and
Anderson (1989, 2003). Ctenis marniana differs
from C. sp. A (below) by the shorter rounded pinnae
and the less dense venation forming irregular and less
elongate areoles.

Two species of Ctenis have been recorded from
the Molteno Formation of South Africa by Anderson
and Anderson (1989, p. 343). Ctenis biloba differs
from the Nymboida specimens by the lobed pinna
apices and denser venation; C. sp. A (of Anderson and
Anderson 1989) differs by the contracted acroscopic
pinna bases and is possibly bipinnate. Detached pinna
fragments occur in the Ipswich Coal Measures of
Queensland. Crenis sp. / of Jones and DeJersey (1948,
figs 27, 28) differs from the Nymboida material by the
acute pinna apices; their C. sp. 2 (Fig. 29) differs by
the irregular, fewer and more elongate anastomoses.

Ctenis sp. A
Figures 23B; 24B, C

121

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Description

Part and counterpart of a mid-portion 90 mm long
of a medium-sized frond, rachis 2 mm wide. Pinnae
well-separated, decurrent, opposite to alternate,
attached semi-dorsally at c. 80°, constricted slightly
near the base, lanceolate to 50 mm long, adjacent
pinnae irregular in width, from 6—12 mm wide, apices
broad-obtuse. Some veins forking close to the rachis
then irregularly throughout the lamina to anastomose
with adjoining veins to form very elongated narrow
areoles to 20 mm long, 0.5—1 mm in width, becoming
shorter and narrower close to the pinna apex; outer
veins running parallel to margin and all veins
terminating around pinna apex. Vein density across
mid-lamina c. 14—-16/10 mm.

Material
AMF134034a and counterpart 134034b,
Reserve Quarry.

Discussion

Ctenis sp. A differs from C. marniana (above)
by the more elongate pinnae and narrower elongate
areoles. However both forms occur in a grey siltstone
from the same horizon, which suggests that they
may belong to a single species bearing pinnae with
extremely variable form and venation pattern.

Order Bennettitales
Family incertae sedis

Genus Halleyoctenis Anderson and Anderson
1989

Type species
Halleyoctenis megapinnata (Anderson and
Anderson 1989).

The genus Halleyoctenis was erected by Anderson
and Anderson (1989) for cycad leaves distinguished by
very fine and closely-spaced veins that bifurcate and
radiate to the lamina margin. The cuticle preserved
on Molteno specimens of H. megapinnata has non-
aligned probably haplocheilic stomata that indicate
an affinity with the Bennettitales. Anderson and
Anderson selected as the genotype the leaf originally
described and illustrated by Shirley (1897, fig. 7a)
from the Ipswich Coal Measures of Queensland as
Pterophyllum multilineatum,|see also Shirley (1898)
and Walkom(1917)]. We have examined this specimen
(correct number QGS161) in the QGS collections in
Brisbane. The venation, where clearly preserved, is
not radiating but straight and parallel and terminates
at the pinna apex. The cuticle is not preserved. We

EZ

also examined the specimens UQF31692, 72854
and 9922 that were described by Webb (1980) in his
unpublished thesis and illustrated by Anderson and
Anderson (1989, p. 328, t. figs 5-8) as Halleyoctenis
multilineata. Two of these specimens show fine veins
(38-40/10 mm) running parallel to the apex and
compare well with Shirley’s original specimen. The
drawing of UQF31692 in Anderson and Anderson
(1989, p. 328, t. fig. 8) is not correct and should be
disregarded. From this evidence we conclude that
Shirley’s type specimen is not a Halleyoctenis and
a new type is required for that genus. We therefore
nominate a new genotype Halleyoctenis megapinnata
specimen BP/2/1817 as illustrated by Anderson and
Anderson (1989, p. 329, t. fig. 1 and Pl. 189, figs 1,
9, 10).

The material from Chile described in Herbst and
Troncoso (2000, pp 285-6; fig. 4G) as Pseudoctenis
multilineata would be best placed in Halleyoctenis
megapinnata. We agree with Anderson and Anderson
(1974, Table 3) and Herbst and Troncoso (2000)
that Pterophyllum multilineatum — based only on
Shirley’s type specimen — is better placed in the genus
Pseudoctenis.

Halleyoctenis brachypinnata Anderson and
Anderson 1989
Figures 25A—C

References

1989 Halleyoctenis brachypinnata, Anderson and
Anderson p. 328, Figs 1-4, Pls 191-194.

2003 Halleyoctenis brachypinnata, Anderson and
Anderson pp 344, 345, figs 1-2.

Description

Part and counterpart of a 95 mm long portion
of a frond with the base and apex missing. Rachis 3
mm wide, striated. Pinnae attachment semi-dorsal,
at c. 90° to rachis. Pinnae well-separated, opposite to
sub-opposite, short to elongated-oblong, increasing
in size distally along the rachis, from 12—25 mm
long and from 6—8 mm wide, base slightly expanded,
margin entire and apex obtuse. Veins emerging
straight from rachis, most forking close to base of
lamina and a few dividing again throughout the
lamina, radiating slightly and terminating around the
lamina margin and apex; vein density in mid-lamina
c. 44/10 mm.

Material
AMF134035, 134036, Reserve Quarry.

Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Discussion

The material available is the part and counterpart
of an incomplete specimen showing the lower and
mid-portion of the frond (Fig. 25A, B). It is placed
tentatively with H. brachypinnata from the Molteno
flora, but differs slightly by the more widely-spaced
pinnae that do not expand distally. It is distinct from
H. megapinnata by the smaller size and lower length/
breadth ratio of the pinnae and from H. symmetrica
by the oblong pinna shape and uncontracted pinna
bases. Anderson and Anderson (2003) suggest that H.
symmetrica may be generically distinct.

CONCLUSIONS

While cycadophyte fronds are infrequent in the
Nymboida flora they are diverse and are placed in
fifteen taxa in the cycadalean genera Pseudoctenis,
Moltenia and Ctenis and one in the bennettitalean
genus Halleyoctenis. Some forms compare well with
material from the Molteno Formation of South Africa
(Pseudoctenis fissa, P. sanipassiensis, Halleyoctenis
brachypinnata), La Ternera Formation of Chile and
the Barreal Formation of Argentina (Pseudoctenis
azcaratei) and the Brady Formation of Tasmania
(Pseudoctenis sp. cf P. strahanii). Fronds considered
distinct are described as the eight new species
Pseudoctenis nymboidensis, P. rigbyi, P. prolongata,
P. cursanervia, P. grandis, P. nettiana, Moltenia
Sparsispinosa and Ctenis marniana. Insufficiently
complete but clearly distinct material is described
as Pseudoctenis sp. A and P. sp. B and Ctenis sp. A.
Pterophyllum azcaratei is transferred to Pseudoctenis
azcaratei and Halleyoctenis megapinnata is
nominated as a new genotype for Halleyoctenis. It
is believed that the pinnae on all fronds are semi-
dorsally attached rather than laterally attached as in
some earlier descriptions.

ACKNOWLEDGEMENTS

W.B.K.H. was greatly assisted by his family over the
many years of collecting at the Nymboida quarries. A grant
from the Betty Mayne Research Fund has contributed to the
progress of this project.

REFERENCES

Anderson, J.M and Anderson, H.M. (1983). Palaeoflora of
southern Africa. Molteno Formation (Triassic). Vol.1.
Part 1. Introduction. Part 2. Dicroidium.. Balkema,
Rotterdam.

Proc. Linn. Soc. N.S.W., 129, 2008

Anderson, J.M and Anderson, H.M. (1989). Palaeoflora of
southern Africa. Molteno Formation (Triassic). Vol.2:
Gymnosperms (excluding Dicroidium). Balkema,
Rotterdam.

Anderson, J.M and Anderson, H.M. (2003). Heyday of the
gymnosperms: systematics and biodiversity of the
Late Triassic Molteno fructifications. Strelitzia 15,
1-398.

Artabe, A.E. (1985). Estudio systematico de la tafoflora
Triasica de Los Menucos, provincial de Rio Negro,
Argentina. Parte 2. Cycadophyta, Ginkgophyta y
Coniferophyta. Ameghiniana 22, 159-180.

Bonetti, M.I-R. (1968). Las especies del género
Pseudoctenis en la flora Triasica de Barreal (San
Juan). Ameghiniana 5, 433-446.

Bonetti, M.I.R. (1972). Las “Bennettitales” de la Flora
Tridsica de Barreal (Provincia Sa Juan). Revista

__ del Institute de Investigatién y Museo Argentino de
Ciences Naturales “Bernadino Rivadavia” 1(10),
307-332.

De Jersey, N.J. (1958). Macro and micro-floras of north-
eastern NSW. Journal and Proceedings of the Royal
Society of NSW 92, 83-89.

Du Toit, A.L. (1927). The fossil flora of the Upper Karoo
Beds. Annals of the South African Museum 22,
289-420.

Flint, J.C.E.,and Gould, R.E. (1975). A note on the fossil
megafioras of the Nymboida and Red Cliff Coal
Measures, southern Clarence-Moreton Basin. Journal
and Proceedings of the Royal Society of NSW 108,
70-74.

Frenguelli, F. (1950). Addenda a la flora del Gondwana
superior en la Argentina. Revista Asociacion
Geologico Argentina 5, 15-30.

Herbst, R. and Troncoso, A. (2000). Las Cycadophyta del
Triasico de las Formaciones La Ternera y El Puquén
(Chile). Ameghiniana 37(3), 283-292.

Hermsen, E.J., Taylor, T.N., Taylor, E.L. and Stevenson,
D.W. (2006). Cataphylls of the Triassic cycad
Antarcticycas schopfii and new insights into cycad
evolution. American Journal of Botany 93, 724-738.

Hill, K. and Osborne, R. (2001). Cycads of Australia.
Kangaroo Press, Kenthurst.

Hill, A., Playford, G. and Woods, J.T. (1965).

Triassic fossils of Queensland. Queensland
Palaeontographical Society, Brisbane. 1-32.

Holmes, W.B.K. (1982). The Middle Triassic flora from
Benolong, near Dubbo, central-western New South
Wales. Alcheringa 11, 165-173.

Holmes, W.B.K. (1992). Glossopteris-like leaves from
the Triassic of eastern Australia. In: Venkatachala,
B.S., Jain, K.P. and Awasthi, N. Eds. Proceedings
of the ‘Birbal Sahni Centenary Palaeobotanical
Conference’, Geophytology 22, 119-125.

Holmes, W.B.K. (2000). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 1. Bryophyta, Sphenophyta.
Proceedings of the Linnean Society of NSW 122,
43-68.

123

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Holmes, W.B.K. (2001). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 2. Filicophyta. Proceedings of
the Linnean Society of NSW 123, 39-87.

Holmes, W.B.K. (2003). The Middle Triassic flora of the
Basin Creek Formation, Nymboida Coal Measures,
New South Wales. Part 3. Fern-like foliage.
Proceedings of the Linnean Society of NSW 124,
53-108.

Holmes, W.B.K. and Anderson, H.M. (2005a). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 4.
Dicroidium. Proceedings of the Linnean Society of
NSW 126, 1-37.

Holmes, W.B.K. and Anderson, H.M. (2005b). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 5.
The Genera Lepidopteris, Kurtziana, Rochipteris and
Walkomiopteris. Proceedings of the Linnean Society
of NSW 126, 39-79.

Holmes, W.B.K. and Anderson, H.M. (2007). The
Middle Triassic flora of the Basin Creek Formation,
Nymboida Coal Measures, New South Wales. Part 6.
Ginkgophyta. Proceedings of the Linnean Society of
NSW 128, 155-200.

ICBN. (2001). International Code of Botanical
Nomenclature, Saint Louis Code (2000). W. Greuter
et al. Eds. Koeltz Scientific Books, KGnigstein,
Germany.

Jones, D.L. (1993). Cycads of the World. Reed Books,
Chatswood.

Jones, O.A. and De Jersey, N.J. (1947). The flora of the
Ipswich Coal Measures —morphology and floral
succession. Papers of the Department of Geology,
University of Queensland. New Series 3, 1-88.

Joshi, R., Rigby, J.F. and Nautiyal, D.D. (2004)
Reinvestigation of some Mesozoic leaves of
Walkom’s collection. Pp 189-197. In Professor
D.D.Pant Memorial Volume, Vistas in Palaeobotany
and Plant Morphology: Evolutionary and
Environmental Perspectives. P.C. Shrivastava Ed.
U.P. Offset, Lucknow.

Johnston, R.M. (1888). Systematic account of the geology
of Tasmania. Government Printer, Hobart.

Klavens, S.D., Taylor, E.L., Krings, M. and Taylor, T.M.
(2003). Gymnosperms from the middle Triassic of
Antarctica: the first structurally preserved cycad
pollen cone. International Journal of Plant Science
164, 1007-1020.

Lele, K.M. (1956). Plant fossils from Parsora in the South
Rewa Gondwana Basin. Palaeobotanist 4, 23-34.

Lindley, J. and Hutton W. (1834). The fossil flora of Great
Britain, or figures and descriptions of the vegetable
remains found in a fossil state in this country.
Ridgeway and Sons, London.

Menendez, C.A. (1951). La flora de la Formacién
Liantenes (Provincia de Mendoza). Revista Instituto
Nacionale de Investigaciones en Ciencias Naturales
(Botanica) 2, 147-261.

124

Ottone, E.G. (2006). Plantas triasicas del Grupo
Rincon Blanco, provincia de San Juan, Argentina.
Ameghiniana 43, 477-486.

Retallack, G.J. (1977). Reconstructing Triassic vegetation
of eastern Australia: a new approach for the
biostratigraphy of Gondwanaland. Alcheringa 1,
247-278. Alcheringa-fiche 1, G1—J16.

Retallack, G.J., Gould, R.E. and Runnegar, B. (1977).
Isotopic dating of a middle Triassic megafossil flora
from near Nymboida, north-eastern New South
Wales. Proceedings of the Linnean Society of NSW
101, 77-113.

Rigby, J.F. (1977). New collections of plants from the Esk
Formation, south-eastern Queensland. Queensland
Government Mining Journal 78, 320-325.

Russel, N.J. (1994). A palaeothermal study of the
Clarence-Moreton Basin. Australian Geological
Survey Organisation Bulletin 241, 237-276.

Seward, A.C. (1903). Fossil floras of Cape Colony. Annals
of the South African Museum 4, 1-122.

Shirley, J. (1897). Two new species of Pterophyllum.
Proceedings of the Royal Society of Queensland 12,
89-91.

Shirley, J. (1898). Additions to the fossil flora of
Queensland. Queensland Geological Survey Bulletin
7, 19-25.

Singh, R. and Radha, P. (2006) A new species of Cycas
from the Malabar Coast, Western Ghats, India.
Brittonia 58, 119-123.

Smoot, E.L., Taylor, T.N. and Delevoryas, T. (1985).
Structurally preserved plants from Antarctica. 1.
Antarcticycas, gen. novy., a Triassic stem from the
Beardmore Glacier area. American Journal of Botany
72, 1310-1423.

Taylor, T.N. and Taylor, E.L. (1993). The biology and
evolution of fossil plants. Prentice Hall, New Jersey.

Troncoso, A. and Herbst, R. (2000). La tafoflora Triasicas
del Cajon Troncoso, Alta Cordillera del Maule,

7" Region, Chile. Revista del Museo Argentino de
Ciencias Naturales, n.s. 2(2), 137-144.

Walkom, A.B. (1917). Mesozoic floras of Queensland.
Part 1 (contd.) The flora of the Ipswich and
Walloon Series. (d) Ginkgoales, (e) Cycadophyta,
(f) Coniferales. Queensland Geological Survey
Publications 259, 1-49

Walkom A.B. (1924). On fossil plants from Bellevue, near
Esk. Memoirs of the Queensland Museum 8, 77-92.

Walkom A.B. (1925). Notes on some Tasmanian Mesozoic
plants.Part 1. Papers and Proceedings of the Royal
Society f Tasmania 1924, 73-89.

Walkom A.B. (1928). Fossil plants from the Esk district,
Queensland. Proceedings of the Linnean Society of
NSW 53, 458-468.

Webb, J.A. (1980). Aspects of the palaeontology of
Triassic continental sediments in South-East
Queensland. Unpublished Thesis. Geology
Department, University of Queensland.

White, M.E. (1990). The Nature of Hidden Worlds. Reed
Books, Balgowlah.

Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure. 1. A, B. Pseudoctenis fissa Du Toit. A. AMF133968; B. AMF133963, Coal Mine Quarry. Scale
bar =1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 1s

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 2. A. Pseudoctenis fissa Du Toit. AMF133962, Coal Mine Quarry. Scale bar = 1 cm.

126 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 3. A-F. Pseudoctenis fissa Du Toit. A. AMF (133966), Reserve Quarry; B. AMF 133960, Coal Mine
Quarry; C. AMF133965, Coal Mine Quarry; D. AMF133961, Coal Mine Quarry; E. AMF 133967, Reserve
Quarry; F. AMF133964, Coal Mine Quarry. Scale bars = 1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 27

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 4. A. Pseudoctenis nymboidensis Holmes and Anderson sp. nov. Holotype AMF 133969, Coal Mine
Quarry. Scale bar = 5 cm.

128 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 5. A, B. Pseudoctenis nymboidensis Holmes and Anderson sp. nov. A. AMF 133972, Coal Mine
Quarry; B. AMF133981, Reserve Quarry. Scale bars = 1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008

1

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 6. A-E. Pseudoctenis nymboidensis Holmes and Anderson sp. nov. A. AMF133976, Coal Mine
Quarry; B. AMF126860, Coal Mine Quarry; C. AMF133977, Coal Mine Quarry; D. AMF133980, Reserve
Quarry. Scale bars = 1 cm.

30 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 7. A. Pseudoctenis nymboidensis Holmes and Anderson sp. nov. AMF 133973, Reserve
Quarry. Scale bar = 1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 Bal

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 8. A. Pseudoctenis nymboidensis Holmes and Anderson sp. nov. AMF 133970, Coal Mine Quarry.
Scale bar = 5 cm.

132 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 9. A. Pseudoctenis rigbyi Holmes and Anderson sp. nov. Holotype. AMF 133982, Coal Mine Quarry.
Scale bar = | cm.

Proc. Linn. Soc. N.S.W., 129, 2008 133

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 10. A. Pseudoctenis rigbyi Holmes and Anderson sp. nov. AMF 133983, Coal Mine Quarry. Scale
loeie = Il Gan,

134 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 11. A. Pseudoctenis sanipassiensis Anderson and Anderson. AMF 133987, Coal Mine
Quarry. Scale bar = 5 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 135

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 12. A—C. Pseudoctenis sanipassiensis Anderson and Anderson. A. AMF 133989; B. AMF133990;
C. AMF133994. Coal Mine Quarry. Scale bars = 1 cm.

136 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 13. A. Pseudoctenis prolongata Holmes and Anderson sp. nov. Holotype. AMF133996, Coal Mine
Quarry. Scale bar = | cm.

Proc. Linn. Soc. N.S.W., 129, 2008 37

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

te

Figure 14. A — C. Pseudoctenis prolongata Holmes and Anderson sp. nov. A. AMF133999, Coal Mine
Quarry; B. AMF133998, Coal Mine Quarry; C. AMF133400, Reserve Quarry. Scale bars = 1 cm.

138 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 15. A-G. Pseudoctenis nettiana Holmes and Anderson sp. nov. A. Holotype. AMF134003, Coal
Mine Quarry; B. AMF134004, Coal Mine Quarry; C. AMF134005, Coal Mine Quarry; D. AMF134008,
Coal Mine Quarry; E. AMF134011, Reserve Quarry; F. AMF134002, Coal Mine Quarry; G. AMF134007,
Coal Mine Quarry. Scale bars = 1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 139

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 16. A, B. Pseudoctenis grandis Holmes and Anderson sp. nov. A. Holotype. AMF134012; B.
AMF 134013. Coal Mine Quarry. Scale bars = 1 cm.

140 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 17. A, B. Pseudoctenis cursinervia Holmes and Anderson sp. nov. A. Holotype AMF 134016; B.
AMF 134017, Coal Mine Quarry. Scale bars = 1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 141

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 18 A—D. Pseudoctenis azcaratei (Herbst and Troncoso) Holmes and Anderson comb. nov. A.
AMF 134020; B. AMF134018; C. AMF134022; D. AMF134019. Coal Mine Quarry. Scale bars = 1 cm.

142 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 19. A. Pseudoctenis azcaratei (Herbst and Troncoso) Holmes and Anderson comb. nov.
AMF 134021, scale bar = 5 cm.; B. Pseudoctenis sp. cf. Pseudoctenis strahanii (Johnston) Anderson and
Anderson, AMF 134023, scale bar = 1 cm.; C, D. Pseudoctenis sp. A; C. AMF 134024, scale bar = 1 cm.;
D. AMF 134025, scale bar = 5 cm. All Coal Mine Quarry.

Proc. Linn. Soc. N.S.W., 129, 2008 143

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 20. A. Moltenia sparsispinosa Holmes and Anderson sp. nov. A. Holotype. AMF134028,
Coal Mine Quarry. Scale bar = 5 cm.

144 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 21. AD. Moltenia sparsispinosa Holmes and Anderson sp. nov. A. AMF 134033, paratype, Reserve
Quarry; B. AMF134030, Reserve Quarry; C. AMF134029, Coal Mine Quarry; D. AMF134027, Coal Mine

Quarry. Scale bars = 1 cm.

Proc. Linn. Soc. N.S.W., 129, 2008 145

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Tats, .
a i ate

Figure 22. A—C. Ctenis marniana Holmes and Anderson sp. nov. A, B. Holotype AMF 134031; C.
AMF 134032, counterpart of Holotype, Reserve Quarry. Scale bars = | cm.

146 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 23. A. Ctenis marniana Holmes and Anderson sp. nov. AMF134032; B. Ctenis sp A.
AMF 134034. Reserve Quarry. Scale bars = | cm.

Proc. Linn. Soc. N.S.W., 129, 2008 147

TRIASSIC FLORA FROM NYMBOIDA - CYCADOPHYTA

Figure 24. A. Pseudoctenis sp. B. AMF 134026, Coal Mine Quarry; B, C. Ctenis sp. A. AMF134034,
Reserve Quarry. Scale bars = | cm.

148 Proc. Linn. Soc. N.S.W., 129, 2008

W.B.K. HOLMES AND H.M. ANDERSON

Figure 25. A-C. Halleyoctenis brachypinnata Anderson and Anderson. A. AMF 134036; B, C.
AMF 134035. Reserve Quarry. Scale bars = | cm.

Proc. Linn. Soc. N.S.W., 129, 2008 149

i” i yey had hy Min mace a

=

ee ee. generg

UE TD yromEd .

150 j PONS Rid WAZ be

Habitat Preferences of Port Jackson Sharks, Heterodontus
portusjacksoni, in the Coastal Waters of Eastern Australia.

Davip M. PowTER AND WILLIAM GLADSTONE

School of Environmental and Life Sciences, Ourimbah Campus, University of Newcastle, Ourimbah NSW

2258 (david.powter@newcastle.edu.au)

Power, D.M. and Gladstone, W. (2008). Habitat Preferences of Port Jackson Sharks, Heterodontus
portusjacksoni, in the coastal waters of eastern Australia. Proceedings of the Linnean Society of New
South Wales 129, 151-165.

The habitat preferences of juvenile and adult Heterodontus portusjacksoni and ovipositing females
were determined from three locations on the central and southern coast of New South Wales. Adults use
shallow coastal rocky reefs in July-November for mating and oviposition, whilst juveniles occupy a seagrass
nursery in a large coastal embayment. The sand/reef interface on the lee side of reefs was preferred by both
sexes, probably as a refuge against strong water movements. Adult females also preferred rocky gutters
when available, possibly as a male avoidance strategy. Preferred oviposition sites were narrow, shallow
crevices (single capsules) or deep, narrow crevices (multiple capsules) which afforded protection against
mechanical dislocation and/or predation. Juveniles exhibited a strong preference for the seagrass bed edge
within a shallow nursery area. The visual complexity of this habitat combined with the juvenile’s disruptive
colouration may provide a refuge from predation, whilst proximity to the seagrass may provide ease of
access for foraging. At a large scale, juveniles preferred areas of moderate slope within the nursery that
provided protection from strong water movement. This study highlights the need for quantitative studies
addressing habitat preferences and a consideration of use-specific factors to fully understand the selection

of habitat by elasmobranchs.

Manuscript received 20 October 2007, accepted for publication 6 February 2008.

KEYWORDS: elasmobranchs, habitat use, habitat preferences, nursery area, oviposition.

INTRODUCTION

‘Habitat’ is the location or environment in which
an organism lives and is determined by a complex
interaction of physical and biotic factors (Sims
2003). Consequently, the range of habitats utilised by
elasmobranchs and the factors contributing to their
selection are many and diverse (Last and Stevens
1994; Goldman and Anderson 1999; Matern et al.
2000; Peach 2002). For example, Heithaus et al.
(2002) found that prey availability was important
to tiger sharks, Galeocerdo cuvier, while bat rays,
Myliobatis californica, made daily movements
between areas of different water temperature to
thermoregulate and influence their metabolic rates
(Matern et al. 2000). Consequently an understanding
of habitat requirements is important for management
and conservation. However, a complete understanding
of the importance of habitat to a species requires the
separation of habitat use from habitat preference
(Carraro and Gladstone 2006). Use relates to the
habitats in which individuals occur, whilst habitat

preference is the level of utilisation of the habitat as a
function of its relative availability.

Few detailed, quantitative studies ofelasmobranch
habitat utilization and preferences have been made
and most have only involved overlaying movement
data over gross habitat characteristics (Simpfendorfer
and Heupel 2004). Despite this there is a clearly
recognized need for a detailed understanding of
habitat requirements for effective conservation
and management, such as the selection and design
of marine protected areas and assessments of the
potential impacts of habitat degradation.

Many sharks show ontogenetic differences in
habitat utilisation (Siindstrom et al. 2001). In most
cases adult and juvenile populations are separated
spatially, with juveniles and neonates occupying
distinct nursery areas associated with decreased
predation risks (Heupel and Hueter 2002) and
possibly abundant food (Castro 1993). Juvenile and
neonate blacktip sharks, Carcharhinus limbatus, in
Terra Ceia Bay, Florida inhabited a core portion of
their nursery area to avoid predation but made regular

HABITAT PREFERENCES OF PORT JACKSON SHARKS

a Cabbage Tree Harbour

Terrigal Haven

Tasman

Dent
Rock

Murray’s
Sandline

50
Kilometers

Figure 1: Map of the central and south coast of NSW, Aus-
tralia showing the location of the study sites. Inset map

macro- and microhabitat features affecting
habitat utilisation by H. portusjacksoni at
adult reproductive grounds, the juvenile
nursery area and oviposition sites and to
elucidate any temporal, spatial or sex-
based patterns. The second goal was to
determine the habitat preferences of adult
and juvenile H. portusjacksoni on the basis
of utilised resting positions.

MATERIALS AND METHODS

Study sites

Observations on adults occurred at
Cabbage Tree Harbour, Terrigal Haven
and Dent Rock on the central and south-
eastern coast of New South Wales (NSW),
Australia (Fig. 1). Oviposition sites
were studied at the latter two locations.
The juvenile nursery area was located at
Murray’s Sandline, Jervis Bay, NSW (Fig.

shows the section of the east Australian coast depicted inthe 1).

main figure.

excursions outside that area which were believed to
be feeding forays (Heupel et al. 2004).

Adult Heterodontus portusjacksoni are demersal
sharks that use a range of habitats throughout their life
cycle (Powter 2006). Adult males and females migrate
annually in the austral winter-spring (July-November)
from deep offshore waters to shallow coastal rocky
reefs for mating, oviposition and feeding. McLaughlin
and O’Gower (1971) and O’Gower (1995) described
the physical characteristics of resting sites as highly
variable, however their studies were not quantitative
and diel variations were not investigated. Additionally,
habitat was considered in terms of use alone and did
not relate habitat utilisation to availability. Similarly
H. portusjacksoni oviposition areas were only
described qualitatively (McLaughlin 1969). Viable
capsules were securely wedged between rocks or in
rock crevices on shallow rocky reefs in sheltered bays
in depths from 1 to 20 m and occasionally on sheltered
areas of some inshore reefs (McLaughlin 1969; Rodda
2000). However, no quantitative understanding
exists of habitat preferences affecting the selection
of oviposition areas by H. portusjacksoni or, in fact,
for any oviparous elasmobranch. Finally, it remains
unclear whether juvenile H. portusjacksoni utilise
distinct nursery areas and, if so, what the specific
features of these may be.

The first goal of this study was to determine the

S52

The three adult sites consisted of
rocky reefs adjoining barren sand flats,
with the junction between the two termed

the interface. All sites were divided into macrohabitat
zones on the basis of topography, substrate type and
biotic characteristics.

Cabbage Tree Harbour (33°16’ S, 151°34’ E) isa
shallow embayment, with water temperatures ranging
from 15° C (mean + S.E.; 16.6° C + 0.37; n=7) in July
to 18°C (16.6° C+ 0.40; n=5) in November during this
study. The surveyed site extended for 270 m and was
divided into three macrohabitat zones (west to east):
(1) the wall zone (3.5—5.4 m deep) was characterised
by a vertical rock wall with urchin-dominated barren
boulders (Edgar 2001) and kelp (Ecklonia radiata
and Phyllospora comosa) habitats at the far eastern
end; (2) the boulder zone (5.4—6.0 m) was a gently
sloping boulder field which is largely urchin barrens
habitat with some E. radiata at the western edge; (3)
the gutter zone (6.0—8.3 m) had a vertical rock wall
above a steeply sloping boulder field with five narrow
rock gutters ranging from 5—15 m long in 6-6.5 m
depth at the eastern end.

Terrigal Haven is a shallow rocky reef in a small
coastal embayment, with water temperatures ranging
from 15° C (mean + S.E.; 16.7° C + 0.14; n=25) in
July to 20° C (18.5° C + 0.22; n=19) in November
during this study. The surveyed reef extended for
280 m and was divided into three macrohabitat zones
(east to west): (1) the shallow zone (3.88.5 m) was

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

a steeply sloping boulder reef characterised by urchin
barrens habitat with sparse E. radiata; (2) the kelp
zone (8.5—-10.6 m) was similar, however the lower
third of the reef (adjoining the interface) was covered
in a dense bed of E. radiata; (3) the barrens zone
(10.6—12.2 m) sloped gently and was dominated by
urchin barrens habitat.

Dent Rock (35°04’ S, 150°41’ E) is located on the
south coast of NSW, with water temperatures ranging
from 14° C (mean + S.E.; 14.3° C + 0.25; n=4) in
July to 20° C (18.3° C + 1.03; n=4) in November
during this study. It is an elliptically shaped, shallow
boulder reef surrounded by sand approximately 160
m offshore within the protected waters of Jervis Bay.
Covering an area of 1935 m’, the reef was divided
into four macrohabitat zones: (1) quadrant 1 (5.3-6
m) sloped steeply and consisted of urchin barrens
habitat replaced by E. radiata nearer the sand, with
several low overhangs and sand-bottomed rock
gutters; (2) quadrant 2 (5.2-5.3 m) sloped gently, with
E. radiata near the sand and around several sand-
bottomed gutters and barrens habitat in the upper
reef; (3) quadrant 3 (5.2-5.5 m) was a gently sloping
boulder reef with barrens habitat; (4) quadrant 4 (5.5-
6 m) increased in slope from north to west and was
mainly urchin barrens habitat with some E. radiata
near the western edge.

Murray’s Sandline (35°08’S, 150°46’ E) is located
within Jervis Bay. The site is a shallow seagrass bed
located approximately 400 m offshore and within 2 km
of the bay’s mouth. The seagrass bed contained a mix
of Zostera capricorni and Halophila ovalis with small

patches of Posidonia australis. The eastern region of
the bed extended for approximately 750 m and had
a continuous cover of seagrass in a depth range of
11.4-4.2 m (west to east). At the western end the bed
sloped steeply. The western region was approximately
750 m in length in depths of 11.4-6 m (east to west).
Seagrass cover decreased and algae cover increased
to the west. The sloping seagrass bed in both regions
was divided into three macrohabitat zones. At the
base of the bed was a gently sloping, barren sand area
(hereafter called the sand zone) separated from the
seagrass bed (the seagrass zone) by the interface zone.
The interface zone was a | m wide transition between
seagrass and sand zones comprised of a fragmented
cover of seagrass.

Surveys

Surveys involved underwater visual census
(UVC) surveys conducted at each location. To
ensure maximum visual coverage, two experienced
divers swam parallel to the reef/sand or seagrass/
sand interface and approximately 1-2 m above the
substrate. Only sharks observed resting when first
sighted were utilised in this study to avoid possible
bias from sharks in transit between locations.

Surveys commenced at Terrigal Haven in January
2002, at Cabbage Tree Harbour in July 2002 and at
Dent Rock and Murray’s Sandline in December 2002.
Surveys concluded at all sites in December 2005.
Terrigal Haven was surveyed twice weekly (one day
and one night) during the adult onshore reproductive
period (July to November; n=131 surveys) (hereafter

Table 1: Microhabitat variables and measurement methods for adult and juvenile resting
site locations. Variables and measurement methods are identical unless specified as.being for

adult* or juvenile* resting sites.

Habitat Variable

Distance (from interface)
Depth (m)

Temperature (°C)

Reef slope

Rock Cover (%)°
Algae/kelp cover (%)

Sand cover (%)

Seagrass cover (%)*

Total vegetation cover (%)*

Sediment grain size composition*

Proc. Linn. Soc. N.S.W., 129, 2008

Method

Visual estimate (nearest 1 m‘; 1 cm*)
Dive computer (nearest 10 cm)

Dive computer (nearest 0.1°C)
Depth/distance measures

4m’ visual quadrat

4m” visual quadrat’; 0.25m?’ photoquadrat*
4m’ visual quadrat’; 0.25m?* photoquadrat*
0.25m* photoquadrat

0.25m? photoquadrat

Sediment analysis

153

HABITAT PREFERENCES OF PORT JACKSON SHARKS

called the ‘season’) and at least monthly outside the
season (n=45). Cabbage Tree Harbour was surveyed
four times per month (2 day and 2 night) during the
season (n=57) and monthly at other times (n=15).
Dent Rock (n=33) and the eastern region of Murray’s
Sandline (n=29) were surveyed monthly during
daylight hours. In addition, every three months from
September 2003 to December 2005, surveys (n=10)
were conducted over the 1.5 km length of the seagrass
bed at Murray’s Sandline.

Habitat Variables

Habitat usage was assessed at two levels.
Macrohabitat was the general landscape-scale features
(macrohabitat zones referred to above) inhabited by
H. portjacksoni, whilst microhabitat variables (Table
1) were the finer-scale elements operating at the scale
of individual sharks (Hall et al. 1997).

The macrohabitat zone occupied by all resting
sharks at the time of first sighting was recorded at all
sites. The total number of resting sharks for which
habitat data was recorded in each year at Terrigal
Haven (2002-2005) was 32, 24, 17 and 9; Cabbage
Tree Harbour (2003-2005) was 26, 21 and 9; Dent
Rock (2003-2005) was 33, 27 and 26; and, Murray’s
Sandline (2003-2005) was 169, 105 and 56.

The microhabitat variables (Table 1) of the
positions used by resting adult sharks were quantified
in a 2 x 2 m area (delimited by a quadrat) centred
on the resting sharks and in haphazardly selected,
unutilized positions of the same area. Microhabitat
features were quantified at Terrigal Haven (35 resting
sharks, 106 unutilised positions), Cabbage Tree
Harbour (38 resting sharks, 75 unutilised positions),
and Dent Rock (27 resting sharks, 121 unutilised
positions). Microhabitat features of the resting
positions of juvenile sharks (Table 1) in the seagrass
nursery were determined at haphazardly selected
(n=90) and utilised resting (n=20) positions spread
over both regions of the seagrass bed using a 0.25 m?
digital photographic quadrat (digital camera mounted
on a preset quadrat frame). The image was analysed
for percent cover (Table 1) using a one hundred point
grid method, in which a 10 x 10 square grid was
superimposed over the photograph and the percentage
contribution of each feature within each square was
visually estimated and individually summed (Foster
et al. 1991). Sediment samples were also taken by
drawing a 50 mm long by 30 mm diameter plastic
container across the sediment to a maximum depth
of 15 mm. The samples were subsequently dried at
60° C for 48 hr and sorted through a stacked series
of graded sieves (1 mm, 500 um, 212 um, 63 um,
<63 um) in a sediment shaker for 10 min. Sample

fractions were weighed individually.

Egg capsules were found throughout the sites at
Terrigal Haven and Dent Rock, however habitat data
at the oviposition sites was only collected in 2004
and 2005. Primary oviposition areas were defined as
concentrated areas (up to 15 m7’) with greater than 10
egg capsules. The habitat and microhabitat features
(depth; reef slope; rock size; crevice size; and, crevice
depth) at the site of each individual egg capsule and
haphazardly selected, unutilised positions in both
Oviposition areas and other portions of the sites were
recorded at Terrigal Haven (14 eggs; 30 unutilised
positions) and Dent Rock (17 eggs; 40 unutilised
positions).

Data Analysis

The macrohabitat zone occupied by individual
resting H. portusjacksoni was recorded during
each survey. Survey data was pooled by factor (e.g.
zone, year, sex, diel period) and Likelihood Ratio
(LR) tests were used to test the null hypothesis for
each site that there was no significant difference in
the proportion of resting sharks that utilised each
macrohabitat zone. Separate LR tests were conducted
for each year, sex and diel period for adults and for
each year for juveniles. G-tests were used for pair-
wise comparisons of significant LR results (Sokal
and Rohlf 2003). The heterogeneity of egg capsule
distribution was tested with G-tests by comparing
the number of viable capsules located in the primary
Oviposition areas and other locations within each
site. The viability of capsules was determined on
the absence of predation or other physical damage
(Rodda 2000; Powter 2006) or the visible presence of
an embryo inside the capsule.

Microhabitat characteristics of individual
adult and juvenile resting positions and the site
of oviposited egg capsules were examined using
Principal Components Analysis (PCA) in PRIMER 5
(PRIMER-E Ltd, UK). Prior to analyses, draftsman
plots were utilised to ensure that variables were not
highly inter-correlated (i.e. R > 0.95) (Clarke and
Warwick 2001). Proportions were arcsine transformed
and distances were log, transformed before analysis
(Sokal and Rohlf 2003).

The relative availability of each macrohabitat
type was quantified from scaled underwater maps of
each site’s terrain (Powter 2006). Habitat preferences
were determined from resource selection ratios
(Manly et al. 1993) using the formula w, = o, / p,,
where w, is the preference score for habitat category 7,
o, is the proportion of habitats used in category 7 and
p, is the proportional availability of habitat category
i, Preference scores were standardised (B,) to sum to

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

1, by dividing each preference score (w,) by the sum
of the preference scores for all habitat categories.
The critical value of B (B,,,,) related to the number
of habitats as follows: B, < (1/number of habitats)
indicated avoidance and B. > (1/number of habitats)
indicated preference.

The null hypothesis that H. portusjacksoni
selected habitats at random at each study site was
tested using the y test with (n-1) degrees of freedom
(Manly et al. 1993):

= 23) in( Ze )

and pair-wise tests of habitat selection ratios were
conducted to determine any significant differences (H,
awl) using the y’ test with 1 degree of freedom:

Dyas (#, a w,F

4 var, — W,)

and

peer teil 2))) Pom wei =e;)
RMR END py Op?

where, uw, is the number of sharks in habitat 7; U is the
total number of observations of sharks; o., 0, is the
proportion of habitats used in category 7 or /; and, p.,
P, is the proportional availability of habitat category
i or j.

RESULTS

Adult Macrohabitat Utilisation
The macrohabitat zone utilised by resting sharks

Table 2: Likelihood Ratio test (LR) results of macrohabitat zone utilisation for
resting adult H. portusjacksoni at Terrigal Haven, Cabbage Tree Harbour and

Dent Rock by year and sex.

differed significantly at Terrigal Haven from 2003 to
2005 (Table 2). In 2002, the greatest proportion of
sharks was located in the shallow zone, however this
proportion declined from 2003 to 2005, with a shift to
the barrens zone (Fig. 2a). There was no evidence of
differential use of macrohabitat zones by males and
females at Terrigal Haven (Table 2).

At Cabbage Tree Harbour there was no significant
difference in the use of macrohabitat zones across
survey years or by sex (Table 2, Fig. 2b). There were
significant differences in the use of macrohabitat
zones at Dent Rock across the survey years, but not
across sexes (Table 2, Fig. 2c). In 2003, there was a
greater proportion of resting sharks in quadrant 2 and
the lowest relative proportion of sharks in quadrant 3
(Fig. 2c). The proportion of sharks in quadrant 1 in
2003 was also less than in 2005.

Diel utilisation of macrohabitat zones at Terrigal
Haven differed only in 2004 (Table 3, Fig. 3a). In
2004 there were significantly more resting sharks
in the shallow zone at night than during the day (G-
test, d.f=1, P=0.02). Diel utilisation of macrohabitat
zones at Cabbage Tree Harbour differed only in
2005 (Table 3, Fig. 3b). However, only one resting
H. portusjacksoni was recorded at night at Cabbage
Tree Harbour in that year. Nocturnal surveys were not
conducted at Dent Rock.

Adult Microhabitat Utilisation

The PCI axis of the Principal Components
Analysis (PCA) biplot for Terrigal Haven represents a
gradient of decreasing cover of sand (left to right) and
indicates selection of resting positions with moderate
sand cover (55.6 + 3.29%; mean + SE) (Table 4, Fig.
4a). The PC2 axis represents a gradient of increasing
reef slope (bottom to top) and indicates selection of
resting positions of low to moderate reef slope. The
tight clustering of resting positions compared to
the unutilised positions, suggests sharks were very
selective in their choice of resting position.

The PC1 axis of
the PCA biplot for
Cabbage Tree Harbour
represents a gradient
of decreasing cover of
rock (left to right) and

Terrigal Haven Cabbage Tree Dent Rock indicates selection of

Harbour resting positions with

Comparison wR P LR P LR P moderate to high rock
cover (87.0 + 4.50%)

Zone x Year Diets L002 yin 2 0.379 14.94 0.021 (Table 4, Fig. 4b). The
PC2 axis represents a

Zone x Sex 3.88 0.143 0.39 0.820 6.17 0.104 gradient of decreasing

Proc. Linn. Soc. N.S.W., 129, 2008

kelp cover (bottom to

155

HABITAT PREFERENCES OF PORT JACKSON SHARKS

ral

2002 2003 2004 2005
Year

(c)

100 -

ee 124
o
=

E690
2
wo

oO 25 4

2003

(b)
100 5

73 +

Percentage
iSai
oO
1

2003 2004 2005

Figure 2: Percentage of resting adult H. portusjacksoni by year and macrohabitat zone at (a) Terrigal
Haven (black bars = shallow; grey bars = kelp; white bars = barrens), (b) Cabbage Tree Harbour (black
bars = wall; grey bars = boulder; white bars = gutter) and (c) Dent Rock (black bars = quadrant one;
grey bars = quadrant two; white bars = quadrant three; stippled bars = quadrant four).

top) and indicates selected resting positions had low
to moderate kelp cover (5.6 + 2.04%). The PCA biplot
for Dent Rock (Fig. 4c) displays a similar pattern to
Cabbage Tree Harbour, with resting positions having
moderate to high rock cover (48.0 + 5.09%) and low
to moderate kelp cover (16.5 + 3.02%).

Adult Habitat Preferences

Although the interface accounted for only 5%
of the available habitat at Terrigal Haven, resting
adult H. portusjacksoni exhibited a highly significant
preference for this habitat (Table 5). This pattern
was consistent across all years and both sexes. The
sand and kelp habitats were avoided in all years (all
B<0.25).

Table 3: Likelihood Ratio test (LR) results for the number of resting adult H. portusjacksoni
by diel period and macrohabitat zone at Terrigal Haven (TH) and Cabbage Tree Harbour

(CTH) by year. *P>0.05; * P<0.05; ** P<0.01

2002 2003
Site Zone % Day LR % Day
TH Shallow 35.0 339
Kelp 20.0 2.84 66.7
Barrens 66.7 46.7
CTH Wall 50.0
Boulder 0.0
Gutter 86.4

2004 2005

LR %Day LR %Day LR
0 0

0.92" 100 DOs 0 0.87"
66.7 37.5
100.0 100.0

1.33" 0.0 0.69" 0.0 6.28"
70.0 100.0

156

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

(a)
100 =
o o B
o 75 o ft
= a rs :
o 1 ® cs
oO j oO 2
5 254 5
a : cl i
0 = =a *
Day Day Night} Day Night | Day Night |
| |
Shallow Kelp Barren Gutters | Boulder Wall |
Macrohabitat zone and diel period Macrohabitat zone and diel period

Figure 3: Diel distribution of resting adult H. portusjacksoni by year and zone at (a) Terrigal Haven and
(b) Cabbage Tree Harbour for 2002 (black bars), 2003 (grey bars), 2004 (white bars) and 2005 (stippled

bars).

PC 2
PC2

4
26 -20 -16 -10 05 0 O05 10 15 20 26 30
PC 1

PC 2

PC2

Figure 4: PCA biplots of microhabitat variables utilised by resting H. portusjacksoni (closed circles) and
random unutilised locations (open diamonds) for adults at (a) Terrigal Haven (PC1, sand cover; PC2,
reef slope), (b) Cabbage Tree Harbour (PC1, rock cover; PC2, kelp cover) and (c) Dent Rock (PC1, rock
cover; PC2, kelp cover) and juveniles at (d) Murray’s Sandline (PC1, 500 um sediment fraction; PC2,

seagrass cover).

Proc. Linn. Soc. N.S.W., 129, 2008 ley)

HABITAT PREFERENCES-OF PORT JACKSON SHARKS

Table 4: PCA results for microhabitat variables utilised by resting adult H. portusjacksoni at
Terrigal Haven, Cabbage Tree Harbour and Dent Rock and juveniles at Murray’s Sandline.
The variables with the highest eigenvalues for each axis are shown.

Site PC Variable
1 Sand Cover
Terrigal
Bacon 2D Reef Slope
3 Kelp Cover
Cabbage 1 Rock Cover
ee 2 Kelp Cover
sap roue 3 Reef Slope

1 Rock Cover
Dent Rock 2
3 Reef Slope

Kelp Cover

1 500 um Sediment
Murray’s

Sandline 2 Seagrass Cover

3 Reef Slope

At Cabbage Tree Harbour the gutter habitat was
significantly preferred by resting adult female H.
portusjacksoni in all years (Table 5). Resting males
generally exhibited a preference for the interface
habitat with the exception of 2003. However, the
number of resting males at Cabbage Tree Harbour
was very low, with only seven observed during the
three study years. No resting individuals of either sex
were ever observed in the sand or kelp habitats.

At Dent Rock adult female H. portusjacksoni
generally preferred resting in the gutter habitat, whilst
males mainly preferred the interface habitat (Table 5).
However, there were some variations to this pattern.
Although there was a moderate preference for the
gutter habitat (6=0.32) by resting males in 2005,
there was a significantly greater preference for the
interface habitat (B=0.68; y’-test, d.f.=3, P<0.001).
In 2004 and 2005, the preferred resting habitat for
adult female H. portusjacksoni was the gutter habitat.
In both years this habitat was significantly preferred
over the remaining four habitat types. However, in
2003 female habitat preferences were approximately
equally shared between the interface (B =0.51) and the

158

Pigenvccion Variation Cumulative
(%) Variation (%)
-0.603 39.8 39.8
0.630 32.2 72.0
-0.776 20.5 QDS
-0.613 45.0 45.0
-0.713 28.8 73.8
-0.762 16.3 90.1
-0.623 46.6 46.6
-0.860 24.6 TAD
0.984 20.3 91.4
-0.468 33.6 33.6
0.211 27.6 61.2
0.623 . 13.7 74.9

gutter (B=0.49) habitats and were not significantly
different (7’-test, d.f=3, P=0.60). A small preference
for the interface habitat was also exhibited by females
in 2005 (B=0.23), but the preference for the gutter
habitat was significantly greater (B—0.77; ’-test,
d.f=3, P<0.001).

Juvenile Macrohabitat Utilisation

Overall there was no significant difference in
the proportion of resting juvenile H. portusjacksoni
located in the east or west macrohabitat zones
(ANOVA, d.f=1,52, P=0.23), however, this was
due to a reversal in proportions observed in 2005
when 65.2% of resting juveniles occurred in the
western zone. In the years 2003 and 2004 there
were significantly more juveniles observed resting
in the east (85.6% and 85.4%, respectively) than the
west macrohabitat zone (f-test, df=1, P<<0.01).
During these years, there was approximately 6
times the number of resting juveniles in the east
macrohabitat zone than the west.

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

Table 5: Macrohabitat type and proportional availability (p,) with preferred habitat type (*) and
standardised preference scores (B,) by year and sex at Terrigal Haven, Cabbage Tree Bay, Dent

Rock and Murray’s Sandline. B, > * (site B

CRIT
Availability Preferences
Habitat ——p. 2002 2003 2004 2005
Sex Habitat* B. Habitat* 8B. Habitat? 2B. MHabitat* B,
Barren 0.34 Male Interface 0.95 Interface 0.94 Interface 0.87 Interface 1.00
Interface 0.05 Female Interface 0.96 Interface 0.92 Interface 0.89 Interface 0.84
Kelp 0.19 Combined Interface 0.96 Interface 0.93 Interface 0.89 Interface 0.93
Sand 0.42
Cabbage Tree Harbour (* 0.2)
Barren 0.41 Male Gutter 0.64 Interface 1.00 Interface 1.00
Gutter 0.01 Interface 0.34
Interface 0.05 Female Gutter 0.98 Gutter 0.99 Gutter 0.93
Kelp 0.08 Combined Gutter 0.95 Gutter 0.98 Gutter 0.87
Sand 0.45
Dent Rock (* 0.2)

Barren 0.38 Male Interface 0.86 Gutter 0.57 Interface 0.68
Gutter 0.06 Interface 0.35 Gutter 0.32
Interface 0.05 Female Interface 0.51 Gutter 0.78 Gutter Ons,
Kelp 0.12 Gutter 0.49 Interface 0.23
Sand 0.39 Combined Interface 0.58 Gutter 0.57 Gutter 0.57

Gutter 0.42 Interface 0.32 Interface 0.43

Murray’s Sandline (* 0.33)
Interface 0.10 Male Interface 0.95 Interface 0.91 Interface 0.93
Sand 0.45 Female Interface 0.95 Interface 0.95 Interface 0.94
Seagrass 0.45 Combined Interface 0.95 Interface 0.94 Interface 0.94
DSIOILUG WIG AC ICE UMTS nts broadly spread unutilised locations, suggesting

The PC1 axis of the PCA biplot represents
a gradient of decreasing percentage of the 500
um sediment fraction (left to right) and indicates
that juveniles utilised resting positions with a low
percentage of this sediment fraction (3.2 + 0.22%;
mean + SE) (Table 4, Fig. 4d). The PC2 axis represents
a gradient of increasing seagrass cover (bottom to
top) and indicates selection of resting positions with
moderate seagrass cover (38.0 + 4.92%). The shark
resting locations are relatively tightly clustered in
respect of the two PC axes in comparison to the

Proc. Linn. Soc. N.S.W., 129, 2008

juveniles were very selective in the choice of resting

value) indicates a significant preference.

locations.

Juvenile Habitat Preferences

Resting juveniles exhibited a very strong preference
for the interface habitat (Table 5). This preference
was demonstrated by both sexes and in all years.
Although the B, values for the seagrass habitat were
small (female: 0.05 + 0.003; male: 0.07 + 0.013; mean
+ SE), the preference scores were significantly

159

HABITAT PREFERENCES OF PORT JACKSON SHARKS

Table 6: Principal Components Analysis results for key oviposition mi-
crohabitat variables at Terrigal Haven and Dent Rock. Variables with

The same primary oviposition
area at each site was used in

the highest eigenvalue shown for each axis. both years.
The PC1 axis of the PCA
am ; biplot for Terrigal Haven
Site PC Vintalblls Eigen- Variation Cumulative represents a gradient of
vector (%) Percentage decreasing reef slope (left to
1 Reef Slope 0.557 47.4 Ajay Mehl) ail uncicaies diet A
Terrigal portusjacksoni placed their
Haven Rock Size 0.795 DBD 70.6 eggs in a steeply sloping
3. Crevice Depth 0.624 14.2 84.7 portion of the reef (25-30°)
@able vos Fics >) 5 thesee2
axis represents a gradient of
1 Crevice Depth -0.589 332) 33.9 increasing rock size (bottom
Deut Crevice Width 0.713 25.5 Oy ae OLD unc ean Ueuisoleciee
Rock Oviposition sites were in an

3 Rock Size 0.850 ipa 76.5

higher than the sand habitat for both sexes in all
years.

Oviposition Site Habitat Characteristics

Viable egg capsules were not uniformly
distributed across the reef. Significantly more
viable egg capsules occurred in the single primary
oviposition area at both Terrigal Haven and Dent
Rock in 2004 (G-tests, df=1, P=0.02; df=1, P=0.02,
respectively) and 2005 (G-tests, df=1, P<0.001; df1,
P=0.03, respectively) than elsewhere on these reefs.

PC2

area with small rocks (51.6 +
5.23 cm; mean + SE). At Dent
Rock, the PC1 axis represents a
gradient of decreasing crevice
depth (left to right) and indicates that oviposition
sites with moderate crevice depth (42.2 + 7.40 cm)
were utilised. The PC2 axis represents a gradient of
increasing crevice width (bottom to top) showing
that utilised oviposition sites had moderate to narrow
crevices (13.2 + 1.11 cm). The utilised oviposition
locations are relatively tightly clustered in respect
of the two PC axes at both Terrigal Haven and Dent
Rock in comparison to the broadly spread unutilised
locations, suggesting high selectivity in the choice of
oviposition locations.

PC 2

“30 -25 -20 -15 -10 05 0
PCI

OS 10 15 20 25

Figure 5: Principal Components Analysis plots of microhabitat variables at positions with egg capsules
(closed circles) and random positions without egg capsules (open diamonds) at (a) Terrigal Haven (PC1,
reef slope; PC2, rock size) and (b) Dent Rock (PC1, crevice depth; PC2, crevice width).

160

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

Table 7: Macrohabitat type, proportional availability (p,) and
standardised preference scores (B,) for oviposition sites at Terrigal
Haven and Dent Rock for all years combined. * indicates preferred

edges. Nelson and Johnson (1970)
and Strong (1989) reported that
adult Heterodontus francisci were

habitat (Terrigal Haven B, > 0.33; Dent Rock B, > 0.25).

mainly found sheltering in caves
and overhangs during the day to
avoid exposure to strong sunlight.

Terrigal Haven Dent Rock It is unlikely that this is a factor

Z for adult H. portusjacksoni at
ai Pa Dies el ctor izle P; &, Terrigal Haven as the east-west
Shallow* 0.357 0.522 Quadrant | 0.126 0.200 running interface is exposed to
Kel 0.304 0.156 AED 0.385 0.010 unobstructed sunlight throughout
ie ee the day. Additionally, there was
Barrens 0.339 0.322 Quadrant 3 0.371 0 no diel variation in the use of,
Quadrant 4* 0.118 0.791 or preference for, this habitat. A

At Terrigal Haven, the shallow macrohabitat
zone was significantly preferred as an oviposition site
(Table 7). The barrens zone was marginally below the
Beir level of habitat preference (0.33), but was still
significantly selected over the kelp zone. Quadrant 4
macrohabitat zone at Dent Rock was the significantly
preferred oviposition habitat (Table 7). Quadrant 1
was marginally below the B,,,, level of preference
(0.25), but was significantly selected over quadrants
2 and 3.

DISCUSSION

To obtain a sufficiently detailed understanding
of habitat requirements for use in conservation and
management, habitat studies must address two key
issues: a determination of how sharks distribute
themselves among the available habitats and an
understanding of the reasons for these choices and
selections (Sims 2003). This study addressed the
issue of distribution across habitats at geographically
separate locations and explored the second issue in
relation to the resting habitat of H. portusjacksoni.
Despite differences in UVC frequency at adult sites,
surveys were conducted over several years, across the
entire breeding season, at different times of day and
across a range of weather conditions to ensure that
observations were representative.

