Diversity in obscurity: fossil flowers and the early history of angiosperms

14 Pages·2009·0.95 MB·English

Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
Phil.Trans.R.Soc.B (2010)365,369–382
doi:10.1098/rstb.2009.0227
Diversity in obscurity: fossil flowers and
the early history of angiosperms
Else Marie Friis1,*, Kaj Raunsgaard Pedersen2 and Peter R. Crane3
1Department of Palaeobotany, Swedish Museum of Natural History, Box 50007,
10405 Stockholm, Sweden
2Department of Geology, University of Aarhus, Universitetsparken, 8000 A˚rhus C, Denmark
3School of Forestry and Environmental Studies, Yale University, 195 Prospect Street,
New Haven, CT 06511, USA
In the second half of the nineteenth century, pioneering discoveries of rich assemblages of fossil
plants from the Cretaceous resulted in considerable interest in the firstappearance of angiosperms
in the geological record. Darwin’s famous comment, which labelled the ‘rapid development’ of
angiosperms an ‘abominable mystery’, dates from this time. Darwin and his contemporaries were
puzzled by the relatively late, seemingly sudden and geographically widespread appearance of
modern-looking angiosperms in Late Cretaceous floras. Today, the early diversification of angio-
sperms seems much less ‘rapid’. Angiosperms were clearly present in the Early Cretaceous,
20–30Myrbeforetheyattainedthelevelofecologicaldominancereflectedinsomemid-Cretaceous
floras,andangiospermleavesandpollenshowadistinctpatternofsteadilyincreasingdiversityand
complexity through this interval. Early angiosperm fossil flowers show a similarorderly diversifica-
tion and also provide detailed insights into the changing reproductive biology and phylogenetic
diversity of angiosperms from the Early Cretaceous. In addition, newly discovered fossil flowers
indicate considerable, previously unrecognized, cryptic diversity among the earliest angiosperms
known from the fossil record. Lineages that today have an herbaceous or shrubby habit were well
represented.Monocotyledons,whichhavepreviouslybeendifficulttorecognizeamongassemblages
ofearlyfossilangiosperms,werealsodiverseandprominentinmanyEarlyCretaceousecosystems.
Keywords: abominable mystery; Darwin; early angiosperms; Early Cretaceous;
fossil flowers; monocots
1. INTRODUCTION partsofEurope(e.g.Saporta1878).Thecomposition
As improved information about the plant fossil record ofthesefloras,whichwereusuallydominatedbyflow-
accumulated during the late nineteenth century, the ering plants, contrasted strongly with that of earlier
history of flowering plants stood out as increasingly floras in the Cretaceous from which angiosperms
anomalous. Darwin’s famous quote, ‘The rapid deve- were apparently absent. To reconcile these obser-
lopment as far as we can judge of all the higher plants vations Darwin speculated that angiosperms perhaps
within recent geological times is an abominable had a long cryptic history in remote areas, which left
mystery’, in a letter written to J. D. Hooker 22 July no trace in the fossil record (Darwin in letter to
1879, captures his own reflections on this topic. Not J. D. Hooker 22 July 1879; see also Friedman 2009).
only was the first appearance of angiosperms during This idea has been reappraised several times since,
the Cretaceous relatively ‘recent’interms ofthe evolu- and it has been suggested that the rapid appearance
tionofothergroupsofplants,buttheapparentrapidity of angiosperms in the Cretaceous fossil record reflects
with which modern-looking angiosperm leaves entered a pattern resulting from migration rather than in situ
the fossil record in several different parts of the world evolution (e.g. Axelrod 1952).
appeared to undermine the possibility of gradual Over the past 120 years, an improved stratigraphic
evolution from some pre-existing form. frameworkcombinedwithdetailedstudiesofCretaceous
Central to the mid-nineteenth century interest in angiosperms has removed some of the tensions at the
early angiosperm evolution was the discovery and heartofDarwin’sconcerns.ThediscoveryofrichEarly
study of rich Cretaceous leaf floras from North Cretaceous palynological assemblages showed that
America, Europe and the Arctic. These leaf floras angiosperms were present in the Early Cretaceous, but
included the flora of the Dakota group in North missing from the very large numbers of Jurassic and
America (Newberry 1863; Lesquereux 1874), floras older samples from all over the world (e.g. Scott et al.
from western Greenland and Spitsbergen (e.g. Heer 1960). It was also recognized that the earliest angio-
1868, 1882; Saporta 1877) and floras from different sperm pollen grains were restricted to monocolpate
rather than triaperturate forms (Muller 1970; Hughes
1994), which was consistent with phylogenetic inter-
*Authorforcorrespondence([email protected]).
pretations that extant magnoliids were most similar to
Onecontributionof16toaDiscussionMeetingIssue‘Darwinand thecommonancestorofalllivingangiosperms.
theevolutionofflowers’.
369 Thisjournalisq2010TheRoyalSociety
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
370 E. M. Friis et al. Diversity in obscurity: fossil flowers
In the mid-1970s, studies of fossil angiosperm of the most convincing detailed information is from
pollen and leaves in the Early and mid-Cretaceous the Early Cretaceous palynofloras described from
Potomac Group of eastern North America demon- southernEnglandbyHughesandcolleagues(seeover-
strated a coordinated increase in the complexity of view in Hughes 1994). These records range in age
both types of organs (Brenner 1963; Doyle & Hickey from the Hauterivian through to the Aptian and have
1976; Hickey & Doyle 1977). The increase in the been studied with both light and scanning electron
architectural complexity of angiosperm leaves was also microscopy. Grains ascribed to angiosperms from
shown to be broadly consistent with interpretations of Israel, which are of possible Late Valanginian–Early
leaf evolution based on extant plants (Doyle & Hickey Hauterivian age (Brenner 1996), are more difficult to
1976; Hickey & Doyle 1977, see also Hickey & evaluate because the available illustrations reveal few
Wolfe 1975). structural details.
