Overview of Archaeological Resource on Forests: A Paper · 2019-09-23 · 4. The Nature of the...

Overview of Archaeological Resource on Forests: ADiscussion Paper

A report undertaken for the NSW CRA/RFA Steering Committee[May 1998]

OVERVIEW OFARCHAEOLOGICAL

RESOURCE ON FORESTS:A DISCUSSION PAPER

CONSULTANCY REPORTPREPARED BY KIM LOMAX,

CONSULTANT ARCHAEOLOGIST TOSTATE FORESTS OF NSW

A report undertaken for the NSW CRA/RFA Steering Committeeproject number NA16/EH

[May 1998]

Report Status

This report has been prepared as a working paper for the NSW CRA/RFA Steering Committee under thedirection of the Environment and Heritage Technical Committee. It is recognised that it may contain errorsthat require correction but it is released to be consistent with the principle that information related to the

comprehensive regional assessment process in New South Wales will be made publicly available.

For more information and forinformation on access to data contactthe:

Resource and Conservation Division, Department ofUrban Affairs and Planning

GPO Box 3927SYDNEY NSW 2001

Phone: (02) 9228 3166Fax: (02) 9228 4967

Forests Taskforce, Department of Prime Minister andCabinet

3-5 National CircuitBARTON ACT 2600

Phone: 1800 650 983Fax: (02) 6271 5511

© Crown copyright [May 1998]

This project has been jointly funded by the New SouthWales and Commonwealth Governments. The workundertaken within this project has been managed by thejoint NSW / Commonwealth CRA/RFA SteeringCommittee which includes representatives from the NSWand Commonwealth Governments and stakeholdergroups.

The project has been overseen and the methodology hasbeen developed through the Environment and HeritageTechnical Committee which includes representatives fromthe NSW and Commonwealth Governments andstakeholder groups.

DisclaimerWhile every reasonable effort has been made to ensurethat this document is correct at the time of printing, theState of New South Wales, its agents and employees,and the Commonwealth of Australia, its agents andemployees, do not assume any responsibility and shallhave no liability, consequential or otherwise, of any kind,arising from the use of or reliance on any of theinformation contained in in this document.

CONTENTS

Executive Summary

1. Introduction 11.1 The Forest Estate 1

2. Cultural Heritage Conservation Policy -1970s -1990s 3

3. Major Studies in Forest Archaeology 73.1 Overview of Major Studies 73.2 Method 83.3 Data Coverage 103.4 Site Types and Distribution 103.5 Methodological Issues for Identification andManagement 11

4. The Nature of the Forest ArchaeologicalRecord 134.1 Overview of the ‘Formation’ of the Record 134.2 Natural Formation Processes 144.3 Aboriginally Induced Geomorphic Processes154.4 Post-Contact Processes 154.5 Future Impacts in Terms of SpecificArchaeological Site Types in Forestry Areas 16

5. Significance of the Resource 195.1 Background 195.2 Behavioural Meaning of Sites 195.3 Scientific Significance 20

6. Current Approaches to Management andIdentification 216.1 Predictive or Correlation Modelling 216.2 GIS Models (East Gippsland Example) 226.3 Tasmanian Management Practices forIdentification 236.4 South East Forest Archaeological ResearchProject 246.5 SFNSW (Northern Region) 246.6 Conclusion 25

7. Recommendations 277.1 Requirements for Further Survey 277.2 Conservation Strategy 277.3 Pre-Harvesting Surveys and Management forDifferent Site Types 28

References 29

1.EXECUTIVE SUMMARY

This report has been prepared for the jointCommonwealth/State Steering Committee whichoversees the comprehensive regional assessmentsof forests in New South Wales.

The comprehensive regional assessments (CRAs)provide the scientific basis on which the State andCommonwealth governments will sign regionalforest agreements (RFAs) for the major forests ofNew South Wales. These agreements willdetermine the future of the State’s forests,providing a balance between conservation andecologically sustainable use of forest resources.

This report was undertaken to present an overviewof the archaeological resource on New SouthWales’ forests and its management requirementsas a context for technical discussions regardingmanagement options for archaeological values inthe CRA/RFAs.

To set the scene, a brief overview of culturalheritage policy in relation to archaeologicalresearch is presented. This is followed by anoverview of archaeological research in New SouthWales’ forests.

The forest archaeological resource is described interms of the relative occurrence of differentarchaeological sites in addition to a more technicaldiscussion about the problems of defining andinterpreting the more intractable surfacearchaeological record (stone artefacts).

This report discusses the ongoing naturallandscape and taphonomic process affecting theresource, in addition to prehistoric, historic andcontemporary cultural impacts, to provide acontext for assessing the relative significance offuture impacts on the resource.

As a prelude to considering current managementrequirements, there is a discussion ofarchaeological significance. In particular, thespatial aspect of the resource is emphasised as aconceptual key for both understanding the resourceand its management.

Following this, approaches used to ‘model’ theforest archaeological resource for managementpurposes, in addition to other current forestmanagement practices in relation to archaeologicalsites, are discussed. Finally, somerecommendations are provided for a strategicprogram of management and research in light ofthe issues raised in this discussion paper.

1

1. INTRODUCTION

In the context of Aboriginal heritage it isimportant to distinguish cultural values,particularly those held by Aboriginal people, fromscientific values established by archaeologists.While there may be some degree of overlap andinterdependence between the two value systems,they nevertheless form discrete systems of belief,each requiring quite different models ofidentification and management.

The issues discussed in this paper relate primarilyto the archaeological or scientific values ofAboriginal archaeological sites mainly of the pre-contact era.While recognising that the orientationof this report is a consequence of the currentlegislation in NSW it should be noted thatAboriginal laws and customs continue to exist inrelation to the appropriate custodianship, use,control and management of Aboriginal culturalheritage. Aboriginal people continue to be boundby these laws and customs regardless of whetherthey are formally recognised by non-Indigenouslaws and policies. Therefore, the setting ofgovernment research and management agendasrelated to Aboriginal heritage must involve advicebeing sought from and negotiation with therelevant Aboriginal people or communities whoconsider themselves the primary custodians andinterpretors of their cultures. Accordingly, thereport should be seen only as one possiblecomponent of the broader discussion andnegotiations that must surround the setting ofresearch and management priorities.

