The fossil record of Elaeocarpus L. fruits

37 Pages·2001·15 MB·English

by  Mary E Dettmann

THE FOSSIL RECORD OF ELAEOCARPUS L. FRUITS
MARY E. DETTMANN AND H. TREVOR CLIFFORD
Dettmann, M.E. & Clifford, H.T. 2000 06 30: The fossil record ofElaeocarpus L. fruits.
Memoirs ofthe QueenslandMuseum 46(2): 461-497. Brisbane. ISSN 0079-8835.
Elaeocarpus L. fruit stones from Australian mid-Tertiary sediments are systematically
described and their stratigraphic and geographic distributions recorded. The fossil fruit
stones comprise2-9-loculate innermesocarps, theiroutersurface with longitudinal sutures
and a sculpture ranging from smooth to pitted, baculate/verrucate/echinate'rugulate, or
fossulate. The fertile locules have a near apical seed and the seedless locules are usually
compressed; thin-walled endocarps dehisce loculicidally and enclosethe locules. The seed
coat is bitegmic and has a multiplicative tegmen, the outer epidermis ofwhich comprises
thick-walled,pittedsclereids. Fivetypesoffruitstonesaredistinguishedonthebasisoftheir
surficial sculpture. Types 1 and 2 have verrucate/rugulate surfaces, Type 1 being
distinguishedbyahigherratio(>0.1 )ofsculpturalbasediametentransversediameteroffruit
stone than that of(<0.1) Type 2 fruit stones. Types 3, 4, and 5 have pitted, smooth, and
fossulate surfaces respectively.
A review offossil fruit stones attributed to Penteune F.Muell., 1874, Pleioelinis F.Muell.,
1882, Phymatocaryon F.Muell., 1871, and Rhytidotheca F.Muell., 1871 confirms these
categoriesarecongenericwithElaeocarpus.Onenewspecies,E. rozefeldsii,isproposedand
typespecimensaredesignated forE. allportii(F.Muell.)comb,nov., E, angularis(F.Muell.)
Selling, 1950, E. bivalve(F.Muell.)comb. nov.,£. brachyclinis(F.Muell.)comb, nov.,
E. couchmanii (F.Muell.) comb, nov., E.johnstonii (F.Muell.) comb. nov. (and itsjunior
synonym, E. bassii Ettingsh.), E. lynchii (F.Muell.) Selling, 1950, E. muelleri Ettingsh.,
1886, E. pleioelinis (F.Muell.) comb, nov., andE. trachyclinis (F.Muell.) Selling, 1950.
The fossil fruit stone record confirms that Elaeocarpus was represented in the eastern
Australian floraasearlyasthe EarlyOligocene. Modificationstothedistribution rangeand
diversity levels ofthe genus have occurred in eastern Australia since the Neogene. These
involvedthelossoftaxawithType5stonesfromAustraliaandashifttomorenortherlyareas
H
ofeastern Australia ofspecies groups with Types 2, 3, and 4 fruit stones. Elaeocarpus,
fossilfruitstones, Australia, mid-Tertiary.
MaryE. Dettmann, DepartmentofBotany, The UniversityofQueensland,StLucia4072;
H. Trevor ( lijford, QueenslandMuseum, POBox3300. SouthBrisbane4101, Australia; 8
November200(1
Elaeocarpus L., agenusof-60or360 species small, psilate, tricolporate pollen are difficult to
(Mabberley, 1987) occurring in tropical and discriminate from those of certain other
warm regions ofthe Old WorldexcludingAfrica Elaeocarpaccae (e.g. Sloanea) and Cunoniaceae,
has an extensive fossil record from Tertiary and detailedcomparative studiesonpollenofthe
sedimentsofeasternAustralia.Thefossiltaxaarc two families are not available.
leaf remains that include compressions, The present paper incorporates a taxonomic
impressions and cuticles (Ettingshausen, 1883; accountoffossilElaeocarpus fruit taxa reported
1886; Christophel, 1994; Carpenter et al., 1994) from Australia, and documents their known
and fruits that occur as permineralised casts and geographic and stratigraphic distributions.
moulds or charcoalified compressions (Mueller, Eighteen species of fossil fruits referable to
1871a,b, 1873, 1874a,b, 1878, 1882; Johnston, Elaeocarpus have been reported from mid-Late
1880a,b, 1882; Ettingshausen 1883, 1886; Cainozoic sediments of eastern Australia, and
Deane, 1925; Berry, 1926; Kirchheimer, 1935; one is newly describedherein.
