The Middle Triassic Megafossil Flora of the Basin Creek Formation, Nymboida Coal Measures, New South Wales. Part 4. Umkomasiaceae. Dicroidium and affiliated Fructifications

2005·18.3 MB·English

by  W B Keith Holmes

The Middle Triassic Megafossil Flora of the Basin Creek
Formation, Nymboida Coal Measures, New South Wales,
Australia. Part 4. Umkomasiaceae. Dicroidium and Affiliated
Fructifications
W.B.Keith Holmes' and H.M.Anderson^
NSW
'46 Kurrajong Street, Dorrigo, 2453, Australia (Hon. Research Fellow, Geology Department,
University ofNew England, Armidale, NSW 2351, Australia);
^Hon. Palaeobotanist, National Botanical Institute, Pretoria 0001, SouthAfrica
Holmes, W.B.K. andAnderson, H.M. (2005). The MiddleTriassic megafossil flora ofthe Basin
Creek Formation, Nymboida Coal Measures, New South Wales,Australia. Part4. Umkomasiaceae.
Dicroidium and affiliated fructifications. Proceedings oftheLinnean Society ofNewSouth Wales 126,
1-37.
The forked leaves of the morpho-genus Dicroidium are the most commonly occurring foliage in the
collections from two quarries in the Middle Triassic Basin Creek Formation at Nymboida, N.S.W. The
extensive Nymboida collections include leaves which, in gross morphology, range from simple with
entire margins to pinnatifid, pinnate, bipinnatifid and bipinnate forms. The wide range ofvariation, which
includesintergradingforms,createsproblemsinestablishingspeciesboundaries.Forcomparisonwithother
Gondwanan material, the Nymboida leaves have been placed in five 'species complexes' distinguished
as 'Dicroidium coriaceum\ '£). odontopteroides\ 'D. lineatum\ 'D. dubium' and 'D. ziiberP. In each
complex there is acontinuum ofvariation ofform andthere are intergradingforms that link each complex.
A
Illustrations of over eighty leaves demonstrate the range of variation present. single leaf only ofD.
elongatum has been collected. An unusual leafis described as ID. nymboidense sp. nov. Fertile material
affiliated with Dicroidium includes three species offemale strobilus, Umkomasia distans, U. sessilis and
U. sp. A. together with dispersed cupulas and ovules. Microsporophylls are placed in Pteruchus sp. cf P.
matatamajorand a single specimen in P. sp. A.
Manuscript received 17 February 2004, acceptedforpublication 18August 2004.
Keywords: Dicroidium. Middle Triassic flora, Nymboida Coal Measures, Pteruchus, Umkomasia.
INTRODUCTION morpho-genus Dicroidium and its affiliated male and
female fructifications, Pteruchus and Umkomasia
This isthe fourth inaseriesofpapersdescribing respectively, from the two Nymboida localities is
the rich and diverse Middle Triassic megafossil floras described and illustrated.
fromtwoquarrieslocatednearthevillageofNymboida
A
in north-eastern New South Wales. A locality map history ofDicroidium
showing the Coal Mine and Reserve Quarries and a Dicroidium appeared first in the Middle Early
summary ofthe geology ofthe Nymboida Sub-basin Triassic (Smithian), becoming more numerous and
was included in Part 1 (Holmes 2000), which also diverse in the Middle and Late Triassic (Ladinian to
Camian) ofthe Gondwana super-continent (Retallack
dealt with the Thallophyta and Sphenophyta. Part 2
1977, Anderson and Anderson 1983). There are no
(Holmes 2001) included descriptions of 14 taxa of
the Filicophyta known from material preserved in a reliable Jurassic records (Anderson and Anderson
fertile state or remains ofsterile material with known 2003, p.243) and the genus is presumed to have
fern relationships. Twenty four morpho-taxa offern- become extinct at the Triassic-Jurassic boundary. The
like leaves ofunknown relationships were described success and diversification of the genus is recorded
in Part 3 (Holmes 2003). in numerous publications since the first description
andillustrationofthree leaffragments fromTasmania
In this paper abundant material of the leaf
TRIASSIC FLORA FROM NYMBOIDA UMKOMASIACEAE
-
by Morris (1845, as Pecopleris odontopleroides). The should also be considered. Since Thomas (1933) first
genus DicruiJium was erected by Gothan (1912), suggested Umkomasia as a possible early angiospenn
who recognised that the forked Gondwanan leaves, this has been much debated and remains unresolved
usually refeired to Thiniijcldiu, were distinct from (see Slavins et ai. 2002 for a recent discussion).
the unforked leaves of the Northern Hemisphere. Thomas (1933) placed Dicroidium
Walkom (1917) retained the name Thinnfeldia and and its affiliated fertile organs in the Family
later erected the genus Johnslonia (Walkom 1925b) Corystospermaceae. This family name has been
for simple forked fronds from Tasmania. widely used but according to ICBN rules a family
In an overview of genera attributed to the must be based on the ovulate genus, in this case
'Thinnfeldia'series, inwhichheacceptedDicroidiiim, Umkomasia and therefore Umkomasiaceae. Meyen
Frenguelli (1943) also erected the new genera (1987) placed the Family Umkomasiaceae and the
Dicroidiopsis, Diplasiophyllum, Xylopteris and Class Umkomasiales in the gymnosperm Order
ZiiberiaforGondwanaforkedleaves.Townrow(1957) Ginkgoopsida. Anderson and Anderson (2003), in
and Archangelsky (1968) synonymised Frenguelli's their recent review of Umkomasia and Pteruchus
new genera with Dicroidium. Retallack (1977), in a from the Molteno Flora, followed this classification.