Adult Habitat Utilisation and Preferences

The interface habitat was strongly preferred
at Terrigal Haven by resting A. portusjacksoni.
McLaughlin (1969) found that 88% of all adult H.
portusjacksoni in captive studies rested within 1.2 m
of the pool sides and suggested that this may have
been to maximise the sunlight shading effects of the

Proc. Linn. Soc. N.S.W., 129, 2008

more plausible explanation for the

preference for the interface habitat

is related to the avoidance of
strong water movement. The Terrigal Haven site was
located on the leeward side of the reef and is afforded
significant protection from all but large seas (personal
observations). Resting positions were concentrated
into two small patches in the shallow zone,
corresponding to the most steeply sloping portion of
the reef, and barrens zone, which is the deepest area.
It is likely that by resting close to the reef interface,
H. portusjacksoni are minimising their exposure to
water movement by using the reef as a form of flow
refuge (Webb 1989). Similarly, resting adult sharks
at Dent Rock were mainly encountered on the reef’s
more sheltered southern side in quadrants | and 2.

Heterodontus francisci are known to migrate
to deeper water during the more storm-prone winter
months around California and shark numbers were
reduced during surveys following high seas (Strong
1989). Similarly, Farina and Ojeda (1993) suggest that
redspotted catsharks, Schroederichthyes chilensis,
migrate into deeper water during winter to avoid
strong water movement and turbulence. Epaulette
sharks, Hemiscyllium ocellatum, on shallow reef flats
were often observed underneath or directly behind
coral heads, which may be used as ‘flow refuges’
from currents (Peach 2002).

The presence of numerous rock and rock/
sand gutters at Dent Rock represents a significant
difference between Dent Rock and Terrigal Haven,
as does the sex-based variation in habitat use at the
former site. Males at Dent Rock were significantly
more likely to prefer resting positions at the interface
than in the gutters, whilst females exhibited a strong
preference for the gutters. Sims et al. (2001) reported
that female dogfish, Scyliorhinus canicula, were often
found in female-only aggregations in refuge habitats
to decrease their accessibility by males and reduce the
energetically demanding activity of mating. This male

161

HABITAT PREFERENCES OF PORT JACKSON SHARKS

avoidance strategy is likely to be a significant factor
in the utilisation of gutters by females at Dent Rock.
The gutters are narrow and provide limited access
from the sides and are often occupied by groups of
females in close contact with each other. Sims et al.
(2001) found that male S. canicula frequented the
entry/exit points of the female refuges at times when
females were more likely to be coming or going. At
Dent Rock, male H. portusjacksoni were more likely
to occupy resting positions outside, but adjacent to,
the gutters, presumably to be near females if mating
opportunities arose. On a number of occasions a
single male H. portusjacksoni was observed resting
several metres from the interface where a group of
females were resting in close proximity to each other.
Although the females were in a mixed orientation to
each other, the male was always perpendicular to, and
facing, the females.

Habitat preferences of resting H. portusjacksoni
at Cabbage Tree Harbour conform to both the water
movement and male avoidance strategies. The reef
face is completely exposed to the prevailing seas
and is often significantly affected by strong swell
and surge (personal observations). As expected by
the water movement avoidance strategy, the number
of adult H. portusjacksoni resting at the interface at
Cabbage Tree Harbour was low. Females exhibited
a highly significant preference for the rock gutters in
the deeper, eastern end of the reef and avoided the
remaining habitats. The strong preference exhibited
by females for the gutters at Cabbage Tree Harbour
also supports the male avoidance strategy, with
over 90% of resting females located in the gutters.
On the contrary, only one male H. portusjacksoni
was observed resting in the gutters during the three
survey years, whilst 50% (n=14) of all males sighted
at Cabbage Tree Harbour were actively swimming in
the gutter zone.

Juvenile Habitat Utilisation and Preferences

The habitat utilised by juvenile H. portusjacksoni
is totally distinct from that of the adults. Despite
three to four years surveying adult habitats at Terrigal
Haven, Cabbage Tree Harbour and Dent Rock,
juvenile sharks were never observed at these sites.
Instead juvenile sharks were located in a shallow
seagrass bed at Murray’s Sandline geographically
isolated from the rocky reefs typically utilised by the
adults. Such spatial separation of adults and juveniles
is common to many elasmobranchs (Merson and
Pratt 2001; Pratt and Carrier 2001; Carlson 2002).
Fulfilling the three criteria specified by Heupel et
al. (2007), this site can be considered a nursery area
for H. portusjacksoni. Juveniles were encountered

162

here regularly, but not at other sites used by adults;
they spend extended periods of time at the site; and,
exhibit strong site fidelity for the site over several
years (Powter 2006).

Juvenile H. portusjacksoni made significantly
greater use of the eastern portion of the seagrass
bed. The eastern portion accounted for 50% of the
total surveyed area, however, 81.7% of sharks were
located within the eastern portion of the seagrass
bed. Although this pattern was consistent in 2003
and 2004, it was reversed in 2005. The exact reason
for this shift is unclear, but may be related to an
avoidance of strong water movement. During a survey
of the entire seagrass bed in March 2004 a lower than
expected number of juvenile H. portusjacksoni was
located (Powter 2006). All of the juveniles located
in the eastern portion of the bed during this survey
were in the easternmost two-thirds of the bed and
closer to Bowen Island than previous surveys. Whilst
surveying the midwestern zone, a 100 m section of
the seagrass bed was missing after being washed
away in heavy seas several weeks previously. An
identical observation was made in December 2005
where an extensive area further east of the previous
section had been washed away. Again the number of
juvenile H. portusjacksoni was significantly lower
than expected.

Murray’s Sandline is afforded some protection
from the prevailing seas by the southern headland of
Jervis Bay (Bherwerre Peninsula) and Bowen Island.
However, the western portion of the bed is less
protected due to its alignment with the gap between
Bowen Island and Bherwerre Peninsula. Hence
the reduction in juvenile numbers after these storm
events indicates that juveniles, like the adults, may
adopt a strategy of avoiding strong water movements.
Further support for this notion is the significant
role sediment grain size and bed slope plays in the
selection of resting sites. Bed slope, greatest in the lee
of Bowen Island, is likely to impact on the intensity
of water movement, with the juveniles occupying
resting positions at the base of the steepest sloping
portion of the eastern seagrass bed, which they may
use as flow refuges to assist in station holding (Webb
1989). The reduced proportion of the 500um grain
size fraction at juvenile resting sites was also likely
to be related to gross water movement, as seagrass
is known to influence both the velocity and direction
of moving water. Low stem densities can lead to
substrate erosion, whilst higher densities can facilitate
the settling out of suspended particles (Edgar 2001).

Although many studies have defined nursery
areas and their importance to elasmobranchs, few
have addressed the issue of habitat preference

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

(Simpfendorfer and Heupel 2004). Both within and
between years and seasons, there was no difference in
the strong preference resting juveniles exhibited for
the interface habitat, where they rested on the sand
substratum in close proximity to the seagrass bed. This
narrow strip connecting the sand flat to the seagrass
bed comprised only 10% of the surveyed area, but
accounted for 83.3% of all juvenile H. portusjacksoni.
Strong (1989) observed a similar pattern in juvenile H.
francisci which exhibited a strong preference for sand
substratum and often sheltered near scattered algae,
debris or other topographic features. Habitat edges
and ecotones have been shown to be important to a
diverse array of organisms for a wide range of reasons
(Meffe and Carroll 1997). However, a possible benefit
of the interface habitat to juvenile H. portusjacksoni
relates to the visual complexity of this region and
the juvenile’s disruptive colouration. Motionless
juveniles at the interface were often difficult to detect
and blended well with their surroundings (personal
observations). The preference may also arise from
a resolution of the conflict between a habitat being
too complex for effective foraging and one not
providing sufficient refuge (Adams et al. 2004). The
diet of juvenile H. portusjacksoni is dominated by
benthic invertebrates, such as decapod crustaceans
and echiurans (Powter 2006) and the seagrass bed
is likely to provide a suitable source of these prey
items (Edgar 2001). During UVC surveys juveniles
were observed on at least four occasions foraging in
the substrate of the seagrass bed or on the epiphytic
organisms on the seagrass blades. However, juveniles
avoided swimming amongst the closely spaced
seagrass blades and were infrequently found resting
amongst them. Accordingly, the interface habitat
may provide an appropriate balance of adequate
refuge with reasonable proximity to prey and ease of
access to foraging, but at a decreased likelihood of
impediments to free movement.

Oviposition Habitat Utilisation and Preferences
As an oviparous species, H. portusjacksoni
deposits its egg capsules in rock crevices in the same
shallow rocky reefs where adults are found resting
(McLaughlin and O’Gower 1971). The choice of
suitable habitat for oviposition is critical as the
developing embryo spends 9-12 months within the
capsule prior to eclosion (Rodda 2000). Oviposition
occurred selectively within primary oviposition areas
at Terrigal Haven and Dent Rock over at least two
consecutive seasons, but there was no evidence of
this activity at Cabbage Tree Harbour. Significant
differences existed between Dent Rock and Cabbage
Tree Harbour in terms of crevice width and depth, but

Proc. Linn. Soc. N.S.W., 129, 2008

no such differences existed between Terrigal Haven
and Cabbage Tree Harbour (Powter 2006). However,
the oviposition areas at Terrigal Haven (southern
reef face) and Dent Rock (west-north-west reef face)
were both located on the reef face less exposed to
prevailing seas and on a downward sloping reef face
behind a raised reef crest, whilst the reef at Cabbage
Tree Harbour faced north and was significantly more
exposed. Despite these similarities, the selected
Oviposition sites at the two locations did vary. The
mean crevice width and depth at Dent Rock were
significantly greater than at Terrigal Haven (Powter
2006) and the method of securing capsules was
influenced by this. At Terrigal Haven capsules were
predominantly located singly tightly wedged into
crevices between rocks, with a maximum of two
capsules observed in the same crevice. Hence, the
narrower and shallower crevices at Terrigal Haven
were suitable for securing individual egg capsules.
Offering less opportunity to secure capsules in this
way, the deeper, wider crevices at Dent Rock often
contained multiple capsules laying relatively loose in
the bottom of deep, narrow crevices, with a maximum
of 18 capsules in the same crevice. Both of these
methods served to hinder mechanical dislocation, but
also offered protection from larger predators (Powter
2006). Additionally, both crevice types appeared to
be equally effective at protecting egg capsules as
demonstrated by the similar levels of embryonic
mortality occurring at Terrigal Haven and Dent Rock
(Powter 2006).

Hence, areas sheltered from the prevailing seas
and with suitable crevices to prevent mechanical
dislocation and predation appear to be a primary
requirement for oviposition areas. However, suitable
oviposition habitat could also be influenced by
water temperature, which Rodda (2000) found had
a significant effect on the early stages of embryonic
development of H. portusjacksoni in laboratory
experiments. The lower thermal limit fell between
15° and 18°C, whilst the upper thermal limit was
approximately 22° C. Consequently the location
of oviposition areas in shallow water may assist in
optimal temperature regulation. Additionally, water
temperature was found to be negatively correlated
with adult H. portusjacksoni numbers at Terrigal
Haven and Dent Rock, but not at Cabbage Tree
Harbour (Powter 2006). McLaughlin and O’Gower
(1971) also reported an inverse relationship between
water temperature and H. portusjacksoni numbers,
but did not offer an explanation for this relationship.
The most likely reason is the narrow temperature
tolerance of developing H. portusjacksoni embryos.
Congregating on shallow coastal reefs for reproductive

163

HABITAT PREFERENCES OF PORT JACKSON SHARKS

purposes (Powter 2006) it is likely that adult numbers
at oviposition reefs, such as Terrigal Haven and Dent
Rock, are related to temperatures within the optimal
range for their developing offspring. The lack of
both reproductive activity at Cabbage Tree Harbour
and a relationship between shark numbers and water
temperature is also consistent with this finding.

A significant preference for the two primary
oviposition areas at Terrigal Haven and Dent Rock
occurred over two consecutive seasons. Although
repeated, or ‘traditional’, use over a number of
breeding seasons by individual females could not
be demonstrated, the high site fidelity exhibited by
mature females at the reefs (Powter 2006) and the
‘traditional reuse’ of the oviposition sites suggests
that females have strong philopatric links to the
oviposition areas. Female H. portusjacksoni have
a long reproductive life and ‘experienced’ females
could reuse the same primary oviposition areas in
subsequent years or dominant females may utilise the
primary oviposition areas and subordinate females use
other locations within the reef. Nonetheless this is the
first quantitative determination of the use of traditional
oviposition sites by an oviparous elasmobranch.

ACKNOWLEDGEMENTS

We wish to thank Australian Geographic and Project
AWARE (PADI Asia Pacific) for their invaluable financial
assistance. All work was conducted under University of
Newcastle Ethics Approval 804 0602, NSW Fisheries
Scientific Collection Permit P02/0042 and Environment
Australia Research Activity Permit BDRO02/00015 and
renewals.

REFERENCES

Adams, A.J., Locascio, J.V. and Robbins, B.D. (2004).
Microhabitat use by a post-settlement stage estuarine
fish: evidence from relative abundance and predation
among habitats. Journal of Experimental Marine
Biology and Ecology 299, 17-33.

Carlson, J.K. (2002). Shark nurseries in the northeastern
Gulf of Mexico. In: Shark nursery grounds of
the Gulf of Mexico and the East Coast waters
of the United States: an overview. An internal
report to NOAA‘ Highly Migratory Species Office
(McCandless, C. T., Pratt, H. L. & Kohler, N. E.
(eds), pp. 165-182. Narragansett: NOAA.

Carraro, R. and Gladstone, W. (2006). Habitat preferences
and site fidelity of the ornate wobbegong shark
(Orectolobus ornatus) on rocky reefs of New South
Wales. Pacific Science 60, 207-223.

164

Castro, J.I. (1993). The shark nursery of Bulls Bay,

South Carolina, with a review of the shark nurseries
of the southeastern coast of the United States.
Environmental Biology of Fishes 38, 37-48.

Clarke, K.R. and Warwick, R.M. (2001). Change in
marine communities: an approach to statistical
analysis and interpretation. Plymouth, U.K.:
PRIMER-E

Edgar, G.J. (2001). Australian Marine Habitats in
Temperate Waters. Sydney: Reed New Holland.

Farina, J.M. and Ojeda, F.P. (1993). Abundance, activity,
and trophic patterns of the redspotted catshark,
Schroederichthys chilensis, on the Pacific temperate
coast of Chile. Copeia 1993, 545-549.

Foster, M.S., Harrold, C. and Hardin, D.D. (1991). Point
vs. photo quadrat estimates of the cover of sessile
marine organisms. Journal of Experimental Marine
Biology and Ecology 146, 193-203.

Goldman, K.J. and Anderson, S.D. (1999). Space
utilization and swimming depth of white sharks,
Carcharodon carcharias, at the South Farallon
Islands, central California. Environmental Biology of
Fishes 56, 351-364.

Hall, L.S., Krausman, P.R. and Morrison, M.L. (1997).
The habitat concept and a plea for standard
terminology. Wildlife Society Bulletin 25, 173-182.

Heithaus, M.R., Dill, L.M., Marshall, G.J. and Buhleier,
B. (2002). Habitat use and foraging behavior of tiger
sharks (Galeocerdo cuvier) in a seagrass ecosystem.
Marine Biology 140, 237-248.

Heupel, M.R., Carlson, J.K. and Simpfendorfer, C.A.
(2007). Shark nursery areas: concepts, definition,
characterization and assumptions. Marine Ecology
Progress Series 337, 287-297.

Heupel, M.R. and Hueter, R.E. (2002). Importance of
prey density in relation to the movement patterns
of juvenile blacktip sharks (Carcharhinus limbatus)
within a coastal nursery area. Marine and Freshwater
Research 53, 543-550.

Heupel, M.R., Simpfendorfer, C.A. and Hueter, R.E.
(2004). Estimation of shark home ranges using
passive monitoring techniques. Environmental
Biology of Fishes 71, 135-142.

Last, P.R. and Stevens, J.D. (1994) Sharks and rays of
Australia. Australia: CSIRO.

Manly, B., McDonald, L. and Thomas, D. (1993).
Resource selection by animals: statistical design and
analysis for field studies. London: Chapman and Hall.

Matern, S.A., Cech, J.J. and Hopkins, T-E. (2000). Diel
movements of bat rays, Myliobatis californica, in
Tomales Bay, California: evidence for behavioral
thermoregulation? Environmental Biology of Fishes
58, 173-182.

McLaughlin, R.H. (1969). The ecology of heterodont
sharks. PhD Thesis, University of New South Wales,
Australia

McLaughlin, R.H. and O’Gower, A.K. (1971). Life
history and underwater studies of a heterodont shark.
Ecological Monographs 41, 271-289.

Proc. Linn. Soc. N.S.W., 129, 2008

D.M. POWTER AND W. GLADSTONE

Meffe, G.K. and Carroll, C.R. (1997). Principles of
conservation biology. Sunderland, Massachusetts:
Sinauer Associates.

Merson, R.R. and Pratt, H.L. (2001). Distribution,
movements and growth of young sandbar sharks,
Carcharhinus plumbeus, in the nursery grounds of
Delaware Bay. Environmental Biology of Fishes 61,
13-24.

Nelson, D.R. and Johnson, R.H. (1970). Diel activity
rhythms in the nocturnal, bottom-dwelling sharks,
Heterodontus francisci and Cephaloscyllium
ventriosum. Copeia 1970, 732-739.

O’Gower, A.K. (1995). Speculations on a spatial
memory for the Port Jackson shark (Heterodontus
portusjacksoni) (Meyer) (Heterodontidae). Marine
and Freshwater Research 46, 861-871.

Peach, M.B. (2002). Rheotaxis by epaulette sharks,
Hemiscyllium ocellatum (Chondrichthyes:
Hemiscyllidae), on a coral reef flat. Australian
Journal of Zoology 50, 407-414.

Powter, D.M. (2006). Conservation biology of the Port
Jackson shark, Heterodontus portusjacksoni, in New
South Wales. PhD Thesis, University of Newcastle,
Australia.

Pratt, H.L. and Carrier, J.C. (2001). A review of
elasmobranch reproductive behavior with a case
study on the nurse shark, Ginglymostoma cirratum.
Environmental Biology of Fishes 60, 157-188.

Rodda, K.R. (2000). Development in the Port Jackson
shark embryo. PhD Thesis, University of Adelaide,
Australia

Simpfendorfer, C.A. and Heupel, M.R. (2004). Assessing
habitat use and movement. In: Biology of sharks
and their relatives (Carrier, J.C., Musick, J.A. &
Heithaus, M.R., eds.), pp. 553-572. Boca Raton: CRC
Press.

Sims, D.W. (2003). Tractable models for testing theories
about natural strategies: foraging behaviour and
habitat selection of free-ranging sharks. Journal of
Fish Biology (Supplement A) 63, 53-73.

Sims, D.W., Nash, J.P. and Morritt, D. (2001). Movements
and activity of male and female dogfish in a tidal sea
lough: alternative behavioural strategies and apparent
sexual segregation. Marine Biology 139, 1165-1175.

Sokal, R.R. and Rohlf, F.J. (2003). Biometry: the
principles and practice of statistics in biological
research. New York: W.H. Freeman and Company.

Strong, W.R. (1989). Behavioural ecology of horn sharks,
Heterodontus francisci, at Santa Catalina Island,
California, with emphasis on patterns of space
utilization. MSc Thesis, California State University,
Long Beach.

Stindstrom, L.F., Gruber, S.H., Clermont, S.M., Correia,
J.P.S., de Marignac, J.R.C., Morrissey, J.F.,
Lowrance, C.R., Thomassen, L. and Oliveira, M.T.
(2001). Review of elasmobranch behavioral studies
using ultrasonic telemetry with special reference
to the lemon shark, Negaprion brevirostris, around
Bimini Islands, Bahamas. Environmental Biology of
Fishes 60, 225-250.

Proc. Linn. Soc. N.S.W., 129, 2008

Webb, P.W. (1989). Station-holding by three species of
benthic fishes. Journal of Experimental Biology 145,

303-320.

165

166

wi veoh aa anne
a fr okie

hats ey “ti, wT

\ i ih see thrice hae hts i PY)
(iz Ley -_" peti ae

cat Oy ree nye:

Lod ua )
t 4 San ce.
7 Ot “si
: ‘ ;
i .
? hae
a ‘
fre
=
= +
. Proluteayt mb
ik ee -
heal burt, ci
. "
ib
Lt
i Pe iy oe
4 i] is ‘> * és ” : Pe 5 r —_= -
' 4 J riL4 ;
¥& APM gM Oy GCE i :

Mah taly ~ emany! Aid wa ean 4

aS 1h eR | naey =

2) lair iy eek Oye) ania ~ >
abi Mahe ¢ sor As Aha)

4 nates #s wales tal '

Pi Ae ia <
AOS ® 3 z VE 2
nie =

Diel Activity of the Endangered Trout Cod (Maccullochella

macquariensis) in the Murrumbidgee River

J.D. Totem, '*" B.C. Epner, '? AND B.T. BROADHURST!

‘Parks, Conservation and Lands, Department of Territory and Municipal Services, ACT Government, GPO
Box 158, Canberra ACT 2601, Australia; "Cooperative Research Centre for Freshwater Ecology, University
of Canberra, ACT 2601, Australia. *(jason.thiem@act.gov.au)

Thiem, J.D., Ebner, B.C. and Broadhurst, B.T. (2008). Diel activity of the endangered Trout Cod
(Maccullochella macquariensis) in the Murrumbidgee River. Proceedings of the Linnean Society of

New South Wales 129, 167-173.

Diel movements and habitat use of most of Australia’s large freshwater fish fauna remain unknown, despite
conservation efforts for many of the threatened species, including re-stocking and habitat protection and
restoration. We used radio-telemetry to monitor diel movements of the endangered trout cod (Maccullochella
macquariensis: Percichthyidae) in a re-stocked population in the Murrumbidgee River, New South Wales,
Australia. Both manual tracking and continuous remote telemetry identified that trout cod activity peaked in
periods of low light; with linear ranges for individuals varying from 6—272 m. Trout cod had strong fidelity
to outer river bends throughout diel periods and this has implications for targeted habitat rehabilitation

efforts.

Manuscript received 1 September 2007, accepted for publication 6 February 2008.

KEY WORDS: diel activity; home range; Percichthyidae; radio-telemetry

INTRODUCTION

Trout cod (Maccullochella macquariensis
(Cuvier): Percichthyidae) is a large freshwater fish
endemic to rivers in the southeast of the Murray—
Darling Basin, Australia (Ingram and Douglas 1995).
The species has undergone a large scale reduction
in distribution from much of its former range
(Cadwallader and Gooley 1984; Douglas et al. 1994)
and is currently classified as endangered (IUCN
2006). Self-sustaining populations of trout cod are
now limited to a single remnant population and two
translocated populations (Douglas et al. 1994; D.
Gilligan, pers. comm.). Conservation efforts to re-
establish the species have been strongly focussed
on: 1) stocking hatchery-produced fingerlings
(Lintermans and Phillips 2005) and, 2) understanding
habitat requirements (Growns et al. 2004; Nicol et al.
2007).

Trout cod in both remnant and stocked lowland
populations exhibit strong preferences for in-stream
wood habitat (Growns et al. 2004; Nicol et al. 2007)
as this generally forms the dominant structural habitat
type in lowland rivers (Koehn et al. 2004). Trout cod

also prefer deep sections of river (Nicol et al. 2007);
often away from the river bank (Growns et al. 2004).
However, both Growns et al. (2004) and Nicol et al.
(2007) only report on trout cod habitat use during
diurnal periods. With river restoration practices
underway in some rivers, including the addition of
structural wood habitat (e.g. Nicol et al. 2004), the
lack of information on trout cod use of space and
habitat over diel periods represents a significant
knowledge gap.

In Australia, radio-tracking has successfully been
used to study localised movement of large freshwater
fishes, primarily percichthyids (Butler 2001; Crook et
al. 2001; Simpson and Mapleston 2002; Crook 2004a,
b; Ebner et al. 2005). Large percichthyids exhibit site
fidelity, have relatively small home ranges (the area
over which the animal normally travels in search of
food (Burt 1943)) as adults over most or all of the year
(Koehn 1997; Simpson and Mapleston 2002; Crook
2004a, b) and are active during periods of low light
(Butler 2001; Simpson and Mapleston 2002; Ebner
et al. 2005). The aim of this study was to determine
diel habitat use and activity of stocked trout cod in a
lowland river.

DIEL ACTIVITY OF TROUT COD

Gogeldrie Weir

Yanco Weir

SS

146°36'E

Narrandera

Study reach

20 Kilometres

34°46'S

Berembed Weir

Figure 1. Location of the study reach in southern New South Wales, Australia.

MATERIALS AND METHODS

Study site

The study was conducted in a lowland reach
of the Murrumbidgee River, 5 km upstream of the
township of Narrandera (173 m ASL) in southern New
South Wales (NSW), Australia (Fig. 1). Narrandera
is one of twelve trout cod stocking sites in the
Murrumbidgee catchment, with 85,000 fingerlings
stocked at this location between 1996 and 2000
(Gilligan 2005). Subsequent surveys have identified
survival and growth of trout cod (Growns et al. 2004;
Gilligan 2005). The river channel has widths of 60—
70 m and maximum water depths of 3—5 m occur on
outside bends of the river. River red-gum Eucalyptus
camaldulensis is common along both banks and fallen
trees or branches comprise the dominant in-stream
structural habitat for trout cod (Growns et al. 2004).

Fish collection and surgery

Movement data were collected from 10 radio
tagged trout cod (370-575 mm Total Length (TL),
599-2587 g, Table 1), probably comprising a mixture
of mature and immature fish (Harris and Rowland
1996). These 10 individuals comprised nine trout
cod captured in the study reach by boat electro-
fishing and one trout cod on-grown in a hatchery
and subsequently released. These individuals were
originally from samples containing 31 trout cod
collected in the study reach by boat electro-fishing
(formerly stocked as fingerlings and subsequently
re-captured) and from 27 trout cod sourced from a
state government hatchery (on-grown two year-old
fish). Radio tags (Gnternal body implants with a 30 cm
trailing whip antenna, models F1830, 35, 40 and 50,
Advanced Telemetry Systems (ATS), Isanti, USA)
were surgically implanted into the peritoneal cavity

Proc. Linn. Soc. N.S.W., 129, 2008

J.D. THIEM, B.C. EBNER AND B.T. BROADHURST

under anaesthesia (0.5 ml Alfaxan (Jurox, Rutherford,
Australia) per litre of water). The weight of radio tags
in air were between 11 g and 25 g to suit a range of
fish sizes and were kept to < 2% of fish mass. Pulse-
coded, two-stage radio transmitters were used on a
frequency of 150-152 MHz and programmed on a
pulse rate of 5 s on and 7 s off to increase battery
life (warranted for between 230 and 504 d). For
external identification, individuals were also tagged
with a dart tag between the second and third dorsal
spines. Tagging procedures were identical to those
described by Ebner et al. (2007), with the exception
that surgical incisions were 2—3 cm in this study.
Individuals were initially recovered in a darkened
enclosure holding 200 | of aerated water at 5 parts
per thousand NaCl. Upon regaining swimming ability
individuals were transferred to large circular concrete
enclosures that held between 500 and 1000 | at 5 parts
per thousand NaCl. Individuals were held for 2-15 d
in the hatchery following surgery and all individuals
were released at the same location in late September
2003. We assume that full recovery from tagging
procedures and resumption of normal behaviour had
occurred since this study was conducted from 14-19
November 2004.

Radio-telemetry
A modified technique of David and Closs (2001)
was used to remotely monitor the activity of a single
trout cod (trout cod no. 5, Table 1) continuously from
2000 h 14 November until 0900 h 19 November 2004
(Australian Eastern Standard Time). A three-element
Yagi antenna (Titley Electronics, Australia) was
fixed to a tree within the home range of trout cod
no. 5, perpendicular to the stream (the home range
location of trout cod no. 5 had been determined in
a previous study (Ebner et al., 2006)). The antenna
was connected to a remote data logger (DCCII Model
D5041, ATS) via a receiver (Model R4100, ATS) and
recorded radio signal strength every 5 s. The standard
deviation of signal strength was plotted in 10 min
grouped intervals to examine signal variability as
a measure of activity, with high signal variability
indicating active periods (David and Closs 2001).
Detection range of the data logger was approximately
80 m in any direction, incorporating any movements
to the opposite river bank.
Ten individuals were manually tracked every
four hours for two consecutive 24 h periods (1400 h
15 November until 1400 h 17 November 2004) from
a power-boat. Trout cod no. 5 was subsequently
tracked hourly for an additional 24 h period (0600 h
18 November until 0700 h 19 November 2004) from
an electric powered boat. Radio-tracking fixes were

Proc. Linn. Soc. N.S.W., 129, 2008

determined using a handheld three-element Yagi
antenna (Titley Electronics) and a receiver (Australis
26k, Titley Electronics). Locations were recorded
using a handheld GPS unit, with three GPS points
taken at each location.

Data analysis

GPS records were averaged to provide a single
location datum for each individual per radio-tracking
fix. Spatial data were plotted in ArcView 3.2™ (ESRI,
USA) over a base-map, generated by walking both
banks of the river with handheld GPS units. A polyline
was generated based on the sequential locations
of each individual using the Animal Movement
Extension in ArcView (Hooge and Eichenlaub 1997)
and the polyline was used to construct a time series of
the distance moved by fish between consecutive radio-
tracking periods (activity). Linear range (the distance
between the most upstream and downstream points)
and area used (Minimum Convex Polygons (MCP))
were also calculated with ArcView. To determine the
proportion of river used (for comparison with MCP
estimates), an ‘available area’ metric was calculated.
This involved constructing a line perpendicular to the
river channel at the most upstream and downstream
fix of each individual and estimating the wetted area
within these limits. Statistical analysis was conducted
using Statistix for Windows (version 2.0) with data
transformed, when necessary, to achieve normal
distribution (Tabachnick and Fidell 1989).

RESULTS

The linear range of trout cod over two consecutive
diel periods ranged from 6—272 m, with a mean (+ SE)
of 83.1 + 30.0 m (Table 1). There was a significant
positive correlation between fish length (mm) and
linear range (m) over the two diel periods (Pearson
correlation co-efficient: 0.7219, P<0.05). The
area of river used (MCP) over the two consecutive
diel periods ranged between 18.5-4603.5 m? and
averaged 1284.1 + 621.1 m*(Table 1). There was also
a significant positive correlation between fish length
(mm) and area used (m7) over the two diel periods
(Pearson correlation co-efficient: 0.7337, P<0.05).
For trout cod no. 5, increasing the temporal resolution
of tracking from four—hourly to hourly increased both
the linear range estimate (75 to 84 m) and the area
used estimate (307 to 1010 m’), over a single diel
period. Trout cod no. 5 used the same section of river
for all diel periods, however, hourly tracking resulted
in an increased lateral range, reflected by a larger area
estimate.

169

DIEL ACTIVITY OF TROUT COD

Table 1. Estimates of length and area of river used by 10 radio tagged trout cod in the Murrumbidgee
River at Narrandera, NSW. Values are based on four—hourly radio-tracking, combined over two con-
secutive diel periods. *Trout cod no. 6 had a weak radio signal and was.excluded from subsequent range

and area calculations. +Denotes on-grown hatchery individual.

Total j Linear Area Available river Proportion of
Trout cod Weight ‘ 5
a length (g) Sex range used area available river
ec aes (m)—— (m’) (m’) used (%)
1 a7 672 2 25 292 2293 13
2 370 599 l@ 26 280 1926 15
3 412 834 la 15 191 1128 V7
4 405 827 2 6 19 585 3
5 430 1066 ? 90 658 6160 11
6* 466 1249 ? N/A N/A N/A N/A
7 461 1292 M 109 925 8374 11
8 481 1384 ? DD 4604 20516 22
97 407 1247 2 31 128 2330, 5
10 DiS 2587 M 174 4463 12672 35
Mean 437.8 1175.7 83.1 1284.1 6220.8 14.7
(SE) (19.4) (179.4) (30.0) (621.1) (2227.7) (3.2)

60

50

40

30

Distance moved (m)

1400- 1800- 2200- 0200- 0600- 1000-
1800 2200 0200 0600 1000 1400

Time interval

Figure 2. The diel activity of nine trout cod in the Murrumbidgee Riv-
er based on the minimum distance moved between four hourly radio-
tracking locations (mean + SE). White and grey sections denote day
and night, respectively. Data are from two consecutive diel periods.

The mean (+ SE) area of river
used by nine individuals over
two consecutive diel periods,
1284.1 + 621.1 m’, represented
only a small proportion (14.7
+ 3.2%) of the river that was
available within the upstream
and downstream limits of total
range (Table 1), suggest-ing
that individual trout cod select
specific habitats in relatively
small areas. All ten individuals
(including trout cod no. 6 that
had a weak radio signal enabling
occasional detection) were
always located in the thalweg
during the day. For nine of ten
individuals this corresponded
to a location in close proximity
(<20 m) to the outer river bend.
An exception was the position of
one individual on the inside of
a bend at the upstream end of a
braided channel section. Nine of
10 individuals did not cross from
one side of the river to the other
during the study. One individual

170 Proc. Linn. Soc. N.S.W., 129, 2008

J.D. THIEM, B.C. EBNER AND B.T. BROADHURST

4-hourly

Logger
test

120
|

Logger
test

100

60
20 |
0

oo 0 © ©

at

M

Signal variability (SD)
BSS
(=)

scan sampling

OOO COO © O

il

Focal Logger
sampling test

‘i i

| ae
| i iil |

120

Time (hours)

Figure 3. Activity of trout cod no. 5, based on continuous logging of signal strength from its radio trans-
mitter, 14-19 November 2004. Data are grouped at 10 min intervals and plotted as the variation in radio
transmitter signal strength. Open circles at the top of the graph denote boat activity (predominantly
investigator related) within the reach. Day and night periods are represented in white and grey respec-

tively.

(trout cod no. 10, Table 1) was recorded crossing the
channel. This occurred in a straight section of river
that was without shallow water associated with either
bank. Additionally, four individuals had overlapping
ranges during this study. Two of these individuals
were found to co-inhabit the same hollow log during
the day for the duration of the study.

Movements of trout cod were generally greater
at night, dusk and dawn relative to daytime (Fig. 2).
The period of greatest movement (mean + SE) 37.27
+ 12.6 m was between 1800 h and 2200 h, with the
period of least movement 8.98 + 4.34 m between 1000
h and 1400 h (Fig. 2). Differences in distance moved
between time periods were non-significant (Kruskal-
Wallis One-way ANOVA, d.f.=107, HA=9.1671,
P>0.05).

Data from the remote logger revealed sporadic
variations in signal strength for the first three daylight
periods (Fig. 3). In comparison, variations in signal
strength were consistently higher and sustained for
longer periods of time during the night or during dusk
or dawn periods, indicating periods of heightened
activity. Additionally, radio transmitter signal
strength varied repeatedly throughout the fourth

Proc. Linn. Soc. N.S.W., 129, 2008

daylight period of remote telemetry logging (Fig.
3). This period coincided with a decrease in water
discharge (4376-3336 ML/day) and river height
(0.21 m decrease) (NSW DNR 2004). The fourth
daylight period also coincided with more intensive
radio-tracking of trout cod no. 5, changing from four
to one hourly (Fig. 3).

DISCUSSION

Trout cod occupied small (<300 m) lengths of
river over consecutive diel periods in this study and
the size of movements (linear range and area used)
were positively correlated with fish length. These
individuals had previously demonstrated fidelity
to the same diurnal locations throughout much of
the previous year (Ebner et al. 2006). Similarly this
species has been shown to be relatively sedentary in
the Murray River (Koehn 1997; Koehn and Nicol
1998). The small movements of trout cod are similar
to that reported for other Australian percichthyids
(Butler 2001; Simpson and Mapleston 2002; Crook
2004a, b).

171

DIEL ACTIVITY OF TROUT COD

Individual trout cod demonstrated a preference
for movements along outer river bends within diel
periods in this study. These outer bends are associated
with deeper areas and contain more structural woody
habitat in lowland rivers (Hughes and Thoms 2002;
Koehn et al. 2004). Efforts to conserve trout cod and
other threatened percichthyids should be aided by an
improved understanding of their lateral movements
in large lowland rivers. This information provides
the basis for strategic placement of structural woody
habitat in stream restoration programs (e.g. Nicol et
al. 2004).

Differences between estimates of available river
area and the area used by each individual reflected
the fidelity of trout cod to outer bends. Where linear
movements of a species predominate, standard
methods to calculate home range often produce
considerable over-estimates (Blundell et al. 2001).
Therefore equating home range size to the area of
river within the upstream and downstream limit of
a radio-tracked individual (e.g. Gust and Handasyde
1995) is an overestimate when applied to trout cod in
a large lowland river. The lateral distribution of trout
cod within a large lowland river is likely to be a direct
response to in-stream habitat differentiation (Hughes
and Thoms 2002; Koehn et al. 2004).

The logger method detected distinct nocturnal
activity of an individual whereas coarse-scale manual
radio-tracking did not. There was an indication of
greater movement during crepuscular and nocturnal
periods, based on manual radio-tracking of ten
individuals. Simpson and Mapleston (2002) found that
the activity of Mary River cod Maccullochella peelii
mariensis Rowland was matinine, based on real-time
manual radio-tracking within one-hour periods. Our
findings indicate that application of the continuous
remote telemetry method of David and Closs (2001)
based on increased sample sizes (e.g. David and Closs
2003) is likely to be an effective means of elucidating
the diel activity patterns of trout cod.

The cause of the shift from nocturnal to both
nocturnal and diurnal activity of an individual in this
study is unknown. The shift corresponded to both the
use of one—hourly boat-based manual radio-tracking
and a change in discharge and river height. This
demonstrates the capacity to use variation in signal
strength from remote loggers to monitor disturbance
(e.g. by the researcher, releases from dams) in
experiments. To date the application of variation
in signal strength has only been used to record diel
activity (see Baras et al. 1998; David and Closs 2001;
Hiscock et al. 2002; David and Closs 2003). Possible
observer effects could be investigated by remotely
monitoring the activity of an entire sample, whilst

v2

conducting manual radio-tracking of a subset of
individuals.

To produce reasonable estimates of the extent
of diel range, four—hourly radio-tracking (of about
five to ten individuals) appears to represent the most
pragmatic solution for a team of two researchers.
Before results of this study are used in a management
context, observations of home range size and shape
should be replicated among seasons to strengthen
the data set. This study indicates that trout cod
inhabit small reaches of river on a scale of tens to
hundreds of metres, within the deeper outer bank
of the Murrumbidgee River, over short periods of
observation. Consequently, the recovery of this
species can probably be conducted within small
reaches of river and specific in-stream habitats can be
prioritised for rehabilitation.

ACKNOWLEDGMENTS

New South Wales Department of Primary Industries
(NSW DPI) staff particularly Ian Wooden assisted with
field collections. Hatchery fish were sourced from the
Victorian Department of Primary Industries. S. Godschalx
provided assistance with surgery. L. Johnston and M.
Lintermans assisted with radio-tracking. J. Prince assisted
with data analysis and M. Evans provided statistical advice.
M. Dunford and L. Johnston provided invaluable GIS
support. D. Crook, K. Frawley, D. Gilligan, L. Johnston,
M. Lintermans and two anonymous reviewers improved
the text. The research was principally funded by the
Fisheries Research and Development Corporation and
Parks, Conservation and Lands (ACT Government) and
benefited from contributions by NSW DPI, DPI Victoria,
the Cooperative Research Centre for Freshwater Ecology
and the Murray—Darling Basin Commission. This study
was performed under NSW Fisheries Animal Care and
Ethics Committee authorisation 03/07.

REFERENCES

Baras, E., Jeandrain, D., Serouge, B. and Philippart,

J.C. (1998). Seasonal variations in time and space
utilization by radio-tagged yellow eels Anguilla
anguilla (L.) in a small stream. Hydrobiologia
371/372, 187-198.

Blundell, G.M., Maier, J.A.K. and Debevec, E.M. (2001).
Linear home ranges: effects of smoothing, sample
size, and autocorrelation on Kernel estimates.
Ecological Monographs 71(3), 469-489.

Burt, W.H. (1943). Territoriality and home range concepts
as applied to mammals. Journal of Mammalogy
24(3), 346-352.

Butler, G. (2001). Age, growth and telemetric trackingof
the eastern freshwater cod, Maccullochella ikei

Proc. Linn. Soc. N.S.W., 129, 2008

J.D. THIEM, B.C. EBNER AND B.T. BROADHURST

(Pisces: Percichthyidae) within the Mann—Nymboida
River System, NSW. Hons. Thesis, Southern Cross
University, Australia.

Cadwallader, P-L. and Gooley, G.J. (1984). Past and
present distributions and translocations of Murray
Cod Maccullochella peelii and Trout Cod M.
macquariensis (Pisces: Percichthyidae) in Victoria.
Proceedings of the Royal Society of Victoria 96(1),
33-43.

Crook, D.A. (2004a). Is the home range concept
compatible with the movements of two species of
lowland river fish? Journal of Animal Ecology 73,
353-366.

Crook, D.A. (2004b). Movements associated with home-
range establishment by two species of lowland river
fish. Canadian Journal of Fisheries and Aquatic
Sciences 61, 2183-2193.

Crook, D.A., Robertson, A.I., King, A.J. and Humphries,
P. (2001). The influence of spatial scale and habitat
arrangement on diel patterns of habitat use by two
lowland river fishes. Oecologia 129, 525-533.

David, B.O. and Closs, G.P. (2001). Continuous remote
monitoring of fish activity with restricted home
ranges using radiotelemetry. Journal of Fish Biology
59, 705-715.

David, B.O. and Closs, G.P. (2003). Seasonal variation in
diel activity and microhabitat use of an endemic New
Zealand stream-dwelling galaxtid fish. Freshwater
Biology 48, 1765-1781.

Douglas, J.W., Gooley, G.J. and Ingram, B.A. (1994).
“Trout cod, Maccullochella macquariensis (Cuvier)
(Pisces: Percichthyidae), resource handbook
and research and recovery plan’. (Department of
Conservation and Natural Resources: Victoria).

Ebner, B., Johnston, L. and Lintermans, M. (2005).
‘Re-introduction of trout cod into the Cotter River
catchment’. (Environment ACT: Canberra).

Ebner, B., Thiem, J., Lintermans, M. and Gilligan, D. Eds.
(2006). ‘An ecological approach to re-establishing
Australian freshwater cod populations: an application
to trout cod in the Murrumbidgee catchment’. (Parks,
Conservation and Lands: Canberra).

Ebner, B.C., Thiem, J.D. and Lintermans, M. (2007).

Fate of 2 year-old, hatchery-reared trout cod
Maccullochella macquariensis (Percichthyidae)
stocked into two upland rivers. Journal of Fish
Biology 71, 182-199.

Gilligan, D.M. (2005). “Fish communities of the
Murrumbidgee catchment: Status and trends’. (NSW
Department of Primary Industries: Cronulla).

Growns, I., Wooden, I. and Schiller, C. (2004). Use of
instream wood habitat by Trout Cod Maccullochella
macquariensis (Cuvier) in the Murrumbidgee River.
Pacific Conservation Biology 10, 261-265.

Gust, N. and Handasyde, K. (1995). Seasonal
variation in the ranging behaviour of the Platypus
(Ornithorhynchus anatinus) on the Goulburn River,
Victoria. Australian Journal of Zoology 43, 193-208.

Harris, J.H. and Rowland, S.J. (1996). Family
Percichthyidae: Australian freshwater cods and
basses. In ‘Freshwater Fishes of South-Eastern

Australia, 2™ edn’ (Ed R.M. McDowall) pp. 150-163.

Proc. Linn. Soc. N.S.W., 129, 2008

(Reed Books: Australia).

Hiscock, M.J., Scruton, D.A., Brown, J.A. and Pennell,
C.J. (2002). Diel activity pattern of juvenile Atlantic
salmon (Salmo salar) in early and late winter.
Hydrobiologia 483, 161—165.

Hooge, P.N. and Eichenlaub, B. (1997). ‘Animal
Movement extension to arcview, version 1.1’. (US
Geological Survey: Anchorage).

Hughes, V. and Thoms, M.C. (2002). Associations
between channel morphology and large woody debris
in a lowland river. In “The structure, function and
management implications of fluvial sedimentary
systems’ (Eds F.J. Dyer, M.C. Thoms and J.M. Olley)
pp. 11-18. (International Association of Hydrological
Sciences: Oxfordshire).

Ingram, B.A. and Douglas, J.W. (1995). Threatened fishes
of the world: Maccullochella macquariensis (Cuvier,
1829) (Percichthyidae). Environmental Biology of
Fishes 43, 38.

IUCN (2006). 2006 IUCN Red List of Threatened Species.
<www.lucnredlist.org>. Accessed 26 August 2007.

Koehn, J. (1997). Habitats and movements of freshwater
fish in the Murray—Darling Basin. In “Proceedings
of the 1995 Riverine Environment Research Forum’
(Eds R.J. Banens and R. Lehane) pp. 27-32.
(Murray—Darling Basin Commission: Canberra).

Koehn, J. and Nicol, S. (1998). Habitat and movement
requirements of fish. In ‘Proceedings of the 1996
Riverine Environment Forum’ (Eds R.J. Banens
and R. Lehane) pp. 1—6. (Murray—Darling Basin
Commission: Canberra).

Koehn, J.D., Nicol, S.J. and Fairbrother, P.S. (2004).
Spatial arrangement and physical characteristics of
structural woody habitat in a lowland river in south-
eastern Australia. Aquatic Conservation: Marine and
Freshwater Ecosystems 14, 457-464.

Lintermans, M. and Phillips, B. (Eds) (2005). Manage-
ment of Murray Cod in the Murray—Darling Basin:
Statements, Recommendations and Supporting
Papers. (Murray—Darling Basin Commission:
Canberra).

Nicol, S.J., Barker, R.J., Koehn, J.D. and Burgman, M.A.
(2007). Structural habitat selection by the critically
endangered trout cod, Maccullochella macquariensis,
Cuvier. Biological Conservation 138(1—2), 30-37.

Nicol, S.J., Lieschke, J.A., Lyon, J.P. and Koehn, J.D.
(2004). Observations on the distribution and
abundance of carp and native fish, and their responses
to a habitat restoration trial in the Murray River,
Australia. New Zealand Journal of Marine and
Freshwater Research 38, 541-551.

NSW DNR (2004). New South Wales Department of
Natural Resources <http://waterinfo.dlwc.nsw. gov.
au>. Accessed 11 April 2005.

Simpson, R.R. and Mapleston, A.J. (2002). Movements
and habitat use by the endangered Australian
freshwater Mary River cod, Maccullochella peelii
mariensis. Environmental Biology of Fishes 65,
401-410.

Tabachnick, B.G. and Fidell, L.S. (1989). “Using
multivariate statistics’. (Harper and Row Publishers:
New York).

73)

(nitetlen % cdeott bo581)

nhl cate celiurtod “1M ctossell?.

re real tety vt ve oe toa 0G i j
are i) (ey besa gh tka | tae ead
Ales drink
j ; (ies. be
' Py pera tf
iariey Ate fais 1s at
ny PPS. et eee “Ut Va H
ithe yf V/ pe Ts Pir eie lamers eaered
ree both aes Pel MITE sal ea eer lel wh ui
i bee SE gT (P e S er ae
adit) hes b camreB yd Mr ide ete) eae
ce rs Aan aa Less 1

CN AFR 2 ee
90) EL puke bie Avil Sete
-) Medicis gn), bho ae
‘wats ony a“ hen eon vous i
a] + Lael
+S Satay pa 7 hin, a iA Pay}, mun: ee 4 Ps We i
Lowy A 2 eto «glen bet ein

lectin 1 YOR aw ei Te ee.
4 /h| eye 1). “Uautigeat bed nah -otleds rail
ote mosh peygegevad @apsie 2)? > sip

Mi acts x Hants, een apd t LH ek

(Tm) ptay acpy) 1 aye 34 boots Cntarne
Troareovine La) EPieta | Aue 4 fac i baie 1 ileenA
eT 4A} at we A advil beri
yak (A eK) rein Mat wal, atten De ele Poke}
rieeer SRST Gari) | Wh Pinned N bey

Lina) aries?
Me") 24 Weekes fine A age ~ CL ole
hag ke aay aidira earned Waka (ects aur Liste
Ok Boye Seat Ch ral te Po aoe Licieer nein
LDL GMa ae Bt Mel at) fee er
ert r bw ny Hin sohbet)
ww OE) sek gill) er hinie bf apabrusio J
TU TTR er ae
iturin? iwi) a4 aura i > Wrap Nil
joie cern es tack ell VwtaT we» eet
Precher >
1A Aer act 14 caltae al ob a SR oe
bie ah cea et eas lint, dirpeet aR LEAD
66 Joe igen bepreteieny

wHaLs
MI ak iti 2 an
TP 100 erwin teed’ ee MY
yi ot nal Wf Gli, Den aie 1: Bll ae
=f ys och 3 r
1 Pato) PP? WETS at Hanae «ee
ny: . fect Aaa eiga, TA RES * OU
2 PEE Wal ait =e
ofmornag! asitttiniviy weet hee at)
iM Dalene Teo Lene es
VAX Logs) Doe d.

Aoheter4 ine AN pagent

Wend CUE ye este ete Ts
; 4 ede ié alt aee 7
TW liahaisleA reas
by tob
+ bet 2 goede
urea
seal wel

a
~
IS

prrae aya

» Not ga Bid)

} a cad
ic

i - iy - ent aa
peste ct Shean ia Pane A A

ir elsaeptaiee NetneNCTN one abt Aira

Selon Saher eet: eh a

: < (peed, tg wi (ita De _calibte®t

aut ati= louian aie p 5 aewentiy ald CRM Adiga

aes Linbbaalat ged vd ‘rhomnr ast ack zo a

Stel, Ware a NN ie Resaney

ERS Aes I
attend peat? ia. : pie he ne
hind tes Vie RN 0 ser
or ved Sew thd Siig falbae inawragier

te Che PEt on wax) aeatt ravin tently

etwruth aaa YO conlvacand grindsine I

cqevenl A474 Yo Voicnigl..qirotaeigvatbay ae beg

zee per sg
arya * La fet ea
Aon Ahged pm po
“1s Vikas ‘open. aay Te:
SELES RUBIN fhe
ees | Pas te ee
ay ae ) zt 7 pon, niga!
Aichi ea soy
ad tl panier a)
partideg ‘eset gad pes he howe f
coe Mj pereenan, > ay eee D a
aye) ipamdtaw.y apebicernauM ant 2) faba ston
a eayineee ee aE ie p
(0k, Bs Preah Sy ahig’ 5 i gaol). Gaae I:
wi wt bowers dzoatindl bicnasers | * to omit .
nage pla erst) par fay Motel aye
Wwe vo Fetus ey ey wa fadon,
Seta re suai
HW Te eth heap as" PONT) ET Se
Nev >) le hal ae pepe Bem i
- (ENS eh aT ele _
1) ey Bee oy et NRE ING oP tl tomy 09;
sad eel Ta Pia T of arene Sono uma
eB ae itt stb lt faa esy HT Mar i.
winery =
=f tf) SE Achat ae ee Maat,
Nabe aT suvtedeae Bataitiae = :
GAM i boda gr ar a7
H badder bea mat

eat:

ol Hea eS TPA
Lobel qo iiqetistt Mlb) ane ‘om

HOG. WY, Sh a

ay

Fragmented Distribution of a Rock Climbing Fish, the
Mountain Galaxias Galaxias olidus, in the Snowy Mountains

KEN GREEN

Snowy Mountains Region, National Parks and Wildlife Service, PO Box 2228, Jindabyne NSW 2627.

Green, K. (2008). Fragmented Distribution of a rock climbing fish, the Mountain Galaxias Galaxias
olidus, mm the Snowy Mountains. Proceedings of the Linnean Society of New South Wales 129, 175-182.

Fish were surveyed visually from 1,500 m elevation to the highest known altitude for Mountain Galaxias
Galaxias olidus of 2,137 m on the slopes of Mt. Kosciuszko (2,228 m). Above 1,500 m, where the species
is the only galaxtid and is physically isolated from all lowland populations, there was further isolation with
76 disjunct populations within the 1,400 km? area surveyed. Trout (Salmonidae) were the main cause of
this isolation because they occupied 95.85 km of the major streams, generally in the main valleys at lower
elevations but reaching up to 1,800 min places, and leaving only the headwaters unoccupied. The distribution
of G. olidus above 1,500 m was, therefore, determined largely by topographic and anthropogenic barriers
to the movements of trout. Despite beg recorded as absent from western drainages in the mountains,
including two of the five glacial lakes, since as long ago as the 19th century, G. o/idus moved into Australia’s
highest lake (Lake Cootapatamba) during the course of the survey with serious implications for biodiversity

in this newly occupied lake.

Manuscript received 15 November 2007, accepted for publication 6 February 2008.

KEYWORDS: Lake Cootapatamba, Mountain Galaxias, Oncorhyychus mykiss, Salmo trutta, Trout.

INTRODUCTION

The Mountain Galaxias Galaxias olidus is
found from sea level to above the treeline in the Snowy
Mountains (McDowall and Frankenberg 1981; Green
and Osborne 1994). Throughout its range it has a
widespread, though fragmented, distribution, largely
due to the effect of alien trout (Salmonidae), and is
most commonly found in the smaller headwaters of
streams or above some obstacle to upward movement
by trout (Cadwallader and Backhouse 1983). The
higher altitudes of the Snowy Mountains that support
a winter snow cover for at least one month per year
(1,500 m a.s.1.) drain into tributaries of the Murray
and Snowy Rivers. Many of these tributaries have been
stocked with trout and most have been impounded at
one or more points along their length for generation of
hydro-electricity and for irrigation. The high altitude
populations of native fish are therefore isolated from
lowland populations by a number of barriers. Within
this high altitude area, there is further subdivision of
the contiguous area above 1,500 m (Fig. 1). This area
is bounded on the north and east by the Eucumbene

River. This, together with the Thredbo River that
drains much of the southern boundary then joins
the Snowy River above Jindabyne Dam where the
Mowamba River also arrives through an aqueduct.
Below the dam, only the Jacobs River, draining areas
above 1,500 m, continues to flow unimpounded into
the Snowy River. On the western side, the main
drainages in the north and west, the Tumut, Tooma,
Geehi and Swampy Plains Rivers and Bogong Creek
are dammed and, of the major catchments, only water
bodies from Leatherbarrel Creek (Fig. 2) southward
flow unimpounded into the lowlands (<400 m
elevation) where they join the Murray River.

In this mountain area above 1,500 m (the
subalpine and alpine zones) there is only one native
species of fish, G. o/idus (Green and Osborne 1994).
Although currently treated as one species, there are
several forms of G. olidus (Raadik 2001; Raadik and
Kuiter 2002). Ogilby (1896) stated that above the
winter snowline, a stout, sombre-coloured form was
found in deep, still pools and smaller ponds while a
slender, brilliant-coloured form occurred in rapidly
moving waters with gravelly or sandy shallows.

MOUNTAIN GALAXIAS IN THE SNOWY MOUNTAINS

Figure 1. The study area showing the largest con-
tiguous area of the Snowy Mountains above the
1,500 m contour and the main rivers and dams es-
sentially isolating Galaxias olidus within this area.

Between the two extremes, ‘every conceivable
variety, both of contour and colour may be found’
(Ogilby 1896). There is also a high level of genetic
distinctiveness of populations of G. olidus across the
landscape (Raadik and Kuiter 2002; T. Raadik pers.
comm. 2006).