&
CrS-ehilrlii•snt*go.phie1l9i,501i;9n9nR6*oaz,efbu,elidns,pres1s9;90Bd;urRroozwesf,el1md9ns9-77)\. ScTTUiDmIiErScrn-O,xN4TFrFOnOCScAScRInPLIFL-nRSnUTITTCS. OAFC
Elaeocarpus-typc pollen has been recorded ex-
tensively from eastern Australian sediments of m 1883 Ettingshausen described fossil fruits
late Eocene-Recent age (Luly et al., 1980; from Tasmania under the name Elaeocarpus
Truswell et al., 1987; Kershaw et al., 1994; bassii and was thus the first to attribute fossil
Blackburn & Sluiter, 1994; Martin, 1998)but the fruits to the genus from the Australian Tertiary.
) )
462 MEMOIRS OF THE QUEENSLAND MUSEUM
However, the earliest reference to fossil fruits resembledrare5-valvedspecimensofPleioclinis
consistent with those ofElaeocarpus dates from couchmanii, the diagnosis of which specified
&
Mueller Smyth's (1870:390) reference to a 7-9, rarely 6 valves. It is also likely that Mueller
'five valved capsule of an unknown genus' had doubts thatPenteuneand Rhytidotkecawere
recovered from deep lead sediments at Haddon, discrete genera as he noted for Penteune clarkei
Victoria. Specimens answering to that that 'some affinity ofthese fossils to those ofthe
description were subsequently attributed to genus Rhytidotkeca is evident (Mueller, 1874b:
Penteune, which name is 'an allusion to the five 41). Thus, over 8 years Mueller directly or in-
valves in which the seeds are imbedded' directly interlinked Rhytidotkeca^ Penteune,
(Mueller, 1874a; 41). Mueller is the authorof10 Phymatocaryon and Pleioclinis all ofwhich are
other fossil fruit species consistent with Elaeo- here associated with Elaeocarpus (Table 1).
carpus and instituted (1871a,b, 1874a, 1882)
Rhytidotkeca, Phymatocaryon^ Penteune and Because Mueller possessed so wide a
Pleioclinis, to accommodate them. Pleioctinis knowledgeoftheAustralian Floraitisofinterest
was firstusedwithoutadescriptionorillustration to speculate as to why he created these 4 fossil
in a note accompanying the description of genera. Indiscussionsaccompanyingthegeneric
Rhytidotkeca pleioclinis (Mueller, 1873), and descriptions, Muellermade it clearhewas ofthe
was formally described later (Mueller, 1882). opinion that it was improper to assign frag-
mentary fossil material to extant genera.
Mueller's generic diagnoses must be treated However, for Penteune, Phymatocaryon and
with caution forthey do not always comply with Rhytidotkeca he suggested affinities with
the characters displayed by the species he Sapindaceae amongst living taxa thereby
attributed to them. For instance, Penteune was strengthening the view the fossil genera were
proposed for fruits with surface sculpture 'very related. HisreasonsforselectingtheSapindaceae
slightlyroughonthedorsalpart'(Mueller, 1874a: are spelled out in the discussion accompanying
41), but on the same page, P. trachyciinis, is the description of Phymatocaryon (Mueller,
described as 'externally very rough, almost 1871a: 47). There he noted that the pendent
vernacular'. For Pleioclinis seed placentation seeds, locular dehiscence, drupaceous fruits
was diagnosed as pendent (Mueller, 1882: 43), ("with a distinct sarcocarp and putamen') are all
but in Rhytidotkecapleioclinis F. Muell., which family characters of the Sapindaceae.
was later transferred to Pleioclinis (Mueller, Furthermore, he observed that the fruits ofsome
18S2),theseedsaresaidtobeerect. Furthermore, members of the family also have surfaces that
fruit-valve number, a character that initially exhibit 'tubercular roughness'. He may have
discriminated the genera, was later found to be beenparticularlyswayedbythischaractersince5
variable; original diagnoses were not emended. years earlier he had described Cupaniopsis
tomentella(F.Muell.ex Benth.)S.Reyn., fruitsof
Mueller had difficulty in discriminating which are described by Reynolds (1985) as
between his genera. For instance, he (Mueller, 'valves, thick rugose, deeply wrinkled outside'.
1874b: 42) emphasised the sculptural similarity
between the 5-valved Penteune trachyciinis and Nonetheless Muellerwas not entirely satisfied
the2-4valved Phymatocaryonmackayi,but later his fossil fruit genera all belonged in the
(Mueller, 1875: 41) reported that fruits off! Sapindaceae for a few months later he noted the
trachyciinis with 4 valves bore much similarity of Rhytidotkeca to the fruits of
resemblance to P. mackavi. Subsequently, he FlindersiaandCh'loroxylon(Mueller, 1871b: 39)
(Mueller, 1882: 43) observed that P. trachyciinis then placed in the Meliaceae. He discounted an
IABLE 1.Thethreeprincipalcharacters stressedbyMuellerin hisoriginaldiagnosesoffossil fruitgeneranow
associated with Elaeocarpus. Bracketed valve numbers designate the less common state(s); bracketed shape
and sculptural designationscomply with terminology used herein.