biostratigraphical review of Dicroidium and allied
genera, retainedJohnstonia andXylopteris. Anderson The question ofDicroidium attachment
and Anderson (1983), in their extensive study of While the leaves of Dicroidium are abundant
Dicroidium in the Molteno Flora (South Africa), and widespread throughout Triassic Gondwana,
provided a detailed historical review, references and the form of a whole plant is still debated. Various
comprehensive lists to 1982 of all the illustrated authors have suggested that Dicroidium ranged from
material from Gondwana that they considered to fall shrubs to tall trees (Retallack 1980; Anderson et al.
within the genus. They also discussed synonyms and 1998; Anderson and Anderson 2003) or was a large
forking fronds distinct fromDicroidium. tree (Taylor 1996). Petriella (1978) reconstructed
Gothan(1912) studiedthecuticleoiDicroidium Dicroidium as a palmiform tree to 10m high. The
andshowedthatitwasdistinctfromThinnfeldia.Jacob reconstructionbyRetallackandDilcher(1988,fig.10)
and Jacob (1950) and Townrow (1957) added fiarther showedatall deciduous foresttree ina seasonallywet
evidence for the separation. Anderson and Anderson lowland.
(1983) carried out a comprehensive cuticular Despitethe largenumbers ofDicroidium leaves
study (light and scarming electron microscopy) preserved in the Nymboida sediments, no leaves
on Dicroidium in the Molteno Formation. Based have as yet been found attached to a stem. The only
on permineralised material from Antarctica, Pigg convincing specimen elsewhere of leaves attached
(1990) and Boucher et al. (1993) have described the to a stem was illustrated by Anderson and Anderson
anatomy of Dicroidium leaves. The permineralised (1983, PI. 88, fig. 1). Another incomplete specimen
fructifications, Pteruchus and Umkomasia, have (AndersonandAnderson 1983,PI.88,fig.2)suggested
been described by Yao et al. (1995) and Slavins et al. that the leaves were attached in a fascicled maimer
(2002) respectively. to a stem. Anderson and Anderson (2003 p. 257, text
While the forked leaves of Dicroidium are fig.5) base their reconstruction on these specimens.
ubiquitous in Gondwana Triassic floras (Retallack In a paper describing Umkomasia uniramia,
1977; Anderson and Anderson 1983) their origin Axsmith et al. (2000, figs 6 and 8) illustrated
is unresolved (Boucher et al. 1993; Axsmith et al. a Dicroidium odontopteroides leaf apparently
2000). Archangelsky (1996) suggested that the contiguous with, or overlain by, a stem of a plant
Carboniferous pteridospemi Botrychiopsis may with long and short shoot morphology. By analogy
provide a plausible ancestral morphology for the with extant plants bearing long and short shoot
Umkomasiaceae. Retallack (1980) noted that the morphology (e.g. Gingko biloba) it would be
multi-forked Lepidopteris callipteroides, which unlikely in the extreme for a plant with this growth
occurs immediately above the Late Pennian coal morphologytobearaleafon the long shoot section of
seams in the Sydney Basin, probably belonged to a stem subsequentto the formation ofwell-developed
the pteridospemn stock that gave rise to Dicroidium, short shoots. On the illustrated Antarctic specimen
most likely through intemiediate forms similar to even the short shoots are in a leafless state. The
'Dicroidium'gopadense from Nidpur in India (Bose close association of the Dicroidium leaf and stems
and Srivastava 1971). Palaeozoic pteridosperms from bearing short shoots suggests an affiliation but this
Siberian and Cathaysian sources that survived the is not exclusive as other leaves (e.g. Heidiphyllum,
end-Pemiian extinction in refugia as yet unidentified Taeniopteris) are present in the same deposit. We do
Proc. Linn. Soc. N.S.W., 126, 2005
W.B.K. HOLMES AND H.M. ANDERSON
notacceptthe claimofAxsmith etal. (2000) thattheir pinnae and typical D. odontopteroides pinnae on
specimen "demonstrates unequivocal evidence" of the same leaf indicate how closely related these
Dicwidium attachment. 'genera' are. It is far more likely that those leaves
Archangelsky (1968) argued a case for the are interspecific rather than intergeneric 'hybrids' or
affiliation of silicified logs of Rhexoxylon and evenonepolymorphic species. Some SouthAmerican
Dicwidium leaves. Meyer-Berthaud et al. (1993) workers continue to use a multiplicity of generic
described permineralised twigs from Antarctica names forDicroidium leaves (Artabe 1990; Artabe et
as Kykloxylon and suggested that the stems bore al. 1998; Gnaedinger and Herbst 1998, 2001).
Dicwidiumfremouwensis Pigg (1990) leaves, but in Early authors appear to have worked only on
neither ofthese cases have the leaves been found in limited museum collections or had very little field
organic connection. experience because they failed to recognise the
variability within their 'species' and the intergrading
Dicroidium and its fertile organs forms that may have linked the 'species'. Holmes
Thomas (1933) described the female organ (1982) discussed the problems created by the
Umkomasia and the male Pteruchus. Based on close variation and intergrading forms of Dicroidium in
associationandsimilarityofcuticlesheregardedthem the Middle Triassic Benolong Flora. He referred to
as the fructifications of Dicwidium. This affiliation the work of Meyen (1979), who had demonstrated
is now generally accepted (Townrow 1962; Holmes fi"om a large population of Permian pteridosperm
1982; Anderson and Anderson 1983, 2003; Crane leaves previously placed in several genera and many
1985; RetallackandDilcher 1988; Yao etal. 1995 and species that they all belonged in the single species
Slavins et al. 2002). Rhaphidopteris praecursoria. Rees and Cleal (1993)
The fertile genera Karibacarpon (Lacey 1976, showed that the leaves of a Jurassic pteridosperm
Holmes and Ash 1979) and Fanewtheca (Anderson previously placed in six species and four genera all
and Anderson 2003) have also been linked to belonged toArchangelskyafurcata.
