lsevier.com/locate/yqres

Quaternary Research 64

A 23,000-yr pollen record from Lake Euramoo,

Wet Tropics of NE Queensland, Australia

Simon G. Haberle *

Resource Management in Asia-Pacific Program, Research School of Pacific and Asian Studies,

Australian National University, Canberra, ACT 0200, Australia

Received 17 January 2005

Abstract

A new extended pollen and charcoal record is presented from Lake Euramoo, Wet Tropics World Heritage rainforest of northeast Queensland,

Australia. The 8.4-m sediment core taken from the center of Lake Euramoo incorporates a complete record of vegetation change and fire history

spanning the period from 23,000 cal yr B.P. to present. The pollen record is divided into five significant zones; 23,000–16,800 cal yr B.P., dry

sclerophyll woodland; 16,800–8600 cal yr B.P., wet sclerophyll woodland with marginal rainforest in protected pockets; 8600–5000 cal yr B.P.,

warm temperate rainforest; 5000–70 cal yr B.P., dry subtropical rainforest; 70 cal yr B.P.–A.D 1999, degraded dry subtropical rainforest with

increasing influence of invasive species and fire.

The process of rainforest development appears to be at least partly controlled by orbital forcing (precession), though more local environmental

variables and human activity are also significant factors. This new record provides the opportunity to explore the relationship between fire,

drought and rainforest dynamics in a significant World Heritage rainforest region.

D 2005 University of Washington. All rights reserved.

Keywords: Lake Euramoo; Pollen; Charcoal; Tropical rainforest; Holocene; Atherton Tablelands; Australia

Introduction

In 1970, the first pollen record from the Australian wet

tropical rainforest (Lake Euramoo: Kershaw, 1970) was

published and demonstrated that tropical palynology was

indeed a viable approach to understanding rainforest dynamics

over timescales only imagined in ecological studies. Since

then, sediments from volcanic craters in the Atherton Table-

lands of northeast Queensland have yielded multi-proxy data

sets that have made an important contribution to our

understanding of Australian palaeoclimates and rainforest

vegetation history (Chen, 1988; Moss and Kershaw, 2000;

Kershaw et al., 2002; Turney et al., 2004). One of the most

outstanding attributes of these data is that they present a picture

of a highly dynamic landscape, sensitive to both climate

change and human activity on timescales ranging from

millennia to decades (Haberle et al. in preparation; Turney et

al., 2004).

0033-5894/$ - see front matter D 2005 University of Washington. All rights reserv

doi:10.1016/j.yqres.2005.08.013

* Fax: +61 2 6125 1635.

E-mail address: simon.haberle@anu.edu.au.

The late Quaternary vegetation history of the Atherton

Tablelands is characterized by the expansion of rainforest taxa

during warm and wet interglacial periods and contraction of

these same taxa, to as yet unknown refugial locations, during

cool and dry glacial periods (Kershaw, 1994). The rate at which

these changes take place and the nature of succession from one

vegetation community to another has been examined in greatest

detail over the late glacial transition. This period is represented

in five pollen records from the Atherton Tablelands and shows

that the transition from sclerophyll woodland to rainforest

began as early as 11,500 cal yr B.P., took from 400 to 1000 yr

to complete and was regionally concluded by 7000 cal yr B.P.

(Hiscock and Kershaw, 1992). An increase in effective

precipitation during the late glacial transition is considered

the primary driver of vegetation change (Kershaw et al., 2002),

though local factors including development of organic soils and

fire frequencies (either natural or anthropogenic) may have

contributed to the apparent inhibition of rainforest advance

during the early Holocene.

Rainforest reach their maximum extent on the Atherton

Tablelands by the around 7000 cal yr B.P. under a climate

(2005) 343 – 356

www.e

ed.

S.G. Haberle / Quaternary Research 64 (2005) 343–356344

regime characterized by the establishment of modern day

temperatures, possibly higher precipitation than present (or at

least reduced seasonality in precipitation: Kershaw and Nix,

1988), and low frequency of fires (every 230 yr or so: Chen,

1988). Mid–Late Holocene rainforest composition was far

from stable with high variability evident in rainforest pollen

assemblages recorded during this period (Hiscock and Ker-

shaw, 1992). Around 5000–6000 cal yr B.P. sustained rises in

the emergent conifer Agathis and the early successional taxa

Urticaceae/Moraceae, point to an opening up of the rainforest

canopy perhaps due to reduced precipitation or increased

disturbance. Rainforest disturbance may have had its origin in

natural climate variability (El Nino-related storm and drought

events: Gagan et al., 2004) or through increased intensity of

human activity in rainforest environments (Haberle and David,

2004; Horsfall, 1987).

This paper describes a new core taken from the centre of

Lake Euramoo in 1999, which is part of a study directed at

understanding the origin and role of disturbance within tropical

rainforests through fine resolution palaeoecology. The new

core represents a complete record of sediment accumulation

from the Last Glacial Maximum to the present (post-European

period), which is unlike any other palaeoecological record from

the Atherton Tablelands (Hiscock and Kershaw, 1992).

The site

Lake Euramoo is located near the western edge of the World

Heritage Wet Tropics Bioregion of North East Queensland (718

m above sea level, 17-10V S, 146-38V E) on the Tertiary uplifted

highlands of the Atherton Tablelands (Fig. 1) and within area

that contains Australia’s most significant expanse of tropical

rainforest. Lake Euramoo is a maar described as an ovate

double explosion crater that formed during the Late Pleistocene

period (Kershaw, 1970), with a relatively small catchment area

of ¨4500 m2 and no inflow or outflow channels. The lake is

warm monomictic with a water depth averaging around 20 m in

Figure 1. Map of northeast Queensland showing locat

the northern basin and 16 m in the southern basin, though there

are seasonal fluctuations in water depth of between 2 and 3 m

(Timms, 1976).

An annual average rainfall of 1500 mm is estimated to occur

at Lake Euramoo with a distinctly dry winter rainfall regime as

more than 60% falls between the months of January to March

(Kershaw, 1970). The lake lies at the lower rainfall end of a

steep east to west rainfall gradient. The dominant source of

precipitation is the Southeast Trades interspersed with occa-

sional Northwesterly monsoonal flows and associated tropical

cyclones which also bring high but infrequent rainfall events

during the austral summer months when the intertropical

convergence zone (ITCZ) is at its most southerly extent

(Godfred-Spenning and Reason, 2002). During El Nino

episodes, a northward movement of the ITCZ and a northeast-

ward migration of the South Pacific convergence zone result in

a significant decrease in summer precipitation (typically 150–

300 mm below seasonal average) over the region (Dai and

Wigley, 2000). Mean daily maximum and minimum tempera-

tures are around 25.9- and 14.4-C and frosts occur infrequently

during the austral winter months at times of weak trade winds

and low cloud cover.