The salmonids Brown Trout (Salmo trutta) and
Rainbow Trout (Oncorhyychus mykiss), native to the
northern hemisphere, were introduced into the Snowy
Mountains in the nineteenth century, with the Brook
Trout (Salvelinus fontinalis) and the Atlantic Salmon
(Salmo salar) introduced later (Cullen and Greenham
1980). However, only the first two are now common in
the Snowy Mountains. The upstream invasion of the
mountain streams from introduced stock has resulted
in a fragmentation of higher elevation populations
of G. olidus. Trout are a major threat to many native
Australian fish species, particularly galaxiids and,
‘completely replace galaxiid populations, leading
to local extinctions’ (Raadik and Kuiter 2002). The

176

consequences of the invasion between 1971 —1974
by O. mykiss of a stream containing only G. olidus
was documented by Tilzey (1976). Later, Lintermans
(2000) documented the re-invasion of a montane
stream by G. olidus once O. mykiss was removed.

The aim of the present study was to survey
the high altitude streams of the Snowy Mountains
(> 1,500 m elevation) to determine the extent of the
invasion by trout and to assess the degree of isolation
of remaining populations of G. olidus.

METHODS

Climbing of wet rocks by G. olidus was recorded
on video camera in a tributary of Dicky Cooper Creek
at the location that climbing by the species was first
described by Green (1979). The video was watched in
slow motion in an attempt to observe the method of
climbing. To examine the mechanism of climbing, the
lateral surfaces of G. olidus collected from the outlet
creek of Lake Cootapatamba were examined beneath
a dissecting microscope and photographed.

Mountain streams above 1,500 m were chosen
for survey from maps at a scale of 1: 50,000 for
the whole Snowy Mountains and at 1: 25,000 for
the Kosciuszko Main Range. These streams and
additional ones observed in the course of the survey
were walked over a period of three years, December
2004—May 2007. Stream waters were extremely clear,
and streams were investigated visually only on calm
days. Fishes clearly above the maximum length of G.
olidus (135 mm -McDowall 2006) were recorded as
trout. No attempt was made to differentiate the species
of salmonid further because their fry are difficult to
identify (Tilzey 1976). Fish in the size range of G.
olidus were inspected with binoculars to determine
species. Differing behaviour and colour pattern were
good indicators of species, but identification of all
fish was based on the location of the dorsal fin, high
on the back in trout and located well back on the
body in G. olidus. Reference specimens of G. olidus
were collected only from the outlet creek of Lake
Cootapatamba. Streams were investigated upwards
from the 1,500 m contour with species of fish and any
barrier to their upstream movement recorded. These
streams were walked below and above these barriers
(such as waterfalls sufficiently high to exclude trout)
until no further trout were seen and the presence or
absence of G olidus was recorded. Because absence
cannot be confirmed, streams in which G olidus was
not recorded were walked until they became a trickle
or dried up completely. Streams checked from higher
altitude downstream were checked only until trout

Proc. Linn. Soc. N.S.W., 129, 2008

Figure 2. One of the two largest stream complexes in the Snowy
Mountains containing Galaxias olidus, together with the five glacial
lakes and the factors isolating the populations. Rivers with dotted
lines contained Galaxias olidus, dashed lines no fish seen, solid line
trout. W = waterfall, i= water intake, D= dam.

were the only fish observed in the stream or until the
1,500 m contour was reached (although some streams
were descended further).

The species ‘trout’ and/or G. olidus, together
with barriers above which one or both species were
missing were recorded on a hand-held GPS. All
locations were plotted onto maps at 1: 50,000 and the
length of streams occupied by trout was measured
using a map wheel.

The presence of G. olidus at its highest elevation
(in the headwaters of Rawsons Creek) was recorded

Proc. Linn. Soc. N.S.W., 129, 2008

regularly throughout the snow-
free seasons to determine date of
arrival and departure of fish. This
was at the upper limit of surface
flowing water in the Snowy
Mountains where the water pooled
at a drain with a depth generally
<30 mm on a rocky substrate
in an area of < 1 m7’, before
being diverted through a culvert
beneath the Mt. Kosciuszko
summit walking track. Presence
or absence was therefore easy to
determine.

RESULTS

Evidence from the video of
climbing by G. olidus showed
that while in contact with the
rock G. olidus lay with pectoral
and anal fins flattened against the
rock. Propulsion was by using
the fins as independent limbs and
‘walking’ (or rather scuttling) up
the rock. Examination beneath
a dissecting microscope and
subsequent photographs (Figs 3
and 4) show that both the pectoral
and anal fins have rugosities on
the ventral surfaces that might be
used in climbing.

Within the. 1,400 km? of
contiguous land above 1,500 m
altitude, salmonids occupied 33
water courses totalling 95.85 km,
generally at the lower elevations
but reaching up to 1,800 m in a
number of tributaries of the upper
Snowy River (Figs | and 2). Only
one stream system, Valentines
Creek, had trout and G. olidus
coexisting in the same reaches. Galaxias olidus was
recorded from 86 stream systems with 76 of these
being isolated from other water bodies containing
G. olidus by trout downstream. There were only two
large complexes of streams containing G. olidus,
both protected by waterfalls at the downstream ends
of their range, these were the upper Geehi River
system and the upper Snowy River system (Figs 1
and 2). Elsewhere, G. o/idus occurred only in isolated
streams. The fine scale disjunct distribution of G.

ev

MOUNTAIN GALAXIAS IN THE SNOWY MOUNTAINS

of 2,137 m, just 91 m lower than
the summit of Mt. Kosciuszko.
Here two streams, tributaries of
Rawsons Creek, narrowed to
trickles 5-10 cm wide and less
than 1 cm deep, and in places
flowed over grass, below culverts
under the Mt. Kosciuszko summit
walking track. Above the culverts,
drains fed by seepages off the side
of Mt. Kosciuszko maintained
semi permanent water bodies.
Galaxias olidus was restricted
in upward movement by waterfalls
that exclude it from Wilkinsons
Creek complex and from many of
the steeper streams on the western
faces of the Main Range (Fig. 2).
Steep streams on the northern side
Figure 3. The lateral surfaces of the pectoral fin of Galaxias olidus. of the Thredbo Valley generally
also lacked fish. These streams
lacked large waterfalls but were
generally small and fast- flowing with few pools.
Additionally many had long areas of continuous cover

olidus is illustrated for the Main Range of the Snowy
Mountains, which comprises 15% of the area above
1,500 m with 28 isolated
populations in about
200 km? of the highest
country in Australia
(Fig. 2).

The lower limits of
distribution of G. olidus
were set by features
thought to be impassable
to trout. Situations where
G. olidus was isolated
from trout included
galaxiid occurrence in
enclosed ponds (three
complexes within areas
otherwise dominated by
trout), anthropogenic
barriers (11) and
waterfalls (33). Barriers
such as waterfalls,
chutes and artificial
interception of flows (for

aqueducts) that blocked
trout movement, blocked Figure 4. The lateral surfaces of the anal fins of Galaxias olidus.

upward movement of
all species at different locations. Anthropogenic
structures were either themselves a barrier, or caused
a barrier to fish movement by eliminating water flow
in some downstream reaches of streams.

Galaxias olidus was recorded up to an altitude

of shrubs, particularly tea tree (Leptospermum spp.)
that shaded the streams. Where G. olidus did occur
in a particular stream few were recorded in similarly
shaded reaches.

For the first time ever, G. ol/idus was recorded in

178 Proc. Linn. Soc. N.S.W., 129, 2008

K. GREEN

Australia’s highest lake, Lake Cootapatamba (2,050
m). During the survey, G. olidus was noted in the
stream draining Lake Cootapatamba in December
2004 and again in mid November 2005 but not in
the lake. It was first noted in the lake on 9 Jan 2006
and by March 2007, a school covering approximately
2 m°* in shallow water near the outlet contained an
estimated 2,000 fish.

Fish were commonly seen above the Kosciuszko
summit walking track in two small ponds at the
headwaters of Rawsons Creek and during this study
were first noted on 9 and 18 November in 2006 and
2005 respectively. In 2005 they were last observed on
2 May. In 2006 they were last observed on 20 March
and, by 11 April, the area was covered in snow and
remained that way for the winter. On 30 Mar 2007,
with an almost full cover of snow, G. olidus was still
present, but had departed by 19 April when none
could be found. Fishes remained in the ice-covered
Blue Lake, Hedley Tarn and Club Lake throughout
winter. One fish was observed swimming beneath the
ice of Club Lake on 15 Jun 2007 and fish were seen
in a part of the lake that thawed early in September
2006, whilst the outlet was still blocked by snow
and ice, indicating that they had wintered over in the
lake.

DISCUSSION

Rock climbing

To understand the distribution of G. olidus we
need to understand its ability to ascend steep streams
and survive in shallow water. McDowall (2006)
considered that G. olidus, unlike G. brevipinnis, does
not have a reputation as a good climber. In fact, G.
olidus is very capable of climbing on vertical or even
overhanging rock, as first reported from the Snowy
Mountains by Green (1979), however, specimens
that were collected at the time were originally
misidentified as G. brevipinnis. The fishes observed
by Green (1979) were 30 to 60 mm long, climbing
being probably aided by their small size giving them
high surface area to weight ratio (McDowall 2006).

Climbing by G. olidus includes at least three
important elements. The ability to jump out of water,
to a height at least four times its length (Green 1979),
allows it to gain contact with surfaces that do not
extend to water level. The next important ability is
to adhere to the rock, whether establishing contact
immediately after a jump or maintaining contact
whilst climbing. The third ability, while retaining
adhesion, is to gain propulsion for climbing. Green
(1979) wrote that G. olidus lacked any specialized

Proc. Linn. Soc. N.S.W., 129, 2008

organs for adhesion and appeared to rely upon the
surface tension of a thin film of water between its
ventral surface and the rock, as also suggested for
diadromous Galaxias by McDowall (2006). As with
G. brevipinnis (McDowall 2006), downward facing
pectoral and anal fins provide surface contact in G.
olidus. In addition to adhesion, these fins were also
observed to be important in propulsion. McDowall
(2003, 2006) illustrated the rugosities on the ventral
surface of pectoral fins of G. brevipinnis, and the
present study has found similar rugosities on the
ventral surfaces of both the pectoral and anal fins of
G. olidus (Figs 3 and 4).

Despite its climbing ability, G. olidus was
absent from a number of streams at high altitude
that apparently have never held trout or galaxiids.
Green (1979) noted that the successful climbers
were those that kept out of the main flow of falls
but remained on rock adjacent to the main flow, that
was only occasionally washed by water. In certain
circumstances wet rock may not occur adjacent to the
main flow (where the water flows over an overhang
and drops sheer) or the moist zone may be unsuitable
(such as where it becomes a haven for moss).

Isolated populations

McDowall (2006) has shown that throughout
the cool temperate southern hemisphere, with few
exceptions, galaxioid fishes are adversely affected
by introduction of trout, with a major decline in the
Galaxiidae in particular. The general pattern of fish
distribution in the Snowy Mountains and Victorian
Alps is one of trout occupying the main stream, while
Galaxias spp. are usually only found in upstream
water bodies inaccessible to trout (McDowall and
Frankenberg 1981). Jackson (1981) stated that, ‘above
the snowline, instances of trout being the only fish
below a waterfall and Galaxias the only fish above
the waterfall are common in our experience.’

Tilzey (1976) sampled a stream in the Lake
Eucumbene catchment in the Snowy Mountains
twice in the course of three years during which time
the stream was invaded by Rainbow Trout. In 1971
the stream contained only G. olidus below a waterfall
but by 1974 trout had spread upstream as far as the
waterfall and G. olidus had disappeared. This same
pattern occurred in the present study. Throughout the
Snowy Mountains in the 1,400 km? of land above
1,500 m, there were 76 isolated populations of G.
olidus. Infiltration of trout was the major cause of
this isolation, and they occupied 33 water courses
totalling 95.85 km downstream of populations of G.
olidus and separated from them mainly by waterfalls.
A secondary isolating mechanism was anthropogenic

179

MOUNTAIN GALAXIAS IN THE SNOWY MOUNTAINS

—dams for aqueduct intakes isolating upstream from
downstream reaches of the same stream. The third
cause was isolation within ponds (such as those near
Mt. Clarke) that are currently well separated from
other water bodies. Connecting water courses may
have existed in the past possibly as erosion runnels,
particularly as a result of cattle grazing activity that
has since ceased, allowing the runnels to revegetate
as recorded elsewhere (Green et al. 2005; Green pers.
obs.).

The failure to find G. olidus in the upper regions
of some trout-accessible streams (where trout are,
however, lacking) may be because the reaches are
uninhabitable in some seasons (under ice) or some
years (drought) or because, as downstream migration
results in predation it cannot be balanced by upstream
migration. Also we cannot now tell which headwaters
have had introductions of, or have been invaded by,
trout in the past resulting in the extirpation of G.
olidus followed by the loss of the trout. Similarly,
some streams had no trout during this survey although
they appeared suitable, and instead had very low
numbers of G. olidus. This may have been caused by
a die-off of trout in warm waters in the drought that
took place over most of the survey period, followed
by a re-invasion by G. olidus of these waters as
observed in Victoria by Closs and Lake (1996). If
this re-invasion also implies continuous downstream
movement by G. olidus when trout were present then
this would result in a continuous loss of individuals to
predation (McDowall 2006). However, where there
were refuges such as shallow, warm anabranches as
in Valentine Creek, both trout and G. olidus occurred.
The trout were present in low numbers and the G.
olidus were in considerably lower numbers than in
nearby streams without trout.

Macleay (1882) described galaxiids from the
higher altitudes of the Snowy Mountains (‘captured

. on Mt Kosciusko’) as G. findlayi, but the type
specimens were subsequently lost (Raadik and Kuiter
2002). Helms (1890) collected specimens possibly in
Diggers or Pipers Creek (Raadik and Kuiter 2002),
and the species was later reported from the headwaters
of the Snowy and Thredbo Rivers (Ogilby 1896). It
is possible that the exact sites where these original
collections came from may still support G. olidus.
One stronghold of the species is the higher altitudes
in the headwaters of the Snowy River (Fig. 2) and G.
olidus was recorded in the headwaters of both Diggers
and Pipers Creek, and although it could not be found
in Thredbo River it did occur in isolated headwaters
of some tributaries.

Some populations have possibly been isolated for
periods of years if not decades and that isolation may

180

be permanent. For example, ponds along Valentines
Creek are as high as two metres above the water level
of the creek. These ponds would presumably flood
exceedingly rarely and with already reduced snowfall
(Green and Pickering 2002) and with predictions of
further reductions in total precipitation (Whetton
1998) the flood conditions needed to inundate these
ponds may not occur again. Ponds near Mt. Clarke
have no nearby water course, and survival of G.
olidus in them demonstrates the resilience of the
species as also demonstrated when depleted oxygen
extirpated Common Carp Cyprinus carpio but not G.
olidus (Lake 2003). As a result of the drought the Mt.
Clarke ponds dried up for an extended period between
17 Feb and 20 Mar 2006, and no water was visible
on the surface on 13 March. However, once rains
came, fish reappeared in the ponds having survived
the intervening weeks, presumably in burrows of
Euastacus sp. Galaxias olidus is known to travel in
underground sections of streams (Lintermans 2000)
and probably does so in the many water bodies in the
Snowy Mountains that have underground sections
incised into a bog. However, the ability to remain
underground during drought for this length of time is
an interesting response to existence in the ephemeral
headwaters of streams and ponds. In the same period,
G. olidus at the highest altitude in the headwaters of
Rawsons Creek was cut off from the main water body
when the outlet from the drains it was 1n dried up, but it
was able to remain in a small seepage until conditions
improved. Its survival in Club Lake over winter is
also of interest because the ice sits on the mud floor
of this lake even in poor snow years constraining the
possible over wintering locations for G. olidus.

Limitations to upstream movement

Because G. olidus can move in small trickles,
it can ascend right to the headwaters of streams and
this makes it impossible to put an upper limit on its
distribution in any network of streams. Galaxias
olidus penetrated to streams above the summit
walking track of Mt. Kosciuszko by November of
each year, and departed once the ice from the first
heavy frosts appeared on the water, at times varying
between April and May. Walford (1928) mentioned G.
olidus seeking lower altitudes in winter “about Mount
Kosciusko’ but gave no other details. However, fish
remained under ice in Hedley Tarn, Blue Lake and
Club Lake, indicating that cold itself was not a major
factor, although in shallow waters, physical contact
with the ice may have to be avoided to prevent
injury.

The valley wall on the southern side of the
Thredbo River, a major trout river, is very steep

Proc. Linn. Soc. N.S.W., 129, 2008

K. GREEN

and, between the confluence with the Little Thredbo
River in the east and Dead Horse Gap in the west,
five isolated populations of G. olidus were recorded,
all above waterfalls. The northern side, with a similar
gradient, however, lacked G. olidus except in one arm
of Bogong Creek and the headwaters of a creek on
Merritts Spur. The isolating factor here may just be
the continuous steepness of the creeks and the general
overgrowth of tea tree (Leptospermum spp.) in the
lower reaches. Although G. o/idus is able to travel in
underground sections of streams (Lintermans 2000),
densely shaded reaches of stream were not normally
inhabited, and open reaches of such streams above
or between shaded areas also often lacked fish. This
shading may have been sufficient to prevent the ascent
of G. olidus into apparently suitable waters arising
between North Ramshead and Ramshead. Galaxias
olidus may also be intolerant of continuous steep
gradients, as its favoured location appears to be in
slow flowing stretches and pools beneath waterfalls
and riffles (pers. obs.). However, not knowing the
history of some of these waters, trout may once have
been introduced. However, between Bogong Creek
in the SW and the lower reaches of Lady Northcote
Creek in the NW, G. olidus has historically been
absent above 1,500 m including the two lakes in
these catchments Lake Cootapatamba and Lake
Albina (Fig. 2). Anthropogenic barriers constructed
in streams above naturally occurring barriers to the
movement of trout have isolated populations of G.
olidus on the western faces. Waterfalls, however,
have been the main limiting factor historically to the
upstream spread of G. olidus on this side. The upward
movement of G. olidus appears to have been barred at
about the 1,500 m contour in a number of streams on
the western side at large waterfalls, possibly due to a
band of rock at that altitude more resistant to erosion.
Such falls occur on Wilkinson’s Creek, which still
has no G. olidus in its upper reaches (although
it must occur lower because this is a tributary of
Swampy Plains River from which the fish in Lake
Cootapatamba ascended). Once above the falls on
Swampy Plains River, G. olidus ascended into two
main tributary arms including that flowing from Lake
Cootapatamba.

Lake Cootapatamba

Helms (1890) wrote of Lake Cootapatamba
and Wilkinsons Valley, ‘the absence of Galaxias at
this elevation struck me as peculiar. It is, however,
remarkable that on the Snowy River side these fishes
are met with almost everywhere.’ Galaxias olidus
was not recorded in Lake Cootapatamba at any time
in the 20th century despite a number of investigations

Proc. Linn. Soc. N.S.W., 129, 2008

of the lake (Green and Osborne 1994). Neither were
fish recorded in Lake Cootapatamba by Timms
(2002). However, in the present study they were
found in Swampy Plains River, which flows out of
Lake Cootapatamba, as early as December 2004.
By mid November 2005 they were still found in the
outlet creek but not in the lake, but in January 2006
they had progressed upstream into the lake. Along
the western fall of the Snowy Mountains from Crags
Creek that empties into the Geehi River (Fig. 2), 24
km south to Cascade Creek that flows into the Murray
River (Fig. 1), no G. olidus were found except in
two catchments. In both Leatherbarrel Creek and
Swampy Plains River, G. olidus was found in 2004.
There is no barrier to movement of these fish from
a waterfall at about 1,500 m to Lake Cootapatamba,
and certainly nothing that would seem likely to delay
the arrival of G. olidus by over 110 years since the
visit by Helms (1890). It would appear that G. olidus
ascended this waterfall sometime between 2002 and
2004. An obvious issue here is the role of a major fire
in the area in 2003. Elsewhere during this survey it
was observed that debris from the fire washed down
creeks and formed dams that raised the water level
below waterfalls. If such an occurrence took place
below a waterfall that had previously prevented
upstream movement it could have the effect of aiding
the ascent by G. olidus.

Translocation of galaxioids into new areas has
caused adverse impacts (McDowall 2006). The
spread of G. brevipinnis from eastern catchments into
the drainage of the Murray River due to diversion of
streams for hydro-electricity and irrigation has been
of concern (McDowall 2006), although the impacts of
this are not clear and the species appear to be cohabiting
(T. Raadik pers. comm. 2007). An expansion of G.
olidus into Lake Cootapatamba might have serious
consequences, and because a lake ecosystem is not
as open as a stream, the effects of changed predation
patterns on biodiversity might be dramatic. Already
the Kosciuszko endemic cladoceran Daphnia nivalis
seems to be in decline and studies of the impact over
the next few years will be crucial in understanding
this impact (Yoshi Kobayashi pers. comm. 2007).

Conclusion

If the mountain galaxias was just one species
also found at different locations down to sea level,
the disjunct distribution in the Snowy Mountains
could perhaps be seen to be of only academic interest.
However, because it is widespread, there is currently
no explicit concern for its conservation (McDowall
2006). Even so, isolation at the local scale can have
serious consequences because small populations

181

MOUNTAIN GALAXIAS IN THE SNOWY MOUNTAINS

are more susceptible to declines from stochastic
effects, be they ecological or genetic (McDowall
2006). Although McDowall and Frankenberg (1981)
brought together a group of poorly defined taxa
under G. olidus (McDowall 2006), it appears that
this is a species complex (Raadik 2001; Raadik and
Kuiter 2002). There is a possibility, therefore, that
some of the species in this complex may occur as
isolated populations and may be at risk owing to the
failure to recognise these new taxa, and so there is a
danger that appropriate conservation measures might
not be considered until it is too late (see McDowall
2006). Pond-bound populations may also die out
in droughts, and groups trapped above permanent
barriers (such as hydro-electric infrastructure) are at
long-term risk from inbreeding. Loss of any of these
isolated populations would at least result in reduction
of the species’ genetic diversity, but may result in
loss of discrete taxa and could have series effects on
ecosystem function (McDowall 2006).

ACKNOWLEDGEMENTS

Thanks to Harvey Marchant for taking the
photographs. Bob McDowall and Tarmo Raadik
commented on the manuscript. Tarmo Raadik also
confirmed the identification of the galaxiids from Lake
Cootapatamba.

REFERENCES

Cadwallader, P.L., and Backhouse, G.N. (1983). °A guide
to the freshwater fish of Victoria’. (Government
Printer: Melbourne).

Closs, G.P. and Lake, P.S. (1996). Drought, differential
mortality and the coexistence of native and an
introduced fish species in a south east Australian
intermittent stream. Environmental Biology of Fishes
47, 17 —26.

Cullen, P. and Greenham, P. (1980). Aquatic ecosystems.
In ‘The Conservation Status of Kosciusko National
Park’. (Ed. A. Turner) pp. 29-34. (Government
Printer: Sydney).

Green, K. (1979). Observations on rock climbing by the
fish Galaxias brevipinnis. The Victorian Naturalist
96, 230-231.

Green, K. Good, R.B. Johnston, S.W. and Simpson, L.A.
(2005). Alpine grazing in the Snowy Mountains
of Australia: degradation and stabilisation of the
ecosystem. In “Land Use Changes and Mountain
Biodiversity’. (Eds E.M. Spehn, M. Liberman, and
C. K6émer) pp. 213-225 (CRC Press: Boca Raton FL,
USA).

Green, K. and Osborne, W.S. 1994. ‘Wildlife of the
Australian Snow-Country’. (Reed: Sydney).

Green, K. and Pickering, C.M. (2002). A scenario
for mammal and bird diversity in the Australian
Snowy Mountains in relation to climate change. In
‘Mountain Biodiversity: a Global Assessment’. (Eds
C. Komer and E.M. Spehn) pp. 241-249. (Parthenon
Publishing: London).

Helms, R. (1890). Report on a collecting trip to Mount
Kosciusko. Records of the Australian Museum 1, 11-
16.

Jackson, P.D. (1981). Trout introduced into south-eastern
Australia: Their interaction with native fishes. The
Victorian Naturalist 98, 18 -24.

Lake, P.S. (2003). Ecological effects of perturbation by
drought in flowing waters. Freshwater Biology 48,
1161-1172.

Lintermans, M. (2000). Recolonization by the mountain
galaxias Galaxias olidus of a montane stream after
the eradication of rainbow trout Oncorhynchus
mykiss. Marine and Freshwater Research 51, 799-
804.

Macleay, W. (1882). On a species of Galaxias found in the
Australian Alps. Proceedings of the Linnean Society
of New South Wales 7, 106-109.

McDowall, R.M. (2003). The key to climbing in the koaro.
Water and Atmosphere 11, 16-17.

McDowall, R.M. (2006). Crying wolf, crying foul, or
crying shame: alien salmonids and a biodiversity
crisis in the southern cool-temperate galaxioid fishes?
Reviews in Fish Biology and Fisheries. 16, 233-422.

McDowall, R.M. and Frankenberg, R.S. (1981). The
Galaxiid fishes of Australia (Pisces: Galaxiidae).
Records of the Australian Museum 33, 443-605.

Ogilby, J.D. (1896). On a Galaxias from Mount
Kosciusko. Proceedings of the Linnean Society of
New South Wales 21, 62-73.

Raadik, T.A (2001). Kosciuszko When is a mountain
galaxias not a mountain galaxias. Fishes of Sahul 15,
785-789.

Raadik, T.A and Kuiter, R.H. (2002). Kosciuszko
Galaxias: a story of confusion and imminent peril.
Fishes of Sahul 16, 829-835.

Tilzey, R.D.J. (1976). Observations on interactions
between indigenous Galaxiidae and introduced
Salmonidae in the Lake Eucumbene catchment,

New South Wales. Australian Journal of Marine and
Freshwater Research 27, 551-564.

Timms, B.V. (2002). Lake Cootapatamba. In “Biodiversity
in the Snowy Mountains’. (Ed. K. Green) pp. 98-101.
(Australian Institute of Alpine Studies: Jindabyne).

Walford, F. (1928). The mountain minnow. Australian
Museum Magazine 3, 274-277.

Whetton, P.H. (1998). Climate change impacts on the
spatial extent of snow-cover in the Australian Alps.
In ‘Snow: A natural history; an uncertain future’.
(Ed. K. Green) pp. 195-206. (Australian Alps Liaison
Committee: Canberra).

Proc. Linn. Soc. N.S.W., 129, 2008

Trilobite-constrained Ordovician Biogeography of China with

Reference to Faunal Connections with Australia

ZHOU ZHI-yt AND ZHEN YONG-Y?

‘Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences, Nanjing 210008, P.R. China
(zyizhou@jlonline.com); *Australian Museum, 6 College Street, Sydney NSW 2010, Australia (yongyi.

zhen@austms.gov.au)

Zhou, Z.Y. and Zhen, Y.Y. (2008). Trilobite-constrained Ordovician biogeography of China with reference
to faunal connections with Australia. Proceedings of the Linnean Society of New South Wales 129, 183-
195.

All the plates and most of the terranes in China exhibit close biogeographic links and may have formed
part of eastern Peri-Gondwana during the Ordovician. Synthetic analysis based largely on the platform/
inner shelf trilobite faunas suggests that the Chinese eastern Peri-Gondwanan plates and terranes may have
belonged to a single biogeographic province during the Tremadocian (Tremadoc) and late Katian-Himantian
(Ashgill), but may be separated into two sub-provinces during the Floian-early Katian (Arenig-Caradoc):
one consists of South China, Tarim and Annamia, and the other may include North China, Sibumasu,
Southern Tibet, Tianshan-Beishan and possibly Hainan. However, the deep-water facies trilobites of the
relevant Chinese geographic units had progressively become more unified from the mid Darriwilian to
early Katian (Llanvirn to Caradoc) before the sub-provinces eventually broke down by the late Katian
(Ashgill). Australian Ordovician trilobite faunas had close affinities with most of the Chinese eastern Peri-
Gondwanan plates and terranes, but closest biogeographic links were in particular with North China and

Middle Tianshan-Beishan.

Manuscript received 6 August 2007, accepted for publication 6 February 2008.

KEYWORDS: Australia, biogeography, China, Ordovician, trilobites.

INTRODUCTION

Eleven Chinese Ordovician geographic units,
corresponding mainly to different allochthonous
continental masses (plates and terranes), are divided
largely on the basis of evidence from regional
tectonics, palaeogeography and stratigraphy (Fig. 1).
Units or regions now in close proximity were distantly
separated from each other during the Early Palaeozoic,
and the assembly of those landmasses likely
underwent a long-sustained process that extended
from the Late Carboniferous to Cenozoic (Zhou et al.
2007). However, trilobite evidence indicates that all
the plates and most of the terranes (with exceptions
noted below) in China exhibit close biogeographic
links to each other and may have formed part of
eastern Peri-Gondwana during the Ordovician (e. g.
Zhou and Dean 1989; Webby et al. 2000; Fortey and
Cocks 2003). Terranes where trilobite faunas show
a strong affinity with those of Siberia and Laurentia
include the Altay Terrane of the Northern Xinjiang
Region (part of the mobile zone between the Siberia
and Tarim plates) (region | in Fig. 1) and the Ergen-

Hinggan Terrane of the Hinggan Region (mobile zone
between the Siberia and North China plates) (region 2
in Fig. 1). The Late Ordovician record of Calyptaulax
and Jsotelus in the Altay Terrane (respectively
described by Zhang (1981) as Calliops taimyricus
Balashova and Fuyunia junggarensis Hsiang and
Zhang; see Zhou et al. 1996a) supports this affinity.
Occurrences of Mid and Late Ordovician forms such
as Parasphaerexochus, Pliomerellus, Quinquecosta
(see Zhao et al. 1997), Cybelurus (see Xiang and
Mao 1986), Eorobergia (see Xiang and Mao 1986, as
?Kainella sp.), Cybeloides, Calyptaulax, Isotelus (see
Nan 1985), and the monorakids Ceratoevenkaspis
and Isalauxina (Zhou Zhiyi unpublished collection)
in the Ergen-Hinggan Terrane also indicate strong
links with Siberia and Laurentia.

In this paper, biogeographic links among the
Chinese plates and terranes of northeastern Peri-
Gondwana are further reviewed on the basis of a
complete dataset available for the Ordovician trilobite
record in China (Zhou and Zhen, in press). As
indicated by Fortey (1975), Fortey and Owens (1978),
Zhou et al. (1989, 1990, 1992) and Fortey and Cocks
(2003), trilobites usually show decreased endemicity

ORDOVICIAN BIOGEOGRAPHY OF CHINA

soi a ime i

cae
GIOIA,
¢

pal

South China Sea

i

&

i

Figure 1. Map showing Ordovician geographic units of China (after Zhou et al. 2007). 1. Northern Xin-
jiang Region (part of the mobile zone between the Siberia and Tarim plates). 2. Hinggan Region (mobile
zone between the Siberia and North China plates). 3. Middle Tianshan-Beishan Region (part of the
Kazakhstan Mid Plate). 4. Tarim Region: 4-1, Bachu-Kalpin Area (the Tarim Block proper, developed
with platform facies (Tremadocian-Dapingian or Tremadoc-late Arenig) and shallow outer shelf facies
(early Darriwilian-early Katian or latest Arenig-Caradoc); its northern boundary employed here is
suitable for the Tremadocian-Dapingian only (see Zhou et al. 1990, 1992); and 4-2, Southern Tianshan
Area (deep-outer-shelf basin or trough fringing the northern margin of the Tarim Block) (Zhou et al.
1995a, 1996a; Ni et al. 2001). 5. North China Region: 5-1, Yellow River Area (North China Platform);
5-2, Ordos Area (western marginal area of the platform during the early Darriwilian-late Katian or lat-
est Arenig-early Ashgill, see Zhou et al. 1989); 5-3, Dunhuang-Alexa Area (a dislocated old land lacking
Early Palaeozoic deposits, possibly derived from the northern margin of the North China Platform, see
Zhou et al. 1996a); and 5-4, Qaidam-Qilian Area (consisting of the Qaidam and Middle Qilian terranes
and the Altun faulted block, see Zhou et al. 1996a). 6. Kunlun-Qinling Region (polycyclic orogenic belt
that crosses over the mainland of China). 7. Northern Qiangtang-Simao Region (a northern extension of
the Indochina Terrane, see Zhou et al. 1998a, 2001a). 8. South China Region: 8-1, Yangtze Area (inner
shelf/platform); 8-2, Jiangnan Area (outer-shelf slope; note its northern boundary varied in the differ-
ent time intervals due to facies shifting, and that employed herein is suitable for the Tremadocian only,
see MK Zhou et al. 1993, fig. 4-1); and 8-3, Cathaysia Area (off-shelf basin) (MK Zhou et al. 1993) (or
Pearl River Area, see Lu et al. 1976). 9. Baoshan-northern Tibet Region (a northern extension of the
Sibumasu Terrane, see Zhou et al. 1998a, 20014). 10. Southern Tibet Region (part of the India Plate, see
Fan et al. 1994). 11. Hainan Region (area subdivision see Wang 1989): 11-1, Wuzhishan Area (deduced
as a mobile zone that edged the shelf); and 11-2, Sanya Area (shelf).

from the onshore towards offshore biofacies belts
in a plate or terrane, and those elements which are
endemic to a particular plate or plates are the most
important indicators in defining biogeographic units.

Therefore, in the following discussion the platform/
inner shelf faunas (especially endemic trilobites)
are emphasized, although, as indicated by Zhou and
Dean (1989), a few endemics restricted to the deep-

184 Proc. Linn. Soc. N.S.W., 129, 2008

ZHOU ZHI-YI AND ZHEN YONG-YI

water facies are also significant enough for defining
faunal provinces.

EASTERN PERI-GONDWANAN TERRANES

Ordovician trilobites are only sporadically
documented in a number of Peri-Gondwanan
terranes in China, where the framework of biofacies
differentiation either in space or in time is usually
poorly known. Therefore, their biogeographic
relationships to the related geographic units or plates
are only briefly discussed herein based on the rather
incomplete faunal data, chiefly those recorded from
shallow marine successions.

Middle Tianshan-Beishan Region/Terrane (part
of the Kazakhstan Mid Plate) (region 3 in Fig. 1):
The only inner-shelf trilobite fauna of the Sandbian
to early Katian (Caradoc) age was described from
Ejin Banner, western Inner Mongolia (Zhou and
Zhou 2006). The dominant forms in this fauna, such
as Bulbaspis, Collis, Pliomerina, Sinocybele and
Basilicus (Basiliella), are also extremely diverse in the
coeval fauna of Kazakhstan, especially of the Chu-Ili
Terrane (see Fortey and Cocks 2003), and they even
share a number of taxa at species level. The region
also has close biogeographic links to the North China
Plate, where coetaneous shallow-water associations
characterized by Pliomerina and Basilicus (s./.) were
recorded from the Altun Mountains, eastern Xinjiang
(5-4) (Zhou et al. 1995b) and southern Ningxia and
central Shaanxi (5-1) (Zhou et al. 1982). However,
a deeper-water trilobite fauna of Sandbian-early
Katian age reported from northernmost Tarim (Zhang
1981) includes Bulbaspis, Sinocybele and Basilicus
(Basiliella), and also exhibits a faunal affinity with
this Kazakhstan fauna. Genera in common with
the inshore sites of eastern Australia (Webby 1971;
Edgecombe etal. 1999a, 2004; Edgecombe and Webby
2006), such as Pliomerina, Sinocybele and Basilicus
(Basiliella), are indicative of a biogeographically
significant faunal province (Webby 1987; Webby et
al. 2000).

Hainan Region/Terrane (region 11 in Fig. 1):
Only a few outer-shelf-facies trilobites were reported
from the Late Ordovician in the Sanya area (Zeng
et al. 1992), of which the genera of a Sandbian to
early Katian age including Ampyxinella, Birmanites,
Dionide and Bulbaspis were mostly widespread
forms. As deep-water faunas of the Tarim and South
and North China plates had already become uniform in
composition during the Sandbian to early Katian (Fig.
2D), their biogeographic affinities with the regions in

Proc. Linn. Soc. N.S.W., 129, 2008

question are difficult to affirm, but Bu/baspis alone
may suggest a faunal link with the Middle Tianshan-
Beishan Terrane and even with the Tarim Plate.

Southern Tibet Region (part of the India Plate)
(region 10 in Fig. 1): Only a few trilobites of the Floian
to early Darriwilian (Arenig) age are known, including
Basilicus (Basilicus) (as Isoteloides bolingensis),
Hystricurus and Pliomerina (as Negaricephalus)
from Ngari, southwestern Tibet (Yang 1990) and
Pseudocalymene (as Eucalymene tuberculata) from
Nyalam of Mount Qomolangma (Chien 1976). All
the trilobite taxa recorded in this region also occur in
the coeval faunas of the North China Plate, where the
genus Pseudocalymene is associated, as at Nyalam,
with the nautiloid Pomphoceras, which is an element
typical of the Yellow River Fauna (Chen et al. 1984).

Baoshan-northern Tibet Region/Terrane
(a northern extension of the Sibumasu Terrane)
(region 9 in Fig. 1): Late Ordovician trilobites have
been described from the rocks of the Pagoda facies
(Lindstrém et al. 1991) in southern Thailand (Fortey
1997) and westernmost Yunnan (Sheng 1974),
representing a relatively widespread fauna of the deep-
water-biofacies, with some forms being identical even
at species level with taxa from South China (Fortey
and Cocks 1998). However, the mid-late Darriwilian
(Llanvirn) fauna from the imner-shelf-facies (Reed
1917; Sheng 1974) in the Baoshan area seems to
be a mixture of trilobites exhibiting two different
biogeographical affinities, with Basilicus (Basilicus)
and Pliomerina typical of the contemporaneous fauna
of North China, and with Hexacopyge, Neseuretus,
Prionicheilus, Reedocalymene, and Sinocybele found
in South China. Therefore, there remain ambiguities in
the explanations for the alignment of this Cimmerian
terrane. Fortey and Cocks (1998, 2003) favoured its
close biogeographic and physical proximity to South
China. However, Zhou et al. (1998a) preferred the
reconstructions proposed by Scotese and McKerrow
(1991) and Metcalfe (1992), in which the West
Malaysia-Thailand Peninsula was located close to the
North China Plate on the palaeo-equator, while Shan
State and the Baoshan-northern Tibet Region may
have been sited in a low latitudinal zone not far from
the South China Plate.

Northern Qiangtang-Simao Region (a northern
extension of the Indochina or Annamia Terrane)
(region 7 in Fig. 1): The Tremadocian-early
Floian shallow-water trilobites from Karakorum,
southwestern Xinjiang include Neopsilocephalina,
Psilocephalina, Psilocephalops and Songtaoia (as
Yinjiangia karakolumensis) described by Zhang
(1991), and Asaphopsoides (as Asaphus elegantulus;
see Jell and Stait 1985) by Gortani (1934). A

185

ORDOVICIAN BIOGEOGRAPHY OF CHINA

less diverse Dapingian-early Darriwilian (late
Arenig) inner-shelf fauna, comprising Hungiodes,
Liomegalaspides, Neseuretus and Ogyginus, was
recorded from Dali, western Yunnan (Zhou et al.
1998a), and Neseuretus and Aristocalymene (as
Neseuretus muliensis; see Turvey 2005b) were
described from Muli, southwestern Sichuan (Lee
1978). Most of the components are in common with the
Yangtze platform, demonstrating a very close faunal
link that existed between this Cimmerian terrane
and the South China Plate. However, occurrences of
Asaphopsoides, Neseuretus and especially Ogyginus
indicate a western Gondwanan and Peri-Gondwanan
faunal affinity, and suggest that the Annamia Terrane
may have been located at higher latitudes or in a more
westerly position as compared with the South China
Plate (Zhou et al. 1998a; Fortey and Cocks 2003).

EASTERN PERI-GONDWANAN PLATES

Ordovician trilobites are well recorded in
three cratonic plates: the South China Plate (region
8 in Fig. 1) [including also the Wudang or South
Qinling Terrane (see Yang 1988; Shang et al. 1994)
of the Kunlun-Qinling Region (region 6 in Fig. 1)
—a polycyclic orogenic belt that crosses over the
mainland of China], North China Plate (region 5
in Fig. 1) [including also West Kunlun, a displaced
landmass of the North China Plate proper (Zhou et
al. 1996a), now located on the western end of the
Kunlun-Qinling Region; the Qaidam and Middle
Qilian terranes and the Altun faulted block (area 5-4
in Fig. 1) (see Zhou et al. 1996a)] and Tarim Plate
(region 4 in Fig. 1). All of them were largely situated
in low latitude zones during the Ordovician (Cocks
and Torsvik 2002; Webby et al. 2000; Turvey 2005a).
In each plate, trilobite faunas display a progressive
onshore to offshore transition in composition and
diversity. Ecological differentiation of the faunas
and their response to biofacies has been documented
along environmental gradients from west to east
and from south to north (platform/inner shelf to off-
shelf basin) of South China (Zhou et al. 1999, 2000,
2001b, 2003, 2004; Z.Q. Zhou et al. 2000; Yuan et al.
2000; Turvey and Zhou 2002, 2004a, 2004b; Yin et
al. 2000; Turvey 2005a), from east to west (platform
to shallow-outer-shelf basin) of North China (Zhou
et al. 1989) and from south to north (platform to
deep-outer-shelf basin/trough) of Tarim (Zhou et al.
1990, 1992). Accordingly, Tremadocian-early Katian
trilobites of the inner shelf/platform and outer shelf/
slope from each of the plates are separately listed in
the Appendix, except for those of the Qaidam-Qilian

terranes (area 5-4 in Fig. 1), as the biofacies patterns
in these areas are not well established yet. They bear
mainly deeper-water forms, but also include a few
shallow-water genera.

Early studies indicated that, as a whole, the
respective trilobite faunas of the South China and
North China plates were closely comparable in
the Tremadocian, but became distinct from each
other afterwards (Zhou and Fortey 1986), while
those from the Tarim Plate mostly extended their
distribution into the South China Plate throughout
the Ordovician, suggesting that both were closely
situated palaeogeographically (Zhou and Chen 1990,
p. tv; 1992, p. 11; Zhou et al. 1996a, pp. 11, 20). This
view is supported by a further comparison among
the inner shelf/platform faunas of the three plates;
those trilobites with significant biogeographical
implications are further reviewed herein.

Tremadocian trilobite faunas of North China
are characterized by having distinctive endemics
— Koraipsis and Penchiopsis, in addition to a few
bathyurids, and those of South China by having
Dactylocephalus, Tungtzuella and Psilocephalina.
However, most of the other trilobites from both
plates are in common, such as Chosenia, Songtaoia,
Wanliangtingia and Yosimuraspis. \t is interesting
to note that some trilobites typical of either of the
two plates co-occur in the contemporaneous faunas
of Australia and northern Iran (the Alborz Terrane).
The Tremadocian trilobites described from the latter
area by Bruton et al. (2004) and Ghobadi Pour (2006)
include Asaphellus inflatus Lu, Psilocephalina lubrica
Hsu, Dactylocephalus, Kayseraspis, Peltabellia, and
Illaenus hinomotoensis Kobayashi. The occurrence
of the first three suggests a close link to South
China, and the others are common elements of the
North China faunas. In Australia, Shergold (1991)
reported Asaphellus cf. trinodosus Chang, Koraipsis,
Kayseraspis and Psilocephalina cf. lubrica Hsu from
the northern part of the Amadeus Basin, Jell and Stait
(1985) described Asaphellus cf. trinodosus Chang,
Asaphopsoides, Chosenia, Dikelokephalina asiatica
Kobayashi, and Aystricurus penchiensis Lu from
Tasmania, and Laurie and Shergold (1996) recorded
Penchiopsis from the Canning Basin. These trilobite
records show a remarkable similarity to those of
North China, but a close link with South China also
existed at either specific or generic rank, as indicated
by the occurrence of Psilocephalina cf. lubrica Hsu,
Asaphopsoides and Chosenia.

From the Floianto early Katian(Arenig-Caradoc),
trilobite faunas of North China with endemic forms
like Hoisotelus, Lonchobasilicus and Pliomerina, 1m
addition to variety of species of Basilicus (s./.), are

Proc. Linn. Soc. N.S.W., 129, 2008

ZHOU ZHI-YI AND ZHEN YONG-YI

quite distinct from those of South China, which are
characterized by having Birmanites, Calymenesun,
Fenghuangchengia, Hexacopyge, Liomegalaspides,
Meitanopsis, Mioptychopyge, Ningkianites, Omeipsis
and Taihungshania. The faunal differences may imply
a significant biogeographical separation between
these two plates. Among the Floian-Darriwilian
(Arenig-Llanvirn) trilobites, Prosopiscus was widely
distributed im Australia, South and North China,
Sibumasu, Himalayas and Argentine Precordillera,
being of important biogeographic significance (Webby
et al. 2000; Paterson 2004). As indicated by Paterson
(2004), the genus originated in the early Floian of
Australia, but its early species, P. /auriei Paterson
from northwestern New South Wales, is closely
related to P magicus Zhou from the North China Plate.
Webby (1971, 1974, 1985, 1987, 1992) and Webby et
al. (2000) suggested a biogeographically distinctive
Eokosovopeltis-Pliomerina Province in the Caradoc
(Sandbian to early Katian) to cover part of the eastern
Peri-Gondwanan regions, including East Australia,
East Asia and Kazakhstan, and probably the Argentine
Precordillera (Edgecombe et al. 1999b). However,
the eponymous forms Eokosovopeltis and Pliomerina
recorded from New South Wales (Webby 1971,
1974; Edgecombe and Webby 2007) and Tasmania
(Corbett and Banks 1974; Edgecombe et al. 1999,
2004) were lacking in all the geographic units of
East Asia during the Sandbian to early Katian, except
the North China Plate, where, as mentioned above,
Pliomerina occurs though with no Eokosovopeltis
associated. A few deeper water taxa with more
restricted distribution including the distinctive three-
segmented raphiophorid genus Nanshanaspis are also
in common between the Qaidam-Qilian Area (5-4 in
Fig. 1) (Chang and Fan 1960; Zhou et al. 1995b) and
southern Tasmania (Burrett et al. 1983), suggestive
again of strong biogeographic relationships between
Australia and North China. Though as noted by Webby
et al. (2000), faunal links based on the available biotic
data from Australia were in general with the Chinese
plates, it is more likely that Australian trilobite faunas
had closest affinities with those of North China during
the Floian-early Katian.

In Tarim, only a bathyurid Aksuaspis is recorded
in the Tremadocian dolomite at Kalpin (Zhou et
al. 1998c), but the occurrence of a few trilobites,
including Asaphopsoides, Paraszechuanella
[=Pseudocalymene, see Bruton et al. 2004], and
Psilocephalina described by Zhou (see Lin et al.
1990) from a shallow outer shelf Nileid Biofacies at
Uligezhitag (Zhou et al. 1990, 1992) suggests a close
faunal link with South China. From the Floian onwards,
the shallow-water trilobites were all associated with

Proc. Linn. Soc. N.S.W., 129, 2008

carbonate buildups, and have been proved to be of
worldwide distribution, except for Liomegalaspides,
a Floian-early Darriwilian form typical of the coeval
fauna in South China, which was described from
the platform facies of Tarim (Zhou et al. 1998c, as
Megalaspides angustus and M. sp.). However, there
was also developed a unique provincial link with the
South China Plate, as evidenced by the occurrence of
a number of common shallow outer shelf elements,
of which characteristic forms are Calymenesun,
Mioptychopyge, Reedocalymene, _—_Xiushuilithus,
Yanhaoia and Zhenganites (see Zhang 1981; Zhou et
al. 1990, 1992, 1998b; Yuan and Zhou 1997).

The late Katian (early-mid Ashgill) inner shelf/
platform trilobite faunas are almost lacking in China,
with the exception of a few forms that were reported
from the latest Katian (mid-Ashgill) carbonate
buildups in the eastern margin of the Jiangnan Area
(Zhou et al. 2004). Occurrence of the Nankinolithus
fauna from the deeper sites (areas 4-2, 5-2, and 8-
2, see Fig. 1) may, however, suggest that the faunal
connection between the South China-Tarim and
North China plates became closer again during this
time interval. Comparatively uniform patterns of
provincialism might continue to exist during the
Hirnantian, when there were only a few trilobites
(largely immigrants from high-latitude Gondwana)
occurring in the Chinese plates, comprising
Dalmanitina (Songxites) (in areas 4-1, 5-4 and region
8, see Fig. 1), Eoleonaspis (in area 5-4 and region 8),
Platycoryphe (in area 5-4 and region 8), Niuchangella
(in area 4-1 and region 8) and So/ariproetus (in area
5-4).

In order to express the preliminary observations
more clearly, cluster analysis of biogeographic links
on the basis of trilobite genera (see Appendix) from
four time intervals respectively (Tremadocian, Floian-
early Darriwilian, mid-late Darriwilian and Sandbian-
early Katian) was conducted using Simpson’s
coefficient (Fig. 2). It reveals that the Chinese plates
belonged to a single biogeographic unit during the
Tremadocian, and that their platform/imner-shelf
and outer-shelf/slope facies areas can be separated
into two distinct clusters (Fig. 2A), each of which
shares a closely similar trilobite fauna. The Floian-
early Darriwilian clusters (Fig. 2B) suggest that the
South China-Tarim and North China plates may
well be referred to two independent biogeographic
units, as evidenced by different faunas distributed
either in shallow or in deep sites. A similar pattern
is also depicted by the mid-late Darriwilian (Fig.
2C) and Sandbian-early Katian (Fig. 2D) faunas,
but deep-water facies trilobites of the North China
Plate progressively become more analogous to those

187

ORDOVICIAN BIOGEOGRAPHY OF CHINA

8-1
5
4

Similarity

Tremadocian

Similarity
oO

mid-late Darriwilian (Lanvirn)

8-2

Similarity
oO
a

Floian-early Darriwilian (Arenig)

Similarity
oO
a

Sandbian-early Katian (Caradoc)

Figure 2. Clusters of Ordovician geographic units on the basis of trilobite faunas (dataset see Ap-
pendix) using Simpson’s coefficient, indicating the biogeographic affinities of Ordovician trilobites
occurring in the shallow-water (4-1, 5-1, 8-1) and deep-water facies (4-2, 5-2, 5-4, 8-2) belts of Tarim
(4), North China (5) and South China (8) plates (Fig. 1). A. Tremadocian; B. Floian-early Darriwilian
(Arenig); C. mid-late Darriwilian (Llanvirn); D. Sandbian-early Katian (Caradoc). Note that only few
Tremadocian and Floian-early Darriwilian platform trilobites were recorded from 4-1, which are not
coded, and, as mentioned in the text, the coded trilobites from 5-4 are mainly deeper-water forms, but

also mixed up with a few from shallow sites.

of South China-Tarim. This suggests that a faunal
exchange between offshore sites of both geographic
units may have started long before the late Katian
when the Nankinolithus fauna testifies to a shared
biogeographic link amongst the Chinese plates.

CONCLUSIONS

Trilobite evidence indicates that all the plates
and most of the terranes in China exhibit a close
biogeographic link and may have formed part of
eastern Peri-Gondwana during the Ordovician, except

Proc. Linn. Soc. N.S.W., 129, 2008

ZHOU ZHI-YI AND ZHEN YONG-YI

for the Altay Terrane of the Northern Xinjiang Region
and the Ergen-Hinggan Terrane of the Hinggan
Region, where trilobite faunas show a strong affinity
with those of Siberia and Laurentia.

Well-defined biogeographic patterns are depicted
mainly by the shallow-water components of the
Ordovician trilobites, especially between the South
China-Tarim and North China plates. Synthetic
analysis suggests that all the Chinese eastern Peri-
Gondwanan plates and terranes may be signified as
belonging to a single biogeographic province during
the Tremadocian and late Katian-Hirmantian, but
exhibit significant faunal differences and therefore
may be separated into two sub-provinces during the
Floian-early Katian: one consists of South China,
Tarim and Annamia, and the other may include North
China, Sibumasu, Southern Tibet, Tianshan-Beishan
and possibly Hainan. However, the deep-water facies
trilobites of the relevant Chinese geographic units
had progressively become more unified from the mid
Darriwilian to early Katian before the sub-provinces
eventually broke down by the late Katian.

Australian Ordovician trilobite faunas had
close affinities with most of the Chinese eastern
Peri-Gondwanan plates and terranes, but closest
biogeographic links were in particular with North
China and Middle Tianshan-Beishan.

ACKNOWLEDGEMENTS

This paper is a contribution to the IGCP project
503 “Ordovician Palaeogeography and Palaeoclimate’.
The research was supported by the Chinese Academy of
Sciences (KZCX3-SW-149), the Ministry of Science and
Technology of China (2006CB806402), and the National
Natural Science Foundation of China (No 40532009). We
are grateful to Zhou Zhigiang for helpful discussions, and to
Ian Percival for his constructive review of the manuscript.

REFERENCES

Bruton, D. L., Wright, A. J. and Hamedi, M. I. (2004).
Ordovician trilobites from Iran. Palaeontographica
Abteilung A 271, 111-149.

Burrett, C., Stait, B. and Laurie, J. (1983). Trilobites and
microfossils from the Middle Ordovician of Surprise
Bay, southern Tasmania, Australia. Memoirs of the
Association of Australasian Palaeontologists 1,
177-193.

Chang, W. T. and Fan J. S. (1960). Ordovician and Silurian
trilobites of the Chilian Mountains. 83-148. Jn
Geological Gazetter of the Chilian Mountains 4 (1),
Science Press, Beijing, 160pp. (in Chinese).

Proc. Linn. Soc. N.S.W., 129, 2008

Chen, J. Y., Zhou, Z. Y., Lin, Y. K., Yang, X. C., Zou, X.
P., Wang, Z. H., Luo, K. Q., Yao, B. Q. and Shen,

H. (1984). Ordovician biostratigraphy of western
Ordos. Memoirs of Nanjing Institute of Geology and
Palaeontology, Academia Sinica 20, 1-31 (in Chinese
with English abstract).

Chien, Y. Y. (1976). Two early Ordovician trilobite species
from Mount Jalmo Lungma Region. 137-138. nA
report of scientific investigations in the Mount Jalmo
Lungma Region (1966-1968): Palaeontology (2).
Science Press, Bejing, 474pp. (in Chinese).

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.

Corbett, K. D. and Banks, M. R. (1974). Ordovician
stratigraphy of the Florentine Synclinortum,
Southwest Tasmania. Papers and Proceedings of the
Royal Society of Tasmania 107, 207-238.

Edgecombe, G. D. and Webby, B. D. (2006). The
Ordovician encrinurid trilobite Sinocybele from New
South Wales and its biogeographic significance.
Memoirs of the Association of Australasian
Palaeontologists 32, 413-422.

Edgecombe, G. D. and Webby, B. D. (2007). Ordovician
trilobite with eastern Gondwanan affinities from
centra-west New South Wales and Tasmania.
Memoirs of the Association of Australasian
Palaeontologists 34, 255-281.

Edgecombe, G. D., Banks, M. R. and Banks, D. M.
(1999a). Upper Ordovician Phacopida (Trilobita)
from Tasmania. Alcheringa 23, 235-257.

Edgecombe, G. D., Chatterton, B. D. E., Waisfeld, B. G.
and Vaccari, N. E. (1999b). Ordovician pliomerid
and prosopiscid trilobites from Argentina. Journal of
Paleontology 73, 1144-1154.

Edgecombe, G. D., Banks, M. R. and Banks, D. M.
(2004). Late Ordovician trilobites from Tasmania:
Styginidae, Asaphidae and Lichidae. Memoirs of the
Association of Australasian Palaeontologists 30,
59-77.

Fan, C. J., Ma, G. Q. and Wang, Z. S. (1994). Geological
features of the Sichuan- Yunnan-Qinghai-Tibet
Region. 239-312. In Cheng, Y. Q. (ed.), An
introduction to the regional geology of China.
Geological Publishing House, Bejing. 517pp. (in
Chinese).

Fortey, R. A. (1975). Early Ordovician trilobite
communities. Fossils and Strata 4, 339-360.

Fortey, R. A. (1997). Late Ordovician trilobites from
southern Thailand. Palaeontology 40, 397-449.