Genus Valvenumber Shape Surface
Penteune (4)5(6) Ovate-globosetobroadlyovate(spherical- Slightly rough-deepiywrinkled(pitted-
prolateellipsoidal) verrucate/ruuulalc)
Rhytidotkeca 5 Ovate(perprolateellipsoidal) Wrinkled(rugulute)
Phymatocaryon 1 [4) Sphericaltoovate(spherical-prolate) vReorurguhc-adtee-erpuigyulwarki-nkled(baculate/
Pleioclinis Ovateglobular-ovate(spherical-prolate) Tubercular-rough verrucatc-rugulate
I
FOSSIL FRUITS OF ELAEOCARPUS 463
affinity with the Sapindaceae because 'the including fossil leaves and other plant organs in
numberoffruit-valves, increasedto live,remains extant genera.
exceptional1. Though most Sapindaceae have The likelyaffinityofPhymatocaryonmackayii
3-loculatc fruits there are many with fewer or F.Mucll., 1871 with Elaeocarpus was noted by
more carpels and so it is surprising Mueller was Deane (1925) and confirmed by Kirchheimer
so concerned on that account. Likewise, in his (1935) after detailed comparison of fossils and
discussion ofthe affinities ofPenteunehe wrote, mesocarps of extant E. angustifalius. Selling
kIt belonged, however, most probably to (1950) formally transferred several ofMueller's
Sapindaceae,although thepossibilityofithaving species to Elaeocarpus without comment. As
formed a genus of the meliaceous order is not noted by Selling (1950) transfer ofRhytidotheca
excluded'(Mueller, 1874a: 41). lynchii, Penteune clarkei, P. allportii, P.
Why Mueller failed to recognise the similarity brreancdhyecrleindis,PaenndtPe.utnreachyacnHdnisRthoyEtliadeooctahrepcuas
ofat least Penteunewith fruits ofElaeocarpus is superfluous. Rozefelds (1990) and Rozefelds &
difficult to understand unless it stemmed from Christophel (1996a, b; in press) provided
mhaitmerbieailngthfaatmilhiaadr,baeseins lpirkeelsys,edonloyr wsittohrefdresihn evidence forreference ofseveral fossil fruit taxa
to Elaeocarpus. These include E. clarkei, E.
preservative. In these circumstances the fruit
spackmaniorum, E. cunnmgiU and E. mackayii.
stones do not disintegrate. They do so after However, they expressed doubts about a
prolongedexposuretowet/dryweatheringcycles
relationship, as suggested by Selling (1950), of
and/ordegradation by fungi and/or insect attack.
Rhytidotheca lynchii, Penteune trachyclims, and
Under these circumstances disintegration ofthe
Phymatocaryon annulare with Elaeocarpus.
stony mesocarp into segments may occur after
loss ofvascular tissue from the central cylinder MATERIAL
and the radial strands that connect the central
cylinder to the segment sutures. It is likely that Fruitsreporteduponhereincludecharcoalilled
fossil fruitsweresubjectedtosimilardegradation material from subsurface sediments near
processes during incorporation into sediments. Moranbah and Blackwaterin central Queensland
Moreover, many of those from deep lead and permineralisedmaterial from Gleneoe (Roz-
&
sediments have been pyritised, and oxidation efelds, 1990; Rozefelds Christophel, 1996b, in
after recovery and storage may cause the meso- press) that are held in the Queensland Museum
carp segments to separate. Thus, Mueller was (QMF); charcoalilled and lignified fruits in the
very likely distracted from the real identityofthe Australian Museum (AMF), the Department of
fossil fruits because many dehisced into Mines, Geological Survey ofNew South Wales
segments as a result ofburial, fossilisation, and (MMF), and the Museum ofVictoria (NMVP).