More recently, the discovery of angiosperm meso- Hughes (1994) distinguished three phases of
fossils containing three-dimensionally preserved angiosperm evolution through the Mid-Hauterivian
angiosperm flowers from the Cretaceous has provided to Early Barremian (phases 0–2) and three phases
a completely new, and previously unimagined, source (phases 3–5) through the Late Barremian to Aptian.
of information about the pattern of early angiosperm Almost all of the angiosperm grains from the six
diversification as well as new insights into the repro- phases recognized by Hughes (1994) are monoapertu-
ductive biology and phylogenetic diversity of rate. Most of the grains have a long straight colpus.
angiospermsduringthefirst70Myroftheirevolution- Morerarelytheapertureisexpandedandmoreorlesscir-
aryhistory(e.g.Friis&Skarby1981;Friis1984;Friis cular.Thepollengrainsareminutetosmallandrangein
et al. 2006a). The more detailed picture that emerges sizefromabout9mmtoabout29mmindiameter.There
is broadly consistent with the pattern of phylogenetic is considerable variation in pollen wall structure and
relationships among extant angiosperms inferred ornamentation. This ranges from continuous tectate,
from studies of molecular sequence data (e.g. The witheitherpunctateorspinulateornamentation,toreti-
Angiosperm Pyhlogeny Group 2003; Soltis et al. culate semitectate, with muri that are smooth or have
2005). Taken together, several lines of evidence from variouskindsofsupratectalornamentation.
the fossil record, combined with inferences from Only two different types of angiosperm pollen are
extant plants, now provide a strong indication that the recorded by Hughes (1994) from phase 0: one has a
first major phylogenetic diversification and ecological continuous tectum and the other has a reticulate
radiationofangiospermstookplaceintheEarlyCretac- (semitectate) pollen wall. Already by phase 1, there
eousoverarelativelyshortintervalofsome20–30Myr is greater diversity among the reticulate monocolpate
(Brenner 1963, 1996; Doyle & Hickey 1976; Hughes grains. Forms that have fine transverse striations on
1976, 1994; Crane et al. 1995; Friis et al. 2006a). the muri also enter the fossil record at this level and
Studiesoffossilflowersandotherangiospermrepro- become very common in later Barremian and Aptian
ductive organs from many different stratigraphic levels palynofloras. Monocolpate and reticulate pollen with
intheCretaceouscontinuetoaddfurtherdetailstothe spiny ornamentation on the muri first appear in
emerging picture of early angiosperm evolution (e.g. phase 2 and are also common through the later
Friisetal. 2006a; Takahashi et al. 2008; Viehofen etal. Barremian and Aptian. Crotonoid pollen and pollen
2008; von Balthazar et al. 2008; Martinez-Millan et al. with trichotomocolpate apertures occur for the first
2009).However,inthispaper,wefocusonfossilassem- time in phase 3.
blagesfromtheEarlyCretaceousthatcontaintheoldest Angiosperm pollen from phases 4 and 5 continues
structurallyandphylogeneticallyinformativeangiosperm to be small and predominantly monoaperturate.
flowers, fruits and seeds. Information from these floras During phase 4, pollen grains with a distinctly
suggests that much angiosperm diversity prior to the graded reticulum are seen in the fossil record for the
mid-Cretaceouswasmainlyamonglineageswithanher- first time. A significant event in angiosperm evolution
baceous or shrubby habit, and that manyofthese early is the first appearance of tricolpate pollen grains at
angiosperms probably grew in wet to fully aquatic around the latest Barremian–Early Aptian. Hughes
environments. Within these assemblages, the presence records a single tricolpate grain in phase 4 and one
of Chloranthaceae, Nymphaeales, Alismatales, Ranun- other tricolpate grain of a different kind in phase 5,
culales,Proteales,Buxalesandothergroupsisnowwell which is dated as Early Aptian (Hughes 1994).
documented, and there is also evidence of many plants Angiospermpollenisalsoknowninmanyotherparts
thatarenotcloselyrelatedtoanyextanttaxa.However, of the world from the Late Barremian–Early Aptian
newresultsreinforcepreviousindicationsthatmonocoty- interval, more or less contemporaneous with phases 4
ledons were also diverse in these ancient angiosperm and5ofHughes,andshowsthesamepatternofdiver-
communities and contributed substantially to the sity as observed for the British sequence. Especially
angiospermcomponentofEarlyCretaceousecosystems. importantaresequencesofpalynoflorasfromthePoto-
mac Group of eastern North America (e.g. Brenner
1963; Doyle 1969; Doyle & Hickey 1976; Doyle &
2. MORPHOLOGICAL DIVERSITY OF EARLY Robbins 1977), from Egypt (e.g. Penny 1986, 1988b)
ANGIOSPERMS: EVIDENCE FROM FOSSIL and from Equatorial Africa (Doyle et al. 1977). The
POLLEN AND FLOWERS evolutionary signal that emerges from these floras is a
(a) Morphological diversity of early pollen considerable expansion of angiosperm pollen diversity,
The earliest unequivocal remains of angiosperms are both in terms of aperture configuration and also
pollen grains from the Early Cretaceous, and some pollen wall sculpture. Scattered records of tricolpate
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
Diversity in obscurity: fossil flowers E. M. Friis et al. 371
pollen grains are also known from contemporaneous There are also a few other Early Cretaceous floras
pollen assemblages from Israel (Brenner 1996), Equa- that contain floral remains of angiosperms from
torial Africa (Doyle 1992), Egypt (Penny 1988b) and other parts of the world. These include the
North America (Doyle 1992). impression/compression flora from the Yixian For-
The first mesofossil floras with angiosperm flowers mation (Late Barremian–Early Aptian) from
and pollen in situ are also from this time interval. Liaoning Province and adjacent areas in northeastern
Mesofossil floras consist of three-dimensionally pre- China (e.g. Sun et al. 2001; Leng et al. 2003) and
served coalified plant remains in the size range from the Late Aptian–Early Albian Crato Formation
between microfossils and macrofossils. The fragments of northeast Brazil (Mohr et al. 2007). Fossils in
are typically a few millimetres long and may include both these assemblages are important in providing
whole flowers with all floral organs preserved (for a whole plant preservation. Angiosperms are rare in
review, see Friis et al. 2006a). Flowers sometimes the Yixian Formation, represented by only few taxa
havestamenswithpolleninsitu.Ultrastructuraldetails and relatively few specimens. Angiosperms are more
of the pollen grains in situ studied with scanning elec- common in the perhaps slightly younger Crato flora.
tron microscopy also indicate that the number of taxa Further, a petrified floral structure was also recently
is higher than would be expected from studies of dis- described from the Late Albian of Australia
persed grains using light microscopy, which has been (Dettmann et al. 2009).
standardinmostpalynostratigraphicstudies.Scanning
electron microscopy of dispersed grains also provides
(i) Size
evidence of cryptic diversity through improved dis-
All early angiosperm flowers known from the Late
crimination of different pollen types based on fine
Barremian–Middle Albian are of small to medium
details of sculpture. Relatively low pollen production
size. Flowers from the many Early Cretaceous meso-
inmanyearlyangiospermsmayalsoimplythatcertain
fossil floras range from about 0.5mm up to about
pollentypesseeninmesofossilsarepoorlyrepresented
5mm, while slightly larger flowers are reported from
in palynofloras that comprise dispersed pollen.
the Crato flora (up to about 6–7mm; Mohr &
Bernardes-de-Oliveira 2004) and from Western
Australia (up to about 16mm; Dettmann et al. 2009).