Non-Indigenous archaeological sites are notcovered in this report and are subject to differentlegislation, management regimes andprescriptions.

This discussion paper raises a number of issuesthat need to be considered in future decisionsregarding the management and researchrequirements for the forest archaeologicalresource in the context of ComprehensiveRegional Assessments / Regional ForestAgreements (CRA/RFAs).

The paper has two main objectives: to give aclear indication of the nature of the forestarchaeological resource and to stimulatediscussion between natural resource and culturalheritage managers on archaeologicalidentification and management issues for theCRA/RFAs.

In describing the forest archaeological resourcethe focus is on providing an appreciation of whatfurther benefits archaeological research canrealistically deliver in the context of theCRA/RFAs. Also, the limitations of particularmethods of identification and approaches tomanagement are discussed.

The report highlights the need for agreed andclearly defined conservation objectives.

1.1 THE FOREST ESTATE

This study refers to forests within the five NewSouth Wales CRA Regions: Upper North East,Lower North East, Sydney Basin, Southern andEden. Together, these areas encompass the bulkof the forests of the Great Dividing Range andcoastal hinterland of eastern New South Wales.

The current forest estate is a subset of the pre -1788 forests of New South Wales. At a verygeneral level, the forest estate is differentiatedfrom the broader pre -1788 forest coverage on thebasis of forested areas characterised by bothlower soil fertility and more rugged terrain.

Most of the forests are encompassed in nationalparks or State forests, although there are somethat are leasehold and a smaller proportion thatare private. This report largely discusses theresults of surveys undertaken on State forests forEnvironmental Impact Assessment (EIA) workand to a lesser extent national parks where therehas not been the same development generatedimpetus to survey. Over the period in which mostsystematic surveys have been undertaken (the last20 years), land tenure has shifted substantially

Overview of Archaeological Resource on Forests May 1998

2

from Crown-timber lands to national parks, aprocess that is continuing. Given this, and theconsiderable overlap of broad forest types andcontiguity between different land tenures, we arefairly safe in taking survey results to date asbroadly indicative of forest in general rather thanspecific tenures.

3

2.CULTURAL HERITAGECONSERVATION POLICY -1970s -1990s

This overview is not intended to be comprehensivebut to indicate the path of forest heritageconservation policy in relation to archaeologicalsites over the last twenty years in New SouthWales.

By the mid 1970s, the then Forestry Commissionof NSW (FCNSW) and the National Parks andWildlife Service (NPWS) had started compilinginventories of Aboriginal sites as a result of theinception of the National Parks & Wildlife Act1974. It was generally thought by the Commissionthat there were relatively few sites in forests. TheCommission did not then perceive that it had abroader role beyond reporting sites chanced uponin the course of its operations and avoidingimpacts on those it knew about. Broadermanagement consideration and the need to surveyfor unknown sites was generally considered thepurview of NPWS.

A raised awareness of other issues relating to theAboriginal heritage and the archaeologicalresource on forests arose as a result of the NewSouth Wales Government investigation into woodchipping of the south coast forests. The Ashtonreport recommended that survey be undertaken toidentify Aboriginal ‘relics’ and areas sacred tolocal Aboriginal people. Archaeological (Hughesand Sullivan 1978) and anthropological (Egloff1979) studies were undertaken that identified arange of archaeological and Aboriginal sites andconfirmed a strong Aboriginal attachment to thesouth coast forest area generally. The results ofEgloff's study, in addition to the passage of theEnvironmental Planning and Assessment Act 1979,further alerted the Commission to the need for

cultural heritage assessments and increased itsresponsibilities in this regard.

In addition to these early forest studies, NPWSinstigated a statewide project to record sites ofAboriginal significance. The project known as the‘Sacred Sites Survey’ was undertaken by RayKelly and Howard Creamer and documented manyhundreds of sites including a significant number inthe north coast forests (Creamer 1980).

In response to the growing appreciation of theextent of the Aboriginal cultural heritage resourceon land that it managed, FCNSW commissioned anoverview of the cultural heritage resource onCrown-timber lands of New South Walesincluding an assessment of associated managementissues. The work was undertaken by SandraBowdler (1983), who recommended that aprogram of surveys be undertaken to locate sitesand that a management regime be instigated foreach site based on its significance (Bowdler 1983).She also recommended that Commission staff betrained in aspects of the management andidentification of cultural resources, and that anarchaeologist and Aboriginal liaison officers beappointed by the Commission to assist in furtherrefining and implementing theserecommendations.

There was not a clear overall response toBowdler's report in the short term, although someof Bowdler's priority areas for survey andrecommendations regarding specific sitesmanagement procedures were adopted during the1980s. Of particular note in this regard is the studyof rainforests, Bowdler's highest priority forsurvey (Byrne 1987). Bowdler's (1983)


4

recommendations for training and for theemployment of an archaeologist and AboriginalSite Officers (in a kind of parallel structure toNPWS) were not implemented until eight to tenyears after the publication of her report, in thecontext of increased EIA responsibilities arisingfrom court cases following forestry environmentaldisputes.

Denis Byrne was the Commissions next principalconsultant after Bowdler and worked on a numberof projects including assessments of the heritagevalues of NSW rainforests (Byrne 1987).

Byrne was to raise a number of key managementissues (Byrne & Smith 1988). Central to hisconcerns was the need to conserve a‘representative’ sample of sites ideally fromundisturbed environmental contexts. He alsoadvocated the need to conserve a sample of theunknown resource, that is sites that were predictedto occur but were as yet documented or unverified.He also recommended that conservation strategiesfor the protection of cultural values be enmeshedin forest operation planning. These ideas weregermane to his later discussion of forestmanagement principles (Byrne 1991) where headvocated the need for a more strategicmanagement approach centred on assessing thepotential of the existing reserve system forprotecting relatively intact archaeologicallandscapes.

In 1990, the Commission commenced an extensiveEIS program covering many of its ManagementAreas, particularly on the north coast. In 1991, anarchaeologist was employed by State Forests ofNew South Wales (SFNSW) Northern Region toproduce and coordinate the cultural heritagecomponents of some 13 Environmental ImpactStatements (EIS) for the North Coast forests.

