Remnants of Antarctic vegetation on King George Island during the early MioceneMelville Glaciation

Sophie Warnya*, C. Madison Kymesa, Rosemary A. Askinb, Krzysztof P. Krajewskic and Philip J. Barta

aDepartment of Geology and Geophysics, Louisiana State University, E-235 Howe-Russell, Baton Rouge, Louisiana 70803, USA;bConsultant, 1930 Bunkhouse Drive, Jackson, Wyoming 83001, USA; cInstitute of Geological Sciences, Polish Academy of Sciences,

Twarda 51/55, 00-818 Warszawa, Poland

Palynological analyses of 12 samples from the Cape Melville Formation, which crops out on easternmost KingGeorge Island, Antarctica, provide new information on the type of vegetation that covered the South ShetlandIslands during the early Miocene Melville Glaciation, c. 23�21 Ma. The assemblage recovered was mostlycharacterised by in situ algae such as leiospheres along with acanthomorph acritarchs, both glacial indicators. Thesparse in situ terrestrial palynomorph assemblage included tundra-indicative moss spores Coptospora sp., rarepodocarp conifer and various angiosperm pollen. The latter includes pollen of several species of Nothofagidites, plusrare Asteraceae, Caryophyllaceae (Colobanthus-type) and Chenopodipollis. The majority of the palynomorphsrecovered are interpreted as reworked, denoting glacial scouring and redeposition from various sites in the AntarcticPeninsula and the South Shetland Islands. These reworked palynomorphs are of Permian to Paleogene age. Thisreworked component provides insight into the potential sources of reworking, and is consistent with multiple cyclesof glacial advances to the Melville Peninsula at the time of deposition. The penecontemporaneous palynomorphsrecovered provide new data on the climatic regime and glacial intensification during the early Miocene on KingGeorge Island.

Keywords: vegetational history; King George Island; Antarctica; Miocene ice-house climate; Melville Glaciation

1. Introduction

King George Island (KGI) is the largest island of the

South Shetland Islands (SSI) magmatic arc that

extends parallel to the northern Antarctic Peninsula

and is detached from it by a young back-arc rift struc-

ture of the Bransfield Strait (Figure 1A and B). The arc

developed as a result of subduction of the Pacific litho-

sphere (Phoenix Plate) beneath the continental crust ofthe Antarctic Peninsula (Barker 1982; Guterch et al.

1985). It is largely composed of Mesozoic and Ceno-

zoic eruptives and associated pyroclastics that are cut

by hypabyssal and abyssal intrusions (Smellie et al.

1984; Haase et al. 2012). A trend from older to younger

magmatic activity is generally observed from southwest

to northeast along the arc (Anderson 1999). This has

led to a concentration of Cenozoic volcanic rocks onKGI (Pankhurst & Smellie 1983; Machado et al. 2005),

with maximum eruptive activity during the Eocene

(Wang et al. 2009; Nawrocki et al. 2011). The Eocene

volcanism has, in part, caused the poor preservation of

Permian to Cretaceous polynomorphs, often found

burnt beyond recognition in many KGI and surround-

ing sequences.

During the Eocene epoch, the island most likely had

a mountainous landscape with stratovolcanoes and lava

fields that were variably covered by Valdivian-type for-ests (a forest found today in Chile and in parts of Argen-

tina), wetlands and freshwater environments (Poole

et al. 2001; Hunt & Poole 2003). A decrease of volcanic

activity close to the Eocene/Oligocene boundary coin-

cided with northward progradation of the Antarctic ice-

sheet and marine transgressions (Baker 2007; Davies

et al. 2012). Parts of KGI were covered by ice and/or

inundated during the Oligocene and early Miocene, giv-ing way to glacial, glaciomarine and marine sedimenta-

tion (Troedson & Riding 2002; Troedson & Smellie

2002). These deposits are known from coastal exposures

and nunataks in the southeastern part of KGI

(Figure 1C). The sedimentary succession embraces the

Oligocene Polonez Cove Formation (Chopin Ridge

Group) occurring between Admiralty Bay and King

George Bay (Figure 1C), and the Late OligoceneDestruction Bay Formation and the earliest Miocene

Cape Melville Formation (Moby Dick Group) occurring

between Sherratt Bay and Destruction Bay (Figures 1C

and 2; Birkenmajer 2001 and references therein).

*Corresponding author. Email: swarny@lsu.edu

� 2015 AASP � The Palynological Society

Palynology, 2015

http://dx.doi.org/10.1080/01916122.2014.999954

The sedimentary succession of the Moby Dick

Group rests on the andesitic�basaltic substratum of

the Sherratt Bay Formation (Figure 2). The Destruc-

tion Bay and Cape Melville formations are of Chattian

(c. 26�24 Ma) and Aquitanian (c. 23�21 Ma) ages,

respectively (Birkenmajer & ºuczkowska 1987; Birken-

majer et al. 1988; Bittner & Crame 2002; Dingle & Lav-

elle 1998; Troedson & Riding 2002; Whittle et al. 2014and references therein). The succession is cut by a

swarm of Miocene basaltic dykes dated c. 20 Ma (Bir-

kenmajer et al. 1985) and by the Quaternary Peak

Melville volcano (Birkenmajer & Keller 1990). The

Destruction Bay Formation provides the only known

land exposure and record of the Late Oligocene warm-

ing in West Antarctica.

The Cape Melville Formation (c. 150 m thick, top

erosional) rests disconformably on the Destruction

Bay Formation (Figure 3). It is composed of a diamic-

tite-dominated interval in its lowermost part, overlainby a succession of fossiliferous clastic sediments con-

taining common dropstones of local and regional prov-

enance (Troedson & Riding 2002). The sediments

Figure 1. A. Sketch map of Antarctica showing the location of the South Shetland Islands (SSI) at the northern tip of the Ant-arctic Peninsula. The red rectangle indicates the area enlarged in B. B. Geological sketch map of the northern Antarctic Peninsulaand the SSI. Bransfield Strait is a young (<4 Ma) back-arc rift structure that did not exist during the sedimentation of the CapeMelville (CM) Formation (early Miocene). At that time, the SSI were a part of the northern Antarctic Peninsula magmatic arc.C. Map of King George Island showing the location of Melville Peninsula (red rectangle) between Sherratt Bay and DestructionBay (see Figure 2).

2 S. Warny et al.

range from sandstones to mudstones. Characteristic of

the succession is the alternating occurrence of grey-

and brown-weathering beds and intervals, which con-

trast with generally dark basaltic sandstones of the

Destruction Bay Formation. The colour difference

reflects the changing contribution of volcanogenicmaterial (brown-weathering) and detrital quartz and

clay minerals (grey-weathering) in the marine sedi-

ments. To facilitate local correlations and reconstruc-

tion of the depositional history, the Cape Melville

Formation has been subdivided into 11 informal lithos-

tratigraphical units (in ascending order, units 1 to 11).

The diamictite-dominated lower part of the formation,

discriminated here as the Crab Creek Diamictite,embraces units 1 and 2, and the basal part of Unit 3

(Figure 3B). This diamictite records an advancement

of the Antarctic ice-sheet to KGI that could correlate

with the Mi-1 glaciation known from the oceanic

record. It is followed by renewed glacio-marine sedi-

mentation over the submerging magmatic arc.

The primary goal of this study was to determine

whether vegetation survived during the Melville Glaci-ation between c. 23 and 21 Ma ago. This investigation

was undertaken via a palynological analysis to charac-

terise what type of vegetation covered the most

northern location of the Southern continent during

this glacial event.

2. Material studied

Twelve samples from the Cape Melville Formationwere selected and analysed for palynomorphs. The

stratigraphical section of the Cape Melville (CM) For-

mation sampled in this study (CM Section of PAS

2007, 2009) was measured during two expeditions by

the Polish Academy of Sciences (PAS) to KGI in Janu-

ary 2007 and January 2009 (Figures 1C and 2). It com-

prises the succession exposed in the southern

promontory of Melville Peninsula east of Crab Creekbetween global positioning system (GPS) points

S 62�01.3910 W 57�37.1390 (bottom of Crab Creek)

and S 62�01.3410 W 57�37.001 (top of Crab Creek),

and S 62�01.4020 W 57�37.0430 (top of Pillar I) and

S 62�01.5520 W 57�35.9760 (top of Pillar VI) (Figure 3).

The detailed location of the 12 samples is shown in

Figure 4. Petrographical composition of the samples

was determined using microscopical examination ofthin sections and x-ray diffraction of disordered pow-

ders. The analytical procedure and instruments used

were identical to those described in Krajewski (2013).

Figure 2. Geological map of the study area studied showing the location of the CM section of the Cape Melville Formation.This section was measured during two expeditions of the Polish Academy of Sciences in 2007 and 2009 (PAS 2007 and 2009).