(b) Morphological diversity of early flowers
Many flowers were borne in dense inflorescences,
and other reproductive organs
mostly spikes, but there are also a few solitary flowers,
There are now a large number of mesofossil floras
both in the mesofossil floras and in the compression/
described from the Late Cretaceous that contain
impression Crato flora. Archaefructus from the Yixian
angiosperm flowers and other reproductive structures.
Formation (Sun et al. 2002) is relatively large, but if
However, the key mesofossil floras from the Early
the inflorescence interpretation of Friis et al. (2003)
Cretaceous are all from Portugal and eastern North
andEndress&Doyle(2009)iscorrectthentheindivid-
America. Some of the fossils in these floras are pre-
ual flowers of Archaefructus are small, simple and also
served as charcoal formed by vegetational fires and
borne in a spike-like inflorescence.
have the three-dimensional structure intact. Others
Early angiosperm fruits and seeds are also typically
arepreservedasligniteformedfromchemicallyaltered
small.InapreliminaryanalysisoftheFamalica˜omeso-
organicmaterialthatmaybemoreorlesscompressed.
fossil flora (Late Aptian), 37 different kinds of
These fossil assemblages were discovered in the late
angiosperm fruits and 64 different kinds of seeds
1980s in soft clays, silts and sands, which range in
were identified (Eriksson et al. 2000b). This flora is
age from the Late Barremian–Early Aptian to
broadly representative of the currently known Early
around the Albian–Cenomanian boundary (for a
Cretaceous mesofossil floras. Fruits range in volume
summary, see Friis et al. 2006a). between 0.12 and 8.34mm3 and seeds vary between
TheTorresVedrasflorafromPortugalispossiblythe 0.02 and 6.86mm3 (Eriksson et al. 2000b). Fruits
oldestofthesemesofossilfloras(LateBarremian–Early
and seeds continue to be small compared with their
Aptian,approximatelycorrespondingtophases4and5
living relatives throughout the Cretaceous (Tiffney
of Hughes). The Torres Vedras assemblage includes
1984; Eriksson et al. 2000a).
three-dimensionally preserved floral structures as well
as fruits and seeds, but also contains diverse isolated
(ii) Sex expression
stamenswithpolleninsituandmanycoprolitesconsist-
AllEarlyCretaceousmesofossilflorasincludebothuni-
ingalmostexclusivelyofpollen.Otherimportantfloras
sexual and bisexual flowers. Unisexual flowers seem
fromPortugalareslightlyyounger.Amongthesearethe
slightly more common in the Early Cretaceous than in
mesofossil floras from Catefica (Late Barremian–
mid-Cretaceous floras. They are certainly much more
Aptian), Famalica˜o (Late Aptian) and Buarcos, Vila
commonthaninLateCretaceousfloraswherebisexual
Verde 2, Vale de Agua and several others (Late
forms predominate, even among flowers that are
Aptian–Early Albian). The most diverse mesofossil
most closely related to extant wind-pollinated taxa
flora from the Early Cretaceous of North America is
with unisexual flowers (Friis et al. 2006b).
from the Puddledock locality (Early–Middle Albian),
but Early Cretaceous mesofossil floras from the Poto-
mac Group sequence range from around the (iii) Position of floral organs
Barremian–Aptian boundary (Drewry’s Bluff and Most of the early angiosperm flowers from the Late
Dutch Gap localities) to the end of the Early Barremian–Early Aptian and through to the Early–
Cretaceous (West Brothers’ locality, latest Albian). Middle Albian are hypogynous, but there are also
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
372 E. M. Friis et al. Diversity in obscurity: fossil flowers
several epigynous forms. This is in contrast to flowers relatively large, the pollen sacs are often very small
from Late Cretaceous floras. Beginning in the compared with the total size of the stamen. Flattened
Turonian,orperhapsevenearlierintheLateCenoma- stamens typically have pollen sacs borne on one sur-
nian, about half of all flowers in typical mesofossil face. Anther dehiscence in the earliest known
floras are epigynous or semiepigynous. The difference angiosperm stamens (Late Barremian–Early Aptian)
mayreflectinitialstrategiesforprotectionoftheovules was probably by longitudinal slits, but none of the
inthefaceofincreasedbioticpollination,aswellasthe available specimens show the mode of dehiscence
high relative diversity of particular groups of angio- very clearly. However, in floras from the Late
sperms (probably early core eudicots) at this stage in Aptian–Early Albian, the variety of anthers is greater
angiosperm evolution. and many have valvate dehiscence (e.g. Friis et al.
2006a).Mostantherswithvalvatedehiscencehavelat-
erally hinged valves over each theca. These open like
(iv) Perianth
shutters to expose the two pollen sacs of each theca.
A surprisingly high proportion of ancient angiosperm
In general, the form of the androecium in
flowers appear to lack a perianth or to have a simple
Turonian–Maastrichtian flowers is markedly different
undifferentiated perianth. None of the flowers from
from that of flowers from the Early Cretaceous. In
theEarlyCretaceoushaveaperianththatisclearlydif-
the Late Cretaceous, dispersed stamens are much
ferentiatedintocalyxandcorolla.Thenumberoffloral
less common and bulky stamen types are rare.
parts is usually also low. Exceptions to this generaliz-
Stamen phyllotaxis is mainly whorled and the
ation are Virginianthus and Carpestella from the
number of androecial whorls is typically one or two,
Early–Middle Albian Puddledock flora, Endressinia
rarely more. Anthers are mostly dithecate with a dis-
from the Late Aptian–Early Albian of Brazil and
tinct band- or thread-shaped filament that is
Lovellea from the Albian of Australia, all of which
separated from the anther by a distinct joint.
have multipartite flowers with numerous floral organs
(Friis et al. 1994; Mohr & Bernardes-de-Oliveira
2004; von Balthazar et al. 2008; Dettmann et al. (vi) Gynoecium
2009).InVirginianthus,andalsoinTeixeiraealusitanica MostEarlyCretaceous angiospermflowersapparently
(von Balthazar et al. 2005), the tepals grade from had monocarpellate flowers, judging from the many
sepaloid on the outside to petaloid on the inside. unilocular, often one-seeded fruits in the early floras.