Expectations varied within the Commission,however, there was a general assumption that thesite inventories produced for the EIS programwould be definitive and would provide for all thefuture cultural heritage management requirements.According to Gollan, this was an understandableexpectation because there had not yet beensufficient investigation in the northern forests toindicate the extent of the resource nor didconservation models from NPWS at the timesignificantly enlighten that view (Gollan 1992).

During the course of the EIS program, it becameincreasingly obvious that the extent of the resourcewas such that it defied definitive treatment in an

EIS. The focus of consultants recommendationsrelated more to the need for various measuresrelating to ongoing identification of the resourcethan management of the sites identified. Variousapproaches were recommended by differentconsultants and developed by SFNSW’archaeologist. These are discussed later in thisreport.

Once SFNSW, through the EIS program, was fullyengaged in identifying sites and then having toactively manage them, they were in effectcommitted indefinitely to providing resources tosustain this effort. The archaeologist's position onthe north coast has become permanent and fourAboriginal Cultural Heritage Officers have beenemployed in the north east of the State.

The policy situation is less clear in the southernpart of the State, but nevertheless, a considerablecommitment to Aboriginal cultural heritagemanagement has been maintained through the EISprogram and requirements to manage known sites.

In 1992, Gollan undertook an analysis ofAboriginal cultural heritage policy development inrelation to forests, with regard to the practices ofNPWS, SFNSW, and heritage experts/bodies,including the Australian Heritage Commission(AHC). He looked at the methods employed toidentify and characterise the cultural resource onforests. Gollan argued for critical reappraisal ofthe existing policy which had at its centre the ‘site’as the principal focus for management and greaterconsideration given to area based management topreserve the environmental and spatial context ofthe archaeological resource (Gollan 1992:17). Insupport of this view he cited Byrne's then recentpaper which argued similarly for a moreconsidered conservation strategy based on reserveselection to ensure that representative areas ofarchaeological variability are maintained whichwill enable future researchers to examine bothsystemic and symbolic aspects of human action(Byrne 1991:386).

The basic weakness of the assumption [that is, sitebased management], as conservation policy, isthat it does not allow for the consideration of thesystemic values of cultural information, that is,archaeology in environmental context, and that thearchaeological information system is dynamic withchanging impacts and threats (Gollan 1992:17).

Generally speaking he was critical of what heperceived to be SFNSW' failure to proactivelymanage for cultural resource values and at the

May 1998 Overview of Archaeological Resource on Forests

5

same time, NPWS’s failure to provide more viableconservation models rather than concentrating onsite inventorying as part of EISs. He argued thatNPWS's position was too focused on its regulatoryfunction rather than the provision of programmaticdirection that would provide conservationalternatives to agencies such as SFNSW (Gollan1992:13).

Perhaps NPWS’s legislation provides littleimpetus to produce conservation strategies beyondthe level of site protection and identification giventhat all sites are lawfully protected regardless oftheir significance. While operational guidelinesexist in relation to licensing activities and issues ofAboriginal involvement they are rarely articulatedwithin the context of an overarching conservationstrategy (Zone Team Accountabilities andOperating Guidelines 1995). However pragmaticdecisions are and will continue to be made on thebasis of ‘significance’ and the process for thiswould be greatly enhanced if the context fordecision making was broader than the confines ofa specific EIA project.

While NPWS's conservation agenda may not havebeen operationalised beyond the immediateproblems of identifying and managing sites, thereis evidence that for some time it has includedconsideration of the need to preserve and protect arepresentative sample of all sites types within aparticular environmental region, in addition toany sites of particular Aboriginal and scientificsignificance (Ross 1986:61).

The need to provide conservation leadership isrecognised by NPWS's efforts in the early 1990s toproduce ‘Cultural Area Plans’ or standing researchdesigns for specific NPWS regions. Unfortunately,the initial enthusiasm for these plans has wanedand they still await completion.

Again, NPWS willingness to develop innovativeapproaches to management is evidenced in theiroverview of the archaeological resource of theNatural Resources Audit Council (NRAC)northeast forest area (James & Conyers 1995). Thestudy provided a detailed summary description ofAboriginal archaeological site types in public landin the study area, identified gaps in the currentinformation base, and provided recommendationsfor future management. The report concluded witha broad range of management recommendations,including the need to formulate and test deductivepredictive models based on ‘human decisionmaking’ and to adopt ‘representative’ landscapebased units in favour of ‘sites’ for management

purposes. However, while the document articulatesa conservation strategy, it gives no indication howdeductive predictive models may be formulated ortested or their relationship to representative areasfor the purposes of conservation.

7

3.MAJOR STUDIES INFOREST ARCHAEOLOGY

3.1 OVERVIEW OF MAJOR STUDIES

To give the reader an appreciation of the amountof forest survey that has been undertaken, a briefoutline of some of the larger regional scale forestsurveys and/or regional synthesis is discussed.However, not included are the numerous sitespecific and the earlier smaller scale surveys oranthropological studies undertaken.

3.1.1 State-wide

Statewide studies of forest archaeological valuesinclude Byrne's (1987) rainforest study andBowdler's (1983) overview of archaeological sitesin Crown-timber lands.

3.1.2 Upper North East, Lower North Eastand Sydney Basin CRA Regions

The Upper North East and Lower North East CRARegions have had the most systematic surveycoverage of any of the NSW CRA areas. Unlikeother CRA regions, much of the forest data wasrecorded under a single program and recorded to astandard format. This program recorded over 500sites in the far north coast area alone.

In Upper North East and Lower North East CRARegions there has been two regional studiesencompassing forest areas. These are McBryde's(1974) pioneering regional prehistory of the NewEngland Region and Godwin's (1990) PhD thesisof the New England Tablelands and escarpmentforests. In addition to these academic studies, therehas been archaeological surveys undertaken ofeach of SFNSW' north coast management areas.These include studies of the Dorrigo (Comber1991), Glen Innes (Hall 1992a; Lomax & Fife inprep.), Grafton (Hall & Lomax 1993a), Tenterfield(Byrne 1995), Urbenville (Smith 1993), DorrigoThree Year EIS (Kuskie 1994), Coffs

Harbour/Urunga (Davies & Stewart-Zerba 1995),and Walcha/Nundle and Styx River (Davies et al1995; Lomax 1995) Management Areas. Surveyshave also been taken of SFNSW’ Casino District(Hall & Lomax 1993b), Duck Creek forestry EISArea (Collins 1991) and compartments within theChaelundi forests (Collins 1991).

