Fire regime and vegetation change in the transitionfrom Aboriginal to European land management in a Tasmanianeucalypt savanna

Louise M RomaninAC Feli HopfB Simon G HaberleB and David M J S BowmanA

ASchool of Biological Sciences University of Tasmania Private Bag 55 Hobart Tas 7001 AustraliaBANU College of Asia and the Pacific Australian National University Canberra ACT 0200 AustraliaCCorresponding author Email LouiseRomaninutaseduau

Abstract Using pollen and charcoal analysis we examined how vegetation and fire regimes have changed over the last600 years in the Midlands of Tasmania Sediment cores from seven lagoons were sampled with a chronology developed atone site (Diprose Lagoon) using 210Pb and 14C dating Statistical contrasts of six cores where Pinus served as a marker ofEuropean settlement in the early 19th Century and showed significant changes in pollen composition following settlementwith (a) influx of ruderal exotic taxa including Plantago lanceolata L Brassicaceae Asteraceae (Liguliflorae) and Rumex(b) increase in pollen of the aquatics Myriophyllum spp and Cyperaceae (c) a decline in native herbaceous pollen taxaincluding Chenopodiaceae and Asteraceae (Tubuliflorae) and (d) a decline in Allocasuarina and an initial decline and thenincrease of Poaceae The presence of Asteraceae (Liguliflorae) in the pre-European period suggests that an important rootvegetable Microseris lanceolata (Walp) SchBip may have been abundant Charcoal deposition was low in the pre-European period and significantly increased immediately after European arrival Collectively these changes suggestsubstantial ecological impacts following European settlement including cessation of Aboriginal traditions of firemanagement a shift in hydrological conditions from open water lagoons to more ephemeral herb covered lagoons andincreased diversity of alien herbaceous species following pasture establishment

Additional keywords Aboriginal fire management grassland grassy woodland historical ecology landscape ecologymacro-charcoal palynology

Received 24 February 2016 accepted 11 July 2016 published online 2 August 2016

Introduction

The transition from Aboriginal to European management is acritical period in the history of Australian ecosystems Europeancolonisation of Australia which commenced in 1788 resulted inthe introduction of new plants animals and microorganismssome of which initiated a wave of native plant and animalextinctions (MacPhee and Flemming 1999 Woinarski et al2015) Aboriginal people were removed from their traditionallands causing the loss of ancient socio-ecological systems andnative vegetation was modified for farmland urban areas andmines (Fensham 1989 Lunt 1991) Particularly hard hit weretemperate environments especially those dominated by nativegrasslands and woodlands with a grassy understorey thatcollectively form a savanna biome These regions were idealfor early pastoral settlement which lead to rapid changes inland management (Benson and Redpath 1997) The oldestagricultural landscapes in Australia are woodlands on theCumberland Plain to the west of Sydney settled in 1791 andthe Midlands of Tasmania settled from the early 1820s (Phillip1892Morgan 1992) The conversion of temperate grassland and

woodland ecosystems which we collectively describe aslsquotemperate savannarsquo has been almost complete with less than20 of these ecosystems remaining on mainland Australia andTasmania (Kirkpatrick et al 1988 Fischer et al 2009)

There are sparse data on Aboriginal land use practices intemperate savannas given that these ecosystems were impactedso early in the colonial historical of Australia Archaeologicalevidence has illuminated patterns of Aboriginal land occupancyand there are early colonial reports of Aboriginals burning thelandscape however the ecological consequences and frequencyof burning is unclear (Kee 1990 Brown 1991) Aboriginal landmanagement continues with relatively little disruption intropical Australian savannas but there is uncertainty about theapplicability of these contemporary management practices in totemperate Australian systems (Bowman 1998 Russell-Smithet al 2013) For these reasons prehistorical reconstructions ofAboriginal land management regimes are of prime importancein these temperate ecosystems

Broadly there are two different approaches to reconstructionof past environments environmental history stemming from

CSIRO PUBLISHING

Australian Journal of Botanyhttpdxdoiorg101071BT16032

Journal compilation CSIRO 2016 wwwpublishcsiroaujournalsajb

the humanities and historical ecology from the sciences(Bowman 2002) The former relies on interpretation ofhistorical documents and images (Ryan et al 1995 Bowman2002 Butzer and Helgren 2005 Mactaggart et al 2007Gammage 2008) and the latter environmental reconstructionsusing a range of proxies the most important beingdendrochronology (Banks 1982 von Platen et al 2011) andpollen and charcoal in sediments (Swetnam et al 1999Whitlock and Larsen 2001 Turner and Plater 2004 Higueraet al 2005 Black et al 2007) These approaches are not mutuallyexclusive indeed sometimes they motivate and inform eachother and can be effectively merged to provide more robustreconstructions of vegetation cover (Fensham 1989 Batek et al1999 Benson and Howell 2002 Lunt 2002)

In temperate Australia there is a suite of direct scientificanalysis of the transition from Aboriginal to Europeanmanagement (summarised by Dodson and Mooney 2002) butin temperate savanna landscapes there has been a far greaterreliance on the analysis and interpretation of historical records(Fensham 1989 Thomas 1994 Griffiths 2002 Lunt 2002Gammage 2008 2011) These qualitative studies have beeninfluential on contemporary management practices especiallytree clearing in rural landscapes and bushfire management(Rolls 1981 Benson and Redpath 1997 Jurskis 2000Mooney et al 2007 Attiwill and Adams 2013) A particularlyprominent view is that the amount of landscape burning declinedabruptly following colonisation resulting in a thickening oflandscapes that were formerly much more open (Rolls 1981Jurskis 2000 Franco and Morgan 2007 Gammage 2011)Despite their popular appeal there has been limitedindependent verification using robust environmental proxiesleading to ongoing debate between scientists and historiansabout landscape history (Benson and Redpath 1997 Flannery1997 Jurskis 2000 Bowman 2001)

The Midlands of Tasmania are important in discussionsabout the nature of landscape changes that followed Europeansettlement The temperate savannas which supported highdensities of kangaroos and an endemic and now extinct emuare assumed to have been maintained by frequent landscapeburning (Jones 1969 Duncan 1990 Bowman 1998 Jackson1999 Gammage 2008) a practice that is likely to reach back tothe beginning of Aboriginal colonisation some 35 000 years ago(Colhoun and Shimeld 2012 OrsquoConnell and Allen 2015) Theopen vegetation proved ideal for sheep grazing so the grassylowlands of the Midlands were rapidly allocated to free settlersand by 1830 Aboriginal people were completely removed fromthe area (Fensham 1989 Morgan 1992 Benson and Redpath1997) This transition resulted in tree clearance establishmentof exotic taxa and the transformation of a focal region of theindigenous economy (Ryan 2012 Prior et al 2013 Romaninet al 2015)

There is a diverse range of evidence to support the hypothesisthat Aboriginal people used fire to create or maintain openlandscapes in wetter regions of Tasmania (Ellis and Thomas1988 Jackson 1999 Fletcher and Thomas 2010 Wood andBowman 2012 Bowman et al 2013) However there is lessevidence to support this hypothesis in the Midlands wherethere has been a great reliance on the veracity of early colonialpaintings that depict vast savanna landscapes with open crown

eucalypts (Duncan 1990 Kirkpatrick 2007 Gammage 2008)The accuracy of these paintings is questionable because theywereexecuted with different motivations Some artists were employedby the government to create immigration propaganda which waspicturesque and had a strong resemblance to English landscapessome were trained as surveyors and produced highly accuratelandscape representations still others were artistically trained soused artistic licence to create aesthetically-pleasing scenes(Fensham 1989 Farag-Miller et al 2013) The landscapes ofGlover are often referred to as accurate pre-European depictionsof the Midlands his works were criticised by contemporariesfor their lsquohideousfidelity to naturersquo (Smith 1960 Fensham1989)Nonetheless he portrayed groups of Aboriginal people inlandscapes free of European disturbance despite them beingexpatriated years previously and independent geographicvalidation using surveying techniques demonstrates thatGloverrsquos landscape features are of low topographic fidelity(Farag-Miller et al 2013) Considering these factors it isreasonable question his artworksrsquo ecological and botanicalfidelity as well

The scientific evaluation of the environmental changes inthe Midlands is limited to a dendrochronological study in thedry sclerophyll forest of the eastern tiers (von Platen et al 2011)preliminary palaeoecological examinations of the long-termpollen and charcoal record in a Midlands lagoon (Sigleo andColhoun 1981) an unpublished thesis looking at the pollenand charcoal record of an additional Midlands lagoon (Jones2008) and inferences from current demographic structure andopen branching habit of old woodland trees (Duncan 1990Kirkpatrick 2007 Gammage 2008) These studies togethersuggest that Aboriginal fire was infrequent or low in severity(vonPlaten et al 2011) and the dominant vegetationwas eucalyptsavanna with a substantial Allocasuarina component (Sigleoand Colhoun 1981 Jones 2008) before European settlementImmediately after Europeans arrived fire frequency as recordedin fire scars decreased dramatically but increased again infrequency between 1850 and 1980 after 1989 there were fewfires recorded (von Platen et al 2011) The single analysis ofmacro-charcoal in the Midlands showed sparse levels in boththe pre- and post-European periods however micro-charcoalseems to have been more abundant pre-settlement (Jones2008) Pictorial evidence used by Duncan (1990) andGammage (2008) suggest that the elimination of Aboriginalland management resulted in increased density of trees inuncleared fragments with widely spaced wide-canopied treesbeing replaced by narrow crowned pole saplings and youngertrees

In this study we sampled sediments from seven lagoons andused pollen and charcoal analysis to examine how vegetationand fire regimes at a landscape-level have changed followingEuropean colonisation of the temperate savannas of theMidlandsof Tasmania This is the most extensive sampling of this regionto date According to previous research we expect to see a sharpdecrease in fire activity post settlement and then a subsequentincrease Trends in the floristic composition of post settlementvegetation are expected to include a mixed signal of increasedruderals and (pasture) grasses and non-native trees and a loss ofnative species particularly herbs and fire sensitive woody taxaincluding Allocasuarina

B Australian Journal of Botany L M Romanin et al

Materials and methods

Geographic context

Our study locus was the Midlands of Tasmania ndash a region that ischaracterised by fertile river valleys and lateritic plains lyingbetween the Central Plateau to the west and the Eastern Tiersto the east (Fensham and Kirkpatrick 1989 Doyle 1993) Theregion is in a rain shadow (Fig 1a) and is the driest part of theisland with rainfall between 397 and 626mm annually evenlydistributed throughout the year (Fig 1b) Average maximumtemperatures range between 9 and 11C inwinter and 21 and 24Cin summer The regional climate is classified as temperate withwarm summer according to Koumlppen-Geiger classification (Peelet al 2007) Frosts can occur during any month and the averagenumber of frost days across the region is between 75 and 150per year (Bureau of Meteorology 2008)

The original vegetation of the area based on historicalreconstructions was a mosaic of lowland grasslands and grassyeucalypt woodland dominated by Eucalyptus viminalis LabillEucalyptus pauciflora Sieber ex Spreng and Eucalyptusamygdalina Labill with understorey trees such asAllocasuarina verticillata (Lam) LASJohnson Acaciadealbata Link and Banksia marginata Cav and a tussockgrassland dominated by Poa labillardieri Steud and Themedatriandra Forssk (Fig 1c) (Fensham 1989 Fensham andKirkpatrick 1989) Approximately 37 of the region remainsunder natural vegetation most of which is highly fragmented(Norton and Lacey 2012) due to extensive clearing for pastureand crops (Fig 1d) that occurred over the last ~200 years ofEuropean settlement Small wetlands with varying levels ofsalinity occur throughout the region (David and Browne 1950Buckney and Tyler 1976 Kirkpatrick and Harwood 1983)

High 1602 m

Low ndash155 m

600 mm800 mm1000 mm1200 mm1600 mm

Annual rainfall

Launceston

Campbell Town

Ross

Launceston

Hobart

Launceston

Campbell Town

Ross

0 125 25 km

(b)(a)

(c) (d )

1788 current

41degS

42degS

143degE 144degE 145degE 146degE 147degE 148degE 149degE

Northern Midlands

Elevation

Sample sites

Diprose Lagoon

Towns

Major rivers

Vegetation typeurban

cleared

wetland

grassland

scrub

woodland

forest

N

Fig 1 Geographic context of the study area showing (a) elevation across Tasmania with the study area outlined in red (b) annual rainfall across the studyarea which ranges from 600 to 1000mm annually (c) a reconstruction of 1788 vegetation patterns (adapted from Fensham 1989) and (d) current vegetationpatterns (adapted from TASVEG 20 Department of Primary Industries and Water 2009) with locations of sampling sites marked as black crosses in both the1788 and current vegetation maps The core from Diprose Lagoon that was 210Pb and 14C dated to obtain calibrated chronology is circled in red

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany C

supporting sedge and rushlands dominated by a suite ofgenera including Juncus Baumea Carex Lepidosperma andPhragmites and herblands with ephemeral aquatic plants such asMyriophyllum spp and Triglochin spp where open water ispresent The most saline wetlands support salt marshcommunities including Chenopodium spp (Kirkpatrick andGlasby 1981)

Sediment samplingSediment cores were collected from seven lagoons along a~40 km long northndashsouth transect spanning the study area(Fig 1c d) The lagoons varied in condition and aquaticvegetation therefore a variety of methods were used to collectsediment samples sample resolution was necessarily low at onesite due to the dryness of the sediments (Table 1) A d-sectioncorer was used where sediments were soft enough for corerpenetration A hammer-driven polycarbonate piston corer wasused where sediments were dominated by stiff clays All coreswere wrapped in plastic and stored in PVC tubes They werekept cool and subsequently subsampled for macro-charcoal andpollen analysis

Charcoal and pollen analysisMacro-charcoal particles (gt125mm) are likely to be derivedfrom local fire events (Whitlock and Larsen 2001 Duffin et al2008) To prepare samples for macro-charcoal analysis asubsample of 125 cm3 of material from each contiguous05 cm increment was extracted Subsamples with high claycontent were deflocculated in Calgon solution for 2ndash3 daysafter which samples were rinsed gently with water Charcoalsubsamples were then treated with bleach for at least 24 h toremove the pigment from organic matter and aid in identifyingcharcoal (Black et al 2007) Bleached samples were gentlywashed through a 125mm sieve to separate macro-charcoal

from organic matter and finer particles Total macro-charcoalparticles were counted under a low power stereo-microscope

Broad vegetation patterns over time were examined throughpollen analysis Sediment recordswere sampled at 2 cm intervals125 cm3 subsamples of these were analysed at AustralianNational University (ANU) using standard preparation andcounting techniques by Bennett and Willis (2001) Pollenidentification and nomenclature follows species lists includedby Zacharek et al (1997) plant surveys at Tom Gibson NatureReserve (Parks and Wildlife Service Tasmania 2006) andreference collections held at ANU (httpapsaanueduauaccessed 5 May 2014) Eucalyptus and Allocasuarina typepollen is assigned based on current vegetation patternsTerrestrial or dryland pollen counts (including fern spores butexcluding aquatic taxa) were expressed as percent pollen of eachsample and used in statistical analyses where (a) taxa thatoccurred in less than four samples per core or made up lessthan 1 of the total were excluded from analysis (n= 17) (b) allRumex types were treated as one taxon as distinguishing betweennative and exotic Rumex taxa that coexist in Tasmania isproblematic (c) indeterminate pollen grains were alsoexcluded from analysis

Sediment chronologyDiprose Lagoon was selected for 210Pb and 14C dating atAustralian Nuclear Science and Technology Organisation(ANSTO) (Lucas Heights NSW Australia) The upper 12 cmof the core was analysed for 210Pb activity and the constant rateof supply (CRS) model was used to develop a chronology forDiprose Lagoon (Oldfield and Appleby 1984) The CRS modelassumes a constant flux of unsupported 210Pb to the surfacesediments through time (Appleby and Oldfield 1978 Binford1990) This model is suitable for many temperate-zone lakeswhere 210Pb activity has been diluted by recent high sedimentaccumulation rates (Binford 1990) and is appropriate in this

Table 1 Sampling details of the seven lagoons in the Midlands of TasmaniaThe location and landscape setting sampling methods employed for extracting sediment and their total length of each core sampling resolution is summarised

for each site The first occurrence of Pinus is also noted

Lagoon name Location Landscape setting Sampling method Lengthof core

Sampling resolution Pinus1st record

Diprose Lagoon 41480252600S147220162600E

Agricultural land Eucalyptusamygdalina woodland

Universal coresampler

215 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Smiths Lagoon 41490229500S147260116400E

Eucalyptus amygdalinawoodland

Shovel excavation 14 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Near Lagoon 42 40118500S147260158300E

Themeda and Poa grasslandsEucalyptus viminalis grassywoodlands and Eucalyptuspauciflora woodland

Auger 36 cm 2 cm (pollen)2 cm (charcoal)

22 cm

White Lagoon 42 50559200S147260551200E

Themeda triandra grasslandE viminalis woodland

Universal coresampler

26 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Bells Lagoon 42 50102100S147210148900E

Agricultural land nativelowland grassland complexE viminalis woodland

Universal coresampler

28 cm 2 cm (pollen)05 cm (charcoal)

4 cm

Township Lagoon 42 80501300S1472503600E

Agricultural land nativelowland grassland complex

Universal coresampler

34 cm 2 cm (pollen)05 cm (charcoal)

26 cm

Cleveland Lagoon 41480127100S147240381600E

Agricultural landE amygdalina andE viminalis woodland

Shovel excavation 21 cm 2 cm (pollen)05 cm (charcoal)

Throughout

D Australian Journal of Botany L M Romanin et al

Materials and methods

Geographic context

Our study locus was the Midlands of Tasmania ndash a region that ischaracterised by fertile river valleys and lateritic plains lyingbetween the Central Plateau to the west and the Eastern Tiersto the east (Fensham and Kirkpatrick 1989 Doyle 1993) Theregion is in a rain shadow (Fig 1a) and is the driest part of theisland with rainfall between 397 and 626mm annually evenlydistributed throughout the year (Fig 1b) Average maximumtemperatures range between 9 and 11C inwinter and 21 and 24Cin summer The regional climate is classified as temperate withwarm summer according to Koumlppen-Geiger classification (Peelet al 2007) Frosts can occur during any month and the averagenumber of frost days across the region is between 75 and 150per year (Bureau of Meteorology 2008)