InmanyEarlyCretaceousfossilflowers,thephyllo- However, multicarpellate flowers were also present
taxis of the floral organs, including the perianth, is early in angiosperm history. In the Torres Vedras
difficult to establish. In some cases this may be flora, both apocarpous fruits with many free carpels
because of distortion during fossilization, but in and syncarpous gynoecia formed from two or three
other cases floral phyllotaxis may be irregular. This is carpels are present. From the contemporaneous
in contrast to Late Cretaceous flowers that predomi- Yixian flora, Archaefructus has a gynoecium of one or
nantly have distinct heterochlamydous perianth of two free carpels (inflorescence interpretation) and
sepals (calyx) and petals (corolla) that are clearly Sinocarpus is syncarpous.
arranged in whorls. None of the early flowers have a distinct style with
an elevated stigma. Instead, the stigmatic surface is
often extended or indistinct. This is in strong contrast
(v) Stamens
to flowers from the Late Cretaceous, which normally
Evidence from mesofossils shows that there is con-
have distinct styles that elevate the stigma. From the
siderable diversity in the organization of the
Turonian through to the Maastrichtian, there is also
androecium among the earliest angiosperm flowers.
a clear dominance of flowers with syncarpous gynoe-
Early Cretaceous floras also include a surprisingly
cia, which are mainly related to groups of core
large number of dispersed stamens with pollen in situ.
eudicots.
In the Torres Vedras mesofossil flora (Late
Barremian–Early Aptian) stamens typically have
simple anthers with only modest development of the (vii) Fruits and seeds
connective between the pollen sacs. The filament is Fruit types among Early Cretaceous angiosperms are
not well developed and the stamen base is short. diverse. There are several apocarpous-follicular fruits
Anthersvaryinshapeandsizefromshortandsagittate that have more than one seed per carpel in the
to narrow and elongate. Sometimes there is an apical Torres Vedras assemblages, but the vast majority of
expansion of the connective, but typically it is small. fruits at this level are indehiscent, unilocular nuts or
On average, anthers in Early Cretaceous floras are drupes with one or very few seeds. For instance, in
relatively large compared with those from Late Creta- the Famalica˜o mesofossil flora (Late Aptian), seed
ceousfloras.ManyoftheLateAptian–Albianstamens number could be determined for 18 fruits. Of these,
are also bulky with extensive development of the con- 12 had only a single seed, while the remaining six
nective, both between the pollen sacs and at the hadtwo,threeormoreseedsperfruit.Theproportion
stamen apex. This apical extension may be strongly offruitswithafleshyouterlayerissurprisinglyhigh.In
developed and in some cases it may be as long as the the Famalica˜o flora, about 25 per cent of the fruits
pollen sacs. The lower part of the stamen may be were either drupes or few-seeded berries. This prob-
short, or longer and developed into a filament, but it ably indicates the early establishment of animal
is often flattened and always grades into the anther dispersal among angiosperms (Eriksson et al. 2000b).
without a distinct joint. Even though anthers are Because of their small size, the fleshy portion of each
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
Diversity in obscurity: fossil flowers E. M. Friis et al. 373
fruit was small, but there are several indications that the Turonian of New Jersey, USA (Gandolfo et al.
the fruits were borne in dense infructescences. 2004), has been questioned (Endress 2008).
Collectively, the amount of fleshy tissue may have Despite their sparse Late Cretaceous record, and
been sufficient for attracting animal dispersers. eventhoughtheextantgeneramaynothavediversified
Early Cretaceous angiosperms typically have ana- until the Cenozoic, Nymphaeales were clearly present
tropous ovules, but rare orthotropous ovules are also at an early stage in angiosperm evolution. A key early
known from the earliest mesofossil floras including record of Nymphaeales is the flower of Monetianthus
the Torres Vedras flora. Campylotropous ovules are mirus, from the Late Aptian–Early Albian flora of
known from the Late Aptian flora of Famalica˜o. Vale de Agua (Friis et al. 2001, 2009b). Monetianthus
Most early angiosperm seeds are apparently formed provides unequivocal evidence of crown group
from bitegmic ovules. Most are exotegmic with a Nymphaeales, even though it cannot be
hard outer seed coat (Friis et al. 1999). accommodated in any extant genus.
The Monetianthus flower is perigynous, about
3mm long and 2mm in diameter, with remains of
3. PHYLOGENETIC DIVERSITY OF perianth, androecium and gynoecium. The gynoe-
EARLY ANGIOSPERMS cium is particularly informative, with a whorl of 12
All angiosperm pollen grains from the earliest phases carpels that are fused for most of their length and
of the angiosperm diversification described by free only in the apical part. There is a small central
Hughes (1994) are monocolpate. There are also projection of the floral axis between the carpels.
some records of inaperturate pollen from the Valangi- Each carpel contains many small ovules, which do
nian–Hauterivian of Israel (Brenner 1996), but these not fill out the locules completely. Placentation is
need further documentation. Among extant laminar. The combined characters of the gynoecium
angiosperms, monoaperturate pollen, in which the are unique for Nymphaeaceae, and phylogenetic
aperture is generally colpate, is exclusively produced analysis suggests a position of Monetianthus at the
byangiospermsattheANITAgrade(theearliestdiver- root of the BarclayaþNymphaeoideae clade (Friis
ging lineages of extant angiosperms including et al. 2009b).
Amborella, Nymphaeaceae, Illicium, Trimeniaceae Pluricarpellatia peltata from the Early Cretaceous
and Austrobaileyaceae sensu Qiu (1999)) as well as by (LateAptian–EarlyAlbian)ofBrazilalsoprovidesevi-
Chloranthaceae, monocots and eumagnoliids. Within dence of Early Cretaceous Nymphaeales. It is a
these groups, some families and genera have very dis- fruiting structure preserved in organic connection
tinctive pollen features and can be identified based on with peltate nymphaealean leaves (Mohr et al. 2007).
isolatedpollenalone.However,despitetheveryexten- Other Early Cretaceous leaf fossils with distinct nym-
sive record of dispersed angiosperm pollen from the phaealean features also support the presence and
Early Cretaceous, very few of these grains can be diversity of the group early in angiosperm history.