Palynology 3

Figure 3. A. Melville Peninsula seen from the summit of Peak Melville showing a section of the upper part of the DestructionBay Formation and the Cape Melville Formation (Moby Dick Group). Numbers the indicate lithostratigraphical units of the

4 S. Warny et al.

The samples were taken at the following depths: at

65 m (M34 � Unit 4), 66 m (M31 � Unit 5), 67 m

(M30 � Unit 5), 88 m (M20 � Unit 6), 89 m (M19A �Unit 6), 100 m (M17 � Unit 6), 131 m (M7 � Unit 8),142 m (M5 � Unit 9), 155 m (M3 bottom � Unit 9),

157 m (M3 top � Unit 9), 170 m (M2 top � Unit 10)

and 175 m (M1A � Unit 11) of elevation from the base

of the outcrop.

3. Methods

These samples were chemically processed following a

palynological technique summarised by Brown (2008).

For each sample, about 20 g of dried sediment was

weighed. The sediment was spiked with a known quan-

tity of Lycopodium spores to allow assessment of the

absolute abundance of palynomorphs in the sample.

Acid-soluble minerals were digested in hydrochloricacid (HCl) and hydrofluoric acid (HF) to remove carbo-

nates and silicates respectively. The palynomorphs were

then concentrated by filtration through a 10-mm mesh

sieve. Samples were not sufficiently rich to allow the

counting of 300 palynomorphs per sample, but up to

two full slides were counted for each sample. Reworked

and in situ species were separated primarily on the basis

of stratigraphical range (for all species with a knownrange of Oligocene or older), and/or preservation of the

grains and thermal maturity of the walls. A database of

all palynomorphs recovered was prepared, and key spe-

cies were documented photographically.

4. Palynological results from the Cape MelvilleFormation

Although not rich (total counts ranged from 84 to 272

specimens per sample with an average of 158 specimens

per sample), the Cape Melville Formation samples stud-

ied yielded a diverse group of palynomorphs (Figure 4).

The assemblage is extremely complex because it containsa mix of presumed in situ and reworked microfossils of

diverse provenance. But despite the complexity of the

yield, the palynomorphs provide important information

about the type of environment that existed during the

deposition of the Cape Melville Formation. Because of

the complexity of the yield, the information provided

below lists the species found per sample, from the oldest

to the youngest sample, without discussion of their

reworked versus in situ status. The interpreted reworked

status is indicated by ‘Rw’ on Figure 4, next to the raw

count numbers.

4.1. Cape Melville: sample M34 (at 65 m above base)

Sample M34 was collected from a grey mudstone with

interbedded brown sandstone�mudstone. The sample

yielded ?Baculatisporites sp., Cicatricosisporites hughesii,

Converrucosisporites sp. cf. C. cameronii, Cyathidites

spp. including Cyathidites minor, Dictyophyllidites sp.,

Laevigatosporites ovatus, Leiosphaera spp., Micrhystri-

dium sp., Microcachyridites antarcticus, Nothofagidites

sp. (fusca gp.)., Nothofagidites fortispinulosus, Nothofa-

gidites sp. (brassii gp.), Osmundacidites wellmanii, Phyl-

locladidites spp. including Phyllocladidites mawsonii,

Podocarpidites spp. including Podocarpidites ?exiguus,

Retitriletes sp. cf. R. austroclavatidites, Ruffordiaspora

australiensis and Stereisporites antiquasporites.

4.2. Cape Melville: sample M31 (66 m)

Sample M31 was collected from a brownish mudstone

with grey intercalations. The sample yielded Cyathi-

dites spp. including Cyathidites minor, ?Deltoidospora

directa, Dictyophyllidites sp., Laevigatosporites ovatus,

Leiosphaera spp. Manumiella druggii, Microcachryi-

dites antarcticus, Nothofagidites saraensis, Nothofagi-

dites flemingii, some extremely baked and broken

pollen of possible proteaceous affinity, Protohaploxypi-nus spp. and Punctatisporites sp.

4.3. Cape Melville: sample M30 (67 m)

Sample M30 consists of a brownish mudstone transi-

tioning to sandstone. The majority of specimens recov-

ered are definitely reworked, and many are damaged

beyond classification. As is the case for the two previ-

ous samples, most of the identifiable species recoveredinclude the in situ leiospheres that literally inundate the

slides. Other than the Leiosphaera spp., few recognis-

able specimens were recovered; most specimens

included some spore fragments displaying a very strong

maturation index, most likely resulting from proximity

to magma. Those we could identify include Cyathidites

sp., ?Marsupipollenites striatus, Phyllocladidites sp. and

Punctatisporites sp.

Cape Melville Formation. Units 1 and 2, and the lower part of Unit 3, comprise a diamictite-dominated succession, which is dis-criminated here as the Crab Creek Diamictite (CCD). The ‘Lower Fork (LF)’ and the ‘Upper Fork (UF)’ are informal names forUnits 4 and 6, respectively. These units have been used as local correlation horizons of parts of the geological section measured inPillars I to VII in the southern promontory. B. Southern promontory of Melville Peninsula seen from Sherratt Bay. The red lineindicates the CM section of the Cape Melville Formation on both photographs (PAS 2007, 2009). The upper parts of the Destruc-tion Bay Formation and the Cape Melville Formation are well exposed in the steep rocky walls of the peninsula. The height of thesouthern promontory facing Sherratt Bay (left part of the photograph) is c. 200 m.

Palynology 5

4.4. Cape Melville: sample M20 (88 m)

Sample M20 was collected in a grey ‘gravity flow slope’

mudstone just below a basal unit of brown sandy mud-

stone. It is characterised by Alisporites sp., Cyathidites

minor, Leiosphaera spp., a very corrodedMicrocachryi-

dites antarcticus specimen and Nothofagidites saraensis.

4.5. Cape Melville: sample M19A (89 m)

Sample M19A was collected just above a grey ‘gravity

flow’ mudstone layer in a brown sandy mudstone inter-

val. That sample has mainly leiospheres, with reworked

palynomorphs and rare possibly in situ species. The

assemblage recovered includes Araucariacites australis,

Figure 4. Panoramic geological section (CM section) of the Cape Melville Formation (PAS 2007, 2009) showing the location ofsamples analysed in this paper. See the text for additional explanations.

6 S. Warny et al.

Chenopodipollis sp., Leiosphaera spp., Nothofagidites

sp. (fusca gp.), Nothofagidites ?rocaensis, Phyllocladi-

dites sp. and dinoflagellate cysts reworked and cor-

roded beyond recognition.

4.6. Cape Melville: sample M17 (100 m)

Sample M17 was taken c. 10 m above the previous

sample and consists mostly of grey mudstones. It con-

tains many fungal spores, Alisporites sp., Cyathidites

minor, Leiosphaera spp., Nothofagidites flemingii,

Nothofagidites saraensis, Nothofagidites sp. (fusca gp.),Phyllocladidites sp. and possibly Wanaea sp. (but only

one partial and damaged specimen was recovered).

4.7. Cape Melville: sample M7 (131 m)

Sample M7 was collected towards the upper portion of

the formation in a grey mudstone layer. Although pre-

dominantly characterised by an abundance of leio-spheres, this sample yielded various reworked and in

situ pollen and spores and some large reworked Meso-

zoic dinoflagellate cysts. Main species recovered

include an Asteraceae pollen, ?Converrucosisporites

sp., Coptospora sp., Cribroperidinium spp., Cyathidites

australis, Cyathidites sp., Diconodinium spp., Leios-

phaera spp., Osmundacidites wellmanii, Phyllocladidites

mawsonii, Ruffordiaspora sp., ?Rugulatisporites sp. andpossibly Wanaea sp. (but only partial and damaged

specimens were recovered).

4.8. Cape Melville: sample M5 (142 m)

Sample M5 was collected in a grey mudstone unit

above a major depression. Again, this sample consists

mostly of leiospheres, along with a few whole dinofla-gellate cysts, all of which are reworked. Many indeter-

minate dinoflagellate cyst fragments were observed

though not counted. The sample includes Baculatispor-

ites comaumenis, Coptospora sp., Cyathidites australis,

Cyathidites sp., Diconodinium spp., Gleicheniidites

senonicus, Leiosphaera spp., Manumiella seymourensis,

Nothofagidites ?asperus, Nothofagidites saraensis and

Nothofagidites rocaensis.

4.9. Cape Melville: samples M3 (bottom � 155 m, andtop � 157 m)

The two M3 samples were collected from a grey mud-

stone unit. The first sample shows a slight increase in

whole specimens of dinoflagellate cysts. The lowermost

sample contains Alisporites sp., Diconodinium spp.,Laciniadiniium sp., Leiosphaera spp., Nothofagidites

sp., Oligosphaeridium sp., Podocarpidites sp. cf. P. exi-

guus and Ruffordiaspora ?australiensis.