The original vegetation of the area based on historicalreconstructions was a mosaic of lowland grasslands and grassyeucalypt woodland dominated by Eucalyptus viminalis LabillEucalyptus pauciflora Sieber ex Spreng and Eucalyptusamygdalina Labill with understorey trees such asAllocasuarina verticillata (Lam) LASJohnson Acaciadealbata Link and Banksia marginata Cav and a tussockgrassland dominated by Poa labillardieri Steud and Themedatriandra Forssk (Fig 1c) (Fensham 1989 Fensham andKirkpatrick 1989) Approximately 37 of the region remainsunder natural vegetation most of which is highly fragmented(Norton and Lacey 2012) due to extensive clearing for pastureand crops (Fig 1d) that occurred over the last ~200 years ofEuropean settlement Small wetlands with varying levels ofsalinity occur throughout the region (David and Browne 1950Buckney and Tyler 1976 Kirkpatrick and Harwood 1983)

High 1602 m

Low ndash155 m

600 mm800 mm1000 mm1200 mm1600 mm

Annual rainfall

Launceston

Campbell Town

Ross

Launceston

Hobart

Launceston

Campbell Town

Ross

0 125 25 km

(b)(a)

(c) (d )

1788 current

41degS

42degS

143degE 144degE 145degE 146degE 147degE 148degE 149degE

Northern Midlands

Elevation

Sample sites

Diprose Lagoon

Towns

Major rivers

Vegetation typeurban

cleared

wetland

grassland

scrub

woodland

forest

N

Fig 1 Geographic context of the study area showing (a) elevation across Tasmania with the study area outlined in red (b) annual rainfall across the studyarea which ranges from 600 to 1000mm annually (c) a reconstruction of 1788 vegetation patterns (adapted from Fensham 1989) and (d) current vegetationpatterns (adapted from TASVEG 20 Department of Primary Industries and Water 2009) with locations of sampling sites marked as black crosses in both the1788 and current vegetation maps The core from Diprose Lagoon that was 210Pb and 14C dated to obtain calibrated chronology is circled in red

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany C

supporting sedge and rushlands dominated by a suite ofgenera including Juncus Baumea Carex Lepidosperma andPhragmites and herblands with ephemeral aquatic plants such asMyriophyllum spp and Triglochin spp where open water ispresent The most saline wetlands support salt marshcommunities including Chenopodium spp (Kirkpatrick andGlasby 1981)

Sediment samplingSediment cores were collected from seven lagoons along a~40 km long northndashsouth transect spanning the study area(Fig 1c d) The lagoons varied in condition and aquaticvegetation therefore a variety of methods were used to collectsediment samples sample resolution was necessarily low at onesite due to the dryness of the sediments (Table 1) A d-sectioncorer was used where sediments were soft enough for corerpenetration A hammer-driven polycarbonate piston corer wasused where sediments were dominated by stiff clays All coreswere wrapped in plastic and stored in PVC tubes They werekept cool and subsequently subsampled for macro-charcoal andpollen analysis

Charcoal and pollen analysisMacro-charcoal particles (gt125mm) are likely to be derivedfrom local fire events (Whitlock and Larsen 2001 Duffin et al2008) To prepare samples for macro-charcoal analysis asubsample of 125 cm3 of material from each contiguous05 cm increment was extracted Subsamples with high claycontent were deflocculated in Calgon solution for 2ndash3 daysafter which samples were rinsed gently with water Charcoalsubsamples were then treated with bleach for at least 24 h toremove the pigment from organic matter and aid in identifyingcharcoal (Black et al 2007) Bleached samples were gentlywashed through a 125mm sieve to separate macro-charcoal

from organic matter and finer particles Total macro-charcoalparticles were counted under a low power stereo-microscope

Broad vegetation patterns over time were examined throughpollen analysis Sediment recordswere sampled at 2 cm intervals125 cm3 subsamples of these were analysed at AustralianNational University (ANU) using standard preparation andcounting techniques by Bennett and Willis (2001) Pollenidentification and nomenclature follows species lists includedby Zacharek et al (1997) plant surveys at Tom Gibson NatureReserve (Parks and Wildlife Service Tasmania 2006) andreference collections held at ANU (httpapsaanueduauaccessed 5 May 2014) Eucalyptus and Allocasuarina typepollen is assigned based on current vegetation patternsTerrestrial or dryland pollen counts (including fern spores butexcluding aquatic taxa) were expressed as percent pollen of eachsample and used in statistical analyses where (a) taxa thatoccurred in less than four samples per core or made up lessthan 1 of the total were excluded from analysis (n= 17) (b) allRumex types were treated as one taxon as distinguishing betweennative and exotic Rumex taxa that coexist in Tasmania isproblematic (c) indeterminate pollen grains were alsoexcluded from analysis

Sediment chronologyDiprose Lagoon was selected for 210Pb and 14C dating atAustralian Nuclear Science and Technology Organisation(ANSTO) (Lucas Heights NSW Australia) The upper 12 cmof the core was analysed for 210Pb activity and the constant rateof supply (CRS) model was used to develop a chronology forDiprose Lagoon (Oldfield and Appleby 1984) The CRS modelassumes a constant flux of unsupported 210Pb to the surfacesediments through time (Appleby and Oldfield 1978 Binford1990) This model is suitable for many temperate-zone lakeswhere 210Pb activity has been diluted by recent high sedimentaccumulation rates (Binford 1990) and is appropriate in this

Table 1 Sampling details of the seven lagoons in the Midlands of TasmaniaThe location and landscape setting sampling methods employed for extracting sediment and their total length of each core sampling resolution is summarised

for each site The first occurrence of Pinus is also noted

Lagoon name Location Landscape setting Sampling method Lengthof core

Sampling resolution Pinus1st record

Diprose Lagoon 41480252600S147220162600E

Agricultural land Eucalyptusamygdalina woodland

Universal coresampler

215 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Smiths Lagoon 41490229500S147260116400E

Eucalyptus amygdalinawoodland

Shovel excavation 14 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Near Lagoon 42 40118500S147260158300E

Themeda and Poa grasslandsEucalyptus viminalis grassywoodlands and Eucalyptuspauciflora woodland

Auger 36 cm 2 cm (pollen)2 cm (charcoal)

22 cm

White Lagoon 42 50559200S147260551200E

Themeda triandra grasslandE viminalis woodland

Universal coresampler

26 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Bells Lagoon 42 50102100S147210148900E

Agricultural land nativelowland grassland complexE viminalis woodland

Universal coresampler

28 cm 2 cm (pollen)05 cm (charcoal)

4 cm

Township Lagoon 42 80501300S1472503600E

Agricultural land nativelowland grassland complex

Universal coresampler

34 cm 2 cm (pollen)05 cm (charcoal)

26 cm

Cleveland Lagoon 41480127100S147240381600E

Agricultural landE amygdalina andE viminalis woodland

Shovel excavation 21 cm 2 cm (pollen)05 cm (charcoal)

Throughout

D Australian Journal of Botany L M Romanin et al

supporting sedge and rushlands dominated by a suite ofgenera including Juncus Baumea Carex Lepidosperma andPhragmites and herblands with ephemeral aquatic plants such asMyriophyllum spp and Triglochin spp where open water ispresent The most saline wetlands support salt marshcommunities including Chenopodium spp (Kirkpatrick andGlasby 1981)

Sediment samplingSediment cores were collected from seven lagoons along a~40 km long northndashsouth transect spanning the study area(Fig 1c d) The lagoons varied in condition and aquaticvegetation therefore a variety of methods were used to collectsediment samples sample resolution was necessarily low at onesite due to the dryness of the sediments (Table 1) A d-sectioncorer was used where sediments were soft enough for corerpenetration A hammer-driven polycarbonate piston corer wasused where sediments were dominated by stiff clays All coreswere wrapped in plastic and stored in PVC tubes They werekept cool and subsequently subsampled for macro-charcoal andpollen analysis

Charcoal and pollen analysisMacro-charcoal particles (gt125mm) are likely to be derivedfrom local fire events (Whitlock and Larsen 2001 Duffin et al2008) To prepare samples for macro-charcoal analysis asubsample of 125 cm3 of material from each contiguous05 cm increment was extracted Subsamples with high claycontent were deflocculated in Calgon solution for 2ndash3 daysafter which samples were rinsed gently with water Charcoalsubsamples were then treated with bleach for at least 24 h toremove the pigment from organic matter and aid in identifyingcharcoal (Black et al 2007) Bleached samples were gentlywashed through a 125mm sieve to separate macro-charcoal

from organic matter and finer particles Total macro-charcoalparticles were counted under a low power stereo-microscope

Broad vegetation patterns over time were examined throughpollen analysis Sediment recordswere sampled at 2 cm intervals125 cm3 subsamples of these were analysed at AustralianNational University (ANU) using standard preparation andcounting techniques by Bennett and Willis (2001) Pollenidentification and nomenclature follows species lists includedby Zacharek et al (1997) plant surveys at Tom Gibson NatureReserve (Parks and Wildlife Service Tasmania 2006) andreference collections held at ANU (httpapsaanueduauaccessed 5 May 2014) Eucalyptus and Allocasuarina typepollen is assigned based on current vegetation patternsTerrestrial or dryland pollen counts (including fern spores butexcluding aquatic taxa) were expressed as percent pollen of eachsample and used in statistical analyses where (a) taxa thatoccurred in less than four samples per core or made up lessthan 1 of the total were excluded from analysis (n= 17) (b) allRumex types were treated as one taxon as distinguishing betweennative and exotic Rumex taxa that coexist in Tasmania isproblematic (c) indeterminate pollen grains were alsoexcluded from analysis

Sediment chronologyDiprose Lagoon was selected for 210Pb and 14C dating atAustralian Nuclear Science and Technology Organisation(ANSTO) (Lucas Heights NSW Australia) The upper 12 cmof the core was analysed for 210Pb activity and the constant rateof supply (CRS) model was used to develop a chronology forDiprose Lagoon (Oldfield and Appleby 1984) The CRS modelassumes a constant flux of unsupported 210Pb to the surfacesediments through time (Appleby and Oldfield 1978 Binford1990) This model is suitable for many temperate-zone lakeswhere 210Pb activity has been diluted by recent high sedimentaccumulation rates (Binford 1990) and is appropriate in this

Table 1 Sampling details of the seven lagoons in the Midlands of TasmaniaThe location and landscape setting sampling methods employed for extracting sediment and their total length of each core sampling resolution is summarised

for each site The first occurrence of Pinus is also noted

Lagoon name Location Landscape setting Sampling method Lengthof core

Sampling resolution Pinus1st record

Diprose Lagoon 41480252600S147220162600E

Agricultural land Eucalyptusamygdalina woodland

Universal coresampler

215 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Smiths Lagoon 41490229500S147260116400E

Eucalyptus amygdalinawoodland

Shovel excavation 14 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Near Lagoon 42 40118500S147260158300E

Themeda and Poa grasslandsEucalyptus viminalis grassywoodlands and Eucalyptuspauciflora woodland

Auger 36 cm 2 cm (pollen)2 cm (charcoal)

22 cm

White Lagoon 42 50559200S147260551200E

Themeda triandra grasslandE viminalis woodland

Universal coresampler

26 cm 2 cm (pollen)05 cm (charcoal)

10 cm

Bells Lagoon 42 50102100S147210148900E

Agricultural land nativelowland grassland complexE viminalis woodland

Universal coresampler

28 cm 2 cm (pollen)05 cm (charcoal)

4 cm

Township Lagoon 42 80501300S1472503600E

Agricultural land nativelowland grassland complex

Universal coresampler

34 cm 2 cm (pollen)05 cm (charcoal)

26 cm

Cleveland Lagoon 41480127100S147240381600E

Agricultural landE amygdalina andE viminalis woodland

Shovel excavation 21 cm 2 cm (pollen)05 cm (charcoal)

Throughout

D Australian Journal of Botany L M Romanin et al

landscape characterised by recent vegetation clearance andsubsequent erosion (Sigleo and Colhoun 1982) To estimatethe ages of the deeper sediments AMS radiocarbon (14C)dating was used Two samples were analysed however onlythat from 20 cm depth gave an age consistent with the 210Pbchronology whereas the sample at 15 cm gave an age that isbest explained as contaminated with older carbon or otherwiseunreliable Age-depthmodelling of the 210Pb and 14C chronologywas performed using the CLAM package in R (R DevelopmentCore Team 2008 Blaauw 2010) a smoothed spline curve wasfitted to the 210Pb and 14C dates In the remaining six cores theoccurrence of Pinus spp was used as an unambiguous albeitconservative marker to divide into pre- and post-Europeanperiods (Mooney et al 2001) One core (Cleveland Lagoon)was found to have Pinus throughout the entire core so wasexcluded from further analysis

Data analysisChanges in macro-charcoal and major terrestrial and aquaticpollen types were investigated for the dated core fromDiprose Lagoon through graphical analysis and detrendedcorrespondence analysis (DCA) To determine whether macro-charcoal had changed since the first appearance of Pinus spp(ie pre- and post-European periods) all sites except Clevelandwere combined log-transformed then analysed by aMannndashWhitney U test Using the percentage pollen data forthe six sediment cores boxplots were generated to examinethe distribution of pollen abundance in the pre- and post-European periods for the following ecological groups of taxa(a) dry sclerophyll forest trees ndash Allocasuarina spp andEucalyptus spp (b) native herbaceous taxa ndash PoaceaeAsteraceae (Tubuliflorae) and Chenopodiaceae and (c) exoticruderal taxa ndash Brassicaceae Asteraceae (Liguliflorae) andPlantago lanceolata The difference in pollen percentagebetween the two periods for these ecological groups wasmodelled using generalised linear models with a quasibinomialdistribution to account for the overdispersed proportion dataAll analyses were performed using the R statistical softwarepackage (R Development Core Team 2008) ordinations wereperformed using the vegan package (Oksanen et al 2015)

Results

Diprose Lagoon

The geochronology using 210Pb and 14C of the 22 cm DiproseLagoon core indicates a continuous and surprisingly linearsediment record over the past ~600 years (Tables 2 3) Thelowest 210Pb measurement at 12 cm depth coincides withEuropean arrival in Tasmania (1803) (Fig 2a) There arechanges in the physical composition of the sediments with thepost-European sediments being loamy whereas the pre-European sediments are clay mixed with organic matter to18 cm after which clays are less enriched in organic matterThere is a corresponding strong decline in pollen preservationdown the profile with a concentration of 32 pollen grains percm3 at the lowest depth (215 cm)

There were clear changes in floristic composition throughthe core with a conspicuous decline in the abundance ofthe two dominant tree taxa Eucalyptus and Allocasuarina

throughout the post-European period (Fig 2b) At the time ofsettlement Eucalyptus formed ~50 of the pollen assemblagefalling to ~10 in the upper (modern) sediments In the sameperiod Allocasuarina declined from 15 to 4 (Fig 2b) Poaceae(grass) pollen was the most plentiful herbaceous group beingrelatively common (7ndash25) in the pre-European period but itsabundance dropped immediately after European arrival (Fig 2c)Since AD1990 Poaceae has become the most abundant (35) inthe terrestrial pollen record Asteraceae (Tubuliflorae) or lsquonativedaisyrsquo pollen also showed a decline following Europeansettlement from being the most abundant herbaceous taxon(25 of the pollen record) in the 14th century (Fig 2c)declining to 1ndash5 of the terrestrial pollen in the 20th centuryChenopodiaceae first appeared in the record at around the time ofEuropean settlement and has maintained low levels (~1)throughout the modern period Amongst the aquatic pollenthere is a striking transition from Botryococcus spp toMyriophyllum spp at around AD1940 (Fig 2d) Cyperaceaepollen first appears at AD1870 and increases in abundancetowards the surface Pinus spp also first occurs at aroundAD1870 reaching peak abundance at AD1960 and declining inmodern sediments Asteraceae (Liguliflorae) pollen is often usedas indicator of the European period but is not diagnostic here asthere are native Tasmanian species within this pollen typeincluding the important edible plant Microseris lanceolatalsquoyam daisyrsquo which may have formed part of the diet ofAboriginal people in the region (Gott and Murray 1982) Itspresence in both the pre- (AD1630) and post-European sectionsof the core indicates there were native species in the region thatwere replaced by exotic taxa (Fig 2e) A full pollen diagram is

Table 2 210Pb dating for Diprose Lagoon core from the in theMidlands of Tasmania

Dating was performed by Australian Nuclear Science and TechnologyOrganisation (ANSTO) Depth of sample level of radioactivity andcalculated age based on a constant rate of supply (CRS) model are supplied

Depth (cm belowsurface)

Total 210Pb(Bq kgndash1)

Massaccumulation

rates (g cmndash2 yearndash1)

CalculatedCRS ages(years)

0ndash3 138plusmn 6 0091 plusmn 0005 9 plusmn 33ndash4 139plusmn 7 0053 plusmn 0004 26 plusmn 54ndash5 129plusmn 6 0043 plusmn 0004 35 plusmn 65ndash6 109plusmn 5 0035 plusmn 0003 47 plusmn 76ndash7 82 plusmn 4 0028 plusmn 0004 66 plusmn 87ndash8 53 plusmn 2 0026 plusmn 0005 88 plusmn 99ndash10 33 plusmn 2 0025 plusmn 0009 134 plusmn 1211ndash12 23 plusmn 1 0014 plusmn 0012 192 plusmn 29

Table 3 14C AMS dates for Diprose Lagoon core in the Midlandsof Tasmania

Depth of sample d13C of samples and calculated sample ages with 1d errorare provided

Depth (cm belowsurface)

d(13C)per mil

Conventional 14C age14C year before present 1d error

15 cmAndash303 1932 28

20 cm ndash350 538 22

ASample excluded from agendashdepthmodel due to suspected contaminationof sample with older reworked carbon

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany E

1992

1915

1426

0

200

400

600

60

40

20

0

40

20

0

80

40

0

30

20

10

0

80

60

40

20

0

05

00

ndash05

ndash10

ndash15

cal B

PP

erce

ntag

eP

erce

ntag

eP

erce

ntag

e

Per

cent

age

Par

ticle

s (gt

125

microm)

cmndash3

DC

A 1

st a

xis

(a)

(b)

(c)

(d )

(e)

(f )

(g)

0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22

Depth (cm)

Asteraceae (tubuliflorae)Plantago spp (native)

Allocasuarina spp

BrassicaceaeSalixPinus sppCoprosma spp

Allocasuarina sppEucalyptus spp

BrassicaceaePinus sppPlantago lanceolataAsteraceae (liguliflorae)

PoaceaeChenopodiaceaeAsteraceae (tubuliflorae)