The terrestrial vegetation surrounding Lake Euramoo is a

remnant of moist sub-montane rainforest surrounded by

previously cleared land that within the last 50 years has been

either planted with endemic (Araucaria cunninghamii) and

exotic confers or has undergone secondary colonization

towards re-establishing rainforest. Over 100 species of rain-

forest tree and shrub have been recorded within 100 m of Lake

Euramoo (Kershaw personal communication). Typical moist

sub-montane rainforest species found near Lake Euramoo

include members of the Arailiaceae (e.g., Polyscias australi-

ana, Schefflera actinophylla), Araucariaceae (e.g., Agathis

robusta), Moraceae (e.g., Ficus sp.) Elaeocarpaceae (e.g.,

Elaeocarpus grandis), Euphorbiaceae (e.g., Aleutites mollu-

cana, Macaranga sp.), Myrtaceae (e.g., Austromyrtus sp.,

Eugenia cormiflora), and Rubiaceae (e.g., Flindersia bray-

ion of wet tropical rainforests and Lake Euramoo.

S.G. Haberle / Quaternary Research 64 (2005) 343–356 345

leyana, Euodia bonwickii). The nearby (<1 km) drier

vegetation communities include tall open forest (sclerophyll

woodland) dominated by Casuarina sp. and Eucalyptus sp. At

higher altitude, cool lower montane rainforest is increasingly

dominated by members of families such as the Cunoniaceae

(e.g., Caldcluvia australiensis, Schizomeria whitei) Podocar-

paceae (e.g., Podocarpus sp., Prumnopitys sp.) and Rhamna-

ceae (e.g., Alphitonia sp.).

The vegetation around the lake margin is made of

conspicuous zones of aquatic plant communities governed by

water depth and seasonal fluctuations in water level (Kershaw,

1978). Rainforest lianas (e.g., Parsonia sp.) intertwine with tall

swamp grass Phragmites sp. around the shallow lake edge.

Hibiscus sp. and Ludwigia sp. become more common outside

the influence of canopy shade and liana growth in the deeper

(less than 1 m water depth) swamp margin. A zone of rooted

emergent aquatics and floating vegetation mats is characterized

by Cyclosorus gongylodes and Oenanthe sp., with Blechnum

sp. and Elaeocharis equiseteum becoming common on less

stable floating mats. As water depth increases to greater than 1

m, there is a transition from rooted emergent aquatics to

floating aquatics, mainly Nymphoides sp. which can occur as

much as 30 m from the lake shore.

Methods

Fieldwork was conducted in the southern hemisphere

winter of 1999. Parallel cores were taken from the center of

the lake with a modified Livingstone corer with the mud–

water interface section taken with a 1-m plastic piston corer

and extruded at 1-cm intervals on site. Each core was

described in detail and scanned by a magnetic susceptibility

core-scanning loop at 1-cm intervals that identifies changes in

the abundance of ferrimagnetic minerals, which was used as

an indication of changing erosional input to the lake. The

cores were also sampled at 1-cm intervals for loss on ignition

analysis (at 550-C) to identify changes in the amount of

combustible organic matter. Sampling for pollen analysis

ranged from 2- to 8-cm intervals. Chronological control is

provided by using 210Pb dating and AMS 14C analysis of bulk

organic and pollen preparation samples extracted from the

central portion of the sediment core (Table 1, calibrated years

before present, where 0 cal yr B.P. = 1950 A.D, Stuiver et al.,

1998).

Pollen analysis follows the standard KOH, HF, HCl and

acetolysis method described by Bennett and Willis (2001).

Samples from these lake deposits are generally organic-rich

with high pollen concentrations calculated volumetrically with

the addition of an exotic spike (Lycopodium clavatum). Pollen

identification and nomenclature follows regional reference

collections currently held at the School of Geography and

Environmental Science, Monash University and the Depart-

ment of Archaeology and Natural History at the Australian

National University. Ecological information on the taxa

represented in the pollen diagrams can be found in Kershaw

(1970, 1971, 1975, 1983). Additional taxa identified in this

analysis include Melalueca comp. (Myrtaceae; a wet scler-

ophyll woodland tree in poorly drained soils), and Colocasia

(Araceae, herb of poorly drained soils). Pollen counts are

expressed as percentages of the total pollen sum (excluding

pollen of aquatic vascular plants and spores of ferns and fern

allies), which reaches a minimum of 300 in all samples.

Microscopic charcoal particle accumulation rates were

calculated as concentrations of particles 10–125 Am encoun-

tered during pollen counting compared to exotic Lycopodium

spike divided by the sediment accumulation rate. Macroscopic

charcoal particle accumulation rates were calculated as

concentrations of particles >125 Am determined at contiguous

1-cm intervals by treating 4 cm3 sediment with KOH, mild

bleach (6% Potassium Hypochlorite) overnight, sieved

through a 125-Am sieve and counting total particles encoun-

tered under a low-powered (¨10�) stereomicroscope. This

provides basic data on the abundance of charcoal of two

different size fractions, the finer of which is most likely to be

derived from long distance as well as local fires and the

coarser fraction is likely to be derived from more local fire

events.

Numerical zonation, rarefaction and Principle Components

Analysis (PCA) were performed with only major taxa whose

pollen or spore values exceeded 5% at least once. Numerical

zonation employed optimal splitting by sum-of-squares anal-

ysis to partition the stratigraphically constrained fossil pollen

assemblage data into significantly different pollen zones (after

Bennett, 1996). Rarefaction analysis provides an estimate of

the diversity of pollen types or palynological richness of each

sample (Birks and Line, 1992). PCA is used to reduce the

pollen and spore data to a two-dimensional plot and the

resulting data set is displayed as a plot for samples and taxa

(Birks and Gordon, 1985). All numerical analyses have been

implemented within psimpoll, a program for plotting pollen

data, developed by Bennett (1994).

Results and discussion

Lithology, sediment analysis and chronology

A total of 840 cm of lake muds and clays were recovered

from the centre of the northern basin of Lake Euramoo from

1600 to 2440 cm bws (below water surface). The basal core

encountered solid bedrock at a depth of 2440 cm bws.

Overlying the bedrock are sediments composed of grey-green

massive silty clays to 2365 cm bws that are characterized by

low LOI values and peak magnetic susceptibility values (Fig.