Dicwidium. The species described by Axsmith et al. The revision by Retallack (1977) of the
(2000) as Umkomasia uniramia from Antarctica is an Dicroidium genus (plus Xylopteris, Johnstonia and
ovulate organ with singlepedicillate cupules arranged Tetratilon, which he retained as allied but distinct)
in a terminal radial whorl on a peduncle attached to a was a significant attempt to provide a taxonomic
leafless short shoot on a mature stem bearing further guide to the genus that would mainly be usefiil for
leafless short shoots. From the significant differences stratigraphic purposes. However, his descriptions and
in the architectural arrangement ofthis fiiictification stylised sketches of the taxa he recognised failed to
we believe it should be placed in a morpho-genus demonstrate the range of variation within a single
separate from Umkomasia. population of each taxon. The necessity under the
IBCN Code (ICBN 2001) for typification has also
Problems with Dicroidium taxonomy exacerbated the problem of dealing with variable
In many early records (Walkom 1917, fossil populations.
Frenguelli 1943) the forked leaves now known as The photographic record by Anderson and
Dicwidium (Gothan 1912) were placed in several Anderson(1983)of1133 individualDicroidiumleaves
genera and particularly in Thinnfeldia. Walkom arranged in 'palaeodemes'was anotable achievement
(1917) consideredhis '"Thinnfeldia'"'' leavestobe ferns and clearly demonstrated the diversity within single
as they were closely associated with fi-ond fragments populations of Dicroidium leaves. They were the
bearing sori. Townrow (1957) separated these fertile first to use palaeo-gamma taxonomy forDicroidium,
fi"agments fromDicwidium as they were indeed ferns i.e. using the data of comprehensive collections
which Holmes (2001) has placed in the new genus from many localities. They defined a 'palaeodeme'
Herbstopteris. Several genera of forked fronds have as a "collection of specimens representing a single
been described as separate fi-om Dicwidium based breedingpopulation showing anormal distribution of
mainly on morphological differences of pinna or variation and derived from a single fossil assemblage
pinnule shape (eg Johnstonia, Xylopteris, Zuberia from a discrete lithological unif The same authors
.
and others). Townrow (1957), Archangelsky (1968), did not attempt to define precise boundaries between
Holmes (1982) and Anderson and Anderson (1983, taxa as they considered "it would be unproductive
2003) have regarded most ofthese additional genera considering the general morphological fluidity within
as synonyms of Dicwidium. Leaves illustrated by the genus" and they had "little doubt thatDicroidium
Anderson andAnderson (1983, PI. 74, figs 1-9) which speciation was anything but simple and came closer
showed typical D. elongatum (=Xylopteris elongata) to the reticulate speciation model of Sylvester-
Proc. Linn. Soc. N.S.W., 126, 2005
TRIASSIC FLORA FROM NYMBOIDA UMKOMASIACEAE
-
Bradley (1977)". Based on the palaeodeme approach, palaeo-beta taxonomy and based ourclassification on
Anderson and Anderson (1983, 1989) listed a total our comprehensive collection of over 2600 selected
ol"32 taxa comprising 10 species, 17 subspecies and and catalogued slabs that have been accumulated
15 formae. In their recent revision (Anderson and during many trips to the Nymboida quarries over a
Anderson 2003) Dicroidium now comprises 21 taxa period ofalmost 40 years. The bulk ofthe collection
ofspecific rank plus 15 fonnae. was made from fallen blocks that had been blasted or
The reasons listed by Boucher et al. (1993) bulldozed from the quarry faces. The exact source of
provided a good summary ofwhy Dicroidium is not the in situ material is mostly unknown. Except in rare
a well-understood genus: there is a large amount of cases such as the slabs containing large numbers of
morphological variation; there is inconsistency in similar leaves (e.g. Figs 2 and 6A), we are unable to
naming; variations in cuticular features are not well identify specific populations or palaeodemes as was
enough known; complete plants are not known; achieved by Anderson and Anderson (1983, 1989,
sample sizes are small especially from remote areas; 2003). Future work involving in situ collecting from
and often only fragments are preserved. specific assemblages (i.e. from discrete lithological
units) would allow palaeo-gamma taxonomy and the
material to be documented into palaeodemes.
MATERIALS Dicroidium foliage is the most commonly
preservedfossilatNymboidaandalsointhecatalogued
In the Holmes' catalogued collection of over collection. The forked leaves occur in a diversity of
2600 slabs ofNymboida material, approximately one forms ranging and intergrading from simple leaves
third of the slabs bear leaves that can be attributed with entire margins to pinnate to bipirmate. Faced
to Dicroidium. The female and male fertile organs with this continuum one could place all the leaves
Umkomasia andPteruchus affiliatedwithDicroidium in one polymorphic species or, on the other hand,
are rare and have been found at a ratio of one describe numerous species each based on a 'single'
identifiable specimen to c. 70 leaves. This is closely type specimen. Due to this range of variation and
similar to the ratio observed at the Benolong locality intergradation we are unable to determine reliable
(Holmes 1982). IntheMoltenoassemblagesAnderson diagnostic features. This makes it difficult to identify
and Anderson (2003, p. 240, 250) also noted that the the Nymboida Dicroidium material to a specific
fertile organs were rare and gave the return for each level.