CyperaceaeMyriophyllum sppBotryococcus spp

1804

Fig 2 Variation in pollen and charcoal in the dated core from Diprose Lagoon in the Midlands of Tasmania (a) An age-depth model was developed usingCLAM in R (Blaauw 2010) by fitting a smoothed spline curve to the 210Pb (black) and 14C (coloured purple) dates estimated ages were converted to calendaryears with 1950 as the baseline The x-axis is depth in cm with surface sediments at the left The scale of the y-axis varies in each panel The blue vertical lineindicates the depth at which the first grain of Pinus pollen was found the red vertical line indicates the approximate year that Europeans arrived in Tasmania(b) percentage of the major dry forest tree taxa pollen Allocasuarina spp and Eucalyptus spp (based on dryland pollen sum) (c) percentage of nativeherbaceous groups Poaceae Asteraceae (Tubuliflorae) and Chenopodiaceae (based on dryland pollen sum) (d) percentage of aquatic taxa CyperaceaeMyriophyllum spp and Botryococcus spp (based on the aquatic pollen sum excluding dryland taxa) (e) percentage of exotic indicator groups including thetree species Pinus and the ruderal herbs Brassicaceae Plantago lanceolata and Asteraceae (Liguliflorae) (based on dryland pollen sum) (f) contiguous recordof macro-charcoal (particles gt125 microns with sample resolution of 05 cm) (g) the first axis of the DCA plotted against depth of the core a steep slope of theaxis indicates periods of rapid vegetation change Pollen was analysed at 2 cm resolution

F Australian Journal of Botany L M Romanin et al

included in Fig S1 available as Supplementary Material to thispaper

Charcoal was present throughout the sediment recordthough the uppermost sediments after AD1990 have very lowbut fairly continuous macro-charcoal deposition A major peakin macro-charcoal preserved in the sediments was found in themid-19th and early 20th century (Fig 2f) After the 1950scharcoal deposition dropped to near pre-European levels Inthe pre-European period there is only a small amount ofcharcoal deposited with gaps in the record

The first axis of a DCA ordination of the terrestrial pollen taxareflects the rate and direction of vegetation change followingEuropean settlement (Figs 2g 3) In the early 20th century there israpid floristic transition largely driven by the establishment ofexotic taxa (Fig 2e) The taxa that drive this transition are thosewith the strongest positive DCA scores and include Coprosmaspp as well as exotic taxa Brassicaceae Salix spp Pinusspp (Fig 3) This transitional period also corresponds to thepeak macro charcoal (Fig 2f) In the earliest part of the record(AD1320ndashAD1530) there is also rapid taxonomic turnover thatcorresponds to the transition from Allocasuarina to Eucalyptusdominance (Fig 2b) The native species that comprise thenegative portion of the DCA first axis are Asteraceae(Tubuliflorae) Plantago spp (native) and Allocasuarinaspp (Fig 3)

Regional network

Total pollen preservation varied substantially between thewetlands Township and Cleveland Lagoons had very good

preservation throughout the length of the core Diprose WhiteNear and Smiths Lagoons had excellent pollen preservation in theupper sections of the sediment profile but much poorer pollenpreservation in the deepest sediments particularly so for Smithsand as noted previously Diprose Lagoon Bells Lagoon hadhigh pollen preservation only in the surface sediment samplethe remainder of this core had poor preservation Summaries ofpollen and charcoal found at the six lagoons not discussed indetail are included in Fig S2 and a list of all pollen and spore taxarecorded in this analysis is included in Table S1 available asSupplementary Material to this paper

We used the first incidence of Pinus spp pollen toapproximate European settlement (Kodela and MacKillop1988) in all of the cores except Cleveland Lagoon because ithad Pinus pollen throughout and overall pre- and post-Europeanperiods were examined Among the remaining 6 cores Pinusconstitutes 6 of the total terrestrial pollen with an interquartilerange of 05ndash8 In these cores the first appearance of Pinusvaried from 4 to 26 cm depth indicating different rates ofsedimentation in the wetlands (Table 1) We used these sixcores for statistical contrasts in pollen between the pre(ndash Pinus) and post (+ Pinus) European periods For these coresthere was no statistically significant change in the proportion ofEucalyptus pollen after the arrival of Europeans across the region(P= 0484) (Fig 4) however there was a statistically significantdecrease in Allocasuarina pollen (P = 001) (Fig 4) In thedeepest sections of four cores Allocasuarina pollen was atleast as common as Eucalyptus There were marked increasesin Poaceae pollen abundance after settlement with Asteraceae(Tubuliflorae) and Chenopodiaceae pollen abundance

3

2

1

0

ndash1

ndash2

ndash4 ndash2 0 2 4

DCA1

DC

A2

Fig 3 Detrended correspondence analysis (DCA) of sediment record from Diprose Lagoon Green squares indicate samples from the pre-European period(as determined by the absence of Pinus spp pollen) red squares indicate samples from the post-European period All the taxa with a strong influence on thefirst two axes are indicated where the following abbreviations are used AP Apiaceae AsteraceaeL Asteraceae (Liguliflorae) AsteraceaeT Asteraceae(Tubuliflorae) AX Amperea xiphoclada Car Caryophyllaceae Ch Chenopodiaceae Cup Cupressaceae DA Dicksonia Antarctica Dod Dodonea sppFab Fabaceae HI Histiopteris incise Indet indeterminate pollen grains Isp Isoetes spp Lin Lindseae LF Lagarostrobos franklinii Mon monolete psilatespore Msp Micrantheum spp NC Nothofagus cunninghamii PA Phyllocladus aspleniifolius Plantago Plantago spp (native) PlantagoL Plantagolanceolata Pom Pomaderris apetala TL Tasmannia lanceolata

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany G

significantly decreasing (Fig 4) Two of the remainingherbaceous taxa examined probably represent a mix of bothexotic and native species native Asteraceae (Liguliflorae) andBrassicaceae species are rare today but they have a strongrepresentation in the early period Both of these groups havebeen replaced by exotic ruderal species as indicated by asignificant increase in Asteraceae (Liguliflorae) type pollenand an increase in the median value of Brassicaceae pollen inthe post-settlement period (Fig 4) Plantago lanceolata is anexclusively exotic species that appeared earlier in the sedimentrecord than the exotic indicator taxa Pinus at two wetlands(Fig 4)

The most abundant aquatic taxon was Botryococcus spp itwas present at each of the six wetlands examined and made up~80 of the aquatic record with no significant change inabundance found between the pre- and post-European periods(P = 049) There was a significant increase in Myriophyllumspp (P = 002) after European settlement (Fig 4)

Minor taxa

Amongst the 6 wetlands rainforest components such asNothofagus cunninghamii Pomaderris apetala type andPhyllocladus aspleniifolius were a minor background element(lt9) of total terrestrial pollen Dry forest species such asAcacia spp Banksia marginata and undifferentiated Fabaceae

species made up a very low proportion of the total terrestrialpollen typically individual species comprised less than 6of the total terrestrial pollen in each core The highest diversitywas found in herbaceous taxa and these were in aggregatemore abundant than the previously mentioned groups Rumexwas the only taxon to show a statistically significant(P = 0003) increase (mean 001 to 17) from the pre- topost-European period (Fig 4) Pteridophyte taxa had constantlow background levels (3) in all the cores Isoeteswas presentin all wetlands in both time periods albeit with highly variablelevels

There were only seven aquatic or wetland species representedin the pollen records None of the minor aquatic species changedin abundance between the two periods Restionaceae was theonly minor wetland family found at all wetlands it made up03ndash6 of the aquatic pollen assemblage

Charcoal

The amount of charcoal preserved at the individualwetlands varied considerably The amount of macro charcoalpreserved in all six sampled wetlands increased from the pre-European (median = 16 pieces cmndash3) to the post settlementperiods (median = 68 pieces cmndash3) (U = 3215 Plt 0001)(Fig 5)

100

80

60

40

20

0P = 001 P = 048 P = 014 P = 001 P lt 005 P = 018 P lt 005 P lt 0001 P lt 001 P = 049 P = 002

pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post

Allo Euc Poa AsterT Cheno Brass AsterL Plan Rum Bot Myr

Per

cent

age

Fig 4 Differences in the pre- and post-European patterns for pollen data derived from a combined analysis of the 6 sediment cores showing the dominantdry forest tree taxa Allocasuarina (Allo) and Eucalyptus (Euc) native herbaceous taxa Poaceae (Poa) Asteraceae (Tubuliflorae) (AsterT) and Chenopodiaceae(Cheno) exotic ruderals Brassicaceae (Brass) Asteraceae (Liguliflorae) (AsterL) Plantago lanceolata (Plan) and Rumex spp (Rum) and aquatic taxaBotryococcus spp (Bot) and Myriophyllum spp (Myr) Pollen values are proportions relative to selected pollen sums (see Fig 2) and have been convertedto percentages for presentation Probablility values based on generalised linear models are shown for each contrast

H Australian Journal of Botany L M Romanin et al

Discussion

Until now there have been few attempts to examine the interplayof vegetation and fire in the dry eastern side of Tasmaniaparticularly the Midlands (Fensham 1989 Jones 2008 vonPlaten et al 2011) Most of the historical research in southernAustralia has relied upon the accounts of early settlers to evaluatethe fire regime of Aboriginal peoples The paucity ofpalaeoecological research in this region has been in part dueto poor pollen preservation and the difficulties in examiningsediment records from seasonally dry lagoons because pollenpreservation can be inadequate (Thomas 1991) and 210Pb datingcan become problematic because drying of sediments mayfacilitate the downcore movement of 210Pb which can lead toan underestimate of age at any given depth (Higuera et al2005) Our approach of using a network of cores enabled us toselect a site with good pollen preservation that was suitable forbuilding a chronology and use other less sensitive sites to drawlandscape-level inferences based on statistical analysis of pollenand charcoal in samples amalgamated using a Pinus pollen as amarker of European settlement

Our study demonstrates marked ecological changes in theTasmanian Midlands following the establishment of a neo-European agricultural landscape These include a change inwoody and herbaceous species abundance associated with

native vegetation clearing and the introduction of exoticpasture species In addition we found a dramatic increase infire activity in the terrestrial environment and changedhydrological conditions in wetlands These findings concurwith our research expectations and the findings of previousresearch (Kershaw et al 1994) In contrast to previousresearch however we found no change in sedimentation rate(Dodson and Mooney 2002) We avoided problems with datingby using Pinus as a proxy for the post-settlement period exoticindicator species have been used to demarcate the arrival ofEuropeans in both Australia (Kershaw et al 1994 Graysonet al 1998 Kenyon and Rutherfurd 1999) and America (vanZant et al 1979)

Palaeoecological studies in southern Australia using pollenand charcoal to reconstruct pre-European fire regimes arebeginning to reveal the dynamic nature of fire frequency andimpact on vegetation during the Holocene although theinterpretation of these data is not clear cut (Mooney et al2001 Fletcher et al 2014) Charcoal peaks in sediments aremore likely to reflect high intensity fires (Whitlock and Larsen2001 Higuera et al 2010) However the absence of charcoal in asedimentary record does not necessary equate to an absence offire Low intensity fires produce high levels of incompletelyburned particulate matter but this effect is counteractedbecause these fires are typically small in extent leaving asmall signature in the pollen record (Whitlock and Larsen2001 Higuera et al 2010) Indeed Mooney et al (2011)argued that the presence or absence of sedimentary charcoalmay be a poor indicator of low intensity fires thought tocharacterise Aboriginal burning patterns Studies of traditionalAboriginal fire usage in northern Australia have demonstrated aregime of frequent and patchy fire (Bowman et al 2001 2004Bliege Bird et al 2008 Trauernicht et al 2015) They used fireto maintain open grassy landscapes to increase the abundanceof game species (Murphy and Bowman 2007 Codding et al2014) and motivations are assumed to be similar in southernAustralia Dendrochronological evidence shows that dry foreststo the east of the study region persisted through increasedburning after Europeans arrival from ~07 fires per decadeunder Aboriginal management to 17 per decade in the postsettlement period (von Platen et al 2011)

Previous studies on mainland south-east Australia havefound an increase in charcoal after European settlement(Johnson 2000 Mooney and Dodson 2001 Mooney et al2001 2011) Likewise the charcoal from Diprose Lagoonrecord showed that there was a sharp increase in charcoalproduction concurrent with the appearance of Pinus in thesediment record We suggest the source of this charcoal is thecombined effects of using fire to clear land and settinggeographically large grass fires to stimulate pasture production(Kirkpatrick 2007) In theDiprose core charcoal production in the1950s declined possibly reflecting a dwindling of tree-clearingand associated burning and greater use of mechanisedagriculture

The greater abundance of Allocasuarina in the period beforeEuropean settlement suggests that fire regimes were favourablefor this species and other relatively fire sensitive species andthat clearing grazing and an increase in burning by Europeanscaused Allocasuarina to decline This inference is supported

30

25

20

15

10

05

00

pre post

log(

char

coal

(no

cm

ndash3))

Fig 5 Boxplots of the differences in the pre- and post-European periodsfor the macro-charcoal record of six wetlandsrsquo sediment cores Charcoalconcentrations of the two periods are log-transformed and differ at the 0001level according to a MannndashWhitney test The pre-European abundance isshown on the left post-European on the right

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany I

by the post-fire regeneration strategies of the two Allocasuarinaspecies found in theMidlandsA littoralis is a serotinous obligateseeder that has low fire resistance (Morrison and Renwick 2000Burley et al 2007) while A verticillata is also serotinous but hasa limited capacity to resprout from rootstock (Singh and Geissler1985Ladd1988)Both species regenerate prolifically afterfire asthey release a heavy rain of seed onto disturbed ground (Ladd1988) but require at least 5ndash6 years fire free to begin producing itsserotinous seed (Hueneke 1976 Singh and Geissler 1985) Asimilar pattern of early and rapid decline of Allocasuarina hasbeen found in other temperate savannas of southern Australia(Kershaw et al 1994 Mooney and Dodson 2001 Bickford et al2008 Jones 2008)

We found no significant change in the abundance ofEucalyptus pollen following European settlement This lack ofchange inEucalyptuswas unexpected given thewell documentedloss of tree cover in the Midlands which has occurred sinceEuropean settlement (Fensham and Kirkpatrick 1989 Prior et al2013 Romanin et al 2015) and may be an artefact of theamalgamation of data from multiple cores In highly disturbedagricultural areas of New SouthWales a similar lack of change inthe pollen signature of Eucalyptus has been identified with thesuggestion that clearing of woodland trees has been compensatedby theplantingof other species of eucalypts aroundhomesteadsoras windbreaks (Kershaw et al 1994) A lack of overall changemay also be evidence of thickening of native remnant vegetationafter Aboriginal fire management was disrupted (Gammage2008) Or alternatively as tree density decreased floweringand therefore pollen production may have increased and leftthe pollen signature unchanged (Williams et al 2006) therelevance of this to woodland eucalypt species is howeverlargely unknown

These sediment cores provide further circumstantial evidencefor a change in the species composition of understorey plantsfollowing settlement including a sharp reduction in abundance ofthe Aboriginal staple food of southern Australia Microserislanceolata or yam daisy (Gott and Murray 1982 Gott 19832005 Clarke 1986) Asteraceae (Liguliflorae) type pollen cannotbe identified to species but except for M lanceolata Sonchusmegalocarpus and Picris angustifolia all members in Tasmaniaare exotic species S megolocarpus and P angustifolia are onlyfound in coastal regions (httpwwwutaseduaudicotkeydicotkeyASTgAsteraceaehtm accessed 17 February 2016)so we attribute this pollen type in the pre-European period toM lanceolataM lanceolata numberswould have declinedwhenactive Aboriginal management ceased and introduced sheepbegan eating them (Gott and Murray 1982) Gott and Murray(1982) point out that the invading European species from theAsteraceae (Liguliflorae) group such as Hypocharis andLeontodon are adapted to living with sheep and cattle andwould have outcompeted M lanceolata

We found an initial decline then increased abundance ofPoaceae in the Midlands following European colonisation Theabundance of Poaceae in association with Eucalyptus andAsteraceae pollen before Europeans arriving supports thenotion that eucalypt savanna was the dominant vegetation atthe time of colonisation (Ellis and Thomas 1988 Fensham 1989Kershaw et al 1994) Thedramatic loss of native grasslands in theMidlands that has occurred in the 20th century (Gilfedder 1990)

cannot be seen in these sediment records because native grassspecies have been replaced by exotic pastures with higher pollenproduction (Smart et al 1979) especially in the post SecondWorld War period when pasture production was sharplyincreased (Kirkpatrick 2007) Potentially quantification of thischange from native to exotic grasses could be approximated infuture work by measuring the size of Poaceae grains in thesediment record larger grains may be produced by introducedgrass taxa (Kenyon and Rutherfurd 1999 Bickford et al 2008)

Changes in the composition of aquatic taxa were mostapparent at Diprose Lagoon where a shift in composition fromBotryococcus spp toMyriophyllum spp in the mid-20th centurywas detected Botryococcus is an algal genus which requiresstanding water (Harle et al 1993) whereasMyriophyllum spp isassociatedwith slowflowingwater or themargins of streamsWeinterpret this transition as reflecting the practise of drainage ofwetlands to increase agricultural land thus producing suitablehabitat for Myriophyllum to establish in impounded andoccasionally flowing artificial drainage channels Across theMidlands 34 of wetland area has been drained and a further23 has been affected by artificial changes in water level(Fensham and Kirkpatrick 1989) In addition to changes inwater levels Botryococcus is vulnerable to eutrophicationwhich may have occurred with increased application ofartificial fertilisers and associated increased sheep density inthe second half of the 20th century (Kellaway 1989Kirkpatrick 2007) A shift towards Myriophyllum and anoverall increase in its abundance throughout the region issomewhat surprising given its sensitivity to salinity anddocumented salinity increases in the Midlands (Kirkpatrickand Harwood 1983 Bastick and Walker 2000 Bastick andLynch 2003) However some species such as M salsugineumare known to be tolerant of shallow saline waters (Thomas 1991)

We found no change in sedimentation rate at Diprose LagoonThis was unexpected becausemost sediment cores from southernAustralia show that a clear increase in sedimentation occurredshortly after the arrival of Europeans in the region (Dodson et al1994a 1994b Mooney and Dodson 2001) Sediment depositionis increasedwhen vegetation cover is removed throughfire or treeremoval which allows erosion to occur An apparent lack ofincreased sedimentation at Diprose Lagoon could be associatedwith increased compaction of upper sediments after sheep wereintroduced and the impact of artificial drainage on catchmenthydrology and sedimentation

Broadly we support the notion that Aboriginal landmanagement maintained eucalypt savanna and that Europeandisruption of Aboriginal management resulted in changed fireregimes and associated ecological changes most notably anincrease in fire a reduction in Allocasuarina and increases inexotic taxa However we cannot resolve the question of thehypothesised switch from a finer grained to coarse grained spatialpattern of landscape burning that caused the loss ofgrasslandndashforest mosaics that lies at the heart of manyenvironmental histories (eg Rolls 1981 Jurskis 2000Gammage 2011) It is possible that integrated studiescombining dendrochronology palynology and charcoalsediment analysis and historical records in the montanegrassland-forest mosaics in Tasmania that are rich in sedimenttraps andhave tree specieswith annual growth rings could further