subsequent retrieval and storage. The last mentioned collection includes most of
Mueller'sVictorian material originally housed in
By 1884 Muellerwas aware that Rhytidolheca the Geological Survey of Victoria. Mueller's
fruits were considered consistent with those of material came from deep lead sediments in Vic-
Elaeocarpus because R. jolmstonii F.Muell.(in toria and Newr South Wales and specimens
Johnston, 1882) was based on material assigned illustrated by Mueller have been identified for
by Ettingshausen (1883) to E. bassii Ettingsh. most of his Victorian species. Mueller's New
Nonetheless, hedidnotcommenton orchallenge SouthWalesmaterialmayhavebeendestroyed in
Ettingshausen's assignment. Likewise it is the Garden Palace fire in 1882 which included
surprising that Ettingshausen (1883: 63), having 'the palaeonlolouical specimens ofthe recently
recognised that E. bassii and R.jolmstonii were deceased Reverend W.B. Clarke' (Gilbert, 1986:
conspecific, relegated other species ofRhytido- 107). Clarke collectedseveral ofthe fruitsdescribed
theca (R. lynchii, R.pleioclinis)toIncertaesedis by Mueller from New South Wales. The reposit-
withoutcomment. Furthermore,hedidnotsuggest ory oftypematerialofE. muelleriEttingshausen,
that Penteune and Elaeocarpus might be 1886 is also unknown. Tasmanian material
congenericand followedMuellerinassigningthe attributed to E. bassii by Ettingshausen (1883)
former to Sapindaceae (Ettingshausen,1883:16) and to Penteune allportii and Rhytidotheca
rather than Tiliaceae in which at that time jolmstoniiby Mueller(in Johnston, 1882) has not
Elaeocarpus was included. In a lecture to the been located. Other material not examined
Royal Society of Tasmania, Mueller (1884) includesRhytidothecama&jorDeane, 1925 and E.
restated his argument (Mueller, 1871a) for not cerebriformis Rozefelds Christophel, 1996b.
S
.
464 MEMOIRS OF THE QUEENSLAND MUSEUM
10"s
LOCALITIES. (Fig. 1).
Queensland, a) Picardy Station, near Moranbah
21°5'17.6"S 147°50'34.3"E. HolesRDPD98MA
17, 1 11-133m and RDPD98MA21, 123-133m.
Fruits were recovered from both boreholes from
nearthebaseofathin sequenceofsandsandsilts
with interbeds oflignites that occur beneath bas-
alts and overlie Permian Coal Measures (Fig.
2A). The basalts are probably related to those of
20'
the Nebo Province dated as 30-34 Ma. (Wellman
& McDougal. 1974, Sutherland pers. comm.),
but stratigraphic relationships betweenthese and
those intersected in the boreholes have not been
established. Host sediment was not retained for
palynological assessment. Thus, the minimum
ageistentativelysuggestedasEarlyOligocene.
b) Near Blackwater 24°1*13"S 148°48'50"E.
South Blackwater Coal Pty Ltd Hole R8736,
82m. Fruitsare from ligneousbandswithinsands
that underlie basalts and overlie Permian
SW
sediments (Fig. 2B). Basalts to the in the
Springsurc area are dated at 27-33Ma. (Suther-
land et al„ 1977), but stratigraphic relationships
between dated basalts and those in borehole
R8736are uncertain. Moreover,nosedimentwas
available for palynological analysis and a
minimum age of Early-Late Oligocene is
tentatively indicated. 40°S'
c) Glencoe Station (23°36'S 148°06'E), near
Capella (Rozefelds, 1990; Rozefelds &
Christophel, 1996a, in press). These occur in
silcretes that overlie undated basalts. However,
volcanics near Emerald are dated as 30-32Ma. FIG I. Map of Australia showing localities from
and those southwest ofCapellaat26Ma. imply a which fossil fruits of Elaeocarpus have been
reported.
maximum Oligocene age, possibly Late Oligo-
cene-Early Miocene according to Rozefelds
mesocarp fossils from the separate localities
(1990). derived from the same plant taxon.
New South Wales. Elaeocarpus-typQ fruits are c) Gulgong district (32°12'S, 149°32'E) sites
known from ascatteroflocalities on the western include Home Rule Lead and Black Lead
flanks of the eastern highlands ofnorthern and (Mueller, 1876, 1877, 1879). The leads are
central New South Wales (Fig. 1). sediment fills of palaeodrainage channels on
a) Newstead near Elsmore (29°47'S, Palaeozoic basement and are overlain by basalt
151°17'E) Ettingshausen (1886) described E. flows. Isotope ages (K-Ar) ofthe basalts range
3
muelleri fruits from ironstones overlain by from 13.8± i.2Mato 14.8 ± 1.2Ma.(Dulhunty,
basalts. Pickett et al. (1990) concluded an Early 1971; Meakin &Morgan, 1999). TheHomeRule
Miocene oryounger age based on K-Ar dates of Lead is assigned to the Middle Miocene
20.5±0.2Ma. for a nearby basalt flow Triporopollenites bellits Zone (McMinn, 1981)
b)At Witherden'sTunnel, nearEmmavilleleaf in agreement with isotope dates.