locality only in terms of man-hours spent chipping Although the sediments in the Nymboida
open fossiliferous slabs. quarries were deposited over a relatively short period
TheNymboidafossilplantmaterial ispreserved of geological time (Holmes 2000) they do represent
mostly as carbonaceous compressions, but a tectonic many different sedimentary facies, e.g. coal seams,
heating event during the Cretaceous (Russell 1994) shales, fossil soils, siltstones and sandstones. The
has destroyed the cuticle of otherwise beautifully- enclosed fossils certainly would have been derived
preserved specimens. fi^omarangeofhabitatssoitismostlikelythatvarious
In addition to the material held in the Holmes species ofDicroidium plants had evolved or adapted
collection we have examined other Nymboida to occupy particular environmental niches. Only one
specimens housed in the University ofNew England variable species ofPteruchus plus a possible second
Geology Department and theAustralian Museum. species are recognised in the Nymboida collections
while at least three separate species of the female
Dicroidium fructification Umkomasia are present. It
OUR METHOD OF NOMENCLATURE is therefore most likely that several discrete species
ofplants bearing Dicroidium leaves were growing in
Earlier attempts to classify Nymboida the region in the Middle Triassic.
Dicroidium leaves were made from limited and For the Nymboida Dicroidium leaves to be
usually fragmentary specimens, i.e. using mainly compared with collections ixom other geographical
palaeo-alpha taxonomy (De Jersey 1958, Flint and and stratigraphical localities we have separated
Gould 1975, Retallack et al. 1977). Retallack (1977 the leaves into informal complexes based on leaf
Microfiche Frames G11-G13) listed 12 Dicroidium morphology, which range from simple to pinnate,
taxa plus one each of Johnstonia and Tetraptilon bipirmatifid and bipirmate. Except for the single
(both here included in Dicroidium) and two ovulate specimen placed in D. elongatum (Figs. 19, 20A) and
species from the Coal Mine Quarry. the few rare and problematic specimens (Figs 20B-E)
In this present revision we have used mainly wehavegroupedthefoliagematerialintofive 'species
Proc. Linn. Soc. N.S.W., 126, 2005
W.B.K. HOLMES AND H.M. ANDERSON
complexes' each named for a previously-described PI. 91, fig. 1
D
species that represents the core of the complex. 1982 Dicroidium coriaceum. Holmes, figs 3C,
Specimens in the Selected References lists also have 1983 Dicroidium coriaceum subsp. coriaceum,
a gross morphology identifiable with the core ofthe Anderson andAnderson, PI. 36, figs 3-6; PI.76,
complex. Each complex includes a particular section figs 1-6
oftherange ofdiversity ofthe NymboidaDicwidium 1983 Dicroidium coriaceum subsp. dutoitii,
leaf collections and is linked by intergrading forms Anderson andAnderson, PI. 41, figs 1-28; P1.76,
with adjacent complexes. We illustrate over 80 figs 12-17
reasonably well-preserved leaves (Figs 1-18), that
demonstrate the variation within each complex and Description
the obvious forms intergrading between successive Forked leaves ofvariable length and width,
complexes. While we acknowledge that our form of usually entire, sometimes lobed and grading to
classification is subjective, it is based on a large and pinnatifid.
representativecollectionofreasonablywell-preserved
material. Discussion
Due to constraints imposed by the lack of D. coriaceumwasfirstdescribedfromTasmania
storage facilities at the Australian Museum only by Johnston (1887) as RhacophyUum coriaceum.
the specimens illustrated in this paper have been Walkom (1925b) re-examined Johnston's material
allocatedAMP numbers. All other specimens remain and transferred it to the new genus Johnstonia. He
in the Holmes Collection. noted that the margins were entire or slightly lobed.
mm
Specimens attained a length up to 100 above the
dichotomy and the breadth of the larger specimens
SYSTEMATIC PALAEOBOTANY was 10 mm, though in general they were narrower.
Antarctic material with leaves having a broad lamina
Order Ginkgoopsida Meyen 1987 and the fork closer to the apex were described by
Class Umkomasiales Meyen 1984 Townrow (1967) as D. dutoitii. He selected as the
Family Umkomasiaceae Meyen 1984 holotype a leaf illustrated by DuToit (1927, text fig.
12D) from the Molteno Formation of South Africa,
Genus Dicroidium Gothan 1912 which differed quite significantly from the Antarctic
specimens. Retallack (1977) placed the Antarctic
Type species material in Johnstonia coriacea var obesa with the
Dicroidium odontopteroides (Morris 1845) South African material retaining the epithet dutoitii.
Gothan 1912 Anderson and Anderson (1983) have separated D.
dutoitii from D. coriaceum on palaeodeme evidence
Synonymy [Synonymised by]: at localities in the Molteno Formation although their
Johnstonia 'WaXkom 1925b [Townrow 1957] illustrations show some overlap in leaf dimensions.
Dicroidiopsis Yxeng^itlW 1943 [Archangelsky 1968] From SouthAmerica, Jain and Delevoryas (1967, PI.
Diplasiophylliim Frenguelli 1943 [Archangelsky 91,fig. 1)illustratedasingleslabthatencompassesthe
1968] size range ofboth above species. Holmes (1982) also
Zuberia Frenguelli 1943 [Townrow 1957] noted a wide variation in size range ofD. coriaceum
^(y/op/em Frenguelli 1943 [Archangelsky 1968] in the Benolong flora ofNSW. The Nymboida leaves
Tetraptilon Frenguelli 1950 [Anderson andAnderson illustrated in Figure 1 are a larger form, ranging
mm mm
1983] from 120-140 long and 5-10 wide and from
//oeg/a Townrow 1957 [Archangelsky 1968] entire to lobed, and couldbe regarded as a form ofZ).
dutoitii Townrow (1967). As only three slabs ofthis
larger form are present in the Nymboida collections
'Dicroidium coriaceum complex' andtheirfull rangeofvariation isnotknownweplace
Figures lA-C; 2A-E them in theD. coriaceum complex.