J Australian Journal of Botany L M Romanin et al

1992

1915

1426

0

200

400

600

60

40

20

0

40

20

0

80

40

0

30

20

10

0

80

60

40

20

0

05

00

ndash05

ndash10

ndash15

cal B

PP

erce

ntag

eP

erce

ntag

eP

erce

ntag

e

Per

cent

age

Par

ticle

s (gt

125

microm)

cmndash3

DC

A 1

st a

xis

(a)

(b)

(c)

(d )

(e)

(f )

(g)

0 2 4 6 8 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22

Depth (cm)

Asteraceae (tubuliflorae)Plantago spp (native)

Allocasuarina spp

BrassicaceaeSalixPinus sppCoprosma spp

Allocasuarina sppEucalyptus spp

BrassicaceaePinus sppPlantago lanceolataAsteraceae (liguliflorae)

PoaceaeChenopodiaceaeAsteraceae (tubuliflorae)

CyperaceaeMyriophyllum sppBotryococcus spp

1804

Fig 2 Variation in pollen and charcoal in the dated core from Diprose Lagoon in the Midlands of Tasmania (a) An age-depth model was developed usingCLAM in R (Blaauw 2010) by fitting a smoothed spline curve to the 210Pb (black) and 14C (coloured purple) dates estimated ages were converted to calendaryears with 1950 as the baseline The x-axis is depth in cm with surface sediments at the left The scale of the y-axis varies in each panel The blue vertical lineindicates the depth at which the first grain of Pinus pollen was found the red vertical line indicates the approximate year that Europeans arrived in Tasmania(b) percentage of the major dry forest tree taxa pollen Allocasuarina spp and Eucalyptus spp (based on dryland pollen sum) (c) percentage of nativeherbaceous groups Poaceae Asteraceae (Tubuliflorae) and Chenopodiaceae (based on dryland pollen sum) (d) percentage of aquatic taxa CyperaceaeMyriophyllum spp and Botryococcus spp (based on the aquatic pollen sum excluding dryland taxa) (e) percentage of exotic indicator groups including thetree species Pinus and the ruderal herbs Brassicaceae Plantago lanceolata and Asteraceae (Liguliflorae) (based on dryland pollen sum) (f) contiguous recordof macro-charcoal (particles gt125 microns with sample resolution of 05 cm) (g) the first axis of the DCA plotted against depth of the core a steep slope of theaxis indicates periods of rapid vegetation change Pollen was analysed at 2 cm resolution

F Australian Journal of Botany L M Romanin et al

included in Fig S1 available as Supplementary Material to thispaper

Charcoal was present throughout the sediment recordthough the uppermost sediments after AD1990 have very lowbut fairly continuous macro-charcoal deposition A major peakin macro-charcoal preserved in the sediments was found in themid-19th and early 20th century (Fig 2f) After the 1950scharcoal deposition dropped to near pre-European levels Inthe pre-European period there is only a small amount ofcharcoal deposited with gaps in the record

The first axis of a DCA ordination of the terrestrial pollen taxareflects the rate and direction of vegetation change followingEuropean settlement (Figs 2g 3) In the early 20th century there israpid floristic transition largely driven by the establishment ofexotic taxa (Fig 2e) The taxa that drive this transition are thosewith the strongest positive DCA scores and include Coprosmaspp as well as exotic taxa Brassicaceae Salix spp Pinusspp (Fig 3) This transitional period also corresponds to thepeak macro charcoal (Fig 2f) In the earliest part of the record(AD1320ndashAD1530) there is also rapid taxonomic turnover thatcorresponds to the transition from Allocasuarina to Eucalyptusdominance (Fig 2b) The native species that comprise thenegative portion of the DCA first axis are Asteraceae(Tubuliflorae) Plantago spp (native) and Allocasuarinaspp (Fig 3)

Regional network

Total pollen preservation varied substantially between thewetlands Township and Cleveland Lagoons had very good

preservation throughout the length of the core Diprose WhiteNear and Smiths Lagoons had excellent pollen preservation in theupper sections of the sediment profile but much poorer pollenpreservation in the deepest sediments particularly so for Smithsand as noted previously Diprose Lagoon Bells Lagoon hadhigh pollen preservation only in the surface sediment samplethe remainder of this core had poor preservation Summaries ofpollen and charcoal found at the six lagoons not discussed indetail are included in Fig S2 and a list of all pollen and spore taxarecorded in this analysis is included in Table S1 available asSupplementary Material to this paper

We used the first incidence of Pinus spp pollen toapproximate European settlement (Kodela and MacKillop1988) in all of the cores except Cleveland Lagoon because ithad Pinus pollen throughout and overall pre- and post-Europeanperiods were examined Among the remaining 6 cores Pinusconstitutes 6 of the total terrestrial pollen with an interquartilerange of 05ndash8 In these cores the first appearance of Pinusvaried from 4 to 26 cm depth indicating different rates ofsedimentation in the wetlands (Table 1) We used these sixcores for statistical contrasts in pollen between the pre(ndash Pinus) and post (+ Pinus) European periods For these coresthere was no statistically significant change in the proportion ofEucalyptus pollen after the arrival of Europeans across the region(P= 0484) (Fig 4) however there was a statistically significantdecrease in Allocasuarina pollen (P = 001) (Fig 4) In thedeepest sections of four cores Allocasuarina pollen was atleast as common as Eucalyptus There were marked increasesin Poaceae pollen abundance after settlement with Asteraceae(Tubuliflorae) and Chenopodiaceae pollen abundance

3

2

1

0

ndash1

ndash2

ndash4 ndash2 0 2 4

DCA1

DC

A2

Fig 3 Detrended correspondence analysis (DCA) of sediment record from Diprose Lagoon Green squares indicate samples from the pre-European period(as determined by the absence of Pinus spp pollen) red squares indicate samples from the post-European period All the taxa with a strong influence on thefirst two axes are indicated where the following abbreviations are used AP Apiaceae AsteraceaeL Asteraceae (Liguliflorae) AsteraceaeT Asteraceae(Tubuliflorae) AX Amperea xiphoclada Car Caryophyllaceae Ch Chenopodiaceae Cup Cupressaceae DA Dicksonia Antarctica Dod Dodonea sppFab Fabaceae HI Histiopteris incise Indet indeterminate pollen grains Isp Isoetes spp Lin Lindseae LF Lagarostrobos franklinii Mon monolete psilatespore Msp Micrantheum spp NC Nothofagus cunninghamii PA Phyllocladus aspleniifolius Plantago Plantago spp (native) PlantagoL Plantagolanceolata Pom Pomaderris apetala TL Tasmannia lanceolata

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany G

significantly decreasing (Fig 4) Two of the remainingherbaceous taxa examined probably represent a mix of bothexotic and native species native Asteraceae (Liguliflorae) andBrassicaceae species are rare today but they have a strongrepresentation in the early period Both of these groups havebeen replaced by exotic ruderal species as indicated by asignificant increase in Asteraceae (Liguliflorae) type pollenand an increase in the median value of Brassicaceae pollen inthe post-settlement period (Fig 4) Plantago lanceolata is anexclusively exotic species that appeared earlier in the sedimentrecord than the exotic indicator taxa Pinus at two wetlands(Fig 4)

The most abundant aquatic taxon was Botryococcus spp itwas present at each of the six wetlands examined and made up~80 of the aquatic record with no significant change inabundance found between the pre- and post-European periods(P = 049) There was a significant increase in Myriophyllumspp (P = 002) after European settlement (Fig 4)

Minor taxa

Amongst the 6 wetlands rainforest components such asNothofagus cunninghamii Pomaderris apetala type andPhyllocladus aspleniifolius were a minor background element(lt9) of total terrestrial pollen Dry forest species such asAcacia spp Banksia marginata and undifferentiated Fabaceae

species made up a very low proportion of the total terrestrialpollen typically individual species comprised less than 6of the total terrestrial pollen in each core The highest diversitywas found in herbaceous taxa and these were in aggregatemore abundant than the previously mentioned groups Rumexwas the only taxon to show a statistically significant(P = 0003) increase (mean 001 to 17) from the pre- topost-European period (Fig 4) Pteridophyte taxa had constantlow background levels (3) in all the cores Isoeteswas presentin all wetlands in both time periods albeit with highly variablelevels

There were only seven aquatic or wetland species representedin the pollen records None of the minor aquatic species changedin abundance between the two periods Restionaceae was theonly minor wetland family found at all wetlands it made up03ndash6 of the aquatic pollen assemblage

Charcoal

The amount of charcoal preserved at the individualwetlands varied considerably The amount of macro charcoalpreserved in all six sampled wetlands increased from the pre-European (median = 16 pieces cmndash3) to the post settlementperiods (median = 68 pieces cmndash3) (U = 3215 Plt 0001)(Fig 5)

100

80

60

40

20

0P = 001 P = 048 P = 014 P = 001 P lt 005 P = 018 P lt 005 P lt 0001 P lt 001 P = 049 P = 002

pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post

Allo Euc Poa AsterT Cheno Brass AsterL Plan Rum Bot Myr

Per

cent

age

Fig 4 Differences in the pre- and post-European patterns for pollen data derived from a combined analysis of the 6 sediment cores showing the dominantdry forest tree taxa Allocasuarina (Allo) and Eucalyptus (Euc) native herbaceous taxa Poaceae (Poa) Asteraceae (Tubuliflorae) (AsterT) and Chenopodiaceae(Cheno) exotic ruderals Brassicaceae (Brass) Asteraceae (Liguliflorae) (AsterL) Plantago lanceolata (Plan) and Rumex spp (Rum) and aquatic taxaBotryococcus spp (Bot) and Myriophyllum spp (Myr) Pollen values are proportions relative to selected pollen sums (see Fig 2) and have been convertedto percentages for presentation Probablility values based on generalised linear models are shown for each contrast

H Australian Journal of Botany L M Romanin et al

Discussion

Until now there have been few attempts to examine the interplayof vegetation and fire in the dry eastern side of Tasmaniaparticularly the Midlands (Fensham 1989 Jones 2008 vonPlaten et al 2011) Most of the historical research in southernAustralia has relied upon the accounts of early settlers to evaluatethe fire regime of Aboriginal peoples The paucity ofpalaeoecological research in this region has been in part dueto poor pollen preservation and the difficulties in examiningsediment records from seasonally dry lagoons because pollenpreservation can be inadequate (Thomas 1991) and 210Pb datingcan become problematic because drying of sediments mayfacilitate the downcore movement of 210Pb which can lead toan underestimate of age at any given depth (Higuera et al2005) Our approach of using a network of cores enabled us toselect a site with good pollen preservation that was suitable forbuilding a chronology and use other less sensitive sites to drawlandscape-level inferences based on statistical analysis of pollenand charcoal in samples amalgamated using a Pinus pollen as amarker of European settlement

Our study demonstrates marked ecological changes in theTasmanian Midlands following the establishment of a neo-European agricultural landscape These include a change inwoody and herbaceous species abundance associated with

native vegetation clearing and the introduction of exoticpasture species In addition we found a dramatic increase infire activity in the terrestrial environment and changedhydrological conditions in wetlands These findings concurwith our research expectations and the findings of previousresearch (Kershaw et al 1994) In contrast to previousresearch however we found no change in sedimentation rate(Dodson and Mooney 2002) We avoided problems with datingby using Pinus as a proxy for the post-settlement period exoticindicator species have been used to demarcate the arrival ofEuropeans in both Australia (Kershaw et al 1994 Graysonet al 1998 Kenyon and Rutherfurd 1999) and America (vanZant et al 1979)

Palaeoecological studies in southern Australia using pollenand charcoal to reconstruct pre-European fire regimes arebeginning to reveal the dynamic nature of fire frequency andimpact on vegetation during the Holocene although theinterpretation of these data is not clear cut (Mooney et al2001 Fletcher et al 2014) Charcoal peaks in sediments aremore likely to reflect high intensity fires (Whitlock and Larsen2001 Higuera et al 2010) However the absence of charcoal in asedimentary record does not necessary equate to an absence offire Low intensity fires produce high levels of incompletelyburned particulate matter but this effect is counteractedbecause these fires are typically small in extent leaving asmall signature in the pollen record (Whitlock and Larsen2001 Higuera et al 2010) Indeed Mooney et al (2011)argued that the presence or absence of sedimentary charcoalmay be a poor indicator of low intensity fires thought tocharacterise Aboriginal burning patterns Studies of traditionalAboriginal fire usage in northern Australia have demonstrated aregime of frequent and patchy fire (Bowman et al 2001 2004Bliege Bird et al 2008 Trauernicht et al 2015) They used fireto maintain open grassy landscapes to increase the abundanceof game species (Murphy and Bowman 2007 Codding et al2014) and motivations are assumed to be similar in southernAustralia Dendrochronological evidence shows that dry foreststo the east of the study region persisted through increasedburning after Europeans arrival from ~07 fires per decadeunder Aboriginal management to 17 per decade in the postsettlement period (von Platen et al 2011)

Previous studies on mainland south-east Australia havefound an increase in charcoal after European settlement(Johnson 2000 Mooney and Dodson 2001 Mooney et al2001 2011) Likewise the charcoal from Diprose Lagoonrecord showed that there was a sharp increase in charcoalproduction concurrent with the appearance of Pinus in thesediment record We suggest the source of this charcoal is thecombined effects of using fire to clear land and settinggeographically large grass fires to stimulate pasture production(Kirkpatrick 2007) In theDiprose core charcoal production in the1950s declined possibly reflecting a dwindling of tree-clearingand associated burning and greater use of mechanisedagriculture

The greater abundance of Allocasuarina in the period beforeEuropean settlement suggests that fire regimes were favourablefor this species and other relatively fire sensitive species andthat clearing grazing and an increase in burning by Europeanscaused Allocasuarina to decline This inference is supported

30

25

20

15

10

05

00

pre post

log(

char

coal

(no

cm

ndash3))

Fig 5 Boxplots of the differences in the pre- and post-European periodsfor the macro-charcoal record of six wetlandsrsquo sediment cores Charcoalconcentrations of the two periods are log-transformed and differ at the 0001level according to a MannndashWhitney test The pre-European abundance isshown on the left post-European on the right

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany I

by the post-fire regeneration strategies of the two Allocasuarinaspecies found in theMidlandsA littoralis is a serotinous obligateseeder that has low fire resistance (Morrison and Renwick 2000Burley et al 2007) while A verticillata is also serotinous but hasa limited capacity to resprout from rootstock (Singh and Geissler1985Ladd1988)Both species regenerate prolifically afterfire asthey release a heavy rain of seed onto disturbed ground (Ladd1988) but require at least 5ndash6 years fire free to begin producing itsserotinous seed (Hueneke 1976 Singh and Geissler 1985) Asimilar pattern of early and rapid decline of Allocasuarina hasbeen found in other temperate savannas of southern Australia(Kershaw et al 1994 Mooney and Dodson 2001 Bickford et al2008 Jones 2008)

We found no significant change in the abundance ofEucalyptus pollen following European settlement This lack ofchange inEucalyptuswas unexpected given thewell documentedloss of tree cover in the Midlands which has occurred sinceEuropean settlement (Fensham and Kirkpatrick 1989 Prior et al2013 Romanin et al 2015) and may be an artefact of theamalgamation of data from multiple cores In highly disturbedagricultural areas of New SouthWales a similar lack of change inthe pollen signature of Eucalyptus has been identified with thesuggestion that clearing of woodland trees has been compensatedby theplantingof other species of eucalypts aroundhomesteadsoras windbreaks (Kershaw et al 1994) A lack of overall changemay also be evidence of thickening of native remnant vegetationafter Aboriginal fire management was disrupted (Gammage2008) Or alternatively as tree density decreased floweringand therefore pollen production may have increased and leftthe pollen signature unchanged (Williams et al 2006) therelevance of this to woodland eucalypt species is howeverlargely unknown

These sediment cores provide further circumstantial evidencefor a change in the species composition of understorey plantsfollowing settlement including a sharp reduction in abundance ofthe Aboriginal staple food of southern Australia Microserislanceolata or yam daisy (Gott and Murray 1982 Gott 19832005 Clarke 1986) Asteraceae (Liguliflorae) type pollen cannotbe identified to species but except for M lanceolata Sonchusmegalocarpus and Picris angustifolia all members in Tasmaniaare exotic species S megolocarpus and P angustifolia are onlyfound in coastal regions (httpwwwutaseduaudicotkeydicotkeyASTgAsteraceaehtm accessed 17 February 2016)so we attribute this pollen type in the pre-European period toM lanceolataM lanceolata numberswould have declinedwhenactive Aboriginal management ceased and introduced sheepbegan eating them (Gott and Murray 1982) Gott and Murray(1982) point out that the invading European species from theAsteraceae (Liguliflorae) group such as Hypocharis andLeontodon are adapted to living with sheep and cattle andwould have outcompeted M lanceolata

We found an initial decline then increased abundance ofPoaceae in the Midlands following European colonisation Theabundance of Poaceae in association with Eucalyptus andAsteraceae pollen before Europeans arriving supports thenotion that eucalypt savanna was the dominant vegetation atthe time of colonisation (Ellis and Thomas 1988 Fensham 1989Kershaw et al 1994) Thedramatic loss of native grasslands in theMidlands that has occurred in the 20th century (Gilfedder 1990)

cannot be seen in these sediment records because native grassspecies have been replaced by exotic pastures with higher pollenproduction (Smart et al 1979) especially in the post SecondWorld War period when pasture production was sharplyincreased (Kirkpatrick 2007) Potentially quantification of thischange from native to exotic grasses could be approximated infuture work by measuring the size of Poaceae grains in thesediment record larger grains may be produced by introducedgrass taxa (Kenyon and Rutherfurd 1999 Bickford et al 2008)

Changes in the composition of aquatic taxa were mostapparent at Diprose Lagoon where a shift in composition fromBotryococcus spp toMyriophyllum spp in the mid-20th centurywas detected Botryococcus is an algal genus which requiresstanding water (Harle et al 1993) whereasMyriophyllum spp isassociatedwith slowflowingwater or themargins of streamsWeinterpret this transition as reflecting the practise of drainage ofwetlands to increase agricultural land thus producing suitablehabitat for Myriophyllum to establish in impounded andoccasionally flowing artificial drainage channels Across theMidlands 34 of wetland area has been drained and a further23 has been affected by artificial changes in water level(Fensham and Kirkpatrick 1989) In addition to changes inwater levels Botryococcus is vulnerable to eutrophicationwhich may have occurred with increased application ofartificial fertilisers and associated increased sheep density inthe second half of the 20th century (Kellaway 1989Kirkpatrick 2007) A shift towards Myriophyllum and anoverall increase in its abundance throughout the region issomewhat surprising given its sensitivity to salinity anddocumented salinity increases in the Midlands (Kirkpatrickand Harwood 1983 Bastick and Walker 2000 Bastick andLynch 2003) However some species such as M salsugineumare known to be tolerant of shallow saline waters (Thomas 1991)