In addition to these regional forest surveys, therehas been a program of open site excavation in theDorrigo Three Year EIS Area (Sullivan et al 1996)and a desk top data audit of archaeological sites onpublic lands of the NRAC north-east study area(James & Conyers 1995). Currently Hall andLomax (1996a) are synthesising some of theresults of the north coast work (this is discussed indetail below).

Areas within the Lower North East and SydneyBasin CRA Regions have been intensivelyinvestigated by both ground survey andexcavation, particularly parts of the Hunter Valleyand the Sydney region sandstone plateaux. Awealth of comprehensive archaeological dataincluding several thousand ‘sandstone sites’ arerecorded within or in close proximity to Stateforests or other forested lands (see Kinhill 1995for comprehensive review of this work).

Recent archaeological surveys undertaken on Stateforests in the Lower North East CRA Regioninclude regional surveys of Wingham (Collins &Morwood 1991), Kempsey/Wauchope (Packard1992), Gloucester/Chichester (Byrne 1992), andMorisset (Kinhill 1995) Management Areas.

Detailed systematic surveys have also beenundertaken of McPherson State Forest (McDonald1988) and Yarrahapinni and Way Way StateForests (Morwood & Collins 1991) in addition to aregional survey of the forests of the NewnesPlateau area (Gollan 1987).


8

Recently open site excavation in forest uplandswas undertaken in SFNSW’ KempseyManagement Area (Lomax 1994).

3.1.3 Southern and Eden CRA Regions

Some of the earliest systematic forest surveys wereundertaken in the forests of the Southern and EdenCRA Regions with reasonable survey coverage fora range of environmental contexts, particularly inSFNSW’ Eden Management Area.

Regional scale academic research undertaken inforested areas of this region include Flood's (1980)southern highlands prehistory, and recently,research on coastal forest rockshelter and opensites (Boot 1993). A large number of AustralianNational University (ANU) student surveys havealso been undertaken on low altitude forests of themid south coast recording literally thousands ofsites within a relatively small area of a fewhundred square kilometres. The results of thesestudies were recently compiled by ANU graduatestudent Tom Knight (1996).

Regional surveys undertaken on State forests andnational parks in Southern and Eden CRA Regionsinclude surveys of Wandella-Dampier (Byrne1983a), Five Forests (Sullivan & Hughes 1978;Byrne 1993b), Duea and Wadbilliga National Park(Byrne 1983c & d), Eden Woodchip AgreementArea (Byrne & Smith 1987; Byrne 1990) andSFNSW' Queanbeyan and Badja ManagementAreas (Grinberg Knight & Associates 1995).

Extensive field surveys of forests were alsoundertaken in the early 1990s for NPWS andSFNSW' South East Forests Projects (Packard1991; Packard in prep). Extensive and systematicsurvey coverage was achieved, unfortunately therehas been no report produced on the results of thiswork, although the results have been summarisedin the Eden EIS (SFNSW 1994). A total of 173artefact scatters were located at a density of 1.3sites per km of survey transect. Artefact scatterswere located in all environmental contextssurveyed and correlations noted between sitelocation and terrain characteristics (Packard, P.,NSW NPWS, pers. comm., 1997)

NPWS is currently undertaking a regional scaleprogram of research and field survey of landsystems in the Eden CRA Region. This work willinclude the data compiled from the South EastForests Project (Heffernen, K., NSW NPWS, pers.comm., 1997).

In addition to forest surveys, a desk top overviewof Aboriginal sites for NPWS south-east regionswas undertaken by Byrne & Smith (1988).

Regional scale survey of Victorian foresteduplands bordering the Eden CRA Region includethe Lower Snowy River (Geering 1981), SnowyRiver Management Unit (Hall 1990) and Far EastGippsland (Hall 1991) areas.

3.2 METHOD

Most of the archaeological surveys of forestedenvironments in New South Wales have beenundertaken in the last 20 years. Over this periodthere are a number of discernible trends both in themethod employed in undertaking field surveys,general aims of studies, data interpretation and inthe style of management recommendationsprovided.

A significant proportion of forest survey in NewSouth Wales has been undertaken as a result ofEIA on crown timber lands. Much of the work hasbeen regional in scale and has focused ondeveloping predictive models from sample surveydata and characterising the relative archaeologicalsensitivities of different environmentalcomponents in terms of artefact and/or sitedensities.

Noted trends include an increased attention inrecording variation in archaeological visibilityconstraints and relatively fine grainedenvironmental and taphonomic data pertinent toexamine correlations between site locations andformally defined landscape characteristics. Inaddition, early attempts to utilise formalstatistically based sampling frameworks have beenabandoned in favour of more ‘opportunistic’ non-probablistic approaches.

One of the earliest systematic forest survey wasundertaken in the early 1980s in the southernforests of New South Wales by Denis Byrne andBrian Egloff. Of central concern to this early workwas establishing formal sampling frameworks andthe need to devise means to overcome problems ofaccess and varying site visibility conditions. Onemethod devised to overcome these problems inforested upland areas was to locate surveytransects along ridgeline toposequences (Byrne1984).

Formal landscape definitions were not utilised forthe early survey work and site location wasdescribed in terms of composite topographic


9

elements such as ridgelines. While this level ofrecording was perfectly adequate forcharacterising general site location, laterresearchers have favoured the use of more formaldefinitions to permit a greater level of comparativeanalysis and more detailed predictions of sitelocations.

The results of the early forest work addedconsiderably to the then little known archaeologyof the New South Wales’ coastal hinterland,indicating a strong correlation between landscapecharacteristics and site location. The relationshipbetween terrain characteristics and constraints onhuman movement in upland coastal forests wasnoted and characterised in terms of a ‘ridgelinemodel’ of movement (Byrne 1984:70).

Subsequent regional scale field investigationundertaken in East Gippsland, Victoria by Hall(1991; 1992b) gave greater attention to a more finescale plotting of the occurrence of archaeologicalmaterials in terms of a range of environmental andtaphonomic parameters. This was undertaken tomore carefully characterise factors critical indetermining site location. Unlike previousapproaches, the site concept was abandoned andreplaced with the artefact as the minimal analyticalunit. This was undertaken principally to enable thecomparison of archaeological materials betweensampling strata at a range of spatial scales.