2). Between 2365 to 2265 cm bws, the sediments are black lake

muds with occasional bands (<1 cm thick) of light brown

clayey muds decreasing in frequency up-core. Magnetic

susceptibility values show a decreasing trend to minimum

values at around 2320 cm bws. The LOI values increase up

core with a sharp peak (60% organics) from 2320 to 2290 cm

bws, followed by a reversal to lower LOI (30% organics)

between 2290 to 2265 cm bws. Above 2265 cm bws to the mud

water interface at 1600 cm bws, the sediments are uniformly

dark brown lake muds with LOI values around 80–90%. A

zone of coarse detritus and wood is encountered at 2000 to

Table 114C Radiocarbon AMS sample results from the ANSTO radiocarbon lab

ANSTO code Sample Type and ID

(cm below water level)

y(13C) per mil Conventional 14C age Calibrated age, cal yr B.P.

(CALIB v4.4, 2j error)14C yr B.P. 1y error

OZE682 EU 1695–1696 �27.9 570 50 490–570 (80%)

590–630 (20%)

OZE683 EU 1790–1790 �30.1 1320 40 1080–1110 (5%)

1120–1160 (9%)

1170–1290 (86%)

OZE684 EU 1857–1858 �31.8 2410 40 2320–2490 (85%)

2630–2710 (15%)

OZE685 EU 1925–1926 �33.5 3510 40 3630–3840 (100%)

OZE686 EU 2040–2041 �31.7 4090 40 4420–4630 (92%)

4760–4800 (8%)

OZE687 EU 2157–2158 �33.9 6210 50 6900–7210 (99%)

7220–7230 (1%)

OZH310 EU 2201–2202 �31.1 5690 50 6640–6580 (9%)

6570–6400 (86%)

6370–6350 (4%)

6330–6320 (1%)

OZH311 *EU 2219–2220 �31.7 7860 60 8980–8910 (11%)

8900–8880 (4%)

8870–8830 (7%)

8810–8540 (78%)

OZH312 EU 2246–2247 �21.3 9400 80 11,070–10,940 (11%)

10,860–10,830 (2%)

10,810–10,800 (1%)

10,790–10,400 (86%)

OZH313 EU 2255–2256 �25.5 9530 60 11,110–10,660 (99%)

10,620–10,610 (1%)

OZH688 EU 2263–2264 �28.5 9640 60 11,190–11,050 (40%)

11,040–10,780 (60%)

OZH314 *EU 2270–2271 �23.2 9800 90 11,550–11,500 (4%)

11,450–11,390 (3%)

11,360–11,060 (81%)

10,950–10,840 (9%)

10,830–10,760 (3%)

OZH315 *EU 2275–2276 �29.1 10,130 60 12,280–12,220 (3%)

12,130–11,540 (83%)

11,520–11,400 (10%)

11,390–11,340 (4%)

OZH316 Eu 2282–2283 �27.2 10,640 60 12,950–12,600 (76%)

12,500–12,340 (24%)

OZE838 EU 2292–2293 �24.0 11,100 60 12,670–12,720 (3%)

12,870–13,190 (96%)

13,310–13,340 (1%)

OZH317 *EU 2303–2304 �26.6 9110 60 10,470–10,460 (2%)

10,430–10,180 (98%)

OZH318 EU 2315–2316 24.6 11,230 60 13,760–13,700 (3%)

13,460–13,000 (97%)

OZE689 EU 2341–2342 �26.4 12,600 70 14,240–15,540 (100%)

OZH319 EU 2361–2362 �25.0 15,820 120 19,550–18,270 (100%)

OZH320 EU 2389–2390 �24.8 11,310 70 13,310–13,070 (100%)

OZH321 EU 2413–2414 �18.6 18,960 160 22,940–22,170 (100%)

OZE690 EU 2430–2435 �24.1 19,130 90 21,930–23,430 (100%)

All samples are derived from bulk organic detritus with the exception of 4 pollen preparation samples*. Calibration results from CALIB v4.4 (Stuiver and Reimer,

1993). The midpoint of calendar yr range marked in bold are used in age-depth model calculations (3 samples excluded from the age-depth model are not in bold).

S.G. Haberle / Quaternary Research 64 (2005) 343–356346

1940 cm bws and is considered to be debris from a fallen tree

or floating root mat and is excluded from the analysis.

Core chronology is based on a total of 22 AMS 14C dates

on bulk organic detritus and pollen preparation samples

(Table 1, derived from 5-mm-thick sample extractions).

Fourteen sediment samples from the upper 46 cm of the

core were dried overnight at 75-C, finely ground and

analysed for 210Pb by ANSTO (Lucas Heights, Sydney).

Sedimentation rates for the top 46 cm were derived from the

analysis of 210Pb activity using the constant initial concen-

tration (CIC) model (Appleby and Oldfield, 1983) and

produced a basal 210Pb age of A.D 1880 at 1646 cm bws.

The final age-depth model adopted for this study (Fig. 2) is

derived by linear interpolation between a mud–water inter-

Figure 2. Lake Euramoo stratigraphy, sediment description, 14C age yr B.P., water content, loss on ignition and age-depth model.

S.G. Haberle / Quaternary Research 64 (2005) 343–356 347

face age of A.D 1999, the first appearance of exotic pollen

types taxa (Lantana, Mimosa, Pinus and Plantago, details in

Haberle et al. in prep), the 210Pb profile, and the 14C age

determinations. Three AMS 14C dates were excluded from the

age-depth profile on the basis of poorest fit with the linear

interpolation and may be the result of down-core contamina-

tion with younger carbon.

Pollen and charcoal record

The pollen record for Lake Euramoo is presented in Figures

3a–d and is divided into five zones that are described below. A

summary interpretation of each zone is presented in Table 2.

Lake Euramoo 23,000–16,800 cal yr B.P.

(Zone Eu-A1, 2432–2340 cm bws)

Grasses and the sclerophyll woodland taxa, mainly

Casuarina, Myrtaceae (dominated by Melalueca comp.) and

to a lesser extent Callitris dominate the diagram forming

>95% of the pollen sum. Melalueca comp. and the grasses

alternate peak values with Casuarina. There is no significant

presence of rainforest taxa and the palynological richness

reaches a minimum in this zone. Locally, the presence of

Cyperaceae, increasing sharply towards the top of the zone,

and aquatic taxa such as Sparganium, Botryococcus and

Pediastrum suggest that a swampy margin developed around

a small open water body. The Cyperaceous swamp expanded

markedly after about 18,000 cal yr B.P. and included a greater

presence of Lycopodiaceae, Sparganium and the aquatic fern

Selaginella. Fires appear to be infrequent throughout this

zone.

Lake Euramoo 16,800–8700 cal yr B.P.