fossils attributed to E. muelleri were recorded d) At Orange (33°17'S. 149°06'E), Carcoar
fromcarbonaceousclaysandsiltsbeneathbasalts (33°37'S, 149°08,E), and Bathurst (33°25'S,
(30.4±0.3Ma.); palynological dates confirm a 149°351E) charcoalified fruits are known from
Late Eocene age (Pickett et al., 1990). However, beneath basalts in numerous deep leads and reef
there is no evidence to surest that the leafand mines. As at Gulgongthe sediments are overlain
YOS$lLVKUrTSO¥£LAEOCARPl
s
KwTintoExploration SouthBiackwaterCoalPty Ud
B
HoleROPD98MA21, PicardySUtkx. HoleR8736,
near Moranbah nearBlaekwatet
m rivpthim)
U=R1IARY
TERTIARY » volcanic?
nlk-ii
foil
slttycliyt sand-ftli
liontatmnda rruribtwtftf
horizon
PERMIAN
basalt Rllll-illill!;
interval
wnn
fruits
FIG. 2, Straiigraphic sequence* and sampling horizons in: A, Hole RDPD998MAJ7* Picard) Station near
Moranbah, Queensland; and B, South Blackwater Coal Ply Lid Hole R8736 near Blackwater, Queensland.
b\ basalts thai were extruded from Mt. Crucible Co. Shalt, -23.2m (76IU) are desig-
(.anaholosSW ofOrange. Basaltsthatoverliethe nated as (he source ofthe fruits (Mueller. I S74e:
sediments at Karesl Reefs and m the Lucknow 2l>. Map I), and in the Reform Shaft the sedi-
Mine have pro\ tded dates of 11.2-loMa., late ments are beneath basalts. Dates of2-5Ma. have
Middle-early Late Miocene(Johnson, 1989). been obtained from basalts linn ovcilie alluvial
Victoria. Fossil Quits have been collected from sediments in the Ballaral district (Sutherland,
buried plaeer deposits in ibe Murray, Cippsland 1995) and palynollorus recovered from sedi
and Otway Basins, and on the northern Hanks of meats intercalated between two of the Flows
Hie Great Divide. Hiere are lew data on precise indicate the Tubulifhtiiiltfcs pUt$teC£fth US
ages ofthe d^^-p lead sediments. spore-pollen /one oflate Pliocene to Pleistocene
a)The EldoradlIdeeplead(TedOvens(i & I \1 age l Partridge. L995), fhe minimum age ofthe
Co, Shaft), Beeebworth(37°l8,S 146 32'U in iubbasaltic sediments is Lhus Late Miocene «igct
the Murray Basin, (Mueller. 1874c: fig. 2) is of butRoyelclds&Christophelt 1996b)fugue foran
unknown age. Palynological dates on olber Mur- arlv-Middle Miocene age based on association
1
ray BasindeepleadsnearSlavsell, \V Victonaand ofthe fruit E mackayii (KMuell.) Kjrchheimer,
inthfi Woodstock 10008 bore near Bendigo. Vic- E. tiiirkfi (hMucll.) Selling, and Spotnhio-
toria, indicate Oligoccnc and late Parly Miocene sfmbtts sxmihii F.Muell. (see also Greenwood et
ages based on refewn, , . i rhc formertothe lower al. 2000), Unfortunately. Muellerdid not al
P?ote&cidite& u<hcicidalus spore-pollen Zone specify precise depth or mine shaft ^m' localities
and Of the latter lo the upper paTt of that iouc at lladdon.
(Partridge&Wilkinson,l982; Partridge, 993)-
I
h) The bulk offruitsdescribed by Muellerwcie c) ( iippsland Basin deep lead sediments at
obtained from deep lead sediments at Smythc's Taihol (37 LO'S 146 14 I ). Foster (37°I0*S
Creek (Reform Co. Shaft) and ftirttiflgbool l4o l-N-iand Tanjil ( 38»OrS l46°i4'E). thus
(Crucible Co. Shaft), near lladdon (37aI8'S far, the ages ofthese have not been resolved, hut
32'E), SW of Ballarat. Sihv sands ofthe as for the Haddon sediments Ro/efeld
'wash dirt* overlying basement near the bottom Christophel (1996b) suggest an Early-Middle
ul'ihe shafts ( Reform ( 0, Shalt. --47.5m ( 15611); Miocene a
466 MEMOIRS OF THE QUEENSLAND MUSEUM
b) At Schultz Creek (42°16'48"S 171°
07' 2"E),northofGreymouth, South Island lake
1
sediments deposited during the last interglacial
( 1OOKa.)haveyielded fruitscomparabletothose
q[' extant E. dentatus (Burrows, 1997). The
material illustrated has been deposited at the
Queensland Herbarium.