We found no change in sedimentation rate at Diprose LagoonThis was unexpected becausemost sediment cores from southernAustralia show that a clear increase in sedimentation occurredshortly after the arrival of Europeans in the region (Dodson et al1994a 1994b Mooney and Dodson 2001) Sediment depositionis increasedwhen vegetation cover is removed throughfire or treeremoval which allows erosion to occur An apparent lack ofincreased sedimentation at Diprose Lagoon could be associatedwith increased compaction of upper sediments after sheep wereintroduced and the impact of artificial drainage on catchmenthydrology and sedimentation

Broadly we support the notion that Aboriginal landmanagement maintained eucalypt savanna and that Europeandisruption of Aboriginal management resulted in changed fireregimes and associated ecological changes most notably anincrease in fire a reduction in Allocasuarina and increases inexotic taxa However we cannot resolve the question of thehypothesised switch from a finer grained to coarse grained spatialpattern of landscape burning that caused the loss ofgrasslandndashforest mosaics that lies at the heart of manyenvironmental histories (eg Rolls 1981 Jurskis 2000Gammage 2011) It is possible that integrated studiescombining dendrochronology palynology and charcoalsediment analysis and historical records in the montanegrassland-forest mosaics in Tasmania that are rich in sedimenttraps andhave tree specieswith annual growth rings could further

J Australian Journal of Botany L M Romanin et al

included in Fig S1 available as Supplementary Material to thispaper

Charcoal was present throughout the sediment recordthough the uppermost sediments after AD1990 have very lowbut fairly continuous macro-charcoal deposition A major peakin macro-charcoal preserved in the sediments was found in themid-19th and early 20th century (Fig 2f) After the 1950scharcoal deposition dropped to near pre-European levels Inthe pre-European period there is only a small amount ofcharcoal deposited with gaps in the record

The first axis of a DCA ordination of the terrestrial pollen taxareflects the rate and direction of vegetation change followingEuropean settlement (Figs 2g 3) In the early 20th century there israpid floristic transition largely driven by the establishment ofexotic taxa (Fig 2e) The taxa that drive this transition are thosewith the strongest positive DCA scores and include Coprosmaspp as well as exotic taxa Brassicaceae Salix spp Pinusspp (Fig 3) This transitional period also corresponds to thepeak macro charcoal (Fig 2f) In the earliest part of the record(AD1320ndashAD1530) there is also rapid taxonomic turnover thatcorresponds to the transition from Allocasuarina to Eucalyptusdominance (Fig 2b) The native species that comprise thenegative portion of the DCA first axis are Asteraceae(Tubuliflorae) Plantago spp (native) and Allocasuarinaspp (Fig 3)

Regional network

Total pollen preservation varied substantially between thewetlands Township and Cleveland Lagoons had very good

preservation throughout the length of the core Diprose WhiteNear and Smiths Lagoons had excellent pollen preservation in theupper sections of the sediment profile but much poorer pollenpreservation in the deepest sediments particularly so for Smithsand as noted previously Diprose Lagoon Bells Lagoon hadhigh pollen preservation only in the surface sediment samplethe remainder of this core had poor preservation Summaries ofpollen and charcoal found at the six lagoons not discussed indetail are included in Fig S2 and a list of all pollen and spore taxarecorded in this analysis is included in Table S1 available asSupplementary Material to this paper

We used the first incidence of Pinus spp pollen toapproximate European settlement (Kodela and MacKillop1988) in all of the cores except Cleveland Lagoon because ithad Pinus pollen throughout and overall pre- and post-Europeanperiods were examined Among the remaining 6 cores Pinusconstitutes 6 of the total terrestrial pollen with an interquartilerange of 05ndash8 In these cores the first appearance of Pinusvaried from 4 to 26 cm depth indicating different rates ofsedimentation in the wetlands (Table 1) We used these sixcores for statistical contrasts in pollen between the pre(ndash Pinus) and post (+ Pinus) European periods For these coresthere was no statistically significant change in the proportion ofEucalyptus pollen after the arrival of Europeans across the region(P= 0484) (Fig 4) however there was a statistically significantdecrease in Allocasuarina pollen (P = 001) (Fig 4) In thedeepest sections of four cores Allocasuarina pollen was atleast as common as Eucalyptus There were marked increasesin Poaceae pollen abundance after settlement with Asteraceae(Tubuliflorae) and Chenopodiaceae pollen abundance

3

2

1

0

ndash1

ndash2

ndash4 ndash2 0 2 4

DCA1

DC

A2

Fig 3 Detrended correspondence analysis (DCA) of sediment record from Diprose Lagoon Green squares indicate samples from the pre-European period(as determined by the absence of Pinus spp pollen) red squares indicate samples from the post-European period All the taxa with a strong influence on thefirst two axes are indicated where the following abbreviations are used AP Apiaceae AsteraceaeL Asteraceae (Liguliflorae) AsteraceaeT Asteraceae(Tubuliflorae) AX Amperea xiphoclada Car Caryophyllaceae Ch Chenopodiaceae Cup Cupressaceae DA Dicksonia Antarctica Dod Dodonea sppFab Fabaceae HI Histiopteris incise Indet indeterminate pollen grains Isp Isoetes spp Lin Lindseae LF Lagarostrobos franklinii Mon monolete psilatespore Msp Micrantheum spp NC Nothofagus cunninghamii PA Phyllocladus aspleniifolius Plantago Plantago spp (native) PlantagoL Plantagolanceolata Pom Pomaderris apetala TL Tasmannia lanceolata

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany G

significantly decreasing (Fig 4) Two of the remainingherbaceous taxa examined probably represent a mix of bothexotic and native species native Asteraceae (Liguliflorae) andBrassicaceae species are rare today but they have a strongrepresentation in the early period Both of these groups havebeen replaced by exotic ruderal species as indicated by asignificant increase in Asteraceae (Liguliflorae) type pollenand an increase in the median value of Brassicaceae pollen inthe post-settlement period (Fig 4) Plantago lanceolata is anexclusively exotic species that appeared earlier in the sedimentrecord than the exotic indicator taxa Pinus at two wetlands(Fig 4)

The most abundant aquatic taxon was Botryococcus spp itwas present at each of the six wetlands examined and made up~80 of the aquatic record with no significant change inabundance found between the pre- and post-European periods(P = 049) There was a significant increase in Myriophyllumspp (P = 002) after European settlement (Fig 4)

Minor taxa

Amongst the 6 wetlands rainforest components such asNothofagus cunninghamii Pomaderris apetala type andPhyllocladus aspleniifolius were a minor background element(lt9) of total terrestrial pollen Dry forest species such asAcacia spp Banksia marginata and undifferentiated Fabaceae

species made up a very low proportion of the total terrestrialpollen typically individual species comprised less than 6of the total terrestrial pollen in each core The highest diversitywas found in herbaceous taxa and these were in aggregatemore abundant than the previously mentioned groups Rumexwas the only taxon to show a statistically significant(P = 0003) increase (mean 001 to 17) from the pre- topost-European period (Fig 4) Pteridophyte taxa had constantlow background levels (3) in all the cores Isoeteswas presentin all wetlands in both time periods albeit with highly variablelevels

There were only seven aquatic or wetland species representedin the pollen records None of the minor aquatic species changedin abundance between the two periods Restionaceae was theonly minor wetland family found at all wetlands it made up03ndash6 of the aquatic pollen assemblage

Charcoal

The amount of charcoal preserved at the individualwetlands varied considerably The amount of macro charcoalpreserved in all six sampled wetlands increased from the pre-European (median = 16 pieces cmndash3) to the post settlementperiods (median = 68 pieces cmndash3) (U = 3215 Plt 0001)(Fig 5)

100

80

60

40

20

0P = 001 P = 048 P = 014 P = 001 P lt 005 P = 018 P lt 005 P lt 0001 P lt 001 P = 049 P = 002

pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post

Allo Euc Poa AsterT Cheno Brass AsterL Plan Rum Bot Myr

Per

cent

age

Fig 4 Differences in the pre- and post-European patterns for pollen data derived from a combined analysis of the 6 sediment cores showing the dominantdry forest tree taxa Allocasuarina (Allo) and Eucalyptus (Euc) native herbaceous taxa Poaceae (Poa) Asteraceae (Tubuliflorae) (AsterT) and Chenopodiaceae(Cheno) exotic ruderals Brassicaceae (Brass) Asteraceae (Liguliflorae) (AsterL) Plantago lanceolata (Plan) and Rumex spp (Rum) and aquatic taxaBotryococcus spp (Bot) and Myriophyllum spp (Myr) Pollen values are proportions relative to selected pollen sums (see Fig 2) and have been convertedto percentages for presentation Probablility values based on generalised linear models are shown for each contrast

H Australian Journal of Botany L M Romanin et al

Discussion

Until now there have been few attempts to examine the interplayof vegetation and fire in the dry eastern side of Tasmaniaparticularly the Midlands (Fensham 1989 Jones 2008 vonPlaten et al 2011) Most of the historical research in southernAustralia has relied upon the accounts of early settlers to evaluatethe fire regime of Aboriginal peoples The paucity ofpalaeoecological research in this region has been in part dueto poor pollen preservation and the difficulties in examiningsediment records from seasonally dry lagoons because pollenpreservation can be inadequate (Thomas 1991) and 210Pb datingcan become problematic because drying of sediments mayfacilitate the downcore movement of 210Pb which can lead toan underestimate of age at any given depth (Higuera et al2005) Our approach of using a network of cores enabled us toselect a site with good pollen preservation that was suitable forbuilding a chronology and use other less sensitive sites to drawlandscape-level inferences based on statistical analysis of pollenand charcoal in samples amalgamated using a Pinus pollen as amarker of European settlement

Our study demonstrates marked ecological changes in theTasmanian Midlands following the establishment of a neo-European agricultural landscape These include a change inwoody and herbaceous species abundance associated with

native vegetation clearing and the introduction of exoticpasture species In addition we found a dramatic increase infire activity in the terrestrial environment and changedhydrological conditions in wetlands These findings concurwith our research expectations and the findings of previousresearch (Kershaw et al 1994) In contrast to previousresearch however we found no change in sedimentation rate(Dodson and Mooney 2002) We avoided problems with datingby using Pinus as a proxy for the post-settlement period exoticindicator species have been used to demarcate the arrival ofEuropeans in both Australia (Kershaw et al 1994 Graysonet al 1998 Kenyon and Rutherfurd 1999) and America (vanZant et al 1979)

Palaeoecological studies in southern Australia using pollenand charcoal to reconstruct pre-European fire regimes arebeginning to reveal the dynamic nature of fire frequency andimpact on vegetation during the Holocene although theinterpretation of these data is not clear cut (Mooney et al2001 Fletcher et al 2014) Charcoal peaks in sediments aremore likely to reflect high intensity fires (Whitlock and Larsen2001 Higuera et al 2010) However the absence of charcoal in asedimentary record does not necessary equate to an absence offire Low intensity fires produce high levels of incompletelyburned particulate matter but this effect is counteractedbecause these fires are typically small in extent leaving asmall signature in the pollen record (Whitlock and Larsen2001 Higuera et al 2010) Indeed Mooney et al (2011)argued that the presence or absence of sedimentary charcoalmay be a poor indicator of low intensity fires thought tocharacterise Aboriginal burning patterns Studies of traditionalAboriginal fire usage in northern Australia have demonstrated aregime of frequent and patchy fire (Bowman et al 2001 2004Bliege Bird et al 2008 Trauernicht et al 2015) They used fireto maintain open grassy landscapes to increase the abundanceof game species (Murphy and Bowman 2007 Codding et al2014) and motivations are assumed to be similar in southernAustralia Dendrochronological evidence shows that dry foreststo the east of the study region persisted through increasedburning after Europeans arrival from ~07 fires per decadeunder Aboriginal management to 17 per decade in the postsettlement period (von Platen et al 2011)

Previous studies on mainland south-east Australia havefound an increase in charcoal after European settlement(Johnson 2000 Mooney and Dodson 2001 Mooney et al2001 2011) Likewise the charcoal from Diprose Lagoonrecord showed that there was a sharp increase in charcoalproduction concurrent with the appearance of Pinus in thesediment record We suggest the source of this charcoal is thecombined effects of using fire to clear land and settinggeographically large grass fires to stimulate pasture production(Kirkpatrick 2007) In theDiprose core charcoal production in the1950s declined possibly reflecting a dwindling of tree-clearingand associated burning and greater use of mechanisedagriculture

The greater abundance of Allocasuarina in the period beforeEuropean settlement suggests that fire regimes were favourablefor this species and other relatively fire sensitive species andthat clearing grazing and an increase in burning by Europeanscaused Allocasuarina to decline This inference is supported

30

25

20

15

10

05

00

pre post

log(

char

coal

(no

cm

ndash3))

Fig 5 Boxplots of the differences in the pre- and post-European periodsfor the macro-charcoal record of six wetlandsrsquo sediment cores Charcoalconcentrations of the two periods are log-transformed and differ at the 0001level according to a MannndashWhitney test The pre-European abundance isshown on the left post-European on the right

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany I

by the post-fire regeneration strategies of the two Allocasuarinaspecies found in theMidlandsA littoralis is a serotinous obligateseeder that has low fire resistance (Morrison and Renwick 2000Burley et al 2007) while A verticillata is also serotinous but hasa limited capacity to resprout from rootstock (Singh and Geissler1985Ladd1988)Both species regenerate prolifically afterfire asthey release a heavy rain of seed onto disturbed ground (Ladd1988) but require at least 5ndash6 years fire free to begin producing itsserotinous seed (Hueneke 1976 Singh and Geissler 1985) Asimilar pattern of early and rapid decline of Allocasuarina hasbeen found in other temperate savannas of southern Australia(Kershaw et al 1994 Mooney and Dodson 2001 Bickford et al2008 Jones 2008)

We found no significant change in the abundance ofEucalyptus pollen following European settlement This lack ofchange inEucalyptuswas unexpected given thewell documentedloss of tree cover in the Midlands which has occurred sinceEuropean settlement (Fensham and Kirkpatrick 1989 Prior et al2013 Romanin et al 2015) and may be an artefact of theamalgamation of data from multiple cores In highly disturbedagricultural areas of New SouthWales a similar lack of change inthe pollen signature of Eucalyptus has been identified with thesuggestion that clearing of woodland trees has been compensatedby theplantingof other species of eucalypts aroundhomesteadsoras windbreaks (Kershaw et al 1994) A lack of overall changemay also be evidence of thickening of native remnant vegetationafter Aboriginal fire management was disrupted (Gammage2008) Or alternatively as tree density decreased floweringand therefore pollen production may have increased and leftthe pollen signature unchanged (Williams et al 2006) therelevance of this to woodland eucalypt species is howeverlargely unknown

These sediment cores provide further circumstantial evidencefor a change in the species composition of understorey plantsfollowing settlement including a sharp reduction in abundance ofthe Aboriginal staple food of southern Australia Microserislanceolata or yam daisy (Gott and Murray 1982 Gott 19832005 Clarke 1986) Asteraceae (Liguliflorae) type pollen cannotbe identified to species but except for M lanceolata Sonchusmegalocarpus and Picris angustifolia all members in Tasmaniaare exotic species S megolocarpus and P angustifolia are onlyfound in coastal regions (httpwwwutaseduaudicotkeydicotkeyASTgAsteraceaehtm accessed 17 February 2016)so we attribute this pollen type in the pre-European period toM lanceolataM lanceolata numberswould have declinedwhenactive Aboriginal management ceased and introduced sheepbegan eating them (Gott and Murray 1982) Gott and Murray(1982) point out that the invading European species from theAsteraceae (Liguliflorae) group such as Hypocharis andLeontodon are adapted to living with sheep and cattle andwould have outcompeted M lanceolata

We found an initial decline then increased abundance ofPoaceae in the Midlands following European colonisation Theabundance of Poaceae in association with Eucalyptus andAsteraceae pollen before Europeans arriving supports thenotion that eucalypt savanna was the dominant vegetation atthe time of colonisation (Ellis and Thomas 1988 Fensham 1989Kershaw et al 1994) Thedramatic loss of native grasslands in theMidlands that has occurred in the 20th century (Gilfedder 1990)

cannot be seen in these sediment records because native grassspecies have been replaced by exotic pastures with higher pollenproduction (Smart et al 1979) especially in the post SecondWorld War period when pasture production was sharplyincreased (Kirkpatrick 2007) Potentially quantification of thischange from native to exotic grasses could be approximated infuture work by measuring the size of Poaceae grains in thesediment record larger grains may be produced by introducedgrass taxa (Kenyon and Rutherfurd 1999 Bickford et al 2008)

Changes in the composition of aquatic taxa were mostapparent at Diprose Lagoon where a shift in composition fromBotryococcus spp toMyriophyllum spp in the mid-20th centurywas detected Botryococcus is an algal genus which requiresstanding water (Harle et al 1993) whereasMyriophyllum spp isassociatedwith slowflowingwater or themargins of streamsWeinterpret this transition as reflecting the practise of drainage ofwetlands to increase agricultural land thus producing suitablehabitat for Myriophyllum to establish in impounded andoccasionally flowing artificial drainage channels Across theMidlands 34 of wetland area has been drained and a further23 has been affected by artificial changes in water level(Fensham and Kirkpatrick 1989) In addition to changes inwater levels Botryococcus is vulnerable to eutrophicationwhich may have occurred with increased application ofartificial fertilisers and associated increased sheep density inthe second half of the 20th century (Kellaway 1989Kirkpatrick 2007) A shift towards Myriophyllum and anoverall increase in its abundance throughout the region issomewhat surprising given its sensitivity to salinity anddocumented salinity increases in the Midlands (Kirkpatrickand Harwood 1983 Bastick and Walker 2000 Bastick andLynch 2003) However some species such as M salsugineumare known to be tolerant of shallow saline waters (Thomas 1991)

We found no change in sedimentation rate at Diprose LagoonThis was unexpected becausemost sediment cores from southernAustralia show that a clear increase in sedimentation occurredshortly after the arrival of Europeans in the region (Dodson et al1994a 1994b Mooney and Dodson 2001) Sediment depositionis increasedwhen vegetation cover is removed throughfire or treeremoval which allows erosion to occur An apparent lack ofincreased sedimentation at Diprose Lagoon could be associatedwith increased compaction of upper sediments after sheep wereintroduced and the impact of artificial drainage on catchmenthydrology and sedimentation