The specimens illustrated in Figures 2A-E and
Selected References Figure 4C were all collected from a single slab of
1925bJohnstonia coriacea, Walkom, figs 6,7 whitish siltstone. We believe that, with the exception
1927 Johnstonia coriacea, DuToit, fig. 12D, 13B only ofthe pinnate fragment on the lower left ofFig.
1932 Johnstonia coriacea, Du Toit, text fig. 2A 2A, all the other leaves represent a single population.
mm
1967 Dicroidium coriacium, Jain and Delevoryas, They range in length from c. 40-80 long and 3-
Proc. Linn. Soc. N.S.W., 126, 2005
TRIASSIC FLORA FROM NYMBOIDA UMKOMASIACEAE
-
mm
7 wide. The leaf margins arc mostly entire but illustrated and Morris noted that the specimen (PI.
lobecl and pinnalifid forms (Fig. 2E and Fig. 4C) are 6, fig. 4) with more elongate pinnae was probably
present. The latter forms intergrade with, and are very a variety of the species. When compared with
close to, the pinnatifid and pinnule coalescing forms the range of variation encompassed in our D.
ofthe 'D. odontopteroides complex'. odontopteroides complex, Morris's three specimens
show a very limited range of variation as would be
expected for this morpho-species. However, due to
'Dicroidiuni odontopteroides complex' the lack of understanding of the range of variation
Figures 3A-E; 4A-G; 5A-D; 6A; 7A-C; 8A; 9A-C; that commonly occurs in a single population, leaves
10A,D,E similar to Morris's elongated pinna form were raised
to specific rank i.e. Dicroidium lancijolium (Gothan
Selected References 1912; Walkom 1917; Frenguelli 1943; and others, see
1 845 Pecopteris odontopteroides, Morris, PI. 6, fig. Hypodigm Lists of Anderson and Anderson 1983).
3 The presence of a midrib (alethopteroid venation)
1890 Thinnfeldia odontopteroides, Feistmantel, PI. in the more elongated pinnae has been regarded
26, figs 2, 2a by some authors as a specific or varietal diagnostic
1917 Thinnfeldia hncifolia, Walkom, PI. 7, fig. 2; feature. The form of venation is usually dependent
Pi. 3, fig. 1 on the length and shape of the pinnae, which varies
1975 Dicroidium odontopteroides, Flint and Gould, according to the position on the frond and whether on
PI. 3, figs 10, 11 the inside or outside ofthe fork. Both odontopteroid
1982 Dicroidium odontopteroides var. moltenense, and alethopteroid forms ofvenation may be observed
Holmes, figs 4A, B on a single frond. Leaves placed in D. lancifolium
1983 Dicroidium odontopteroides subsp. by several authors are best regarded as long-pinnaed
orbiciiloides, Umk 111 palaeodeme,Anderson forms ofD. odontopteroides.
andAnderson, PI. 42 The range of variation of fronds within this
1983 Dicroidium odontopteroides forma Nymboida complex encompasses forms that accord
odontopteroides, Umklll palaeodeme, Anderson with several published species, subspecies, varieties
andAnderson, PI. 43 or forms. On the classification of Retallack (1977,
1992 Dicroidium odontopteroides, Taylor et al., fig. Microfiche Frames II to 17) these varieties include D.
1 odontopteroides var. moltenense, D. odontopteroides
2000 Dicroidium odontopteroides,Axsmith et al., var. obtusifolium and D. odontopteroides var.
fig. 3, leafonly odontopteroides.
2001 Dicroidium odontopteroides, Gnaedinger and Specimens with short broad pinnae (Figs 3A-
Herbst, fig. 4E (in gross morphology but with D; 4C, D, G; 7C) agree with leaves from the D.
denser venation) odontopteroides subsp. orhiculoides palaeodeme
of Anderson and Anderson (1983, PI. 42), which
Description included D. crassinervis forma obtusifolium,
mm
Usually once-forked pinnate frond to 160 D. odontopteroides subsp. orhiculoides and D.
mm mm
long; pinnae to 40 long and 15 wide, not crassinervis forma crassinen'is. Boucheret al. (1993)
basally contracted, ranging from semi-orbicular to separated compression and impression material from
broadlytriangular, to elongatedrectangularorslightly Mt Falla in the Beardmore Glacier area ofAntarctica
tapering, with rounded or broadly obtuse apex; intoD. odontopteroides,D. lancifoliumandD. dubium
the longer pinnae with a midrib and alethopteroid based on minormorphological and cuticular features,
venation; shorter rounded pinnae with odontopteroid which may have represented a nonnal range of
venation as in Fig. 7C. The specimens illustrated in variation within a single species complex. Frenguelli
Figs 7A and B represent the core of this Nymboida (1950) erected the genus Tetraptilon ior Dicroidium
complex. The leaf in Fig. 8A is the largest specimen odontopteroides-like leaves from Argentina in which
from this complex in the collection. the frond had a double fork. This feature has been
recorded also in D. odontopteroides assemblages
Discussion from Australia (Flint and Gould 1975, PI. 3, fig. 10;
Taxonomic confiision has arisen since the Retallack 1977) and South Africa (Anderson and
original description and illustrations of Dicroidium Anderson 1983, PI. 87, figs 1.2,4 and 6, PI. 88. fig.
odontopteroides (as Pecopteris odontopteroides) 1). Anderson and Anderson (1983) synonymised the
by Morris (1845). Only three frond fragments were genus Tetraptilon with Dicroidium. In ourcollections
Proc. Linn. Soc. N.S.W., 126, 2005
W.B.K. HOLMES AND H.M. ANDERSON
these multiple forked fronds comprise c. 1% of the This complex is closely allied to the D.