(Zone Eu-A2, 2340–2220 cm bws)

High values for sclerophyll taxa are maintained though

there is an increase in the importance of Mrytaceous taxa,

including the wet sclerophyll Eucalyptus types. Callitris

increases to its highest values in the record. Grasses and

Compositae decrease with minor increases recorded in Alter-

nathera and Liliaceae between 16,000 and 10,600 cal yr B.P.

The first appearance of 36 rainforest taxa, dominated by

Urticaceae/Moraceae, is recorded in this zone which is

S.G. Haberle / Quaternary Research 64 (2005) 343–356348

reflected in a doubling of the palynological richness by 15,500

cal yr B.P. The rainforest gymnosperms Agathis and Podo-

carpus appear for the first time by 15,500 cal yr B.P. but then

are absent between 12,600 and 9600 cal yr B.P. Rainforest

taxa as a whole make up between 10% and 20% of the pollen

sum between 16,800 and 11,500 cal yr B.P., but then show a

reversal to dry sclerophyll taxa dominance (mainly Casuarina)

between 11,500 and 9600 cal yr B.P. Regional pollen dispersal

from the lowland mangrove swamps is also incorporated into

the Lake Euramoo sediments and may reflect increased

convection or more extensive mangrove forests along the

coastal margins as sea level stabilizes during the early

Holocene (Clark and Guppy, 1988). The Cyperaceae swamp

dominates the lake margin until 15,500 cal yr B.P. but then

becomes a less significant feature of the lake edge vegetation

after this time. Ferns also begin to become important in this

zone. Fire is much more prominent in the landscape after

16,800 cal yr B.P. The fire record is dominated by fine

charcoal particles derived from regional as well as local

(probably grass dominated) fires. By around 11,500 cal yr B.P.

the coarser charcoal fraction begins to increase perhaps in

response to reduced grassy understorey and the development

of a more dominant woody understorey.

Figure 3. Lake Euramoo pollen diagram. (a) Summary pollen diagram with rarefactio

rates (2 size fractions: 10–125 Am and >125 Am). (b) % diagram of herbaceous, scle

<1%). (c) % diagram of minor rainforest taxa included within the total pollen s

palynomorphs excluded from the total pollen sum.

Lake Euramoo 8700–5000 cal yr B.P.

(Zone Eu-A3, 2220–2050 cm bws)

At 8700 cal yr B.P. there is a rapid shift from sclerophyll

to rainforest dominance in the pollen spectrum. Casuarina,

Myrtaceae, Callitris and the grass percentages all fall to low

values, reaching a minimum by 7300 cal yr B.P., as the

total rainforest taxa contribution to the pollen sum rise

sharply to around 80–90%. Urticaceae/Moraceae, Mallotus/

Macaranga, Trema, Cunoniaceae, Elaeocarpus and ferns

(mainly Filicales-psilate type) all show sustained increase in

percentages through this zone. Podocarpus shows an initial

increase and then declines by around 7300 cal yr B.P. A

further 19 rainforest taxa make their first appearance in this

zone, though the palynological richness reaches a high level

at the base of this zone that is maintained for the rest of the

record, reflecting a balance of taxa losses as well as gains

during this transition. The swamp herbs Colocasia, Halor-

agis and an undetermined Umbelliferae are incorporated into

the local cyperaceous swamp. The charcoal record is

dominated by the coarse charcoal fraction through to about

7300 cal yr B.P. possibly reflecting continued burning of the

rainforest-sclerophyll woodland margin during the early

Holocene. Total charcoal reaches a minimum between about

n analysis and principal components analysis results, and charcoal accumulation

rophyll woodland and rainforest taxa included within the total pollen sum (dot =

um (dot = <1%). (d) % diagram of swamp herbs, aquatics, ferns and other

Figure

3(continued).

S.G. Haberle / Quaternary Research 64 (2005) 343–356 349

Figure

3(continued).

S.G. Haberle / Quaternary Research 64 (2005) 343–356350

Figure

3(continued).

S.G. Haberle / Quaternary Research 64 (2005) 343–356 351

Table 2

Interpretation of Lake Euramoo pollen zones and equivalent zones represented in the Lake Euramoo edge core (Kershaw, 1970)

Pollen zone Age cal yr B.P. Interpretation of pollen spectra Equivalent pollen

zone Kershaw 1970

Eu-A1 23,000–16,800 Dry sclerophyll woodland dominated by fire-sensitive Casuarina with a grassy understorey. A sedge

swamp surrounds a small permanent open water body with local Melalueca swamp around the lake

edge margins.

Not evident

Eu-A2 16,800–8700 Wet sclerophyll woodland with marginal rainforest in protected pockets or forming rainforest mosaics

with drier woodland formations. Fire becoming increasingly prevalent with increase in fire-promoting

woody flora.

Zone 1

Eu-A3 8700–5000 Lower montane rainforest with relative dominance by Cunoniaceae and Urticaceae/Moraceae with fire

virtually absent.

Zone 2, 3

Eu-A4 5000–70 Sub-montane rainforest dominated by Elaeocarpaceae, Agathis and Trema with fire increasingly

prevalent.

Zone 4, 5

Eu-A5 AD 1880–1999 Degraded sub-montane rainforest with increasing influence of invasive species and fire. Surface samples only

S.G. Haberle / Quaternary Research 64 (2005) 343–356352

7300 and 6300 cal yr B.P. at the same time as rainforest

taxa attain maximum percentage representation of close to

100% of the pollen sum. The coarse charcoal fraction then

makes a sustained increase after 6300 cal yr B.P. and is

accompanied by rises in Trema a pioneer rainforest taxon.

Lake Euramoo 5000–70 cal yr B.P.

(Zone Eu-A4, 2050–1646 cm bws)

This zone is separated from the previous zone by a sustained

rise in Elaeocarpus, closely associated with slight rises in

Agathis and Mallotus/Macaranga. Cunoniaceae generally

decrease while Urticaceae/Moraceae, Cunoniaceae and Podo-

Figure 4. Principle components analysis of the Lake Euramoo pollen data from the K

paper. The analysis included only taxa with >5% representation common to both rec

1 = 63% and axis 2 = 8%. Sample symbols distinguish pollen zones from each site a

dashed (lake edge core) arrow.

carpus maintain constant percentage values in this zone. Fine

and coarse charcoal particles are present throughout with a

series of peaks in course charcoal deposition recorded from

4000 cal yr B.P. onwards and a rise in accumulation rate

between 2700 and 1200 cal yr B.P. At around 1200 cal yr B.P.,

a further rise in Elaeocarpus is accompanied by slight

increases in sclerophyll taxa and a decrease in Agathis,

Mallotus/Macaranga and Trema. A decrease in Cyperaceae

and an increase in the aquatics, Sparganium, Nymphoides and

Botryococcus, suggest a return to more open water conditions

at this time. The appearance of the swamp fern Thelypter-

idaceae (most likely Cyclosorus gongylodes), a weedy fern

ershaw (1970) lake edge core and the lake centre core taken in 1999 from this

ords (Rut. = Rutaceae/Arailiaceae, Cunon. = Cunoniaceae) with eigenvalue axis

nd the direction of change is indicated by grey solid (lake centre core) and grey

S.G. Haberle / Quaternary Research 64 (2005) 343–356 353

recorded in disturbed swampland across the Pacific (Powell,

1982; Mueller-Dombois and Fosberg, 1998) around 1000 cal yr

B.P. suggests the floating mat habitat may have expanded as

well in the last millennia.