Broadly we support the notion that Aboriginal landmanagement maintained eucalypt savanna and that Europeandisruption of Aboriginal management resulted in changed fireregimes and associated ecological changes most notably anincrease in fire a reduction in Allocasuarina and increases inexotic taxa However we cannot resolve the question of thehypothesised switch from a finer grained to coarse grained spatialpattern of landscape burning that caused the loss ofgrasslandndashforest mosaics that lies at the heart of manyenvironmental histories (eg Rolls 1981 Jurskis 2000Gammage 2011) It is possible that integrated studiescombining dendrochronology palynology and charcoalsediment analysis and historical records in the montanegrassland-forest mosaics in Tasmania that are rich in sedimenttraps andhave tree specieswith annual growth rings could further

J Australian Journal of Botany L M Romanin et al

significantly decreasing (Fig 4) Two of the remainingherbaceous taxa examined probably represent a mix of bothexotic and native species native Asteraceae (Liguliflorae) andBrassicaceae species are rare today but they have a strongrepresentation in the early period Both of these groups havebeen replaced by exotic ruderal species as indicated by asignificant increase in Asteraceae (Liguliflorae) type pollenand an increase in the median value of Brassicaceae pollen inthe post-settlement period (Fig 4) Plantago lanceolata is anexclusively exotic species that appeared earlier in the sedimentrecord than the exotic indicator taxa Pinus at two wetlands(Fig 4)

The most abundant aquatic taxon was Botryococcus spp itwas present at each of the six wetlands examined and made up~80 of the aquatic record with no significant change inabundance found between the pre- and post-European periods(P = 049) There was a significant increase in Myriophyllumspp (P = 002) after European settlement (Fig 4)

Minor taxa

Amongst the 6 wetlands rainforest components such asNothofagus cunninghamii Pomaderris apetala type andPhyllocladus aspleniifolius were a minor background element(lt9) of total terrestrial pollen Dry forest species such asAcacia spp Banksia marginata and undifferentiated Fabaceae

species made up a very low proportion of the total terrestrialpollen typically individual species comprised less than 6of the total terrestrial pollen in each core The highest diversitywas found in herbaceous taxa and these were in aggregatemore abundant than the previously mentioned groups Rumexwas the only taxon to show a statistically significant(P = 0003) increase (mean 001 to 17) from the pre- topost-European period (Fig 4) Pteridophyte taxa had constantlow background levels (3) in all the cores Isoeteswas presentin all wetlands in both time periods albeit with highly variablelevels

There were only seven aquatic or wetland species representedin the pollen records None of the minor aquatic species changedin abundance between the two periods Restionaceae was theonly minor wetland family found at all wetlands it made up03ndash6 of the aquatic pollen assemblage

Charcoal

The amount of charcoal preserved at the individualwetlands varied considerably The amount of macro charcoalpreserved in all six sampled wetlands increased from the pre-European (median = 16 pieces cmndash3) to the post settlementperiods (median = 68 pieces cmndash3) (U = 3215 Plt 0001)(Fig 5)

100

80

60

40

20

0P = 001 P = 048 P = 014 P = 001 P lt 005 P = 018 P lt 005 P lt 0001 P lt 001 P = 049 P = 002

pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post pre post

Allo Euc Poa AsterT Cheno Brass AsterL Plan Rum Bot Myr

Per

cent

age

Fig 4 Differences in the pre- and post-European patterns for pollen data derived from a combined analysis of the 6 sediment cores showing the dominantdry forest tree taxa Allocasuarina (Allo) and Eucalyptus (Euc) native herbaceous taxa Poaceae (Poa) Asteraceae (Tubuliflorae) (AsterT) and Chenopodiaceae(Cheno) exotic ruderals Brassicaceae (Brass) Asteraceae (Liguliflorae) (AsterL) Plantago lanceolata (Plan) and Rumex spp (Rum) and aquatic taxaBotryococcus spp (Bot) and Myriophyllum spp (Myr) Pollen values are proportions relative to selected pollen sums (see Fig 2) and have been convertedto percentages for presentation Probablility values based on generalised linear models are shown for each contrast

H Australian Journal of Botany L M Romanin et al

Discussion

Until now there have been few attempts to examine the interplayof vegetation and fire in the dry eastern side of Tasmaniaparticularly the Midlands (Fensham 1989 Jones 2008 vonPlaten et al 2011) Most of the historical research in southernAustralia has relied upon the accounts of early settlers to evaluatethe fire regime of Aboriginal peoples The paucity ofpalaeoecological research in this region has been in part dueto poor pollen preservation and the difficulties in examiningsediment records from seasonally dry lagoons because pollenpreservation can be inadequate (Thomas 1991) and 210Pb datingcan become problematic because drying of sediments mayfacilitate the downcore movement of 210Pb which can lead toan underestimate of age at any given depth (Higuera et al2005) Our approach of using a network of cores enabled us toselect a site with good pollen preservation that was suitable forbuilding a chronology and use other less sensitive sites to drawlandscape-level inferences based on statistical analysis of pollenand charcoal in samples amalgamated using a Pinus pollen as amarker of European settlement

Our study demonstrates marked ecological changes in theTasmanian Midlands following the establishment of a neo-European agricultural landscape These include a change inwoody and herbaceous species abundance associated with

native vegetation clearing and the introduction of exoticpasture species In addition we found a dramatic increase infire activity in the terrestrial environment and changedhydrological conditions in wetlands These findings concurwith our research expectations and the findings of previousresearch (Kershaw et al 1994) In contrast to previousresearch however we found no change in sedimentation rate(Dodson and Mooney 2002) We avoided problems with datingby using Pinus as a proxy for the post-settlement period exoticindicator species have been used to demarcate the arrival ofEuropeans in both Australia (Kershaw et al 1994 Graysonet al 1998 Kenyon and Rutherfurd 1999) and America (vanZant et al 1979)

Palaeoecological studies in southern Australia using pollenand charcoal to reconstruct pre-European fire regimes arebeginning to reveal the dynamic nature of fire frequency andimpact on vegetation during the Holocene although theinterpretation of these data is not clear cut (Mooney et al2001 Fletcher et al 2014) Charcoal peaks in sediments aremore likely to reflect high intensity fires (Whitlock and Larsen2001 Higuera et al 2010) However the absence of charcoal in asedimentary record does not necessary equate to an absence offire Low intensity fires produce high levels of incompletelyburned particulate matter but this effect is counteractedbecause these fires are typically small in extent leaving asmall signature in the pollen record (Whitlock and Larsen2001 Higuera et al 2010) Indeed Mooney et al (2011)argued that the presence or absence of sedimentary charcoalmay be a poor indicator of low intensity fires thought tocharacterise Aboriginal burning patterns Studies of traditionalAboriginal fire usage in northern Australia have demonstrated aregime of frequent and patchy fire (Bowman et al 2001 2004Bliege Bird et al 2008 Trauernicht et al 2015) They used fireto maintain open grassy landscapes to increase the abundanceof game species (Murphy and Bowman 2007 Codding et al2014) and motivations are assumed to be similar in southernAustralia Dendrochronological evidence shows that dry foreststo the east of the study region persisted through increasedburning after Europeans arrival from ~07 fires per decadeunder Aboriginal management to 17 per decade in the postsettlement period (von Platen et al 2011)

Previous studies on mainland south-east Australia havefound an increase in charcoal after European settlement(Johnson 2000 Mooney and Dodson 2001 Mooney et al2001 2011) Likewise the charcoal from Diprose Lagoonrecord showed that there was a sharp increase in charcoalproduction concurrent with the appearance of Pinus in thesediment record We suggest the source of this charcoal is thecombined effects of using fire to clear land and settinggeographically large grass fires to stimulate pasture production(Kirkpatrick 2007) In theDiprose core charcoal production in the1950s declined possibly reflecting a dwindling of tree-clearingand associated burning and greater use of mechanisedagriculture

The greater abundance of Allocasuarina in the period beforeEuropean settlement suggests that fire regimes were favourablefor this species and other relatively fire sensitive species andthat clearing grazing and an increase in burning by Europeanscaused Allocasuarina to decline This inference is supported

30

25

20

15

10

05

00

pre post

log(

char

coal

(no

cm

ndash3))

Fig 5 Boxplots of the differences in the pre- and post-European periodsfor the macro-charcoal record of six wetlandsrsquo sediment cores Charcoalconcentrations of the two periods are log-transformed and differ at the 0001level according to a MannndashWhitney test The pre-European abundance isshown on the left post-European on the right

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany I

by the post-fire regeneration strategies of the two Allocasuarinaspecies found in theMidlandsA littoralis is a serotinous obligateseeder that has low fire resistance (Morrison and Renwick 2000Burley et al 2007) while A verticillata is also serotinous but hasa limited capacity to resprout from rootstock (Singh and Geissler1985Ladd1988)Both species regenerate prolifically afterfire asthey release a heavy rain of seed onto disturbed ground (Ladd1988) but require at least 5ndash6 years fire free to begin producing itsserotinous seed (Hueneke 1976 Singh and Geissler 1985) Asimilar pattern of early and rapid decline of Allocasuarina hasbeen found in other temperate savannas of southern Australia(Kershaw et al 1994 Mooney and Dodson 2001 Bickford et al2008 Jones 2008)

We found no significant change in the abundance ofEucalyptus pollen following European settlement This lack ofchange inEucalyptuswas unexpected given thewell documentedloss of tree cover in the Midlands which has occurred sinceEuropean settlement (Fensham and Kirkpatrick 1989 Prior et al2013 Romanin et al 2015) and may be an artefact of theamalgamation of data from multiple cores In highly disturbedagricultural areas of New SouthWales a similar lack of change inthe pollen signature of Eucalyptus has been identified with thesuggestion that clearing of woodland trees has been compensatedby theplantingof other species of eucalypts aroundhomesteadsoras windbreaks (Kershaw et al 1994) A lack of overall changemay also be evidence of thickening of native remnant vegetationafter Aboriginal fire management was disrupted (Gammage2008) Or alternatively as tree density decreased floweringand therefore pollen production may have increased and leftthe pollen signature unchanged (Williams et al 2006) therelevance of this to woodland eucalypt species is howeverlargely unknown

These sediment cores provide further circumstantial evidencefor a change in the species composition of understorey plantsfollowing settlement including a sharp reduction in abundance ofthe Aboriginal staple food of southern Australia Microserislanceolata or yam daisy (Gott and Murray 1982 Gott 19832005 Clarke 1986) Asteraceae (Liguliflorae) type pollen cannotbe identified to species but except for M lanceolata Sonchusmegalocarpus and Picris angustifolia all members in Tasmaniaare exotic species S megolocarpus and P angustifolia are onlyfound in coastal regions (httpwwwutaseduaudicotkeydicotkeyASTgAsteraceaehtm accessed 17 February 2016)so we attribute this pollen type in the pre-European period toM lanceolataM lanceolata numberswould have declinedwhenactive Aboriginal management ceased and introduced sheepbegan eating them (Gott and Murray 1982) Gott and Murray(1982) point out that the invading European species from theAsteraceae (Liguliflorae) group such as Hypocharis andLeontodon are adapted to living with sheep and cattle andwould have outcompeted M lanceolata

We found an initial decline then increased abundance ofPoaceae in the Midlands following European colonisation Theabundance of Poaceae in association with Eucalyptus andAsteraceae pollen before Europeans arriving supports thenotion that eucalypt savanna was the dominant vegetation atthe time of colonisation (Ellis and Thomas 1988 Fensham 1989Kershaw et al 1994) Thedramatic loss of native grasslands in theMidlands that has occurred in the 20th century (Gilfedder 1990)

cannot be seen in these sediment records because native grassspecies have been replaced by exotic pastures with higher pollenproduction (Smart et al 1979) especially in the post SecondWorld War period when pasture production was sharplyincreased (Kirkpatrick 2007) Potentially quantification of thischange from native to exotic grasses could be approximated infuture work by measuring the size of Poaceae grains in thesediment record larger grains may be produced by introducedgrass taxa (Kenyon and Rutherfurd 1999 Bickford et al 2008)

Changes in the composition of aquatic taxa were mostapparent at Diprose Lagoon where a shift in composition fromBotryococcus spp toMyriophyllum spp in the mid-20th centurywas detected Botryococcus is an algal genus which requiresstanding water (Harle et al 1993) whereasMyriophyllum spp isassociatedwith slowflowingwater or themargins of streamsWeinterpret this transition as reflecting the practise of drainage ofwetlands to increase agricultural land thus producing suitablehabitat for Myriophyllum to establish in impounded andoccasionally flowing artificial drainage channels Across theMidlands 34 of wetland area has been drained and a further23 has been affected by artificial changes in water level(Fensham and Kirkpatrick 1989) In addition to changes inwater levels Botryococcus is vulnerable to eutrophicationwhich may have occurred with increased application ofartificial fertilisers and associated increased sheep density inthe second half of the 20th century (Kellaway 1989Kirkpatrick 2007) A shift towards Myriophyllum and anoverall increase in its abundance throughout the region issomewhat surprising given its sensitivity to salinity anddocumented salinity increases in the Midlands (Kirkpatrickand Harwood 1983 Bastick and Walker 2000 Bastick andLynch 2003) However some species such as M salsugineumare known to be tolerant of shallow saline waters (Thomas 1991)

We found no change in sedimentation rate at Diprose LagoonThis was unexpected becausemost sediment cores from southernAustralia show that a clear increase in sedimentation occurredshortly after the arrival of Europeans in the region (Dodson et al1994a 1994b Mooney and Dodson 2001) Sediment depositionis increasedwhen vegetation cover is removed throughfire or treeremoval which allows erosion to occur An apparent lack ofincreased sedimentation at Diprose Lagoon could be associatedwith increased compaction of upper sediments after sheep wereintroduced and the impact of artificial drainage on catchmenthydrology and sedimentation

Broadly we support the notion that Aboriginal landmanagement maintained eucalypt savanna and that Europeandisruption of Aboriginal management resulted in changed fireregimes and associated ecological changes most notably anincrease in fire a reduction in Allocasuarina and increases inexotic taxa However we cannot resolve the question of thehypothesised switch from a finer grained to coarse grained spatialpattern of landscape burning that caused the loss ofgrasslandndashforest mosaics that lies at the heart of manyenvironmental histories (eg Rolls 1981 Jurskis 2000Gammage 2011) It is possible that integrated studiescombining dendrochronology palynology and charcoalsediment analysis and historical records in the montanegrassland-forest mosaics in Tasmania that are rich in sedimenttraps andhave tree specieswith annual growth rings could further

J Australian Journal of Botany L M Romanin et al

Discussion

Until now there have been few attempts to examine the interplayof vegetation and fire in the dry eastern side of Tasmaniaparticularly the Midlands (Fensham 1989 Jones 2008 vonPlaten et al 2011) Most of the historical research in southernAustralia has relied upon the accounts of early settlers to evaluatethe fire regime of Aboriginal peoples The paucity ofpalaeoecological research in this region has been in part dueto poor pollen preservation and the difficulties in examiningsediment records from seasonally dry lagoons because pollenpreservation can be inadequate (Thomas 1991) and 210Pb datingcan become problematic because drying of sediments mayfacilitate the downcore movement of 210Pb which can lead toan underestimate of age at any given depth (Higuera et al2005) Our approach of using a network of cores enabled us toselect a site with good pollen preservation that was suitable forbuilding a chronology and use other less sensitive sites to drawlandscape-level inferences based on statistical analysis of pollenand charcoal in samples amalgamated using a Pinus pollen as amarker of European settlement

Our study demonstrates marked ecological changes in theTasmanian Midlands following the establishment of a neo-European agricultural landscape These include a change inwoody and herbaceous species abundance associated with

native vegetation clearing and the introduction of exoticpasture species In addition we found a dramatic increase infire activity in the terrestrial environment and changedhydrological conditions in wetlands These findings concurwith our research expectations and the findings of previousresearch (Kershaw et al 1994) In contrast to previousresearch however we found no change in sedimentation rate(Dodson and Mooney 2002) We avoided problems with datingby using Pinus as a proxy for the post-settlement period exoticindicator species have been used to demarcate the arrival ofEuropeans in both Australia (Kershaw et al 1994 Graysonet al 1998 Kenyon and Rutherfurd 1999) and America (vanZant et al 1979)

Palaeoecological studies in southern Australia using pollenand charcoal to reconstruct pre-European fire regimes arebeginning to reveal the dynamic nature of fire frequency andimpact on vegetation during the Holocene although theinterpretation of these data is not clear cut (Mooney et al2001 Fletcher et al 2014) Charcoal peaks in sediments aremore likely to reflect high intensity fires (Whitlock and Larsen2001 Higuera et al 2010) However the absence of charcoal in asedimentary record does not necessary equate to an absence offire Low intensity fires produce high levels of incompletelyburned particulate matter but this effect is counteractedbecause these fires are typically small in extent leaving asmall signature in the pollen record (Whitlock and Larsen2001 Higuera et al 2010) Indeed Mooney et al (2011)argued that the presence or absence of sedimentary charcoalmay be a poor indicator of low intensity fires thought tocharacterise Aboriginal burning patterns Studies of traditionalAboriginal fire usage in northern Australia have demonstrated aregime of frequent and patchy fire (Bowman et al 2001 2004Bliege Bird et al 2008 Trauernicht et al 2015) They used fireto maintain open grassy landscapes to increase the abundanceof game species (Murphy and Bowman 2007 Codding et al2014) and motivations are assumed to be similar in southernAustralia Dendrochronological evidence shows that dry foreststo the east of the study region persisted through increasedburning after Europeans arrival from ~07 fires per decadeunder Aboriginal management to 17 per decade in the postsettlement period (von Platen et al 2011)

Previous studies on mainland south-east Australia havefound an increase in charcoal after European settlement(Johnson 2000 Mooney and Dodson 2001 Mooney et al2001 2011) Likewise the charcoal from Diprose Lagoonrecord showed that there was a sharp increase in charcoalproduction concurrent with the appearance of Pinus in thesediment record We suggest the source of this charcoal is thecombined effects of using fire to clear land and settinggeographically large grass fires to stimulate pasture production(Kirkpatrick 2007) In theDiprose core charcoal production in the1950s declined possibly reflecting a dwindling of tree-clearingand associated burning and greater use of mechanisedagriculture

The greater abundance of Allocasuarina in the period beforeEuropean settlement suggests that fire regimes were favourablefor this species and other relatively fire sensitive species andthat clearing grazing and an increase in burning by Europeanscaused Allocasuarina to decline This inference is supported