The uppermost sample is unique in that it contains

the only likely in situ dinoflagellate cyst recovered. A

single specimen of Selenopemphix nephroides was

found in this layer. Other than this occurrence, thesample was composed of unidentifiable reworked

(based on maturation) dinoflagellate cysts, Cycadopites

spp. including Cycadopites follicularis, Gleicheniidites

senonicus, Leiosphaera spp., Nothofagidites rocaensis

(with very distinct oval apertures, rather than dumb-

bell-shaped as in the flemingii types), Nothofagidites sp.

(fusca gp.) and Nothofagidites sp. (brassii a).

4.10. Cape Melville: sample M2 (170 m)

Sample M2 was collected near the uppermost part of

the Cape Melville Formation in a brown mudstone

with paler sandstone intercalations. This sample, much

like the others, shows a predominance of leiospheres,

Haloragacidites harrisii and Nothofagidites spp.,

including Nothofagidites rocaensis.

4.11. Cape Melville: sample M1A (175 m)

The uppermost sample, sample M1A, was taken at the

top of the section in a grey mudstone. It yielded some

Asteraceae, a Caryophyllaceae of Colobanthus type,

Cyathidites minor, Gleicheniidites senonicus, Leios-

phaera spp., Nothofagidites sp. (fusca gp.), Pediastrumsp., Phyllocadidites sp., Podocarpidites ?exiguus,?Rugu-

latisporites sp., Tricolpites ?confessus, Pediastrum and

scolecodonts.

5. Environmental significance of the assemblages

recovered

In order to extract a palaeoenvironmental signal fromthe complex assemblage recovered, the first task is to

identify whether a species is reworked or in situ. When

age-diagnostic species older than Miocene are found,

the task is simple. For long-ranging species, discerning

between reworked and in situ is less straightforward

because palynomorphs are resilient microfossils that

can withstand heat, ageing and some depositional pres-

sure. The task is particularly difficult when workingwith Antarctic sediments post-dating the inception of

latest Eocene glacial conditions, as reworking (some-

times multiple cycles) is always a consideration. The

approach taken herein was to identify reworking first

on the basis of certain known ranges to avoid circular

reasoning. When ranges were too long to determine the

reworking status, the preservation of the grains and

the thermal maturity of their wall were considered.Figure 5 summarises our interpretation and illustrates

the difference in thermal maturation observed between

reworked and presumed in situ specimens.

Palynology 7

The summary presented in Figure 5 shows the distri-

bution of the main types of palynomorphs recovered,

divided into what we know to be definitely reworked,and what we evaluated to be most likely in situ. The first

observation is that the amount of in situ is higher than

the reworked. But if the leiospheres are excluded from

the in situ category, then it is obvious that remnants of

penecontemporaneous vegetation (estimated from the in

situ spores, gymnosperm and angiosperm pollen grains)

are very rare, ranging from 0 to 10.3% of the entire

assemblage recovered and averaging 4.6% of the totalrecovery. The sparse occurrences of likely in situ species

indicate that there was some limited vegetation growing

on the nearby land during the Melville Glaciation. The

rare in situ cryptogam spores recovered comprise a sev-

eral few moss spores (Coptospora). The flora also

includes rare specimens of podocarp conifer pollen

(Podocarpidites spp.), and rare angiosperms such as some

Asteraceae, a single Caryophyllaceae pollen grain of theColobanthus type � one of two vascular plants that can

still be found on the northern Antarctic Peninsula today

(Kom�arkov�a et al. 1985) � a single Chenopodipollis sp.

and a few Nothofagidites spp. (most likely low-growing

species). All of the belong to the colder temperate fusca-

type (including N. lachlaniae and N. saraensis). Interest-

ingly, the most common pollen type in this very sparse

assemblage is Nothofagidites, consistent with a continua-tion of a Nothofagus-rich Eocene�Oligocene flora on

KGI, as described by Torres & M�eon (1998).

Other than these indications of the type of vegeta-

tion that was surviving on the nearby land, the major-

ity of the assemblage is characterised throughout the

section by Leiosphaera spp., a group of acritarchs that

is known to be mostly abundant at the limit between

pack ice and sea ice (Mudie 1992; Troedson & Riding2002; Warny et al. 2006; Warny et al. 2009). Of inter-

est, and confirming this environmental trend, is the

recovery of a single specimen of the cold-adapted

marine dinoflagellate cyst Selenopemphix nephroides, a

species that was also found by Troedson & Riding

(2002) when these authors analysed four samples from

Cape Melville, three of them in Miocene sediments.

These authors stated that Selenopemphix nephroides isindicative of deposition in a glacial setting. Our data

(the rarity of in situ dinoflagellate cysts and the abun-

dance of leiospheres) confirm that the section was

deposited under marine conditions influenced by sea-

ice at the site, or in close proximity to the site.

Figure 5. Palynological diagram summarising all species recovered, graphed by actual counts. The species are organised by theirbiological affinity. Rw indicates reworked palynomorphs.

8 S. Warny et al.

6. Provenance of reworking sources and implications

for glacial advances

6.1. Age of the reworked assemblage

As previously mentioned, the most difficult aspect of

working on the Cape Melville Formation is how to

assess the reworked component and understand what

the reworked species mean. Ice-rafted dropstones vary-

ing in size from sand fraction up to blocks a few metres

across are derived from various continental sources,including the local substratum of the SSI, the Antarctic

Peninsula, the Ellsworth Mountains and possibly also

the Pensacola�Theron Mountains (Birkenmajer 2001

and references therein).

If Leiosphaera spp. are excluded, the assemblage

recovered is mostly characterised by reworked speci-

mens of various ages and thermal maturities. This

diversity makes it almost impossible to decipher thesource of the reworking with certitude. The ranges of

the reworked species indicate that reworked palyno-

morphs came from at least three different sources, a

mostly Permian source (or sources), a Jurassic to Cre-

taceous source (or sources) and a Paleogene source (or

sources). Some species, because of their longer range,

could be in two of these groups, which makes the inter-

pretation extremely complicated. There is no cleartrend in the distribution of reworking. In most sam-

ples, a mix of various ages is found together with the

in situ material. Nonetheless, we noted that most of the

Permian reworking is found in the base of the section,

while most of the reworked Cretaceous dinoflagellate

cysts are present in the upper portion of the outcrop.

This indicates possible changes in the source of the

reworking between the base and the top of the outcrop.This may denote some transitional or punctuated

events occurring at the time of deposition, such as

changes in the intensity and direction of ice-sheet

advances, or the incorporation of various ice-rafted

debris during the deposition of the formation. The

reworked palynomorphs and their possible sources are

discussed below under their broad age grouping.

Figure 6 presents a summary of all the formations dis-cussed, organised by ages.

6.2. Potential source of the reworked assemblages

Because of the complexity (various ages and prove-

nance) of the reworked assemblage, trying to tie the

palynological signal to sites with published palaeoflora

information was difficult, at best. One of the problems

is that, for Permian to Cretaceous plants, the biological

affinity between a plant and its pollen or spore is notalways established. Many of the studies on the SSI

have dealt with the macrofloral fossil record, and there

is not always a clear link between a plant and its pollen

or spores. Below is our evaluation of potential sources

(see Figure 7 for locations discussed) based on the pub-

lished literature.

6.3. Permian

Protohaploxypinus spp. and Marsupipollenites striatus

are Permian indicators, though the former taeniate

bisaccate pollen type can range into the Triassic. These

reworked Permian pollen occur in the lower part of the

Cape Melville section, in samples M30 and M31. It is

possible that a few of the other less diagnostic palyno-morphs are also derived from Permian strata. These

included some specimens of Alisporites sp., or the

unidentified spores and unidentified taeniate pollen

such as in sample M17 (100 m), which would spread

the inclusion of reworked Permian into younger parts

of the formation. For purposes of discussion, we

include here the two first-mentioned age-restricted

taxa. The specimen listed as ?Marsupipollenites striatus

is poorly preserved and dark brown in colour; how-

ever, the specimens of Protohaploxypinus, although

somewhat damaged with torn sacs, are a lighter colour

and not as heavily metamorphosed.

The source of these Permian (to Triassic?) speci-

mens might possibly be some strata broadly included

in the Trinity Peninsula Group exposed on Trinity Pen-

insula between Hope Bay and Paradise Harbour, andalso on Livingston Island close to KGI and also part

of the SSI (Birkenmajer 1992a; Doktor et al. 1996). We

note that glacial deposition of the Cape Melville For-

mation, with its reworked palynomorphs, predates the

formation of the Bransfield Strait. The Bransfield Rift,

which now separates the Antarctic Peninsula and the

SSI, is less than 4 million years old, with oceanic crust

in the Bransfield Strait from c. 1.3 Ma (Barker & Aus-tin 1998).