METHODS
Charcoalified specimens were photographed
after whitening with ammonium chloride;
internal characters including those ofthe locules
werephotographedwithoutwhitening. Mueller's
type specimens are figured together with
reproduction of the original lithographic
illustrations, the latter ofwhich represent mirror
images. Available seed coats were examined in
transmitted light after clearing in chromium
trioxide for several hours, followed by thorough
FIG.3.StylisedtransversesectionofElaeocarpusfruit washing in distilled water after which they were
sincuwlhpitcuhreeaocfh otfhethein5necrarpmeelssohcaasrdpi.ffear,entaxsiusr;ficcall, mounted onglass microscope slides in glycerine
compressed locule; e, endocarp; en, endosperm; em, jelly.
embryo; ex, exocarp; f, fibres; im, inner mesocarp; Fruit stones ofextant E. angustifolius Blume
om, outermesocarp; I, locule; r, raphe; s, suture. and E. reticulatus Smith were collected from
trees growing at the Brisbane Botanic Gardens,
d) Yallourn Coal Measures, Yallourn (38°10'S Mt Coot-tha and the University ofQueensland.
146°2l'E) inthe Yallourn Formation whithinthe SYSTEMATIC PALAEOBOTANY
Tripompoilenites bellus spore-pollen Zone of
late Early-Late Miocene age (Rozefelds & Elaeocarpus L.1753
Christophel, 1996b).
RhytidothecaF.Muell., 1871:39.
South Australia. Fruits from subsurface sedi- Pkymatocaryon F.Mucll, 1871:41.
ments near Bethany (34°32'S 139°00,E) in the PentemeF.Mucll., I874a:3941.
BarossaBasin(PatersoninHossfeld, 1949)occur PleioclmisF.Muell., 1882:43
in the Rowland Flat Sand, a lignitic fades over- TYPESPECIES.Elaeocarpusserratus L
lain by coarse-grained fluviolacustrine sand and
gravel fining upwards to bedded silt and clay. FRUITS OF EXTANT ELAEOCARPUS. In
Palynological dates are Early Oligocene-Early extant Elaeocarpus fruits develop from flowers
Miocene for the basal sediments and Early-Late with a superior 2-8 (usually 3-5) loculate ovary
Miocene for coarser upper facies (Alley, 1995). that terminates in a single style surmounted by a
lobed stigma. There are 2-12 anatropous ovules
Tasmania. Deep leadsedimentsatBrandy Creek,
per locule and these are attached to an axile
Beaconslleld (41°12'S 148°49'E) and outcrops
at Launceston (41°27'S 147°10'E). A mid- placenta. When there are few ovules per locule
they are attached high up on the axis and so
Tloecratliiatriyes(bOalisgeodceonne)paalgyneofisloirnadsicoaftesdub-fobrastahletisce appear to be subapical (Figs 4, 6B,C). Only one
sedimentsintheTamarGraben(Forsyth, 1989). ovuleperloculedevelops intoaseedandinsome
species there is only 1 seed per fruit. Expansion
(4b1)°1C2al,Sca1r4e8o°u4s9f'rEui)tsocfcruormiGneitrlasvteorntiBnaey,whHiocbharits othfetahidsjo1insienegdsueseudallelsys rleoscuulltessi(nFicgosm4p,re5sDs,iGo)n. of
equivalent in age to nearby basalts dated as 22.4 The fruit is a drupe, a useful term, ill-defined
Ma, Early Miocene (Tedford et al., 1975). but widely used in the literature (Clifford &
New Zealand, a) At Mangonui (35°00'S 173° Dettmann, in press). In most species the fruit
20*24"E), north ofAuckland, charcoalfied fruits surface is iridescent blue,the colourdue knottoa
occur Mangonui Formation lignites (Late Mio- blue pigment, but by the structure ofthe cuticle
cene) associated with Cocas zeylandica Berry which reflects blue light' (Lee, 1991). The outer
(Berry, 1926;Thompson, 1978;Isaacetal. 1994). fleshy mesocarp is usually thin and densely
OSSIL FRUITS OF EL IEOCARPUS
I
em
--.
PIG, 4.sectionsthrough Eiwficarpuxangmtifvlitu Slume fruits. A. transverse; B, vertical. Labelsas for hi|
llbrous. fibres intertwined with the sculptured (Ro/cfelds & Chnstophel, 1996a). How.
surface of, or extending into, the woody inner such features have been little utilised and remain
mesocarp (Fig, 5A)« The latter encloses the largely undocumented in current infrageneric
loctilesandtheirsurroundingeiidoeaipstot'onna classifications, Moreover, liltle attention has
sloncwhoseoutersurface may be smooth, pilled, heen accorded anatomical features of (lie inner
fpssulate. cumulate, baculate. veimeate, or mesocarp wall and the enclosed endocarps.