D. odontopteroides complex. The forking may result odontopteroides complex but has been separated on
in three, four or five branches, with four branches the basis of the decurrent pinnae which taper to an
being the most common (Figs 5A-D and Fig. 6A). acute apex.
On some bedding planes this is the only form ofleaf Retallack (1977, Microfiche Frames HI7,
preserved. This suggests that for some trees double- HI8) placed leaves ofthis complex as a variety ofD.
forked leaves were the normal frond and not aberrant lancifolium, which we regard as belonging in the 'D.
as was surmised by Anderson and Anderson (1983, odontopteroides complex'. Leaves comparable to our
Pis 71, 87) because of their rare occurrence. The 'D. lineatum complex' were described by Anderson
slab illustrated in Figure 6Ais a good example ofthe andAnderson (1983) from the Molteno Formation of
range of variation in a single population of double- South Africa as D. odontopteroides subsp. lineatum
forked leaves. A single specimen ofa double-forked but were later raised to specific rank (Anderson and
D. dubium (Fig. 6B) has also been collected. Anderson 2003).
In a recent paper, Gnaedinger and Herbst IntheNymboidacollection,formswithundulate
(2001) recognisedthe greatmorphological variability margins or incipient lobes (Figs IIC; 12A,D) are
in Dicroidium leaves from three Upper Triassic linking forms with the 'Z). dubium complex' below.
formationsinnorthernChile.Theyillustratedanumber
ofexamples that indicated an intergrading 'line' from
D. obtusifolium to D. odontopteroides (varieties ''Dicroidium dubium complex'
moltenense and remotum) to D. odontopteroides var. Figures 6B;.l3A-C; 14A-E; 15A-D; 16A
odontopteroides to D. lancifolium var. lancifoliitm.
This range ofmorphology is here included in our 'D. Selected References
odontopteroides complex'. 1890 Gleichenia dubia, Feistmantel, PI. 26, fig. 3
Leaves illustrated in Figures 9A, B are forms 1908 Thinnfeldia odontopteroides, Seward, fig. 4
intergrading with extremes from the D. lineatum 1928 Thinnfeldia talbragarensis, Walkom, PI. 27,
complex. fig. 1
1947 Thinnfeldia talbragarensis, Jones and
deJersey, PI. 1, fig. 5
''Dicroidium lineatum complex' 1965 'Thinnfeldia'talbragarensis. Hill et al., PI. T4,
Figures 10B,C; 11A,B,D; 12A-C fig. 6
1982 Dicroidium dubium var. dubium. Holmes, fig.
Selected References 7D
1883 Gleichenia lineata, Tenison-Woods, PI. 3, fig. 1983 Dicroidium dubium subsp. dubium, Anderson
6; PI. 8, fig. 2 andAnderson, PI. 33, figs 21-31
1898 Thinnfeldia indica var. falcata, Shirley, PI. 7, 1983 Dicroidium dubium subsp. tasmaniense,
fig. 2 Anderson andAnderson, PI. 44, figs 7-16; PI. 53,
1917 Thinnfeldia acuta, Walkom, PI. 3, fig. 4 figs 15-20; PL 59, figs 1-14; PI. 60, figs 1-5
1977 Dicroidium lancifolium var. lineatum, 1983 Dicroidium dubium subsp. switzifolium,
Retallack, Microfiche Frame H17 Anderson andAnderson, PI. 44, figs 17-20
1983 Dicroidium odontopteroides subsp. lineatum,
Anderson andAnderson, PI. 64, figs 12-29; PI. Description
65, figs 1-3; PI. 79, figs 4, 6 Forked bipinnatifid leaf; pirmae elongate,
1985 Dicroidium lancifolium var. lineatum, Artabe, tapering to acute apex; margin variously lobed,
PL 3, fig. 5 proximal lobes sometimes separated to the pinna
1992 Dicroidium lancifolium, Taylor et al., fig. 2 rachis but coalescing distally. Leaves variable in size
mm
2001 Dicroidium lancifolium var. lineatum, from 70 - 200 long; pinnae in mid-portion ofleaf
Gnaedinger and Herbst, fig. 2, A-E, fig. 3 L from 15-90 mm long, 4 - 15mm wide.
Description Discussion
mm
Forked pinnate leafto 200 long; pinnae Fronds of the 'D. dubium complex' are the
mm
elongated-triangularto 35 long, tapering to most commonly occurring Dicroidium fossils at
acute apex, broad base, usually decurrent. the Nymboida quarries. Leaves that represent the
core of this complex (Figs 13C, 14D,E, 15B,C) are
Discussion bipinnatifid to partially bipinnate and agree well with
Proc. Linn. Soc. N.S.W., 126, 2005
TRIASSIC FLORA FROM NYMBOIDA UMKOMASIACEAE
-
the MoltenopalaeodemesofD. cliihiumsubsp. diihiiim varietal types (Figs 16B, 18A). The range ofpinnule
as recognised by Anderson and Anderson (1983, shapes in the Nymboida 'D. ziihericomplex' supports
p. 106). Pinnae on the same frond may vary from the decision of Anderson and Anderson (1983) to
entire to lobed and pinnate. Specimens with narrower recognise the previously described D. zuheri varieties
pinnae and undulate margins are intergrading forms feistmantelii, sahnii, harrealemis and papillatum as
with the 'D. linecitiim complex' while the bipinnate nornial variations within a single species.
forms intergradewith those ofthe 'D. ziihericomplex' We agree with Archangelsky (1968), Retallack
(Fig. 5A). The apical leaf fragment showing large (1977) and Anderson and Anderson (1983, 2003)
1
bipinnatifid pinnae (Fig. 16A) may be compared that the continued placement ofthese leaves into the
with D. duhiiim var. australe (sensu Retallack 1977, separate genus Zuberia (Artabe 1990; Gnaedinger
Microfiche Frame H8). and Herbst 2001 ) is not warranted.