Trinity Peninsula Group sediments, including tur-

bidites and submarine fan deposits, are exposed in the

SSI, on Livingston Island as the Miers Bluff Forma-

tion (Arche et al. 1992; Doktor et al. 1994) and on the

Trinity Peninsula as the Hope Bay, View Point and

Legoupil formations (e.g. Hyden & Tanner 1981; Bir-

kenmajer 1992a, 1992b). Although generally given anage of ?Permian�Triassic, the only definitive and reli-

able fossils found are of Triassic age, such as marine

invertebrates from the Legoupil Formation (Thomson

1975), and Late Triassic palynomorphs from the Miers

Bluff Formation (Deng et al. 2002). Radiometric ages

also indicate a Triassic age for some localities (Trouw

et al. 1997). A maximum age for the Trinity Peninsula

Group is < 350 Ma, Carboniferous, from ura-nium�lead (U�Pb) detrital zircon dating (Loske et al.

1988; Barbeau et al. 2009). No definitive Permian fos-

sils have been found in this region, and certainly no

Palynology 9

barely-metamorphosed non-marine Permian sedi-

ments. The high thermal maturity of much of the Trin-

ity Peninsula Group, including at Hope Bay, hasmostly precluded the extraction of identifiable palyno-

morphs, with previous attempts yielding only black

high-rank organic matter (Askin & Elliot 1982; Birken-

majer 1992b).

We note that a source of the glacially-derived

Permian palynomorphs far to the south beyond the

base of the Antarctic Peninsula (the Transantarctic

Mountains�Ellsworth Mountains) is not ruled out,as suggested for the provenance of ice-rafted drop-

stones of varying sources (Birkenmajer 2001 and

references therein). Although the thermal matura-

tion of Permian spores and pollen from these

regions is generally much higher than seen in our

specimens. The fact remains, however, that Permian

non-marine palynomorphs are consistently found in

almost all Mesozoic and Cenozoic samples evalu-ated for palynomorphs in the Antarctic Peninsula

region (e.g. this study; Askin & Elliot 1982; Warny

& Askin 2011a, 2011b; Bowman et al. 2014). These

span varying ranks of thermal alteration, from well-

preserved yellow specimens that are derived from

barely buried sediments, right through to black,

barely recognisable and broken remnants. Askin &

Elliot (1982) discussed the problems associated withidentifying the provenance of the well-preserved

spores and pollen, and suggested a possible source

from the forearc terrain. A key point regarding

these (albeit rare) Permian palynomorphs is that

they provide further indirect evidence for the former

presence of undeformed/unmetamorphosed non-

marine (to shallow marine?) Permian sediments in

the Antarctic Peninsula.Another likely source of Permian palynomorphs in

the Cape Melville Formation is a secondary cycle of

reworking of sediments from the James Ross Basin

(see discussion below), northern Antarctic Peninsula.

Well-preserved Permian palynomorphs occur sporadi-

cally reworked throughout the shallow marine James

Ross Basin sediments, and they are especially common

in the upper part of the Eocene La Meseta Formation(Askin & Elliot 1982). Their appearance, particularly

in the lower part of the Cape Melville Formation, is

consistent with a reverse stratigraphy seen with unroof-

ing of youngest sediments at the top of the sedimentary

pile in the earliest stage of glaciation.

Figure 6. Summary diagram showing key palynomorphs grouped as reworked species: a. Nothofagidites sp. 22 mm (M34C U11/2), b. Diconodinium sp. 31 mm (M5 S35/2), c. Cribroperidinium sp. 67 mm (M7 U14/4), d. Podocarpidites sp. 30 mm (M34C T27/0), e. Ruffordiaspora australiensis 33 mm (M34C U31/1), f. Converrucosisporites sp. 28 mm (M34C W14/2), and in situ species: g.Nothofagidites sp. (fusca gp.) 24 mm (M19A S38/0), h. Asteraceae 16 mm (M7 U21/2), i. Selenopemphix nephroides 20 mm (M3V41/3), j. Podocarpidites sp. cf. P. exiguus 35 mm (M3 T27/0), k. Coptospora sp. 23 mm (M5 W14/1), l. Leiosphaera sp. 20 mm(M31 Y15/0) with thin wall, m. Leiosphaera sp. with thick wall 26 mm (M19 X46/4). The species name is followed by the largestwidth of the specimen illustrated and the England Finder (EF) coordinate. Note that each palynological group (e.g. dinoflagellatecysts, leiospheres, spores, gymnosperm pollen, angiosperm pollen) is colour coded in the distribution diagram and matches thecolour of the microphotograph frames for easy association between the palynomorph illustrations and the abundance curves.

10 S. Warny et al.

6.4. Upper Jurassic�Cretaceous

Most of the reworked palynomorphs are grouped within

the broad category of Upper Jurassic�Cretaceous.

Most of the taxa in this category are long-ranging

throughout the Cretaceous, with some (e.g. fern

spore Ruffordiaspora australiensis, podocarp pollen

Microcachryidites antarcticus) ranging from the Upper

Jurassic, which is why a broad, inclusive Upper Jurassic�Cretaceous category is used here. Some range into thePaleogene, and comments on those are included below

in the more age-restricted reworked Paleogene section.

Very long-ranging types, such as the moss spore

Figure 7. Sketch map of the northern Antarctic Peninsula region showing known locations that are potential sources of rework-ing and redeposition into the early Miocene Cape Melville (CM) Formation. One of the major differences between the present-day configuration and the time of deposition of the CM section is the fact that the Bransfield Strait was closed during the Mio-cene. See the text for additional explanations.

Palynology 11

Stereisporites antiquasporites, the fern spores Osmunda-

cidites wellmanii and Gleicheniidites senonicus, and the

gymnosperm pollen Araucariacites australis, which are

Triassic through Paleogene (and through Neogene insome austral areas), are also included in this grouping.

Some taxa are more restricted, and represent

reworking of more specific strata within this category

(such as the Maastrichtian dinoflagellate cysts Manu-

miella seymourensis and M. ?druggii). These are dis-

cussed below.

There are various exposed outcrops of Jurassic�Cretaceous age in the SSI and northern Antarctic Penin-sula that could provide the source of reworked marine

and terrestrial palynomorphs grouped within this broad

age category. Strata of this age on the northern Antarc-

tic Peninsula include the well-known non-marine beds

with well-preserved leaf fossils from the Mount Flora

Formation exposed at Hope Bay in the northern Trinity

Peninsula, a location that was discovered during Otto

Nordenskj€old’s expedition between 1901 and 1904. Theycould also come from slightly further south at Botany

Bay. In numerous attempts (Askin, unpublished; other

studies listed in Rees & Cleal 2004, p. 74), it has not

been possible to extract palynomorphs from these sedi-

ments due to their high thermal maturity, and the age of

the plant fossil assemblages has been contentious, vary-

ing in more recent studies from Early Jurassic (Rees &

Cleal 2004; see discussion and references therein) toEarly Cretaceous/Berriasian (Gee 1989). A Jurassic age

is generally accepted, and supported by indirect radio-

metric evidence from Botany Bay that indicates the plant

beds are older than late Middle/Late Jurassic (152 § 8

Ma, Millar et al. 1990). With the high thermal maturity

of these currently exposed sediments, and their Early (to

early Mid?) Jurassic age, and as no restricted Early

Jurassic palynomorphs occur among our reworkedassemblage, these plant beds are not considered a proba-

ble source of our reworked assemblage.

Another possibility is that this group of reworked

palynomorphs is more locally-derived, from shallow

marine to non-marine sediments from the SSI. Notable

occurrences of plant-bearing strata of this general age,

with poorly to moderately preserved palynomorphs,

are found in the shallow marine to non-marine volcani-clastic Byers Group sediments on Byers Peninsula,

westernmost Livingston Island and immediately to the

south on President Head, easternmost Snow Island.

Aside from the several papers on the plant macroflora

(e.g. Hernandez & Azc�arate 1971; Fuenzalida et al.

1972; Torres et al. 1997; Cantrill 1998), the palynology

papers are of direct relevance to this study. Preliminary

papers (Askin 1983; Duane 1994) and more compre-hensive studies (Duane 1996; 1997) described spores,

pollen and dinoflagellate cysts of Late Jurassic (Titho-

nian) to Early Cretaceous age. Some of the long-

ranging taxa noted above are common to both the

Byers Group samples and the Melville reworked

assemblage. However, taxa that characterised at least

some of the Byers samples, and were abundant in theyoungest lacustrine samples, such as the distinctive

fern spore Cyatheacidites annulatus (which now

includes C. tectifera), were not observed in the

reworked assemblage. But this provenance is only one

of the possible scenarios, as a similar trend is observed

with some of the dinoflagellate cysts recovered. Dino-

flagellate cyst species in Batioladinium, Cribroperidi-

nium and Oligosphaeridium found by Duane (1996)from the Byers Group on Livingston and Snow Islands

were also found in the Cape Melville reworked assem-

blage, but these constitute only a small fraction of the

Byers Group dinoflagellate cysts.