Fortey, R. A. and Cocks, L. R. M. (1998). Biogeography
and palaeogeography of the Sibumasu terrane in the
Ordovician: a review. /n Hall, R. and Holloway, J. D.
(eds), Biogeography and Geological Evolution of SE
Asia. Backhuys Publishers, Amsterdam. 43-56.

Fortey, R. A. and Cocks, L. R. M. (2003). Palaeontological
evidence bearing on global Ordovician-Silurian
continental reconstructions. Earth-Science Reviews
61, 245-307.

189

ORDOVICIAN BIOGEOGRAPHY OF CHINA

Fortey, R. A. and Owens, R. M. (1978). Early Ordovician
(Arenig) stratigraphy and faunas of the Carmarthern
district, southwest Wales. Bulletin of the British
Museum (Natural History), Geology Series 30,
225-294.

Ghobadi Pour, M. (2006). Early Ordovician (Tremadocian)
trilobites from Simeh-Kuh, eastern Alborz, Iran. In
Bassett, M. G. and Deisler, V. K. (eds), Studies in
Palaeozoic Palaeontology. National Museum of Wales
Geological Series 25, 93-118.

Gortani, M. (1934). Fossili Ordoviciani del Caracorum.
Spedizione Italiana del Filippi nell’ Himalaia,
Caracorum e Turechestan Cinne (1913-14). Series 2,
Vol. 5, 1-97.

Jell, P. A. and Stait, B. (1985). Tremadoc trilobites from
the Florentine Valley Formation, Tim Shea area,
Tasmania. Memoirs of the Museum of Victoria 46,
1-34.

Laurie, J. R. and Shergold, J. H. (1996). Early Ordovician
trilobite taxonomy and biostratigraphy of the
Emanuel Formation, Canning Basin, Western
Australia. Part 2. Palaeontographica Abteilung A
240, 105-144.

Lee, S. J. (1978). Trilobita. 179-284. In Palaeontological
atlas of southwest China, Sichuan volume, part 1,
Sinian to Devonian. Geological Publishing House,
Beijing, 617pp. (in Chinese).

Lin, H. L., Zhou, Z. Y. and Luo, H. L. (1990). Trilobita,
109-122. Jn Sinian to Permian stratigraphy and
palaeontology of the Tarim Basin, Xinjiang (1):
Kuruktag Region. Nanjing University Press, Nanjing,
252pp. (in Chinese with English summary).

Lindstrém, M., Chen, J. Y. and Zhang, J. M. (1991).
Section at Daping reveals Sino-Baltoscandian
parallelism of facies in the Ordovician. Geologiska
Féreningens i Stockholm Férhandlingar 113 (2-3),
189-205.

Lu, Y. H., Zhu, Z. L., Qian, Y. Y., Zhou, Z. Y., Chen, J. Y.,
Liu, G. W., Yu, W., Chen, X. and Xu, H. K. (1976).
Ordovician biostratigraphy and palaeozoogeography
of China. Memoirs of the Nanjing Institute of
Geology and Palaeontology, Academia Sinica 7,

1—83 (in Chinese).

Metcalfe, I. (1992). Ordovician to Permian evolution of
Southeast Asian terranes: NW Australian Gondwana
connections. Jn Webby, B. D. and Laurie, J. R. (eds),
Global Perspectives on Ordovician Geology. pp.
293-305. A. A. Balkema, Rotterdam.

Nan, R. S. (1985). Upper Ordovician trilobites from the
Wulongtun Formation of eastern Yilehuli Shan,
Heilongjiang Province. Bulletin of the Shenyang
Institute of Geology and Mineral Resources, Chinese
Academy of Geological Sciences 12, 56—67 (in
Chinese with English abstract).

Ni, Y. N., Geng, L. Y., Wang, Z. H., Zhao, Z. X., Chen, T.
N., Zhang, Y. B., Wang, H. F., Zhang, S. G., Yuan,
W. W., Zhang, S. B., Gao, Q. Q. and Li, J. (2001).
Ordovician. 39-80, 343-344. In Zhou, Z. Y. (ed.),
Stratigraphy of the Tarim Basin. Science Press,
Beijing, 358pp. (in Chinese with English summary).

Paterson, J. (2004). Palaeogeography of the Ordovician
trilobite Prosopiscus, with a new species from
western New South Wales. A/cheringa 28, 65—76.

Reed, F. R. C. (1917). Ordovician and Silurian fossils from
Yun-nan. Palaeontologica Indica, N.S. 6, 1-84.

Scotese, C. R. and McKerrow, W. S. (1991). Ordovician
plate tectonic reconstructions. /n Barnes, C. R.
and Williams, S. H. (eds), Advances in Ordovician
Geology. Paper of the Geological Survey of Canada
90-9, 271-282.

Shang, R. J., Chen, J. Y., Lao, Z. Q. and Wu, X. N. (1994).
Geological features of the Kunlun-Qinling Region.
165-238. In Cheng, Y. Q. (ed.), An introduction to
the regional geology of China. Geological Publishing
House, Beijing. 517pp. (in Chinese).

Sheng, X. F. (1974). Ordovician trilobites from western
Yunnan and its stratigraphical significance. 96-140.
In Subdivision and correlation of the Ordovician
System in China. Geological Publishing House,
Beijing, 153pp. (in Chinese).

Shergold, J. H. (1991). Late Cambrian and Early
Ordovician trilobite faunas of the Pacoota Sandstone,
Amadeus Basin, Central Australia. Bureau of Mineral
Resources, Geology and Geophysics, Bulletin 237,
15-75.

Turvey, S. T. (2005a). Early Ordovician (Arenig) trilobite
palaeontology and palaeobiogeography of the South
China Plate. Palaeontology 48 (3), 549-575.

Turvey, S. T. (2005b). Reedocalymenine trilobites from the
Ordovician of central and eastern Asia, and a review
of species assigned to Neseuretus. Palaeontology 48
(3), 519-547.

Turvey, S. T. and Zhou, Z. Y. (2002). Arenig trilobite
associations of Daping, Yichang, Hubei, South China.
Acta Palaeontologica Sinica 41, 10-18.

Turvey, S. T. and Zhou, Z. Y. (2004a). Arenig trilobite
associations from the Jiangnan Transitional Belt of
northwestern Hunan, China. Journal of Asian Earth
Sciences 23, 47-61.

Turvey, S. T. and Zhou, Z. Y. (2004b). Arenig trilobite
associations and faunal changes in Southern Shaanxi,
China. Journal of Asian Earth Sciences 23, 91-103.

Wang, X. F. (1989). Palaeogeographic reconstruction
of Ordovician in China and characteristics of its
sedimentary environment and biofacies. Acta
Palaeontologica Sinica 28 (2), 234-248 (in Chinese
with English summary).

Webby, B. D. (1971). The trilobite Pliomerina Chugaeva
from the Ordovician of New South Wales.
Palaeontology 14, 612-622.

Webby, B. D. (1974). Upper Ordovician trilobites from
central New South Wales. Palaeontology 17, 203—
US,

Webby, B. D. (1985). Influence of a Tasmanide Island-
Arc on the evolutionary development of Ordovician
faunas. New Zealand Geological Survey Record 9,

99-101.

Webby, B. D. (1987). Biogeographic significance of
some East Australian Ordovician faunas. 103-117.
In Leitch, E. C. and Scheibner, E. (eds), Terrane

Proc. Linn. Soc. N.S.W., 129, 2008

ZHOU ZHI-YI AND ZHEN YONG-YI

accretion and orogenic belts. American Geophysical
Union, Monograph Series 19.Washington.

Webby, B. D. (1992). Ordovician Island-Arc Biotas.
Journal and Proceedings of the Royal Society of New
South Wales 124, 51-77.

Webby, B. D., Percival, I. G., Edgecombe, G. D.,
Cooper, R. A., VandenBerg, A. H. M., Pickett, J.
W., Pojeta Jr, J., Playford, G., Winchester-Seeto,

T., Young, G. C ., Zhen, Y. Y., Nicoll, R. S., Ross,
J. R. P. and Schallreuter, R. (2000). Ordovician
palaeobiogeography of Australia. Jn Wright, A. J.,
Young, G. C., Talent, J. A. and Laurie, J. R. (eds),
Palaeobiogeography of Australasian faunas and
floras. Memoirs of the Association of Australasian
Palaeontologists 23, 63-126.

Xiang, L. W. and Mao, Y. H. (1986). Some Ordovician
trilobites from Sunid Zuogi, Inner Mongolia.
Professional Papers of Stratigraphy and
Palaeontology 14, 125—130 (an Chinese with English
abstract).

Yang, J. L. (1988). A survey of Cambrian
paleotectonogeography in East Qinling. Earth
Science Journal of China University of Geosciences
13 (5), 473-480 (in Chinese with English abstract).

Yang, J. L. (1990). Ordovician trilobites from Negari,
Xizang (Tibet). 23-32, 269-272. In Yang, Z. Y. and
Nie, Z. T. (eds), Palaeontology of Ngari, Xizang
(Tibet). China University of Geosciences Press,
Wuhan (in Chinese with English summary).

Yin, G. Z., Tripp, R. P., Zhou, Z. Y., Zhou, Z. Q. and
Yuan, W. W. (2000). Trilobites and biofacies of the
Ordovician Pagoda Formation, Donggongsi of Zunyi,
Guizhou Province, China. Zransactions of the Royal
Society of Edinburgh, Earth Sciences 90, 203-220.

Yuan, W. W. and Zhou, Z. Y. (1997). Some Ordovician
trilobites from the northern Tarim Basin, Xinjiang.
Acta Palaeontologica Sinica 36 (Supplement), 168—
181 (in Chinese with English summary).

Yuan, W. W., Zhou, Z. Y., Zhang, J. M., Zhou, Z. Q.,
Sun, X. W. and Zhou, T. M. (2000). Tremadocian
trilobite biofacies in western Hunan-Hubei. Journal
of Stratigraphy 24 (4), 275—282 (in Chinese with
English abstract).

Zeng, Q. L., Wu, W., Lin, J. M., Cai, D. G. and Wu, T.

F. (1992). Study on foundational geology in Sanya
area, Hainan Island, China. China University of
Geosciences Press, Wuhan, 174pp. (in Chinese with
English sammary).

Zhang, T. R. (1981). Trilobita. 134-213. In
Palaeontological atlas of Northwest China. Xinjiang
(1). Geological Publishing House, Beijing (in
Chinese).

Zhang, T. R. (1991). Some Ordovician trilobites from
Karakorum Mountain of Xijiang. Xijiang Geology 9
(1), 31-39 Gn Chinese with English abstract).

Zhao, D., Zhang, M. S., Cheng, L. R. and Zhu, H. S.
(1997). Ordovician strata, trilobite fauna and its
tectonic setting of Hinngan region, China. Jilin
People’s Press, Changchun, 181pp. (in Chinese with
English abstract).

Proc. Linn. Soc. N.S.W., 129, 2008

Zhou, M. K., Wang, R. Z., Li, Z. M. et al. (1993).
Ordovician and Silurian lithofacies, paleogeography
and mineralization in South China. Geological
Publishing House, Beijing, 111pp. (in Chinese with
English abstract).

Zhou, Z. Q. and Zhou, Z. Y. (2006). Late Ordovician
trilobites from the Zhusilenghaierhan area, Ejin
Banner, western Inner Mongolia, China. Memoirs of
the Association of Australasian Palaeontologists 32,
383-411.

Zhou, Z. Q., Li, J. S. and Qu, X. G. (1982). Trilobita. 215—
294. In Palaeontological Atlas of Northwest China,
Shaanxi-Gansu-Ningxia Volume (1): Precambrian-
Lower Palaeozoic. Geological Publishing House,
Beijing (in Chinese).

Zhou, Z. Q., Zhou, Z. Y. and Yuan, W. W. (2000). Middle
Caradoc trilobite biofacies of the Micangshan Area,
northwestern margin of the Yangtze Block. Journal
of Stratigraphy 24 (4), 264-274 (in Chinese with
English abstract).

Zhou, Z. Y. and Chen, P. J. (1990). Preface. In Zhou, Z. Y.
and Chen, P. J. (eds), Biostratigraphy and geological
evolution of Tarim. Science Press, Beijing. i1i—v (in
Chinese).

Zhou, Z. Y. and Chen, P. J. (1992). Preface. Jn Zhou, Z.
Y., Chen, P. J. (eds), Biostratigraphy and geological
evolution of Tarim. Science Press, Beijing, i—111.

Zhou, Z. Y. and Dean, W. T. (1989). Trilobite evidence for
Gondwanaland in East Asia during the Ordovician.
Journal of Southeast Asian Sciences 3, 131—140.

Zhou, Z. Y. and Fortey, R. A. (1986). Ordovician trilobites
from North and Northeast China. Palaeontographica
Abteilung A 192, 157-210.

Zhou, Z. Y. and Zhen, Y. Y. (eds), in press. Trilobite record
of China. Science Press, Beijing.

Zhou, Z. Y., Chen, X., Wang, Z. H., Wang, Z. Z., Li, J.,
Geng, L. Y., Fang, Z. J., Qiao, X. D. and Zhang, T. R.
(1990). Ordovician of Tarim. 56-130. In Zhou, Z. Y.
and Chen, P. J. (eds), Biostratigraphy and geological
evolution of Tarim. Science Press, Beijing, 366pp. (in
Chinese).

Zhou, Z. Y., Chen, X., Wang, Z. H., Wang, Z. Z., Li, J.,
Geng, L. Y., Fang, Z. J., Qiao, X. D. and Zhang, T. R.
(1992). Ordovician of Tarim. 62-139. Jn Zhou, Z. Y.
and Chen, P. J. (eds), Biostratigraphy and geological
evolution of Tarim. Science Press, Beijing, 399pp.

Zhou, Z. Y., Dean, W. T. and Luo, H. L. (1998a). Early
Ordovician trilobites from Dali, west Yunnan,

China, and their palaeogeographical significance.
Palaeontology 41 (3), 429-460.

Zhou, Z. Y., Dean, W. T., Yuan, W. W. and Zhou, T. R.
(1998b). Ordovician trilobites from the Dawangou
Formation, Kalpin, Xinjiang, north-west China.
Palaeontology 41 (4), 693-735.

Zhou, Z. Y., Zhou, T. R. and Yuan, W. W. (1998c).
Ordovician trilobites from the Upper Qiulitag Group,
western Tarim, Xinjiang, Northwest China. Acta
Palaeontologica Sinica 37 (3), 269-282.

Zhou, Z. Y., Luo, H. L., Zhou, Z. Q. and Yuan, W. W.
(2001a). Palaeontological constraints on the extent

19]

ORDOVICIAN BIOGEOGRAPHY OF CHINA

of the Ordovician Indo-China Terrane in western
Yunnan. Acta Palaeontologica Sinica 40 (3), 310—
317 (in Chinese with English summary).

Zhou, Z. Y., Zhou, Z. Q. and Yuan, W. W. (2001b).
Llanvirn-early Caradoc trilobite biofacies of western
Hubei and Hunan, China. A/cheringa 25, 69-86.

Zhou, Z. Y., Lin, H. L. and Ni, Y. N. (1996a). Early
Palaeozoic plate tectonics and geological evolution.
3-21. In Zhou, Z. Y. and Dean, W. T. (eds), A
Series of Solid Earth Sciences Research in China:
Phanerozoic Geology of Northwest China. Science
Press, Beijing, 316pp.

Zhou, Z. Y., Ni, Y. N., Lin, H. L., Zhou, Z. Q. and Yu, F.
(1996b). Palaeogeographic development during the
Ordovician. 71-82. In Zhou, Z. Y. and Dean, W. T.
(eds), A Series of Solid Earth Sciences Research in
China: Phanerozoic Geology of Northwest China.
Science Press, Beijing, 316pp.

Zhou, Z. Y., Ni, Y. N., Lin, H. L., Zhou, Z. Q. and Yu,

F. (1996c). Ordovician. 149-169. Jn Zhou, Z. Y.

and Dean, W. T. (eds), A Series of Solid Earth
Sciences Research in China: Phanerozoic Geology of
Northwest China. Science Press, Beijing, 316pp.

Zhou, Z. Y., Yuan, W. W., Han, N. R. and Zhou, Z. Q.
(2004). Trilobite faunas across the Late Ordovician
mass extinction event in the Yangtze Block.
127-152, 1042. In Rong, J. Y. and Fang, Z. J. (eds),
Mass extinction and recovery — evidences from
the Palaeozoic and Triassic of South China. China
University of Science and Technology Press, Hefei,
Vol. 1, 1-472; Vol. 2, 473-1087 (in Chinese with
English abstract).

Zhou, Z. Y., Ni, Y. N. and Yuan, W. W. (1995a). Outline
of Ordovician Palaeogeography, Tarim, Northwest
China. /n Cooper, J. D., Droser, M. L. and Finney, S.
C. (eds), Ordovician Odyssey (Short papers for the
Seventh International Symposium on the Ordovician
System). Pacific Section SEPM Book 77, 207-210.
Fullerton, California.

Zhou, Z. Y., Zhang, T. R., Yuan, W. W. and Yuan, J. L.
(1995b). Trilobita. 137-143. Jn Wang, Q. M. (ed.),
Sinian to Permian stratigraphy and palaeontology
of the Tarim Basin, Xinjiang (IV): Altun Mountains
Region. The Petroleum Industry Press, Beijing.
285pp. (in Chinese with English summary).

Zhou, Z. Y., Zhen, Y. Y., Zhou, Z. Q. and Yuan, W. W.
(2007). A new approach to the division of Ordovician
geographic units of China. Acta Palaeontologica
Sinica 46 (Supplement), 558-563.

Zhou, Z. Y., Zhou, Z. Q. and Yuan, W. W. (1999). Middle
Caradoc trilobite biofacies of western Hubei and
Hunan, South China. Acta Universitatis Caroline-
Geologica 43 (1/2), 385-388.

Zhou, Z. Y., Zhou, Z. Q. and Zhang, J. L. (1989).
Ordovician trilobite biofacies of North China
Platform and its western marginal area. Acta
Palaeontologica Sinica 28 (3), 296-313.

Zhou, Z. Y., Zhou, Z. Q., Siveter, D. J. and Yuan, W. W.
(2003). Latest Llanvirn to early Caradoc trilobite
biofacies of the north-western marginal area of

the Yangtze Block, China. Special Papers in

Palaeontology 70, 281-291.
Zhou, Z. Y., Zhou, Z. Q., Yuan, W. W. and Zhou, T. M.

(2000). Late Ordovician trilobite biofacies and
palaeogeographical development, western Hubei-
Hunan. Journal of Stratigraphy 24 (4), 249-263 (in
Chinese with English abstract).

Proc. Linn. Soc. N.S.W., 129, 2008

ZHOU ZHI-YI AND ZHEN YONG-YI

APPENDIX

Based on the dataset compiled by Zhou and Zhen (in press), following are listed, in descending stratigraphic
order, Ordovician trilobite genera that have been recorded from the inner shelf/platform (areas 4-1, 5-1, 8-1 in
Fig. 1) and outer shelf/slope (areas 4-2, 5-2 and 8-2 in Fig. 1) of the Tarim, North and South China plates. The
late Katian-Hirnantian (Ashgill) trilobites are excluded from the list, as the early Ashgill inner shelf/platform
trilobite faunas are almost absent in China, and so are the early Darriwilian-early Katian (latest Arenig-Caradoc)
shallow-water taxa from Tarim, because they were associated with carbonate buildups, and are cosmopolitan
in distribution. The biofacies patterns are not well established yet in the Qaidam and Middle Qilian terranes
and the Altun faulted block (area 5-4 in Fig. 1), but, judging from the faunal sequences and palaeogeographic
framework (Zhou et al. 1996b, c), in addition to a few genera that may belong to the shallow-water dwellers,
most of the listed trilobites were associated with slope facies. Note that the slope facies trilobites from the
western marginal area (5-2 in Fig. 1) of the North China Platform occur only from the early Darriwilian
onwards. Also only a few Sandbian-early Katian (Caradoc) offshelf pelagic trilobites were recorded in the
Cathaysia Area (8-3 in Fig. 1) of South China, and most of them extended their distribution to the adjacent shelf
slope (the Jiangnan Area, 8-2 in Fig. 1). Therefore, in the following list they are incorporated into the fauna of
Area 8-2 (Fig. 1).

Sandbian-early Katian (Caradoc)

4-2: Alceste, Amphilichas, Amphitryon, Ampyx, Ampyxinella, Basilicus (Basiliella), Birmanites, Bulbaspis,
Calymenesun, Corrugatagnostus, Cyclopyge, Degamella, Dicranurus, Dionide, Dividuagnostus,
Ellipsotaphrus, Endymionia, Illaenus, Kanlingia, Kongqiaoheia, Lisogolites, Lonchodomas, Microparia
(Heterocyclopyge), Microparia (Microparia), Microparia (Quadratapyge), Nanshanaspis, Nileus,
Ovalocephalus, Parisoceraurus, Penderia, Pricyclopyge, Pseudosphaerexochus, Reedocalymene,
Remopleurides, Rhombampyx, Robergia, Sagavia, Scotoharpes, Shumardia, Sinocybele, Sphaerexochus,
Stenopareia, Taklamakania, Telephina, Trinodus, Xiushuilithus

5-1: Basilicus (Basilicus), Lamproscutellum, Lonchobasilicus, Metopolichas, Pliomerina, Pseudostygina,
Sphaerexochus

5-2: Birmanites, Chedaoia, Cyclopyge, Cyphoniscus, Geragnostus, Glaphurina, Kodymaspis, Lichas,
Lisogolites, Lonchodomas, Microparia (Quadratapyge), Nileus, Ovalocephalus, Paraphillipsinella,
Paratiresias, Phorocephala, Pliomerina, Pseudostygina, Rorringtonia, Shumardia, sss Stenopareia,
Telephina, Trinodus, Xenocybe

5-4: Ampyxinella, Basilicus (Basilicus), Corrugatagnostus, Cyclopyge, Elongatanileus, Endymionia,
Hemiarges, Lonchobasilicus, Madygenia, Mendolaspis, Nanshanaspis, Nileus, Ovalocephalus, Pliomerina,
Poronileus, Porterfieldia, Remopleurides, Rhombampyx, Shumardia, Taklamakania, Tarimella, Telephina,
Toernquistia, Trinodus, Yumenaspis

8-1: Agerina, Amphilichas, Ampyx, Annamitella, Birmanites, Bumastoides, Calymenesun, Diacanthaspis,
Dicranurus, Dulanaspis, Hexacopyge, Illaenus, Lamproscutellum, Lonchodomas, Metopolichas,
Ovalocephalus, Parillaenus, Phorocephala, Prionocheilus, Prosopiscus, Pseudosphaerexochus,
Rhombampyx, Sinocybele, Telephina

8-2: Agerina, Alceste, Amphitryon, Ampyx, Ampyxinella, Aspidaeglina, Birmanites, Calymenesun, Cekovia,
Corrugatagnostus, Cyamella, Cyclopyge, Decoroproetus, Degamella, Diacanthaspis, Dionide, Dionidella
(Huangnigangia), Dislobosaspis, Dubhglasina, Effnaspis, Ellipsotaphrus, Elongatanileus, Encrinurella,
Gastropolus, Girvanopyge, Hanjiangaspis, Hexacopyge, Holdenia, Jianxilithus, Lamproscutellum,
Leiagnostus, Lichas, Lisogolites, Lonchodomas, Madygenia, Megatemnoura, Miaopopsis, Microparia
(Microparia), Microparia (Quadratapyge), Nanillaenus, Nileus, Niuchangella, Oedicybele, Ogmasaphus,
Ovalocephalus, Panderia, Paraphillipsinella, Parillaenus, Parisoceraurus, Pentacopyge, Phillipsinella,
Phorocephala, Placoparia, Platyptychopyge, Pricyclopyge, Pseudampyxina, Pseudopetigurus,

Proc. Linn. Soc. N.S.W., 129, 2008 193

ORDOVICIAN BIOGEOGRAPHY OF CHINA

Pseudosphaerexochus, Psilacella, Quyuania, Reedocalymene, Remopleurides, Rhombampyx, Rorringtonia,
Sagavia, Sarkia, Shumardia, Sinocybele, Sphaeragnostus, Sphaerexochus, Spinillaenus, Stenoblepharum,
Stenopareia, Symphysops, Taklamakania, Telephina, Trinodus, Xenocyclopyge, Xiushuilithus

Mid-late Darriwilian (Llanvirn)

4-2: Amphitryon, Birmanites, Endymionia, Gog, Illaenus, Liomegalaspides, Mendolaspis, Mioptychopyge,
Nanillaenus, Nanshanaspis, Nileus, Ogmasaphus, Ovalocephalus, Pseudocalymene, Rhombampyx,
Shumardia, Taklamakania, Tarimella, Telephina, Yanhaoia, Zhenganites

5-1: Ampyx, Basilicus (Basilicus), Basilicus (Basiliella), Basilicus (Parabasilicus), Glaphurina,
Lamproscutellum, Lonchobasilicus, Pliomerina, Sphaerexochus

5-2: Abulbaspis, Basilicus (Basiliella), Birmanites, Conophrys, Dulanaspis, Endymionia, Nileus,
Ovalocephalus, Paraptychopyge, Poronileus

5-4: Ampyxinella, Basilicus (Basilicus), Basilicus (Basiliella), Elongatanileus, Geragnostus, Hemiarges,
Illaenus, Leiagnostus, Lonchobasilicus, Lonchodomas, Mendolaspis, Nanshanaspis, Nileus, Ovalocephalus,
Paradionide, Plasiaspis, Pliomerina, Poronileus, Porterfieldia, Prosopiscus, Rhaombampyx, Shumardia,
Symphysurus, Taklamakania, Tarimella, Telephina, Toernquistia

8-1: Agerina, Amphilichas, Ampyx, Annamitella, Birmanites, Bumastoides, Calymenesun, Calymenia,
Diacanthaspis, Dicranurus, Hexacopyge, Illaenus, Lonchodomas, Metopolichas, Neseuretus, Ovalocephalus,
Parillaenus, Phorocephala, Prionocheilus, Prosopiscus, Pseudosphaerexochus, Rhombampyx, Sinocybele,
Telephina, Vietnamia

8-2: Agerina, Ampyx, Bathycheilus, Birmanites, Calymenesun, Carolinites, Cyamella, Cyclopyge,

Dionide, Gog, Hemisphaerocoryphe, Hexacopyge, Illaenus, Leiagnostus, Liomegalaspides, Lonchodomas,
Megatemnoura, Miaopopsis, Microparia (Microparia), Microparia (Quadratapyge), Mioptychopyge,
Nanillaenus, Nileus, Ogmasaphus, Ovalocephalus, Panderia, Paraphillipsinella, Paratiresias, Parillaenus,
Parisoceraurus, Platyptychopyge, Pricyclopyge, Prionocheilus, Pseudocalymene, Pseudopetigurus,
Pseudosphaerexochus, Pytine, Reedocalymene, Rhombampyx, Rorringtonia, Sagavia, Sinocybele,
Spinillaenus, Stenoblepharum, Stenopareia, Telephina, Trinodus, Yanhaoia, Zhenganites

Floian-early Darriwilian (Arenig)
4-1: Liomegalaspides

4-2: Birmanites, Carolinites, Eccoptochile, Hemisphaerocoryphe, Illaenus, Liomegalaspides, Mioptychopyge,
Nanillaenus, Nileus, Ogmasaphus, Ovalocephalus, Pseudocalymene, Yanhaoia, Zhenganites

5-1: Basilicus (Basilicus), Basilicus (Basiliella), Basilicus (Parabasilicus), Eoisotelus, Illaenus, Pliomerina

5-2: Abulbaspis, Annamitella, Basilicus (Basiliella), Geragnostus, Glaphurina, Gog, Lonchodomas,
Mendolaspis, Phorocephala, Placoparina, Poronileus, Pytine, Tongxinaspis

5-4: Annamitella, Basilicus (Basiliella), Cybelopsis, Geragnostus, Homalopyge, Illaenus, Omuliovia,
Porterfieldia, Rhombampyx, Scotoharpes, Tsaidamaspis

8-1: Agerina, Ampyx, Carolinites, Ceratolithus, Chengkouella, Fenghuangchengia, Geragnostus,
Guizhouhystricurus, Guizhoupliomerops, Hagiorites, Hanchungolithus, Hungioides, Liomegalaspides,
Madygenia, Meitanopsis, Metopolichas, Mioptychopyge, Neopsilocephalina, Neseuretus, Ningkianites,
Ningkianolithus, Niobella, Omeipsis, Omuliovia, Ovalocephalus, Phorocephala, Pseudocalymene,
Psilocephalina, Psilocephalops, Rhombampyx, Saltaspis, Scotoharpes, Taihungshania, Yinpanolithus

194 Proc. Linn. Soc. N.S.W., 129, 2008

ZHOU ZHI-YI AND ZHEN YONG-YI

8-2: Agerina, Ampyx, Alloillaenus, Annamitella, Aristocalymene, Aulacopleura (Paraaulacopleura),
Birmanites, Caputrotundum, Carolinites, Celmus, Ceratocephalina, Cyclopyge, Degamella, Diacanthaspis,
Dikelocephalina, Dionide, Eccoptochile, Euloma, Geragnostus, Gog, Han, Hanchungolithus,
Hemisphaerocoryphe, Hexacopyge, Hungioides, Illaenus, Incaia, Liomegalaspides, Loganopeltis,
Madygenia, Microparia (Microparia), Microparia (Quadratapyge), Mioptychopyge, Nanillaenus, Neseuretus,
Nileus, Ningkianites, Ningkianolithus, Niobe, Niobella, Opipeuterella, Ovalocephalus, Paraphillipsinella,
Paratiresias, Phorocephala, Pricyclopyge, Proscharyia, Prosopiscus, Pseudocalymene, Pseudopetigurus,
Rhombampyx, Sagavia, Scotoharpes, Shumardia, Sinocybele, Symphysurus, Taihungshania, Toernquistia,
Trinodus, Xystocrania, Yinpanolithus, Zhenganites

Tremadocian
4-1: Aksuaspis

4-2: Acrocephalina, Asaphopsoides, Bienvillia, Borthaspidella, Diceratopyge, Dichelepyge, Dividuagnostus,
Euloma, Harpides, Hysterolenus, Leiagnostus, Lotagnostus (Semagnostus), Macropyge, Niobella, Norinia,
Platypeltoides (Troedssonia), Prospectatrix, Proteuloma, Pseudocalymene, Psilocephalina, Rhadinopleura,
Scotoharpes, Shumardops, Trilobagnostus, Trinodus

5-1: Annamitella, Apatokephalops (Lulongia), Apatokephalus, Asaphellus, Asaphopsoides, Chosenia,
Dikelocephalina, Hystricurus, Ilaenus, Jiia, Jiliaoaspis, Jujuyaspis, Kainella, Kayseraspis, Koraipsis,
Leiostegium (Euleiostegium), Leiostegium (Jinanaspis), Leiostegium (Leiostegium), Omuliovia, Parapilekia,
Peltabellia, Penchiopsis, Platypeltoides (Troedssonia), Protopliomerops, Pseudorhaptagnostus,
Remopleuridiella, Scotoharpes, Sinobathyurus, Songtaoia, Strigigenalis, Trilobagnostus, Wanliangtingia,
Yosimuraspis

5-4: Bienvillia, Ceratopyge, Conophrys, Geragnostus, Harpides, Hystricurus, Kainella, Leiostegium
(Leiostegium), Nileus, Omuliovia, Onychopyge, Parabolinella, Platypeltoides (Troedssonia), Pseudokainella
(Pseudokainella), Symphysurus, Szechuanella, Yinaspis

8-1: Annamitella, Apatokephalops (Apatokephalops), Asaphellus, Asaphopsoides, Aulacopleura
(Paraaulacopleura), Chashania, Chengkouella, Chosenia, Chungkingaspis, Conophrys, Dactylocephalus,
Dikelocephalina, Geragnostus, Goniophrys, Guizhouhystricurus, Harpides, Hungioides, Hystricurus, Iduia,
Illaenus, Jiia, Jinshaella, Lohanpopsis, Parapilekia, Pseudocalymene, Psilocephalina, Remopleuridiella,
Scotoharpes, Songtaoia, Trinodus, Tungtzuella, Wanliangtingia, Yosimuraspis

8-2: Acrocephalina, Akoldinioidia, Amzasskiella, Anglagnostus, Apatokephalus, Asaphopsoides, Bienvillia,
Birmanites, Brachyhipposiderus, Ceratopyge, Chosenia, Ciliocephalus, Clavatellus, Conophrys, Degamella,
Diceratopyge, Dichelepyge, Dividuagnostus, Euloma, Geragnostus, Gymnagnostus, Harpides, Hospes,
Hunanopyge, Hypermecaspis, Hysterolenus, Illaenopsis, Leiagnostus, Leiostegium (Leiostegium), Levisaspis,
Liexiaspis, Lotagnostus (Semagnostus), Macropyge, Metayuepingia, Micragnostus, Niobe, Niobella,
Onchonotellus, Onychopyge, Orometopus, Palaeoharpes, Parabolinella, Parapilekia, Pharostomina,
Platypeltoides (Troedssonia), Proscharyia, Prospectatrix, Protarchaeogonus, Proteuloma, Protopliomerops,
Pseudocalymene, Pseudokainella (Pseudokainella), Pseudokoldinioidia, Pseudorhaptagnostus,
Rhadinopleura, Scotoharpes, Sinoparapilekia, Songtaoia, Strictagnostus, Symphysops, Szechuanella,

Symphysurus, Taoyuania, Trilobagnostus

Proc. Linn. Soc. N.S.W., 129, 2008 195

be
7 ne
on 7
HOON veo UNDER
is ee a ne
eid: Uowiiewtiea Kew Al emer, ee fe TA ‘ pil i Vf
Oykinnrer ly pay log Hoey) nena saa Agate toe
Gtlarny, 3 capac) datidveny ideo antisioe
wiyiiteeivlaeehk hebl S60) cepeeronsae) ssiiGhe Perens
here a2 eae Wea ttt | IAD! yor ehh mL ony t sprees pene:
cute yom sree). too eerie aC yy ik meyewaly, Ananth ornoery pale
went abd otk, casa elliiy Ob a) bevel alll cohaef Feanliaenedtigna wtp

‘dena Neto my te Arwend in) Wey eee arin A mor aro | Nea ORNC eH sien ae; Pri
bw tact l minwuding lt. icin ont) Sieg? renee” Savgeeestrsa,” Sana

vega § eect, yireri oie

A

= Ora ~ af HE we:
i ’ :
9 ‘een, MAnkiqra el
seta
wii er olawC) wade). Willis Co Yel tierce Dh ated QED, sarin

Jadiind pmben ives Me ny “inked. KO cdboynal ty: iy sabvictibientiers Lieven), actors
ry wl Oe ave: cal Kargapiaaliagh: jampyhayehachil. sowhliateet Nabi om mercial
“y HAytteerm ; + relhcanar Vy var RE?
talet eal’ > eelsyortqohagee ky carlwhegoal es hisitesoe hy Seas Ue Taylor mpeg sllsiienaitaly
Hoe CHL saetresn rt Ayal oho ung tan.ftaka tian ald a focracinur Epona

nto aren Ls cA aed A tire) ile alee ‘ctw aration A (verge Neth
worsens! (Chaemre) eryeochiic lion Uterine AeA reviseskanys
ied wien lols) leesyiaid?, aeoniped abu intend, ear ie ulbatirweih
‘ fi viet Qarol ihe ait at whdiiee. :
aemiry ivieteSoniovie woe cei god putt annua ie aaa shiguighihall Sash
Vineet “chawrieu(T Pesta ules WD aso Ay areca ey tees ye ir g
itl oo ‘wa\aeeipat Ula rane 2)
wh eens, yO ony gneiig, ;
Laags yw talboite: exilicadiohiethe scNivaeih shh it sinha
Liev die ol) Apuees ld emanate ub Goa vy eel Qi aww, Yavin) convene
with yrds lian) 2a aN Grarel the itucitaseed 1 ae ia vat AD at
blefabplanie Tale ie sarsneloodhoe) oltslnsat deeonnado’ Siletl bak Sine
unpwarical loupe lid wonsigel) 2ubentTT pitas aaegwnthe

slbrmi A once bh. owls tGliah Saat anh al\erizzisia hk anti ane ant welt

sD \Wevkeuuat aoe vel row il soa isvepl otal een Loe rer iagcay 3. iagrelpinngalay er ney vant

Sr) Avice ys aaingn cin: oiieelvgadticn slenuinda eure diac peel su arees

iets ,' Want \)\ tifibiedie) Jeigotigpied cede scan abl) 2aveshayei ew ll PASTAS RY eqs app eusy ie
wteniee 7) Sota Autdrucosel. aha daserr sat fin dona tan oupenetey 1 pesto) Gy Fae

: Lay gy Mary [oder ) sareceatals ny Pay Lele ates) aula .

Wh de cunihedty Fonotom tai aaalprsanentl 4 JshinntatnnenW Aayebnciesens aah soy ie

foun anepoher a aires A cbhand (in beraboes 72 Koen) ation) alan ae can sol *

uw COCIIPLE TEOMA: elboteHhd OlaliQEin Ont 2x ne hres Awe

' GOOG E nei

i F : wr bey ene a, Cae HeaER):

Fits, TTAaee Sh Trem eed Ba ce

atetes igen i) i y mee eae

14 eG ee Le

pre ana Owe “lipeh Ae Unreal

196 hues PE MOL.

The Upper Ordovician Kenyu Formation in the Boorowa
District, Southeastern New South Wales

I.G. Percivau!, Y.Y. ZHEN”, D.J. PoGson? AND O.D. THomas*

‘Geological Survey of New South Wales, Department of Primary Industries, 947-953 Londonderry Road,
Londonderry, N.S.W. 2753 (ian.percival@dpi.nsw.gov.au);
Australian Museum, 6 College Street, Sydney, N.S.W. 2010 (yongyi.zhen@austmus.gov.au);
388 Tallean Road, Nelson Bay, N.S.W. 2315 (poggo@optusnet.com.au);
4Geological Survey of New South Wales, Department of Primary Industries, 161 Kite Street, Orange, N.S.W.
2800 (owen.thomas@dpi.nsw.gov.au)

Percival, I.G., Zhen, Y.Y., Pogson, D.J. and Thomas, O.D. (2008). The Upper Ordovician Kenyu
Formation in the Boorowa District, Southeastern New South Wales. Proceedings of the Linnean Society

of New South Wales 129, 197-206.

Conodonts obtained during mapping of the Boorowa 1:100000 geological sheet indicate a late Gisbornian
to earliest Eastonian age (Late Ordovician: late Sandbian to earliest Katian) for allochthonous limestone
in the Kenyu Formation. This age is based on co-occurrence of Belodina compressa, Phragmodus undatus
and Yaoxianognathus wrighti, associated with Drepanoistodus suberectus, Panderodus gracilis, Periodon
aculeatus, Protopanderodus liripipus, Scabbardella sp. cf. altipes and Yaoxianognathus sp. The faunal
association, including acrotretide, discinide and lingulide brachiopods in addition to the conodonts,
indicates that the limestone was probably originally deposited on the shelf edge, prior to being dislodged
down the flanks of a volcanic island in a mass flow. The late Gisbornian to earliest Eastonian age recognised
for the Kenyu Formation provides an important constraint on the age and cessation of contemporaneous
volcanism in the central Macquarie Arc, represented more extensively further north by the Walli Volcanics
and Fairbridge Volcanics. No significant break intervened between the end of this volcanism and ensuing
deposition of widespread limestones of Eastonian age on the Molong Volcanic Belt.

Manuscript received 9 November 2007, accepted for publication 6 February 2008.

KEYWORDS: Conodonts, Fairbridge Volcanics, Kenyu Formation, Late Ordovician, Macquarie Arc,

Walli Volcanics

INTRODUCTION

The Kenyu Formation was described by Stevens
(1955) as a sequence of sediments and volcaniclastics
together with andesite, located in a narrow belt
extending roughly north-south adjacent to the
western edge of the Wyangala Batholith between
Boorowa and Cowra. The southern limit of outcrop
is near the Boorowa to Gunnary road (Fig. 1). Best
exposures (though these are far from complete) occur
in the Boorowa River valley, 10-15 km northeast of
Boorowa on the Boorowa 1:100 000 map sheet, and
further north between the Lachlan Valley Way and the
Wyangala Batholith in the vicinity of Godfreys Creek.
Due to the problematic outcrop, lack of internal age
control, and faulted boundaries with adjoining rock
units, relationships of the Kenyu Formation were not
previously satisfactorily understood. Stevens (1955)

inferred an Ordovician age, whereas Offenberg (1974)
assigned to the Kenyu Formation an age between Late
Ordovician and Middle Silurian. Inferences about
possible time equivalence of this unit to the Walli
Volcanics (itself lacking age constraints) further north
on the Cowra 1:100 000 geological sheet (Krynen
and Pogson 1998) were made without the benefit of
palaeontological evidence.

During remapping of the Goulburn 1:250 000
map sheet by the Geological Survey of NSW, the
Kenyu Formation has been examined in detail, the
concept of the unit refined to exclude derived rocks
of Early Silurian age, and a representative section
defined in the northern part of the Boorowa 1:100 000
geological sheet (Pogson et al. in press). Fossils found
in the formation for the first time establish its age and
depositional environment, and enable placement in
its correct temporal and tectonic setting. Preliminary
observations of a very small conodont fauna obtained
from an allochthonous limestone block during the

=n

be

148°40'E

AAAAAA
ANNKAKA
ANAAAA
ANAAA
AANAAA
ANAAA
AAAAA
AAAAN
AANAAA
AANAARK

ANANAAAA
ANAKRAAAAA
ANANAAAAAA AY

AKNAKAAAKAAA
ANAAAARAA
NANAANAAA

Xe
ANAANARA ADA A AAG
AANAAA DRUNK NAA

ANAKAKKRAAARAAA S

AKRANANAKANAKAAAAAA

\O
oO

UPPER ORDOVICIAN KENYU FORMATION

148°50'E

33°50'S 32°30'S

Wellingtone

rs Bakers
Swamp

33°00'S

33°30'S INSET REFERENCE

[2 Fairbridge Volcanics
ee Cargo Volcanics
Walli Volcanies

HB Kenyu Formation

34°00'S

°
Boorowa

34°30'S
148°40'E

149°00'E

NEW SOUTH WALES
7] 34°10’'S

Broken Hill raater

Interest
\
ACT(P

Newcastle
Sydney
Wollongong

REFERENCE
Upper Silurian

Wyangala Batholith
Lower Silurian

Douro Group

E+} Lower Silurian sediments
Upper Ordovician

34°20'S

Kenyu Formation
& Bendoc Group

Lower Middle Ordovician
fr Adaminaby Group

ANAAARAAAA \ A A
: * Fossil location

0 10km

2008_10_0181

Proc. Linn. Soc. N.S.W., 129, 2008

I.G. PERCIVAL, Y.Y. ZHEN, D.J.

early phase of the remapping project enabled Percival
(2001) to deduce a Gisbornian age. Subsequent
resampling of this limestone yielded a significantly
larger and more diverse fauna, documented in this
paper, that has allowed precise correlations with
other volcanic and volcaniclastic-dominated units
in the Macquarie Arc, thus providing important age
constraints on the cessation of Phase 2 volcanism
(Crawford et al. 2007) in the Molong Volcanic Belt.

LITHOLOGICAL CHARACTERISTICS

For much of its exposure the Kenyu Formation
occupies a narrow fault-bounded belt, in contact with
Early Silurian volcanic rocks of the Douro Group to
the west and faulted against the Wyangala Batholith
to the east (Johnston et al. 2001). On the Cowra 1:100
000 geological sheet south of the Lachlan River
outcrop is very poor, being restricted to a handful of
exposures that are surrounded by alluvium. In this
area, the extent of the formation is largely defined by
its strong aeromagnetic response (Krynen and Pogson
1998). Further south in areas of better exposure, the
Kenyu Formation is intruded by the Licking Gully
Granite (of the Wyangala Batholith). In the Frogmore
area (Fig. 1) several fault-bounded blocks of Kenyu
Formation abut the Early to Middle Ordovician
Adaminaby Group, Late Ordovician Bendoc Group
and the Early Silurian Douro Group. Other Early
Silurian sedimentary rocks, previously included
within the Kenyu Formation, but now recognised
as a new formation (to be defined by Pogson et al.
in press), occupy three fault-bounded blocks near
Frogmore, Gunnary, and immediately west of the
southern belt of Kenyu Formation (Fig. 1).

Faulting and the lack of continuous outcrop
limits the potential for a type section. The area
where the allochthonous limestone blocks occur is
to the southeast of Godfreys Creek (Fig. 1). Here,
representative lithologies of the Kenyu Formation (as
briefly described below — more detail is presented in
Pogson et al. in press) are exposed along Right Hand
Creek and tributary gullies (GR 661065 6218310
to GR 657960 6219380) and along Narrallen Creek
in the vicinity of GR 661073 6215719. Isolated dip
directions and depositional younging trends in this

POGSON AND O.D. THOMAS

area imply that the limestone blocks lie in the upper
part of the formation. Total thickness of the Kenyu
Formation is unknown, as both the top and bottom of
the unit are faulted out.

The bulk of exposures of the Kenyu Formation
consist of dark green to grey, very thinly bedded
to medium bedded cherty mudstone (commonly
silicified) and siltstone often interbedded with
fine-grained mafic volcaniclastic sandstone. Large
pavements along Narrallen Creek display parallel
bedding laminations, grading and erosional bases,
consistent with deposition by turbidity currents.
Examples of soft-sediment deformation in this
area include flame structures and slump folds. The
stratigraphic relationship between these sedimentary
rocks, and nearby sequences dominated by primary
volcanics and volcaniclastic conglomerates is
unclear.

A variety of plagioclase-phyric and pyroxene-
phyric andesites and pillow basalts is present in the
Kenyu Formation. Black, aphanitic to porphyritic,
flow-banded basalt occasionally displaying pillow
structures occurs in several locations. One of these
pillows (at GR 660628 6218239) is partially enveloped
by chert which unfortunately lacks microfossils.

In the Right Hand Creek area, the Kenyu
Formation is dominated by volcaniclastic deposits
including polymictic conglomerate with subrounded
to angular clasts up to cobble and small boulder size
of volcanic sandstone and siltstone, and occasionally
enclosing masses of porphyritic (feldspar-phyric)
andesite/basalt, indicating primary volcanism.
Rare allochthonous limestone blocks of various
dimensions (the largest up to 250 x 50 m) occur
within volcaniclastic sediments in the Right Hand
Creek and Narrallen Creek areas. They are all strongly
recrystallised, with a definite tectonic foliation.
Conodont CAI values of 5+ attest to the effects of
considerable post-depositional heat and pressure on
these limestones.

PALAEONTOLOGY

Insoluble residue of an allochthonous strongly
recrystallised limestone sampled at GR 660085
6217818 (conodont sample C1961 — locality shown

Figure 1 (LEFT). Main map depicts geological units simplified from the Boorowa 8629 1:100 000 geo-
logical sheet south of 34°S, together with data interpreted from the Cowra 8630 1:100 000 geological
sheet above this latitude, showing inferred extent of Kenyu Formation and localities mentioned in the
text. Note that all boundaries between the Kenyu Formation and adjacent units are faulted. Fossil local-
ity C1961 is identical in position to C2362. Inset map shows spatial relationship of Kenyu Formation to
Walli Volcanics, Fairbridge Volcanics, and Cargo Volcanics on Molong Volcanic Belt.

Proc. Linn. Soc. N.S.W., 129, 2008

199

UPPER ORDOVICIAN KENYU FORMATION

[ees ae
%
Lu O. robustus
ZZ
ZI S& i j
ZB Nok <z| ° O. velicuspis
Ze | eile
Olecs | SAIS
S Q iw B. confluens
mn
S a
M@ inean SS Sl
| O uw P. tenuis =
lle P. undatus 2
Lu Zz Zz a B. compressa
fr < ES © | E quadridactylus
= QO 6s P. aculeata
PS
x |
18
C.sweeti
| + eee.
t © ges
< es:
O = 7 || 2 SY ees
S eT < ja) ses Oo
=|e= £ 8a
Oy= | = SESS
= = fs} Ke)
a na nN o> GO
m~ eS ~ s ao & G a
ing < ja) ££ Oo
Oy] =| a Snaes
. = oO O
Q g “North Anerican a zs
= Midcontinent”

(Shallow Sea
condont zones)

Taogupognathus philipi
Taoqupognathus blandus

Borowa - Walli Bakers Swamp -
Cowra area area Molong area

CHEESEMANS
CREEK
FORMATION

MALONGULLI
FORMATION

ECLIEFDEN CAVES
F LIMESTONE —

VVOLCANICS/ V
Vw wy VY VW WY W
v KENYU vVivvvvvvvy

FORMATION V v v v v v v Y WAHRINGA
VvVVVVVMV VV VV vv LIMESTONE
VV YY YY YW YW we Ww MEMBER

VV AN EENIIIN, TRV NEY SN VN,
ME Np NE MBNIN ENON N/E NIE NISON/.
WV NORE NIN NIN NEN ENS NEN SN

= WA ONS WEN NYS IN NE ONE NONE ONIN

iS WE WA WAS NYE NAINA NEON ME WN NW NON NWN NY
S VON ES NN TONNE

% | (base unknown)| (base unknown) Jy. wv ap, VJ
= | FAIRBRIDGE ,,

g vy VOLCANICS.
dD NY NP NA SEN SW

8 VVVNVVV WY
= VAAN VN

fo)

x (base unknown)

2007_11_0182

Figure 2. Stratigraphic correlations along the Molong Volcanic Belt (right-hand columns) and ranges of
key species (central column), plotted against segment of the Middle to Upper Ordovician timescale and
conodont zonation (after Webby et al. 2004 and Goldman et al. 2007). Note that the Yuranigh Limestone
Member of the Fairbridge Volcanics is not shown; it equates to the uppermost part (late Gisbornian) of

the Wahringa Limestone Member.

on Fig. 1), obtained after acid dissolution in 10%
acetic acid and separation in sodium polytungstate,
yielded 14 elements referrable to Protopanderodus
liripipus and Panderodus gracilis. Resampling of this
limestone block (conodont sample C2362) produced
99 identifiable elements, distributed amongst nine
species including Belodina compressa, Drepanoist-
odus suberectus, Panderodus gracilis, Periodon
aculeatus, Phragmodus undatus, Protopanderodus
liripipus,Scabbardellasp.cf.altipes, Yaoxianognathus
wrighti and Y. sp., indicating a late Gisbornian to
earliest Eastonian age (Fig. 2). This age corresponds
to the late Sandbian to earliest Katian stages of the
Late Ordovician, approximately 457-455 Ma on
the age scale of the International Commission on
Stratigraphy (Cooper and Sadler 2004).

The most significant species supporting
this age determination are Belodina compressa

200

and Phragmodus undatus. Belodina compressa 1s
mainly differentiated from the succeeding species,
B. confluens, in having a distinctly flattened section
of the anterior margin at the antero-basal corner in
inner lateral view (Fig. 3A). Absence of B. confluens
from the assemblage in the Kenyu Formation
allochthonous limestone implies that the latter was
not contemporaneous with the widespread Eastonian
limestones of the Macquarie Arc in which B.
confluens is prolific. Grandiform and compressiform
elements of B. compressa, represented by eight
specimens recovered from sample C2362 (Fig. 3A,
B), are identical with those obtained from the upper
part of the Wahringa Limestone Member (in the
Bakers Swamp area further north on the Molong
Volcanic Belt) and from allochthonous limestone
blocks in the immediately overlying Fairbridge

Proc. Linn. Soc. N.S.W., 129, 2008

I.G. PERCIVAL, Y.Y. ZHEN, D.J. POGSON AND O.D. THOMAS

Volcanics (Zhen et al. 2004). In the North American
Midcontinent succession typical of warm shallow
seas, B. compressa is recognized as an index zonal
species with a stratigraphic range from the compressa
Zone of late Gisbornian age-equivalence to the tenuis
Zone of the early Eastonian (Sweet 1988).

Co-occurrence of Phragmodus undatus,
represented by four S and three M elements in
sample C2362, further restricts this age range from
latest Gisbornian to earliest Eastonian (wndatus to
tenuis zones). In the North American Midcontinent
succession, P. undatus defines the eponymous Zone
immediately above the compressa Zone with a range
extending from the wndatus Zone to the end of the
Ordovician (Sweet 1988). The species has been widely
recorded in Eastonian-age carbonates of central NSW
(Zhen and Webby, 1995; Trotter and Webby 1995;
Zhen et al. 1999, 2003).

Broad support for a late Gisbornian to earliest
Eastonian age assignment for the Kenyu Formation
limestone is provided by the co-occurrence of
Protopanderodus liripipus, Periodon aculeatus and
Yaoxianognathus wrighti in the fauna. P. liripipus
has an age range from the late Gisbornian to near the
end of the Ordovician (Sweet 1988). It was recorded
in the upper Wahringa Limestone Member also in
association with B. compressa, Periodon aculeatus
and Panderodus gracilis (Zhen et al. 2004), and from
the Bowan Park Limestone Subgroup (Zhen et al.
1999) in central New South Wales. Yaoxianognathus
wrighti, widely distributed in the Fossil Hill Limestone,
Bowan Park Limestone Subgroup, and other time
equivalent limestones in central New South Wales,
is represented only by the Pa element (Fig. 3U) with
seven specimens recovered from sample C2362. Its
presence in the Kenyu limestone most likely indicates
an extension of its age range into slightly older rocks
than was previously known.

Two fragmentary specimens from the same
sample are referred to Pb elements (Fig. 3V) of
Yaoxianognathus sp. and an additional specimen
from sample C2366 is a bipennate ramiform element
assignable to the S (most likely Sc) position (Fig.
3W). These specimens show marked differences
from corresponding elements of Y. wrighti, but due
to the limited material it is uncertain whether they
represent elements of a single additional species of
Yaoxianognathus.

DISCUSSION

Age connotations
Although the age range deduced for the conodont

Proc. Linn. Soc. N.S.W., 129, 2008

fauna from the Kenyu Formation is relatively
restricted, the fact remains that these fossils were
extracted from allochthonous limestone which is
poorly constrained stratigraphically within this
unit. The late Gisbornian to earliest Eastonian age
merely indicates the maximum age (for the originally
deposited limestone) and the allochthonous blocks
may either have been penecontemporaneously
redeposited, or else could be significantly younger.
As discussed below, regional correlations strongly
support the former view. The widespread volcanic
hiatus (Packham et al. 2003, Percival and Glen
2007) that extended throughout the western Molong
Volcanic Belt in Eastonian time precludes Kenyu
Formation volcanicity during that interval. Large
feldspar phenocrysts, observed in Kenyu Formation
andesites located in the vicinity of the allochthonous
limestone blocks, are also characteristic of other pre-
Eastonian lavas in the Molong Volcanic Belt, such
as the Walli, Fairbridge and Cargo Volcanics, and do
not suggest similarities with post-Eastonian Phase 4
volcanics.

Previously proposed correlations (Krynen and
Pogson 1998, Percival and Glen 2007) equated the
Kenyu Formation with the upper part of the Walli
Volcanics, exposed in the Walli-Cliefden Caves
area between Mandurama and Canowindra, 90
km north of Boorowa (Fig. 1). The late Gisbornian
to earliest Eastonian age now established for part
of the Kenyu Formation supports this correlation.
The Walli Volcanics contain no internal evidence
of age; however, this unit is overlain (with minor
disconformity) by shallow water sedimentary rocks of
the Fossil Hill Limestone (Cliefden Caves Limestone
Subgroup) at Fossil Hill, near Cliefden Caves
(Webby and Packham 1982). As these strata contain
early Eastonian (middle Late Ordovician: early
Katian Stage) conodonts (Zhen and Webby 1995),
any time break between deposition of volcaniclastic
conglomerates at the top of the Walli Volcanics
(eroded from immediately underlying andesitic
lavas), and the intertidal mudstones and impure
limestones which infilled the irregular topography
of the underlying volcanic island, must have been
minimal. Correlation between part of the Kenyu
Formation of late Gisbornian to earliest Eastonian
age and the upper Walli Volcanics does not preclude
both formations from having a depositional history
extending back into the Darriwilian, in common with
other volcanics in the Macquarie Arc (Percival and
Glen 2007).