The survey strategy employed by Hall (1991,1992b) entailed quantifying areas of groundsurface exposure within each sampling strata andrecording the presence/absence of archaeologicalmaterial in terms of a range of environmental andtaphonomic considerations. This ‘controlled non-random’ sampling approach was later furtherformalised by Packard (1991) for the New SouthWales’ south east forests and used in a suite ofsubsequent forestry EIA studies (Byrne 1992;Byrne 1995; Davies et al 1995; Davies & Stewart -Zerba 1995; Grinberg Knight & Associates 1995;Hall & Lomax 1993 a & b; Kinhill 1995; Kuskie1994; Packard 1992; Smith 1993).

There have been various arguments for and againstprobabilistic and non-probabilisitic samplingdesigns such as that outlined above (Bird 1993).On a regional scale the former is costly andunproductive in terms of producing sufficientlocational data for modelling purposes. Also, inpractice the actual field survey is restricted toareas of high visibility within a given samplingunit, a bias which is not generally adequatelyaccounted for in the interpretative framework.

Non-probabilistic sampling designs focus on‘windows’ of archaeological visibility and havebeen characterised as non-representative and atworst circular (Bird 1993). However, the approachis particularly useful in forest EIA surveys as itfacilitates high site capture rates by focusing onareas of visibility and at the same time permitsquantitative assessments of sample biases inherentin such a data collection strategy.

In 1991, as part of their EIS program, SFNSWcommenced a program of systematicarchaeological survey and research (most of thesestudies are referred to above). This program wasundertaken over a five year period and involvedregional scale survey, intensive localised surveysand limited open site excavation.

The regional scale archaeological surveys wereundertaken according to a standardised recordingsystem based on that initially developed by Hall(1991). As a result, a considerable amount ofcomparable and detailed site locational data wasrecorded.

Management recommendations for thearchaeological resource have also evolved as ourknowledge of the resource has increased. Inparticular, the increasing recognition of the almostcontinuous distribution of archaeological surfacematerials throughout the forest environmentchallenges the logic of the automatic legalprotection afforded to the often highly disturbedexamples of these sites identified during forestsurveys. For this reason recommendationsincreasingly advocate strategies to protect the‘unknown resource’, that is, sites predicted tooccur but as yet unverified.

Recommendations for how this is achieved aregenerally couched in terms of the need to maintaina ‘representative sample’ but varies according towhether the analytical potential of the resource isseen principally in terms of its spatial or‘systemic’ nature, or alternatively, whethersignificance is associated with individual sites.Advocates of the former view tend to seemanagement in terms of defining representative‘landscape areas’ on the basis of environmentalparameters, while advocates of the latter prefer amore circuitous path and emphasise the need todirect efforts to further refine predictive models toidentify archaeologically sensitive areas.


10

3.3 DATA COVERAGE

The forested areas under investigation coverhundreds of thousands of hectares. In terms of areasurveyed, coverage is minuscule. However, thework to date can be considered a reasonablestarting point for formulating an understanding ofthe resource as a basis for its management. Firstly,our existing knowledge of the resource indicatesthat there are broad regularities in the record (thisis discussed further below). Much of the work thathas recently been undertaken is now producingredundant data in this respect (ie similar patterningnoted in respect to correlations between sitelocation/ artefact density and landscapecharacteristics etc.). Also, through forestry EIAwork, much of the data has been compiled in away that is relevant to providing regionaloverviews - (use of comparative data collectiontechnique and regional research designs). Furtherregional survey work, if undertaken within thecontext of the CRA, is unlikely to add significantlyto the existing corpus of survey data and wouldneed justification beyond that of a ‘gap’ fillingexercise.

An additional factor is that, aside from thesystematic surveys of the last 20 years, there hasbeen a good deal of informal ‘surveys’ of forestedareas through the work of NPWS, private oruniversity researchers and forestry workers overthe years (reflected in the NPWS and SFNSWdatabases). These researchers would have tappedinto local knowledge regarding the presence of themore ‘spectacular’ sites, so we can assumecoverage for these rare site types, althoughunmeasurable, is much greater than the coverageprovided by systematic surveys to date.

3.4 SITE TYPES AND DISTRIBUTION

This section will give a very general description ofthe forest archaeological resource in terms of sitetype and occurrence.

A wide range of archaeological site types havebeen recorded in forested areas. These includerockshelter occupation and art sites, quarries, axegrinding grooves, scarred and carved trees, stonearrangements, bora rings, rock engravings, burialsand artefact scatters. Sites types other than stoneartefact scatters are generally rare in forests, mostextremely rare. The main exceptions to this arerockshelter occupation/art sites andgrooves/engravings associated with sandstone

formations in the Sydney Basin, Lower North Eastand Upper North East CRA Regions.

Site types such as burial grounds, scarred trees andbora grounds are generally located in those areaswhich were the primary focus of Aboriginalexploitation, that is productive riverine areas andcoastal and estuarine resource areas, rather thanthe more marginal areas occupied by present daycommercial forests. Bora and burial grounds inparticular are often associated with riverinelandforms. In addition, these sites are highlyvulnerable to processes of natural and humanlyinduced attrition and the surviving sample of suchsites will be relatively small.

The impacts of fire, vegetation removal, grazingand timber harvesting would account for the rarityof vulnerable site types such as stone arrangementsand carved trees. Recently, a number of stonearrangements have been recorded in the forests ofSFNSW' Glen Innes Management Area (Lomax &Fife in prep.). However, these sites are located inprecipitous and rocky terrain on the edge of theNew England Tablelands escarpment and arelikely to have undergone only minimal disturbance- and are unlikely to be exposed to any futuredisturbance. While forests were undoubtably usedfor ceremonial activities, particularly mountainsand prominent topographic features,archaeological evidence of these activities is rare.

The occurrence of rockshelter occupation, art sitesand axe grinding grooves is in a large partdetermined by the presence of suitable rockoutcrops or shelters for occupation. These sitetypes most commonly occur where suitablesandstone formation is present and in associationwith basalt tors (Byrne 1995; Lomax & Fife inprep.).