cchinatc. At matutih the vascular strands ofthe The only record of fossil mesocarps attribut-
avis decay, resulting m a hollow cylinder which able to an extant species is thai off.- dentalUS
extends for - two-thirds the length o\ the stones in interglacial (lOOKa) sediments from
mesocarp from its proximal end (Figs \ 4. 5F ). New Zealand (Burrows, 1997),
Further deca\ ot radially disposed strands be-
neath thesegment sutures results in themesocarp FOSSIt FRUITS, The following fossil laxa arc
and enclosed endocarps splitting vertically considered congeneric with Elaeocarpus.
between the septa to produce segments that P/:\m Nl F.MuelK. 1874; lypc species
expose seeds or aborted ovules on their radial (designated here) PenteU#8 dorkel F.MuclI.,
walls (Fig 5G-I), As was recognised by Mueller, I874A: 41 from l.lsmorc, NSW; Larlv Miocene:
such dehiscence is locuhcidal. The lines along Neplype [designated Ro/elelds & Chnstophel,
wmhairckhedtbhye vmeertsioccalarsputusprleistoanretheussuuarlflaycecloelartlhve M1u99e6lal:er4':s.». fNiMgVurPe5d39s1p3e,ciFmiegn.s1 1fAr.oBm; Haddon
mesocarp dig. 5A-C)and each segment consists tSmythc's Creek. '.'Kch-rni Co. Shaft. 47.5m)
ofa woody, inner mesocarp. half of 2 adjacenl arc believed lost (Ro/efelds & Christophcl,
endocarps and associated loctiles(Fig. 5H,I). The
1996a: 4 •>.
inner surface ol the endocarp may bear multi- Specie;s included: P aWpnrtii l.Mttelk, in
cellular scales and hairs (Cornet; 1976). Johnston 1882: Rg840,41,44-47..P hnuhvclinis
Seeds are fusiform bitegmic with glabrous or F.MuelL 1874a:4LpK8 f\z> 1-9, p frachydmh
hairy surfaces. The testa is sfiygral cells thick F. MuelL 1874a: 41, pi. K, Tigs 10-17.
with an outer epidermis of elongate to iso- rHYMlTOCARWN I.Muell., 1871; type
adrieamelitgrniicriccedl.lsTh(Feiglse6g1m,e7nF)isthmautltinipsliocmaetisvpee,ciet:h-e; 4sp1eci(Lb.ys/ m//io/n;/o;//t..y-,pv/e/-v1 'f/rMo;um;r/Huan/dTd.oVUnicl(lS.t,ny1tXh7<la.
outer epidermis ofwhich has fibrifonn. bull L'reck. Reform ^'o.i Shaft, -47.5m), sediments
or dumb-belt shaped, lignilied selereids ihat arc beneath basalts; -karly-Middle Miocene;
arranged longitudinally (Ftg. 5G.II), Both the LccLoivpe (dcsmnaied Ro/elelds & Chnstophel
wtchsilachandfotregmmsenthaerew\oasocdulyaribsaesdalin athnedchaacluat/ea (inpress))NMV33562, Mueller I871a;41apit2,
prominence ol the seed (Comer, 1976). Species included:^ annulare F.MuelL IS74b:
Shape, structure and surface sculpture of the 4UpI. ]() figs 1-4. r BtofiwF.MudT.,1877: I8D.
fruit stones are useful species discriminators . 59,p] | , HgJ fr-9
468 MEMOIRS OF THE QUEENSLAND MUSEUM
j &
Cl ^r w* ;F
~\m D q^
FIG. 5. FruitstonesofElaeocarpusangustifoliusBlume.A-C,apical, lateral,andbasal viewsshowingsculpture
ofwoodyinnermesocarpwithattachedfibre(f)bundlesandsutures, x 2.D,photographoftransversesectionin
Fig. 4Aoffresh, mature fruit stone showing sculptured woody inner mesocarp with fibre (f) bundles on the
surface, enclosedendocarps,twowith fertile loculescontainingseeds,each showingthe raphe,theotherthree
loculescompressed, X 2.E,photographofverticalsectioninFig.4Bshowingvascularstrandsintheaxis, X 2.
F, vertical section portion ofinner mesocarp removed to reveal closely adpressed endocarp, X 4. G, internal
view oftwo segments ofpartially rotted mature fruit stone showing seed in fertile locule (left), compressed
sterilelocule(right),andhollowaxis, X 4. H,I,internal viewsofdehiscedsegmentcomprisinginnermesocarp
with halves oftwo adjacent locules and endocarps showing ridges and grooves on their inner surface, x 4.
Labels as for Fig. 3.
FOSSIL FRUITS OF ELAEOCARPUS 469
!