Elsewhere in the northern Molong Volcanic
Belt of the Macquarie Arc, strata contemporaneous
with limestone in the Kenyu Formation include

201

UPPER ORDOVICIAN KENYU FORMATION

the Yuranigh Limestone Member of the Fairbridge
Volcanics in the vicinity of Molong (Percival et al.
1999). Of the depauperate conodont fauna obtained
from this very shallow water limestone, only Belodina
compressa (initially identified as B. confluens by
Percival et al. 1999, but reassigned by Zhen et al.
2004) is in common with the deeper water Kenyu
assemblage. Approximately 1300 m of volcaniclastic
sandstones and conglomerates, intruded by felsic
igneous rocks and hornblende-bearing dykes,
intervenes between the Yuranigh Limestone Member
and the Reedy Creek Limestone of early Eastonian
age. A comparable situation occurs in the Bakers
Swamp area, 35 km north of Molong, where the
upper beds of the Wahringa Limestone Member of
the Fairbridge Volcanics (Zhen et al. 2004) contains,
inter alia, the conodonts B. compressa, Panderodus
gracilis and Periodon aculeatus, again from a
relatively shallow water setting. The Wahringa
Limestone Member is overlain by a substantial but
unknown thickness of volcaniclastics, rare lavas
and allochthonous limestone pods, which equate to
the section of Fairbridge Volcanics overlying the

Yuranigh Limestone Member.

In the Bowan Park area between Orange and
Cudal, on the western side of the Molong Volcanic
Belt, the Cargo Volcanics is overlain by the Bowan
Park Limestone Subgroup which is of early to late
Eastonian age (Zhen et al. 1999). Pickett (1974)
recorded a solitary specimen of Belodina from a
limestone lens within the Cargo Volcanics south of
Cargo, though unfortunately Simpson et al. (2007)
were unable to relocate this lens to determine whether
its relationship to the enclosing volcanics was
conformable or allochthonous. Simpson et al. (2007)
recognised a significant discordance between bedding
orientations intheuppermostvolcaniclastic-dominated
successions of the Cargo Volcanics and those in the
overlying Bowan Park Limestone Subgroup. This
observation, combined with correlations (based on
time planes recording the cessation of terrigenous
erosional input) between the Bowan Park and
Cliefden Caves carbonate successions which suggest
that clastic deposition commenced as much as half
a zone earlier in the Fossil Hill Limestone (Webby
and Packham 1982; Packham et al. 2003; Simpson et

Figure 3 (RIGHT). Scanning electron microscope photomicrographs of conodonts from allochthonous
limestone in the Kenyu Formation. MMMC and C refer to registered specimen numbers and sample
numbers, respectively, in the Microfossil Collection of the Geological Survey of NSW, Londonderry;
numbers commencing with IY indicate digital photofiles from the Electron Microscope Unit, Austral-
ian Museum, Sydney.

A, B. Belodina compressa (Branson and Mehl, 1933); A, compressiform element, MMMC4367,
C2362, inner lateral view ([Y102001); B, grandiform element, MMMC4368, C2362, outer lateral view
(TY102005). ;

C, D. Scabbardella sp. cf. altipes (Henningsmoen, 1948); c element, MMMC4369, C2362, C, fur-
rowed side (1Y102010), D, unfurrowed side (1Y 102009).

E-G. Drepanoistodus suberectus (Branson and Mehl, 1933); E, Sb element, MMMC4370, C2362,
inner lateral view (TY 102023); F, M element, MMMC4371, C2362, posterior view (LY 102022); G, Sd
element, MMMC4372, C2362, inner lateral view (1Y 102024).

H-J. Panderodus gracilis (Branson and Mehl, 1933); H, falciform element, MMMC4373, C1961,
furrowed side (LY037006); I, graciliform? element with distally recurved cusp, MMMC4374, C2362,
furrowed side (LY102028); J, graciliform element, MMMC4375, C1961, furrowed side (1Y037007).

K-N. Periodon aculeatus Hadding, 1913; K, M element, MMMC4376, C2362, anterior view
(1Y102017); L, M element, MMMC4377, C2362, posterior view (LY102016); M, Sc element,
MMM.0C4378, C2362, inner lateral view (LY102018); N, Sc element, MMC4379, C2362, outer lateral
view (1Y102019).

O, P. Phragmodus undatus Branson and Mehl, 1933; O, M element, MMMC4380, C2362, posterior
view (1Y¥ 102021); P, Sc element, MMMC4381, C2362, inner lateral view (LY102020).

Q-T. Protopanderodus liripipus Kennedy, Barnes and Uyeno, 1979; Q, R, Sa element, MMMC4382,
C2362, lateral views ([Y102025, 1Y102026); S, Sb? element, MMMC4383, C1961, inner lateral view
(TY037004); T, P element, MMMC4384, C2362, inner lateral view (1Y 102027).

U. Yaoxianognathus wrighti Savage, 1990; Pa element, MMMC4385, C2362, outer lateral view
(TY102011);

V, W. Yaoxianognathus sp.; V, Pb? element, MMMC4386, C2362, outer lateral view (LY102015); W,
Sc? element, MMMC4387, C2366, inner lateral view ([Y102036). Scale bars are 100 um.

i)

Proc. Linn. Soc. N.S.W., 129, 2008

I.G. PERCIVAL, Y.Y. ZHEN, D.J. POGSON AND O.D. THOMAS

Proc. Linn. Soc. N.S.W., 129, 2008 203

UPPER ORDOVICIAN KENYU FORMATION

al. 2007), imply that the Cargo block had a different
geological history from that of the contiguous central
Molong Volcanic Belt to the east.

Depositional environment

The association of large allochthonous
limestone blocks with poorly sorted volcaniclastic
conglomerates in the upper Kenyu Formation suggests
that the limestones were redeposited downslope via
mass slumping. Evidence for an original depositional
environment on an unstable shelf edge comes from
the microfauna extracted from the residues (samples
C1961 and C2362) of the limestone. The conodont
fauna includes some forms (eg Belodina compressa
and Phragmodus undatus) that typically characterise
shallow waters, associated with Periodon aculeatus
and Protopanderodus liripipus that tend to inhabit
interpreted deeper water environments. The fauna
also includes acrotretide (Scaphelasma?) and
discinide (Orbiculoidea) brachiopods together
with large indeterminate thick-shelled lmgulides,
flat-spired gastropods, hyolithids and associated
opercula. Unfortunately these specimens are all
either fragmented or tectonically strained, preventing
precise identifications (and detracting from their
documentation by illustration in this paper). This
ecological association is somewhat reminiscent
of the fauna from allochthonous limestones in the
Malongulli Formation (overlying the Cliefden Caves
Limestone Subgroup), which are believed to represent
periplatform deposits originally deposited on the shelf
edge or upper slope and subsequently dislodged into
deeper water (Webby 1992).

The association in the Kenyu Formation of
allochthonous limestones with volcanics is typical of
successions accumulating around oceanic islands and
seamounts, where the steep flank gradients ensure
proximity of a variety of rock types and allow for
intermixing of otherwise disparate facies. Much of the
Kenyu Formation can be interpreted as volcaniclastic
and sedimentary deposits forming an apron proximal
to a source of primary volcanism, represented in the
Kenyu Formation by pillow basalts and minor flows
issuing from localised vents. Submarine eruption of
the correlative Walli Volcanics to the north, indicated
by the presence of pillow basalts (Smith 1967, 1968),
built the southern region of the Molong Volcanic Belt
up to depths where limestones were deposited on the
submarine flanks of this major volcanic centre.

204

CONCLUSIONS

Limestone within the Kenyu Formation is
determined as having a late Gisbornian to earliest
Eastonian depositional age, primarily on the basis of
co-occurrence of the conodonts Belodina compressa
and Phragmodus undatus. This age constraint
supports previously conjectured correlation of part
of the Kenyu Formation with the upper part of the
undated Walli Volcanics. Furthermore, a consistent
age relationship is recognisable along the entire length
of the central Molong Volcanic Belt, where limestones
of late Gisbornian age (Yuranigh and upper Wahringa
Limestone Members of the Fairbridge Volcanics) are
overlain by subsequent volcanic and volcaniclastic
deposits that are directly succeeded in the Molong
area by early Eastonian limestones. The Walli
Volcanics, that formed the volcanic island around
which the volcaniclastic-dominated apron of the
Kenyu Formation accumulated, are in turn overlain
by the early Eastonian age Fossil Hill Limestone.
The time interval between cessation of volcanism and
subsequent erosion, leading eventually to extensive
carbonate deposition, must have been very brief
in geological terms. Any suggestion of an angular
unconformity between the Walli Volcanics and the
Cliefden Caves Limestone Subgroup (cf Smith 1968)
is almost certainly not due to contemporaneous
tectonic activity, but rather results either from
differences in original depositional gradient, or else is
due to subsequent faulting. The situation in the Cargo
block on the western flank of the Molong Volcanic
Belt appears to differ, with both tectonic activity and
a significant depositional hiatus intervening between
the Cargo Volcanics and the overlying Bowan Park
Limestone Subgroup.

ACKNOWLEDGMENTS

Initial remapping of the Kenyu Formation and sample
collection was undertaken by Anthony Johnston
(formerly with the Geological Survey of NSW). We
thank Gary Dargan for conodont sample preparation
and Cheryl Hormann for cartography. SEM
photographs were prepared in the Electron Microscope
Unit of the Australian Museum. Constructive reviews
of the manuscript were provided by Gordon Packham
and John Pickett. Ian Percival and Owen Thomas

Proc. Linn. Soc. N.S.W., 129, 2008

I.G. PERCIVAL, Y.Y. ZHEN, D.J. POGSON AND O.D. THOMAS

publish with permission of the Deputy Director-
General, NSW Department of Primary Industries —
Minerals. This is a contribution to IGCP Project 503:
“Ordovician Palaeogeography and Paleoclimate”

REFERENCES

Branson, E.B. and Mehl, M.G. (1933). Conodont studies.
University of Missouri Studies 8, 1-349.

Cooper, R.A. and Sadler, P.M. (2004). The Ordovician
Period, pp.165-187. In Gradstein, F.M., Ogg, J.G.,
and Smith, A.G., (eds). “A Geologic Time Scale
2004’. 589 pp. (University Press, Cambridge).

Crawford, A.J., Meffre, S., Squire, R.J., Barron, L.M. and
Falloon, T.J. (2007). Middle and Late Ordovician
magmatic evolution of the Macquarie Arc, Lachlan
Orogen, New South Wales. Australian Journal of
Earth Sciences 54, 167-179.

Goldman, D., Leslie, S.A., Nolvak, J. and Young, S.
(2007). The Black Knob Ridge section, southeastern
Oklahoma, USA: the Global Stratotype-Section and
Point (GSSP) for the base of the Katian Stage of the
Upper Ordovician Series. In Li Jun, Fan Junxuan
and Percival, I.G., (eds.). The Global Ordovician
and Silurian. Acta Palaeontologica Sinica 46
(Supplement, June 2007), pp. 144-154.

Hadding, A.R. (1913). Undre dicellograptusskiffern
1 Skane jamte nagra dagra darmed ekvivalenta
bildningar. Lunds Universitets Arsskrift, Ny Féljd,
Afdelning 2, 9(15), 1-90.

Henningsmoen, G. (1948). The Tretaspis Series of
the Kullatorp Core. Bulletin of the Geological
Institutions, University of Uppsala 32, 374-432.

Johnston, A.J., Pogson, D.J., Thomas, O.D., Watkins,

J.J. and Glen, R.A. (2001). Boorowa 1:100 000
Geological Sheet 8629, Provisional 1“ Edition.
Geological Survey of New South Wales, Orange.
(Draft updated version, September 2007).

Kennedy, D.J., Barnes C.R. and Uyeno T.T. (1979).

A Middle Ordovician conodont faunule from the
Tetagouche Group, Camel Back Mountain, New
Brunswick. Canadian Journal of Earth Sciences 16,
540-551.

Krynen, J.P. and Pogson, D.J. (1998). Kenyu Formation.
p. 28. In Pogson, D.J. and Watkins, J.J. (compilers).
“Bathurst 1:250 000 Geological Sheet SI/55-8:
Explanatory Notes’. 430 pp. (Geological Survey of
New South Wales, Sydney).

Offenberg, A.C. (1974). ‘Explanatory notes on the
Goulburn 1:250,000 Geological Sheet SI/55-12’.

57 pp. (Geological Survey of New South Wales,
Sydney).

Packham, G.H., Keene, J.B. and Barron, L.M.

(2003). Middle to early Late Ordovician
hydrothermal veining in the Molong Volcanic Belt,
northeastern Lachlan Fold belt: sedimentological

Proc. Linn. Soc. N.S.W., 129, 2008

evidenceAustralian Journal of Earth Sciences
50,257-69.

Percival, I.G. (2001). Palaeontological determinations
from the Boorowa 1:100 000 sheet, between Cowra
and Yass. Palaeontological Report 2001/1. Geological
Survey of New South Wales Report 2001/438, 10 pp.
(unpubl.).

Percival, I.G. and Glen, R.A. (2007). Ordovician to
earliest Silurian history of the Macquarie Arc,
Lachlan Orogen, New South Wales. Australian
Journal of Earth Sciences 54, 143-165.

Percival, I.G., Morgan, E.J. and Scott, M.M. (1999).
Ordovician stratigraphy of the northern Molong
Volcanic Belt: new facts and figures. Geological
Survey of New South Wales, Quarterly Notes 108,
8-27.

Pickett, J.W. (1974). Ordovician and Silurian
conodonts from central-western New South Wales.
Palaeontological Report 74/11. Geological Survey of
New South Wales Report GS1974/151 (unpubl.).

Pogson, D.J., Simpson, C.J. and Johnston, A.J. (an press).
Kenyu Formation. In Thomas, O.D. (compiler).
“Goulburn 1:250 000 Geological Sheet SI/55-12
Explanatory Notes, 2"! Edition’. (Geological Survey
of New South Wales, Maitland).

Savage, N.M. (1990). Conodonts of Caradocian (Late
Ordovician) age from the Cliefden Caves Limestone,
southeastern Australia. Journal of Paleontology 64,
821-831.

Simpson, C.J., Scott, R.J., Crawford, A.J. and Meffre, S.
(2007). Volcanology, geochemistry and structure of
the Ordovician Cargo Volcanics in the Cargo-Walli
region, central New South Wales. Australian Journal
of Earth Sciences 54, 315-352.

Smith, R.E. (1967). Segregation vesicles in basaltic lava.
American Journal of Science 265, 696-713.

Smith, R.E. (1968). Redistribution of major elements
in the alteration of some basic lavas during burial
metamorphism. Journal of Petrology 9, 191-219.

Stevens, N.C. (1955). The petrology of the northern part of
the Wyangala Bathylith. Proceedings of the Linnean
Society of New South Wales 80, 84-96.

Sweet, W.C. (1988). “The Conodonta: Morphology,
Taxonomy, Paleoecology, and Evolutionary History
of a Long-Extinct Animal Phylum’. 212 pp.
(Clarendon Press, Oxford).

Trotter, J.A. and Webby, B.D. (1995). Upper Ordovician
conodonts from the Malongulli Formation, Cliefden
Caves area, central New South Wales. AGSO Journal
of Australian Geology and Geophysics 15(4), 475-
499.

Webby, B.D. (1992). Ordovician island biotas: New South
Wales record and global implications. Journal and
Proceedings of the Royal Society of New South Wales
125, 51-77.

Webby, B.D., Cooper, R.A., Bergstrém, S.M. and Paris,

F. (2004). Stratigraphic framework and Time Slices,
pp. 41-47. In Webby, B.D., Paris, F., Droser, M.L.

205

UPPER ORDOVICIAN KENYU FORMATION

and Percival, I.G. (eds.) ‘The Great Ordovician
Biodiversification Event’. 484 pp. (Columbia
University Press, New York).

Webby, B.D. and Packham, G.H. (1982). Stratigraphy and
regional setting of the Cliefden Caves Limestone
Group (Late Ordovician), central-western New South
Wales. Journal of the Geological Society of Australia
29, 297-317.

Zhen, Y.Y., Percival, I.G. and Farrell, J., (2003). Late
Ordovician allochthonous limestones in Late Silurian
Barnby Hills Shale, central western New South
Wales. Proceedings of the Linnean Society of New
South Wales 124, 41-63.

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. 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, Australia. 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, central New South Wales, Australia.
Géobios 32, 73-104.

Proc. Linn. Soc. N.S.W., 129, 2008

Age Determination and Growth in the Male South African Fur
Seal Arctocephalus pusillus pusillus (Pinnipedia: Otariidae)
Based upon Skull Material

C. L. Stewarpson!, T. Prvan?, M. A. MeYer? AND R. J. Rivcuie**

‘Botany and Zoology, Australian National University, Canberra, ACT, Australia
(Present Address, Fisheries and Marine Sciences Program Bureau of Rural Sciences, The Department of
Agriculture, Fisheries and Forestry, CANBERRA ACT 2601 Australia)
; Department of Statistics, Macquarie University, NSW 2109;
3Marine and Coastal management (MCM), Rogge Bay, Cape Town, South Africa; ‘School of Biological
Sciences, The University of Sydney, NSW 2006

*Corresponding Author (rrit3 143 @usyd.edu.au)

Stewardson, C.L., Prvan, T., Meyer, M.A. and Ritchie, R.J. (2008). Age determination and growth in the
male South African Fur Seal Arctocephalus pusillus pusillus (Pinnipedia: Otartidae) based upon skull
material. Proceedings of the Linnean Society of New South Wales 129, 207-252.

Skull remains are the most commonly found material of marine mammals and the most likely to be
kept in natural history collections. Morphology, relative size and growth of the skull in 83 South African fur
seals, Arctocephalus pusillus pusillus, from the coast of southern Africa are described. The South African
or Cape fur seal is very closely related to the Australian fur seal (Arctocephalus pusillus doriferus). Age
structure of populations is important in understanding the conservation status of an animal population
and the impacts of human activity upon the survival of viable wild populations of animal species. Skull
measurements (7 = 32 variables) were examined in relation to standard body length (SBL - defined as the
length from the nose to the tail in a straight line with the animal on its back), condylobasal length (CBL) and
chronological age (y) using linear regression. Animals ranged from 10 months to > 12 y (12* y). Twenty four
animals were of known-age, while 39 were aged from counts of incremental lines observed in the dentine
of tooth sections. Morphological observations were generally consistent with earlier studies. Condylobasal
length was highly, positively correlated with SBL and age. Overall, skull variables grew at a slower rate
than SBL, apart from height of mandible at meatus and angularis to coronoideus, which expressed isometry
relative to SBL. Condylobasal length continued to increase until at least 12 y, with no obvious growth spurt
between 8-10 y, when social maturity (full reproductive capacity) is attained. Mean CBL was 19.4% of
SBL in yearlings; 15.5% in subadults, and 13.7% in adults. Apart from the dentition, all variables of the
facial skeleton followed a somatic growth trajectory. Most variables expressed positive allometry relative
to CBL, with greatest growth occurring in the vertical part of the mandible. Mastoid breadth, and gnathion
to middle of occipital crest, expressed a strong secondary growth spurt at 10 y. Breadth of brain case, and
basion to bend of pterygoid, followed a neural growth trajectory, scaling with negative slope relative to
CBL. Sutures of the brain case (1.e., basioccipito-basisphenoid, occipito-parietal, interparietal and coronal)
closed before those of the facial skeleton. Condylobasal length was found to be a ‘rough indicator’ of SBL
and age group (adult, subadult), but not of absolute age. Suture age was not a good indicator of absolute
age or age group. A comparison is finally made between skull data on the South African fur seal (4. pusil/us
pusillus) with available data on the Australian fur seal (A. pusillus doriferus).

Manuscript received 12 November 2007, accepted for publication 6 February 2008.

KEY WORDS: allometry,Arctocephalus pusillus doriferus, Arctocephalus pusillus pusillus, Australian fur
seal, Otartidae, skull growth, skull morphology, South African fur seal.

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

INTRODUCTION

Skull remains are the most commonly found
material of marine mammals and the most likely
to be kept in natural history collections. It would
be useful to be able to gain as much information
as possible about sex, age, probable size, breeding
status and even in many cases positive identification
of such material in terms of modern taxonomy and
nomenclature (Brunner 2003). The South African
or Cape fur seal (Arctocephalus pusillus pusillus)
occurs on the Namibian and South African coasts and
nearby offshore islands (Schaffer, 1958; King, 1983;
Warneke and Shaughnessy, 1985) but does not occur
on Subantarctic Islands between Africa and Australia.
The South African fur seal and the Australian fur seal
(Arctocephalus pusillus doriferus) are now regarded
as closely related varieties of the same species
(Arctocephalus pusillus) (King 1972; King 1983;
Wynen et al., 2001; Brunner et al., 2002; Brunner
2003). Historically, the Australian fur seal was found
on the southern Australian coast from Kangaroo
Island (South Australia) to Seal Rocks (mid coast
NSW) with its distribution centred on Bass Strait
and Tasmania (King 1969). The identity of the fur
seals that originally inhabited Macquarie Island until
wiped out in the early 19" century is uncertain. Today
breeding colonies are more or less restricted to islands
of the Bass Strait region and Tasmania (Kirkwood et
al., 1992; Arnould, and Warneke, 2002; Shaughnessy
et al., 2002; Arnould et al., 2003).

It is useful to as fully as possible investigate
morphometric measurements of seal skulls to
correlate with age and maturity and breeding status.
Earlier cranial growth studies in pinnipeds were
based on unreliable age determination techniques,
including: (1) the extent of closure of cranial sutures;
(11) body length, colour of vibrissae, pelage and
general appearance; (ii1) ovarian structure; and (iv)
baculum development (e.g. Doutt, 1942; Rand,
1949a, b, 1950, 1956; King 1969; King 1983;
Brunner 1998ab; Brunner et al., 2002; Brunner et
al., 2004; Daneri et al., 2005). A common feature
of most of these studies is the limited number of
specimens available and the unknown age of most
of the material. Therefore, observed growth patterns
could not be quantified with any real meaning (King,
1972); generally a division into mature and immature
animals was all that was possible (King 1969; King
1983). Fortunately a reasonably precise method of
age determination of untagged South African fur
seals has been established from incremental growth
lines in the pulp of teeth (Scheffer, 1950; Laws, 1953;
McCann, 1993; Oosthuizen, 1997) but this involves
destructive sampling of material which might not be
negotiable on museum material. Dentition was also
used for aging 69 male and 163 female Australian fur
seals collected from an island in Bass Strait in 1970-
1972, but unfortunately the skulls were not measured
as part of the study (Armould and Warneke 2002).

Within the Otariidae, information on cranial

and mandibular growth based on animals aged from
tooth structure, or on animals of known-age (i.e.,
animals tagged or branded as pups), is only available
for a small number of species including Callorhinus
ursinus, northern fur seal (Scheffer and Wilke, 1953);
Zalophus californianus, California sea lion (Orr et
al., 1970); and Ewmetopias jubatus, northern (Steller)
sea lion (Fiscus, 1961; Winship, Trites and Calkins
2001). Sometimes very few skull variables have
been recorded. Currently, there is limited information
on cranial growth according to age (y) in southern
hemisphere fur seals: the main problem being small
sample sizes (King 1969; King 1983; Brunner 1998ab;
Brunner et al., 2002; Brunner et al., 2004; Daneri et
al., 2005), particularly the small numbers of tagged
individuals of known age that are available.

In the mammalian skull, there are two growth
models, neural and somatic, each with two types of
growth, monophasic and biphasic (Toddand Schweiter,
1933; Scott, 1951; Moore, 1981; Sirianni and
Swindler, 1985). In neural growth, skull components
associated with the nervous system (i1.e., braincase,
orbital and otic capsules) grow rapidly during
prenatal and early postnatal life, completing most of
their growth well before the rest of the body (Moore,
1981). In somatic growth, all other skull components
(1.e., facial skeleton) follow a more protracted growth
course (Moore, 1981). After the initial growth spurt
experienced during early development, growth may
be reasonably constant (monophasic growth), or there
may be a secondary growth spurt in older animals
(biphasic growth) when they reach sexual maturity.
Brunner (1998a) and Brunner et al. (2004) drew the
overall conclusion that growth patterns in fur seal
skulls were similar to that found in other carnivores
such as canids (dogs) (Wayne, 1986; Morey, 1990;
Evans, 1993) and other marine mammals (Bryden,
ISD).

Current populations of South African fur seals
number more than 2 million individuals although they
reached a low level of about 100,000 early in the 20 th
century. Their larger populations, occasional culling,
drowning in fishing nets and shooting of “problem”
animals, have made more specimens available for
study than their Australian relatives (4. pusillus
doriferus). Furthermore, another consequence of
the much smaller populations of Australian fur seals
(about 35,000 — 60,000: Kirkwood et al., 1992; 67,000
— 200,000, Shaughnessy et al., 2002; Arnould and
Warneke 2002; Arnould et al., 2003) and more limited
accessibility is that very little cranial morphometric
data are available on Australian fur seals (King 1969;
Brunner 1998ab), particularly of definitively known
ages based on tagged individuals (Brunner 1998ab,

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Arnould and Warneke 2002).

Here we examine the skulls of 83 male South
African fur seals, Arctocephalus pusillus pusillus,
from southern Africa. Specific objectives were to:
(1) describe the general morphology of the skull;
(11) quantify growth of skull measurements (” = 32
variables) relative to standard body length (n = 74
animals), condylobasal length (n = 83 animals) and
chronological age (7 = 63 animals); and (11) determine
if condylobasal length and suture closure are useful
indicators of age and/or standard body length. This
is a very large data set compared to recent studies on
the Antarctic fur seal (Arctocephalus gazella) and
Southern fur seal (A. australis) (Daneri et al., 2005),
the Australian fur seal (A. pusillus doriferus) (King
1969; Brunner 1998ab) and the New Zealand fur seal
(A. australis forsteri: King 1969; Brunner 1998ab)
and the recent review of cranial ontogeny of otariid
seals by Brunner et al. (2004).

A limited comparison is also made between the
available data on the South African fur seal from the
present study with published material from King
(1969) and from Brunner (1998ab, 2000) on Australian
fur seals. Modern multivariate morphometric analyses
of skull parameters complete data sets rather than
just means and variances of variables need to be
available.

MATERIALS AND METHODS

Collection of specimens

South African fur seals were collected along
the Eastern Cape coast of South Africa between
Plettenberg Bay (34° 03’S, 23° 24’E) and East
London (33° 03’S, 27° 54’E), from August 1978
to December 1995, and accessioned at the Port
Elizabeth Museum (PEM). The circumstances under
which most specimens were obtained are listed in
Appendix 1. Apart from specimens collected before
May 1992 (n = 16), all specimens were collected by
the first author. One animal (PEM2238) was collected
NE of the study area, at Durban. From this collection,
skulls from 59 males were selected for examination
(Appendix 1).

Thesamplewas supplemented withmeasurements
from 24 skulls collected by staff from Marine and
Coastal Management (MCM), Cape Town, South
Africa. These skulls were from males that had been
tagged as pups, and were therefore of known-age (1—
12 y). They were collected from the west coast, south
west coast, south coast and the Eastern Cape Province
of South Africa, between February 1984 and July

Proc. Linn. Soc. N.S.W., 129, 2008

1997. The date of collection, method of collection
and approximate location of specimens are listed in
Appendix 1. MCM seal specimens are accessioned as
AP followed by a number in Appendix 1.

East Coast and West Coast Animals

Additional skulls from Sinclair Island (West
coast of southern Africa, 27° 40’S, 15° 31’E) were
measured (condylobasal length only) to determine if
Eastern Cape seals (n = 28 males) were of similar size
to those inhabiting west coast waters (7 = 12 males).
PEM animals were adults 7— = 12 y. West coast
animals were adults of undocumented-age. West
coast animals were collected by Dr R. W. Rand in
the 1940s and housed in the South African Museum,
Cape Town. These skulls were divided into sub-adult
or adult classes based upon size and suture criteria
(see below).

Preparation and measurement of skulls

Skulls were defleshed and macerated in water
for 2-3 months. Water was changed regularly. Skulls
were then gently washed in mild detergent (or brushed
with water), and air dried at room temperature. A
small number of skulls were defleshed and gently
boiled. Dry specimens were measured ( = 32 linear
measurements) to the nearest 0.5 mm using a vernier
calliper (Table 1, Figure 1).

Variables used largely correspond to those
reported in earlier otariid studies (Sivertsen, 1954;
Orr et al., 1970; Repenning et al., 1971; Kerley
and Robinson, 1987; Daneri et al., 2005; Brunner
et al., 2004; Brunner et al., 2004) and in particular
the studies of Brunner (1998ab) focusing on the
Australian fur seal (A. pusillus doriferus). Care was
taken to measure standard parameters, measured
in the same way as described in previous studies.
Variables were grouped by region in an attempt to
reflect a functional cranial analysis and to assess
overall skull size (Hartwig, 1993) (Table 1). All PEM
measurements (and measurements taken form Sinclair
Island skulls) were recorded by the first author. The
majority of MCM measurements were recorded by
the third author (M.A. Meyer).

Suture index

Eleven cranial sutures (Figure 2) from 48 skulls
were examined and assigned a value 1-4, according to
the degree of closure (1 = suture fully open; 2 = open;
3 = suture half closed; 4 = suture closed), according
suture index (SI), ranging from 11 (all fully open) to
44 (all sutures closed). These values were added to
give a total suture index (SI), ranging from 11 (all
fully open) to 44 (all sutures closed). The suture

209

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

Table 1: Linear skull measurements (n = 32) taken from male South African fur seals in this study. Meas-
urements illustrated in Figure 1. Note that L26 and L27 were difficult to measure accurately. Param-
eters are broadly associated with the following functions; A- articulation, BC — braincase, F — feeding,
R/V — respiration/vocalisation.

Code Variable Region of Skull Function

Condylobasal length (posterior point on the occipital
condyles to the most anterior point on the premaxilla)

Gnathion to middle of occipital crest

Gnathion to posterior end of nasals splanchnocranium
Greatest width of anterior nares (distance between the
. L nasal R/V
anterior margins of the nares)
Greatest length of nasals (distance between the anterior
g : nasal R/V
and posterior margins of nasals)

Breadth at preorbital processes

Ss)
S

D3
D4
D5

Least interorbital constriction frontal

Greatest breadth at supraorbital processes frontal F

Breadth of brain case (at the coronal suture, anterior to ;
; neurocranium BC
the zygomatic arches)
Palatal

Palatal notch to incisors (posterior margin of first
Bag incisor alveolus to palatal notch, excluding cleft) RE

4

Length of upper postcanine row (anterior margin of

Pll postcanine one alveolus to the most posterior margin of | palate (dentition) F
postcanine six alveolus)

P12 Greatest bicanine breadth R/V, F

P13 Gnathion to posterior end of maxilla (palatal) R/V, F

P14 Breadth of zygomatic root of maxilla zygomatic arch F

PIS Breadth of palate at postcanine | (excluding the cele F
alveoli)

P16 Breadth of palate at postcanine 3 (excluding the aril F
alveolt)

P17 Breadth of palate at postcanine 5 (excluding the saline F
alveol1)

P18 Gnathion to posterior border of postglenoid process R/V, F

P19 Bizygomatic breadth (maximum distance between the 1 Dee
lateral surfaces of the zygomatic arches) Ye

es
P20 Basion to zygomatic root (anterior)

P2| Calvarial breadth (greatest transverse width across of pat ee i
; the skull base, anterior to the mastoid)
Mastoid breadth (width across the processes) A,F

P23 Basion to bend of pterygoid (anterior of basion to een BC, A, F
anterior of pterygoid)

210 Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Table 1 (continued)

face length)

Height of sagittal crest

Mandibular

postcanine five alveolus)

of post- canine five alveolus)

M31
M32 process to the top of the angularis)

numbering system and the method of judging degree
of closure and calculation of the suture index follows
Moore (1981). The suture index has been frequently
and successfully used as a criterion for separating
immature and mature skulls of mammals (Moore
1981), including seals (eg. Rand 1949; King 1969;
Orr et al., 1970; Bryden, 1972; King 1972; Brunner
1998ab; Brunner et al., 2004; Daneri et al., 2005).

Age determination

The age of animals was estimated from counts
of growth layer groups (GLGs) observed in the
dentine of thin tooth sections (Scheffer, 1950). Upper
canines were sectioned longitudinally using a circular
diamond saw. Sections were ground down to 280-320
um, dehydrated, embedded in resin and viewed using
a stereomicroscope in polarised light (Oosthuizen,
1997). Each section was read by one individual
five times, without knowledge of which animal was
being examined (repeated blind counts). Ages were
rounded off to the nearest birth date. The median date
of birth was assumed to be 1 December (Arnould
and Warneke 2002; Shaughnessy et al., 2002). The
median of the five readings was used too estimate
age. Outliers were discarded as reading errors.

The total number of aged animals (tagged known-

Proc. Linn. Soc. N.S.W., 129, 2008

Variable Region of skull
oo | a ee

Gnathion to anterior of foramen infraorbital (= lateral

Gnathion to posterior border of preorbital process i
Height of skull at base of mastoid (excluding crest) Ao. SS ae
Eom

Length of mandible (posterior margin of condyle to
anterior margin of the first incisor alveolus)

Length of mandibular tooth row (anterior margin of the
first incisor alveolus to the most posterior margin of

Length of lower postcanine row (anterior margin of
post- canine one alveolus to the most posterior margin

Height of mandible at meatus (dorsal margin of
coronoid process to the base of the angularis)

Angularis to coronoideus (dorsal margin of coronoid

Function

splanchnocranium

splanchnocranium

neurocranium

BC, F
F

mandible

mandible (dentition)

mandible (dentition)

mandible

mandible

age animals and canine aged animals) was 63. All
MCM skulls (7 = 24) were of known-age based on
tagging. Of the 59 PEM animals in the study: (i) 28
were aged from counts of incremental lines observed
in the dentine of upper canines as described in Schaffer
(1950), i.e., range 1-10 y; (ii) 11 were identified as
adults > 12 y(1.e., pulp cavity of the upper canine
closed); and (111) 20 could not be aged because of
missing or decayed teeth. In South African fur seals,
animals > 13 y can not be aged from counts of growth
layer groups in the dentine of upper canines because
the pulp cavity closes at that age which terminates
tooth growth, hence the age group “=12 y or 12° y)’.
Dentition has also been used to age Australian fur
seals (Armmould and Warneke 2002) who claim that it
is possible to age some male Australian fur seals to 17
years old based on upper canine dentition.

For this study, the following age groups were
used: yearling (10 mo to 1 y 6 mo); subadult (1 y 7
months to 7 y 6 months); and adult (= 7 y 7 months)
(Table 2). Unfortunately, no South African fur seals
skulls were available from wild-tagged individuals
with definitive ages greater than 12 y. The estimated
longevity of South African fur seal males based upon
zoo animals is c. 20 y so it appears that they have
a similar potential lifespan to the Australian variety

Zi

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

ee Sr r—“—

i
t
i
i
j
1
1
1
1
1
1
1
1
!
1
1
1
1
1
1
ui
1

1
1
4
'
1
1
1
1
1
1
1
1
1
1
1
1
1
1
J
1
1
1
1
1
1
u
1
L

Figure 1: Diagram of a South African fur seal skull (PEM554) indicating individual measurements

taken. Measurements are defined in Table 1.

(Armould and Warneke, 2002).

Currently, examination of canine tooth structure
is the most precise method of age determination in
untagged pinnipeds; however, pulp cavity ring counts
are not without error. In principle, other seal teeth

DAD

such as postcanines, can be aged using growth ring
counts in the pulp cavity up to the point where the
pulp cavity closes up. For recent assessments of the
reliability of this method see Arnbom et al., (1992),
McCann (1993) and Oosthuizen (1997).

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

ll

Figure 2: Diagram of a South African fur seal skull (PEM554) showing the position of sutures examined
in this study. 1. Occipito-parietal; 2. interparietal; 3. coronal; 4. interfrontal; 5. internasal; 6. premaxil-
lary-maxillary; 7. basioccipito-basisphenoid; 8. basisphenoid-presphenoid; 9. squamosal-parietal; 10.

squamosal-jugal; 11. maxillary.

Classification of growth patterns

In the present study, neural and somatic growth
patterns were distinguished as follows: [(mean
skull measurement for adults > 12 y — mean skull
measurement for subadults at 7 y) x 100%]/(mean

Table 2: The age distribution of Cape Fur Seals

a group “ (y)

Proc. Linn. Soc. N.S.W., 129, 2008

skull measurement for subadults at 7 y). Where the
percent increase in variable size was < 6%, growth
was classified as neural, i.e., most growth was
completed as subadults. Where percent increase was
> 6%, growth was classified as somatic, i.e., growth
continued to increase in adults. Percentage increase
for each variable is given in Table 3.

Pages 214-218 comprise Table 3

Summary statistics for dorsal (D1-9),
palatal (P10—23), lateral (L24—27) and
mandibular (M28-—32) skull variables
according to age (y) and age group.
Data presented as mean skull variable
in mm + S.E., followed by coefficient of
variation in round brackets, and skull
variable expressed as a percentage of
skull length. Maximum value of each
variable (males of unknown-age), and
classification of growth pattern, are
also presented.

JAS)

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

%0 ‘TeINON %O] “OneUog %~] OneUIosg %6 “OBWIOg %L] OeUloS %Vy] oneUosS Ap] OWeWoS Mp] “OeUlog | 8 ‘news | uJoyyed YIMOIDH
: 5 > 5 : : oyeur
[el Ov FECL

Wd (007
[eloorzss ETES CO SOF BE | oveee 6TFTOS T1¥90€ Ly Lo 7'6 F 8'PET OL FT OST S sojetu 10} weap

N
S
%9'EE %7' EC (Y'CI) %7 ST (96) %6 LC (8) %8' st (101) %6 TT %0' LE (L's)
lv = 6 6r7

(7S) € 1 F0'¥8 17¥O8S | loll7i+ose |] lollst+169 VIFOLr | (06) 80+96c | (18) 7716
%7L8 (19) (€v)

lo1lS€ +6017 STHEL IV

%1 06 (EL)
[11] o's + €'S7

%S SI (BL) %8L7 (TL) %S LI (ZT) ¢ 11 (16) %6°CE (69)
[stlso+ozLe | lorlereclo | (stltistc7w | lo1l9o+s2z | lolly 1 +z ss

%0' Ez (9°9)
[ellol +L ss

AYE
($7) 60+ 0'r8

%L 88 (19) (tp)
19 +1007 LY ¥FT8be

%1°S8 (1S) (ST)
Ip] es #5 °907 OL FT CHC

“TE? MEST (9'€) %/ 17 (E01) %6 LI (L'01) “wT %7OE
(SLI6T#FELS [p] LOF9'8E TEFL BI TTFEPY | (18)01F 062 | (€9)57+ 668

%HGEEE
(€7) 60478

ME VE %Z ET (9'8) %8ST (66) %LLT (Ep) %/L91 (66) %O TI 8 re (Ob)
(O'€) ECL FTES le] 87+ 95 [pl61T#HY8E Ip] ST #COL9 [plOT#ES OF | (6) 71 F787 [rp] 61 FE +8

a

%yL8 (WL) (0's)
95 +6907 Sv 8 9ET

LCE %8'7 (81) y's] %ELT %9'L (6'6) %7 TT (L'8) MEE
(€9) 0'7 = 9'¥8 [slvoF9ES | (S8)71FS9E | (0'9)51 +099 [9] LT +9 Ir [9] 604592 | (CL) ETFI'ES

%6'06 (06) (€8)
[1€] 67 FO'E81 TEFLSIZ

YL ve (VID)
[Te] ST FEL

%6 Ol (01)
[lel pO+F VET

%ITLT (O'El)
[0] 60 OLE

HERE %1 7 (OCI) ME HI (8°71) %6 97 (€ IT)
G)90FL78 | Lloclol+oLr LO+8'0€ T1+0'8S

%> 58 (7) (7 €)
[11] 61 + 8°61 17 ¥9'0€7

Hy OE %L CC (LL) YL HI %9 LT py LI %B OI (8°L) MECE
(ee) SOF TPs | [171 +S 7 | 9) LOFGEEE | OS)OT+9ED | (15)90FT0r | [11]90F0S7 |] (9)ST FETS

%0'S8 (Lp) (1'€)
PVFOL8I VE +9 0CC

HELE XO IYI VLC %L LI (6 ED) %L OI (671) %0'SE (91)
OMLIFLES | OS)ETFES | OH) LOFTIE | O1)S0F66S LTFO0G6E S1EOET [Ee] LOFV9L

a]
1H
<
Be
oD
cfs
BE
Be
cl iE
v 9

%9'r8 (L'€) (0p)
VEFS EST EV FOLIZ

HELE YC YE YI AY %VLI Ml %6 YE
GE)XLTFETCS | (8OTELOV | (TL) TIFOTE | OB)ETFELS | GH SOFTLE | WH)SOFTHE | (BH) SIFLSL

(79)
Ly +FTS0C

%9' 8 (9'9)
OV F9ELI

%86E %7 CC (701) Ay HI %9'9C %8OT (b'6) %lIT LE
(L€) VIFL 18 91*=9St | (08) 8045S 6c | (49) €1 F9'HS lS) €LFIre | (Cr) €O0F87 | TL) SITFIIL

“OI ileee %9 TI YRC %G OT %O TT AT EE %0'€8 (ZT) (€1)
(89) 07#LI8 (15) 60FL4 | (LO)EOF98r | (L8)GIFIEE | OOOLFSIZ | (ry) 614619 618° C91 SIFT 9IGI

ES %L IC %9 TI %8'SC (811) %L't1 %E OI WC CE %CS8 (00) (8's) NAA
(TOCTELBL | (TL)6LFSLE | 8) E1 FI SEFOSH | (6E)LOFLSZ | (CE) POFELI | (9)9TE EDS LCL 8b1 CLESYLI

%8 8V %6 EC %S El %9'ST %L‘SI %L TI %O EE %L ‘06 (EC) (€@) 3
Bet. : a As : ee ; baa ; aac at d ret : BR ; : , j G I UUTALE).
QO) POFOSL | (SI)POFTS8E | OLDEOF9IZ | (CO) TOFS Or | (88)9T FIST | (LO) 1O0FEO? | (DOI FI +S VCFS8 PHI TFL 6SI

i ee ee

Proc. Linn. Soc. N.S.W., 129, 2008

214

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

O*PET

7x9] 90S)

8 OBULOS

O} (0) B JO 9OUSPIAT

%6 OeVUIOg

PIL+68 CTEL OLELLI Let LE
pS bs

%8 TT (8) %~O TT (€ 11) %89 %L Op (89)

{i1lgso+o0e | Litlso+osc | (LL) r0+F7LI VCFVS8IL

%8°O1 (9°01) %6 6 (€ 01) %19 (ECT) %E Ly (Lb)

Istlzo+79c | l[rllLo+ 1 v2 PvOFS rl | [IlEl+rrII

My I~
(89) TL +1s

(9°8) 9° 1 + 0°79

%0' Cr (€9)
OTF E9IOI

%0'0C
(68) 01+ 8r

%6 01 (871) %9 OT (18) %*r'9 (6 F1) %~ Ly (9'€)

%9'0¢ (C1)
STFIIS

(£9) 60+ 68S

%v VC

%6 €C

(Lb) 71 = £66

%I Ev (69)
SIFT HOL

%O' rr (9'L)
LEE 601

SL FOLT [pl V1 #497 TI +091 61LFTLIL
%E OL (TL) %E6 (5 9) %6'S (011) %E Ly (Ss 1)
[pr] 60 +057 [pr] L0+ 9°77 LOFYVHI [rp] 60 +0'SII

%7 61
(L€) 80 +99b

%T TT (€8) %6 6 (3'8) %I9 %E Ly (0'9)
[9] 60+ £97 lo] 80+ Ee | (L6)S0Fr FI 97TFOTII

%7 OC
(TLE LEO LY

(78) 8TFLLS

%6 VC

(vb) 71 #709

Yv VCE

%9' Ev (ST)
L0+¥9°S0I

Yl Cr
(99) $7 +966

%9'OT (S 11) %L 8 (C91) %79 (901) 6 9r (901)
loq]so+sze | ([zz7l90+9°81 €OFV El 61 E101

%LO1 (Sv) %6'8 (0:91) %0'°9 (701) %0' 8h (0'r)

[8] p0 = 0'SZ [gs] 71 = L07 v0OFS8 El €1+9 011
% OI %6'8 ia) %0'Lb (0'¢)

(19) 90FVEC | (Cr) VO+F96I | (88)90FI 41 VIF6E0I

%8'OI (6L) %*%r'6 (3°6) %L9 %0' Lp (79)
[e€] 1 #8°€Z% lel71+L0c | ($6) L0FS 41 TE FS IO!

%9'OI
(TL) 90 +617

y's %79 %6 SY
(LL)SOFTLI | OL)EOFLTI | (18) L776

%S 61 (€ OD)
[1el80+6 Ir

%8'6I (€'€)

YI SE

(€6) 60+ 7S

%8 VC

[11]S0F6 Sr | (Sb) 80FTLS

%9 61

YL VC

(VE) LOFOED | (€8)80F EHS

%9 61

%0 SC

(19 €1l#Str | (6) PTF THS

%E Ol

YE VCE

(SS) 80496E | (SC) P08 6r

VE 8 HED %S Sv

[1] 9°61 (LDTOFES | WE)EOFHTZI | (TO) TOF E68

%9'6 My L %>9 %o by
(COCOFLII | WE)VOFOE! | EVEOFETI | (TOTIFLLL

%e Cr (901)
[Ts] L1 +606

2S
=
=)
us)
<x

iratG
(0°99) L 1 = 8°L6

%E CH
(LO VII ¥6

%9 Tr (vb)
lel ©7=9'88

%8 Cy
(18) S7#6L8

%6 LI

@E)SOFTSE | (DL

%9' ST
(89) 91 #S7ZE

isan %B's %1'8 YL Sr
(SESOFPSI | (9) LOFI HI | (SS5)S0FETI | OL) S0FOEL

Ml?
(8b) TL #8 EE

%9' 1

O+F1 79

YL VC

(80) €0* 7 Et

%0 Cr
(8b) 87 ET

%6 LE
(79) 0 € #799

%7 Ov
(VI) LOFEP9

\© >

ynpeqns
| sea]

INS

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

aaa
Hs “enon YL] “OeWog

SI+Crs

19

%B' TE (5°)
[111604 +08

%0'7E (8'p)
[91] 60+9LL

ML TE
(€L)9TF9OBL

%G TE (SZ)

Cv +L 6€1

MESS (0'S)
[11] 77 = v'6El

%6 TS (69)
lot) 77F9LTI

*I Ss (yl)
VVFEVEL

%GE 1S (S'S)

(suons) A OT

eval

MET IIeulos

%OI “ewuog

yest

LEFT OCI

6 +8 PLT

19

6881

%OT ‘OHeUIOg

%EL is aE

al 8 4

0c +0 OFT

%8 Ly (Ev)
91 +O ICI

%589 (L'€)
S1FOELI

%8 Op (8p)
VI+FOEII

%9 Or (L'S)
6TFSSII

%E Op (0'€)

[plOT#L9OL | [pl Seosz SLFOCII
ALTE %S TS (8's) py Lp (v's)

(Vb) CLEP LL LTFV VE CTFSIII
AL EE %EOS (E11) %E Op

($9) 804L7L VZFL801 | (86) L1 +666
%OEE %L TS (1S) %9 Or (Ep)

(TE) LOFTIL

SIF Ol

elL+s Lol

%7'89 (8b)
[stIloc+7 S91

%L'89 (1'€)
Ip] LTE ELI

%0'89 (01)
L0+9 P91

%0'89 (9's)
lol v€ FO 191

%7 89 (78)
L7#TLHPI

%E89 (0'€)
iP CWSI

%MEOS (7 €)

%6'6S
(Ly)V 79051

€9+0Z6l

%'9L pean
[it] 17+ S61

OTS 9E

ea
($9) 9048°SE

%NLS (7S)
(

%TLS (9b)
6TF8 Ib

%9'9¢ (67)
S1FULEL
%TLS (79)
CEFISEL
%S HS (8'6)
OTFOLII

%O HS
(LS)V CFL STI

%S rs (0'9)
TEFO ITI

%7 96 (6S)
EFC ITI

%O'EE | %S Or (8'0) %0'89 (6 1)
(C1) 708 EL IS III 70 +901 €LFTOSI
Wy ye %9'0S (€9) %6 Lr (tb) p39 (S'€)
($7) 604 9'PL C€+F8601 €CTFO POI OTF E SHI
YO'VE %7 6b (6'L) %9' Sr % 1°89 ($9)
(0S) 71 = 8°69 87F6OOI | (HL) 17+ 9'E6 TE FSC
KG EE %B Ly %S bh %8'L9 (+0)
(PDL0FS99 | O1ILO0F6E6 | OI)LOFELB 0 + O'€E1
NESE YO LY %9 by %6'89 (19)
(76) Lv F971 (HS)TEFOTS | CDSTESLL LS+00CI
MELE %6 OF br Sp %¢'89 (8°)
(072) 6096S | OP ITTE8HL | (LEVOTFELL TTF EOI
€7d @Zd Iz

%Is (39)
LTFO TI

%6'TS (CC)
L1I+8 0!

%T ES
(S1) OF =9'C6

CHS
(80) $0 +998

SDS & Jo souspiag

jSOSIe] UI ON[VA “xe

Z SOLU IO UBD

VE+FO 681

%6 SL (17)
[plOT FE PSI
py CL (€9)
TV FS B8LI

%B Eel (€ O01)
[Sz] OE F HSI

%6 VL (€'€)
[Ol] 81 +TELI

%T rl (17)
[pl L179

%T VL (0'S)
[el L479

%T EL (8'L)
1y+00S1

%6 IL (€ 1)
€1+O0I+1

%7OL (6'€)
PEF CC

%9'°69 (€'0)
€OFTIIM

%8°CL (7S) %py El (€6) wi Or
[oI] pZFTESI lol] 80497 =8
5 ; ek ;

%C OL (1h) “II (6 1) ¢ ay

Ir] €0F LSE

asrall
(1'8)1' 1 = 8°67
%8 ET
(CL) 60+ 97E

%O'ET (El)
[Te] LOFT87

%7 EI (9'L)
[11] 20+ S5°0¢

(16) 71 #767

%6 EI (ZIT)
LI+FV0€

%L TI (€ CI)
V1 #097

(8°8) 91 #0'SZ

(9b) LOFL0C

%0' ET
(€0) 10 +802

Proc. Linn. Soc. N.S.W., 129, 2008

216

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

(7x9} 998) (yeom) A OT SDS 2 JO s0uEpIAg

%TB Wjeurosg %ET OYLULOS %G] oOeulog WE] Wjeuros ulo}ed YyMoID

La

Q[LUI JSOBIL] UI ONTeA “xey|

£88

ZEF 068 674308 | M007 S SoTPUI 10y Ua]

CN FESO)

Sr ITI

©9

MEE (vy y)I 1 F9'S8 %G OE OPTI FC8L cl<
ae
YE7TE(TS) (91 OLFV6L | %66C(S br) 80F HCL

l6l80478 | LITIOL#S SII

[ZI] 804TL 87FS EOL

E16 89+0 801 YlEE (S'S) OT FIT %8' 6c (€'S) L 1 FOL

_—

[gE] 91 +09 €S+#8 101 MOTE(CE PIE TFS BL | %I6T(TCLOFS IL

Ir] LOFTS €€+F1 001 YECTE(WS)STFEOLL | “KEOE(TS) VIF9IL Wnpy

NX La
(39) i va) Va) =

I
N

[p71 90 = 9'1 [17] 81 +298 AE TE(SIIYIF689 | %S 67 (871) I +9'E9

[SIZT1 +S 7 l6l SL #16 ACTE(SHOTLFOSL | “WIOES) TIL F169

Ir] S0FL0 [e] 804598 AOCE(EHVIF6IL | %86C(79) BIS H9

[1] 00 604 ECL YE6T(SHILIFTCS | %Cve~ (06) LTCC

eat go [i] 1€z HEOE(TIPFOFE Sr | KISTOL)STEL SH

LTI 971 scl Zant

nN
—

le] 17 +098 METE(TS)SIFESD | %167(6E) TI FI'E9

Ip] 9OFOLL YWETE(TIN STEED | NOTH CLD LTFV 19

[o]— AYLOE(LOEOFIT6S | %l67(68)9EFILS

ict)
<
a=
=
5
>

217

Proc. Linn. Soc. N.S.W., 129, 2008

09 19 6S 9¢ c9

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

%9'¥¢ (901)
61+LI19

(s6) [91]
C1 O'S

%0'SZ (FET)
VTFLIO

%0 €¢ (FIT)
S1+9'SS

py Ec (O'EI)
[rp] SEF OBS

%0'7 (89'S)
VLFCES
%0'€7 (76)
61 Fr bS

6) WY

0.
(Let) lie]
LLEL bp

%O' re (CHI)
8'€ S65

%~ CC
(S3)17#LES

%6 CC (S56)
61+ CPS

%9°81 (6'L)
Lit] 11 +994

%E 61
(Tr) S0 +994
%9' 81
(LS) TL #€ 9p
%E6I
(€€) L0+9 9b

%L'61
(+) 80FL9r

%/ 61 (9°ST)
[te] 71 +9

%9'6T (0°8)
log] 90 +S tr

%E IT (E9)
60+ £ 6b

%7~ 17 ($9)
CLEC

%8' 07 (59)
60FT8L

%py OC (LL)
SI4C bb

%5 61
(60 r0OFTSY

%9 61
(6€) LOFS EP

%BL7 (8)
[s]07#6L9

%/ 87 (6'€)
lol] 20+ 169

%p 8 (T'S)
[rp] 81169

WAG

(61) 90119
%9'6C

(L'€) 01 =0'0L
%6'8z (TOT)
foe] 11 #179
% (67

(TE) 90FSL9

%9'LZ (9b)
[ely 1FE19

%7 OL
(LL) 6€ FVSLI

%> OL (9's)
loll y7FVILI

ETL
(91) 09+ 0ZLI

%/.69
(V1) V1 #8891

%S OL (1's)
lo] S€ +1891

%9'LO
(SOD LTE 9FI

%L89
(L'€) L 1 =9'6S1

%0'89
(670) 61 = L6¢1

%T 17 (19)
VIFLSY

%7 61 (66)
VO#LIp

%9'61
(Or) 80S tr

%8°8z (62)
[elo1FE19

%/' 8 (6'€1)
[IL] 17 +S Or

%L8t (8°71)
[L161 FV'8E

%> 61 (L'9)
lo] TI FV 6E

%6' 87
(OIL) V7 # €'6S

%8'L9
($9) SvFCLYI

%5°99
(6L) SEFS9EL

%88I (99)
SILFVSE

%ILI (€6)
VTFISE

%E SI (09)
vl FOE

%L9T (8ST)
€€F 16S

%9'07
(L7) 8016

%O8T
(TE) TL FES

%TH9
(€1) 71 +0971

%9'07
(F'0) LOF6SE

%TZ 6
(8°72) 01 = 60S

%8'8zZ
(Sb) SL#8 94

%8'S9
(TID 16 F811

%6'C9
(10) 10 +686

ympeqng

SuI[AvO,

dnois
asV

Proc. Linn. Soc. N.S.W., 129, 2008

218

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Statistical analyses
Skull measurement error

For most PEM skulls, duplicate measurements
were taken of 7 randomly selected variables to assess
measurement error. The l-sample sign test was used
to test the null hypothesis that the true median was
equal to the hypothesised median. The Wilcoxon
sign-rank test requires the assumption that the parent
population is symmetric (Gibbons and Chakraborti,
1992, p. 155). However, the distribution of data was
not symmetric for all variables, thus the slightly less
powerful sign test was used. Inter-observer error was
not assessed but care was taken to follow standard
measurement protocols as described by previous
authors (Sivertsen, 1954; Orr et al., 1970; Repenning
et al., 1971; Kerley and Robinson, 1987; Brunner
1998ab).