Thus, there is no definitive evidence for a south-

westerly sourced component from the Byers Group

sediments, although they could be a source of the

unidentifiable dinoflagellate cyst remnants. There areother possible Cretaceous beds (or equivalents) from

Livingston Island, such as the Williams Point Beds

(once assumed to be Triassic, but now assigned to the

Cretaceous (e.g. Rees & Smellie 1989; Poole & Cantrill

2001 and references therein) and possibly Coniacian

(Poole et al. 2005), and the Coniacian�Santonian

plant beds at Hannah Point on Livingston Island

(Leppe et al. 2007).There are also local occurrences of Upper Creta-

ceous volcano-sedimentary successions containing

plant megafossils and palynomorphs, such as the Cam-

panian strata on Fildes Peninsula and the Zamek For-

mation at the Zamek locality in Admiralty Bay (e.g.

Birkenmajer & Zastawniak 1989; Dutra & Batten

2000, Zastawniak & Szafer 1990, 1994 and references

therein), that could provide a source for some of thereworked material. It should be noted, however, that

recent geochronological investigation suggest an

Eocene age for the Zamek Formation.

One of the most likely sources of the reworked

Upper Jurassic�Cretaceous group of marine and ter-

restrial palynomorphs, with mixed though not high

thermal maturities, is via a northwards-flowing glacier

from the southeastern flank of the northern AntarcticPeninsula and James Ross Basin. A succession of dino-

flagellate cyst floras, along with associated spores and

pollen, has been described from marine sediments of

the Nordenskj€old Formation at Sobral Peninsula, and

Dundee and Joinville islands (Kimmeridgian�Berria-

sian), and the Gustav Group on James Ross Island

(Aptian to Coniacian), by Riding et al. (1998), Riding

& Crame (2002 and references therein), and from theSantonian to early Danian Marambio Group on James

Ross Island and adjacent islands (e.g. Dettmann &

Thomson 1987; Askin 1988; numerous authors in

12 S. Warny et al.

Duane et al. 1992; Bowman et al. 2014). These beds

include all the taxa we find reworked into the

Cape Melville Formation, including the age-restricted

dinoflagellate cysts Manumiella seymourensis (Maas-trichtian) and Manumiella druggii (late Maastrich-

tian�earliest Danian) which occur in the L�opez de

Bertodano Formation on Seymour Island (Askin 1988,

1999; Thorn et al. 2009; Bowman et al. 2012). These

are the most proximal reworking sources, but there is

no telling how far south along the Antarctic Peninsula

reworked assemblages could have travelled via ice-

rafted dropstones.We note that Cretaceous reworking has not only

been observed from the palynological assemblages.

Recycled Cretaceous belemnites (Birkenmajer &

ºuczkowska 1987) and calcareous nannoplankton

(Dudziak 1984) provide a substantial addition to the

indigenous marine fossil spectrum in the Cape Melville

Formation. The latter embraces various bivalves, gas-

tropods, brachiopods, solitary corals, crabs, echinoids,bryozoans, and calcitic and arenaceous foraminifera

(Birkenmajer & ºuczkowska 1987). The exact location

of the Cretaceous sediments that acted as a source for

the recycled belemnites and nannofossils is also

unknown.

6.5. Paleogene

The source of the recycled Paleogene palynoflora may

be simpler to identify. This grouping includes taxa of

broad Paleogene age, with some, as noted below, more

long ranging, and some more restricted. As stated

above, some taxa grouped within the Upper Jurassic-

�Cretaceous group range into the Paleogene (and even

younger in some cases), though these have been

included in the older reworked grouping because oftheir darker colour (higher thermal maturation) and

poorer preservation. Examples of these longer ranging

taxa include spores such as Baculatisporites comaumen-

sis, Cyathidites australis, C. minor, Gleicheniidites seno-

nicus and Laevigatosporites ovatus, and gymnosperm

pollen Araucariacites australis, Microcachrydites ant-

arcticus, Osmundacidites wellmanii, Phyllocladidites

mawsonii and Stereisporites antiquasporites.Taxa grouped in the Paleogene reworked cate-

gory are largely species of Nothofagidites, which is

consistent with the diverse and abundant Nothofa-

gus species apparently thriving in the region during

the Paleogene (e.g. Stuchlik 1981; Torres & M�eon1998). This group likely also included cryptogam

spores and gymnosperm pollen as noted above. For

a source of these palynomorphs, there are manyplant-bearing outcrops of Paleogene age locally on

KGI. There are numerous published studies on the

macrofossils (leaves, wood) and occasionally the

palynomorphs. This paper does not attempt to

cover them all, but instead gives a few representa-

tive studies. The plant beds include the Fossil Hill

Formation and wood flora in moraine blocks ofCollins Glacier on Fildes Peninsula (e.g. Poole et al.

2001, 2005, and references therein); the Cytadela

Plant Beds, Point Thomas Formation (Birkenmajer

& Zastawniak 1989; Mozer 2012, 2013); and various

plant beds (including the Point Hennequin flora in

moraine blocks of the Dragon Glacier) in the

Mount Wawel Formation (Birkenmajer & Zastaw-

niak 1989; Zastawniak et al. 1985). These papersdescribe plant megafloras spanning the Paleocene to

Oligoceneage. Palynomorph assemblages are

described (Stuchlik 1981) from the Petrified Forest

Member of the Arctowski Cove Formation, of Lute-

tian ([middle] Eocene) age (not Oligocene, as stated

in the paper), and a diverse assemblage of Nothofa-

gidites pollen of Eocene�Oligocene age from the

Fildes Peninsula (Torres & M�eon 1998).As with the Mesozoic reworked material discussed

above, the James Ross Basin remains a viable more

southern source of the Paleogene reworked material.

Paleogene beds in the James Ross Basin include the

mostly shallow marine uppermost L�opez de BertodanoFormation, Sobral, Cross Valley and La Meseta for-

mations of Seymour Island, which contain rich dinofla-

gellate cyst, acritarch, spore and pollen assemblages(e.g. Askin 1988; Wrenn & Hart 1988; Bowman et al.

2012; 2014). Indirect evidence of this James Ross Basin

source lies in the presence of moderately well-preserved

Permian pollen, possibly derived from secondary

reworking as discussed in the Permian section above.

There is a lack, however, of the distinctive Transan-

tarctic Suite of Eocene dinoflagellate cysts from the La

Meseta Formation reworked into the Melville Forma-tion. Perhaps the reworked material is a mixture of the

more southern James Ross Basin sediments and locally

derived KGI material.

The complexity of the reworking signal (Per-

mian�Oligocene) seen in the Miocene samples is in

line with the complex glacial history known from

seismic records collected in the Antarctic Peninsula

regions (Bart & Anderson 1995; Bart et al. 2005).Although the precise number of ice sheet advances in

the region studied between c. 23 and 21 Ma is

unknown, we have some good indications for the

middle Miocene. In their 1995 paper, Bart & Ander-

son used intermediate-resolution seismic-reflection

data to show that ice sheets advanced onto the conti-

nental shelf in the middle Miocene time during at

least 18 short-duration glacial episodes. Such a com-plex glacial history makes it impossible to pinpoint

precisely the source of reworking, but it gives a clear

indication of the former and current existence of

Palynology 13

pristine Permian to Oligocene sections, some of

which, possibly ice-covered, have yet to be discov-

ered and sampled in the region.

7. Conclusions

The samples provided by the Institute of Geological

Sciences, Polish Academy of Sciences, from KGI allowed

us to evaluate palaeoenvironments on two fronts. First,

the presumed in situ component gave us an indication of

the type of vegetation that survived on the landscape dur-

ing the early Miocene Melville Glaciation (c. 23�21 Ma).Second, the reworked component gave us some insight,

even if not definitive, into the potential sources of

reworking, and provided evidence of glacial advances to

the Melville Peninsula at that time.

Palynomorphs recovered from the Cape Melville

Formation reveal a nearshore marine depositional

environment dominated by sea-ice as indicated by the

abundance of leiospheres in all samples. The abun-dance of various Permian�Mesozoic reworked species

mixed with Paleogene and in situ Miocene species indi-

cates that several episodes of glacial advances brought

various types of reworked pollen, spores and marine

palynomorphs from a suite of locations in the area.