assigned confidently to extant lineages at the order, Among these are Brasenia-like leaves from the Early
family or generic level. Recent discoveries of fossil Cretaceous (Late Aptian or Early Albian) flora of
flowers and other angiosperm reproductive organs Buarcos-para-Tavarede, Portugal, assigned to the fos-
allow much more detailed evaluation of the phyloge- sil genus Braseniopsis (Saporta 1894), and Scutifolium
netic relationships of early angiosperms to extant jordanicum from the Early Cretaceous (Albian) of
plants. These fossils show that although angiosperms Jordan, which was assigned to the Cabombaceae
are already diverse by the Barremian–Aptian, this (Taylor et al. 2008). There are also many seeds with
diversity is restricted to early diverging lineages of distinct nymphaealean features among the Early
extant angiosperms, as well as other lineages that Cretaceous mesofossils from Portugal and eastern
appear to be extinct. North America. Their first occurrences are in the
Late Barremian–Early Aptian Torres Vedras flora, as
(a) ANITA lineages wellasinthemoreorlesscontemporaneousmesofossil
(i) Nymphaeales flora from Drewry’s Bluff, Virginia (Friis et al. 1999,
Fossil remains of Nymphaeales are very well rep- 2006a).
resented in Palaeogene and Neogene floras. Carpestella lacunata from the Early–Middle Albian
Nymphaealean seeds are particularly common and Puddledock flora of Virginia, USA (von Balthazar
easily recognized by their distinct bitegmic, exotestal et al. 2008) is another Early Cretaceous flower prob-
organization, the presence of a micropylar lid and ably related to ANITA grade angiosperms. The fossil
the often strongly undulating anticlinal walls of the is a small fragmentary gynoecium, about 0.65mm
cells on the surface of the exotesta. The Cretaceous long and 0.45mm in diameter, that shows many
fossil record of Nymphaeales is much more meagre. helically arranged scars from detached floral organs.
Dispersed Nymphaea-like pollen, described as The gynoecium is syncarpous, consisting of 13
Zonosulcites scollardensis and Zonosulcites parvus, is carpels arranged radially around a central column in
known from the Maastrichtian of Canada (Srivastava a star-shaped pattern, as is seen in extant Nymphaea-
1969; Muller 1981) and unequivocal nymphaealean ceae and Illicium (Schisandraceae). The presence of
seeds, described as Symphaenale futabensis, are known septal slits suggests a link between the fossil and
from the Late Cretaceous (Early Santonian) Gokura- extant Nymphaeaceae, but the precise position of
kuzawa locality of northeastern Honshu, Japan Carpestella cannot be established securely with the
(Takahashi et al. 2007). The nymphaealean affinity information currently available (von Balthazar et al.
suggested for flowers of Microvictoria, described from 2008).
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
374 E. M. Friis et al. Diversity in obscurity: fossil flowers
There are other fossils of putative relationship with particularly similar to extant Eupomatiaceae and
ANITA grade angiosperms from Early Cretaceous Himantandraceae, but cannot be placed in any
mesofossil floras (Friis et al. 1999), but none can be extant family and may represent an extinct lineage.
attributedtoanextantfamily.Theyprobablyrepresent Archaeanthus linnenbergeri from latest Albian–earliest
extinct lineages that diverged close to the base of the Cenomanian Dakota Formation is very similar in
angiosperm phylogenetic tree. floral structure to extant Magnoliaceae, but the fruits
differ from all extant taxa in producing many small
seeds (Dilcher & Crane 1984). Archaeanthus possibly
(b) Chloranthaceae
represents stem-group Magnoliaceae. Flowers of
Asteropollis is a genus of distinctive early angiosperm
crown group Annonaceae are known from the Early
pollen grains that is closely comparable with pollen
Coniacian of Japan (Takahashi et al. 2008).
of the extant genus Hedyosmum (Chloranthaceae)
(Walker & Walker 1984). Asteropollis includes species
characterized by their star-shaped distal aperture, (ii) Laurales
reticulate pollen wall and muri that have beaded Laurales have an extensive Cenozoic fossil record and
ornamentation. In addition, some species of are becoming increasingly well documented from the
Clavatipollenites, which have a monocolpate or Cretaceous based on numerous remains of flowers
trichotomocolpate aperture and reticulate pollen wall and inflorescences (e.g. Drinnan et al. 1990; Eklund
with beaded muri, are very similar to grains of extant 2000; Takahashi et al. 2001). The first of these fossils
Ascarina (Chloranthaceae) (Couper 1960; Walker & to be described was Mauldinia mirabilis from the
Walker 1984). Mauldin Mountain flora of Maryland (Early
Studiesofmesofossilsconfirmindicationsfromdis- Cenomanian) (Drinnan et al. 1990). Mauldinia and
persed pollen that Chloranthaceae entered the similargenerahavesincebeenrecognizedfromseveral
angiosperm fossil record very early. The family was fossil floras in Europe and Central Asia (Eklund &
probably already diverse in the earliest mesofossil Kvacˇek 1998; Frumin et al. 2004; Viehofen et al.
floras (Friis et al. 2006a). In the Torres Vedras meso- 2008), and Mauldinia currently includes four distinct
fossil flora (Late Barremian–Early Aptian) species.
inflorescences,fruits,dispersedstamensandcoprolites In the Early Cretaceous, lauralean flowers are
suggest the presence of several different species of known only from the Puddledock flora
Chloranthaceae. Hedyosmum-like staminate inflores- (Early–Middle Albian) of Virginia, USA. Laurales
cences often associated with Hedyosmum-like fruits known from this flora are the calycanthaceous flower
are also known from several of the Aptian–Early Virginianthus (Friis et al. 1994), the lauraceous flower
Albian floras from Portugal (Friis et al. 2006a). Potomacanthus lobatus (von Balthazar et al. 2007) and
Hedyosmum-like fruits with adhering Asteropollis several other undescribed forms. The fossil flower
pollen have also been recorded from the Early– Lovellea wintonensis, described by Dettmann et al.
Middle Albian Puddledock flora (Friis et al. 1997). (2009)fromthemid-CretaceousofWesternAustralia,
Unequivocal androecia of Chloranthus-like plants are is another Early Cretaceous lauralean element. Its
known from Late Cretaceous mesofossil floras closest similarities are to extant Gomortegaceae and
(Crane et al. 1989; Eklund et al. 1997). Monimiaceae,but its precise position among Laurales
remains to be established (Dettmann et al. 2009).