''Dicroidium zuberi complex' Dicroidium elongatum (Carruthers) Archangelsky
Figures 16B,C; 17A-D; 18A 1968
Figures 19, 20A
Selected References
1890 Thinnfeldia odontopteroides, Feistmantel, PI. Holotype
24, figs 1,2; PI. 25, figs 1,2 Designated by Retallack (1977) Sphenopteris
1917 Thinnfeldiafeistmantelii, Walkom, PI. 2, figs elongata Carruthers 1872, PL 27, fig.1
1,2
1944 Zuberiazuberi, Frenguelli, PI. 4 Selected References
1975 Hoegiapapillata, Flint and Gould, PI. 2, figs 1883 Trichomanidesspinifolium, Tenison-Woods,
4,5 PL 3, fig. 7
1977 Dicroidium zuberi, Retallack et al., FigsA-F 1898 Trichomanides elongata var. spinifolia,
1979 Dicroidium zuberi, Petriella, PL 2, fig. 5 Shirley, PL 5, fig. 2
1983 Dicroidium zuberi, Anderson andAnderson, 1917 Stenopteris elongata, Walkom, PL 6, figs 1, 3
PL 61, figs 1-13; PL 62, figs 1-4; PL 81, figs 1-5 1965 Xylopteris spinifolia. Hill et al., PL T5, fig. 7
1985 Dicroidium zuberi var.papillatum, Artabe, PL 1977 Xylopteris spinifolia, Retallack, Microfiche
4, fig. 2 Frames J13, 14
1982 Dicroidium spinifolium. Holmes, Fig. 5B,C
Description 1983 Dicroidium elongatum forma remotifolium,
mm
Frond bipinnate, small to large, 150 - 600 Anderson andAnderson, p.115, PL 48, figs 24-32
long; pinnules variable in shape, rounded to blunt
rhomboid or falcate, inclined towards pinna apex, Description
coalescing distally and apically. Basiscopic pinnules Knownonlyfrom asingle incompletespecimen
often decurrent on main rachis. from the Coal Mine Quarry (AMF125082). Frond
mm
bipinnate, c. 120 long; pinnae linear at c. 45°
Discussion to main rachis; pirmules well-spaced, elongated
Leaves of this complex are widespread but triangular to linear.
not common at the Nymboida localities. Solitary
large leaves are sometimes preserved intact in beds Discussion
of sandstone thus indicating they were tough and This solitary specimen is preserved in siltstone
resistant to damage during transport. D. zuberi is togetherwithotherfragments thatappeartohavebeen
differentiated from D. dubium by the presence of subjected to long distance transport before burial. In
pinnules divided to the pinna rachis, but there are the probable contemporary Benolong Flora (Holmes
numerous intergrading fornis with deeplybipinnatifid 1982) leaves of a 'Dicroidium elongatum complex'
fronds (Fig. 17A,B). comprised 20% of the preserved plant remains.
Retallack (1977, Microfiche Frame G12) listed Holmes reconstructed the Benolong environment as
three varieties ofZ). zuberi from Nymboida based on a dry sclerophyll woodland on low fertility sandy
the pinnule shape of mostly fragmentary material. soils. The presence of a rich fern and fern-like leaf
Our extensive collections clearly demonstrate flora at Nymboida (Holmes 2001, 2003) suggests
intergrading fonns that link the varieties. Indeed a vegetation growing in a moist environment with
pinnules on the same leaf may include different rich soils. This solitary specimen of D. elongatum
Proc. Linn. See. N.S.W., 126, 2005
W.B.K. HOLMESAND H.M. ANDERSON
together with associated plant fragments was most essentially odontopteroid, is somewhat similarto that
Ukely transported from a drier less fertile upstream in the decurrent, ovate, sometimes lobate pinnules
source. of Nymboidiantum spp (Holmes 2003, Figs17-
D. elongatum is a species complex ofpinnate to 19). However the size and texture of the pinnae of
tripinnate intergrading forms. In the past the various IDicroidium nymboidense is unlike that of any
forms have been separated into species, subspecies Nymboidiantum spp. While we are uncertain of its
and formae. Jones and de Jersey (1947) considered generic placementit is certainly a distinct species and
the pinnate to tripinnate forms to represent an is described and illustrated to draw attention to its
evolutionary series but all forms are now known to presence in the Nymboida Flora.
occurtogether. Holmes (1982) noted that leaves with
intergrading characters ofmost forms werepresent in
the Benolong Flora. Dicroidium sp. A
Figures 20C,D
Description
? Dicroidium nymboidense Holmes sp. nov. Apical portions of two fronds with extremely
Figure 20B widely spaced pinnae; Fig. 20C is pirmate, pinnatifid
Diagnosis withslightlylobedorentirestronglydecurrentpinnae;
A small pinnate frond; pinnae few, large Fig. 20D is bipirmatewithbroadlytriangularpinnules
elongate-ovate, rounded, irregularly lobed; venation and basiscopic pinnules decurrent on or attached
odontopteroid, two decurrent veins entering each directly to the main rachis.
pinna then dividing acutely up to five times to form
parallel venation to the pinna apices. Discussion
The wide spacing of the pinnae of these
Description incomplete fronds sets them apart from all othermore
This new morpho-taxon is based on a single completely preserved fronds. They may possibly
frond impression and counterpart. Frond pinnate, represent aberrant forms ofD. dubium and D. zuberi
mm
115 long with base missing; curving rachis with respectively. D. pinnis-distantibus (Kurtz) Frenguelli
M
five opposite pairs of pinna and a terminal pinna, (Retallack 1977, Microfiche Frame 8) also has
increasing in size apically; fifth pair ofpinnae to 55 very widely spaced piimae but with entire pinnae
mm mm
long, 25 wide, elongate-ovate with two or margins.
three broad irregular obtuse lobes; texture ofpinnae
thin, venation conspicuous; two veins enter each Dicroidium sp. B
decurrent pinna at c. 30°, arch and divide from three Figure 20E
to five times then run parallel to each other to pinna
apex; the upper of the two primary veins forming a Description
short median vein. The apical portion ofa very large bipinnatifid
Dicroidium-Wks, frond with deeply dentate pirma
Holotype margins.