30

25

20

15

10

05

00

pre post

log(

char

coal

(no

cm

ndash3))

Fig 5 Boxplots of the differences in the pre- and post-European periodsfor the macro-charcoal record of six wetlandsrsquo sediment cores Charcoalconcentrations of the two periods are log-transformed and differ at the 0001level according to a MannndashWhitney test The pre-European abundance isshown on the left post-European on the right

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany I

by the post-fire regeneration strategies of the two Allocasuarinaspecies found in theMidlandsA littoralis is a serotinous obligateseeder that has low fire resistance (Morrison and Renwick 2000Burley et al 2007) while A verticillata is also serotinous but hasa limited capacity to resprout from rootstock (Singh and Geissler1985Ladd1988)Both species regenerate prolifically afterfire asthey release a heavy rain of seed onto disturbed ground (Ladd1988) but require at least 5ndash6 years fire free to begin producing itsserotinous seed (Hueneke 1976 Singh and Geissler 1985) Asimilar pattern of early and rapid decline of Allocasuarina hasbeen found in other temperate savannas of southern Australia(Kershaw et al 1994 Mooney and Dodson 2001 Bickford et al2008 Jones 2008)

We found no significant change in the abundance ofEucalyptus pollen following European settlement This lack ofchange inEucalyptuswas unexpected given thewell documentedloss of tree cover in the Midlands which has occurred sinceEuropean settlement (Fensham and Kirkpatrick 1989 Prior et al2013 Romanin et al 2015) and may be an artefact of theamalgamation of data from multiple cores In highly disturbedagricultural areas of New SouthWales a similar lack of change inthe pollen signature of Eucalyptus has been identified with thesuggestion that clearing of woodland trees has been compensatedby theplantingof other species of eucalypts aroundhomesteadsoras windbreaks (Kershaw et al 1994) A lack of overall changemay also be evidence of thickening of native remnant vegetationafter Aboriginal fire management was disrupted (Gammage2008) Or alternatively as tree density decreased floweringand therefore pollen production may have increased and leftthe pollen signature unchanged (Williams et al 2006) therelevance of this to woodland eucalypt species is howeverlargely unknown

These sediment cores provide further circumstantial evidencefor a change in the species composition of understorey plantsfollowing settlement including a sharp reduction in abundance ofthe Aboriginal staple food of southern Australia Microserislanceolata or yam daisy (Gott and Murray 1982 Gott 19832005 Clarke 1986) Asteraceae (Liguliflorae) type pollen cannotbe identified to species but except for M lanceolata Sonchusmegalocarpus and Picris angustifolia all members in Tasmaniaare exotic species S megolocarpus and P angustifolia are onlyfound in coastal regions (httpwwwutaseduaudicotkeydicotkeyASTgAsteraceaehtm accessed 17 February 2016)so we attribute this pollen type in the pre-European period toM lanceolataM lanceolata numberswould have declinedwhenactive Aboriginal management ceased and introduced sheepbegan eating them (Gott and Murray 1982) Gott and Murray(1982) point out that the invading European species from theAsteraceae (Liguliflorae) group such as Hypocharis andLeontodon are adapted to living with sheep and cattle andwould have outcompeted M lanceolata

We found an initial decline then increased abundance ofPoaceae in the Midlands following European colonisation Theabundance of Poaceae in association with Eucalyptus andAsteraceae pollen before Europeans arriving supports thenotion that eucalypt savanna was the dominant vegetation atthe time of colonisation (Ellis and Thomas 1988 Fensham 1989Kershaw et al 1994) Thedramatic loss of native grasslands in theMidlands that has occurred in the 20th century (Gilfedder 1990)

cannot be seen in these sediment records because native grassspecies have been replaced by exotic pastures with higher pollenproduction (Smart et al 1979) especially in the post SecondWorld War period when pasture production was sharplyincreased (Kirkpatrick 2007) Potentially quantification of thischange from native to exotic grasses could be approximated infuture work by measuring the size of Poaceae grains in thesediment record larger grains may be produced by introducedgrass taxa (Kenyon and Rutherfurd 1999 Bickford et al 2008)

Changes in the composition of aquatic taxa were mostapparent at Diprose Lagoon where a shift in composition fromBotryococcus spp toMyriophyllum spp in the mid-20th centurywas detected Botryococcus is an algal genus which requiresstanding water (Harle et al 1993) whereasMyriophyllum spp isassociatedwith slowflowingwater or themargins of streamsWeinterpret this transition as reflecting the practise of drainage ofwetlands to increase agricultural land thus producing suitablehabitat for Myriophyllum to establish in impounded andoccasionally flowing artificial drainage channels Across theMidlands 34 of wetland area has been drained and a further23 has been affected by artificial changes in water level(Fensham and Kirkpatrick 1989) In addition to changes inwater levels Botryococcus is vulnerable to eutrophicationwhich may have occurred with increased application ofartificial fertilisers and associated increased sheep density inthe second half of the 20th century (Kellaway 1989Kirkpatrick 2007) A shift towards Myriophyllum and anoverall increase in its abundance throughout the region issomewhat surprising given its sensitivity to salinity anddocumented salinity increases in the Midlands (Kirkpatrickand Harwood 1983 Bastick and Walker 2000 Bastick andLynch 2003) However some species such as M salsugineumare known to be tolerant of shallow saline waters (Thomas 1991)

We found no change in sedimentation rate at Diprose LagoonThis was unexpected becausemost sediment cores from southernAustralia show that a clear increase in sedimentation occurredshortly after the arrival of Europeans in the region (Dodson et al1994a 1994b Mooney and Dodson 2001) Sediment depositionis increasedwhen vegetation cover is removed throughfire or treeremoval which allows erosion to occur An apparent lack ofincreased sedimentation at Diprose Lagoon could be associatedwith increased compaction of upper sediments after sheep wereintroduced and the impact of artificial drainage on catchmenthydrology and sedimentation

Broadly we support the notion that Aboriginal landmanagement maintained eucalypt savanna and that Europeandisruption of Aboriginal management resulted in changed fireregimes and associated ecological changes most notably anincrease in fire a reduction in Allocasuarina and increases inexotic taxa However we cannot resolve the question of thehypothesised switch from a finer grained to coarse grained spatialpattern of landscape burning that caused the loss ofgrasslandndashforest mosaics that lies at the heart of manyenvironmental histories (eg Rolls 1981 Jurskis 2000Gammage 2011) It is possible that integrated studiescombining dendrochronology palynology and charcoalsediment analysis and historical records in the montanegrassland-forest mosaics in Tasmania that are rich in sedimenttraps andhave tree specieswith annual growth rings could further

J Australian Journal of Botany L M Romanin et al

by the post-fire regeneration strategies of the two Allocasuarinaspecies found in theMidlandsA littoralis is a serotinous obligateseeder that has low fire resistance (Morrison and Renwick 2000Burley et al 2007) while A verticillata is also serotinous but hasa limited capacity to resprout from rootstock (Singh and Geissler1985Ladd1988)Both species regenerate prolifically afterfire asthey release a heavy rain of seed onto disturbed ground (Ladd1988) but require at least 5ndash6 years fire free to begin producing itsserotinous seed (Hueneke 1976 Singh and Geissler 1985) Asimilar pattern of early and rapid decline of Allocasuarina hasbeen found in other temperate savannas of southern Australia(Kershaw et al 1994 Mooney and Dodson 2001 Bickford et al2008 Jones 2008)

We found no significant change in the abundance ofEucalyptus pollen following European settlement This lack ofchange inEucalyptuswas unexpected given thewell documentedloss of tree cover in the Midlands which has occurred sinceEuropean settlement (Fensham and Kirkpatrick 1989 Prior et al2013 Romanin et al 2015) and may be an artefact of theamalgamation of data from multiple cores In highly disturbedagricultural areas of New SouthWales a similar lack of change inthe pollen signature of Eucalyptus has been identified with thesuggestion that clearing of woodland trees has been compensatedby theplantingof other species of eucalypts aroundhomesteadsoras windbreaks (Kershaw et al 1994) A lack of overall changemay also be evidence of thickening of native remnant vegetationafter Aboriginal fire management was disrupted (Gammage2008) Or alternatively as tree density decreased floweringand therefore pollen production may have increased and leftthe pollen signature unchanged (Williams et al 2006) therelevance of this to woodland eucalypt species is howeverlargely unknown

These sediment cores provide further circumstantial evidencefor a change in the species composition of understorey plantsfollowing settlement including a sharp reduction in abundance ofthe Aboriginal staple food of southern Australia Microserislanceolata or yam daisy (Gott and Murray 1982 Gott 19832005 Clarke 1986) Asteraceae (Liguliflorae) type pollen cannotbe identified to species but except for M lanceolata Sonchusmegalocarpus and Picris angustifolia all members in Tasmaniaare exotic species S megolocarpus and P angustifolia are onlyfound in coastal regions (httpwwwutaseduaudicotkeydicotkeyASTgAsteraceaehtm accessed 17 February 2016)so we attribute this pollen type in the pre-European period toM lanceolataM lanceolata numberswould have declinedwhenactive Aboriginal management ceased and introduced sheepbegan eating them (Gott and Murray 1982) Gott and Murray(1982) point out that the invading European species from theAsteraceae (Liguliflorae) group such as Hypocharis andLeontodon are adapted to living with sheep and cattle andwould have outcompeted M lanceolata

We found an initial decline then increased abundance ofPoaceae in the Midlands following European colonisation Theabundance of Poaceae in association with Eucalyptus andAsteraceae pollen before Europeans arriving supports thenotion that eucalypt savanna was the dominant vegetation atthe time of colonisation (Ellis and Thomas 1988 Fensham 1989Kershaw et al 1994) Thedramatic loss of native grasslands in theMidlands that has occurred in the 20th century (Gilfedder 1990)

cannot be seen in these sediment records because native grassspecies have been replaced by exotic pastures with higher pollenproduction (Smart et al 1979) especially in the post SecondWorld War period when pasture production was sharplyincreased (Kirkpatrick 2007) Potentially quantification of thischange from native to exotic grasses could be approximated infuture work by measuring the size of Poaceae grains in thesediment record larger grains may be produced by introducedgrass taxa (Kenyon and Rutherfurd 1999 Bickford et al 2008)

Changes in the composition of aquatic taxa were mostapparent at Diprose Lagoon where a shift in composition fromBotryococcus spp toMyriophyllum spp in the mid-20th centurywas detected Botryococcus is an algal genus which requiresstanding water (Harle et al 1993) whereasMyriophyllum spp isassociatedwith slowflowingwater or themargins of streamsWeinterpret this transition as reflecting the practise of drainage ofwetlands to increase agricultural land thus producing suitablehabitat for Myriophyllum to establish in impounded andoccasionally flowing artificial drainage channels Across theMidlands 34 of wetland area has been drained and a further23 has been affected by artificial changes in water level(Fensham and Kirkpatrick 1989) In addition to changes inwater levels Botryococcus is vulnerable to eutrophicationwhich may have occurred with increased application ofartificial fertilisers and associated increased sheep density inthe second half of the 20th century (Kellaway 1989Kirkpatrick 2007) A shift towards Myriophyllum and anoverall increase in its abundance throughout the region issomewhat surprising given its sensitivity to salinity anddocumented salinity increases in the Midlands (Kirkpatrickand Harwood 1983 Bastick and Walker 2000 Bastick andLynch 2003) However some species such as M salsugineumare known to be tolerant of shallow saline waters (Thomas 1991)

We found no change in sedimentation rate at Diprose LagoonThis was unexpected becausemost sediment cores from southernAustralia show that a clear increase in sedimentation occurredshortly after the arrival of Europeans in the region (Dodson et al1994a 1994b Mooney and Dodson 2001) Sediment depositionis increasedwhen vegetation cover is removed throughfire or treeremoval which allows erosion to occur An apparent lack ofincreased sedimentation at Diprose Lagoon could be associatedwith increased compaction of upper sediments after sheep wereintroduced and the impact of artificial drainage on catchmenthydrology and sedimentation

Broadly we support the notion that Aboriginal landmanagement maintained eucalypt savanna and that Europeandisruption of Aboriginal management resulted in changed fireregimes and associated ecological changes most notably anincrease in fire a reduction in Allocasuarina and increases inexotic taxa However we cannot resolve the question of thehypothesised switch from a finer grained to coarse grained spatialpattern of landscape burning that caused the loss ofgrasslandndashforest mosaics that lies at the heart of manyenvironmental histories (eg Rolls 1981 Jurskis 2000Gammage 2011) It is possible that integrated studiescombining dendrochronology palynology and charcoalsediment analysis and historical records in the montanegrassland-forest mosaics in Tasmania that are rich in sedimenttraps andhave tree specieswith annual growth rings could further

J Australian Journal of Botany L M Romanin et al

advance this question of the spatio-temporal changes to fireregimes following European disruption of Aboriginal firemanagement (McWethy et al 2013 Bowman et al 2013Holz et al 2015)

Acknowledgements

This research is an output from the Landscape and Policy Research Hub Thehub was supported through funding from the Australian GovernmentrsquosNational Environmental Research Programme and involved researchersfrom the University of Tasmania (UTAS) The Australian NationalUniversity (ANU) Murdoch University the Antarctic Climate andEcosystems Cooperative Research Centre (ACE CRC) Griffith Universityand Charles Sturt University (CSU) Henk Heijnis and Atun Zawadzki fromAustralian Nuclear Science and Technology Organisation assisted with the210Pb analysis

References

Appleby PG Oldfield F (1978) The calculation of lead-210 dates assuming aconstant rate of supply of unsupported 210Pb to the sediment Catena 51ndash8 doi101016S0341-8162(78)80002-2

Attiwill PM Adams MA (2013) Mega-fires inquiries and politics in theeucalypt forests of Victoria south-eastern Australia Forest Ecology andManagement 294 45ndash53 doi101016jforeco201209015

Banks JCG (1982) The use of dendrochronology in the interpretation of thedynamics of the snow gum forest PhD thesis Australian NationalUniversity Canberra ACT Australia

Bastick CH Lynch S (2003) Updated assessment of extent of salinityin Tasmania Tasmanian Salinity Audit Tasmanian Department ofPrimary Industries Parks Water and Environment Launceston TasAustralia

Bastick CH Walker MG (2000) Assessment of extent and impacts ofdryland salinity in Tasmania Tasmanian Salinity Audit TasmanianDepartment of Primary Industries Parks Water and EnvironmentLaunceston Tas Australia

Batek MJ Rebertus AJ Schroeder WA Haithcoat TL Compas E GuyetteRP Haithcoat L (1999) Reconstruction of early nineteenth-centuryvegetation and fire regimes in the Missouri Ozarks Journal ofBiogeography 26 397ndash412 doi101046j1365-2699199900292x

Bennett KD Willis KJ (2001) Pollen In lsquoTracking environmentalchange using lake sedimentsrsquo (Eds JP Smol HJB Birks WM Last RSBradley K Alverson) pp 5ndash32 (Springer Dordrecht The Netherlands)

Benson D Howell J (2002) Cumberland Plain Woodland ecology then andnow interpretations and implications from the work of Robert Brownand others Cunninghamia 7 631ndash650

Benson JS Redpath PA (1997) The nature of pre-European native vegetationin south-eastern Australia a critique of Ryan DG Ryan JR and StarrBJ (1995) The Australian landscape ndash observations of explorers andearly settlers Cunninghamia 5 285ndash328

Bickford S Gell P Hancock GJ (2008) Wetland and terrestrialvegetation change since European settlement on the FleurieuPeninsula South Australia The Holocene 18 425ndash436doi1011770959683607087932

Binford MW (1990) Calculation and uncertainty analysis of 210Pb dates forPIRLA project lake sediment cores Journal of Paleolimnology 3253ndash267 doi101007BF00219461

Blaauw M (2010) Methods and code for lsquoclassicalrsquo age-modelling ofradiocarbon sequences Quaternary Geochronology 5 512ndash518doi101016jquageo201001002

BlackMPMooneySDHaberleSG (2007)Thefire humanand climate nexusin the Sydney Basin eastern Australia The Holocene 17 469ndash480doi1011770959683607077024

Bliege Bird R Bird DW Codding BF Parker CH Jones JH (2008) The lsquofirestick farmingrsquo hypothesis Australian Aboriginal foraging strategiesbiodiversity and anthropogenic fire mosaics Proceedings of theNational Academy of Sciences of the United States of America 10514796ndash14801 doi101073pnas0804757105

Bowman DMJS (1998) Tansley Review No 101 The impact of Aboriginallandscapeburningon theAustralianbiotaNewPhytologist140 385ndash410doi101046j1469-8137199800289x

Bowman DMJS (2001) Future eating and country keeping whatrole has environmental history in the management of biodiversityJournal of Biogeography 28 549ndash564doi101046j1365-2699200100586x

Bowman DMJS (2002) Preface Measuring and imagining exploringcenturies of Australian landscape change Australian Journal of Botany50 indashiii doi101071BTv50n4_PR

Bowman DMJS Garde M Saulwick A (2001) Fire is for kangaroosinterpreting Aboriginal accounts of landscape burning in CentralArnhem Land In lsquoHistories of old ages Essays in honour of RhysJonesrsquo (Eds A Anderson I Lilley S OrsquoConnor) pp 61ndash78 (PandanusBooks Research School of Pacific and Asian Studies Canberra)

Bowman DMJS Walsh A Prior LD (2004) Landscape analysis ofAboriginal fire management in Central Arnhem Land north AustraliaJournal of Biogeography 31 207ndash223doi101046j0305-0270200300997x

Bowman DMJS Wood SW Neyland D Sanders GJ Prior LD (2013)Contracting Tasmanian montane grasslands within a forest matrix isconsistent with cessation of Aboriginal fire management AustralEcology 38 627ndash638 doi101111aec12008

Brown S (1991) Aboriginal archaeological sites in eastern Tasmania(Department of Parks Wildlife and Heritage Hobart Tas Australia)

Buckney RT Tyler PA (1976) Chemistry of salt lakes and other waters ofthe sub-humid regions of Tasmania Australian Journal of Marine andFreshwater Research 27 359ndash366 doi101071MF9760359

Bureau of Meteorology (2008) Gridded high resolution monthly and annualfrost potential Dataset Climate Data Services Victoria AustraliaAccessed online 11 July 2013 (dataset no longer available)

BurleyALPhillipsSOoiMKJ (2007)Canagebepredicted fromdiameter forthe obligate seeder Allocasuarina littoralis (Casuarinaceae) by usingdendrochronological techniques Australian Journal of Botany 55433ndash438 doi101071BT06160