(c) Eumagnoliids
Eumagnoliids are well known in the Cenozoic fossil (iii) Piperales
record where they are represented extensively by Severaldispersed pollen grains from the Early Cretac-
wood,leaves, fruits and seeds. There isalsoa growing eous, such as species assigned to the pollen genera
recordof well-preserved eumagnoliid flowers from the Tucanopollis and Transitoripollis, are monoaperturate
Cretaceous(e.g.Friisetal.1997).Cenozoicfossilsare with a continuous tectum. In these features they are
generally easily assigned to extant genera, but all very similar to pollen grains of extant Piperales. This
Cretaceous eumagnoliids appear to represent extinct kind of pollen has been found on fruits (Friis et al.
genera. The Early Cretaceous record consists mainly 1995) from the Puddledock flora (Early–Middle
of taxa with smaller flowers, and lauralean taxa are Albian). The same kind of pollen has also been found
especially prominent (e.g. Virginianthus, Potoma- on fruits as well as in stamens and in coprolites from
canthus, see below). Larger, multipartite flowers the Torres Vedras locality (Late Barremian–Early
similar to those of extant Magnoliaceae, which were Aptian) (Friis et al. 2006a). The combined fruit, seed
previously believed to be archaic in angiosperms, first and pollen characters suggest a relationship to extant
enter the fossil record in the Late Albian at the very Piperales,butotherfeaturesoftheplantsareunknown
end of the Early Cretaceous (Dilcher & Crane 1984). andtheputativerelationshiptoPiperalesremainstobe
establishedmoresecurely.
(i) Magnoliales
Dispersed permanent tetrads of pollen described as (d) Eudicots
Walkeripollis (Doyle et al. 1990) probably represent The earliest record of triaperturate pollen is a single
stem-group Winteraceae. The earliest flower of prob- grain from Zone 4 of Hughes (Late Barremian–Early
able Magnoliales is Endressinia brasiliana from the Aptian; Hughes & McDougall 1990; Hughes 1994).
Crato Formation (Late Aptian–Early Albian) of AbovethislevelintheWealdensuccessionofsouthern
Brazil (Mohr & Bernardes-de-Oliveira 2004). It is England,triaperturatepollenbecomesmorecommon.
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
Diversity in obscurity: fossil flowers E. M. Friis et al. 375
In palynofloras from other areas tricolpate pollen also flowers, fruits, seeds, pollen and phytoliths. There
enters the fossil record around the Barremian–Aptian arealsoreliablemonocot fossilsfromtheLateCretac-
boundary. Tricolpate grains remain rare until later in eous, including stems, fruits, seeds and pollen (e.g.
the Aptian (e.g. Doyle et al. 1977; Penny 1991) and Herendeen & Crane 1995). However, the record of
first dominate palynofloras in the Late Albian where monocotyledons is much less extensive in the Creta-
they include tricolpate as well as tricolporoidate forms ceous than in the Cenozoic. It is also interesting that
(Doyle & Hickey 1976). Tricolpate pollen grains are some of the key monocotyledons of present day wet-
also known from flowers and dispersed stamens in land communities, which are well represented in the
many Early Cretaceous mesofossil floras. In the Late Cenozoic, are absent in the Cretaceous. For example,
Barremian–EarlyAptianTorresVedrasflora, tricolpate Cyperaceae have fruits that fossilize well, are easily
pollenhasnotsofarbeenfoundinflowersoronthesur- recognizable and that are extremely common in
faceofcarpelsorfruits.Norhasitbeenobservedinthe Palaeogene and Neogene fossil assemblages, but they
dispersedpalynoflorafromthesamesamples.However, have not yet been documented from the Cretaceous.
two different kinds of tricolpate pollen were found in a Untilnow,theEarlyCretaceousrecordofmonocots
coproliteandinastamen.Similarly,intheLateAptian has been especially meagre and, according to
flora of Famalica˜o, no tricolpate pollen grains have Gandolfo et al. (2000), all Early Cretaceous records
been observed in situ. In the younger Early–Middle of monocotyledons are ambiguous.
Albian Puddledock flora, a rough estimate of distri- Recognizing the pollen of monocots among the
bution of angiosperm pollen types in situ shows that monocolpategrainsofearlyangiospermshasbeenapar-
aboutone-thirdaretriaperturate. ticular challenge because the morphological and
AllEarlyCretaceousfloralstructureswithtricolpate structural features that potentially distinguish monocot
polleninsituthathavebeenplacedsystematicallyhave pollen from that of other angiosperms are relatively
relationships to groups of eudicots that diverged very subtle. Walker & Walker(1984)listed sevenfeaturesof
earlyfromthelineagethatgaverisetothebulkofeudi- potential importance for assigning dispersed monocol-
cot diversity. Currently, Early Cretaceous floral pate grains to monocots. These include: (i) reticulate
structures related to Ranunculales, Buxales and Pro- sculpturing differentiated into coarse and fine areas
teales have been identified. Probable Ranunculales (graded reticulum), (ii) smooth muri, (iii) lumina of
arerepresentedbyasingle staminateflower,Teixeiraea different sizes, (iv) ‘frilled’ muri caused by laterally
lusitanicum,fromtheValedeAguaflora(LateAptian– extended columellae, (v) regular and clearly polygonal
Early Albian) (von Balthazar et al. 2005). Also from lumina, (vi) a thin pollen wall combined with a very
the Vale de Agua flora are four different kinds of thinnon-aperturalnexineand(vii)absenceofendexine.
pistillate structures (Aguacarpus hirsutus, Lusicarpus Whiletherearemanyexceptionstothese generaliz-
planatus, Silucarpus camptostylus, Valecarpus petiolatus) ations, among extant angiosperms monoaperturate
and one kind of staminate flower (Lusistemon striatus) pollen with a graded reticulum is known only for
that are related to the Buxales (Pedersen et al. 2007). monocots.Thereticulummayhaveluminadecreasing
Unequivocal buxalean flowers and inflorescence frag- insizeeithertowardsthepolarareas(e.g.somepalms)
ments are also known from the Potomac Group. or towards restricted areas of the equatorial region
Spanomera marylandensis is from the Late Albian (e.g. some Liliales and Asparagales), but there are also
West Brothers locality and Spanomera mauldinensis is other variations. Based on these features, Walker &
from the Early Cenomanian Mauldin Mountain Walker (1984) suggested that many of the dispersed
locality(Drinnanetal.1991),bothinMaryland,USA. pollen grains from the Early Cretaceous that are
Proteales are known from several Potomac Group assigned to extinct pollen genera,such as Retimonocol-
mesofossil floras. Especially common are a variety of pites and Liliacidites, were produced by monocots.
flowers and inflorescences of platanoid affinity. These Support for a monocot affinity of some Liliacidites
include pistillate flowers and inflorescences of grains was also inferred by Doyle et al. (2008) from a
Friisicarpus brookensis and the associated staminate recent phylogenetic analysis.
material of Aquia brookensis from the Early– Another character that, as far as we are aware, is
Middle Albian Bank near Brooke locality in Virginia restricted to monocot pollen is acolumellate-reticulate
(Crane et al. 1993), as well as pistillate Friisicarpus wall structure, in which the reticulum is only loosely
marylandensis and associated staminate Platananthus attached to the foot layer and where the infratectal
potomacensis from the Late Albian West Brothers layer is reduced to an ultrathin granular layer. This
locality (Friis et al. 1988). kind of wall structure is particularly common among
ThepresenceofProtealesintheEarlyCretaceousis extant Alismatales (Grayum 1990). It may reflect an
further supported by the occurrence of Nelumbo-like evolutionary transition to the predominantly atectate
leaves described as Nelumbites from the Bank near condition seen inextantaroidandother monocots.In
Brooke (Early–Middle Albian) and the Quantico thefossilrecord,transitionalstatesbetweencolumellate
localities (Late Albian) (Doyle & Hickey 1976; and acolumellate grains reported for Pennistemon-type
Upchurch et al. 1994), both from Virginia, USA. pollen (Penny 1988a) have also been observed for
variousRetimonocolpites-typepollengrains.