Lake Euramoo A.D 1880–1999

(Zone Eu-A5, 1646–1600 cm bws)

This zone is separated from the preceding zone by the

appearance of exotic species Lantana, Mimosa, Plantago,

Pinus, Solonaceae and Rumex, introduced during the period

of European settlement, beginning in the early A.D 1880s.

Sclerophyll taxa maintain constantly low values whereas

several rainforest taxa show marked declines including

Urticaceae/Moraceae, Elaeocarpus and Agathis, most likely

as a result of removal of the taxa due to selective logging

during the 19th Century (Birtles, 1988). Increases are

recorded in Rutaceae/Araliaceae, Trema, Cunoniaceae, Sloa-

nea and Eugenia. Charcoal particle accumulation rates return

to high values comparable with those recorded in the early

Holocene, though there is little apparent impact on the total

rainforest taxa percentage of the pollen sum.

Comparison between lake edge and lake centre pollen record

The PCA ordination (Fig. 4) incorporates two fossil pollen

datasets from Lake Euramoo. The diagram compares the

extended Lake Euramoo pollen record from the lake centre

(23,000 cal yr B.P. to present) with the lake edge core

collected by Kershaw (1970) that spanned the Holocene from

11,000 to 1500 cal yr B.P. and was also divided into five

pollen zones. Axis 1 contrasts the sclerophyll woodland taxa,

Casuarina and Gramineae, with rainforest taxa and appears to

be related primarily to precipitation. The intermediate position

of Eucalyptus taxa along axis 1 reflects the widespread

association of Eucalyptus with both woodland and rainforest

habitats. Axis 2 explains much less variance though it appears

to distinguish mid–late Holocene dominance of sub-montane

rainforest gymnosperms and secondary rainforest angiosperms

such as Mallotus/Macaranga and Trema, from the early-mid

Holocene lower montane rainforest taxa dominated by

Cunoniaceae and Urticaceae/Moraceae. Axis 2 may relate to

disturbance factors such as increased burning in the mid–late

Holocene or a decrease in dry season precipitation proposed

by Kershaw and Nix (1988) bioclimatic analysis of the Lake

Euramoo edge core record.

The two records are readily comparable through PCA

analysis (Fig. 4, Table 2) though the lake edge core samples

are overall more negative on axis 2 than the lake centre core

samples. One explanation may be that the catchment area

for pollen in the edge core is more sensitive to sub-canopy

and surface wash dispersed pollen including taxa such as

Trema, Mallotus/Macaranga and Podocarpus and the local

swamp Gramineae. The lake centre core is likely to be more

sensitive to pollen wind dispersed from the local emergent

and canopy taxa such as the Cunoniaceae and Urticaceae/

Moraceae, as well as regionally dispersed taxa such as

Eucalyptus.

Discussion

An extended sclerophyll woodland phase at Lake Euramoo

The new pollen record from Lake Euramoo extends the old

record by around 12,000 cal yr B.P. to include the Last Glacial

Maximum. Kershaw (1970: 797) originally hypothesized that

the persistence of sclerophyll woodland in the early Holocene

part of the record was a consequence of recent (early

Holocene) volcanic activity at the site and that the earliest

vegetation then evident from the pollen record represented

pioneer vegetation establishing on immature soils. The new

core from the lake centre clearly refutes the possibility of early

Holocene volcanic activity from the Euramoo crater, placing a

minimum age for crater formation of >23,000 cal yr B.P. There

is no indication in the pollen record that vegetation went

through successional phases of establishment on maturing

basalt soils around the site after 23,000 cal yr B.P. As

sediments accumulated over basalt bedrock the regional

vegetation remains dominated by fire-sensitive Casuarina with

a grassy understorey, though the alternating peaks in Mela-

lueca comp. and the grasses with Casuarina, possibly reflect

millennial-scale shifts in regional soil moisture or salinity

during this period (Crowley, 1994). Swamp vegetation persists

locally throughout this early period.

Pollen and sediment evidence from Lynch’s Crater (Turney

et al., 2004; Kershaw, 1994), sea surface temperature

reconstructions from the Coral Sea (Barrows and Juggins, in

press; Anderson et al., 1989) and fluvial geomorphology from

NE Queensland (Thomas et al., 2001) all point to conditions

on the Atherton Tablelands at this time being much drier (up to

50% reduced precipitation) and cooler (2-–5-C) than today.

The low levels of charcoal preserved in the Lake Euramoo

sediments suggests that the maintenance of the sclerophyll

woodland community around Lake Euramoo was most

strongly associated with the climatic controls of low precip-

itation rather than a high burning frequency. Similar conditions

are reflected in the Lynch’s Crater (zone L2 lower section) and

Strenekoff’s Crater (zone S2 lower sample) record, where

samples closest to the LGM in these records show relatively

high Casuarina–low Eucalyptus pollen percentages accom-

panied by low charcoal densities (Kershaw, 1994). This is at

odds with reconstructions of biomass burning further to the

north in the equatorial highlands of New Guinea where

cumulative charcoal records from wet montane forest/grass-

land settings suggest a high rate of charcoal production

between 20,000 and 10,000 cal yr B.P. The New Guinea

records have been interpreted as resulting from greater

variability in rainfall patterns, during the late glacial period,

enhancing the potential for frequent fires (Haberle et al., 2001).

The combined effect of relatively lower annual precipitation

and the lower biomass of sclerophyll woodland in the Atherton

Tablelands may have been sufficient to reduce charcoal

deposition during the late glacial period in Lake Euramoo

and it is not until increased woodiness, in response to increased

precipitation of the late glacial transition phase, that charcoal

production also increased.