Condylobasal length expressed in relation to SBL
Growth in condylobasal length (CBL), relative

to standard body length (SBL), was calculated as
follows, using paired samples only:

[CBL (mm)/SBL (mm)] x 100%

The SBL is defined as the length from the nose to
the tail in a straight line with the animal on its back.
As the approximate variance of the ratio estimate is
difficult to estimate, percentages must be interpreted
with caution (Cochran, 1977, p. 153).

Condylobasal length as an indicator of SBL and age
The degree of linear relationship between log

CBL, log SBL and age (y) was calculated using the
Spearman rank-order correlation coefficient. Linear
discriminant function analysis (Mahalanobis squared
distance) was used to predict the likelihood that an
individual seal will belong to a particular age group
(subadult, adult) using one independent variable,
skull length (see Stewardson, 2001 for further
details). Yearlings were not included in the analysis
because of the small number of yearling skulls
available, i.e., n = 2 yearlings.

Suture index as an indicator of age

The degree of linear relationship between suture
age and canine age (y) was calculated using the
Spearman rank-order correlation coefficient (Draper
and Smith, 1981). Linear discriminant function
analysis was used to differentiate between subadult
and adult skulls using one independent variable,
suture age.

Proc. Linn. Soc. N.S.W., 129, 2008

Bivariate allometric regression
The relationship between value of skull

measurement and: (1) SBL, (ii) CBL, (ii) age (y),
was investigated using the logarithmic (base e)
transformation of the allometric equation, y = ax’
which may equivalently be written as log y = log a
+ b log x. ‘Robust’ regression (Huber M-Regression)
was used to fit straight lines to the transformed data
(Draper and Smith, 1981). The degree of linear
relationship between the transformed variables was
calculated using the Spearman rank-order correlation
coefficient, + (Gibbons and Chakraborti, 1992).
Testing of model assumptions, and hypotheses about
the slope of the line, followed methods described
by Cochran (1977), Draper and Smith (1981) and
Gibbons and Chakraborti (1992).

Comparisons between South African and Australian
material

The South African fur seal data from
Stewardson (2001) were compared to published
material from King (1969) and from Brunner
(1998ab) and Brunner (2000) on Australian fur seals.
In the case of King (1969) the condylobasal length,
postorbital width and zygomatic width of male
skulls were read off graphs in her paper (accuracy
about + 1 mm). The data from King’s study was then
compared to similar data for South African fur seals
from the present study. Brunner published mean,
variance and number of measurements data for most
of the standard seal skull parameters set out in the
legend for Figure 1 but did not provide sets of the
raw data on individual South African or Australian
fur seal skulls in her papers (Brunner 1998ab) or her
PhD thesis (Brunner 2000). Multiple comparison t-
tests (Cochran 1977) could be used to compare the
means of South African and Australian material but
more sophisticated multivariate analyses were not
possible. Multivariate principal component analyses
such as those performed by Brunner et al. (2002) and
Daneri et al. (2005) would have required access to the
full data sets to determine the interdependence of the
standard skull parameters between individual skulls.
The large numbers of Australian fur seals collected
by Arnould and Warneke (2002) provides a large data
set on SBL vs. dentition-based-age for both males
and females. Unfortunately, the animals were killed
by shooting in the head and skull measurements were
not taken.

Statistical Analysis Software
Statistical analysis and graphics were
implemented in Minitab (Minitab Inc., State College,

219

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

1999, 12.23), Microsoft Excel 97 (Microsoft Corp.,
Seattle, 1997) and S-PLUS (MathSoft, Inc., Seattle,
1999, 5.1).

RESULTS

Skull measurement error

Of the 7 variables that were measured twice,
measurements were reproducible at the 5%
significance level (range for p-values: 1.0—0.08).
Heights of sagittal crest (L27), and heights of the
skull at base of mastoid (L26), were both difficult
to measure reproducibly. Hence the sagittal crest
(L27) and base of mastoid (L26) data needs to be
interpreted with caution. All errors are quoted as
standard deviations (SD) or standard errors (SE) as
appropriate.

Skull morphology

The youngest animals in the sample were 10
months of age (Figure 3a; Table 3). Inthese individuals,
the skull was c. 160+2.6 (SE, n=2) mm in length (D1)
and 87 + 0.5 (SE, n = 2) mm wide (P19). The brain
case was relatively long, measuring approximately
120 mm (75% of CBL) from the most posterior end of
nasals to the most posteriorly projecting point on the
occipital bone, in the mid-sagittal plane. There were
no signs of bony ridges or prominences. Relative to
CBL, the face and mandible were short (Figure 3a).
Milk dentition had not been completely replaced by
permanent teeth in animal AP4999. It was clear that
the deciduous canines persist in at least some animals
until their tenth month.

In adults 10 y of age, the skull was rugose, with
heavy bony deposits (Figure 3b; Table 3). Mean
length (D1) and breadth (P19) was 248 + 4.7 (SE,
n=5) mm and 142 + 2.9 (SE, n= 5) mm, respectively.
The braincase was approximately 157 mm in length
(or 63% of CBL), and a sagittal crest was always
present but varied greatly in height (range 4.4-12.0
mm) or 9.2 + 1.3 mm in height (SE, m = 5). The
forehead was convex at the supraorbital region.
Relative to CBL, the face was long, with long nasals
that flared anteriorly. The ratio of nasal breadth to
length was 1: 1.5. The palate was long, moderately
broad and arched. The ratio of palatal breadth (P15—
17) to palatal length (P10) was 1: 3-4. The maxillary
shelf at the root of the zygomatic process (P14) was
very short in an anterior-posterior direction (16 mm;
6% of CBL). The mandible was long with a broad
coronoid process. The tooth rows were parallel, with
enlarged third incisor; large canines; robust, tricuspid

220

postcanines (PC); anda slight diastema between upper
PC 5 and 6 (Repenning et al., 1971; present study).
Dental formula was (I 3/2 C 1/1 PC 6/5). In seals, the
premolars and molars are similar in appearance and
are collectively termed postcanines.

East Coast and West Coast animals

Available data suggested that skulls from adult
males, 7-12’ y, from Eastern Cape fur seals (mean
246.6 + 2.5 mm (SD); range 213.7—266.8 mm; n =
28) were significantly smaller than skulls from adult
animals inhabiting west coast waters (mean 259.4 +
4.5 mm (SD); range 225.6—282.1 mm; n = 12) (at the
5% significance level two sample t-test: t= -2.48; P=
0.024; df= 17).

However, skulls from adult males, > 12 y, from
the Eastern Cape (mean 255.7 + 2.6 mm (SD); range
239.9-266.6 mm; 7 = 11) were not significantly
smaller than skulls classed as from adult animals from
west coast waters (mean 259.4 + 4.5 mm (SD); range
225.6—282.1 mm; n= 12) (at the 5% significance level
for a two sample t-test: t= -0.71; P= 0.49; df= 17).

Condylobasal length (CBL) expressed in relation
to standard body length (SBL)

In the cases where a seal carcass was the source
of the skull material it was possible to measure
skull size (CBL) and relate it to the standard body
length (SBL) and both to age. Animals seem to cease
growth in length at 10 y. It was found that the relative
condylobasal length decreased with increasing SBL,
i.e., 19.4% (yearlings), 15.5% (subadults), 13.8%
(adults, 8-10 y) and 13.6% (adults > 12 y) (Table 4).

Condylobasal length as an indicator of age
Condylobasal length continued to increase
until at least 12 y, with no obvious growth spurt at
social maturity (8-10 y). In animals 1-10 y, growth
in skull length was highly positively correlated with
age (y) (r = 0.89, n = 51) (Figure 4b). However,
after fitting the straight-line model, the plot of the
residuals versus fitted values was examined, and the
straight-line model was found to be inadequate (the
residuals were not scattered randomly about zero, see
Weisberg, 1985, p. 23). Thus, CBL could not be used
as a reliable indicator of absolute age, particularly in
young and old animals (Figure 4b). The coefficient
of variation in skull length for young males 1—5 y
(12.3%) was considerably higher than in older males
(8-10 y, 4.3%; => 12 y, 5.7%) (Table 3) suggesting that
young males may grow at different rates but survivors
to old age fall into a narrower range of sizes. This
may reflect higher mortality of smaller juveniles.
Although CBL was not an accurate indicator of

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Figure 3a: Size and shape of the South African fur seal skull: juvenile 10 months (AP4999).

Proc. Linn. Soc. N.S.W., 129, 2008 221

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

Figure 3b: Size and shape of the South African fur seal skull: male adult 10 y (AP4992).

220) Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Table 4: Growth in mean condylobasal length relative to mean standard body length (SBL)

Age group

Yearling
Subadult

WSO. 7/ 2s 2
174.5 +7.2
WS D 22 Its
205.2 + 4.7

220.6 + 3.4

n Mean CBL (mm)

219.3+5.1 [3]

232.0 +2.5 [9]

CBL rel. to
SBL

18.5%

Mean SBL (cm)

82.5+2.5
94.5+4.5

121.0+3.0 16.2%
126.0+5.2 16.3%
141.0 + 3.8 [3] 15.6% [3]
149.0+1.7 14.8%
159.0 + 3.4 [9] 14.6% [9]

7

214.7 + 3.6 [28]

238.8 + 6.0 [5]

138.9+4.1 [28] 15.5% [28]
170.4 + 7.6 [5] 14.0% [5]

[eerie 5 242.7 + 2.0 [4] 170.8 + 2.3 [4] 14.2% [4]
ene 0] 00 5 248.2 44.7 187.4465 13.2%
Paes) © (8210 7 243.3 £2.8 [14] 176.6 + 4.0 [14] 13.8% [14]
a boo yual 12 12 250.4+4.5 [11] 183.7+£5.8 [11] 3.6% a1]
Total 63 55

absolute age, it was a ‘rough estimator’ of age group.
When skull length (CBL) is known, the following
linear discriminant functions can be used to categorise
each observation into one of two age groups—adult or
subadult:

Log(y,) = -98.43 + 0.91 x Log(CBL)
Log(y,) = -129.06 + 1.05 x Log(CBL)

where CBL is the skull length (mm); subscript 1
= subadult; and subscript 2 = adult. The seal is
classified into the age group associated with the linear
discriminant function which results in the minimum
value. Of the 61 observations in this study 85% could
be correctly classified using this method (Table 5).

Condylobasal length as an indicator of SBL

Skull length (CBL) was highly, positively
correlated with SBL (r = 0.93, n = 74; Figure 4a).
When CBL is known, the following equation (linear
least squares fit; log-transformed data) can be used as
an ‘estimator’ of SBL:

Log(y) = -4.11 + 1.69 x Log(CBL)

Proc. Linn. Soc. N.S.W., 129, 2008

which may equivalently be written as SBL = Cui

x CBL’” or 0.01641 x CBL'®, where the standard
error (SE) of the intercept is 0.28 and the SE of the
slope is 0.05 (n = 74).

Suture index as an indicator of age -

The sequence of suture closure according to
age (y) and age group is shown in Table 6. Sutures
i-xi showed signs of partial closure at different
times, and the time taken to reach full closure
was different for each suture. The cranial sutures
(basioccipito-basisphenoid, coronal, occipito-parietal
and interparietal) were the first to partially close.
The squamosal-jugal, squamosal-parietal, maxillary,
premaxillary-maxillary, and interfrontal were the
last to show signs of partial closure (order of partial
closure unknown), with the basisphenoid-presphenoid
and internasal remaining fully open in all specimens
examined.

The exact sequence of full closure forall 11 sutures
could not be established because animals of known-
age, => 12 y, were not available (The definitive ages
of such old individuals can only be determined from
tagging). However, the basioccipito-basisphenoid
and occipito-parietal were the first sutures to fully

223

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

close in animals 3 and 4 y, respectively; followed by
the interparietal in some animals > 7 y; and then the
coronal or squamosal-jugal in animals > 12 y.

In animals 1-10 y, suture age was highly
positively correlated with age (r= 0.81, p< 0.001, n=
38) (Figure 5c). However, after fitting the straight line
model, the plot of the residuals versus fitted values
was examined, and the straight line model was found
to be inadequate.

Furthermore, linear discriminant function
analysis could not be used satisfactorily to categorise
each observation into age groups, i.e., of the 46
animals examined, all subadults (7 = 26) were
correctly classified; however, seven (35%) of the 20
adults were incorrectly classified as subadults.

Suture age was highly positively correlated with
SBL (r = 0.89, p < 0.001, n = 63) (Figure Sa).
Further information on suture age as an indicator
of physiological maturity is presented elsewhere
(Stewardson 2001).

Bivariate allometric regression

Regression statistics for skull measurements
for sub-adults (yl) and adults (y2) are given in
Appendices 2, 3 & 4. The log-transformations
of the parameters (DI—M32) are regressed on
Log(SBL), Log(CBL) and Log(age) respectively.
Overall, correlation coefficients were moderately
to strongly positive, 1.e., most points on the scatter
plot approximate a straight line with positive slope,
r 20.70. Exceptions included breadth of brain case
(D9) on SBL, CBL and age (y) (r = 0.3-0.4); length
of upper PC row on age (y) (r = 0.59), and breadth of
zygomatic root of maxilla on age (y) (r= 0.57). SBL
was strongly positively correlated with age (y) (r =
0.87). Although correlation coefficients indicate that
linearity was reasonably well approximated for most
variables after log-log transformations of the data, a
linear relation (Log y vs. Log x) did not necessarily
best describe the relationship. A larger data set would
be needed to find an optimum relationship using more

complex models such as the Logistic growth curve,
which has an asymptotic maximum.

Growth of skull variables according to region

Most variables within a given region of a skull
were significantly positively correlated with each
other, r = 0.70 (Appendix 4.5). Exceptions were: (1)
breadth of palate at PC 5 (P17) with length of upper
PC row (P11) (7 <0.7; significant at p < 0.01); and (1)
breadth of brain case (D9) with height of sagittal crest
(L27) (r = 0.25; not significant).

Neurocranium region (D9, L27)

Breadth of brain case (D9) followed a neural
growth pattern, with most growth completed by 6
y (84 mm) (Figure 6). Overall growth scaled with
negative slope (b= 0.17) relative to CBL. In yearlings,
the brain case was proportionally long, 1.e., 75% of
CBL in yearlings, and 63% of CBL at 10 y. Growth
in length of the brain case (31% at 10 y relative to
yearlings abbreviated to RTY) was much greater than
growth in breadth (8% at 10 y, RTY). The ratio of
breadth to length increased from 1: 1.5 (yearlings) to
Is Oy CIO sp).

Height of sagittal crest (L27) appeared to follow
a somatic growth pattern; however, there was great
variation among individuals of similar age. The
crest was absent in juveniles and young subadults.
Evidence of crest formation was apparent in one 4
y old (n = 7), two 6 y olds (n = 4), eight 7 y olds (n
= 8), and all males = 8 y. Maximum crest height was
11-12 mm (nv = 4). There was some evidence of a
very slight secondary growth spurt in some males at
c. 10 y, but sample size was too small to confirm this
observation.

Basicranium region (P21, P22, P23)
Calvarial breadth (P21) followed a somatic,

monophasic growth pattern. Overall growth in
variable size increased in proportion (6 = 1) to skull
size, increasing by 49% at 10 y (RTY). Mastoid

Facing page: Figures 4-6

Figure 4a, b: Bivariate plot of log condylobasal length (mm) on: (a) log body length (cm) and (b) age (y).
Solid triangles, known-age animals (MCM) based on tagging. Squares, canine aged animals (PEM).

Figure 5a, b, c: Bivariate plots of: (a) log body length (cm) on suture age; (b) log condylobasal length
(mm) on suture age; (c) suture age on age (y). Solid triangles, known-age animals (MCM) based on tag-

ging. Squares, canine aged animals (PEM).

Figure 6: Example of neural growth. Log breadth of brain case (mm) on age (y). Solid triangles, known-
age animals (MCM) based on tagging. Squares, canine aged animals (PEM).

224

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

O
S) ni: ee
OB
IB. AE. 1
S S
© © a] < po o
xt <x
<add
+ a <a —
qa <
<4
Ge 02 Gl os 8'Y ov vv ov Or
abe aunins (ww) esed uleig JO yypeeig UT
io)
fap)
a
®o
2 2
@ 5)
= 5
—J _
a ag
ae
Zs os ey oF vY S's es plies
(w9) y}6ua} Apog u7 (ww) uiGHus| j|eseqo|Apuod uz
O
OO < re
imps.
2 2) e]ae) jo) -
aS)
= OM) i
=) Ss
& Oo «a< © o
> Zz
3 Oo <@
Cc
= 4<0 «a << b +
<<
Oo
= < < rN
<<
GS es ie 9S oe) VS es os LS
(ww) uj6u9] jeseqo|Apuoo u7 (ww) u}6Hu9] feseqo|Apuod u7

Proc. Linn. Soc. N.S.W., 129, 2008 225

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

Table 5: Discriminant analysis for seal age group (sub-adult, adult) inferred from skull length

Known Age Classification
Group into age group Subadult (y,) Adult (y,)
(1 y 7 months to 7 y 6 months) (2 7 y 7 months)

26 (81%)

26 (90%)

2g)

a

Basioccipito-

vill | basisphenoid
brain case

3

Coronal (brain
case)

Squamosal-jugal
x
(face-zygomatic)
Premaxillary-
i | maxillary (face-
maxilla

Maxillary (face-
maxilla)

Squamosal-
parietal (brain

v1
1
1x
IV
Basisphenoid-
vill | presphenoid
brain case

Internasal (face-
Vv

nasal)

Suture index

1

Total no. skulls
= 48

226 Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

breadth (P22) followed a somatic growth pattern.
Overall growth scaled with positive slope (6 = 1.29)
relative to CBL, increasing by 80% at 10 y (RTY).
A prominent secondary growth spurt was apparent at
10 y (Figure 7). Basion to bend of pterygoid (P23)
followed a neural growth pattern, with most growth
competed by 7 y (76 mm). Overall growth in variable
size was negatively allometric relative to CBL.

Frontal region (D7, D8)

Least interorbital constriction (D7) followed
a somatic, monophasic growth pattern. Overall
growth expressed positive allometry relative to CBL,
increasing by 79% at 10 y (RTY). Most growth was
completed by 9 y. Greatest breadth at supraorbital
processes (D8) followed a somatic growth pattern.
Overall growth scaled with a very slight positive
slope (6 = 1.03) relative to CBL, increasing by 50%
at 10 y (RTY). A weak secondary growth spurt was
apparent at 7 y.

Zygomatic arch (P14, P19)

Breadth of zygomatic root at maxilla (P14)
followed a somatic, monophasic growth pattern.
Overall growth was isometric relative to CBL,
increasing by 24% at 10 y (RTY). Zygomatic
breadth (P19) followed a somatic, monophasic
growth pattern. Overall growth scaled with positive
slope (6 = 1.12) relative to CBL, increasing by 64%
at 10 y (RTY). The ratio of zygomatic breadth to CBL
was 1:1.8 in yearlings and adults. Zygomatic breadth
was generally the widest part of the skull; however,
mastoid breadth exceeded zygomatic breadth in 10
animals (7 subadults; 3 adults).

Splanchnocranium region (D3, L24, L25)

Gnathion to posterior end of nasals (D3) followed
a somatic growth pattern. Overall growth expressed
positive allometry relative to CBL, increasing by
66% at 10 y (RTY). A weak secondary growth spurt
was apparent at 10 y. Gnathion to foramen infraorbital
(L24) and gnathion to posterior border of preorbital
process (L25) followed a somatic, monophasic
growth pattern. Overall growth scaled with positive
slope (6 = 1.26, 1.25) relative to CBL, increasing by
62% and 70% at 10 y (RTY), respectively.

Nasal region (D4, D5)

Width of anterior nares (D4) followed a somatic,
monophasic growth pattern. Overall growth was
isometric relative to CBL (Figure 8), increasing by
43% at 10 y (RTY). Greatest length of nasals (D5)
followed a somatic growth pattern. Overall growth
expressed positive allometry relative to CBL,
increasing by 76% at 10 y (RTY). There was some

Proc. Linn. Soc. N.S.W., 129, 2008

evidence of a very slight secondary growth spurt at 10
y, but this may have been an effect of sampling. The
ratio of nasal breadth to length increased from 1:1.2
(yearlings) to 1:1.5 y (10 y).

Palatal region (P10, P11, P12, P13, P15, P16, P17)
Palatal notch to incisor (P10) and gnathion to

posterior end of maxilla (P13), followed a somatic
growth pattern. Overall growth scaled with a very
weak positive slope (b = 1.07, 1.06) relative to
CBL, increasing by 70% and 61% at 10 y (RTY),
respectively.

Length of upper PC tooth row (P11) followed a
somatic, monophasic growth pattern. Overall growth
scaled with negative slope (b = 0.84) relative to CBL,
increasing by 46% at 10 y (RTY). Greatest bicanine
breadth (P12) followed a somatic growth pattern.
Overall growth expressed positive allometry relative
to CBL, increasing by 51% at 10 y (RTY). There was
some evidence ofa very slight secondary growth spurt
at 10 y, but this may have been an effect of sampling
biases.

Breadth of palate at PC 1 (P15), 3 (P16) and 5
(P17) followed a somatic growth pattern, increasing
by 89%, 47% and 72% at 10 y (RTY), respectively.
Overall growth expressed strong positive allometry
for breadth at PC1; positive allometry for PC5; and
isometry for breadth at PC3, relative to CBL. There
was some evidence of a very slight secondary growth
spurt in breadth at PCS at 10 y, but this may have been
an effect of sampling. The ratio of palatal breadth at
PCS (P17) to palatal length (P10) was 1:3 in both
yearlings and adults (10 y).

Mandible (M28, M29, M30, M31, M32)

Length of mandible (M28) followed a somatic,
monophasic growth pattern. Overall growth scaled
with positive slope (b = 1.22) relative to CBL,
increasing by 79% at 10 y(RTY). Length of mandibular
tooth row (M29) and length of lower post-canine row
(M30) followed a neural growth pattern, with most
growth completed by 7 y (table 3; 68 + 0.6 (SE, n
= 12) mm; 45 + 0.4 (SE, n = 12) mm). Growth was
negatively allometric relative to CBL (Figure 9). The
ratio of the length of the lower PC row (M30) to upper
PC row (P11) was 1:1.1 (yearlings) and 1:1.3 (10 y).
Height of mandible at meatus (M31), and angularis
to coronoideus (M32), followed a somatic growth
pattern, with a weak secondary growth spurt at 10 y.
Overall growth expressed strong positive allometry
relative to CBL (Figure 10), with variables increasing
by 130% and 105% at 10 y (RTY), respectively.
Growth in vertical height of the mandible was
considerably greater than that of length.

227,

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

Ln mastoid breadth (mm)
Ln length of lower postcanine row (mm)

5.1 5.3 5.5
Ln condylobasal length (mm)

Ln greatest width of anterior nares (mm)
Ln height of mandible at meatus (mm)

5.1 5.3 5.5 5.1 5.3 5.5
Ln condylobasal length (mm) Ln condylobasal length (mm)

Figure 7: Example of somatic, biphasic growth. Log mastoid breadth (mm) on age (y). Solid triangles,
known-age animals (MCM) based on tagging. Squares, canine aged animals (PEM).

Figure 8: Example of isometric growth. Log greatest width of anterior nares on log condylobasal length
(mm). Solid triangles, known-age animals (MCM) based on tagging. Squares, canine aged animals
(PEM).

Figure 9: Example of negative allometry. Log length of lower postcanine row (mm) on log condylobasal
length (mm). Solid triangles, known-age animals (MCM) based on tagging. Squares, canine aged ani-
mals (PEM).

Figure 10: Example of strong positive allometry. Log height of mandible at meatus (mm) on log condylo-

basal length (mm). Solid triangles, known-age animals (MiCM) based on tagged animals. Squares, canine
aged animals (PEM).

228 Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Abnormalities

Of the 60 PEM skulls examined, 14 (23%) had a
cleft palate varying in severity from slight clefts to
moderate deformity (PEM: 951, 958, 1453, 1560,
SS eZ05 05220515 20525.2053, 2132, 2137, 2141,
2197, 2253), and abnormal bony deposits were
observed on the occipital bone, at the base of the
parietal in PEM2049. In addition, it was clear that the
deciduous canines persist in some animals until their
tenth month (i.e., AP4999). Hence the generalisations
of Rand (1950) that (i) deciduous canines of pups are
lost by the end of March (about 3 to 4 months old);
(11) permanent canines do not erupt from the gums
before 4 months of age and are well developed by 8
months old (end of July) are not universal.

Comparison of South African and Australian Fur
Seal Skulls
Comparisons were made between data on male
South African fur seal skulls from the present study
with published information on male Australian fur
seals. King (1969) published data on condylobasal
length (CBL), zygomatic width and supraorbital
width (postorbital width) of male Australian fur seal
skulls judged to be adult male based on their suture
indices. Figures 11 and 12 are plots of King’s data
compared to data on South African material with a
Condylobasal length greater than 180 mm. Linear
egression analysis showed that both Zygomatic Width
vs. CBL and Supraorbital width vs. CBL are linearly
related and the Australian and South African material
fall on the same regression lines:
Zygomatic Width vs. CBL (n = 78, South
African n = 36, Australian n = 42)
Slope 0.678 + 0.037 (Standard Error, SE)
Intercept = -25.8 + 9.99 mm (SE) (marginally
different to zero based on t-test)
INK —A 207] SX CIE 25.8, f= 0.92525 p<
0.001
Supraorbital Width vs. CBL (n = 73, South
African n = 32, Australian n = 41)
Slope 0.212 + 0.026 (SE)
Intercept = 3.98 + 7.09 mm (SE) (not
significantly different to zero)
SOB 0212 CB + 3.98.1 = 0.6898, p <
0.001
However, although the regression analyses show that
the South African and Australian skulls share the same
geometry it is obvious that the South African skulls
are significantly smaller than the Australian material
from King (1969) although there is considerable
overlap: South African; CBL = 248 + 10.7 mm (SD,
n= 36), Zygomatic width = 141 + 10.1 mm (SD, n=
36), Supraorbital width = 57 + 5.4 mm (SD, n = 32);

Proc. Linn. Soc. N.S.W., 129, 2008

Australian; CBL = 283 + 10.7 (SD, n= 42, Zygomatic
width = 166 + 10.3 (SD, n = 42), Supraorbital width
= 64+5.8 (SD, n= 41). These values agree well with
those published by Brunner (1998ab, Brunner 2000).

Table 7 compares the mean values fora wide range
of skull parameters of South African and Australian
fur seals. The Australian fur seal measurements are
consistently larger than the South African skulls.
However, there is one significant exception. In the
present study, adult South African fur seals were
found to have a braincase width (D9) of about 84
+ 2 mm (SD, n = 46) (see Table 3); this does agree
with values published for South African fur seals by
Brunner (1998b) (84 + 4.6 mm, SD, n = 17) but not
with the value found in her thesis (Brunner 2001) (77
+ 2, SD, n = 38). Brunner (1998b) reports the brain
case width of Australian fur seals to be 78 + 2.25 (SD,
n = 45). If the braincase data of the present study
shown in Table 3 was correct, then braincase size
was larger in South African fur seals than that of the
Australian variety. This seems exceptional and is not
consistent with the differences found in the other skull
parameters (Table 7). Furthermore, if the braincase
was smaller in Australian fur seals, this would affect
the zygomatic width measurements (Figure 1, P19)
and hence the geometry of the skull. In the present
study, braincase width (D9) was measured across
the same section of the skull as the zygomatic width
(P19) (see Figure 1). However, the graph shown in
Figure 11 shows that South African and Australian fur
seals have similar skull geometry. We conclude that
the braincase width has been measured in a different
way in the present study compared to the method used
by Brunner (2001). The braincase width measure
shown on the skull diagrams in the present study (D9
in Figure 1) and parameter 9 in Figure 2 in Brunner
(1998a) appear to be the same but cannot have been
measured in the same way. If we take the braincase
width data from Brunner (2001) for both the South
African and Australian fur seals we find there is no
significant difference in braincase width in the two
populations (Table 7).

DISCUSSION

Skull size

Arctocephalus pusillus is the largest of the fur
seals, therefore the skull is correspondingly large.
In the present study of the South African fur seal
(A. pusillus pusillus), the maximum CBL was 275.4
mm (PEM898); however, skulls up to 307 mm
have been reported in the Australian fur seal (A.

229

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

200

+ Australian Fur Seals

o South African Fur Seals “

180

160

140

zygomatic Width (mm)

120

100

200 220 8240

260 280 300 320

Condylobasal Length (mm)

Figure 11: Comparisons of the skulls of male South African and Australian fur seals: Zygomatic arch
width vs. Condylobasal length (CBL). When fitting two linear regression models to the Australian and
South African fur seal data it was found that the intercepts and the slopes were not significantly different
using a F-test (p = 0.08) (Draper and Smith, 1981). A single straight line could be fitted to all of the data

(r = 0.9252, n = 78, p < 0.0001).

pusillus doriferus: Cruwys and Friday, 1995). As
with all southern hemisphere fur seals, the skull is
considerably larger in males than in females, reflecting
pronounced sexual dimorphism (Bryden, 1972; King,
1972, 1983; Cruwys and Friday, 1995; Arnould and
Warneke, 2002; Brunner et al., 2004). Table 7 and
Figures 11 and 12 clearly indicate that male South
African fur seal skulls are on average smaller than
Australian fur seal skulls but Figures 11 and 12 show
that they share the same geometry.

230

South African Fur Seals from the East Coast and
West Coasts of South Africa

It has been suggested that marine mammal
species inhabiting warmer waters may be smaller in
body size than marine mammal species inhabiting
cooler waters (Ross and Cockcroft, 1990) reflecting
generally higher productivity and hence food supply
in cooler waters. Long-term climatic data in Algoa
Bay (South Africa), based on daily measurements,
indicate that the mean water temperature is 16—17°
C in winter and 21—22° C in summer. For Luderitz

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

80

60

A0

supraorbital Width (mm)

20

¢ Australian Fur Seals

o South African Fur Seals

200

220 240

2760 280 300 320

Condylobasal Length (mm)

Figure 12: Comparisons of the skulls of male South African and Australian fur seals: Supraorbital width
vs. Condylobasal length (CBL). Fitting two linear regression models to the Australian and South Afri-
can fur seal data showed that the intercepts and the slopes were not significantly different based upon a
F-test (p = 0.35) (Cochran, 1977). A single straight line could be fitted to all of the data (r = 0.6848, n =

73, p < 0.0001).

(near Sinclair Island, South Africa), mean water
temperature is 12—13° C in winter and 14-15° C in
summer, considerably cooler than Eastern Cape waters
(Dr M. Grundlingh, pers. comm.). When comparing
CBL from adult South African fur seals from these
two geographic locations, we did not find sufficient
reason to reject the hypothesis that the population
means for skull length were equal using PEM
animals = 12 y. However, it is not clear if this result
was influenced by a larger number of older adults in

Proc. Linn. Soc. N.S.W., 129, 2008

the PEM sample. When younger PEM animals were
included in the adult sample (7-12’ y), Eastern Cape
seals were found to be significantly smaller than west
coast seals. Further testing using a larger sample of
aged animals is required.

The larger size of the Australian fur seal (A.
pusillus doriferus) which mainly lives in Bass Strait in
Australia (Arnould and Warneke, 2002), is an argument
against a straightforward relationship between
body size and water temperatures (Warneke and

231

*(100Z) J0UUNAG WO. Udye} S}UDUIEMSvoU ATVUIUINS [vos IN} UvITAyY yYNOS,,
“UBILIFY YNOS — uL[e.YsNy suOstIvduIOd [TV *(S9I
-[2}1 UI) S[eag AMY ULOLAFY YINOS [NPY ILIA WIM (UOJ pys) sjeag Any UETEYSsNY INpy FRA Jo Uostredui0D */ aqQuI,

HOVd ONIOVA GCHONILNOD

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

ca © OE SE EE coeyipe os 198 proRaod yo puog —uorsea
0@d Ss ae: i eae S0000> Lr EST Joo1 oyeut0SAz — uoIseg
ea ee ee eae A
tld i snc ae BAe S000°0> SV 9L Cl (jeqeyed) eljrxeur 1o119jsod — uoryyeun
ee oe aoe | Sosa oo
Pe a i ae a i S000, 2 16°01 [eyqrooid fo 1ap10q ioe
vol e Bae aa ne S0000> 19 co vl S[eqroeyur UStUeIO; — UOTYyeUt)
co Ecos evs ese Jo FU
ce 969 sesou jo pus ouasod—uonpoup
a a
Id 2 Sie ae ore S000 0> vr OS el yysuey jeseqo]ApuoD
spon siu asuvy GS uvayy onjea-d jp § (9) sduaa07FI1q [MAS JO sJUSWIIINSvay] YSsUIT

Proc. Linn. Soc. N.S.W., 129, 2008

E337)

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

cCd

I@d

6Id

Lid

91d

vid

Cld

LCT

9C1

6d

8d

Ld

9d

vd
9po)

6L7IC OTT
VOLI-L LVI
T9CI-9 VOL
0'6L1-C OCI
6851-4 ICL
O'6LI-L ESI
VOE-ELC
OlSp-IS Se
UM CEDAWG,
67 Vv-06 8C
V6I-C CI
90 CC-9S El
119-6 1P
9L69-LE CS
0-C1-00

8h 91-889
CSCI-ES8
LOrl-0'L11
L0 18-0 €L
L8°€8-09 CL
169-1 9F
LO'9L-99'SS
8 CV8 OE
O08 Ly-6S SE
OLLE SS
IL L8-vr 69
8 EET VC
O09 EP-SL'TE
osuLy

£66
GUS
SY)
€$°8
906
66S
89T&
8LC
DEIR
NOt
9861
v3 1
LO8 F
09'€
88L°E
IvVé
CO IT
6e S
COC
eC
EOS
OSES
6cl €
vO€
9C°
617
INL
tell ©
ds

GS
vO 8SI
VOTT
OSET
VCHI
LYy91
96 €&
LL 6€
fie, JUG
€S'8E
ES ST
10°61
98 05
819
CCLV
IZ Ol
6r SOT
6 0S1
OF LL
JEM Sie
VE 9S
L199
O08 LE
IL CY
0289
6¢ 6L

89° 8C
VOLE
Uva

£000°0>

£000 0>

£000 0>

£000 0>

£000 0>

£000°0>

$000 0>

$000 0>

$000 0>

10) 0)

£000 0>

S000 0>

£000 0>

£000°0>

onjea-d

9€

c9

(G4

gS

es

9¢

LV

EXE

Se

08

cs

es

SV

6V

61 Cl

cs Ol

Sc Il

86 L

NS fe!

c6 9

eL Ol

cO'L

66 81

7) Il

98°9

Ig9

ce 6

8Er1
(3) sduaI0FFIG

Uypearq proyseyAl

ypeeq Atoypny

Uypesig syewWo0sA7Z

¢ ouTuRsjsod ye ojeyed Jo yypeolg

p-¢ souruesjsod ye oye7ed Jo yIpRog

RI[IXCUI JO JOOI SIjeWIOSAZ Jo YpRog

UIPIM [e.ysoy

JSOJO [BYISES JO JYUSIO

Prlo}seul — jsoI9 [eyId1I990

,OSBouIelg Je YIpeoig

sassoooid jeyIqioeidns ye ypesig

UONILYSUOD [e}1G10.19}U]

sassoooid [ey1q10o1d je Yypeorg

soieu JO Ypeog

[MAS JO syusWIANsvayA] YpPLog

233

Proc. Linn. Soc. N.S.W., 129, 2008

Code

n

SD Range

Mean

df p-value

Difference (t)

Mandible and Teeth

234

M29

24

69.23-84.17
3.876 59.8-75.7

3.20

WIMS
68.70

<0.0005

41

6.89

Length of mandibular tooth row

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

M30

4]
29

51.61 2.56 43.86-57.53
2.836 40.8-52.5

46.77

56 <0.0005

7.32

Length of lower postcanine row

M31

41
30

47.66-78.04
39.5-68.8

5.30
7.65

69.03
58.35

48 <0.0005

6.58

Height of mandible at meatus

M32

41
29

56.33-76.81
42.3-67.1

4.35
6.65

68.77
58.07

46 <0.0005

7.69

Angularis - coronoideus

M36

198.1 2 181.7-214.6 4]
11.50 141.6-192.0 2

M33

43 <0.0005

10.26

Length of mandible

Table 7 continued

Shaughnessy, 1985; Brunner 1998ab; Brunner et al.,
2002) because Bass Strait waters are considerably
warmer than the South African waters where the South
African variety occurs (Stewardson 2001). Arnould
and Warner (2002) also point out that Bass Strait
waters are also far less productive than the waters
inhabited by the South African fur seal. Other closely
related fur seals found in Australian, New Zealand,
Subantarctic and Antarctic waters (A. australis
forsteri, A. tropicalis, A. gazella and A. australis) are
all smaller than A. pusillus (Kerley and Robinson,
1987; Brunner 1998ab; Brunner et al., 2002; Daneri
et al., 2005; McKenzie et al., 2007).

Skull shape

Morphological observations of the skull were
generally consistent with earlier studies by Rand
(19496, 1950, 1956) and Repenning et al. (1971). As
for all otariids, the frontal bones project anteriorly
between the nasal bones; supraorbital processes
are present; the tympanic bulla are small and fiat,
comprised primarily of the ectotympanic; the
alisphenoid canal is present; the mastoid processes are
massive; the jugal-squamosal joint of the zygomatic
arch overlap; and deep transverse grooves occur on
the occlusal surface of the upper incisors (Burns and
Fay, 1970; King, 1983; present study).

Within the species, the forehead is convex at the
supraorbital region; the snout is long; the nasals are
long and flared anteriorly; the palate is moderately
broad and arched; the maxillary shelf at the root of
the zygomatic process is very short in an anterio-
posterior direction; the tooth rows are parallel, with
robust, tricuspid PC, and a slight diastema between
upper PC 5 and 6 (Repenning et al., 1971; present
study).

As with other species of this genus, the interorbital
region (D7) was less than 20% of CBL in adults (..e.,
15%); palatal notch to the incisors (P10) was more
than 37% of CBL (i.e., 43%); and nasal length (D5)
fell within 14% (smallest fur seal, A. galapagoensis)
and 18% (largest fur seal, A. pusillus pusillus and
A. pusillus doriferus) of CBL (i.e., 18%) (Scheffer,
1958; Cruwys and Friday, 1995; present study).

Growth Curves

Mammals typically exhibit a determinant growth
pattern — as juveniles and subadults they grow in size
at a fast rate when very young then growth gradually
slows until it finally more or less stops as adults. Such
growth patterns can usually be described by various
variations of exponential or logistic growth curves of
the form: exponential saturation or von Bertalanffy
curve, y = y_. *(1-e“') or logistic growth curve, y

max

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

= y,,,,/(1-be™). Both these curves can be difficult to
fit to experimental data and obtain useful estimates
of y,,, and k. Fitting usually requires values for
accurately aged fully grown animals over a range
of adult ages otherwise the asymptotic y,,, may not
be estimated reliably. More complex models with
another unknown, for example to allow for length as
infants, are even more difficult to fit.

The present study has shown that South African
male fur seals continue to grow to about 10 y, only
just short of the age of the oldest definitively aged
individuals (tagged animals) and also just short of the
maximum age that can be determined from dentition
(12 y). Arnould and Warneke (2002) working on
Australian fur seals had enough fully grown and
aged seals to be able to fit asymptotic growth curves
to their data for both males and females. Similarly,
Winship et al. (2001) working on the Steller sea lion
(Eumetopias jubatus) had access to data on hundreds
of individuals aged on the basis of dentition and were
able to fit SBL vs. age to both types of asymptotic
growth curve.

In the present study SBL and skull parameters
such as CBL when plotted against age did not show
obvious asymptotes (Table 4, Figure 4b) and little
curvature so maximum SBL and CBL could not be
accurately determined from such curves. There were
not many aged adult animals and the oldest known
age was 13 years so the range of ages of fully-grown
adults was small. When we did fit these exponential
saturating curves to the data the residuals versus
fitted values plot were not random scatters about zero
indicating that the curves were not adequate models
for the data. The data was better described by simple
linear or Log/Log relationships.

Condylobasal length as an indicator of SBL and
age

In male South African fur seals, CBL continued
to increase until at least 12 y, with no obvious growth
spurt at social maturity (8-10 y). The absence of very
old skulls of known-age (18-20 y), made it difficult
to determine overall growth in CBL. In contrast
CBL continues to increase until at least 13 y in
male C. ursinus (Scheffer and Wilke, 1953, but see
Scheffer and Kraus, 1964) and slows at 10 y in male
Eumetopias jubatus (Fiscus, 1961).

Condylobasal length was found to be a
reasonable indicator of SBL and age group, but not
of absolute age. The classification criteria for SBL
developed in this study will be particularly useful
when a seal is decomposed/scavenged (total SBL
can not be measured), and/or the skull is incomplete/
absent (total SBL can not be extrapolated from skull

Proc. Linn. Soc. N.S.W., 129, 2008

length). The classification criteria for age group will
be particularly useful when teeth are not available
for age determination; or museum records have been
misplaced or destroyed. As more specimens become
available, particularly very old tagged individuals of
known age, the classification criteria could be made
more precise.

Suture index as an indicator of age

Although cranial sutures close progressively
with age, suture age was not considered to be a
good indicator of chronological age (y) or age group
(sub-adult/adult). Similar observations have been
made in other male otariids, e.g., in C. ursinus, the
rate of suture closure is highly variable, and like
SBL and CBL, is a poor indicator of chronological
age (Scheffer and Wilke, 1953) compared to teeth
(Scheffer, 1950; McCann, 1993; Oosthuizen, 1997).
Brunner et al. (2004) in their study of suture closure
sequences in several fur seal species was also
hesitant to use suture closure indices to indicate
chronological age although they were perhaps more
confident in using it to assign skulls to age groups
than is warranted from the conclusions drawn from
the present study. A particular problem is that canine-
tooth sections are only useful for aging individuals
up to 12 years old and yet some individuals at that
age still have some incomplete closure of sutures.
Some individuals probably live considerably longer
than 12 years but skulls from very old animals are not
currently available.

Function and growth
Neurocranium region

In mammals, growth of the protective brain case
corresponds closely to that of the enclosed brain
(Moore, 1981). The brain/brain case grows rapidly
during prenatal and postnatal life; attains full size
early in development before that of the basicranium
or face; and scales with negative slope relative to
skull size (Moore, 1966, 1981; Bryden, 1972; King,
1972; Gould, 1975; Moore and Lavelle, 1975; Enlow,
1982; Shea, 1985; Wayne, 1986; Hartwig, 1993;
Morey, 1990; present study). Early maturation of the
brain/brain case is essential for nervous control of the
body.

The sagittal crest strengthens the skull, and
provides an increased surface area for muscle
attachment. In adults, large crest size is advantageous
in combat behaviour between breeding bulls, and in
feeding (increases bite force). Sagittal crest height
begins to increase in size at 4—7 y (highly variable),
reaching at least 12 mm in some adult males. In male
Zalophus californianus, the sagittal crest begins to

235

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

develop at 5 y, with height ranging from 11—36.5
mim in adults (Orr et al., 1970). In male Ewmetopias
jubatus, sagittal crest height ranged from 7-37 mm
in adults (Fiscus, 1961). The wide variance in crest
height in older South African fur seal males is possibly
an artefact of the unknown actual age of bulls classed
as >12y based on dentition. It is also possible that it
might reflect differences between breeding and non-
breeding bulls in the collection of skulls used in the
present study.

Basicranium region
The basicrantum accommodates the hearing

apparatus (Enlow, 1982). As with other mammals,
growth of the otic capsule (and associated structures)
appears to follow a neural growth pattern (Bast and
Anson, 1949; Hoyte, 1961; Moore, 1981). Early
development of the otic capsule enables juveniles to
recognise the ‘pup-attraction call’ of their mothers.
Mother-pup recognition is critical for pinnipeds
living within a colony where separation is frequent,
and mother-pup pairs are numerous (see Rand,
1967; Trillminch, 1981; Oftedal et al., 1987; Bowen,
1ODi)).

Unlike the otic capsule, calvarial breadth and
mastoid breadth mature much later in life (present
study). In adults, enlarged mastoids are advantageous
in combat behaviour between breeding bulls, and
in feeding (large head size/increases bite force);
and facilitate directional hearing (provides a greater
surface area of specific orientation for selective
reflection of sound) (Repenning, 1972).

Frontal region
The interorbital region provides the structural

base for the snout (Enlow, 1982). The dimensions
of this region increase with age to accommodate the
development of the proportionally large snout.

The supraorbital processes strengthen the skull
(very thick in adults), protects the orbital region,
and increases bite force. In adult seals, this enlarged
structure is advantageous in feeding, and in combat
behaviour between breeding bulls.

Zygomatic arch

The zygomatic arch protects the eye, provides a
base for the masseter and part of the temporal muscle,
accommodates conductive hearing (squamous root)
and is the poimt of articulation for the mandible
(Evans, 1993; Repenning, 1972). As with other
mammals, the zygomatic arch enlarges laterally and
inferiorly to accommodate enlargement of the head,
and a correspondingly greater temporal muscle mass
(Moore, 1981; present study).

236

The orbital border of the zygomatic bone forms
the ventral margin of the eye socket. As with other
pinnipeds, the orbits were large to accommodate
large eyes (King, 1972). In South African fur seals,
the horizontal diameter of the eye is c. 40 mm
(e.g., animals AP5215, 2 y 4 months; AP5210, 3 y).
Although large eyes are potentially advantageous in
the detection of benthic and/or fast moving pelagic
prey (David, 1987), vision is not necessary to locate/
capture prey in seals (see King, 1983).

Splanchnocranium region
In South African fur seals, lateral face length and

width of snout at the canines, scaled with positive
slope relative to CBL, similar to that of wild canids
(Lumer, 1940; Wayne 1986; Morey, 1990; Evans,
1993). As the face and snout increased in length, the
brain case and orbits became proportionally smaller.

In mammals, the size and shape of the brain
establishes boundaries that determine the amount of
facial growth; and special sense organs housed within
the face influence the direction of growth (Enlow,
1982). In adult South African fur seals, the brain is
relatively large and more spherical than in terrestrial
carnivores (Lumer, 1938; Harrison and Kooyman,
1968; King, 1983; Evans 1993), yet long and narrow
compared to humans and other primates (i.e., small
cerebrum) (Scott 1951; Gould 1975; Moore and
Lavelle 1975; Shea 1985; Sirianni and Swindler,
1985). Therefore, the snout is correspondingly long
and narrow. The wide nasal openings were aligned
in a horizontal plane with the nerves of the olfactory
bulb; and the orbital axis is pointed straight forward
in the direction of body movement (Enlow, 1982;
present study).

Nasal and palatal region
The naso-maxillary complex is the facial part

of the respiratory (nasal cavity) and alimentary (oral
cavity) tracts, which also facilitates sound production
and the sense of smell. The floor of the nasal cavity
forms the roof of the oral cavity, thus growth of the
two cavities was highly coordinated. Growth was
predominantly somatic, with similar allometric trends
to those of wild canids (Lumer, 1940; Wayne, 1986;
Morey, 1990). Progressive growth of this region is
needed to accommodate the large dental battery.
Growth of dentition has been described by Rand
(1950, 1956). At 6 to 12 months, South African
fur seals gradually transfer from milk to solids
(fish, crustacean and cephalopod) (Warneke and
Shaughnessy, 1985). Although the small, deciduous
teeth are usually lost by the end of the first 5 months
(Rand, 1956), deciduous canines may persist for 10
months (present study). The permanent teeth are used

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

to hold slippery prey (gripping), and to reduce prey
size (biting and shearing) and of course in fighting
between rival males. Growth of the permanent teeth
is a gradual process, with diet becoming more varied
with age and experience (Rand, 1959). In the upper
jaw, the canines protrude beyond the tip of the 3rd
upper incisor only in the 2nd y (Rand, 1956).

In male South African fur seals, the ability to
produce sound is evident at birth, with vocal skills
broadening with increased age (Rand, 1967). In
otariids, the production of sound is important in
mother-pup recognition; communicating within a
colony; and affirmation of territorial boundaries and
social status (e.g., Stirling and Warneke, 1971).

Although the olfactory area is reduced when
compared to terrestrial carnivores, the sense of smell
appears to be well developed, and plays an important
role in the detection of sexually receptive females,
and land predators (Harrison and Kooyman, 1968;
Peterson, 1968; King, 1983; Renouf, 1991; Wartzok,
1991).

Mandible

Using human anatomy as a model, the horizontal
part of the mandible (corpus) provides the structural
basis for tooth formation, and the vertical part
(ramus = condyle, angular process, coronoid process,
masseteric fossa) provides areas for articulation and
muscle attachment (Scott, 1951).

As with other carnivores, the ramus increased
substantially in height to accommodate implantation
of the teeth, and expansion of the nasal region
(Evans, 1993; Enlow, 1982; present study). The
coronoid process grew upwards and backwards
increasing in thickness on the anterior borders; the
condyles grew backwards, beyond the level of the
coronoid process; and the masseteric fossa formed a
large, deep depression for jaw muscle (masseter and
temporalis) attachment. Large jaws and jaw muscles
are advantageous in feeding and in combat behaviour
between breeding bulls (increases bite force/increases
gape).

In mammals, the mandible of newborns is
proportionally smaller than the upper jaw, and
therefore must grow at a slightly faster rate to provide
anatomical balance (Scott, 1951; Enlow, 1982).
In order to achieve correct occlusal relationships
between upper and lower dentition, the rate of growth
between the mandible and maxilla needs to be highly
coordinated (Scott, 1951; Bryden, 1972; Moore, 1981;
Enlow, 1982; Hartwig, 1993; Brunner et al., 2004).
In South African fur seals, the PC teeth are robust,
therefore the tooth row is long compared to other fur
seals and other carnivores such as dogs (Rand, 1950;

Proc. Linn. Soc. N.S.W., 129, 2008

Scott, 1951; Scheffer and Kraus, 1964; Bryden, 1972;
Burns and Fay, 1970; Enlow, 1982; Wayne, 1986;
Hartwig, 1993; McCann, 1993; Cruwys and Friday,
1995; Oosthuizen, 1997; Brunner, 1998ab; Brunner et
al., 2002; Brunner et al., 2004; Daneri et al., 2005).

Growth rate of the lower PC row (6 = 0.7) was
similar to that of the upper PC row (b = 0.8), relative
to CBL. Overall percent increase in growth was
greater in the upper jaw because there are 6 PC in the
upper jaw and only 5 in the lower jaw. The ratio of
length of the lower PC row to upper PC row was 1:1.1
in yearlings, and increased to 1:1.3 in adults (at 10
y). Growth of the anterior dentition was considerably
greater than that of the PC, due to development of the
large canines.

Information presented in this study confirms
earlier descriptions of the South African fur seal skull
(Rand, 19495, 1950, 1956; Repenning etal., 1971), and
provides new information on skull growth according
to age (y), not available for most seal species. In male
South African fur seals, CBL continued to increase
until at least 12 y, with no obvious growth spurt at
social maturity (8-10 y). Growth of the skull was a
differential process and not simply an enlargement of
overall size. Components within each region matured
at different rates and grew in different directions.
Apart from the dentition, all variables of the facial
skeleton followed a somatic growth trajectory, and
most variables were positively allometric with CBL.
Breadth of braincase and basion to bend of pterygoid
followed a neural growth trajectory and scaled with
negative slope relative to CBL. Condylobasal length
and suture age were found to be poor indicators
of absolute age. However, CBL was a reasonable
indicator of SBL and age group.

Further information is needed on cranial capacity;
orbital size; tooth eruption; and the development of
the sagittal crest in relation to chronological age and
social status. Meaningful biological, evolutionary
and functional inferences on skull growth can only
be made when similar quality data is available for
other pinniped species of known-age in particular the
Australian fur seals (A. pusillus doriferus) and New
Zealand (A. australis forsteri). Multivariate statistical
procedures can then be employed to summarise
morphometric relationships within and among
populations.

Table 7 and Figures 11 and 12 clearly indicate that
the male South African fur seal skulls in the present
study are smaller than available material from their
Australian relatives. Some caution is necessary in
drawing the conclusion that the male South African fur
seal is indeed always smaller than the male Australian

D3 i]

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

fur seal. There is considerable overlap in skull sizes,
particularly the ranges of the measurements (Table
7). In the present study, the South African skull
material represented material collected from breeding
colonies, stranded animals, animals killed by fishing
boat crews and animals that had been tagged as pups.
Most of the material used by King (1969), Crowys
and Friday (1995), Brunner (1998ab), Brunner (2000)
and Brunner et al. (2002) were museum specimens;
mainly “beachmaster’” breeding males shot at breeding
colonies. The Australian skull material is therefore
biased towards large breeding males and probably
also by the so-called “trophy effect” where collectors
tend to choose the largest specimens. The trophy effect
is compounded in the case of seals by their social
behaviour: breeding male fur seals “beachmasters”
exclude smaller non-breeding “bachelor” males
from colonies. Arnould and Warneke (2002) made a
deliberate effort to collect a range of sizes of seals to
avoid this problem.

There are also good biological reasons for
supposing that Australian fur seals are better fed
than their South African relatives. The modern South
African fur seal population has grown to near the
estimated population before commercial exploitation
but the Australian fur seal is still rapidly recovering
from near-extinction and so it is unlikely that
individuals are limited in size by resource limitations
(Armould et al., 2003). Today mass starvation of
South African fur seals occurs at irregular intervals
as a knock-on effect of failures of upwelling currents
(Anselmo etal., 1995). Itis also possible that Australian
fur seals are longer lived (see Arnould and Warneke
2002), which might again reflect a population not
yet fully recovered to their original numbers. From
these considerations it is reasonable to conclude that
the South African material is more representative of a
stable population of male South African fur seals than
the corresponding male Australian fur seal material.

Historically the separation of the South African
and Australian fur seals into subspecies was mainly
based on them having non-overlapping geographical
distributions and only minor differences such as a
slight difference in size (Australian fur seals seem to
grow slightly larger (& longer lived?), Cruwys and
Friday, 1995; Brunner 1998ab; Arnould and Warneke,
2002; Brunner et al., 2002; Brunner 2003; Brunner et
al., 2004). Recent molecular evidence supports their
varietal status as two very closely related but distinct
populations (Wynen et al., 2001). This implies that
the Australian population is of geologically recent
origin from South African immigrants.

Do stragglers from South Africa reach Australian
waters today? Identification of fur seals until recently

238

was largely based onprovenance because it was difficult
to separate some species based on classical taxonomy
(Brunner 2003). Thus, where distributions overlap,
for example the Australian (A. pusillus doriferus) and
New Zealand fur seal (A. australis forsteri), there can
be difficulties in positive identification, particularly
of immature individuals (King 1969; Brunner 2003).
Understandably, stragglers outside their normal range
can be difficult to identify. Recently some molecular
biological information on the interrelationships of
otariid seals has become available (Wynen et al.,
2001; Lancaster et al., 2006). Both the South African
and Australian fur seals are thought to be less wide-
ranging than their Subantarctic and Antarctic cousins
(A. tropicalis and A. gazella) and largely remain in
coastal waters. On the other hand, Warneke and
Shaughnessy (1985) state that South African fur
seals are known to forage at least 220 kilometres
offshore. Molecular evidence (Wynen ef al., 2001)
does not refute the possibility that contempory South
African fur seals do occasionally reach Australia and
breed with the local population. Lack of evidence of
stragglers from South Africa turning up in Australian
waters should not be construed as proof that this does
not occur today. Given an A. pusillus skull of unknown
provenance it would not be possible to confidently
assign it to A. pusillus pusillus or A. pusillus doriferus
on the basis of current morphometric or molecular
taxonomy (Wynen et al., 2001; Brunner et al., 2002;
Brunner 2003). The two varieties of A. pusillus are
so similar that only finding a tagged individual from
South Africa in Australia would settle the issue.