Particularly, in samples M3 (at 155 and 157 m), a

diversity of most likely Cretaceous species indicates a

different provenance of material for the top of the sec-tion. To work on provenance, a detailed palynological

sampling of all the sections mentioned above would be

needed, but, as mentioned, numerous past efforts to do

so were hindered by the high thermal maturity of some

of the Permian�Cretaceous strata. Such an effort

would also require several months of field seasons on

the SSI and the Northern portion of the Antarctic Pen-

insula. The success of the mission would have to relyon strong international collaboration, such as one

between the Arctowski Polish Antarctic Station and

the U.S. Palmer Station.

But despite our inability to better constrain the

source of the reworking, some clear indication about

the type of vegetation that dominated the landscape

during the deposition of the Cape Melville Formation

was established. Potentially contemporaneous terres-trial species include mostly Nothofagidites spp. of the

fusca group (cool�cold climate southern beech), with

sporadic and rare pollen of Podocarpidites sp., Astera-

ceae, Caryophyllaceae (Colobanthus-type), Chenopodipol-

lis and moss spores Coptospora sp. This is a similar

assemblage to that described from Miocene SHALDRIL

cores off Seymour Island (Warny & Askin 2011). The

assemblage indicates the sparse vegetation that survivedduring the early Miocene Melville Glaciation was strictly

limited to periglacial tundra species with likely shrubby

southern beech and podocarps. This is in agreement

with the vegetation that was able to recolonise and pro-

liferate in parts of Antarctica during the Mid Miocene

Climatic Optimum (Warny et al. 2009; Feakins et al.

2012) and survived in the northern Antarctic Peninsuladuring the early Miocene (Warny & Askin, 2011a;

Anderson et al. 2011).

Acknowledgements

The authors would like to thank the Institute of Geological Sci-ences, Polish Academy of Sciences, for providing the samples.Field assistance by Marcin Klisz and Mariusz Potocki duringthe 2007 expedition and Grzegorz Zieli�nski during the 2009expedition is greatly appreciated. The authors are grateful tothe anonymous reviewers for their review of this manuscript.Thanks are extended to HESS corporation (D. Pocknall) forthe donation of the Stratabugs software and to HESS consul-tant (P. Griggs) and intern (M. Thomas) for training the Centerfor Excellence in Palynology (CENEX) research group.

Disclosure statement

The work presented here is strictly academic in nature and wasdone as part of the training of a graduate student. The authorsdo not have any financial interest arising from this study.

Funding

The fieldwork was financially supported by the Polish Ministryof Science and Higher Education [Grant No. DWM/N8IPY/2008]. This is a contribution to the Antarctic Climate Evolu-tion (ACE) program. We are grateful to the US National Sci-ence Foundation (NSF) Polar and CAREER programmes forfunding this research. Funding for this research was providedby the US NSF grant ANT-1048343 to SW.

Author biographies

SOPHIE WARNY is an associate professor of palynology inthe Department of Geology and Geophysics, and a curatorat the Museum of Natural Science at Louisiana State Univer-sity (LSU) in Baton Rouge, Louisiana, USA. Sophie has along history with American Association of ShragigraphicPalynologists (AASP) as she won the AASP Student Awardin 1996. Sophie received her PhD from the Universit�e Catho-lique de Louvain, in Belgium, working with Jean-Pierre Suc;her doctoral dissertation was on the Messinian Salinity Cri-sis. Since graduating, Sophie has been working on Antarcticsediments that were acquired via the ANDRILL SMS andSHALDRIL projects. In 2011, she received the National Sci-ence Foundation CAREER award to support her research inAntarctica. In addition to her research, Sophie teaches histor-ical geology, palaeobotany and micropalaeontology. Sophiecurrently has a research group that is composed of three PhDstudents, three masters students and one undergraduate, allworking on sections ranging in age from Cretaceous to Ceno-zoic. Since being hired at LSU, Sophie has graduated ninestudents; all of which are now employed in the oil and gasindustry with Hess, BP, Devon, Chevron, BHP BillitonPetroleum and EOG.

C. MADISON KYMES is currently pursuing an MSc in pal-ynology in the Department of Geology and Geophysics atLouisiana State University, Baton Rouge, Louisiana, USA

14 S. Warny et al.

under the direction of Sophie Warny. Madison was born inShreveport, Louisiana, but spent the majority of his lifegrowing up in Madison, Mississippi, where he recentlyinterned as an environmental geologist.

ROSEMARY A. ASKIN is currently pursuing research in thepalynology of Antarctica and consulting in Jackson, Wyoming,USA. She earned her PhD from Victoria University, Welling-ton, New Zealand. Rosemary has researched and taught in sev-eral US universities, most recently at the Byrd Polar ResearchCenter, Ohio State University. Her most recent project therewas establishing the US Polar Rock Repository.

KRZYSZTOF P. KRAJEWSKI is a professor of earth scien-ces at the Institute of Geological Sciences, Polish Academy ofSciences, in Warsaw, Poland. He obtained his PhD in 1986,and continued research at the Polish Academy of Sciences,the University of Oslo and the University of Oldenburg.Krzysztof completed his DSc in sedimentary petrology andgeochemistry in 2001. His scientific interests are on the inter-actions between biological activity and mineral deposition incarbonate, phosphate and organic carbon-rich sedimentarysystems. For more than 20 years, Krzysztof has undertakengeological research in the Arctic, where he has concentratedon Mesozoic phosphorites and petroleum source beds. He iscurrently involved in research on the evolution of Cenozoicsedimentary systems in Antarctica.

PHILIP J. BART is an associate professor in the Departmentof Geology and Geophysics at Louisiana State University inBaton Rouge, Louisiana, USA. He earned his PhD fromRice University in Houston, working with John Anderson.Philip received the Presidential Early Career Award for Sci-entists and Engineers in 2001 for his sequence stratigraphicalresearch in Antarctica. He is one of the four current US Rep-resentatives to the Standing Scientific Group on GeoSciencesof the Scientific Committee on Antarctic Research. Philip’sresearch is on understanding the Cenozoic history of thecryosphere from the perspective of outer continental shelfseismic stratigraphy. The major emphasis of his work hasbeen to constrain the time, frequency, duration and locationsof past ice sheet dynamics within the context of possible forc-ing mechanisms.

References

Anderson JB. 1999. Antarctic Marine Geology. CambridgeUniv. Press, Cambridge, U. K. 297 pp.

Anderson JB, Warny S, Askin RA, Wellner JS, Bohaty SM,Kirshner AE, Majewski W. 2011. Progressive Cenozoiccooling and the demise of Antarctica’s last refugium.Proc Natl Acad Sci. 108(28):11356�11360.

Arche A, L�opez-Martinez J, Martinez de Pison E. 1992. Sedi-mentology of the Miers Bluff Formation, LivingstonIsland, South Shetland Islands. In: Yoshida Y, Kami-numa K, Shiraishi K, editors. Recent progress in Antarc-tic Earth science, Tokyo: Terra Antartica Publication;p. 357�362.

Askin RA. 1983. Tithonian (uppermost Jurassic)�Barremian(Lower Cretaceous) spore, pollen and microplanktonfrom the South Shetland Islands, Antarctica. In: OliverRL, James PR, Jago JB, editors. Antarctic Earth Science,Canberra: Australian Academy of Science; p. 295�297.

Askin RA. 1988. Campanian to Paleocene palynological suc-cession of Seymour and adjacent islands, northeasternAntarctic Peninsula. In: Feldmann RM, Woodburne

MO, editors. Geology and paleontology of SeymourIsland, Antarctic Peninsula, Memoir: Geological Societyof America; 169:131�153.

Askin RA, Elliot DH. 1982. Geologic implications ofrecycled Permian and Triassic palynomorphs in Tertiaryrocks of Seymour Island, Antarctic Peninsula. Geology10(4):547�551.

Barbeau DL Jr, Davis JT, Murray KE, Valencia V, GehrelsGE, Zahid KM, Gombosi DJ. 2009. Detrital-zircon geo-chronology of the metasedimentary rocks of north-west-ern Graham Land. Antarctic Sci. 22:65�78.

Baker PF. 2007. The history of Antarctic Peninsula glacia-tion. In: Cooper AK, Reymond CR et al., editors. Ant-arctica: A Keystone in a Changing World�OnlineProceedings of the 10th ISAES. USGS Open-File Report2007�1047, Short Research Paper 042, 5p.; doi: 10.3133/0f2007-1047.srp042.

Barker PF. 1982. The Cenozoic subduction history of thePacific margin of the Antarctic Peninsula: ridge-crust-trench interactions. J Geol Soc London 139: 787�801.

Barker DHN, Austin JA. 1998. Rift propagation, detach-ment faulting, and associated magmatism in BransfieldStrait, Antarctic Peninsula. J Geophysical Res.103:24017�24043.