S.G. Haberle / Quaternary Research 64 (2005) 343–356354

Four phases of rainforest development

The transition from sclerophyll woodland to rainforest as

recorded at Lake Euramoo can be considered to have taken

9500 yr, with the initial indication that rainforest was re-

invading the Atherton Tablelands from glacial refugia begin-

ning at 16,800 cal yr B.P. and being completed by 7300 cal yr

B.P. The development of rainforest appears to occur in four

phases.

The first phase begins at 16,800 cal yr B.P., when a limited

suite of rainforest angiosperm and gymnosperm taxa and wet

sclerophyll taxa first appear near the site, possibly as insipient

rainforest patches forming mosaics with wet sclerophyll

woodland. This is almost 4000 yr earlier than similar changes

noted in other Atherton Tablelands records (Walker and Chen,

1987; Hiscock and Kershaw, 1992) and may reflect closer

proximity of Lake Euramoo to glacial rainforest refugia, more

favorable geology (basalts) for early establishment of rain-

forest, or the wet swampy lake margins providing restricted

refugia for rainforest taxa as precipitation and temperature

increased during the late glacial transition. Increased organic

accumulation and biomass burning are associated with this

phase, which supports the suggestion that a rise in woody plant

biomass may have been a factor in rising charcoal particle

accumulation rates in the sediments.

The second phase spans the period from 12,600 to 9600 cal

yr B.P. and is marked by the disappearance of a number of

rainforest taxa from the record, including Agathis and

Podocarpus, and a return to peak representation of Casuarina

in place of wet scleophyll woodland. This may represent a

reversal towards drier climatic conditions, however, rainforest

taxa continue to make initial appearances in the record

suggesting that taxa disappearances may represent shifts in

canopy dominance through competitive advantage rather than a

reversal in climatic conditions necessarily restricting rainforest

advancement into the area.

The third phase occurs between 9600 and 8700 cal yr B.P.

when the maximum rate of increase in rainforest, led by taxa

typical of lower montane rainforest, is recorded at Lake

Euramoo. This is most likely to represent the local establish-

ment of closed canopy rainforest under the influence of

increasing precipitation at the site. This occurs within a 900-

yr period and is comparable to the period of time recorded at

other sites on the Atherton Tablelands where the transition

from sclerophyll woodland to rainforest dominance has been

estimated to have taken between about 400 and 1000 yr on the

Atherton Tablelands (Hiscock and Kershaw, 1992).

The final phase of rainforest development involves the

gradual exclusion of sclerophyll woodland prior to the peak

development of lower montane rainforest at 7300 cal yr B.P.

Local dominance of rainforest was achieved despite the

persistent recurrence of fires that appear to have maintained

local patches of sclerophyll woodland and may have retarded

the encroachment of rainforest onto the site.

The non-synchronous and non-linear nature of the scler-

ophyll to rainforest transition recorded at five sites on the

Tablelands points to local as well as regional environmental

variables influencing the development of rainforest through

time. The apparent contradiction that the most intensive fire

period occurs during the phase of most rapid increase in fire

sensitive rainforest suggests that fire may have become

decoupled from the natural climate regime and was at least

partially controlled by aboriginal fire practices. The motivation

for maintaining open sclerophyll woodlands in an ever

encroaching rainforest landscape may have stemmed from the

need to maintain open habitats that were so prevalent during

the late glacial period for hunting, mobility and to exploit

woodland plant diversity for food procurement (Hill and Baird,

2003; Horsfall, 1987).

Rainforest dynamics during the mid–late Holocene

Around 7300 to 6300 cal yr B.P. rainforest achieved its

maximum extent across the Atherton Tablelands under condi-

tions of higher than present wet season temperatures and higher

dry season precipitation (Kershaw and Nix, 1988). This

effectively would have reduced annual climate seasonality

resulting in low potential for dry-season burning of sclerophyll

woodland and limited moisture loss from rainforest habitats

reduced its vulnerability to fire. Reduced seasonality may have

been partially driven by orbital forcing according to the

Milankovitch theory, which predicts a minimum in Southern

Hemisphere summer insolation from ¨12,000 to 9000 cal yr

B.P., with an increase toward the present (Berger and Loutre,

1991). In the Lake Euramoo record, the maximum rainforest

extent of rainforest occurs at least 2000 yr after the insolation

minimum. One explanation for this apparent lag in vegetation

response to Milankovich forcing lies in the strong influence of

southeast trade winds on precipitation in the Atherton Table-

lands today. In the past precipitation at Lake Euramoo would

have been influenced by the proximity of the coastline. Rising

sea levels during the late-glacial transition flooded the

continental shelf adjacent to the Atherton Tablelands as well

as flooding the Torres Strait, between northern Australia and

New Guinea, with current conditions only being attained by

around 8000 to 7000 cal yr B.P. (Woodroffe et al., 2000). The

combined impact of reduced continentality, increased moisture

convection over the Torres Strait leading to greater southern

penetration of the ITCZ during the southern summer, may have

been the primary driver increasing delivery of rainfall to Lake

Euramoo after 8700 cal yr B.P. and during the mid Holocene

period.

The evidence for a sustained onset of increased burning

after 6300 cal yr B.P. and the appearance for major infrequent

(every 250–1000 yr) peaks of charcoal after 4000 cal yr B.P. is

accompanied by evidence for greater disturbance of the

rainforest, suggesting that the charcoal source was not

restricted to patches of sclerophyll woodland (currently within

1 km from the site) but may indicate fires occurring within the

rainforest resulting in an opening up of the canopy. Similar fire

frequencies are suggested for the Lake Barrine catchment area

during the period of rainforest encroachment (Chen, 1988) and

this may represent a general fire return time for rainforests in

the region. Increased rainforest disturbance may have been

S.G. Haberle / Quaternary Research 64 (2005) 343–356 355

driven by overall drier conditions or increased rainfall

seasonality (Kershaw and Nix, 1988). In addition, the potential

influence of intensified El Nino activity after ¨5000 cal yr B.P.

that has been suggested from eastern (Sandweiss et al., 1996;

Rodbell et al., 1999) and western Pacific (Haberle et al., 2001;

Hayne and Chappell, 2001) proxy records and would likely

have a significant impact on rainforest dynamics in the

Atherton Tablelands region. The shift in the Lake Euramoo

pollen assemblage towards greater representation of short-lived

taxa such as Mallotus/Macaranga, after 5000 cal yr B.P. may

well be indicative of an overall increase in ecosystem turnover

rates under the influence of more frequent El Nino-related

disturbance events. Alternatively, or at least overlaying these

climatic factors there may have been a greater intensity of

occupation and resource management by aboriginal popula-

tions after about 4000 cal yr B.P. facilitated through a

broadening of the range of foodstuffs consumed and the

commencement of toxic plant exploitation (Horsfall, 1987;

Haberle and David, 2004). Destruction of the swamp forest on

Lynch’s Crater at about 5000 cal yr B.P. has also been linked to

aboriginal exploitation and intensified use of fire at this time

(Hiscock and Kershaw, 1992). The slight rise in total charcoal

accumulation rate and increased frequency of major charcoal

peaks between 2500 and 1500 cal yr B.P. may also point to a

period of intensified burning activity near Lake Euramoo. The

relative contribution of climate change and human activity to

these changes remains a complex question that ultimately may

not be resolved if we assume an interdependence of human

response to environmental conditions (Haberle and Chepstow-

Lusty, 2000; Haberle and David, 2004).