Recorded stone quarry sites in forests are rare.While surveys have been undertaken in areaswhere suitable rock outcrops are common, fewquarry sites have been located (Hall & Lomax1993; Kuskie 1994). The use of specific rockoutcrops as quarries appears to be extremelyselective and it is anticipated that rarity of suchsite types is a real reflection of their occurrence.Other factors which may account for the rarity ofstone quarry sites is the use of pebble shingle bedsas sources of stone raw materials. The examinationof stone artefact assemblages from open siteexcavations in forests suggest that pebble shinglebeds are likely to have been a significant source ofstone raw materials even in locations where


11

suitable stone sources are known to occur(Sullivan et al 1996; Lomax 1994).

In contrast, archaeological evidence in the form ofstone artefact scatters is present in all forest typesand in many, if not most areas, occurs more or lesscontinuously across the landscape at an averagedensity of one to two artefact occurrences perkilometre of survey transect (Hall & Lomax1996:35). There have been thousands of such sitesnow recorded in forests, mostly through theforestry EIA work described above. The presenceof these stone artefact occurrences throughoutforested areas is not surprising given that theypotentially represent the accumulated debris ofmany thousands of years of occupation.

3.5 METHODOLOGICAL ISSUESFOR IDENTIFICATION ANDMANAGEMENT

There is now a large amount of forestarchaeological work that has been undertaken inNew South Wales. While the views of agenciesand practitioners on identification andmanagement have so far been divergent, it seemswe are now in a position to define some essentialcharacteristics of the archaeological record thatshould provide a common baseline for futurework. These relate to recognising the extent andformation history of the record, and realisticallyaddressing these factors in terms of researchprograms, significance assessment and approachesto management.

13

4. THE NATURE OF THEFORESTARCHAEOLOGICALRECORD

4.1 OVERVIEW OF THE‘FORMATION’ OF THE RECORD

It has been established that the archaeologicalresource on forests is present in all forest typesand in many areas occurs more or lesscontinuously across the landscape. Generallyspeaking, what are termed sites merely reflect‘windows’ of archaeological visibility. These‘sites’ are generally characterised by small shallowdeposits of stone artefacts.

At the same time, forest research has demonstratedthe connections between landscape configurationand landscape scale and the formation of sucharchaeological deposits. In forested uplands inparticular, site location can be predictedprincipally on the basis of terrain characteristics.This is a direct reflection of the topographicconstraints imposed on human movement, wheremovement is constrained to specific pathways suchas ridges.

Due to the lessening of constraints on movement,conversely the less dissected a region is, there is acorresponding weakening of the tendency forarchaeological materials to cluster on linearlandforms away from water sources, such asridgelines. In these landscapes the focus may movemore to areas adjacent to waterways and swamps,or other resource considerations. For example, inthe north coast forests on the coastal plains andtablelands, where planar landforms dominate,there is little landform focus for site formation and

it is correspondingly more difficult to intersectsites by surveys.

Any interpretations of regional patterning in areasof the Great Dividing Range needs to beunderstood firstly in terms of the opportunities andconstraints imposed by the physical configurationof the regional landscape. This requiresconsidering the ‘connectivity’ of the landscape -that is how patterns of movement in one area areinfluenced by the channelling of movement bylandscape constraints imposed in all other areas(cf. Byrne 1991).

While landscape configuration and terraineffectively constrains the range of options for thespatial deposition of materials in a landscape, landform stability determines the possibility for, andtemporal scale of deposition. Generally speaking,landforms in forested environments of the GreatDividing Range are fairly stable with the notableexception of areas associated with drainage linesand colluvial wasting associated with some steeplydissected areas.

Stable landscapes typically result in the long termaccumulation of cultural remains. In landscapeareas where there is both the long termaccumulation of materials and constraints onspatial deposition, the expectation is thatarchaeological variability will be characteristicallycoarse grained (cf. Binford 1980).

Recently, data from detailed site surveys (Kuskie1994) and open site excavations in the north eastof New South Wales (Sullivan et al 1996; Lomax1994) in eastern Chaelundi State Forest and Nulla


14

Five Day Forest were examined specifically todetermine how landscape variation and surfaceprocesses affect the spatial and temporal scale ofarchaeological variability (Hall & Lomax 1996a).Most of the local area and site specific dataexamined related to eastern Chaelundi State Forest(Sullivan et al 1996). This study area was a goodstarting point for obtaining a baselineunderstanding of these issues, as it is designatedold growth and therefore had been subject tominimal disturbance from logging.

Looking at the Chaelundi data, and a range ofother regional survey and excavation data, Halland Lomax (1996a) argued (on the basis ofcomparison between five excavated siteassemblages) that there was limited subregionalvariation between assemblages, in terms of theproportion representation of either raw material orartefact categories other than that related directlyto variation in local geology. In addition, theyargued that variation noted between sites, in termsof spatial structure and artefact density, was notrelated to functional considerations, but proximityto stone raw material source and variation in localand micro-topography (Hall & Lomax 1996a).

Geomorphic examination of sediments at a numberof the Chaelundi ridge crest sites confirmed a highdegree of stability to the depositional matrix ofsites, indicating that these landform surfaces, andpresumably others like them, throughout the eastcoast of New South Wales have retained completerecords of stone artefact discard (Sullivan et al1996).

These results generally reinforce the view (inrespect to archaeological variability) that there is,overall, a fairly low level of spatial and temporalresolution to the archaeological record in forests.More specifically, occurrences are generally notamenable to interpretation as discrete behaviouralentities. Rather they reflect patterns of movementabout the landscape and to a large extent theirformation is determined by broad scale regularitiesand patterning of the physical environment (cf.Stafford & Hajic 1990:142).

4.2 NATURAL FORMATIONPROCESSES

Archaeologists often refer to sites as beingundisturbed or ‘in situ’. Generally, by this theymean that sites have not been disturbed by culturalprocesses or impacts that might alter their spatial‘integrity’. In reality, however, there are a range of

natural and cultural processes which alter thespatial arrangement of archaeological materialfrom the time they are discarded. How rapidly thisprocess happens and how significant the affectsare depends on the nature of thelandscape/sediment units in which the material isdeposited.

To understand the effect of natural formationprocess on the structure of the archaeologicalrecord, it is useful to consider cultural materials interms of their landscape/sediment assemblagematrix. Such units are useful for ordering andcharacterising cultural materials, as they havemeasurable spatial and temporal boundaries, thatreflect relatively long term processes, especiallywhere vegetation and climate have limitedapplicability to the past.