Bart PJ, Anderson JB. 1995. Seismic record of glacial eventsaffecting the northwestern Antarctic Peninsula continen-tal shelf. In: Cooper AK, Barker PJ, Brancolini G, edi-tors. Antarctic research series. Washington, D.C.:American Geophysical Union; p. 74�95.

Bart PJ, Egan DC, Warny S. 2005. Direct constraints on Ant-arctic Peninsula Ice Sheet grounding events between 5.12and 7.94 Ma. J Geophysical Res�Earth Surface (110).doi: 10.1029/2004JF000254.

Birkenmajer K. 1992a. Trinity Peninsula Group (Permo-Tri-assic?) at Paradise Harbor, Antarctic Peninsula. StudiaGeologica Polonica 101:7�25.

Birkenmajer K. 1992b. Trinity Peninsula Group (Permo-Tri-assic?) at Hope Bay, Antarctic Peninsula. Polish PolarRes. 13(3-4):215�240.

Birkenmajer K. 2001. Mesozoic and Cenozoic stratigraphicunits in parts of the South Shetland Islands and NorthernAntarctic Peninsula (as used by the Polish Antarctic Pro-grammes). Studia Geologica Polonica 118:5�188.

Birkenmajer K, Keller RA. 1990. Pleistocene age of the PeakMelville volcano, King George Island, West Antarctica.Bull, Polish Acad Sci Earth Sci. 38:17�24.

Birkenmajer K, ºuczkowska E. 1987. Foraminiferal evidencefor a Lower Miocene age of glaciomarine and relatedstrata, Moby Dick Group, King George Island (SouthShetland Islands, Antarctica). Studia Geologica Polonica90:81�123.

Birkenmajer K, Ga�zdzicki A, Kreuzer H, M€uller P. 1985. K-Ar dating of the Melville Glaciation (Early Miocene) inWest Antarctica. Bull Polish Acad Sci Earth Sci.33:15�23.

Birkenmajer K, Soliani E, Kawashita K. 1988. Early Mio-cene K-Ar age of volcanic basement of the Melville Gla-ciation deposits, King George Island. West Antarctica 36(1):25�34.

Birkenmajer K, Zastawniak E. 1989. Late Cretaceous-earlyTertiary floras of King George Island, West Antarctica:their stratigraphic distribution and palaeoclimatic signifi-cance. In: Crame JA, editor. Origins and evolution of theAntarctic Biota, Geological Society Special Publication,47:227�240.

Palynology 15

Bittner MA, Crame JA. 2002. Brachiopods from the LowerMiocene of King George Island, West Antarctica. PolishPolar Res. 23:75�84.

Bowman VC, Francis JE, Askin RA, Riding JB, SwindlesGT. 2014. Latest Cretaceous�earliest Paleogene vegeta-tion and climate change at the high southern latitudes:palynological evidence from Seymour Island, AntarcticPeninsula. Palaeogeography, Palaeogeography, Palaeo-climatology 408:26�47.

Bowman VC, Francis JE, Riding JB, Hunter SJ, HaywoodAM. 2012. A latest Cretaceous to earliest Paleogene dino-flagellate cyst zonation from Antarctica, and implicationsfor phytoprovincialism in the high southern latitudes.Rev Palaeobotany Palynology 171:40�56.

Brown CA. 2008. Palynological Techniques. Riding JB andWarny S, editors. AASP Special Publications. 137pp.

Cantrill DJ. 1998. Early Cretaceous fern foliage referable toLophosoriaceae from President Head, Snow Island, Ant-arctica. Alcheringa. 22:241�258.

Davies BJ, Hambrey MJ, Smellie JL, Carrivick JL, GlasserNF. 2012. Antarctic Peninsula Ice Sheet evolution duringthe Cenozoic Era. Quaternary Sci Rev. 31:30�66.

Deng X, Zheng X, Liu X, Ouyang S, Shen Y. 2002. Terres-trial palynomorphs from the Miers Bluff Formation ofLivingston Island, West Antarctica. In: Gamble JA,Skinner DNB, Henrys S, editors. Antarctica at the closeof a millennium. Royal Soc N Z Bull. 35:269�273.

Dettmann ME, Thomson MRA. 1987. Cretaceous palyno-morphs from the James Ross Island area, Antarctica�apilot study. Br Antarctic Surv Bull. 77:13�59.

Dingle RV, Lavelle M. 1998. Antarctic Peninsula cryosphere:Early Oligocene (c. 30 Ma) initiation and a revised glacialchronology. J Geol Soc London 155:433�437.

Doktor M, Swierczewska A, Tokarsi, AK. 1994. Lithostra-tigraphy and tectonics of the Miers Bluff Formation atHurd Peninsula, Livingston Island (West Antarctica).Studia Geologica Polonica 104:41�104.

Duane AM. 1994. Preliminary palynological investigion ofthe Byers Group (Late Jurassic-Early Cretaceous), Liv-ingston Island, Antarctic Peninsula. Rev PalaeobotanyPalynology 84:113�120.

Duane AM. 1996. Palynology of the Byers Group (LateJurassic�Early Cretaceous) of Livingston and Snowislands, Antarctic Peninsula: its biostratigraphical andpalaeoenvironmental significance. Rev Palaeobotany Pal-ynology 63:1�11.

Duane AM. 1997. Taxonomic investigations of palyno-morphs from the Byers Group (Upper Jurassic�LowerCretaceous), Livingston and Snow islands, AntarcticPeninsula. Palynology 21:123�144.

Duane AM, Pirrie D, Riding JB. 1992. Palynology ofthe James Ross Island area, Antarctic Peninsula, Numer-ous authors, Special Issue. Antarctic Sci. 4(3):259�362.

Dudziak J. 1984. Cretaceous calcareous nannoplankton fromglaciomarine deposits of the Cape Melville area, KingGeorge Island (South Shetland Islands, Antarctica). Stu-dia Geologica Polonica 79:37�51.

Dutra TL, Batten DJ. 2000. Upper Cretaceous floras of KingGeorge Island, West Antarctica, and their palaeoenviron-mental and phytogeographic implications. CretaceousRes. 21:181�209.

Feakins SJ, Warny S, Lee JE. 2012. Hydrologic cycling overAntarctica during the Middle Miocene warming. NatGeoscience 5:557�560.

Fuenzalida H, Araya R, Herv�e F. 1972. Middle Jurassic florafrom north-eastern Snow Island, South Shetland Islands.In: Adie RJ, editor. Antarctic Geology and Geophysics,Oslo: Universitetsforlaget, p. 93�97.

Gee CT. 1989. Revision of the Late Jurassic/Early Creta-ceous flora from Hope Bay, Antarctica. Palaeontograph-ica, Abteilung B 213:149�214.

Guterch A, Grad M, Janik T, Perku�c E, Pajchel J. 1985. Seis-mic study of the crustal structure in West Antarctica1979�1980�preliminary results. Tectonophysics114:411�429.

Haase KM, Beier C, Fretzdorff S, Smellie JL, Garbe-Sch€onberg D. 2012. Magmatic evolution of the SouthShetland Islands, Antarctica, and implications for conti-nental crust formation. Contrib Mineralogy Petrology163:1103�1119.

Hernandez PJ, Azc�arate V. 1971. Estudio paleobot�anico pre-liminary sobre restos de una tafoflora de la Pen�ınsulaByers (Cerro Negro), Islas Shetland del Sur, Antartica.Serie Cient�ıfica Instituto Ant�artico Chileno 2(1):15�50.

Hunt RJ, Poole I. 2003. Paleogene West Antarctic climateand vegetation history in light of new data from KingGeorge Island. In: Wing SL, Gingerich PD, Schmitz B,Thomas E, editors. Causes and consequences of Globallywarm climates in the Early Paleogene. Boulder, Colo-rado. Geolog Soc Am Spec Pap. 369:395�412.

Hyden G, Tanner PWG. 1981. Late Paleozoic-Early Meso-zoic fore-arc basin sedimentary rocks at the Pacificmargin in western Antarctica. Geologische Rundschau.70(1):529�541.

Kom�arkov�a V, Poncet S, Poncet J. 1985. Two native Antarc-tic vascular plants, Deschampsia antarctica and Coloban-thus quitensis: A new southernmost locality and otherlocalities in the Antarctic Peninsula area. Arctic AlpineRes. 17(4):401�416.

Krajewski KP. 2013. Organic matter�apatite�pyrite rela-tionships in the Middle Triassic Botneheia Formation ofeastern Svalbard: Relevance to the formation of petro-leum source rocks in the NW Barents Sea Shelf. MarinePetroleum Geol. 45:69�105.