The pollen record of the last few hundred years at Lake

Euramoo, missing in the original lake edge core (Kershaw,

1970), details the first appearances of exotic pollen (Lantana,

Mimosa and Pinus/Plantago) and the dramatic rise in charcoal

accumulation rate in the lake sediments provide a clear

stratigraphic marker for European occupation and land use in

the region. The fine-resolution record of the upper 100 cm

shows that the present composition of rainforest in this region

has been significantly altered over the last 120 yr, most notably

evident through the reduction of key taxa in the pollen record,

including Agathis and Podocarpus (Figs. 3b and 4), both

exploited for the timber industry during the early 19th Century

(Birtles, 1988). The most significant changes are recorded at

the onset of European occupation between about A.D 1880 and

1920, through clearance and burning activities. These changes

were rapid and their impact appears to have persisted through

the last century with no sign of recovery to a pre-European

assemblage despite nearly 50 yr of fire management and

rainforest conservation.

Conclusions

The extended pollen record from Lake Euramoo represents

a complete record of vegetation change from the LGM to

present incorporating the transition from sclerophyll woodlands

to rainforest. Detailed analysis of the transition including a

high-resolution contiguous record of charcoal particle accu-

mulation rates suggests that the progression of rainforest

expansion at the beginning of the Holocene was neither

synchronous with climate change nor unidirectional implying

a much more complex model for rainforest development

incorporating external (regional climatic) as well as local

(topographic, anthropogenic) factors. The process of rainforest

expansion takes about 9500 yr, beginning around 16,800 cal yr

B.P. and proceeding through a non-linear four-phase develop-

ment until maximum rainforest extent is achieved by 7300 cal

yr B.P. Fire appears to play a major role in the development of

early rainforest cover, both through general retardation of

rainforest expansion and by providing competitive advantage

to fire-tolerant and pioneer species. Tropical biomass burning

appears to be at least partly controlled by orbital forcing

(precession), though biomass change and human activity are

also significant factors. Shifts towards competitive dominance

of taxa adapted to drought/fire (e.g., Agathis, Mallotus/

Macaranga, Podocarpus and Trema) are most prevalent in

the Lake Euramoo catchment after 5000 cal yr B.P., which may

be a response to the intensification of El Nino-related climate

variability and/or increased human activity within north

Queensland tropical rainforest.

This new record provides the opportunity to explore the

relationship between fire, drought and rainforest dynamics in a

significant World Heritage rainforest region. That vegetation

changes and fire episodes are not always synchronous with

regional climate change through time supports the notion that

climate thresholds and ecological or vegetation thresholds are

not necessarily the same, nor once they are crossed are they

necessarily reversible (Maslin, 2004).

Acknowledgments

The author thanks the editors of this volume Hermann

Behling and Robert Marchant. Thanks also to Peter Kershaw

for encouraging me to re-visit Lake Euramoo and providing

data and advice along the way. Nick Porch, Henk Heijnis,

Sophia Dimitriadis and Ainsley Noakes helped with fieldwork

and Timme Donders made helpful comments on the manu-

script. Permission to core at Lake Euramoo was granted

through the Queensland Parks and Wildlife Service, Queens-

land Government. Radiocarbon and Lead-210 analysis was

completed at the Australian Nuclear Science and Technology

Organisation funded through an AINSE Grant 00/60. The

research was completed while the author was supported by an

Australian Research Council QEII Fellowship at Monash

University and the Australian National University.

References

Anderson, D.M., Prell, W.L., Barratt, N.J., 1989. Estimates of sea surface

temperature in the Coral Sea at the last glacial maximum. Paleoceanography

4, 615–627.

Appleby, P.G., Oldfield, F., 1983. Assessment of 210Pb from sites with varying

sediment accumulation rates. Hydrobiologia 103, 29–35.

Barrows, T.T., Juggins, S., 2005. Sea-surface temperatures around the

Australian margin and Indian Ocean during the Last Glacial Maximum.

Quaternary Science Reviews 24, 1017–1047.

S.G. Haberle / Quaternary Research 64 (2005) 343–356356

Bennett, K.D., 1994. Fpsimpoll_ version 2.23: a C program for analysing

pollen data and plotting pollen diagrams. INQUA Commission for the

study of the Holocene: working group on data-handling methods.

Newsletter 11, 4–6.

Bennett, K.D., 1996. Determination of the number of zones in a biostrati-

graphical sequence. New Phytologist 132, 155–170.

Bennett, K.D.,Willis, K.J., 2001. Pollen. In: Smol, J.P., Birks, H.J.B., Last, W.M.

(Eds.), Tracking Environmental Change Using Lake Sediments. Terrestrial,

Algal, and Siliceous Indicators, vol. 3. Kluwer, Dordrecht, pp. 5–32.

Berger, A., Loutre, M.F., 1991. Insolation values for the climate of the last 10

million of years. Quaternary Science Reviews 10, 297–317.

Birks, H.J.B., Gordon, A.D., 1985. Numerical Methods in Quaternary Pollen

Analysis. Academic Press, London (317p).

Birks, H.J.B., Line, J.M., 1992. The use of rarefaction analysis for estimating

palynological richness from Quaternary pollen-analytical data. The Holo-

cene 2, 1–10.

Birtles, T., 1988. European interpretation of the Atherton-Evelyn vine scrub of

tropical north Queensland, 1880–1920. In: Dargavel, J., Dixon, K.,

Semple, N. (Eds.), Changing Tropical Forests: Historical Perspectives on

Today’s Challenges in Asia, Australasia and Oceania. Centre for Resource

and Environmental Studies Publication, Australian National University,

Canberra, pp. 197–216.

Chen, Y., 1988. Early Holocene population expansion of some rainforest

trees at Lake Barrine basin, Queensland. Australian Journal of Ecology

13, 225–233.

Clark, R.L., Guppy, J.C., 1988. A transition from mangrove forest to freshwater

wetland in monsoon tropics of Australia. Journal of Biogeography 15,

665–684.