/
-L "S\
* H
gJJ
FIG.6. MicrographsoffruitstonesofElaeocarpusangustifolius Blame. A,transversesectionofinnermesocarp
and endocarp walls, and profile of a concave-crested, arched sculptural element at outer surface of inner
mesocarp, x 10. B,C, detail ofcompressed localewith 2 near-apical, axially attached, aborted ovules(o), and
showingcoalescenceintheapicalregionofloculeofridgesandgroovesoninnersurfaceofendocarp, x 10,and
X 25 respectively. D, scanning electron micrograph ofsection ofinnermesocarp and endocarp, X 30. E, F,
scanningelectron micrographsofoutersurfaceandtransversesectionofinnermesocarp, x 2000and X 4000
respectively. G, outer epidermis oftegmen showing elongated sclereids with thick, pitted walls, x 750. H,
subspherical sclereid from tegmen, x 500. I, outercuticle oftesta, with a sclereid (sc) arrowed derived from
outerepidermis oftegmen, x 250. Labels as for Fig. 3.
470 MEMOIRS OF THE QUEENSLAND MUSEUM
FIG.7.Fruits, fruitstonesandseedsofElaeocarpusreticulums Smith.A,apical(upper)andbasal (lower)views
offruit stone, X 2. B, lateral view of fruit stone, X 2. C, vertical (left) and transverse (right) sections of
immature fruits, X 2. D, transverse section offruit with outer mesocarp partially detached (left) revealing
surfacesculptureofinnermesocarp, x 6. E, vertical sectionofimmature fruitshowingaxile placentationand
axial tissue, x 16. F, outercuticle oftesta, x 250. Labels as for Fig. 3.
PLEIOCLINIS F.MuelL, 1882; type species carpus are distinguished from other Elaeo-
Pleioclinis couchmanii F.MuelL, 1882: 43 (by carpaceae by: a pitted, smooth, fossulate,
original designation) from Haddon (Smythe's verrucate, baculate, echinate or rugose surface
Creek or Nintingbool), sediments beneath with longitudinal sutures that delimit segments;
basalts; ?EarIy-Middle Miocene; 2-9 carpels; passive loculicidal dehiscence into
Lectotype (here designated). NMVP53920, segments; 1-seeded locules; fertile and sterile
Mueller 1882: 41; pi. 19, figs 9, 10; locules, the latter often compressed; ovules
Species included: Rhytidotheca pleioclinis anatropous, pendulous with a ventral raphe;
F.MuelL, 1873), (=R shepherdi F.MuelL, 1882; seeds fusiform with a multiplicative tegmen, the
junior objective synonym) outer epidermis with pitted and lignified
RHYTIDOTHECA F.MuelL. 1871; type species longitudinal cells. Further characters with
apparent interspecific significance include the
Rhytidotheca lynchii F.MuelL, 87lb:39 (by
1 internal organisation and structure ofthe woody
monotypy) from Haddon(Nintingbool, Crucible
innermesocarpandsubjacentloculesasrevealed
Co. Shaft, ^23.2m), deep lead sediments; ?EarIy
by extant E. angustifolius and E. reticulatus
-Middle Miocene.
Lectotype (here designated). NMVP6034, (Figs 5-7). In E. reticulatus young fruits have an
axial column ofelongated vascular strands (Fig.
sMueeglmleenrt,) a1n8d71Nb:M3V9P,60pi3.34,, Mfuiegl.le4r,(r1i8g7h1tb:ha3n9d, 7E). Theinnermesocarpconsistsofisodiametric
pi. 4, fig. 1, centre segmentbroken at apex. stonycells, andtheenclosedendocarpscomprise
Species included:/?, johnstonii F.Muell.in lignified cells, elongated and arranged tangent-
Johnston, 1882: 50, fig. 6.R. majorDeane, 1925: ially. Fruitsof£. angustifoliusaresimilarintheir
491, pi. 60, fig.12 (nomennudum). R.pleioclinis organisation (Figs 5,6). Fruit stones allowed to
rot on the ground decompose slowly, but after
F.MuelL, 1873: 42, pi. 6, figs 1-4. time they may split into segments. Those ofE.
REMARKS. Recognition of fossilised Elaeo- angustifolius were little affected after subjecting
carpusstoneshasbeenbasedonsculpture,shape, them to conditions that may be expected in a
locule number, and position ofseed attachment medium-high energy depositional situation.
(Deane, 1925; Kirchheimer, 1935; Selling 1950; Stones agitated with sand, gravel and water in a
Rozefelds, 1990; Rozefelds & Christophel, closed container on a rotary shaker for 1 week
1996a, b, in press). Woody mesocarps ofElaeo- showed no signs ofbreakage or abrasion ofthe