(e) Monocotyledons
Fossil remains of monocotyledons are common in (i) Alismatales–Araceae
Cenozoic floras where they are represented by many Mayoa portugallica is a small fragment of lignitized
different organs, including stems, leaves, rhizomes, plant tissue with masses of striate (polyplicate) pollen
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
376 E. M. Friis et al. Diversity in obscurity: fossil flowers
adheringtoitsirregularsurface(Friisetal.2004).The assigned to Retimonocolpites or occasionally to
pollen grains are inaperturate and elliptical in outline. Liliacidites. Both of these genera include a variety of
They have a finely striate tectum composed of closely differentdispersedgrainsthatvaryincolpus configur-
spaced, straight ribs that cross each other in a charac- ationaswellasinthepatternofthereticulumandthe
teristic way. The tectum is supported by a very thin ornamentation of the muri. We infer that these other
granular or weakly columellate infratectal layer and forms of Retimonocolpites and Liliacidites are probably
theendexineisextremelythinandgranular.Thesedis- also monocotyledons as suggested previously by
tinctive pollen grains are closely similar to those of Walker & Walker (1984), although they most likely
extant Holochlamys in every respect. Although other belong to different taxa at generic or higher levels.
information on the Mayoa plant has not yet been dis- All characters of the new fossil strongly support a
covered, the unique pollen morphology combined close relationship with the true aroids, subfamily
with the unusual wall ultrastructure strongly suggests Aroideae, which are characterized by their unisexual,
a close relationship of the fossil to the subfamily perianth-lessflowers(Cabreraetal.2008).Theythere-
Monsteroideae, and this supports the presence of fore support earlier inferences based on Mayoa and
crowngroupAraceaeintheEarlyCretaceous.Flowers provide further evidence of crown group Araceae in
of extant Monsteroideae are bisexual and have a the Early Cretaceous. Most extant Aroideae have in-
perianth. aperturate and atectate pollen. However, Calla,
IncontrasttoMayoa,severalnewfossilsfromanew which under some analyses is the sister genus to all
Early Cretaceous locality in Portugal (Vila Verde 2, other truearoids(Cabreraetal.2008),hasaperturate,
northeast of Figueira da Foz, Figueira da Foz For- tectate pollen. Tectate–semitectate pollen, sometimes
mation, Late Aptian–Early Albian) are known from with extended colpi, is also a characteristic of the clo-
much more complete material. These fossils, which sely related Lasioideaeae and Zamioculcadoideae and
comprise inflorescence axes (spadices) and flowers is most likely plesiomorphic within the family.
with pollen in situ, can also be assigned to Araceae. Reduction of columellae as seen in the new fossil
One of these araceous fossils, here informally may be a step towards loss of tectum characterizing
referred to as ‘Araceae fossil sp. A’ (figures 1a–f and most aroids.
2a,b) is known from many small inflorescence frag- Another araceous fossil from the Vila Verde 2
ments bearing simple flowers that are densely packed locality, here informally referred to as ‘Araceae fossil
inalowspiral.Mostofthespecimensarecharcoalified sp. B’ (figure 1g–i), consists of a single inflorescence
and have retained their original three-dimensional with numerous, densely spaced flowers that are
shape. The flowers are unisexual (staminate) and borne in a spiral arrangement. The fossil is lignitic
lack a perianth (in aroid terminology: aperigonate). and slightly compressed, which impedes a full under-
Eachconsistsofone or twofreestamens.The anthers standing of the floral organization. Nevertheless, it is
are tetrasporangiate, basifixed and almost sessile, with clear that the flowers have a perianth (in aroid
a short, inconspicuous filament. Anthers are also terminology, perigonate). Scars left by the flowers,
bulky, broad and almost square apically. They appear and the organization of the surface of the inflores-
to have been fleshy with a very extensively thickened cence, show that there were four tepals. It is unclear
connective between the four small and widely spaced whether they were in opposite pairs or in a whorl.
pollen sacs. The thecae are placed opposite each The flowers appear bisexual, with a massive central
other and the pollen sacs placed along the four edges gynoecium surrounded by free stamens. The stamens
of the anther. The surface of the anthers usually has consist of a long anther and a short, indistinct fila-
distinctive bulges, indicating the presence of densely ment. The apical, exposed surface of the gynoecium
spaced secretory cells. In many specimens, the apical is rhomboidal.
part of the connective shows the remains of an abun- Pollen grains preserved in situ are periporate and
dant secretion, and we infer that the anther semitectate–reticulate. The reticulum is coarse with
connective was odour producing (osmophoric). homogeneous lumina. Muri are smooth with a
Pollen grains found in the stamens are clustered in sharp triangular profile. The tectum is supported
elongated strands. Pollen is monocolpate with a long by scattered, long columellae. The pores are scat-
extended colpus and a semitectate-reticulate pollen tered over the surface of the pollen and the
wall. The reticulum is coarse with dimorphic lumina. aperture membranes are covered with densely
Muri are psilate towards the outside and almost gran- spaced granules. We are not aware that similar
ular towards the inside. The reticulum is supported grains have been described from dispersed Early
only by a few columellae. The columellae are uneven Cretaceouspalynofloras.However,fromthe Famalica˜o
in length and shape. They detach easily from the locality (Late Aptian), Portugal, similar peri-
main body of the grain, leaving scattered scars on the porate grains have been observed in situ in a
outer surface of the foot layer. tetrasporangiate anther (Friis et al. 1999). This pollen
An isolated stamen from the Catefica locality (Late differs in having more densely spaced columellae.