Butzer KW Helgren DM (2005) Livestock land cover and environmentalhistory the tablelandsofNewSouthWalesAustralia 1820ndash1920Annalsof the Association of American Geographers 95 80ndash111doi101111j1467-8306200500451x

Clarke PA (1986) The study of ethnobotany in southern South AustraliaAustralian Aboriginal Studies 2 40ndash47

Codding BF Bliege Bird R Kauhanen PG Bird DW (2014) Conservationor co-evolution Intermediate levels of Aboriginal burning andhunting have positive effects on kangaroo populations in WesternAustralia Human Ecology 42 659ndash669doi101007s10745-014-9682-4

Colhoun EA Shimeld PW (2012) Late-Quaternary vegetation history ofTasmania frompollen records In lsquoPeopled landscapesArcheological andbiogeographic approaches to landscapesrsquo (Eds SG Haberle B David)p 29 (Australian National University Press Canberra)

David TWE Browne WR (1950) The geology of the Commonwealth ofAustralia Vol 2 (Edward Arnold amp Co London)

Department ofPrimary Industries andWater (2009)TASVEG20Tasmanianvegetation monitoring and mapping program Vegetation ConservationSection Hobart Tas Australia

Dodson JR De Salis T Myers CA Sharp AJ (1994a) A thousand years ofenvironmental change and human impact in the alpine zone at MtKosciusko New South Wales The Australian Geographer 25 77ndash87doi10108000049189408703100

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany K

Dodson JR Roberts FK De Salis T (1994b) Palaeoenvironments andhuman impact at Burraga Swamp in Montane rainforest Barringtontops National Park New South Wales Australia The AustralianGeographer 25 161ndash169 doi10108000049189408703115

Dodson J Mooney S (2002) An assessment of historic human impact onsouth-eastern Australian environmental systems using late Holocenerates of environmental changeAustralianJournalofBotany50 455ndash464doi101071BT01031

Doyle RB (1993) lsquoReconnaissance soil map of the South Esk SheetTasmania (Southern half)rsquo (Tasmanian Department of PrimaryIndustries Parks Water and Environment Hobart)

Duffin K Gillson L Willis KJ (2008) Testing the sensitivity of charcoal asan indicator of fire events in savanna environments quantitativepredictions of fire proximity area and intensity The Holocene 18279ndash291 doi1011770959683607086766

Duncan F (1990) Eucalypts in Tasmaniarsquos changing landscape Tasforests 2151ndash158

Ellis RC Thomas I (1988) Pre-settlement and post-settlement vegetationalchange and probable Aboriginal influences in a highland forested areain Tasmania lsquoAustraliarsquos ever changing forestsrsquo (Eds KJ FrawleyN Semple) pp 199ndash214 (Australian Defence Force AcademyPublication Canberra)

Farag-Miller M Miller K Kirkpatrick JB (2013) Determining the accuracyof historical landscape paintings Geographical Research 51 49ndash58doi101111j1745-5871201200763x

FenshamRJ (1989) The pre-European vegetation of theMidlands Tasmaniaa floristic and historical analysis of vegetation patterns Journal ofBiogeography 16 29ndash45 doi1023072845309

Fensham RJ Kirkpatrick JB (1989) The conservation of original vegetationremnants in the Midlands Tasmania Papers and Proceedings of theRoyal Society of Tasmania 123 229ndash246

Fischer J Stott J Zerger A Warren G Sherren K Forrester RI (2009)Reversing a tree regeneration crisis in an endangered ecoregionProceedings of the National Academy of Sciences of the United Statesof America 106 10386ndash10391 doi101073pnas0900110106

Flannery TF (1997) Short communication A reply to Benson and Redpath(1997) Cunninghamia 5 779ndash781

Fletcher M-S Wood SW Haberle SG (2014) A fire-driven shift from forestto non-forest evidence for alternative stable states Ecology 952504ndash2513 doi10189012-17661

Fletcher M-S Thomas I (2010) The origin and temporal development of anancient cultural landscape Journal of Biogeography 37 2183ndash2196doi101111j1365-2699201002363x

Franco JA Morgan JW (2007) Using historical records aerial photographyand dendroecological methods to determine vegetation changes in agrassy woodland since European settlement Australian Journal ofBotany 55 1ndash9 doi101071BT06092

Gammage B (2008) Plain facts Tasmania under aboriginal managementLandscape Research 33 241ndash254 doi10108001426390701767278

Gammage B (2011) lsquoThe biggest estate on earthrsquo (Allen amp Unwin Sydney)Gilfedder L (1990) Threatened species from Tasmaniarsquos remnant grasslands

Tasforests 2 129ndash132Gott B (1983) Murnong ndash Microseris scapigera A study of staple food of

Victorian Aborigines Australian Aboriginal Studies 2 2ndash17Gott B (2005) Aboriginal fire management in south-eastern Australia aims

and frequency Journal of Biogeography 32 1203ndash1208doi101111j1365-2699200401233x

Gott B Murray L (1982) Ecology of root use by the Aborigines of southernAustralia Archaeology of Oceania 17 59ndash67doi101002j1834-44531982tb00039x

Grayson RB Kenyon C Finlayson BL Gippel CJ (1998) Bathymetric andcore analysis of the Latrobe River delta to assist in catchmentmanagement Journal of Environmental Management 52 361ndash372doi101006jema19980181

Griffiths T (2002) How many trees make a forest Cultural debates aboutvegetation change in Australia Australian Journal of Botany 50375ndash389 doi101071BT01046

Harle KJ KershawAPMacphailMKNeylandMG (1993) Palaeoecologicalanalysis of an isolated stand Nothofagus cunninghamii (Hook) Oerst ineastern Tasmania Australia Australian Journal of Ecology 18 161ndash170doi101111j1442-99931993tb00440x

Higuera PE Sprugel DG Brubaker LB (2005) Reconstructing fire regimeswith charcoal from small-hollow sediments a calibration with tree-ringrecords of fire The Holocene 15 238ndash251doi1011910959683605hl789rp

Higuera PE Gavin DG Bartlein PJ Hallett DJ (2010) Peak detection insediment-charcoal records impacts of alternative data analysis methodson fire-history interpretations International Journal of Wildland Fire 19996ndash1014 doi101071WF09134

Holz A Wood SW Veblen TT Bowman DMJS (2015) Effects of high-severity fire drove the population collapse of the subalpine Tasmanianendemic conifer Athrotaxis cupressoides Global Change Biology 21445ndash458 doi101111gcb12674

Hueneke K (1976) Aspects of the autecology and secondary succession ofCasuarina stricta on Canberrarsquos hillsrsquo MSc Forestry thesis AustralianNational University Canberra ACT Australia

Jackson W (1999) The Tasmanian legacy of man and fire Papers andProceedings of the Royal Society of Tasmania 133 1ndash14

Johnson AG (2000) Fine resolution palaeoecology confirmsanthropogenic impact during the late Holocene in the lowerHawkesbury Valley NSW The Australian Geographer 31 209ndash235doi101080713612247

Jones R (1969) Fire-stick farming Australian Natural History 16 224ndash228JonesP (2008)Thepalaeoecologyof theMidlandsTasmaniaHonours thesis

The University of Melbourne Melbourne Vic AustraliaJurskis V (2000) Vegetation changes since European settlement of Australia

an attempt to clear up some burning issues Australian Forestry 63166ndash173 doi10108000049158200010674827

Kee S (1990) Midlands Aboriginal archaeological site surveys (Departmentof Parks Wildlife and Heritage Hobart Tas Australia)

Kellaway RG (1989) Geographical change in Tasmanian agriculture duringthe Great Depression PhD thesis The University of Tasmania HobartTas Australia

Kenyon C Rutherfurd ID (1999) Preliminary evidence for pollen as anindicator of recent floodplain accumulation rates and vegetationchanges the BarmahndashMillewa Forest SE Australia EnvironmentalManagement 24 359ndash367 doi101007s002679900239

Kershaw AP Bulman D Busby JR (1994) An examination of modern andpre-European settlement pollen samples from southeastern Australia ndash

assessment of their application to quantitative reconstruction of pastvegetation and climate Review of Palaeobotany and Palynology 8283ndash96 doi1010160034-6667(94)90021-3

Kirkpatrick JB (2007) History In lsquoPeople sheep and nature conservationthe Tasmanian experiencersquo (Eds JB Kirkpatrick KL Bridle) pp 1-43(CSIRO Publishing Melbourne)

Kirkpatrick JB Glasby J (1981) Salt marshes in Tasmania distributioncommunity composition and conservation (Department of GeographyUniversity of Tasmania Hobart Tas Australia)

Kirkpatrick JB Harwood CE (1983) Plant communities of Tasmanianwetlands Australian Journal of Botany 31 437ndash451doi101071BT9830437

Kirkpatrick JB Gilfedder L Fensham RJ (1988) lsquoCity parks and cemeteriesTasmaniarsquos remnant grasslands and grassy woodlandsrsquo (TasmanianConservation Trust Hobart)

Kodela PG MacKillop RD (1988) Weed pollen a geomorphical application(Introduced pollen taxa as indicators of post-European sediments)Palynological amp Palaeobotanical Association Australian Newsletter17 4ndash6

L Australian Journal of Botany L M Romanin et al

LaddPG(1988)The statusofCasuarinaceae inAustralian forests lsquoAustraliarsquosever changing forestsrsquo (EdsKFrawleyNSemple) pp 63ndash85 (AustralianDefence Force Academy Publication Canberra)

Lunt ID (1991) Management of remnant lowland grasslands and grassywoodlands for nature conservation a review Victorian Naturalist 10856ndash66

Lunt ID (2002) Grazed burnt and cleared how ecologists have studiedcentury-scale vegetation changes in Australia Australian Journal ofBotany 50 391ndash407 doi101071BT01044

MacPhee RDE Flemming C (1999) Requiem AEligternam the last fivehundred years of mammalian species extinctions In lsquoExtinctions neartimersquo (Ed RDE MacPhee) pp 333ndash371 (Springer New York)

Mactaggart B Bauer J Goldney D (2007) When history may lead us astrayusing historical documents to reconstruct swampy meadowschains ofponds in the New South Wales Central Tablelands Australia TheAustralian Geographer 38 233ndash252 doi10108000049180701392782

McWethy DB Higuera PE Whitlock C Veblen TT Bowman DMJS CaryGJ Haberle SG Keane RE Maxwell BD McGlone MS Perry GLWWilmshurst JM Holz A Tepley AJ (2013) A conceptual framework forpredicting temperate ecosystem sensitivity to human impacts on fireregimes Global Ecology and Biogeography 22 900ndash912doi101111geb12038

Mooney SD Dodson JR (2001) A comparison of the environmental changesof the post-European Period with those of the preceding 2000 Years atLake Keilambete south-western Victoria Australian Geographer 32163ndash179 doi10108000049180125614

Mooney SD RadfordKLHancockG (2001) Clues to the lsquoburning questionrsquoPre-European fire in the Sydney coastal region from sedimentary charcoaland palynology Ecological Management amp Restoration 2 203ndash212doi101046j1442-8903200100085x

Mooney SD Webb M Attenbrow V (2007) A comparison of charcoal andarchaeological information to address the influences on Holocene fireactivity in the Sydney basin The Australian Geographer 38 177ndash194doi10108000049180701392774

MooneySDHarrisonSPBartleinPJDaniauALStevenson JBrownlieKCBuckman S CupperM Luly J BlackM Colhoun E DrsquoCosta D DodsonJ Haberle S Hope GS Kershaw P Kenyon CMcKenzieMWilliams N(2011) Late Quaternary fire regimes of Australasia Quaternary ScienceReviews 30 28ndash46 doi101016jquascirev201010010

Morgan S (1992) lsquoLand settlement in early Tasmania creating an antipodeanEnglandrsquo (Cambridge University Press Cambridge UK)

Morrison DA Renwick JA (2000) Effects of variation in fire intensity onregeneration of co-occurring species of small trees in the Sydney regionAustralian Journal of Botany 48 71ndash79 doi101071BT98054

Murphy BP Bowman DMJS (2007) The interdependence of fire grasskangaroos and Australian Aborigines a case study from centralArnhemLand northernAustralia Journal of Biogeography 34 237ndash250doi101111j1365-2699200601591x

NortonTLaceyM(2012)Patch data viewer a tool for planning investment invegetation extent and condition from patch to regional scales InlsquoLandscape logic integrated science for landscape managementrsquo (EdsE Lefroy A Curtis A Jakeman J McKee) pp 173ndash188 (CSIROPublishing Melbourne)

OrsquoConnell JF Allen J (2015) The process biotic impact and globalimplications of the human colonization of Sahul about 47 000 yearsago Journal of Archaeological Science 56 73ndash84doi101016jjas201502020

Oksanen J Blanchet FG Kindt R Legendre P Minchin PR OrsquoHara RBSimpson GL Solymos P Stevens MHH Wagner H (2015) vegancommunity ecology package R package version 20-10 httpCRANR-projectorgpackage=vegan [Verified 15 September 2015]

Oldfield F Appleby P (1984) lsquoEmpirical testing of 210Pb-dating models forlake sedimentsrsquo (Eds E Haworth W Lund) (Leicester University PressLeicester UK)

Parks and Wildlife Service Tasmania (2006) Tom Gibson Nature ReserveAvailable at httpwwwparkstasgovaufileaspxid=19229 [Verified17 July 2016]

PeelM FinlaysonBMcMahonT (2007)Updatedworldmap of the Koumlppen-Geiger climate classification Hydrology and Earth System Sciences 111633ndash1644 doi105194hess-11-1633-2007

Phillip (1892) lsquoHistorical recordsofNewSouthWalesVol 1Part 2rsquo (CharlesPotter Government Printer Sydney)

Prior LD Sanders GJ Bridle KL Nichols SC Harris R Bowman DMJS(2013) Land clearance not dieback continues to drive tree loss in aTasmanian rural landscape Regional Environmental Change 13955ndash967 doi101007s10113-012-0396-0

R Development Core Team (2008) R A language and environment forstatistical computing

Rolls E (1981) lsquoA million wild acresrsquo (Penguin Books Melbourne)Romanin LM Prior LD Williamson GJ Bowman DMJS (2015) Trajectory

of change in land cover and carbon stocks following Europeansettlement in Tasmania Australia Anthropocene 9 33ndash40doi101016jancene201507001

Russell-Smith J CookGDCooke PMEdwardsAC LendrumMMeyerCPWhitehead PJ (2013) Managing fire regimes in north Australiansavannas applying Aboriginal approaches to contemporary globalproblems Frontiers in Ecology and the Environment 11 e55ndashe63doi101890120251

Ryan L (2012) lsquoTasmanian Aborigines a history since 1803rsquo (Allen ampUnwin Sydney)

RyanDGRyan JR Starr BJ (1995) lsquoTheAustralian landscape ndash observationsof explorers and early settlersrsquo (Murrumbidgee Catchment ManagementCommittee Wagga Wagga NSW)

Sigleo WR Colhoun EA (1981) A short pollen diagram from CrownLagoon in the Midlands of Tasmania Papers and Proceedings of theRoyal Society of Tasmania 115 181ndash188

Sigleo WR Colhoun EA (1982) Terrestrial dunes man and the latequaternary environment in southern Tasmania PalaeogeographyPalaeoclimatology Palaeoecology 39 87ndash121doi1010160031-0182(82)90074-8

Singh G Geissler A (1985) Late Cainozoic history of vegetation fire lakelevels and climate at Lake George New South Wales AustraliaPhilosophical Transactions of the Royal Society of London Series BBiological Sciences 311 379ndash447 doi101098rstb19850156

Smart IJ Tuddenham WG Knox RB (1979) Aerobiology of grass pollen inthe city atmosphere of Melbourne effects of weather parameters andpollen sources Australian Journal of Botany 27 333ndash342doi101071BT9790333

Smith B (1960) lsquoEuropean vision and the South Pacific 1768ndash1850rsquo(Clarendon Press Oxford England)

Swetnam TW Allen CD Betancourt JL (1999) Applied historical ecologyusing the past to manage for the future Ecological Applications 91189ndash1206doi1018901051-0761(1999)009[1189AHEUTP]20CO2

Thomas I (1991) The Holocene archaeology and palaeoecology ofnortheastern Tasmania Australia PhD thesis The University ofTasmania Hobart Tas Australia

Thomas I (1994) Ethnohistoric sources for the use of fire by Tasmanianaborigines preliminary compilation and investigation of the evidence(Ed J Jacobs) (Department of Geography and Environmental StudiesUniversity of Melbourne Melbourne Vic Australia)

Trauernicht C Brook BW Murphy BP Williamson GJ Bowman DMJS(2015) Local and global pyrogeographic evidence that indigenous firemanagement creates pyrodiversityEcology andEvolution 5 1908ndash1918doi101002ece31494

Turner S Plater A (2004) Palynological evidence for the origin anddevelopment of recent wetland sediments Mdlanzi Swamp KwaZulu-Natal South Africa South African Journal of Science 100 220ndash229

Fire regime change in Tasmanian eucalypt savanna Australian Journal of Botany M

van Zant KL Webb T Peterson GM Baker RG (1979) IncreasedCannabisHumulus pollen an indicator of European agriculture inIowa Palynology 3 227ndash233 doi1010800191612219799989192

von Platen J Kirkpatrick JB Allen KJ (2011) Fire frequency variation insouth-eastern Tasmanian dry eucalypt forest 1740ndash2004 from fire scarsAustralian Forestry 74 180ndash189doi10108000049158201110676361

Whitlock C Larsen C (2001) Charcoal as a fire proxy In lsquoTrackingenvironmental change using lake sediments Vol 3 Terrestrial algaland siliceous indicatorsrsquo (Eds JP Smol HJB Birks WM Last) pp 1ndash23(Kluwer Academic Publishers Dordrecht The Netherlands)

Williams DR Potts BM Neilsen WA Joyce KR (2006) The effect of treespacing on the production of flowers in Eucalyptus nitens AustralianForestry 69 299ndash304 doi10108000049158200610676250

Woinarski JCZ Burbidge AA Harrison PL (2015) Ongoing unraveling of acontinental fauna Decline and extinction of Australian mammals sinceEuropean settlement Proceedings of the National Academy of Sciencesof the United States of America 112 4531ndash4540doi101073pnas1417301112

Wood SW Bowman DMJS (2012) Alternative stable states and the role offirendashvegetationndashsoil feedbacks in the temperate wilderness of southwestTasmania Landscape Ecology 27 13ndash28doi101007s10980-011-9677-0

Zacharek A Gilfedder L Harris S (1997) The flora of Township LagoonNature Reserve and its management Tunbridge Tasmania Papers andProceedings of the Royal Society of Tasmania 131 57ndash66

N Australian Journal of Botany L M Romanin et al

wwwpublishcsiroaujournalsajb