In forests, soil bioturbation and tree growth arenatural processes which will have had significantimpacts on the spatial integrity of culturalmaterials. Even where soils are very stable andthere is no lateral movement soil bioturbation willrelatively rapidly rework top soils. For this reasonthere is generally no temporal ordering ofmaterials (Mitchell 1995; Sullivan et al 1996).

Tree growth also ensures that over time theentirety of the forest floor will have undergonesome level of disturbance. As Gollan (1992: 44)has suggested, even a simple linear analysis of treegrowth indicates that in forests of 100 trees perhectare, the time taken for every part of the forestfloor to have been effected by just trees is about2 500 years, that is there is arguably no site olderthan that (excluding sites on rock platforms) whichretains depositional integrity.

There are a range of taphonomic factors effectingnot only the formation of archaeological sites, butalso our ability to detect archaeological materialson contemporary landsurfaces. The extent towhich geomorphic processes affect this aspect ofthe resource is often under-estimated byarchaeologists, when they attempt to drawrelationships between the intensity and pattern ofhuman occupation on the basis of relative amountsof archaeological materials located by surfacesurvey.

Except where there is extensive deflation orerosional land surfaces present, the surfaceexpression of archaeological open sites is likely tobe only a fraction of the existing resource.Conditions for the exposure of archaeologicalmaterials in forests vary significantly between


15

different landform/sediment matrices and extremecaution should be exercised in drawingconclusions about human occupation on the basisof correlation data. At one extreme are theconditions presented by forest environments withskeletal soils where materials are conflated withina very thin matrix of sediments and thereforemanifest as high density ‘sites’. Converselymaterials in wet forest environments are extremelydifficult to locate, regardless of the intensity oftheir deposition. This is because materialsdeposited within a significantly greater soilvolume site tend to fall below the threshold ofarchaeological visibility (Hall & Lomax 1993 a &b).

4.3 ABORIGINALLY INDUCEDGEOMORPHIC PROCESSES

In 1969, Rhys Jones coined the phrase ‘fire stickfarming’ (Jones 1969). A radical suggestion at thetime as it challenged the view that there was such athing as a wholly natural ecosystem in Australia(Flannery 1995:223). However, there is ampleevidence to suggest that the early explorer did notenter a continent with pristine, stable landscapes.Rather, the landsurface was being continuouslyand rapidly moulded by the effects of bushfires, aprocess probably intensified by Aboriginal burning(Hughes & Sullivan 1981a).

It has been hypothesised on the basis of geo-archaeological investigations undertaken in easternAustralia, that Aboriginal firing regimes led toepisodic erosion and depositional rates whichgreatly exceeded those under natural firing(Hughes & Sullivan 1981 b: 277). Hughes andSullivan argue that Aboriginal burning practicesremoved the ground cover of grasses, shrubs andprotective leaf litter exposing soils to acceleratedrainwash, sheetwash and rill erosion, especiallyfollowing high intensity rainfall events. Inparticular, areas where soils have low clay content,such as those formed from sandstone and basalt,would have higher rates of instability (Hughes andSullivan 1986:129). Conversely, soils with higherclay and organic content are less likely to havebeen eroded.

4.4 POST-CONTACT PROCESSES

Disturbance as a result of human agency is likelyto have occurred in most forested areas since post-contact. In particular, areas of less marginal andrelatively accessible terrain are likely to have

undergone a range of impacts and vulnerablearchaeological sites in such areas will haveundergone a high degree of disturbance.

Early grazing runs either side of the GreatDividing Range were extensive. The initialpresence of cattle and sheep is likely to haveinitiated the deflation of delicate topsoils evenbefore any purposeful vegetation clearancecommenced. The extent of the impact of grazingwhich has continued to the present day, and soilloss and modification of soil structure throughtrampling stock, will have conflated extensiveareas of the surface archaeological record. Inparticular, sodic duplex soils of south easternAustralian uplands are highly susceptible to sheetwash, gully and tunnel erosion, and deteriorationof landsurfaces in these areas was widespread bythe 1840-50s (Jenkin 1986:136).

As a result of mining activities associated with thegold rushes of the 1850s, slopes were stripped ofsoils, and alluvium in valleys completelyreworked. People generally think of these impactsas being contained to fairly discrete areas, butchanges to hydrological regimes and downstreameffects were often considerable (Jenkin 1986:139).

The 1860s was the era of the Free Selection Acts,designed to put working people on the land.During this period extensive deforestationoccurred as families began to effect obligatory‘improvements’. The extent of vegetationclearance and wastage of timber resources throughring barking was so dramatic that a RoyalCommission was initiated in the early 1900s toinvestigate the future of New South Wales’ timberreserves (Curby 1993:5). Generally, areasconsidered to be 'poor country' or unworthy ofagriculture, were the areas from which New SouthWales’ first timber reserves were recognised.

Former and current Crown-timber lands have beenharvested in most management areas for at least150 years. The forests have been logged and, insome areas, several logging operations haveoccurred over time as demand for different timberproducts arose and sawmilling techniques andlogging equipment improved. In particular,technological advancements in mechanisation inthe late 1930s permitted harvesting of previouslyinaccessible terrain (Byrne 1992:28).

Past harvesting has clearly caused soil erosion,principally gully and sheet wash, andmodifications to hydrological regimes in terms ofincreased run off and infiltration. However, the


16

extent of past impacts depends largely on theintensity of harvesting and the forest type inquestion. Traditional forestry practices involvingselective logging or long rotations have probablyhad only small impacts on soils (Lamb 1986:440).Snigging tracks and forest roads have the mostpotential to cause significant soil erosion, butdamage is generally avoided through providingample water dispersing drainage (Lamb 1986:423).

Fire management and forest fuel reduction areessential to minimise the risk and spread of severewildfires. This is achieved on public lands by themaintenance of an extensive network of fire trailsand in addition to grazing on Crown-timber lands.These activities continue to impact thearchaeological resource on forests, howeveraccelerated erosion through high intensity fire hasa significantly greater negative impact (Lyons, K.,SFNSW, pers. comm., 1997.) .

Recreation and tourism is an important activity inmany of the more accessible areas of public forestand has the potential to impact more obtrusivearchaeological site types through curiosity andvandalism.