Barremian–Aptian)withverybulkyanthersandacon- The characters of this fossil indicate a phyloge-
nectivethatwasprobablyalsoosmophoricmaybelong netic relationship to subfamily Pothoideae. It thus
tothesametaxonastheVillaVerdearoid,indicatinga provides further evidence of crown group Araceae
wider occurrence of this kind of plant. Dispersed in the Early Cretaceous. Extant Pothoideae is the
pollen grains very similar to the pollen grains in situ only extant group of Araceae that has perigonate,
within the stamens from Villa Verde 2 are common bisexual flowers with periporate semitectate–reticulate
in Early Cretaceous palynofloras. They are typically pollen.
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
Diversity in obscurity: fossil flowers E. M. Friis et al. 377
(a) (b) (c)
(d ) (e) ( f )
(g) (h)
(i)
Figure1.ScanningelectronmicrographsoffossilinflorescencesofAraceaefromtheEarlyCretaceous(LateAptian–EarlyAlbian
oftheVilaVerde2locality(sampleVilaVerde2439),Portugal.(a–f)Araceaefossilsp.A,andinsitupollen.(a)Inflorescence
fragmentwithdenselyspacedandspirallyarrangedstaminateflowersonthecentralaxis;S165015.(b)Inflorescencefragment
withmanyspirallyarrangedstaminateflowers;S165005.(c)Singleinflorescencesectionwith10stamens;S165002.(d)Inflor-
escencefragmentwithseveralflowers,centralaxismissing;S165001.(e,f)Pollengrainsfromstaminateflowerininflorescence
fragmentshowninfigure1c;S165005.(g–i)Araceaefossilsp.B,withdenselyspacedandspirallyarrangedflowersandperipo-
ratepollen;S165007.(g)Internalviewofinflorescenceshowingaxisandattachedflowers.(h)Externalviewofinflorescence.(i)
Periporatepollengrainfromanther;arrowsshowthreeofseveralpores.Scalebars:a¼2mm;c¼250mm;b,d,g,h¼1mm;e,
f¼10mm;i¼20mm.AllspecimenshousedinthepalaeobotanicalcollectionsattheSwedishMuseumofNaturalHistory(S).
A further Early Cretaceous fossil that may also buttheflowersappearverysimilar totheAraceaefossil
belong to Pothoideae is a small inflorescence fragment sp. B. The in situ pollen are similarly periporate and
from the Vale de Agua locality (Late Aptian–Early semitectate–reticulate,but the reticulum ismuch more
Albian) of Portugal. The fossil is strongly compressed densewithverysmalllumina,andmurihaveasupratectal
andtheorganizationoftheflowersdifficulttoestablish, ornamentation of minute spinules.
Phil.Trans.R.Soc.B(2010)
Downloaded from http://rstb.royalsocietypublishing.org/ on November 20, 2018
378 E. M. Friis et al. Diversity in obscurity: fossil flowers
(a) (c) (e)
(b) (d) (f)

Figure2.ScanningelectronmicrographsofinsitupollenfromEarlyCretaceousmesofossilflorasfromPortugal.(a,b)Pollen
fromAraceaefossil sp.Ashownin figure1b;S165005(sampleVilaVerde 2439).(a)Pollen showing long opencolpus and
coarse,loosereticulumsupportedbyscatteredcolumellae.(b)Detailofwallshowingmuriwithoutsupratectalornamentation,
butwithagranularstructureontheinternalsurfaceandonthecolumellae.(c,d)Pennipollis-typepollenfromtheLateAptian–
EarlyAlbianValedeAgualocality(sampleValedeAgua139);S105603.(c)Pollenshowinglongcolpusandverycoarse,loose
reticulum with spiny supratectal ornamentation. (d) Detail of reticulum showing spiny supratectal ornamentation externally
(above) and granular internal surface of muri (below). (e,f) ‘Retimoncolpites’-type pollen from Torres Vedras (sample Torres
Vedras 44). (e) Pollen showing long open colpus and coarse, loose reticulum with transverse supratectal striations. (f)
Detail of reticulum from two different grains. The upper reticulum is loosed from the main body of the pollen grain and
shows the granular inner surface of the muri; the reticulum of the grain below shows the striate supratectal ornamentation
of the outside of the muri. Scale bars: a¼10mm; b¼1mm; c, e¼12mm; d¼3mm; f¼1.2mm. All specimens housed in
the palaeobotanicalcollections atthe Swedish Museum ofNaturalHistory (S).
Therearealsoseveralfruitingstructuresintheearly pollen grains from the new Araceae fossil sp. A
mesofossil floras from Portugal that may be linked to (figure 2a,b), which supports the possible alismatalean
Araceae. Some have many fruits borne in a dense affinity for these fossils. This feature also occurs in
spiral along a central inflorescence axis, sometimes many other Retimonocolpites-type pollen grains found in
with rhombic apical faces, others are isolated fruits situindispersedstamensthatrangeinagefromLateBar-
with morphological similarities to extant fruits of remiantoMiddleAlbian(figure2e,f).
AraceaeandhaveRetimonocolpites-typepollenattached. DispersedpollengrainsofPennipollisarediverseand
Alltheseneedtobestudiedinfurtherdetailtoconfirm abundantinEarlyCretaceouspalynologicalassemblages
their systematic affinity. fromsouthernLaurasiaandnorthernGondwana,indi-
cating that the Pennistemon/Pennicarpus/Pennipollis
plants were widespread and probably common in the
(ii) Other Alismatales Barremian–Albianvegetation(Friisetal.2000).
AgeneralrelationshiptoAlismataleshasbeensuggested Several other floral structures that are probably
for the Pennistemon/Pennicarpus/Pennipollis plant (Friis related to extant Alismatales are known from the
etal.2000).Incontrast,anaffinitywithChloranthaceae Early Cretaceous mesofossil floras of Portugal. One
has been suggested by Hesse & Zetter (2007), Doyle is a fragmentary bisexual flower with remains of sta-
et al. (2008) and Doyle (2009). Acolumellate pollen mens around an apocarpous gynoecium of three or
with an extremely loose reticulum is widespread in four free follicular carpels. Pollen grains found in situ
Alismatales and other basal monocotyledons (Grayum are periporate and tectate with echinate tectum orna-
1992),butisnotknownamongextantChloranthaceae. mentation and clusters of small spines covering the
The finely granular inner lining of the muri seen in apertures. These pollen grains are closely similar to
pollen of the Pennistemon/Pennicarpus/Pennipollis plant grains of Alismataceae, and a systematic position
(figure 2c,d) is very similar to that seen in the in situ close to Alismataceae is supported also by the general
Phil.Trans.R.Soc.B(2010)