Leppe M, Michea W, Mu~noz C, Palma-Heldt S, Fernandoy F.2007. Paleobotany of Livingston Island: The first report ofa Cretaceous fossil flora from Hannah Point. In: Antarc-tica: A Keystone in a Changing World. Online Proceed-ings of the 10th ISAES, edited by Cooper AK, RaymondCR et al., USGS Open-File Report 2007-1047, ShortResearch Paper 081, 4 p.; doi: 10.3133/of2007-1047.srp081.

Loske WP, Miller H, Kramm U. 1988. U-Pb systematics ofdetrital zircons from low-grade metamorphic sandstonesof the Trinity Peninsula group (Antarctica). J S AmEarth Sci. 1(3):301�307.

Machado A, Limaa EF, Chemale F Jr, Moratab D, Oteiza O,Almeida DPM, Figueiredod AMG, Alexandre FM,Urrutia JL. 2005. Geochemistry constraints of Meso-zoic�Cenozoic calc-alkaline magmatism in theSouth Shetland arc, Antarctica. J S Am Earth Sci.18:407�425.

Millar IL, Milne A, Whitham AG. 1990. Implications of Sm-Nd garnet ages for the stratigraphy of northern Grahamland, Antarctic Peninsula. Zentrallblatt f€ur Geologie undPal€aontologie 1:97�104.

Mozer A. 2012. Pre-glacial sedimentary facies of the PointThomas Formation (Eocene) at Cytadela, AdmiraltyBay, King George Island, West Antarctica. Polish PolarRes. 33(1):41�62.

16 S. Warny et al.

Mozer A. 2013. Eocene sedimentary facies in a volcanogenicsuccession on King George Island, South ShetlandIslands: a record of pre-ice sheet terrestrial environmentsin West Antarctica. Geol Q. 57(3):385�394.

Mudie PJ. 1992. Circum-arctic Quaternary and Neogenemarine palynofloras: paleoecology and statistical analy-sis. In: Head MJ, Wrenn JH, editors. Neogene and Qua-ternary dinoflagellate cysts and acritarchs. AmericanAssociation of Stratigraphic Palynologists Foundation,Dallas, p. 347�390.

Nawrocki J, Pa�nczyk M, Williams IS. 2010. Isotopic ages andpalaeomagnetism of selected magmatic rocks from KingGeorge Island (Antarctic Peninsula). J Geol Soc London167:1063�1079.

Nawrocki J, Pa�nczyk M, Williams IS. 2011. Isotopic ages ofselected magmatic rocks from King George Island (WestAntarctica) controlled by magnetostratigraphy. Geol Q.55:301�322.

Pankhurst RJ, Smellie JL. 1983. K�Ar geochronology of theSouth Shetland Islands, Lesser Antarctica: apparent lat-eral migration of Jurassic to Quaternary island arc volca-nism. Earth Planetary Sci Lett. 66:214�222.

Poole I, Cantrill DJ. 2001. Fossil woods from Williams PointBeds, Livingston Island, Antarctica: a Late Cretaceoussouthern high latitude flora. Palaeontology 44:1081�1112.

Poole I, Cantrill D, Utescher T. 2005. A multi-proxyapproach to determine Antarctic Terrestrial palaeocli-mate during the Late Cretaceous and Early Tertiary.Palaeogeography, Palaeoclimatology, Palaeoecology222:95�121.

Poole I, Hunt RJ, Cantrill DJ. 2001. A fossil wood flora fromKing George Island: ecological implications for an Ant-arctic Eocene vegetation. Ann Bot. 88:33�54.

Rees PM, Cleal CJ. 2004. Lower Jurassic floras from HopeBay and Botany Bay, Antarctica. Spec Pap Palaeontol-ogy 72:5�90.

Rees PM, Smellie JL. 1989. Cretaceous angiosperms from anallegedly Triassic flora at Williams Point, LivingstonIsland, South Shetland Islands. Antarctic Sci. 1(3):239�248.

Riding JB, Crame JA. 2002. Aptian to Coniacian (Early�LateCretaceous) palynostratigraphy of the Gustav Group,James Ross Basin, Antarctica. Cretaceous Res. 23(6):739�760.

Riding JB, Crame JA, Dettmann ME, Cantrill DJ. 1998. Theage of the base of the Gustav Group, James Ross Basin,Antarctica. Cretaceous Res. 19:87�105.

Smellie JL, Pankhurst RJ, Thomson MRA, Davies RES.1984. The Geology of the South Shetland Islands: VI.Stratigraphy, geochemistry and evolution. Br AntarcticSurv Scientific Rep. 87:1�85.

Stuchlik L. 1981. Tertiary pollen spectra from the EzcurraInlet Group of Admiralty Bay, King George Island(South Shetland Islands). Studia Geologica Polonica72:109�132.

Thomson MRA. 1975. First marine Triassic fauna from theAntarctic Peninsula. Nature 257:577�578.

Thorn VC, Riding JB, Francis JE. 2009. The Late Cretaceousdinoflagellate cyst Manumiella�biostratigraphy, system-atic and palaeoecological signals in Antarctica. RevPaleobotany Palynology 156:436�448.

Torres T, Barale G, M�eon H, Philippe M, Th�evenard F.1997. Cretaceous floras from Snow Island (South Shet-land Islands, Antarctica) and their biostratigraphic

significance. In: Ricci CA ed. The Antarctic Region: geo-logical evolution and processes, p. 1023�1028

Torres T, M�eon H. 1998. Nothofagidites Erdtman ex Potoni�edans le Pal�eog�ene de l’̂Ile Roi Georges, Antarctique.Geobios. 31(4):419�435.

Troedson AL, Riding JB. 2002. Upper Oligocene to Lower-most Miocene strata of King George Island, South Shet-land Islands, Antarctica: stratigraphy, facies analysis,and implications for the glacial history of the AntarcticPeninsula. J Sedimentary Res. 72(4):510�523.

Troedson AL, Smellie JL. 2002. The Polonez Cove Forma-tion of King George Island, Antarctica: stratigraphy,facies and implications for mid-Cenozoic cryospheredevelopment. Sedimentology 49:277�301.

Trouw RAJ, Pankhurst RJ, Ribeiro A. 1997. On the relationbetween the Scotia Metamorphic Complex and the Trin-ity Peninsula Group, Antarctic Peninsula. In: Ricci CA,editor. The Antarctic region: geological evolution andprocesses, Siena, Terra Antartica, p. 383�390.

Wang F, Zheng XS, Lee JIK, Choe WH, Evans N, Zhu RX.2009. An 40Ar/39Ar geochronology on a mid-Eoceneigneous event on the Barton and Weaver peninsulas:Implications for the dynamic setting of the Antarctic Pen-insula. Geochemistry Geophysiscs Geosystems 10:Q12006. doi: 10.1029/2009GC002874.

Warny S, Askin R. 2011a. Vegetation and organic-walledphytoplankton at the end of the Antarctic greenhouseworld: latest cooling events. AGU Spec Publ.63:193�210.

Warny S. Askin R. 2011b. Last remnants of Cenozoic vegeta-tion and organic-walled phytoplankton in the AntarcticPeninsula’s icehouse world. AGU Spec Publ.63:167�192.

Warny S, Askin R, Hannah M, Mohr B, Raine I, HarwoodDM, Florindo F, and the SMS Science Team. 2009. Paly-nomorphs from a sediment core reveal a sudden remark-ably warm Antarctica during the middle Miocene.Geology 37(10):955�958.

Warny S, Wrenn JH, Bart PJ, Askin RE. 2006. Palynology ofthe NBP03-01A transect in Northern Basin, westernRoss Sea, Antarctica: a Late Pliocene record. Palynology30:151�182.

Whittle RJ, Quaglio F, Griffiths HJ, Linse K, Crame JA.2014. The Early Miocene Cape Melville Formation fossilassemblage and the evolution of modern Antarcticmarine communities. Naturwissenschaften 101:47�59.

Wrenn JH, Hart GF. 1988. Paleogene dinoflagellate cyst bio-stratigraphy of Seymour Island, Antarctica. In: Feld-mann RM, Woodburne MO, editors. Geology andpaleontology of Seymour Island, Antarctic Peninsula:Geological Society of America, Memoir, 169:321�447.

Zastawniak E. 1994. Upper Cretaceous leaf flora from theBlaszyk Moraine [Zamek Fromation], King GeorgeIsland, South Shetland Islands, West Antarctica. ActaPalaeobotanica 34(2):119�163.

Zastawniak E, Szafer W. 1990. Late cretaceous leaf flora ofKing George Island, West Antarctica. In: Proceduressymposium: Palaeofloristic and palaeoclimatic changes inthe Cretaceous and Tertiary, p. 81�85.

Zastawniak E, Wrona R, Gazdzicki A, Birkenmajer K. 1985.Plant remains from the top Part of the Point HennequinGroup (Upper Oligocene), King George Island (SouthShetland Islands, Antarctica). Studia Geologica Polonica81:143�164.

Palynology 17