AMF125083 Australian Museum, Sydney.
Discussion
Type Locality This single portion ofan obviously very large
Coal Mine Quarry, Nymboida. Basin Creek fi-ond differs by its size and shape from all other
Formation, Nymboida Coal Measures, Middle bipinnatifid material from Nymboida. D. sp. B differs
Triassic. from the large D. dubium var. australe (Retallack
1977, Microfiche Frame H8) and D. dubium ssp.
Discussion helvetifolium (Anderson andAnderson 1983, p. 105)
A
This rare and problematical frond is placed by the acute triangular pinna lobes. distal fragment
with reservations in Dicroidium. The rachis curves of a somewhat similar-sized but pinnate frond from
sideways basally but the leaf base is missing and the Late Triassic Ipswich Basin was attributed
no fork is preserved. However, on the right margin by Walkom (1917) to Danaeopsis hughesi, now
of the specimen there is a pirmule fragment with Dicroidium hughesi Lele (1962), a species described
venation that divides and runs parallel and is aligned from India.
at an angle appropriate for a pirma ifitwere a portion
of a forked Dicroidium frond. The venation, while
Proc. Linn. Soc. N.S.W., 126, 2005
TRIASSIC FLORA FROM NYMBOIDA UMKOMASIACEAE
-
Fertile organs associated with Dicroidium leaves Discussion
U. sessilis differs from U. distans by the sessile
Genus Vrnkomasia ThomA% 1933 cupules. U. di'ciissalaAnderson andAnderson (2003)
has sessile cupules but differs by the more numerous
Type species pairs of cupules with a decussate arrangement. The
Umkomasia macleannhomas 1933 p.203, figs partial strobilus located on the lower left hand side of
1,2, 56; PI. 23, figs 1-4 Figure 1OA probably belongs here.
iJmkomasia distans Holmes 1987
Figures 21A; 22A Umkomasia sp. A
Figure 22C
Holotype
AMF 63824, Holmes 1987, fig. 3, fig. 4, A-D Description
A portion of a single elongate Umkomasia
Paratypes strobilus with well-separated alternate branches, each
AMF
63825-28 branch with one or more (?) pairs ofcupules borne on
long expanding pedicels; cupules strongly decurved,
mm
Description 8-1 long.
An Umkomasia strobilus with widely spaced
alternate and spirally arranged branches on an Discussion
elongate axis; branches with one or two opposite U. sp. A differs from U. distans and U. sessilis
pairs ofcupules andaterminal pairora single cupule; by the larger decurved cupules and by the elongated
mm
cupules pedicillate, irregularly rounded, 3.5 to 5 andexpandingpedicels.Thecupulesoff/,gracilliaxis
in diameter. Anderson and Anderson (2003) from the Molteno
Formation are decurved and ofa size similarto U. sp.
Discussion A. However, their cupules with four lobes each is a
U. distans was described and illustrated feature notpreserved on this Nymboida specimen.
by Holmes (1987) based on three specimens of
reasonably preserved cupule bearing axes from the Isolated Umkomasia cupules
Coal Mine Quarry, Nymboida. Additional material Figures 22D-F
has been collected (Fig. 21A) but this is fragmentary.
AMF
For comparisons with other material see Holmes Specimen 125092 (Fig. 22D) shows a
(1987). This species was regarded as distinct from the pair of adjacent and opposite dorsally compressed
mm
Molteno species by Anderson and Anderson (2003, semicircular cupules each c. 16 in diameter,
mm
Table 51). with a 4 wide peduncle attached to the proximal
margin. These cupules are closely comparable with
the cupule shown in a dorsal view ofKaribacarpon
Umkomasia sessilis Holmes 1987 (Umkomasia) feistmantelii (Holmes and Ash 1979,
Figures 21B,C;22B Fig. 6.2) from the Early Triassic Lome Basin
assemblage and with cupules from the Sydney Basin
Holotype (Walkom 1925a, PI. 31, fig. 9; Retallack 1980, fig.
AMF63829 and counterpart AMF63830, 21.9E). No additional specimens have been found at
Holmes 1987, fig. 5, fig. 6 A-C Nymboida to show whether these cupules may have
split on dehiscence into several lobes as occurs in U.
Paratype feistmantelii fructifications (Holmes and Ash 1979,
AMF63831 fig. 6. 3-5; Walkom 1925a, PI. 29, fig. 9; 1932, PI. 5,
figs 3-5; Retallack 1980, Fig 21.9F).
AMF
Description Specimen 125093 (Fig. 22E) is a thick
An elongate Umkomasia strobilus with roundedtaperingmass, 15 mmwide, ofcarbonaceous
mm
alternate, spirally arranged branches bearing two material attached to a stout peduncle 2 wide.
pairs ofopposite, sessile cupules 4-5 mm in On opposite sides of the mass are decurved acute
diameter. projections. Thisproblematic specimenmayrepresent
a pair of large fleshy conjoined cupules, perhaps
similar to Umkomasia sp. of Retallack (1980, Fig.
10 Proc. Linn. Soc. N.S.W., 126, 2005