Crowley, G.M., 1994. Quaternary soil salinity events and Australian vegetation

history. Quaternary Science Reviews 13, 15–22.

Dai, A., Wigley, T.M.L., 2000. Global patterns of ENSO-induced precipitation.

Geophysical Research Letters 27, 1283–1286.

Gagan, M.K., Hendy, E.J., Haberle, S.G., Hantoro, W.S., 2004. Post-glacial

evolution of the Indo-Pacific Warm Pool and El Nino-Southern Oscillation.

Quaternary International 118–119, 127–143.

Godfred-Spenning, C.R., Reason, C.J.C., 2002. Interannual variability of

lower-tropospheric moisture transport during the Australian Monsoon.

International Journal of Climatology 22, 509–532.

Haberle, S.G., Chepstow-Lusty, A., 2000. Can climate influence cultural

development?: a view through time. Environment and History 6, 349–369.

Haberle, S.G., David, B., 2004. Climates of change: human dimensions of

Holocene environmental change in low latitudes of the PEPII transect.

Quaternary International 118–119, 165–179.

Haberle, S.G., Hope, G.S., van der Kaars, W.A., 2001. Biomass burning in

Indonesia and Papua New Guinea: natural and human induced fire events in

the fossil record. Palaeogeography, Palaeoclimatology, Palaeoecology 171,

259–268.

Hayne, M., Chappell, J., 2001. Cyclone frequency during the last 5000 years at

Curacoa Island, north Queensland, Australia. Palaeogeography, Palaeocli-

matology, Palaeoecology 168, 207–219.

Hill, R., Baird, A., 2003. Kuku-Yalanji rainforest aboriginal people and

carbohydrate resource management in the Wet Tropics of Queensland,

Australia. Human Ecology 31, 27–52.

Hiscock, P., Kershaw, A.P., 1992. Palaeoenvironments and prehistory of

Australia’s tropical Top End. In: Dodson, J. (Ed.), The Niave Lands:

Prehistory and Environmental Change in Australia and the Southwest

Pacific. Longman Cheshire, Melbourne, pp. 43–75.

Horsfall, N., 1987. Aborigines and toxic north-eastern Queensland rainforest

plants. In: Covacevich, J., Davie, P., Pearn, J. (Eds.), Toxic Plants and

Animals: A Guide for Australia. QueenslandMuseum, Brisbane, pp. 57–63.

Kershaw, A.P., 1970. A pollen diagram from Lake Euramoo, North-East

Queensland, Australia. New Phytologist 69, 785–805.

Kershaw, A.P., 1971. A pollen record from Quincan Crater, north-east

Queensland, Australia. New Phytologist 70, 669–805.

Kershaw, A.P., 1975. Stratigraphy and pollen analysis of Bromfield Swamp,

north-eastern Queensland, Australia. New Phytologist 75, 173–191.

Kershaw, A.P., 1978. The analysis of aquatic vegetation on the Atherton

Tablelands, north-east Queensland, Australia. Australian Journal of Ecology

3, 23–42.

Kershaw, A.P., 1983. A Holocene pollen record from Lynch’s Crater, north-

eastern Queensland, Australia. New Phytologist 94, 669–682.

Kershaw, A.P., 1994. Pleistocene vegetation of the humid tropics of

northeastern Queensland, Australia. Palaeogeography, Palaeoclimatology,

Palaeoecology 109, 399–412.

Kershaw, A.P., Nix, H.A., 1988. Quantitative palaeoclimatic estimates from

pollen using bioclimatic profiles of extant taxa. Journal of Biogeography

15, 589–602.

Kershaw, A.P., van der Kaars, S., Moss, P.T., Wang, X., 2002. Quaternary

records of vegetation, biomass burning, climate and possible human impact

in the Indonesian-Northern Australian region. In: Kershaw, A.P., David, B.,

Tapper, N.J., Penny, D., Brown, J. (Eds.), Bridging Wallace’s Line: The

Environmental and Cultural History and Dynamics of the Southeast Asian-

Australian Region. Catena Verlag, Reiskirchen, Germany, pp. 97–118.

Maslin, M., 2004. Ecological versus climatic thresholds. Science 306,

2197–2198.

Moss, P.T., Kershaw, A.P., 2000. The last glacial cycle from the humid tropics

of Australia: comparison of a terrestrial and marine record. Palaeogeo-

graphy, Palaeoclimatology, Palaeoecology 155, 155–176.

Mueller-Dombois, D., Fosberg, F.R., 1998. Vegetation of the Tropical Pacific

Islands. Springer, New York.

Powell, J.M., 1982. The history of plant use and man’s impact on the

vegetation. In: Gressitt, J.L. (Ed.), Biogeography and Ecology of New

Guinea. Junk, The Hague, pp. 207–227.

Rodbell, D.T., Seltzer, G.O., Anderson, D.M., Abbott, M.B., Enfield, D.B.,

Newman, J.H., 1999. An¨15,000-year record of El Nino-driven alluviation

in southwestern Ecuador. Science 283, 516–520.

Sandweiss, D.H., Richardson III, J.B., Reitz, E.J., Rollins, H.B., Maasch, K.A.,

1996. Geoarchaeological evidence from Peru for a 5000 years B.P. onset of

El Nino. Science 273, 1531–1533.

Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised CALIB

radiocarbon calibration program. Radiocarbon 35, 215–230.

Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Burr, G.S., Hughen, K.A.,

Kromer, B., McCormac, F.G., van der Plicht, J., Spurk, M., 1998.

INTCAL98 Radiocarbon age calibration 24,000-0 cal BP. Radiocarbon

40, 1041–1083.

Thomas, M.F., Nott, J., Price, D.M., 2001. Late Quaternary stream sedimen-

tation in the humid tropics: a review with new data from NE Queensland,

Australia. Geomorphology 39, 53–68.

Timms, B.V., 1976. Morphology of Lakes Barrine, Eacham and Euramoo,

Atherton Tableland, North Queensland. Proceedings of the Royal Society of

Queensland 87, 81–84.

Turney, C.S.M., Kershaw, A.P., Clemens, S.C., Branch, N., Moss, P.T., Fifield,

L.K., 2004. Millennial and orbital variations of El Nino/Southern

Oscillation and high-latitude climate in the last glacial period. Nature 428,

306–310.

Walker, D., Chen, Y., 1987. Palynological light on tropical rainforest dynamics.

Quaternary Science Reviews 6, 77–92.

Woodroffe, C.D., Kennedy, D.M., Hopley, D., Rasmussen, C., Smithers,

S.G., 2000. Holocene reef growth in Torres Strait. Marine Geology 170,

331–346.