The biogeography of ocean roaming fur seals is
not static. For example, Macquarie Island, after its
original fur seal population (specie(s) undetermined?)
was wiped out in the 19" century, has been repopulated
by three species of fur seals (A. australis forsteri, A.
tropicalis and A. gazella). These three species are
known to be hybridising although the breeding success
of the hybrids is not high (Lancaster et al., 2006).
The closest sources of the Antarctic and Subantarctic
fur seal colonists are several thousand kilometres
away. Similarly, straggler Subantarctic fur seals (A.
tropicalis) periodically appear on the South African
coast (Shaughnessy and Ross, 1980) and in Southern
Australia (King, 1983; Kirkwood et al., 1992).

ACKNOWLEDGEMENTS

We wish to express our sincere appreciation to the
following persons and organisations for assistance with
this study: Dr V. Cockcroft (Port Elizabeth Museum), Dr
J. Hanks (WWEF-South Africa) and Prof. A. Cockburn

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

(Australian National University) for financial and logistic
support; Mr B. Rose (Oosterlig Visserye, Port Elizabeth)
who enabled us to collect seals from his commercial fishing
vessels; Dr G. Ross (formerly Port Elizabeth Museum) and
Dr V. Cockcroft for the use of PEM skulls collected before
April 1992 (7 = 16 skulls); Dr J.H.M David (MCM) for the
use of MCM skulls of known-age; Mr H. Oosthuizen for
assistance with aging techniques; Mr S. Swanson (MCM)
for assistance with data extraction and measurement of
MCM specimens; Mr N. Minch (Australian National
University) for photographic editing; Dr C. Groves and Dr
A. Thorne (Australian National University), and Dr J.H.M.
David (MCM) for their constructive comments on an earlier
draft of this manuscript. This paper is based upon a PhD
study by C.L Stewardson compiled on behalf of the World
Wild Fund For Nature — South Africa (project ZA-348, part
1b) and submitted to the Australian National University in
2001.

REFERENCES

Anselmo, S., Hart, P., Vos, H., Groen, J. & Osterhaus,
A.D.M.E. (1995). Mass mortality of Cape Fur Seals.
Arctocephalus pusillus pusillus in Namibia, 1994.
(Seal Rehabilitation and Research Centre Publication.
Pieterburen, Netherlands).

Arnbom, T.A., Lunn, N.J., Boyd, I.L. and Barton, T.
(1992). Aging live Antarctic fur seals and southern
elephant seals. Marine Mammal Science 8, 37 — 43.

Arnould, J.P.Y. and Warneke, R.M. (2002) Growth and
condition in Australian fur seals (Arctocephalus
pusillus doriferus) (Carnivora:Pinnipedia). Australian
Journal of Zoology 50, 53-66

Arnould, J.P.Y., Boyd, I.L. and Warneke, R.M. (2003).
Historical dynamics of the Australian fur seal
population: evidence of regulation by man? Canadian
Journal of Zoology 81, 1428-1436.

Bast, T.H. and Anson, B.J. (1949). The temporal bone and
the ear. (Charles C. Thomas Publ., Springfield).

Bowen, W.D. (1991). Behavioural ecology of pinniped
neonates. In “Behaviour of pinnipeds’, (Ed. Renouf,
D.), pp. 66-127. (Chapman and Hall Publ., London).

Brunner, S. (1998a). Skull development and growth in
the southern fur seals Arctocephalus forsteri and A.
pusillus doriferus (Carnivora: otariidae). Australian
Journal of Zoology 46, 43 — 66.

Brunner, S. (1998b). Cranial morphometrics of the
southern fur seals Arctocephalus forsteri and A.
pusillus (Carnivora: Otariidae). Australian Journal of
Zoology 46, 67-108.

Brunner, S. (2000). Cranial morphometrics of fur seals
and sea lions (family: otariidae) : systematics,
geographic variation and growth. Ph. D. Thesis, Dept.
of Veterinary Anatomy and Pathology, Faculty of
Veterinary Sciences, University of Sydney.

Brunner, S. (2003). Fur seals and sea lions (family
Otariidae) — identification of species and a taxonomic
review. Systematics and Biodiversity 1, 339-439.

Proc. Linn. Soc. N.S.W., 129, 2008

Brunner, S., Shaughnessy, P.D., Bryden, M.M. (2002).
Geographic variation in skull characters of fur seals
and sea lions (family Otariidae). Australian Journal
of Zoology 50, 415 — 438.

Brunner, S., Bryden, M.M., Shaughnessy, P.D. (2004).
Cranial ontogeny of otariid seals. Systematics and
Biodiversity 2, 83 — 110.

Bryden, M.M. (1972). Growth and development of
marine mammals. In “Functional anatomy of marine
mammals’, (Ed. Harrison, R.J.), vol. 1, pp. 58-60.
(Academic Press Publ., London, New York.

Cochran, W.G. (1977). Sampling techniques, 3rd ed. (John
Wiley and Sons Publ., New York).

Cruwys, E., Friday, A.E. (1995). A comparative review
of condylobasal lengths and other craniometric
characters in 30 species of pinniped. Polar Record
31, 45-62.

David, J.H.M. (1987). Diet of the South African fur seal
(1974-1985) and an assessment of competition
with fisheries in southern Africa. In “The Benguela
and comparable ecosystems’ (Eds. Payne, A.I.L.,
Gulland, J.A., and Brink, K.H.). South African
Journal of Marine Science 5, 693-713.

Daneri, G.A., Esponda, C.M.G., de Santis, L.J.M. and
Pla, L. (2005). Skull morphometrics of adult male
Antarctic fur seal, Arctocephalus gazella, and the
South American fur seal A. australis. Iheringia Serie
Zoologie, Porto Alegre 95, 261-267.

Doutt, K.J. (1942). A review of the genus Phoca. Annals
of the Carnegie Museum 29, 61-125.

Draper, N.R. and Smith, H. (1981). Applied regression
analysis, 2nd ed. (John Wiley Publ., New York).

Enlow, D.H. (1982). Handbook of facial growth, 2nd ed.
(W. B. Saunders Publ., Philadelphia).

Evans, H.E. (1993). Miller s anatomy of the dog, 3rd ed.
(W. B. Saunders Publ., Philadelphia).

Fiscus, C.H. (1961). Growth in the Steller sea lion.
Journal of Mammalogy 42, 218-223.

Gibbons, J.D. and Chakraborti, S. (1992). Nonparametric
Statistical inference, 3rd ed. (Marcel Dekker Publ.,
New York).

Gould, S.J. (1975) Allometry in primates, with emphasis
on scaling and the evolution of the brain. In
‘Approaches to primate paleobiology’, (Ed. Szalay,
F.), pp. 244-292. (Krager Publ., Basel).

Hartwig, W.C. (1993). Comparative morphology, ontogeny
and phylogenetic analysis of the Platyrrhine cranium.
PhD thesis, University of California, Berkeley.
Published in 1995 by the UMI Dissertation Services,
A Bell and Howell Company, Michigan. pp. 628.

Harrison, R.J. and Kooyman, G.L. (1968). General
physiology of the Pinnipedia. In, “The behaviour
and physiology of pinnipeds’ (Eds. Harrison, R.J.,
Hubbard, R.C., Peterson, R.S., Rice, C.E. and
Schusterman, R.J.), pp. 211-296. (Appleton-Century-
Crofts Publ., New York).

Hotye, D.A.N. (1961). The postnatal growth of the ear
capsule in the rabbit. American Journal of Anatomy
108, 1-16.

239

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

Kerley, G.I.H. and Robinson, T.J. (1987). Skull
morphometrics of male Antarctic and subantarctic fur
seals, Arctocephalus gazella and A. tropicalis, and
their interspecific hybrids. In “Status, biology, and
ecology of fur seals’ Proceedings of an international
symposium and workshop, Cambridge, England, 23—
27 April. (Eds. Croxall, J.P. and Gentry, R.L.), NOOA
Technical Report NMFS 51, 121-131.

King, J.E. (1969). The identity of the fur seals of Australia.
Australian Journal of Zoology 17, 841-853.

King, J.E. (1972). Observations on phocid skulls. In
‘Functional Anatomy of Marine Mammals’ (Ed.
Harrison, R.J.), vol. 1, pp. 81-115. (Academic Press
Publ., London, New York).

King, J.E. (1983). Seals of the world, 2nd ed. (British
Museum (Nat. Hist.), Oxford University Press Publ.,
London).

Kirkwood, R., Pemberton, D., Copson, G. (1992). The
conservation and management of seals in Tasmania.
(Department of Parks, Wildlife and Heritage:
Hobart).

Lancaster, M.L., Gemmell, N.J., Negro, S., Goldsworthy,
S. and Sunnucks, P. (2006). Menage a trois on
Macquarie Island: hybridization among three species
of fur seal (Arctocephalus spp.) following historical
population extinction. Molecular Ecology 15, 3681—
3692.

Laws, R.M. (1953). The elephant seal (Mirounga
leonina Linn.). 1. Growth and age. Falkland Islands
Dependencies Survey Scientific Reports 8, 1-62.

Lumer, H. (1940). Evolutionary allometry in the skeleton
of the domesticated dog. American Naturalist 74,
439-467.

McCann, T-.S. (1993). Age determination. In “Antarctic
seals, research methods and techniques’ (ed. Laws,
R.M.), pp. 199-227. (Cambridge University Press
Publ., London).

McKenzie, J., Page, B., Goldsworthy, S.D. and Hindell,
M.A. (2007). Growth strategies of New Zealand fur
seals in southern Australia. Journal of Zoology 272,
377 — 389.

Moore, W.J. (1966). Skull growth in the albino rat (Rattus
norvegicus). Journal of Zoology (London) 149,
137-144.

Moore, W.J. (1981). The mammalian skull. (Cambridge
University Press Publ., London).

Moore, W.J. and Lavelle, C.L.B. (1975). Growth of the
facial skeleton in the Hominoidea. (Academic Press
Publ., London).

Morey, D.F. (1990). Cranial allometry and the evolution
of the domestic dog. Ph.D. thesis, University of
Tennessee, Knoxville. Published in 1994 by the UMI
Dissertation Services, A Bell and Howell Company,
Michigan. pp. 306.

Oftedal, O.T., Boness, D.J. and Tedman, R.A. (1987). The
behaviour, physiology and anatomy of lactation in the
Pinnipedia. Current Mammalogy 1, 401-441.

Oosthuizen, W.H. (1997). Evaluation of an effective
method to estimate age of Cape fur seals using
ground fork sections. Marine Mammal Science 13,

240

On, R.T., Schonewald, J. and Kenyon, K.W. (1970). The
Californian sealion: skull growth and a comparison
of two populations. Proceedings of the Californian
Academy of Sciences 37, 381-394.

Peterson, R.S. (1968). Social behaviour in pinnipeds
with particular reference to the northern fur seal. In,
‘The behaviour and physiology of pinnipeds’ (Eds.
Harrison, R.J., Hubbard, R.C., Peterson, R.S., Rice,
C.E. and Schusterman, R.J.), pp. 3-53. (Appleton-
Century-Crofts Publ., New York).

Rand, R.W. (1949a). Studies on the Cape fur seal
Arctocephalus pusillus pusillus \. Age grouping
in the female. Progress report submitted June
1949, Government Guano Islands Administration,
Department of Agriculture, Union of South Africa.

Rand R.W. (19495). Studies on the Cape fur seal
Arctocephalus pusillus pusillus 3. Age grouping
in the male. Progress report submitted November
1949, Government Guano Islands Administration,
Department of Agriculture, Union of South Africa.

Rand, R.W. (1950). On the milk dentition of the Cape
fur seal. Journal of the Dental Association of South
Africa 5, 462-477.

Rand, R.W. (1956). The Cape fur seal Arctocephalus
pusillus pusillus (Schreber): its general characteristics
and moult. Sea Fisheries Research Institute
Investigational Report, South Africa 21, 1-52.

Rand, R.W. (1959). The Cape fur seal Arctocephalus
pusillus pusillus. Distribution, abundance and
feeding habits off the South Western Coast of the
Cape Province. Sea Fisheries Research Institute
Investigational Report, South Africa 34, \—75.

Rand, R.W. (1967). The Cape fur seal Arctocephalus
pusillus pusillus 3. General behaviour on land and at
sea. Sea Fisheries Research Institute Investigational
Report, South Africa 60, 1-39.

Renouf, D. (1991). Sensory reception and processing in
Phocidae and Otariidae. In ‘Behaviour of pinnipeds’
(Ed. Renouf D), pp. 345-394. (Chapman and Hall
Publ., London).

Repenning, C.A. (1972). Underwater hearing in seals:
functional morphology. In, “Functional Anatomy
of Marine Mammals’ (Ed. Harrison, R.J.), vol. 1,
pp. 307-331. (Academic Press Publ., London, New
York).

Repenning, C.A., Peterson, R.S. and Hubbs, C.L. (1971).
Contributions to the systematics of the southern fur
seals, with particular reference to the Juan Fernandez
and Guadalupe species. Antarctic Research Series 18,

1-34.

Ross, G.J.B., Cockcroft, V.G. (1990). Comments on
Australian bottlenose dolphins and the taxonomic
status of Tursiops aduncus (Ehrenberg, 1832). In
‘The Bottlenose Dolphin’ (Eds. Leatherwood, S.L.
and Reeves, R.R.), pp. 101-125. (Academic Press
Publ., London).

Scheffer, V.B. (1950). Growth layers on the teeth of
Pinnipedia as indication of age. Science 112, 309—
Sill.

Proc. Linn. Soc. N.S.W., 129, 2008

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

Scheffer, V.B. (1958). Seals, sea lions, and walruses: a
review of the Pinnipedia. (Stanford University Press
Publ., London).

Scheffer, V.B. and Wilke, F. (1953). Relative growth in the
northern fur seal. Growth 17, 129-145.

Scheffer, V.B. and Kraus, B.S. (1964). Dentition of the
northern fur seal. US Fisheries and Wildlife Services,
Fishery Bulletin 63, 293-315.

Scott, J.H. (1951). The comparative anatomy of jaw and
tooth eruption. Dental Record 71, 149-67.

Shaughnessy, P.D. and Ross, G.J.B. (1980). Records of
the Subantarctic fur seal (Arctocephalus tropicalis)
from South Africa with notes on its biology and some
observations on captive animals. Annals of the South
African Museum 82, 71-89.

Shaughnessy, P.D., Kirkwood, R.J. and Warneke, R.M.
(2002). Australian fur seals, Arctocephalus pusillus
doriferus: pup numbers at Lady Julia Percy Island,
Victoria, and synthesis of the species population
status. Wildlife Research 29, 185-192.

Shea, B.T. (1985). Ontogenetic allometry and scaling:

A discussion based on the growth and form of the
skull in African apes. In ‘Size and scaling in primate
biology (Ed. Jungers, W.L.), pp. 175-205. (Plenum
Press Publ., New York).

Sirianni, J.E. and Swindler, D.R. (1985). Growth and
development of the pigtailed macaque. (CRC Press
Publ., Boca Raton, Florida).

Stewardson, C.L. (2001). Biology and Conservation of the
Cape (South African) fur seal Arctocephalus pusillus
pusillus (Pinnipedia: otariidae) from the Eastern Cape
Coast of South Africa. PhD thesis submitted for the
Degree of Doctor of Philosophy from the Australian
National University.

Stirling, I. and Warneke, R.M. (1971). Implications of a
comparison of the airborne vocalisations and some
aspects of the behaviour of the two Australian fur
seals, Arctocephalus spp., on the evolution and
present taxonomy of the genus. Australian Journal of
Zoology 19, 227-241.

Todd, T.W., Schweiter, F.P. (1933). The later stages of
developmental growth in the hyena skull. American
Journal of Anatomy 52, 81-123.

Trillminch, F. (1981.) Mutual mother-pup recognition
in Galapagos fur seals and sea lions: cues used and
functional significance. Behaviour 78, 21—42.

Warneke, R.M. and Shaughnessy, P.D. (1985).
Arctocephalus pusillus pusillus, the South African
and Australian fur seal: taxonomy, evolution,
biogeography, and life history. In “Studies of Sea
Mammals in South Latitudes’ (Eds. Ling, J.K.
and Bryden, M.M.), pp. 53-77. Proceedings of
a symposium of the 52nd ANZAAS Congress in
Sydney, May 1982. South Australian Museum.

Wartzok, D. (1991). Physiology of behaviour in pinnipeds.
In Behaviour of pinnipeds, (ed. Renouf, D.), pp.
236-299. (Chapman and Hall Publ., London).

Wayne, R.K. (1986). Cranial morphology of domestic
and wild canids: the influence of developmental and
morphological change. Evolution 40, 243-261.

Proc. Linn. Soc. N.S.W., 129, 2008

Weisberg, S. (1985). Applied linear regression, 2nd ed.
(John Wiley and Sons Publ., New York).

Winship, A.J., Trites, A.W. and Calkins, D.G. (2001).
Growth in body size of the Steller sea lion
(Eumetopias jubatus). Journal of Mammalogy 82,
500-519.

Wynen, L.P., Goldsworthy, S.D., Insley, S.J., Adams,
M., Bickham, J.W., Francis, J., Gallo, J.P., Hoelzel,
A.R., Mailuf, P., White, R.W.G. and Slade, R.
(2001). Phylogenetic Relationships within the Eared
Seals (Otartidae: Carnivora): Implications for the
Historical Biogeography of the Family. Molecular
Phylogenetics and Evolution 21, 270-284.

241

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

yoyeo-Aq dod (4.00 ob Z “S.8 oW€-A.8I of Z ‘S.Sh oF €) Spunois men Og ZEAE SZ LHOOTINAd ‘7Z
881 suIpueys dod asnoyyYyst] JO q Wy Z (A.7p 0ST ‘S.7O of €) Ad “Flooy eden = THAIN. G67 S68IINAd ‘IZ
S81 surpueys dod poys yeoq JO | W O08 “(4.67 oS7 ‘S.7O ove) AV VluIpres 76 AINE LZ ~68IWAd 07
LEI “qeyo. Ul perp dod (A.6€ oST7 S.8S o€€) Youog 8, dury Jo FS ‘Yorog sIgoH 7G AINE SI I68INdd ‘61
761 SuIpuens dod (A. oST ‘S.70 ov €) Ad ‘Foo ode «= ZG AINE ET O6SIINAd ‘81
OSI Sulpuelys dod (4.6€ 0ST ‘S.8S o€€) Ad “Youog s.sury 76APIQ C88IWNAd ‘LI
0072 suIpues}s dod (A.7h oSZ ‘S.00 ov€) dd ‘SYOOU IIA §=T6IdVET 6L8INAd 91
SSI suIpueds dod (A.€7 o$Z ‘S.7O of €) dd “Wed sormeT =. 76 IdYVZ LL8IWHd “SI
661 Sulpuens dod (4.7 oSZ “S.70 ov €) dd “Floou ode =: [6 doS pZ + 898IINAd ‘FI
Lr popr09el Jou dod (A. 7p 0ST “S.70 ov €) Ad “aFlooy ode 06 OUNL GT FOLIINAd ‘EI
061 suIpuedys dod (OM) WINoY] JoaTyY shepuns foquycz O6IdVTZI 869IINAd “ZI
TOI suIpueyys dod (A.8€ $7 ‘S.ZS o€€) Yoouwepiojsuty GRAVIS LESIWAd ‘II
10Z Sulpueds dod (S.LT o$Z “Ac10 o£) MOIAPIS «= 88 19D 9T ~OOSIINAd ‘Ol
861 SuIpued)s dod (4.6€ oS7 ‘S.8S o€€) dd “Youog ssury = Bg gays LOSIWAd 6
£61 sulpueys dod Keg shougor “ynoyy Jory snoffoqey quryg gguerog ESpINdd 8
El SuIpue.ys dod (4.6€ ob ‘S.OI ove) APA ISAO §=6p8deSZI ELOINAd —‘L
ZLI SuIpuesys dod OM “YNo] Joary shepung jo quyodp €8190L SL6WHd 9
061 ayyo dod (4.0r oSZ ‘S.66 o€€) dd poomounyH =EZ 90M El 8S6WHd 'S
OLI SuIpuesys dod (DM) dep Apoo,y “yNoY JoATY skepuns jo quryce EZARWOT IS6WHd ‘YF
POI SuIpuedys dod GA WNopy Joary puepepyfomuyz eEguepi] LI6Nad ‘€
16 suIpuesys dod (A.SE o$7 “S.£0 of €) (Ad) Woqezi[Y Hog ‘SMOTTIM E86l uel OI6WHd 2
007 surIpueys dod (q4) Avg stouery 4g “YNoy] JOATY S,Wopiejg ueA FoF UT ~Q%2Gd7~Z 868INdd I

(wd) UOTI9]{[O9 FO UOT}I9][O9
7TqS JO soourjsuIMoID UOISOY uoljed0] SJeuTxo1ddy jo 93eq ‘ON GI

‘pop1odal jOU
SOOULISUINOILD - popxlOdIYy JON “(YOvog poomowIN} JJO uULd90 OY} UI SUTZeOY] PUNOF QC6INAd) YIVEp JO sasnvo Joy}0 — JoNIO ‘(€ = U) WNTIeURDD_ YWoQezI[q WOg

9Y} 1B UONe]IGQeyol SuLINp polp [euUR - ‘quYyoy UI psig ‘(¢ = U) WuLIEd OYUOIOS JopuN pojzoaT[OO [eUIUe - yIUIIEd “19g “(L¢ = U) SUONeIOdo SuIYsy SuLmMp jou
[MCI] [RIOIOUIUIOD & UI YSNed AT[e}UOpTIOUT [eUMTUe - Yd}¥d-Ag ‘(QE = U) Yoo UO peop dn poysemM [eUTIUL - SUIPUB.AYS :UOT}D9][09 JO SooURJSUNIIIO 0} AY]

(A .00 OLE “S .0S 96Z) URGING “(FPS O LZ ‘S .€0 OEE) UOpUOT sey 01 (F HZ O EZ ‘S .€0 O HE) AV Sioquoyoy “(2oUlAoIg
oded wioiseq) OW pue ‘eded uiojseq oy} Surpnyoxe ynq ‘seyjnsy ode_ Jo sea ‘(ysevoo YINOS) DS ‘(q .00 9 07 ‘S .OS O PE) SeUNSY ode_ 0} osnoyYsTT
yuIOg ade_D Jo yyNos “(jsvod jsoM YINOS) OMS “(4.67 OSI ‘SIZ OVE) esnomYySsTT JuIOg ade_ Jo yOu ‘(ysvod ySOM) DAK :UOTIET[OS JO Says 0} Ady

"LOGI Aqne
PUB ZQ6] JoquIoDaq] USEMY4Oq BOLIFY UIOYINOS FO JSBOD OY} WIOA, Pa}d9T[OO O10M STRUIUY ‘ApNjs siy} Ul poulUeXo (¢g = U) Sees IM] ULOTIFY YINOS :] XIGNAddV

242

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

yoqyeo-Aq
yoyeo-Aq
yoqyeo-Aq
yoyeo-Aq
sulpueys

“qeya1 Ul perp

suIpue.s
sulpueys
suIpueys
sulpuels
sulpueys
sulpueys
SUIpUIS

“qeyor Ut palp

sulpueys
SuIpueys
yoyeo-Aq
Ssulpue.ys
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
SuIpueys
sulpuels
yoyeo-Aq
yoqeo-Aq
SuIpue]ys
SuIpueI}s
SuIpue]ys
sulpueys
yoye0-Aq
yoye9-Aq
yoqyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq

dOd
dOd
dod
dOd
dOd
ueging
dOd
dOd
dOd
d0d
dOd
dOd
dOd
dOd
dOd
dOd
dOd
dod
dOd
dOd
dOd
dO
dOd
dOd
dod
dOd
dOd
dOd
d0d
dOd
dod
dO4
dOd
dO”
dOd
dOd
dOd

(stouv1y 1S aded Jo § wu (¢ ‘9) spunoIs [Me] OW
(stouely YS odea Jo S wu O€ )) spunois [Mey OH
(stoueLy 4S odea jo S wu O€ 4) spunois [Med OH
(slouely 1S oded Jo S wu (¢ ‘9) Spunols [Me] OW

(A.LT 0ST “S.10 of €) MATAROS

(4.00 oT € “S.0S 067) ueqing

(4.7Z€ oSZ ‘S.7Z0 oh) doysioyeuusoyos

(A.07 o€7 ‘S.€0 ob €) APY SIoquonold

(A.7 o$7 ‘S.2O of €) Ad aflooy eden

(A.7h o$7 ‘S.70 of €) Ad ‘aflooy ode

(A.LI o$7 ‘S.10 of €) MOIAROS

OM “UNoP] JoATY shepung fo F UP GE

OM “YNoYy JoATY skepung Fo q Ur Op

(4.7r oSZ ‘S.00 o€) Ad ‘PuensiowuINg

(4.61 .97 ‘S.9r o€€) AV dep Apoom

319q90¥ ‘(4.7 o€7 “S.LO of €) Av S1aquopeyd

(stouely 1S ade Jo § wu Y¢ *d) Spunols [Me OW

(4.7 0ST ‘S.7Z0 of €) Ad ‘aFtlooy aden

(4.81 of ‘S. Lv oP €.A.87 of Z “S.Sh of €) SPUNOIS [Mey OF
(4.7E of 2 ‘S. bh oF E.AcIZ of ‘S.9F of) SPUNOIS [ME OY
(4.07 of Z ‘S. Sb of €.4.67 of 2 ‘S. py of €) SPUNOIS [MeN OF
(F.07 of 7 ‘S.S¥ oF £.4.670 of C ‘S. th o£) SpuNoIs [Mey OF
319q90- “(A.S7 o€7 “S.L0 of €) Avg S1oquoyetd

(4.97 o$Z ‘S.10 of €) Youog Woysom “Avg TUby

(F.0b o€7 ‘S.0$ of €-A.LT o€7 “SES of €) Spunoss [men Og
(4.87 ofZ ‘S.60 oS €) Spunois [men OY

(A.ZE $7 ‘S.70 of €) doysroyeuusoyog

S[OOd 190 “(ALT oSZ “S.10 of €) MOTAROS

(A.ST 097 ‘S.0$ o€€) (AV) Avg eosTy ‘syooy youlg

(A..0€ OF oST “S..07.65 o£€€) YORO_ Yo]]od JO AS ‘sodig sy

(A.85 0ST ‘S.€7 oF €-A.H0 097 “S. ET of €) SPUNOIS [Me OF
(4.20 097 ‘S.ST of €-A.0S oSZ ‘S.H7 of E) SPUNOIS [Me OY
(A.S7 of 7 ‘S.9P oF E-AcII of Z ‘S. Lh of €) SPUNOIS [Me OF
(.65 of ‘S.LE oF E-AcSP of Z ‘S. 1h of €) SUNOS [Mey Og
(AS of ‘S.8€ of €-A.Cr of 7 ‘S. 1h of €) Spunols men OF
(A.7p of 7 ‘S.Tr of E-AIS of Z ‘S.7 OE) SPUNOIS [Me OG
(4.0F ofZ ‘S.7h of €-A.ST of Z ‘S. SP of €) SPUNOIS [MEQ OW

v6 0 8 ALScCINAd

76 8NY LZ
76 SNY LZ
76 SNY TZ
76 SNY ZI
7661 Aine
76 Ajne ¢
76 Ajng

76 Ane ZI
76 49d €
76 ULL IZ
76 uel LI
76 UeL LT
76 uel ¢
€6 99C 07
€6 sny LI
€6 Aqng

€6 Aqne 61
€6 OUNL 67
€6 oUNL BZ
€6 OUNL BZ
€6 UNL BZ
€6 UNL g
€6 ung /
€6 API OZ
€6 API 61
€6 APN O€
€6 Ae 87
€6 JEW 61
€6 JEW TI
76 das ¢Z
76 das rI
76 BNY CZ
76 BNY ZZ
76 8NY FI
76 8NY $1
76 BNY €]

vSCCINdd
eScCINdd
CS CCINAd
8VCCINdd
8EcCINdd
LOcCINAd
861 CINdd
L61CINAd
IST CINdd
ev CWNdd
IVI CINdd
OVI CINdd
Le lcNdd
celCINdd
L80CINAd
cC80CINAd
180cINdd
VSOCINAd
eSOCINAd
cSOCINAd
LSOCINdd
OSOCINAd
6V0CINAd
8VOCINAd
9P0CINAd
SvOCWNAd
vrocWadd
9E0CINAd
CeOCWNAd
VIOCINAd
elOcWdd
O10cCINAd
600CINAd
8007CINAd
LOOCINAd
900CINAd

‘6S
“8S
‘LS
9S
wSS
‘VS
te&
GS
‘TS
‘0S
‘60
‘81
LY
‘OV
“Si
VV
“eV
‘Ww
Tv
‘OV
“6E
“BE
LEE
Ne
SE
VE
tele
IGE
alts
‘OE
‘6C
‘87
LG
YG
SG
‘VC
ie

243

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

06

6r 1
OIT
esl
801
08

CII
SM
66

OLT
vel
OST
CON
vcl
srl
Salt
bri
cr
8SI
Lél
CLI
coy
$8

81

yoye9-Aq
yoyeo-Aq
sulpueqs
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoqeo-Aq
yoyeo-Aq
yoyeo-Aq
SuUIpue.)S
yoyeo-Aq
yoyeo-Aq
yoyeo-Aq
yoqeo-Aq
yoyeo-Aq
yoyeo-Aq
yoyeso-Aq
yoyeo-Aq
ued ‘19s
jued ‘19s
SuIpUe.ys
quuied “19s

OM
OM
OM
dOd
OM
OM
OM
MS
OM
OM
MS
OM
OS
OS
OM
IS
IS
OM
OMS
OM
OM
OM
OM
OM

Aeg eusjoH 3S

(A.8€ oLI “S.L7Z o7E) AP PUSTSH 1S SIOYS]FO
asnoyYsrT snoudwuio}s s10ysyjO

(4.0€ o€7 ‘S.0€ of €) AUG s1oquone] gd s10YsJO

(4.L0 oLI ‘S.91 o€€) AbG euYyeples s104syO

Keg 110) (8 | 1S “pueys] [B9S

(4.vI oLI “S.O1 o€€) Abg euyepyes s10ysyO

(4.96 o81 ‘“S.0€ oS€) €6h PHS sulysyY [esiowiogd
euoloH 1S

(4.0 oLI “S.O€ o€€) PULIS] Ussseq Jo YING

(4.6P o1Z ‘S.St ov €) Wlod odeD HO

IMoqieyUMOT, odeg

(A.€S ol7 ‘“S.97 ob €) YINOW JOATY ZjLINOH Fo yNos wu YZ
(A.1S ol7 S.€7 of €) YINOW JOATY Z}IMODH JFo wu ¢-¢

(4.SE oLI ‘S. St oLE) PUrIS] Usssed fo sow
Aeg [Ossoy] JO SOM WU $7

SSOUOPTIM\ JJO SO|TU g

(4.0 oLI “S.1Z o€€) PURIS] USSse SI0YSYO
TOS Pls SuIYsY [esIowied

(A.0€ o91 “S.O€ oI €) Pore Ieeqouliog

JOZUIOT YI

dOZULOL

Aeg euajaH 3S

(4.90 o81 ‘S.81 o€€) Bore SUT[OPUOA OY} STOYS]FO SopTUt Z

16 AON
L6 Ayne ST
L6 Alng €Z
L6 uel pI
96 UNL OT
96 Aine OT
96 dog gz
96 des €1
96 3ny HI
66 das ST
$620 LI
76 UeL LT
06 190 €7@
06 Le €1
Lg Ayne LI
9g ouns ¢
98 Idy ZZ
gg oun 9
gg Ajne LZ
$8 99 O1
78 994 ZI
78 AON +1
76 des 0¢
78 dag ¢Z

SVISWOW
9ETSIOW
SeISWOW
ec I SNOW
COOSIIOIN
666VINOWN
966VINOW
CO6VINON
6867NOW
L6OSVINOW
S6SPNOW
LLSVWOW
88EVINOW
S9EVINOWN
9EIEWOW
L8SEWOW
98SeNOW
C8S ENON
S6LC NOW
c9LCWOW
OI8INOW
608 1LINOW
98L INOW
S9SIWOW

€8
C8
‘18
‘08
‘6L
‘8L
(Ze
OL
GE
VL
ioe
ie
WE
‘OL
‘69
‘89
HS)
Of)
$9
‘V9
{)
1)
AS)
‘09

Proc. Linn. Soc. N.S.W., 129, 2008

244

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

00:0
VN
VN
VN
VN

VN
00:0
00°0

100°0
000
000
00°0

VN

00:0

YN
00:0
00:0

VN
000
000

VN

VN
00°0

OL
VN
VN
VN
VN

VN
69
OL
OL
CL
IL
We

VN
IZ

VN
es)
IY)

VN
69
69

VN

VN
CL

ge

earl

— os =<
ZAAZA

ask
Z

V
aa
BAe oeee

ay

stsoujodAy
SATIOUIOITV

ANOWO]TV

(00°0) 160
(00°0) €6'0
(00'0) 760
(00°0) €6'0
(00°0) 160

(00°0) €6'0
(00°0) £3°0
(00'0) £80
(00'0) 680
(000) ¥Z'0
(00°0) 160
(000) Z6°0
(00°0) LL'0
(00°0) 98°0

(000) 0r'0
(000) 980
(00°0) 88°0
(000) 06°0
(000) $80
(00°0) 88°0
(00'0) 260
(00°0) €6'0
(00°0) €6'0

(d) 1

c0'0 + 9£'0
c0'0 + CL'0
cO'0 + 85:0
100 + £50
c0'0 # 790

c0'0 + $9°0
v0'0 + 89°0
y00 + 09'0
S00+ P80
S00+ br 0
c0'0 + 850
£00 F790
£00 +970
£00+090

c0'0 + O10
vO0+LS0
£00 + 0L0
£00 *S9°0
v0'0 + 69°0
y00 + 190
c0'0 + $90
c0'0 + SS'0
cO'0 + SS'0

‘aS +
adojs

800 + CSC
ILO+FSTT
600+ 9L I
L00 + 9€C
C10 ¥ 691

800 + 881
810 + C0'0-
610+ 170
STO F9I T-
LTO + SV'0
O10 +F PL I
910+ 850
STOOL I
S1TO+F9S'T

60°0 + S6'€
810201
910 + 70'0-
€l0+FS80
0c 0 + 610
610 +910
GOs Cl
IN Ose IS @
800+ S97

‘aS +
jdoo101U]

UOISSOISOI IVOUTT

CL
CL
GE
CL
VL

CG
IL
Gu
GE
vL
ole
€L
tell
tell

We
$9
69
69
OL
We
iI
tell
VL

prosAiojd Jo puog 0} uoIseg ¢Z7dq
YIpeslq Prose Cod
UMS hEe| AMEND) Ec
(1O1IN}UL) JOOI DI}BUTOSAZ 0} UOISeY YZ
Yypesig oneWIOSAZIg 61d

ssod0i1d prous]s}sod
ORS PLO GTP OTOMO OLS UD rst tcl
¢ ouIURD}sod je oyeJed Jo IWpeolg 1d
€ ouIuRdjsod ye aye[ed Jo UIpeoIg OI d
[| oulues}sod ye ojeyed Jo yIpeolg S{d
R][IXPUI FO O01 SVVUIOSAZ JO UIPeolg pd
BI[IXUI JO pUd JOTIN}SsOd 0} UOTIeUD E[
YIpeoig suTURdIG 4S9}VOID ZI d
MOI uIURD}Ssod 1addn Jo yWSUIT [1d
SIOSIOUL 0} Y}OU [RIe[ed Old

jeqeed

asvo UleIg JO YIpeolg 6
sassoooid jeyiqiovidns ye ypeoig Bq

UOTJOLNSUOD [e}IQIOIOUI JseoT LG
sassasoid [eyIqioo1d ye ypeslg 9q
s[eseu JO YSU] Jso}VaID Cq

SOIVU IOLIOJUL JO YIPIM yso}eOIH pq
sjeseu JO pud IOLIA}sod 0} uOTIWeUy Eq
jSO10 [e}1d1990 Jo d[Ppru 0} UOTJeUH Z|
ysug] jeseqojApuoD [qd

[esioq

a[qeiieA yuopuedaq

‘JULITFIUSIS JOU SOYIUSIS SU “JOAS] SQ'd > d ay Ie paisa, “T <§:TH 10 | > 21TH ‘SA | = y:0H{ :sodojs

UOISSOISII UO BUT}SO} SIsoyJodAH ‘YyISua] Apog piepuLjs SO] UO s}USWOINSKOL [[NyS (7E—-8ZN) Je[Nqipuew pue (7-7) [esAIR] ‘(€Z—OT d) [eyed
(6-1) [eS1op So] 107 ANoWOT]e pure ‘s}UdTOYJI09 UOTe[O09 Jopso-yueI ueULIeadS ‘suoenba sul] yYysreNs sorenbs jseo] JSNQOY, *7 XIGNAddV

245

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

000

VN
IL

I>9°H

(00°0) 760
(00°0) €6°0
(00'0) €8°0
(00°0) $80
(000) 760

(00'0) €L'0
(00'0) 060

(00'0) 760
(00'0) 16:0

£00 + S60
700 + 90'T
c0'0 + 9£°0
c0'0 + 670
c0'0 + 69°0

00 + €9°0

c0'0 + 89°0
£00 + L9'0

L10 + 68°0-
610 + Sv I-
NEO 42 2Xo |!
CLO FOL T
60°0 + 8S I

0c 0 +0

I10+680
vVILOFS80

OL
GUE,
69
v9
vL

gs
19

tell
ou

SNOPIOUOIOS 0} SIIe[NSUY TEIN
SNALOUL 4 I[QIpURUL FO JYSIOH [EW
MOLI QuIUeD}SOd JAMO] JO YISUNT OSW
MOL Y}00} Je[NgIpueW Jo SUIT 6ZIN
o[qipuew Jo yysusT 87

FET PUEYN

jSO1O [VISES JO JYSIOY LZ]

PIOJSLUL FO WIO}}0q ye |[NYS JO 1YSI9H OTT
ssooold jeyIg1001d

BfO USING] OMA OL CO EN TD) SEI
[PM GIOVAZUL UOUULIOF 0} UOT}LUD p77]

[e1o}e T

Proc. Linn. Soc. N.S.W., 129, 2008

246

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

00:0
VN
VN
VN
VN

VN
100°0
95°0
000
L10
VN
00°0
VN
VN

VN
VN
00:0
000
00°0
810
000
VN

(We
IV
18
VN
08
VN
VN

VN
VN
OL
SMe,
ie
[Le
8L
VN

a)

ewe
VN
VN
VN
VN

Vv

T<ti)e
su <§f:
te
suT=g:
su. > §J:
<tc
Vv

W

ZZAn een

VN

VN

T<@-
<<
<=
su. <Q:
L<o-
W

agaclacfiiocMectac

stsoyjodAy
SATIVUIOITV

AOWOT]V

(000) 06°0
(000) $60
(00'0) 76°0
(00'0) L6°0
(00°0) 68°0

(00'0) 96°0
(000) $8°0
(00'0) £80
(000) 68°0
(000) SZ’0
(00°0) 960
(00°0) €6':0
(000) LL°0
(00'0) 68°0

(00°0) 6€'°0
(00'0) €8'0
(00'0) 680
(000) 260
(00'0) 880
(00°0) £3'0
(00'0) $6°0
(000) L6°0

(d) a

c0'0 + £9'0
£00 + 67C I
£00 + 00'T
c0'0 + 96'0
yo0+ CIT

00 + OTT
900 + 07 I
900+ 70'I
60°0 + Lv
60°0 = $8 0
c0'0 + 90'T
S00F FIT
v00 + 780
v0'0 + LOT

£00+ZLI10
900 = £01
900 + L71
yO0+9IT
900+ L71
90°0 + 801
yoO+8IT
£0 0+ 10 T

ASF
adojs

C1 0+ 680
MOF CCE
910 + 9L 0-
80°0 + 91 '0-
IC We SE I=

Gl OFF SI I=
ite (Se [Lt c=
S£0+ CHC
80 + 68 F-
8r 0 + L6 I-
11 0+ 90'T-
970 + 8E C-
€COF VS 0-
OC O * CC I=

610 ¥ 87 &
ve 0+ 99 T-
0€ 0 + 9C €-
10 +61 C-
Me Oe IC e=
Ge a= CS) G
61:0 = 10'C-
ENO 3 C6

aS+
jdoo101U]

uOISsoIso1 Ivoury

18
18
18
18
€8

LL
08
tel
ele
€8
C8
C8
c8
(a:

08
vL
8L
8L
8L
6L
08
6L

plosArayd Jo pusq 0} uoIseg ¢7q
YJPeerg Prose Ccd
Uj pecic | EUeATE Ded
(10119}UB) JOOI SVUIOSAZ 0} UOISeg 07d
Ypeeig oyewWosAZIg 61d
ssooo1d prousys}sod
JO JOp1Og) pully Oy UOTFEUD Bid
¢ ouTURdjsod ye ajeyed Jo uIpeorg Ld
€ ouTuKsod je aye[ed Jo Wpeolg 91d
[| ouIuedjsod je ayeyed Jo Wpeolg Sd
PI [LXeUL JO JOOI SeUIOSAZ Jo HWIpeoag pd
PI]IXPU JO pUd IOLIO}SOd 0} UOTHVeUD ET
YIpeoig suTURdIg 1so}VIID TZ d
MOI ouIUeD\sod Jaddn Jo ysus7T [1d
SIOSIOUL 0} YD}OU [eIeIed Old

[eye [ed

Qsevo UIeIG JO IWpeolg 6C

sossoooid [eyIqioeidns ye yIpeolg gq
UONOLYSUOD [e}IGIOIO}UI sea] LG
sossoooid [eyIq1oold je YIpeoig 9q
s[eseu JO YSU] JsaVROID Sq

SOIVU IOLIOVUL JO YIPIM Sa}VOIH pq
s[eseu JO pud JOTIO}Sod 0} UOTJeUDH Eq
jSO10 [e}Id1990 Jo d]pprut 0} UOTyJeUH 7q

jesioq

g[qeliea yuopusdeq

‘Z xipuoddy oj se sodojs uorssoi1se1 Uo Sur}so} stsoyjodAH “YSug] [eseqo;Apuod Bo] UO syUOWIOINSeoU |[NyS (TE—-SZIA) Je[Nqipuew pue (L777 J) [eae] “(EZ
—01d) fexeyed ‘(6-1 d) [esiop Bo] 10} AouO]]e pue ‘syUsTIyJoos UONLIO1I09 Jopio-yueI ueUIeads ‘suonenbe oul] jysIes sorenbs ysevo] ASNGOY, “€ XIGNAddV

247

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

00:0
00:0
00:0
c00°0

VN
VN

VN
VN

VN
VN

(00°0) $60
(00°0) $60
(00°0) ¥8°0
(00°0) $80
(00°0) 86°0

(00°0) 08°0
(00'0) +60

(00°0) L6'0
(00'0) +60

S00 + 99’
90'0 + 881
OOF 890)
00 + 88°0
c0'0 + CCT

900+ 61 T

COO FST!
£00 +97 I

KG (0) = I S-
Ge OF (Me Os
810+ C70
610 + 19°0-
O10 + 8S I-

GE) 3 OS Ie

VIO + 60 C-
AO eAOKGs

6L
18
8L
€L
c8

09
OL

c8
c8

SNOPIOUOIOD 0} SIIVINSUY ZEW
SNVOU Je d[QIPURL JO JYUSIOH [CI
MOI QuUIUeD}SOd IaMO] JO YISUNT YEW
MOI Y}00} Je|NQIpueUl Jo YISUNT 67
o[qipuew Jo ySUST 7

Je[nqipueyy

}SO1O [B}ISeS JO JYSIOY| LZ]

PIO}SCUI JO W10}}0q Je [[NYS JO WSIOx O77
ssoooid [eyIqs1o0o1d

JOMSps0G purty] oy Tomer [C7
[e{IqIOKIZUL UOUTLIOF 0} UOTILUD 7]

[B19}8T

Proc. Linn. Soc. N.S.W., 129, 2008

248

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

(00°0) SL°0 00°0 + Z0'0 OOF IY Ws prosdiayd yo pusg 0} uoiseg €7d
(00'0) 16°0 00°0 + 900 COOF LEY 0S Ypeosq ploysey| 77d
(00°0) 68°0 00°0 + 70°0 70'0 9E'P is yypeorg [eLeATeD 17d
(00°0) 06°0 00°0 + 70°0 COO FLL Y 6 (JOTIO}UL) 1001 SIeUIOSAZ 0} UOISe YZ
(00°0) 680 00°0 + S0°0 Z0'0 + 0S'P IS yypeerq oyewosxzig 61d
(00'0) 260 00°0 + $00 70'0 + 6L'P Op Ssoooid prous]sjsod Jo Jops0g pulY o} UOIIVUD gTd
(00°0) vZ'0 00°0 + S0°0 £00 + S0'E 6 ¢ ouruesysod ye oyeyed Jo ypeorg L1d
(00°0) LL°0 00°0 + 70°0 €0'0 + 68°7 ev € suTURo}sod ye oye[ed JO YIpeolg 91d
(00°0) 68°0 00°0 + L0°0 COO FLST 4 | ourueojsod ye aqeyed yo yipeorg Sid
(00°0) LS°0 00°0 + £0°0 €0'0 + Sv'7 (és EI]IXBUT JO JOOI SIeUIOSAZ Jo YIpeoIg pId
(00°0) €8°0 00°0 ¥ 70°0 700 F9E' Ws P]]IXeW JO pUd JOLIO\sOd 0} UOTYIUD ET d
(00°0) £8°0 00°0 + S0°0 70'0 ¥ 8r'€ Ws YPeorg SuTURII JSO}LOID TI d
(00°0) 65°0 00°0 + £0°0 €00 + 8L'€ Ws MO ouTURD\sod Joddn Jo yysuST [1d
(00°0) 780 00°0 + S0°0 600 Fv r Os SIOSIOUT 0} YO}OU [eIL[ed Old

jeqeyed
(00°0) 1€°0 00°0 + 10°0 10:0 + 8p 1S aseo ules JO YIpeolg 6d
(00°0) $8°0 00°0 + S0°0 €0'0 + 09°€ SY sessoooud jeyqrovidns ye yypeolg 8q
(00°0) 68°0 00°0 + 90°0 €0'0 = 60°€ 6 WOTILYSUOD ]eGIOIOUT seo] LG
(00°0) £80 00°0 + S0°0 ZOO FLL'E Ws sassaooud jeyiqioord ye yypeorg 9d
(00°0) 920 00°0 + S0°0 €0'0 = 0€E'€ LY seseu JO YysUE] Sq}BoID Cq
(00°0) 18°0 00°0 + 70°0 €0'0 + S67 6 SoIVU IOLIO}UL JO YIPIM yso}VoIH) |
(00°0) £80 00°0 + S0°0 20'0 + 70'P 6 sjeseu JO pud IOL1a}sod 0} uoIeUD Eq
(00°0) 760 00°0 + 70'0 70'0 + LOY 6 JSO19 [eIIdI990 JO s[Pprur 0} UOTWeUH Tq
(00°0) 68°0 00°0 + 70°0 700 # SI'S ifs yysua] [eseqojApuoD 1q

[esiod

‘aS F ‘aS F
(d) 1 adojs 1do010}U] u

249

UOISSOISOI IOUT] g[qeliva juopuodsq

(A) ode uo sjuowommseour [[Mys (ZE-SZIN) Je[Nqipuew pue (L7—p7T) [esOIe] “(€Z—OI d) [eseyed “(6-1) [es1op Soy 10y
SJUSTOYJOOS UOTje[OI109 Jopio0-yueI ueuLIveds pue suoyenbo oul] yYsIeNs sorenbs svg] .JSNqoY, ‘’ XIGNAddV

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

(00'0) L380

(00°0) 88°0
(00'0) 060
(00°0) 920
(00°0) SL'0
(00°0) 16°0

(00°0) 98°0
(00'0) 64'0
(00°0) £80
(00'0) 18°0

00°0 + L0'0
00°0 + L0°0
10'0 + 80°0
00°0 + £0°0
00°0 + 700
00°0 + $00

£00°0 + S00
00°0 + $00
£00°0 + S00

cOOF ESL

€0'0 Cre
vOOFIEE
C0'0 + 6S°€
C00 ¥ 68°
c0'0 + OLY

CcOOFLIV
€0'0 + C6'€
€0'0 + 88'€

yysua] Apog 807

SNOPIOUOIOD 0} sIIRNSUY TEIN
SHILOU Je I[QIPUL JO JYSIOH [CW
MOL ouTURD\SOd IaMO] JO YISUST YEW
AMOI YOO} Je[NgIpuew Jo yIsUST 6ZIN
e[qipuewl Jo yysueT BZN

HE] UST LENA

jSO1O [e}ISVS JO JYSIOH LZ]

PIO}seU FO WO}Og ye |[NYS FO 1YSIOH OTT

ssoooid [ejIq10o1d Jo JopIog pulY Oo} UOTUJeUDH SZ]
[eUQIOVIJUI UOUUIIOF 0} UOIYJLUD p77]

jeqoyeT

Proc. Linn. Soc. N.S.W., 129, 2008

250

C.L. STEWARDSON, T. PRVAN, M.A. MEYER AND R.J. RITCHIE

(Z8) 00'1 (18)S8°0 (08) 880 (€Zd) J&A
_ (18) 58'0 = (Z8) 001 ~=—- (08) L6'0 (77d) 1A
(08) 88°0 (08) 460 (78) 00'T (17d) 1eA

(€Zd) HA = (ZZ) RAS (TZ) FRA

winiuesorseg
(€8) 00'T (€8) «ILO (61d) J&A
(€8) xIZ0 (€8) 00'°T (pId) IeA

(61d) 1A = (I'd) FRA

yoie oeul0sA7

(09)001 (6S)Sz70 (LZ) eA
(65)S70 (08) 001 (6d) 1eA
(L717) eA (6) 1eA

WNIUBIOOINON

(84) 001 (6L) «180 (Sd) eA

(9L) «180 (62) 001 (pd) eA
(Sd) eA (7d) 1eA

[esen

(SL) 001 (IL) €6':0 (8d) eA

(IL) €60 (62) 001 (Ld) 1eA
(8d) eA (Ld) eA

jeyUuOL yy

‘() sjoyorsg Ul OZIS o]duues [e}0} SY} YJIM [OAS] SQ'O OU}
J@ JULOIIUSIS SI UOTJLTOIIO OU} IVY} SOJOUSP ,, [OQUIAS “UOITOI 0} SUIPIOION So[QeIIeA [NYAS BO] IO} syUSTOYJZOOS UOeO1I09 Jops1oO-yueI UeULIVSdS *¢ XTGNAddV

251

Proc. Linn. Soc. N.S.W., 129, 2008

AGE AND GROWTH IN SOUTH AFRICAN FUR SEAL SKULLS

(08) 00'I
(TL) +$6°0
(ZL) «€6'0
(6L)%€8'0
(08) «160
(6L) «790
(6) «SL'0
(L1d) 1eA

(TL) «S60
(€L) 00'T
(ZL) «£60
(ZL) «?8'0
(EZ) «S60
(ZL) x9L'0
(ZL) *8L°0
(9Td) 1eA

(ZL) «£60
(ZL) «£60
(€Z) 001
(ZL) «780
(€L) ¥€6'0
(ZL) x9L'0
(ZL) x8L'0
(SId) 1eA

(08) 00'1
(08) £60
(94) 08°0
(TL) 08°0
(6L) 88°0
(ZEW) FEA

(6L) «£80
(ZL) «780
(ZL) «¥8'0
(Z8) 00°I
(18) +06'0
(18) +620
(18) «160
(€1d) eA

(08) L6'0
(Z8) 00°1
(82) 8L'0
(ZL) 780
(18) 38°0
(TE) eA

(08) «160
(€L) «S60
(€L) «£60
(18) «06°0
(Z8) 00'I
(18) #920
(18) x8'0
(ZI d) IA

(9) 08'0
(82) 8L°0
(6L) 00'I
(69) 760
(8Z) 6L'0
(O€W) FRA

(Z8) 00'T
(18) «860
(08) «960
(S71) JeA

(6L) x90
(ZL) «9L'0
(ZL) «9L'0
(18) «64'0
(18) #920
(Z8) 00'1
(18) +920
(11d) 1eA

(IL) 08°0
(CADE HW
(69) 760
(pL) 00'I
(€Z) €8°0
(67) FRA

(18) «86°0

(Z8) 00°1

(6L) «460
(PZT) eA

(6L) xSL'0 (LI1d) 1eA
(ZL) «8L'0 (91d) eA
(ZL) x6L'0 (SId) eA
(18) «160 (€1d) JeA
(18) «?8'°0 (ZI d) J&A
(18) «940 (11d) 1eA
(Z8) 00'1 (OI d) 1eA
(OId) 18A
gee d
(6L) 880 (ZEW) JeA
(18) 88°0 (TEW) 1eA
(82) 6L'0 (ON) 1eA
(€L) €8°0 (6ZN) 8A
(€8) 00'T (SZIN) eA
(SZIN) eA
a]qipuryy
(08) «960 (S77) eA
(6L) x60 (PZT) IeA
(08) 00'1 (€d) eA
(€q) IeA
umnrue1oouyoueyds

Proc. Linn. Soc. N.S.W., 129, 2008

DSD

Records of the Inland Carpet Python, Morelia spilota metcalfei
(Serpentes: Pythonidae), from the South-western Slopes of New
South Wales

DaMIAN R. MICHAEL! AND Davip B. LINDENMAYER

Fenner School of Environment and Society, Australian National University, ACT, 0200
‘Corresponding author: 11 Briwood Court, Albury, NSW, 2640 (Email:michaeldamian@hotmail.com)

Michael, D.R. and Lindenmayer, D.B. (2008). Records of the Inland Carpet Python, Morelia spilota
metcalfei (Serpentes: Pythonidae), from the south-western slopes of New South Wales. Proceedings of
the Linnean Society of New South Wales 129, 253-261.

Location records of the Inland Carpet Python Morelia spilota metcalfei were collated from the south-
western slopes of New South Wales from scientific literature, published reports, landholder questionnaires,
public information sessions, informal conversations and field observations. Fifty-three records, encompassing
a minimum of 95 observations were obtained. Twenty-nine records (58%) and 57 observations (69%)
originated from granite outcrops. High priority conservation areas for this species in the SWS include;
inselbergs such as Goombargana Hill, Gerogery Range and Nest Hill, the granite belt between Kyeamba
and Wagga Wagga, large vegetated ranges such as Yambla Range and the Rock Nature Reserve and the
riverine environment along the Murray and Murrumbidgee Rivers. Future conservation of M. s. metcalfei
habitat in the SWS will require appropriate management of granite land forms with particular focus on
strategic grazing, pest animal programs and fire control.

Manuscript received 20 October 2007, accepted for publication 6 February 2008.

Keywords: conservation, Inland Carpet Python, Inselbergs, granite outcrops, Morelia spilota metcalfei,

south-western slopes of New South Wales.
INTRODUCTION

The Inland Carpet Python Morelia spilota
metcalfei is one of three sub-species of Morelia
spilota that occur in New South Wales. The Diamond
Python M. s. spilota is confined to the east coast of
Australia, ranging from south of the Victorian border
to the northern rivers region of NSW and extending
inland to the Great Dividing Range (Swan et al. 2004).
In the northern part of its range it intergrades with the
Coastal Carpet Python M. s. mcdowelli (formerly part
of M. s. variegata) near Coffs Harbour (Shine 1994,
Swan et al. 2004), extending north to Cape York and
west to the Great Dividing Range (Wilson 2005).
The Inland Carpet Python M. s. metcalfei (formerly
part of M. s. variegata) occurs inland of the Great
Dividing Range, extending from central Queensland
in the north to the Warby Ranges region in Victoria
(Coventry and Swan 1991, Heard et al. 2005) and
west into South Australia to the Eyre Peninsula
(Schwaner et al. 1988). It is geographically isolated
from the other two sub-species, although a potential
contact zone with M. s. spilota may have once existed
in the Hunter Valley region (Shine 1994).

General habitataccounts of M. s. metcalfei indicate
that it occurs in most vegetation types throughout its
range including swamps but excluding treeless plains.
It commonly frequents large trees in River Red Gum
Eucalyptus camuldulensis forest, eucalypt woodland
and mallee vegetation types (Robertson and Hurley
2001), often in association with watercourses (Shine
1994) or granite outcrops (Heard et al. 2005). It utilises
a range of micro-environments including hollows in
trees and logs, rock crevices, disused rabbit burrows
(Heard et al. 2005) and occasionally roof cavities in
rural buildings (Shine 1994).

Morelia spilota metcalfei is the largest snake
species in New South Wales, growing to over 2.5 m in
length and occasionally attaining 3.5 m (Kortlang and
Green 2001). However, basic information regarding
geographical distribution, population density and
habitat preferences remain poorly documented
in NSW (see Heard et al. 2005 for Victorian
populations). Evidence suggests that populations of
M. s. metcalfei have declined considerably over the
last 100 years (Shine 1994, DSE 2003), even though
sightings from new regions are still being reported
(Morris 1993). In Victoria, it is listed as Endangered

RECORDS OF THE INLAND CARPET PYTHON

Legend

e) Towns

eB Python Locations

Wagga Wagga ©

VValoundne

2 Holbrook

Albury
e

e
oe ee
© Gundagai
e

© Tarcutta

10 0 10 20 Kilometers

Figure 1. Location records ( n = 53) of the Inland Carpet Python Morelia spilota metcalfei in the south-
western slopes biogeographical region of New South Wales (including three records from near the Vic-
torian border) based on field observations, published reports, landholder and public questionnaires and

personal communications.

(DSE 2003) and in the western division of NSW it
is considered regionally endangered (Sadlier and
Pressey 1994, Sadlier 1994). In the Victorian wildlife
atlas database, 160 formal location records of the
species exist (DSE 2003) and although anecdotal
reports suggest pythons may be relatively common
along vegetated river systems, there are far fewer
records for the NSW Murray catchment area in the
NSW wildlife atlas database (DECC 2007). However,
during a survey of 105 landholders in the Coleambally
region, 29% of farmers claimed to have seen M. s.
metcalfei on their property (Doody et al. 2004). In
this paper we document historic and current records
of M. s. metcalfei in the upper Murray catchment
area, specifically the south-western slopes (SWS)
biogeographical region of New South Wales (sensu
Benson 1999).

254

METHODS

We collated records of M. s. metcalfei within the
SWS of New South Wales, an area encompassing
major towns such as Albury and Wagga Wagga
and smaller townships such as Walbundrie, Walwa,
Gerogery, Howlong, Tarcutta and Gundagai (Fig.
1). Five distinct methods were used to obtain M. s.
metcalfei records. These were:

(1) Literature review: Records were obtained from
the scientific literature, published reports, the New
South Wales National Parks and Wildlife Service
wildlife atlas database, Victorian Department of
Sustainability and Environment wildlife atlas and
Bionet databases.

Proc. Linn. Soc. N.S.W., 129, 2008

D.R. MICHAEL AND D.B. LINDENMAYER

(2) Landholder_ questionnaire. A total of eighty-
four landholders involved in two long-term wildlife

monitoring programs (Lindenmayer et al. 2001,
Cunningham et al. 2007) and an intensive study
of granite outcrops in the SWS (Michael in prep)
were shown photographs of M s. metcalfei from
field guides and local specimens and asked if they
had ever encountered this or other sub-species on
their property. In addition, an article on python
habitat requirements and a request for information
on sightings in the region was printed in the Murray
Catchment Management Authority (CMA) newsletter
and distributed to members of the West Hume,
Culcairn, Holbrook, Upper Murray and Kyeamba
Creek landcare groups.

(3) Public presentations. As part of an extensive
education program aimed at informing landholders
and the wider community on the habitat requirements
of local wildlife, participants were shown photographs
of M. s. metcalfei and asked if they had encountered
this or other sub-species in the region. Between 2004
and 2007, we held 14 information sessions, mostly
involving local landcare groups, field naturalists and
interested members of the public. Over 500 people
took part in the presentations.

(4) Informal conversations. Between 2000 and 2007,
python sightings were mentioned in conversation
to friends, colleagues and additional landholders
encountered in the region during field work, as a
way of generally collating historical information on
snakes.

(5) Personal field observations. Intensive searches of
50 granite outcrops within the SWS were conducted
to investigate their role in conserving reptile diversity
in modified landscapes. Active searches for pythons
were conducted between 0800 and 1100 hours on clear
sunny days during spring/summer months of 2006
and 2007. Habitats such as hollow logs, trees and rock
crevices were inspected using a hand held torch and
scats and slough skins were recorded. Approximately
1,950 ha of suitable habitat, encompassing 660 ha
of outcropping was surveyed. Spotlighting was
conducted on sites that contained significant amounts
of remnant vegetation (n = 22).

RESULTS
A total of 53 location records, representing a

minimum of 95 observations, were obtained from
the bioregion (Fig. 1), including three records from

Proc. Linn. Soc. N.S.W., 129, 2008

the Victorian border (all from near the Murray River
near Bellbridge, Mount Granya and Pine Mountain).
In some areas of the SWS, M. s. metcalfei has not
been sighted since the 1960’s or 1970’s, however
pythons from 28 locations have been sighted since
1990 (Table 1).

Approximately half of all location records were
obtained from informal conversations. Twenty-three
locality records (43%) and 35 observations (37%)
were collated in this way, although landholder
questionnaires proved successful with 17 location
records (32%) and 35 observations (37%). Nine
location records (17%) and 19 observations (20%)
were obtained from the scientific literature, three
location records (6%) and five observations (5%)
were a result of personal field observations.

Literature review

Most published accounts of M s. metcalfei
came from the Wagga Wagga district. Annable
(1995) recorded 11 pythons between 1976 and 1989,
although specific locations were not documented.
The largest specimen recorded measured almost 4
meters in total length (Annable 1995). Sass (2003)
recorded three pythons between 2001 and 2002
from three locations near Wagga Wagga and a 2.5
m individual was observed basking on the lower
slopes of the Rock Nature Reserve during January
1999 (Murphy and Murphy 2006). Records of two
specimens lodged with the Australian Museum were
obtained from Shine (1994): one from Wagga Wagga
in 1983 and an undated record from near Tumut. A
number of additional surveys for herpetofauna have
been conducted within the SWS, but all lack records
of pythons (Caughley and Gall 1985, Lemckert 1998,
Lindenmayer et al. 2001, Daly 2004). The New
South Wales National Parks and Wildlife Service and
Bionet wildlife atlas databases produced two results;
one from Wagga Wagga (reported in Shine 1994) and
one from the Murray River near Talmalmo (DECC
2007). The Victorian wildlife atlas database contained
a record from Pine Mountain near the state border in
the Upper Murray region (DSE 2003).

Landholder questionnaires

Ofthe sixty-four landholders involved in the long-
term monitoring studies in the SWS, 14 (21.8%) were
aware of pythons being on their properties, although
only two landholders had seen them in recent years
(post 2000). A further 20 landholders involved in the
granite outcrop study revealed two more location
records. In addition, the article published in the
Murray Catchment Management Authority (CMA)
newsletter revealed an additional record near Walla
Walla.

USS

RECORDS OF THE INLAND CARPET PYTHON

(sseg “S) AoAmns sinouoy SulINp punoy uLys Yysnoys S00 ALI LPI S.0€ .SE BIUIPeOOD
(QQan\ ‘d) pous Avy 0} poywoojsuey UoUIDeds [ed0] | $.0961 4.80 .LVI S.9€ .SE CIIWMLUOOD), USAIOJ/\[
(Qq2M\ ‘[) S¥Os JU@d01 ‘S,OL6] Ul MoLING SuLIOWUe T = =§._ LO0T-— ss LO Lp SLE .SE .PIULSUOO, USAIOY\
(ASC) prosarsepe yt LOO~- d.6r .LYI S00 .9€ PHO}IA, UleJUNOI] OUT

MoIING Wqqei ve SuLg\US payysIs— S.0961 A.OS LHI S.9S SE orunO

(SMdN MSN) Auenb Jeou prose sepye | 8661 ACE LHI S.6S SE owpeuuye],

(souor "N) qnd pjo puryoq sen ur paysis— S.OL61 A.67 LPI S.9S SE <qnd pjo, owyeuyey,
(uosMEY “‘D) SMdN ©} popuey uourtoads [[P] peor = S,0661 A.9T LPT S.€0.9€ .Jopsog ueLOJoIA, eAURID II
ITP] peor Ayjequaptooe 100 4.61 .LYI S70 .9€ yeu

(yoord “Y) JOD Suogel|ig wou payoayfoo[ SOL61 A.cS OPI S.ch SE JNOYOO] S,URBIOY\, E[[EA
(Wosio}ed “Y) 8,096] Souls osuRI UO SsUnYsISE COOT- A.€O.LPIS.IS SE AW dovlgey, osuvy ejquiex
(Jeeyorl q) 291) WNH poy UI poyysis | LOOT- ASS OPI S.1S .SE SuoMefsog, A1as0I05)

(siopueg ‘]) poys uress pue [Jy uossunysis+9 L00T- A.bS IPI S.0S .SE Sssuudg vorg, Ateso10H
(Jouo0dg ‘q) MosINg yqqes SuLIajUS pozYysis | ZOOT §8=— Ac v7 =. S St SE jsoroy a8 AYonjuEyy
QiuoAq ‘) Wep wey wou ssuNYsIsZ = 8.0661 ACE OVI S.7S .SE SATOSOY OOS, eplleq

(S]IYOS “D) VOJIA Ul AYN OF popuey [IP peor L661 A.6 OPI S.Ep.SE Peoy sAouojey], apunqie
(SMdN) AeAsns AjIsJoATporq SuLInp punog ysnojs TOOT «=: A cSE HT SEV SE PURSIeQUIOOD, alIpunqeM
(xayxoed “D) [IY JO opis svo ‘peojsouoy wou uses Z = SYOT- Ss ASE OHI S .€b SE BuRsIequIOOD, slIpunqieM
(JILIOA “W)) [114 JO opts som “peojsowoy wou useesp Q0O0T- A.SE.9bIS.€h SE PUBSIeqUIOOD, olpung|eM
(4ayoed “[) [114 JO Opis yOu “peajsouloy wou ueesE —_ENOT- A.SE OPI S.€b .SE PURSIeQUIOOD, o1IpUNg|eM
(uojue,J “Y) Spunosd JooYyos UI 9ey SuIqUITD | S.0661 A.hO.LPIS.p0.9€ Jopsog ueLIO\OIA, espriqijog
(yoyun[d “_) poys Aeyurpsyysis— S.0961 4.90 LPI S.0 .9€ euoosiny |,

(oyunyd A) [IY oyueIs uo payysisZ = S$. 0961 ~=—-« 9S DVI S.LS .SE suey yorlg, Ainqiy

(SoTL WW) Jeary Avsnyy Jeou vjosied Surquiyjo~ S.OL6l A.€S OVI S.v0 .9€ ped uewped, Ainqyy
UOHCUIOJUL SUNYSIS 2 cOURpUNnqYy eq spmisuo7T / opnyey UO}eI07T

Proc. Linn. Soc. N.S.W., 129, 2008

*SUONVITUNUIWIOD [eUOSAId puv SorlvUUOTSonb
dqnd pur A9pjoypury ‘s}10de1 poysiqnd ‘suoneasssgo play UO poseq SoTvAA YINOS MON JO UOIda.1 [RoTYdLAsOsOIg
sodoys U.10}SeM-YINOS 9} Ul UONVULAOJUL SUNYSIS puv UOQLIO] Jafjvrjai vjojids vaso UOYAA, JodAKD puLU *{ eqeL,

256

D.R. MICHAEL AND D.B. LINDENMAYER

(queid ‘W) [[Ty aus yuooelpe JOAL IvU TITY peol
(ouvID ‘J) diysuMo} JO YLIOU JOALI Ie9U [[D] peo.
SUL[SOYJO]O ISIOY S][IY, SUIQUIT|S poyysIs |

(ouvID ‘\) Winssod [reiYysnig Suneisindo1 poyysis |
(1Ox[VYD 4) PROI SUISSOID pozYsIs |

(preMOY “A\) ULL] UO payedojsuRy sudUIdads [ed0] Z
(oueID ‘W) Yoopped suizeis ul pozysis |

(Ves “f) peodJsouoY vou 2 APIAVD JOOI UI pazYsIS ¢
(ueurg “f) Ay[ns Ayoor ur Z “T[IY UO poyysis |
({oeYSI| “C) SuUIpying wey Ul poyysis oyeur WU "|
(oueID ) Auenb 93e]s eo pag yoo10 UI pazyss |
(jevyoI] ‘q) dor9yno oy1uVIs UO puNoO}; sjeds yUd901
(JOBS ‘C) S}VOs JUIEI 2p | [TY s1ULIS UO PoqYysIs Z
(p66 SUIYS) Winesnyy URITeNsnYy UsWIOsds JOYyYONOA
(sseg “S) AdAIns SULINp puNoy UNS Ysnoys

(sseg Ԥ) AdAIns SULINp puNno; UTys Ysnoys

Quen “s{) Y8e19 ePUIOOT IeaU 9907] UI SUYSIS |
Quelng “y) Wing [oUOS SuLIMp poh] Aljewoeprs9e
(Aal0D “g) AeMYsSIPY SUN UO [LY peor ojeuoy Wiz
(uddIH ‘q) ssunYsIs Jepnso1 +¢

(sses ‘S) SSunYsIS Ie[Ndo1 +7

(766] OUIYS) Wnesnyy URTTeENSNY UoUTIOSds JoyYoNOA,
(S661 914eULY) JOLNSIp oy} UI Uses []

(sseg “S) Ssunysis iepnsor

(sseg ‘S) ssunysis 1epnso1

(9007 Aydinyy 29 Aydin) So] Mozjoy Uo poyysis |
(enboeyesseg ‘N) Us]o}s Udy} Poys Ul pozysis |
(ysng ‘q) oul AT[NS AyOoI ul JoYyIES0} pozysIs Z
(uelIg.O ‘d) peajsowloy 2 poys WuIeF UI poyYsis +¢

~ 9007
Z00Z
S.0861
9002
0002
S.066
0002
0007 -
S,0661
9007
9007
L007
S007
0861
Z0/100Z
L007
6661
S007
9002
LOOT -
0002
£861
68-961
LOOT -
S00C -
6661
S007
LOOT -
LOOT -

A .60 .8h1 S00 .SE
A.L0 .8hl $.10 SE
A.91 8hl S.6S re
d.67 .8h1 S$ .6S .r€
d Zp Lvl S £0 SE
Alp Lyl S.10 SE
4.95 Lp S10 SE
A.LS LPL S.6 ve
A.9S .Lyl S.6S ve
ASS AIRS MIOIRGS
A.80 .8h1 $.90 SE
Gi Re AAS EI SE
GE Li SAC ASE
A.€1 8hl S81 Se
Ale Ly SOC Se
A .9€ Lyl S97 SE
A.£0 81 S.8S SE
ALE LPI S.0€ .SE
A 87 Lvl S$ .9€ SE
ance Eyes (ole Se
A 97 Lbl SLI SE
at Pee glial Sell SSE
di.90 LVS .€1 .S€
A 0 Lyl Sl SE
A.ZE .Lyl S90 SE
A .b0 Lpl S91 SE
A 7 Ly S.ZE SE
A £7 Lol S.€€ SE
ai (Se lel SE SE

eIpueqee y

Tesepuny,

.2[BP3901S, Vo1suevuRN]
AoeA euuod, soisueueN
ITH SH]UOoW,, Atospeqeiue\
ped joujny, Alospeqriue,
snsueN

snsueN

dues uosspig, snsueny
snsueN

-Aisenb a1e]S, sseyeyore

_ UISUODSIA\, YUISApe’T
.ASON SMOID, YUSApeT
yun [,

Alpueyjeg, equievaAyy
QAIBSAI JOOS, VquIRaAY]
julog winssog, BWwOo],
JSoIOf 9181S dIplNseLnyy
yoorD Suogelyig 1117
sssullds 31g, esse
AOUL] A YA, PSSeM
TEXoFoLINN

JOLYSIp, esse

sulppnhd wn] d, esse
BIOL WAL, [HH 189104
DAIOSOY SINjJeN YOO oY,
Alosinn sppoydAer, dojoyjng
ITH ISON, doyoy[ng

IIH 1S0N, dojoyng

257

Proc. Linn. Soc. N.S.W., 129, 2008

RECORDS OF THE INLAND CARPET PYTHON

Public presentations

Many landholders involved in the ‘questionnaire’
also attended presentations; therefore their records
were already incorporated in the list of sightings. The
information sessions revealed an additional python
location near Wymah. Unfortunately, the landholder
identified it as being a python after accidentally driving
over it. He had lived on the property all his life and
had never encountered a python before. Interestingly,
another participant claimed to have seen a large
python descending a tree near Mudgegonga during
the late 1980’s, which, 1f authentic would prove to
be a significant record and range extension for this
species in Victoria.

Personal field observations

On the western side of Gerogery Range
(Stringybark Hill), a 2 m python was observed in a
grain shed during 2001, whilst in May 2006, on the
opposite side of the range, a 1.8 m female was found
coiled in the canopy branches of a Blakely’s Red
Gum Eucalyptus blakelyi. Landholders from both
sides of the range were aware of pythons inhabiting
the hill, although only the farmer on the western side
of the range had regularly sighted pythons. Two were
accidentally killed during routine farming activities
during the summer of 2005.

During the summer of 2006 we were informed
of a python residing within a farm shed near Nangus.
This specimen measured 1.4 m in length and was
identified as a sub-adult male. In addition, during the
granite outcrop study python scats were identified near
Morvan (an area known to harbour the species based
on the landholder questionnaire) and on a property in
the Kyeamba Valley with no previous known records
of the species.

Informal conversations

Feedback from colleagues and other sources
produced 22 extra location records for the region as
well as securing additional sightings from some of the
more familiar python locations such as Wagga Wagga,
Yambla Range, Gerogery Range and Goombargana
Hill near Walbundrie. Six of these records were of
road killed specimens, of which two were handed in
to local wildlife authorities for formal identification.

Location of pythons based on landform type
Landform type had a significant effect on the
location and abundance of M. s. metcalfei records
within the SWS (Fig. 2). Forty two percent of
location records and 50% of observations were from
well vegetated ‘inselbergs’ and other granite land
formations. Lowland remnant vegetation accounted

45

CO Location

40

4 Observation

35

30

725)

20

15

10

5
0O-

Water course Vegetated
granite

inselberg

formations

Cleared granite

Low land Undefined
remnant

vegetation

Vegetated
sedimentary
ranges

Figure 2. Distribution of Inland Carpet Python Morelia spilota metcalfei location records (n = 53) and
observations (n = 95) from the south-western slopes of New South Wales, classified by topographic

position or vegetation condition.

258

Proc. Linn. Soc. N.S.W., 129, 2008

D.R. MICHAEL AND D.B. LINDENMAYER

for 18% of records and 13.4% of observations, while
granite hills, devoid of native overstorey vegetation,
accounted for 16% of records and 19.5% of
observations. Watercourses and sedimentary ranges
accounted for 18% and 8% of records, 9.7% and 7.3%
of observations, respectively (Fig. 2).

DISCUSSION

Understanding what factors affect a species’
distribution and abundance provides an important
foundation in mitigating human impacts on
biodiversity (Lindenmayer and Burgman 2005).
However, with limited information on past population
densities for many reptile species, reference to
present day patterns of diversity must be made with
caution (Sadlier and Pressey 1994). Nevertheless,
M. s. metcalfei was once considered widespread
in woodlands along major watercourses and rock
outcrops in Victoria (LCC 1987, DSE 2003) and
presumably was similarly abundant in southern New
South Wales. Anecdotal reports however, suggest
population densities of M s. metcalfei have been
significantly reduced in many parts of south-eastern
Australia (Robertson et al. 1989, Sadlier and Pressey
1994, Sadlier 1994, Shine 1994).

Habitat loss has undoubtedly had a significant
effect on the distribution and abundance of M. s.
metcalfei. However, a number of other causal factors
have contributed to population declines, including
illegal collection for the pet trade (Hoser 1993, Shine
1994), changes in prey availability and composition
(Heard et al. 2004, Shine 1994), predation by feral
animals, particularly the European Fox Vulpes vulpes
(Heard et al. 2006) and deliberate or accidental killing
by humans (Shine 1994). All of these factors are
likely to have influenced the present day distribution
and abundance M. s. metcalfei in the SWS.

Box-Gum Woodland (Eucalyptus albens, E.
melliodora and E. blakelyi) once occurred extensively
throughout the fertile lowland parts of the SWS
but has been reduced to 4% of its original extent
(NSW NPWS 2002). In contrast, the less fertile hills
and granitic woodlands are much less cleared and
appear to have played an extremely important role
in buffering pythons from the effects of broad scale
vegetation loss. Over 40% of all records collated in
this study came from vegetated granite inselbergs
(Fig. 2). The remaining records originate from lowland
remnant vegetation, vegetated watercourses such as
the Murray and Murrumbidgee Rivers, Billabong
and Little Billabong Creek or ranges such as the
Rock Nature Reserve and Yambla Range (including

Proc. Linn. Soc. N.S.W., 129, 2008

Tabletop Mountain). Although few observations stem
from these sedimentary ranges, given the large area
of remnant vegetation and rugged terrain, their role in
conserving viable populations of M. s. metcalfei may
be important.

A study in the Coleambally region of the
Murray and Murrumbidgee irrigation area found M.
s. metcalfei to be significantly associated with large
patches of remnant vegetation, independent of the
presence of creeks or river systems (Doody et al.
2004). In this study, 16 observations came from areas
devoid of native overstorey vegetation (Fig. 2). In
summer, pythons are often attracted to rural buildings
in search of commensal prey items (Shine 1994, Fearn
et al. 2001) but also native birds which nest in garden
vegetation (e.g. one landholder watched a python
raiding a Superb Fairy Wren Malurus cyaneus nest
from a garden shrub). Pythons exhibit seasonal shifts
in habitat use depending on thermoregulatory needs
and prey availability (Shine and Fitzgerald 1996,
Slip and Shine 1988, Heard et al. 2004). Therefore,
it is likely that pythons in the SWS are persisting in
modified landscapes by supplementing their diet with
commensal prey species and using the cover of rocks,
native grass and road side vegetation when returning
to elevated outcrops during the cooler months.

The practice of lodging specimens with the
Australian Museum or observations with the National
Parks and Wildlife Service is not strong in the region,
as evidenced by the number of sightings provided
by landholders and subsequent lack of records in
the wildlife atlas databases (Fig. 2). Similarly, the
transportation of pythons from the local environment
into grain and hay sheds appears to have ceased in
the region, probably as an artefact of extirpation
and population declines, but also as a response to
improved rodent control, better shed designs and
tighter laws governing the handling and movement
of reptiles. Similarly, one landholder near Gerogery
reported Queensland pythons being historically
stocked on his property, however recent observations
from pythons on this property and adjacent granite
outcrops resemble M. s metcalfei. How commonly
M. s. mcdowelli was translocated to the SWS is now
difficult to determine as properties are increasingly
changing ownership meaning this information may
no longer exist in the region.

This study highlights significant populations of
M. s. metcalfei may still occur in the SWS. Priority
conservation areas for this species in the region
include inselbergs and other granite land forms, such
as Goombargana Hill, Gerogery Range, Nest Hill, the
granite belt between Kyeamba and Wagga Wagga,
vegetated ranges such as Yambla Range and the Rock

259

RECORDS OF THE INLAND CARPET PYTHON

Nature Reserve and the riverine environment along
the Murray and Murrumbidgee Rivers. Sadly, due to
extensive habitat loss, degradation and fragmentation,
it appears many potentially suitable granite outcrops
in the production parts of the landscape no longer
support populations of M. s. metcalfei in the SWS.

ACKNOWLEDGEMENTS

We gratefully acknowledge the landholders involved
in the study for freely providing information on python
sightings in the region and for giving us permission to
survey granite outcrops on their properties. We would also
like to thank colleagues: Mason Crane, Rebecca Montague-
Drake, Lachie McBurney, Chris MacGregor and volunteers:
Greg Slade, Hugh MacGregor and Nigel Jones for assisting
with python surveys. Steve Sass and Kylie Durrant helped
provide additional records.

REFERENCES

Annable, T.J. (1995) Annotated checklist of reptiles of
Wagga Wagga and district, NSW. Herpetofauna 25
(1), 22-27.

Benson, J. (1999) ‘Setting the scene - the native vegetation
of New South Wales’. (Native Vegetation Advisory
Council, Royal Botanic Gardens: Sydney).

Caughley, J. and Gall, B. (1985) Relevance of
zoogeographical transition to conservation of fauna:
amphibians and reptiles in the south-western slopes of
New South Wales. Australia Zoologist 21, 513-529.

Coventry, A.J. and Robertson, P. (1991) ‘The snakes of
Victoria: a guide to their identification’. (Department
of Conservation and Environment: East Melbourne).

Cunningham, R.B., Lindenmayer, D.B., Crane, M.,
Michael, D. and MacGregor, C. (2007) Reptile and
arboreal marsupial response to replanted vegetation
in agricultural landscapes. Ecological Applications
17, 609-619.

Daly, G. (2004) Surveys of reptiles and amphibians
on the south-western slopes of New South Wales.
Herpetofauna 34, 2-16.

DECC (2007) New South Wales National Parks and
Wildlife Service wildlife atlas database. http://
wildlifeatlas.nationalparks.nsw.gov.au/wildlifeatlas/
watlas.jsp

Doody, S., Osbourne, W., Bourne, D., Rennie, B.
and Simms, R.A. (2004) Vertebrate diversity on
Australian rice farms: an inventory of species,
variation among farms and proximate factors
explaining that variation. (Rural Industries Research
and Development Corporation: ACT).

DSE (2003) Department of Sustainability and
Environment Action statement: flora and fauna

260

guarantee Act 1988, No. 175 - Inland Carpet
Python Morelia spilota metcalfei. (Department of
Sustainability and Environment: Victoria).

Fearn, S., Robinson, B., Sambono, J. and Shine, R. (2001)
Pythons in the pergola: the ecology of ‘nuisance’
carpet pythons (Morelia spilota) from suburban
habitats in south-eastern Queensland. Wildlife
Research 28, 573-579.

Heard, G.W., Black, D. and Robertson, P. (2004) Habitat
use by the inland carpet python (Morelia spilota
metcalfei: Pythonidae): seasonal relationships with
habitat structure and prey distribution in a rural
landscape. Austral Ecology 29, 446-460.

Heard, G.W., Robertson, P., Black, D., Barrow, G.,
Johnson, P., Hurley, V. and Allen, G. (2006) Canid
predation: a potentially significant threat to relic
populations of the Inland Carpet Python Morelia
spilota metcalfei (Pythonidae) in Victoria. The
Victorian Naturalist 123, 68-74.

Hoser, R. (1993) ‘Smuggled: the underground trade in
Australia’s wildlife’. (Apollo Books: NSW).

Kortlang, S. and Green, D. (2001) ‘Keeping Carpet
Pythons’. (Australian reptile keepers publications:
Australia).

LCC (1983) Report on the Murray valley area. (Land
Conservation Council: Melbourne).

Lemkert, F. (1998) A survey for threatened herpetofauna
of the south-west slopes of New South Wales.
Australian Zoologist 30, 492-499.

Lindenmayer, D.B., Cunningham, R.B. Tribolet, C.R.,
Donnelly, C.F. and MacGregor, C. (2001) A
prospective longitudinal study of landscape matrix
effects on fauna in woodland remnants: experimental
design and baseline data. Biological Conservation
101, 157-169.

Lindenmayer, D. and Burgman, M. (2005) ‘Practical
conservation biology’. (CSIRO Publishing: Victoria).

Morris, P. (1993) The occurrence of the Carpet Snake
Morelia spilota variegata in northwestern New
South Wales. In “Herpetology in Australia: a diverse
discipline’ (Eds D. Lunney and D. Ayers) pp. 67-68.
(Surrey Beatty and Sons: Sydney).

Murphy, M. J. and Murphy, S. (2006) Additions to the
herpetofauna of the Rock Nature Reserve near Wagga
Wagga, New South Wales. Herpetofauna 36, 99-101.

NSW NPWS (2002) ‘White Box — Yellow Box- Blakley’s
Red Gum (Box Gum) Woodland: fact sheet for
NSW’. (National Parks and Wildlife Service: NSW).

Robertson, P., Bennet, A.F., Lumsdun, L.F., Silveira, C.E.,
Johnson, P.G., Yen, A.L., Milledge, G.A., Lillywhite,
P.K. and Pribble, H.J. (1989) Fauna of the Mallee
study area north-western Victoria. (Arthur Rylah
Institute for Environmental Research: Victoria).

Robertson, P. and Hurley, V.G. (2001) Report on Habitat
of the Inland Carpet Python (Morelia spilota
metcalfei) in the Mildura Forest Management Area.
(Department of Natural Resources and Environment:
Melbourne).

Sadlier, R.A. (1994) Conservation status of the reptiles
and amphibians in the Western Division of New

Proc. Linn. Soc. N.S.W., 129, 2008

D.R. MICHAEL AND D.B. LINDENMAYER

South Wales — an overview. In ‘Future of the fauna of
Western New South Wales’ (Eds D. Lunney, S. Hand,
P. Reed and D. Baker) pp. 161-167. (Surrey Beatty
and Sons: Sydney).

Sadlier, R.A. and Pressey, R.L. (1994) Reptiles and
amphibians of particular conservation concern in the
Western Division of New South Wales: a preliminary
analysis. Biological Conservation 69, 42-54.

Schwaner, T., Francis, M. and Harvey, C. (1988)
Identification and conservation of carpet pythons
(Morelia spilota imbricata) on St. Francis Island,
South Australia. Herpetofauna 18 (2), 13-20.

Shine, R. (1994) The biology and management of the
Diamond Python (Morelia spilota spilota) and Carpet
Python (M. s. variegata) in NSW. (NSW National
Parks and Wildlife Service: NSW).

Shine, R. and Fitzgerald, M. (1996) Large snakes in
a mosaic rural landscape: the ecology of Carpet
Pythons Morelia spilota (Serpentes: Pythonidae) in
coastal eastern Australia. Biological Conservation 76,
113-22.

Slip, D.J. and Shine, R. (1998) Habitat use, movements
and activity patterns of free-ranging Diamond
Pythons, Morelia spilota spilota (Serpentes: Boidae):
a radiotelemetric study. Australian Wildlife Research
15, 515-31.

Swan, G., Shea, G. and Sadlier, R. (2004) ‘A field guide to
reptiles of New South Wales’. (Reed New Holland:
Sydney).

Wilson, S. (2005) ‘A field guide to reptiles of
Queensland’. (Reed New Holland: Sydney).

Proc. Linn. Soc. N.S.W., 129, 2008

261

a re ; ‘ans ai of
ARNE Px ES * ei a 1226 Rea ab Mh? i
. eRe SSC AAG UA Mele te MORE AAV AY
: ft aishe bye ner: vdien.) Nerhore
leay (6 RE GA apr ean hl panes +
Hoods rip aegis apt rmntetey bee, amathe ‘
Ayah, eigen at aot alot saan eps iy ies)
bh do 8 esl peed ae? |e
oho urine ied bas yoolerd ad E(peOrt Rs
deplore rua ualine aie oh boty paar
lwrgsten’, W224 1 kage oda ae
7 ek Het oath SEAM den,
=" pntente red TN) blows kid bane
Mer 116, eyalood-od! pagupebast: pita
(b iqalaadhe$arcaagphy pel iy aistGih’ anon
20 aie wneeD Beagalntt, sideman hs srr tao
« vinita’a wildlife’ Agee Bion 5 Wi
BRITO AR oe jatidda! F402) ih cooadek at r
+ RREALIG gokart eR ay koaatiag av
N OOS Autrey noc PvE whey east
i scan A Wan vh, ghp bc sipiie
eran eS nec Msticamel ARG
atic let GM oaOR A eibbab ead wifey
jet ie cen fremtokl vail Saad 5° ee eee
> vat valli Fee et 1, WIT Ae: i:
‘been ee seman a, As
ae st pr ‘op kt rad. ees) ete bie 7
Yi ya leone eter tal dander ripe se 2
eis cx batanie, Salam eotnantic ote
Actes tx rim ay Bignong ie vil wir ee

(mike ‘. we. fo, Opes Me «lO a octet

mw el he ven” ila ts
DOT hw eect. 10 Thad & ~van Shree
ey 4h op! wy pettek SPUiriy 1 vhs saree ;
eit he ee ek eke. Ga’ Verran mers FP Ataralle a dies

wr ES ey ot 5 owes) qrge ti? Bi

—“ ght ba heck beers =

9 By nd Weeds, 0 So io ate iy,

. (omng @ Seay » Reset yc. neme

Tg Sevit yok hci hui ok We :
We Suk = Velo Bg AR

mus (eae a as\ 1
Senet Fess oat ain
Pheuey) BF genes ao” V. Z Atermee «
Pe fi 2 Piet 47AS sation ihe
3 pe Aa egtl las x seu ea, Hit 9 (ete of tee Malle OER
j — di eeecere 6 bape OAT n Ry tee :
ote Trmecrmereitnaewh: Viele
Y Sat beg ¥ 1G i) Pepa ae
ips Cis © Siam } Moe lls MB IAR oes = io
cote dee hillcrerg S45 1! Maneervietit ees,” a
ni sp Ppa? fat sees Rye ied mel ih rachel ha
— ae 7 '
rie ( ‘Gamer heey aa oie wait ee
“ral Vi ibe? Paw) 7h i i een Faden par 7 a. ;
: ni i _ bata

“ws ln MB,

Proceedings of CAVEPS 2005
Alcheringa Special Issue 1, 2006
L. Reed, S. Bourne, D. Megirian, G. Prideaux, G. Young and A. Wright (eds)

I have always considered ‘Proceedings’
volumes to be a particularly valuable form of
publication. For starters, they cover a relatively
narrow field and almost all papers will be of interest
to people in that field. They usually include reviews
as well as examples of current trends in research,
which allow anyone looking for an overview to find
such information in one place. Indeed I have often
found “Proceedings’ volumes to be a good resource
for university students.

The volume under consideration, the
Proceedings of CAVEPS 2005 (a Conference on
Vertebrate Evolution, Palaeontology and Systematics
held in Naracoorte in 2005)) is an excellent example of
a ‘Proceedings’ volume with content of both specialist
and general interest. This is particularly the case with
papers dealing with the Pleistocene extinction of the
Australian megafauna, since this topic has even made
it into daily newspapers recently.

Not surprising, given the venue for the
conference, there are three papers dealing with
Naracoorte Caves (stable isotope record, guano-
derived deposits and fossil deposits). A fourth
paper indirectly concerns Naracoorte Caves. Elery
Hamilton-Smith reviews the life of Tenison Woods.
The paper is fascinating, mainly because Tenison
Woods was a very complex and extraordinarily active
man (in several fields, including geology). Most
people would know of him because of his relationship
to Mary McKillop. Or is it Saint Mary McKillop?
Last time I visited the Mary McKillop Museum in
North Sydney, she was still one miracle short of the
full deck.

Ernie Lundelius gives an interesting summary
of the contributions to vertebrate palaeontology made
by studies of cave sites throughout the world.

CAVEPS meetings are usually dominated
by mammal and fish papers, but in this case there
is only one fish paper, dealing with Devonian
placoderms from New South Wales. However a paper
by Gavin Young straddles the fish/tetrapod boundary.
He discusses the status of two trackways and one
jaw that have previously been accepted as the only
Australian evidence of Devonian tetrapods. The
paper also includes a good review of tetrapod origins
and tetrapod interchanges between Gondwana and
Laurussia.

Birds get a look-in, with one paper dealing
with a Cenozoic songbird from Riversleigh and
another dealing with a New Zealand late-Pleistocene
cave avifauna. Peter Murray and Dirk Megirian
describe one dromornithid, along with reptiles and
mammals, from a presumed Oligocene fauna at
Pwerte Marnte Marnte in the Northern Territory.

Except for a paper by Sue Turner on the
UNESCO Geoparks program (well worth reading
by anyone involved with paleontological as well as
strictly geological aspects of tourism), and a paper
by Roslyn Stemmler highlighting the use of fossils
in educating and inspiring school children, the rest of
the volume belongs exclusively to mammals.

Oliver Brown presents a thoughtful analysis
of Tasmanian Devil extinction on the mainland,
proposing a role for ENSO intensification. Most
importantly, he suggests the mainland extinction
occurred between 3,000 and 4,000 years ago and
convincingly rejects the dates of 430 and 620 for
mainland devils suggested by Archer and Baynes
in 1972. Those dates were always very suspect, but
they managed to get into the general literature as
absolute.

Along the same lines, Peter Murray and Dirk
Megirian provide, in a second paper, a great deal of
data and a strong, thorough analysis concerning the
origin of the thylacinids, based mainly on a Miocene
thylacinid they describe herein. Steve Wroe and others
have proposed that dasyurids were derived from a
thylacinid relatively late in geological time. I was
never comfortable with that unlikely scenario, and
I think Murray and Megirian correctly highlight the
difficulty of determining polarities. Their argument
that thylacinids are a sister group to the plesiomorphic
dasyurids is very convincing.

A new species of palorchestid is described
by Katarzyna Piper, and Neville Pledge presents the
first fossil record of sirenians in southern Australia.

A data-filled paper by Kenny Travouillon
and the usual Riversleigh crowd is harder to place.
It deals with mammal faunas from Riversleigh, but
is mainly concerned with a detailed analysis of the
many individual sites at Riversleigh. This is important
material to have on record, bot not recommended
for light bedtime reading. The same is true of a
companion paper by Mike Archer and 19 co-authors

BOOK REVIEW

which presents species-level lists of the fauna from
80 Cenozoic sites at Riversleigh. This is done in a
huge table which tabulates an immense amount of
work, by many people over a long time.

The remaining papers deal with the
fascinating and popular topic of Australian megafaunal
extinctions, but first Richard Tedford, Rod Wells and
Gavin Prideaux set the scene by discussing marsupial
evolution and the turnover in species preceding the
last glacial cycle (120-20 ka).

Rod Wells and nine co-authors report on
the excavations at Black Creek Swamp on Kangaroo
Island. An excellent table that summarises most of
the species involved and their habitat is a good place
to start for anyone not familiar with the megafauna.
But this is also a paper that should be read by all
graduate students because it is an excellent example
of a complete study of an excavation,. Not only the
fauna, but also the stratigraphy, taphonomy and dating
techniques are thoroughly and clearly set out. This
paper may not represent good career management, as
the authors could have spun at least four minor papers
out of it, but it is terrific science.

In relation to the debate about megafaunal
extinction, the opening sentence of a paper by
historian Kirsty Douglas sums it up: “Debate about
Pleistocene extinction was and is inflected by history,
convention, politics and rhetoric”. Despite the sloppy
use of the word ‘inflected’, presumably the author
meant ‘modulated’ not ‘bent’, the idea is good and well
developed in this paper. The science is occasionally
suspect (I don’t think anyone believes human
influences lead to dwarfing of megafaunal species for
example), but there are some very interesting stories
and insights in this paper.

Lyndall Dawson brings a physiological
perspective to the extinction of large marsupial
herbivores in middle and late Pleistocene Australia.
This excellent review is not “inflected” or even
“bent” by anything other than good, ecophysiological
data provided by the work of, amongst others, Terry
Dawson. Two short, but extremely useful, appendices
summarise the position of marsupial herbivores in
time and space (fossil site).

I have saved the best for last. There are
three papers that make significant contributions to
the megafaunal extinction debate. Firstly, Donald
Pate and others present new carbon dates and review
previous carbon dates from Naracoorte Wet Cave
and conclude that “. . results support other published
data sets in relation to a continent-wide extinction of
megafauna at ca. 46,000 years ago, and reject a late
survival of megafauna at the site”.

264

I have always thought that those who
insist that prehistoric Aboriginal people were not
responsible for the megafaunal extinctions put
up a ‘straw man’ when arguing that blitzkrieg by
overhunting was unlikely and a data-free concept.
Tim Flannery of course draws the ire of this group
by pointing out that environmental changes resulting
from the use of fire could have been the mechanism,
rather than “big game hunting”. In this volume, Barry
Brook and Christopher Johnson provide a model
for the extinction of large species, Diprotodon in
particular, as a result of low levels of exploitation of
juveniles. Zoologists have long recognized that for
most mammalian species the pressure by predators
is against the young and juveniles more than adults.
Brook and Johnson conclude that evidence for a
sophisticated hunting toolkit and massive kill sites
are not a necessary adjunct to overkill.

Finally, Richard Gillespie, Barry Brook
and Alex Baynes use the GIGO (garbage in,
garbage out) principle to cull the radiocarbon data
set. Considerable reliable results stand the test to
establish a human/megafauna overlap of about 3,900
years centred around 44,000 BP. They conclude that,
“Our results rule out climate and environmental
changes associated with the Last Glacial Maximum
as contributing factors in Australian late Pleistocene
megafauna extinctions”. Sir Richard Owen, who
clearly saw the essential nature of this debate 140
years ago, would have loved this paper.

M.L. Augee
April 2007

Proc. Linn. Soc. N.S.W., 129, 2008

Platypus - 4th Edition (2007)
Tom Grant
Illustrated by Dominic Fanning
CSIRO Publishing
RRP $39.95

I remember when the first edition of this
book appeared in 1984. I wish I still had my copy,
and if the person to whom I loaned it some years ago
is reading this, it is now time to return it.

The first edition had a hard cover and
looked at first glance like a children’s book. It was
not really in the ‘children’s book’ genre, although it
was a terrific book for young naturalists interested in
native animals. However, like the current edition, it
was aimed primarily at field naturalists and tertiary
students.

The second edition was one of the first
issues in the UNSW Press Natural History Series.
The continuing success of the book is clearly shown
by the fact that the current edition is the fourth, with
a number of reprints along the way. The Natural
History Series also lives on, although now published
by CSIRO Publishing.

One of the memorable features of the earlier
editions was that they were organized not by topic, but
by the seasons of the year. I thought this was a very
good idea and even attempted to follow it in writing
the first edition of Echidnas. However I found it too
difficult to write within the seasonal framework and
reverted to the usual division into topical chapters. The
problem is that many basic features, such as anatomy
and physiology, don’t really fit into a seasonal pattern.
I see that Tom has given up this format in the 4th
edition and has gone for the usual division of chapters
(ecology, energetics, etc.). There is, I suspect, a good
reason for this, since almost all new material to be
included in the new edition comes from studies in
structure and function rather than natural history. And
Tom has done a very good job in incorporating new
material into this edition. I particularly recommend
“Electroreception: the ‘sixth sense’ revisited”.

The best innovation in this edition is one
which I will most certainly follow if there are future
editions of Echidnas: a “Questions and Answers”
section. This is a very good way of incorporating
those questions that almost everyone wants to ask
about platypuses but are very hard bits of information
to fit into the text. For example “What do you call a

baby platypus?”. In answering this Tom dismisses the
obnoxious term ‘puggle’, although I think he is much
too gentle in doing so. I also like the answer to the
“What is the plural of platypus?” question (“platyp1’
is indeed the only logical answer).

There are good illustrations, good pictures
and good science. This is a book well worth buying,
even if you have a previous edition. Congratulations to
CSIRO Publishing for keeping this valuable Natural
History Series alive.

M.L. Augee
Co-author of “Echidna - extraordinary egg-laying
mammal”

a ®
BOOK REVIEWS
~

' ENOB EY: edits a ah. uayti grec
Mas} ai fe? then. prehiabaattes ley inyet gaat

ve Suid) vd bale «
| pisces ravkee he's 2

+

ht egy ae aS
at ft Sie Wed Alii:

HW aki Sy wil ac Pay oie? on
WeamreyS aT Rar phy es teenie Galt PT

ml = TYNE ES Vile aay Deh
whe Soe, pe ca bwarky Ab peat?

See Ouch ee tise adh shi ALT Sgeeseay Leute ire
9 annie iro toa MUTA C Sad OVW nds
ost stellate tsi a Nal aR CMTS

4 a ae remy rents eat) Ath

ot ees Ae tarred vies

AP athe) Vi7 agente

Pees teers, OS ae ae _—
2) ba wana Soaks” to sotwe-uD
Tet doivard 7 rhe aia
"he! : he , aie ae ee
al i] a ' ; tf if) a
uu
j 1 (yr
i] aut)
4 Pay 7 1
' weritl))
A ' ha
me Tal Gig. ' u
4 cE AeIL'*
‘ ay
rr)
afi
Pr i
rT
Ty
ut t
j
Bis
; am Hi
CoCeeae
| | | 4 Vs 7
t ee
; Tt 7h
i EAD Gi
( i Sel @ wie

preter 26 ncisyils lunar add vat 900g end bas a6

pang fae le us# heey evn tT at

> agai ey mint
whigo waguing ait ite
em sulhinty Gd a dita. fray exon

Sow

PZ FLSA anueesl Mocs? hows em be ot ther grog
nevi ow Bhi is ert ipeniat chases eesnitiny
wth tied dee a. ae aay herh the mech ha

= Waa bis past: Me St hes itt id DATE a
caceh os ; a (pes pei wide. a wintel

as By MUNN ari hay? insets aorta ‘i
eee Vets hen Hae) WR \eRkls Wi tae ee
ear patie Bh aaa Y intros haaian lie fe

m) Hof inet oberily oatsaye Aith &
tories cnitte, Petals fal pee” Phar Caden
meweel: Cte A mel gts aw OMe Shanig nano Be

TYAN REA eH” ie We neg Cine eat Rg

aGinbb) fredixy: isha Solna tia
(ve be aonb :
wget Oat) Piette Sy Gab Ke Boe ef ‘ey. .

ai a SLY seinen TAR es |
eels ice awe sree w BP WED by un
reads pineisl sunita eel it Gals FIR UREN

dil Se ee ee
ae anh eb Sef: sprig
beth eer Aigeieette x, rH
jifoen tf A Vina «Sig ‘Seat p
theltay oath ratte ices Si.
Lindh 2D) de neh At ROD Nt oa ly
eo 0 sev) wid) idguodt 1 agsgcani to enoekee
Lihkov Ge" wolial on botamem: osvo baa 6
oh Wah] sevawolt sonia to seilibe
bie AowoEreTh lmonnse sft niriive oti Mie
ott eonqiiria lnoinged ont sore ils lesoatk ait OBS) vO)

cones ee Mole aves! lend -venen Jedd 2irielde
mk Unie Laie gieka i Ode HiT (isn tae yaotoay
di} silt 0 Jeri 2idi qu seyin eet cout id

buck a tompaie ts2t SdT «ote zobesis ween
odo) laneetatnwon lie teectdn oon lilt Ad Ae
i osbas sve) soeiou riots waa oni of tebolsat
oak csived ated med odie corte! besciiiie ro
vin 4yOLAOtaga nt iot bony aay ¥ Sivohy ant
Losunneosy Ybaluained Lancidibe 2ult out te te
.“boaliatwiyy Séteon riixbe* orks ‘nouqeasINis ae
ant vi aovtihe wut et nullevoem Jest ott. oe
wuout smn oxi a wolidh yiuanies teomeili. toi 9
“grows A bee gedhacedy” p oMtatsd, Oe ol

GY Aitw Sites ome hath: enouD sant im
honest 20 atid byvast wise Te tad’ eseorpesely Ww :
w liso pty, Mihail Ye neces mT a

ar

Animals of Arid Australia: out on their own? (2007)
Chris Dickman, Daniel Lunney and Shelley Burgin (eds)
Royal Zoological Society of New South Wales
RRP $30.00

The Royal Zoological Society of NSW
is gaining a strong reputation for publishing useful
volumes arising from their symposia. In recent years
these volumes include a great deal more material
than was actually part of the relative symposium.
This is the case with the present volume, including
an excellent foreword by Gordon Grigg, covering his
years of research in arid and semi-arid zones.

I will start by pointing out a major gap in
coverage. Reptiles are missing, which is odd given
that reptiles are a major component of arid vertebrate
faunas. There is the occasional mention, and a gekko
makes the front cover, but basically this is a reptile-
free zone.

As with most symposium volumes, there is
a mix of papers; some are primarily research reports
and others are essentially reviews. Actually there are
only three that I would classify as research papers,
two of which are based on state museum records: W.
Ponder and C. Slater deal with freshwater molluscs;
and C. Slatyer, W. Ponder, D. Rosauer and L. Davis
deal with land snails. H. Jones deals with mussels,
at the same time presenting some interesting data on
hydrology.

The rest of the papers are reviews, and that
is not a bad thing. As I have often said, reviews are of
great value for students and for researchers working
in different but related fields. That is not to say the so-
called ‘experts’ do not read such volumes carefully.
We do, mainly to be sure our own research has been
covered.

Reviews cover:

the Aboriginal Dreaming track system in relation to
the arid conditions of Australia (D. Witter);

the need to understand the arid landscape in order to
maintain biodiversity and management (J. Kerle,
M. Fleming and J. Foulkes);

the impact of European settlement, especially the
direct effect of livestock grazing, on native species
in arid zones (M. Letnic);

the landscape approach, particularly dealing with
spatial and temporal variability and fragmentation
(C. McAlpine, S. Phinn, T. Pople, N. Menke and
B. Price);

competition between kangaroos and stock, particularly
when conditions in the arid zone are poor and
survival of kangaroos depend on juveniles (T.
Dawson and A. Munn);

assessing impacts of vegetation management on fauna
in south western NSW, a paper which includes a
massive table with good data relating various
tetrapod species with habitat (vegetation) and an
estimate of habitat quality and permeability (M.
Ellis, M. Drielsma, L. Mazzer and E. Baigent);
and

the interesting question of whether man-made
watering points should be removed or preserved
in national parks (D. Croft, R. Montague-Drake
and M. Dowle).

The final section, entitled “Current
perceptions”, contains four papers, a summary of
the plenary session and a summary of the entire
symposium by the editors.

The first paper in this section, by R.T.
Kingsford, examines four case studies that deal
with the relationship between policy functions of
conservation bureaucracies and scientific information.
The case studies give hope that policy shifts towards
conservation can occur and can be influenced by
science and the media. The important role of the
media and the way in which scientific research can
reach a wide audience are examined by P. Willis in
“Taking the arid zone to TV”.

Contributions from authors working in other
fields can provide refreshing new insights or can be
simply perplexing. The paper by L. Robin and M.
Smith certainly provides a different perspective. I
found it very interesting that they opted for using the
term ‘desert’. I have never understood the Australian
hesitancy to use the term except for specific places
(the Simpson Desert, the Little Desert, etc.), and this
paper gives a brief discussion of the relation of ‘desert’
to ‘arid zone’ and ‘rangeland’. Some clashes occur
between the mindset of biologists and archaeologists.
A biologist with a paleontological bent would hardly
consider 100,000 years ago as “deep time”. That
was yesterday. A biologist might also be bemused
by the statement that “Its focus on the humans in the

BOOK REVIEW

landscape creates a space whether other knowledge
systems can converse with it”. Sounds good, but
could mean anything.

Finally, J. Pickard reinforces the point that
Australian citizens, especially politicians, rapidly
forget the lessons of the past. He does this in regard
to predator-proof fences, which is important in regard
to the great expense of maintaining these structures.

My summary: a mixed volume; mixed in subject
and quality, but well worth consultation by anyone
interested in the arid and semi-arid zones (deserts
even) of Australia.

M.L. Augee
December 2007

268

Proc. Linn. Soc. N.S.W., 129, 2008

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 (M.L. 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, ‘and’ is spelled out (not &), and et al. is not in italics.

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, W.K. (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 autoformatting
or track changes.

INSTRUCTIONS FOR AUTHORS

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 fi les scanned at 600 dpi. Jpeg is not
acceptable and figures of any kind set in WORD are unacceptable. 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 WORD 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
from the website or from the Secretary.

11. It is helpful 1f authors suggest a running head of less than 40 characters.

270 Proc. Linn. Soc. N.S.W., 128, 2007

PROCEEDINGS OF THE LINNEAN SOCIETY OF N.S.W.

VOLUME 129

SMITHSONIAN INSTITUTIO

TT

UL

3 9088 01433 8

Issued 19 March 2008
CONTENTS

1

151

167

WAS

183

197

207

253

Timms, B.V.
The ecology of episodic saline lakes of inland eastern Australia, as exemplified by a ten year study of the
Rockwell-Wombah Lakes of the Paroo.
McAlpine, D.K.
New extant species of ironic flies (Diptera: lronomyiidae) with notes on ironomyiid morphology and relationships.
Mahoney, K.S. and Harris, J.M. Early natural history of the greater glider Petauroides volans (Kerr, 1792).
Zhen, Y.Y. and Pickett, J.
Ordovician (Early Darriwilian) conodonts and sponges from west of Parkes, central New South Wales.
Wright, A.J.
Emsian (Early Devonian) tetracorals (Cnidaria) from Grattai Creek, New South Wales.
Chalson, J.M. and Martin, H.A.
A 38,000 year history of the vegetation at Penrith Lakes, New South Wales.
Holmes, W.B.K. and Anderson, H.M.
The middle Triassic megafossil flora of the Basin Creek Formation, Nymboida Coal Measures, New South Wales,
Australia. Part 7. Cycadophyta.
Powter, D.M.and Gladstone, W.
Habitat preferences of Port Jackson sharks, Heterodontus portjacksoni in the coastal waters of eastern Australia.
Thiem, J.D., Ebner, B.C. and Broadhurst, B.T.
Diel activity of the endangered trout cod (Maccullochella macquariensis) in the Murrumbidgee River.
Green, K.
Fragmented distribution of a rock climbing fish, the Mountain Galaxias, Galaxias olidus, in the Snowy Mountains.
Zhou, Z.Y. and Zhen, Y.Y.
Trilobite-constrained Ordovician biogeography of China with reference to faunal connections with Australia.
Percival, |.D., Zhen, Y.Y., Pogson, D.J. and Thomas, O.D.
The Upper Ordovician Kenya Formation in the Boorowa district, southeast New South Wales.
Stewardson, C.L., Prvan, T., Meyer, M.A. and Ritchie, R.J.
Age determination and growth in the male South African Fur Seal Arctocephalus pusillus pusillus (Pinnipedia:
Otaridae) based upon skull material.
Michael, D.R. and Lindenmayer, D.B.
Records of the inland carpet python Morelia spilota metcalfei (Serpentes: Pythonidae) from the south-western
slopes of New South Wales.
Book review: Proceedings of CAVEPS 2005.
Book review: Platypus - 4"" Edition, M.L. Augee December 2007.
Book review: Animals of arid Australia.
Instructions for authors.

Printed by Ligare Pty Ltd,
138 Bonds Rd. Riverwood 2210

a

€
hi

= aaa > $

eae

i hy » Cin ny

Ve) LD,

Basel
igh sh 8,

SMITHSONIAN INSTITUTION LIBRARIES '

WA

9088 014815716 _ J

ees icwE i * 2 our Bae d uo . * es ot
aese tue, i ; ' diate, Bae Mintoharpat ee
tei peach be Fai ’ ‘ i ‘ ‘ , on aed Pane A
, ‘ae as tyngtatarclee ,
eee rear orteed : 1 Pant ‘
pra utanelat sad ne
i :
tale PEA Ab Men eet Poon hoe
Brag 4 Aas ed oe ad oN Pr a ia t
' trast 4 . ea
H Penn) ae toate
avenge ' ‘Gea
yarebelpath Tires 2 oe
Cagrs \ ney mee?
: De Se Par det :
Bees . . qin . icearterrne
Sissy ces Prerefy: t
tM th rf .
Every sy ;
ARAN LEN
Recscanes '
rey!
Pena ee
OPE, ms a 7 ' 7
Peo ? iar MeCN ‘ease,
Peventnty a te : :
' : fe ‘
eee ys ee: "4 . :