Given the long history of processes affecting thedepositional integrity of the surface archaeologicalrecord, it is reasonable to question the efficacy ofsite survey as a process to mitigate further impactsin areas other than those that have undergone‘minimal’ disturbance.

4.5 FUTURE IMPACTS IN TERMS OFSPECIFIC ARCHAEOLOGICALSITE TYPES IN FORESTRYAREAS

Archaeological sites can be regarded as having twodimensions from which their value derives. Theirphysical elements and their structure. Some sitetypes are more vulnerable to disturbance thanothers. For example, sites with large physicalelements (i.e. scarred trees and stonearrangements) are likely to be destroyed by asingle impact, whereas stone artefact sites whichare less vulnerable to disturbance, may withstand anumber of impacts before they are completelydestroyed (Byrne 1992; Hall & Lomax 1993 a &b).

4.5.1 Stone arrangements and boragrounds

These sites are vulnerable to processes of naturalattrition such as tree growth, however sites formed

in the recent past may still be detected as culturalformations. With the exception of stonearrangements located on rock platforms/ledges,extant sites are likely to have been impacted bypost-contact land use.

4.5.2 Scarred and carved trees

These are relatively uncommon site types awayfrom major valleys and riverine areas and arevulnerable to high intensity fire, vegetationclearance and relatively rapid natural attrition.

Remnant trees may be present in timber productionforests as a result of their poor timber quality orbecause they are non-commercial species. Whilethe trees may not directly be targeted in futureharvesting activities due to their likely lowcommercial values, they could be inadvertentlyaffected through forestry operations.

4.5.3 Rockshelter sites

These sites are likely to be relatively common inforests in areas where sandstone shelter formationor granite tors are present.

It is often assumed by archaeologists thatrockshelter sites have the potential to be impactedfrom run off as a result of vegetation clearance orharvesting activities. However, the detailedexamination of rockshelter sites in the UpperMangrove Creek catchment, which has beenperiodically logged, failed to report any damage toshelters that could be specifically attributed tologging (Kinhill 1995:105).

Perhaps the greatest threat to rockshelter sites isthrough direct human disturbance in popularrecreation areas or if roading was to be undertakenin close vicinity.

4.5.4 Rock engravings and axe grindinggrooves

These are relatively rare in forests with theexception of sandstone areas.

Timber harvesting is unlikely to have a directimpact on these sites on steeper slopes, includingcliff lines with rock outcrops and rock shelters.Nor should it affect sites in or adjacent to creeklines given the requirements to maintain bufferzones along creek lines. Indirect impacts to thesesites as a result of sediment wash is likely to bemitigated by current soil managementprescriptions (Kinhill 1995:106).


17

For sites located in the open, contexts damagecould be caused any where that heavy machinery isused.

4.5.5 Stone quarry sites

Stone quarry sites in general comprise areas ofartefact debitage and loose rock floaters that aresusceptible to ground surface impacts fromforestry activities. Larger quarry sites associatedwith extensive rock outcrops are extremely rare.Generally, it can be expected that these sites willbe located in areas of low timber site quality, giventhat their existence depends on largeconcentrations of rock. For this reason, they areunlikely to be impacted directly from harvestingoperations but may be impacted by gravelextraction or roading (Collins & Morwood1991:56).

4.5.6 Surface archaeological record

Apart from what can be determined on the basis oflogical argument, it would be difficult toobjectively measure the effects of differentimpacts of harvesting on the resource. This isbecause the resource has continuously undergonetransformation (according to the above formationprocesses) at varying rates and there is no obviouscondition against wish to measure disturbance.

General predictions about the likely condition ofthe surface archaeological resource could be madeon the basis of disturbance history and byreference to the stability of different soil types andlandforms. As a general rule of thumb, we canassume that the surface record will haveundergone a high degree of disturbance in areasthat have had high intensity harvesting operations,particularly those undertaken in areas where soilshave low clay content and/or are inherentlyunstable following ground surface disturbance.

Impacts to sites can be of two kinds, the effects onartefact themselves and secondly the effects on sitestructure.

Damage to artefacts occurs mainly as a result ofdirect pressure from machine tracks or tyres ontracks, especially on compact surfaces such asgravel where they could potentially be crushed.Off tracks this will generally only apply to arelatively small part of each operation, althoughthe effect will be cumulative with each subsequentharvesting operation (Byrne 1992).

The movement of artefacts is the most pervasivecause of degradation to sites. Apart from naturalprocesses of disturbance, the dislocation ofartefacts resulting from roading is a new order ofdisturbance which virtually destroys the spatialpatterning of sites (Byrne 1992). It is axiomaticthat forestry activities in unlogged areas have ahigher potential to disturb sites than the sameactivities in previously logged areas (Packard1992).

The actual processes associated with harvestingwhich disturb the structure of the surface recordare ground churning, compaction and subsequentsoil erosion. This will occur most intensively inassociation with snigging tracks, log dumps andduring road construction. However, the intensity ofpotential impacts will vary with terrain. In areas ofdissected terrain, logging operations will be mostintensive on the upper parts of toposequences.Most of these areas have relatively higharchaeological sensitivity and the impacts ofoperations is potentially high. In areas whereplanar landforms dominate, archaeologicalmaterials will tend to be more dispersed. In theseareas there will be less direct correspondencebetween operations and site locations and impactswill be correspondingly lower (Hall & Lomax1993 a & b).

The effect of this on the archaeological resourcewill have been cumulative degradation, rather thanthe complete destruction of the resource. Eachcutting cycle leaves some areas intact or onlypartially degraded. This degradation will increaseuntil a hypothetical end point is reached, themaximum possible disturbance of all areas (Byrne1992).

Even though there has been a long history ofdisturbance to sites in forests, large numbers of‘intact’ or partially intact sites will remain (Hall &Lomax 1993b:83). In addition, the lateraldisplacement of artefacts will not have affected allpotential information associated with surface sites.For example, the ratio of different artefactcategories and raw materials at site assemblages isunlikely to be significantly affected.

19

5.SIGNIFICANCE OF THERESOURCE

5.1 BACKGROUND

Our understanding of the significance ofarchaeological sites is fundamental in shaping ourapproach to identification and management.

Significance, of course, has many differentdimensions: scientific, social and aesthetic. Asdiscussed previously, this paper is primarilyconcerned with archaeological or scientificaspects, although there is some congruence withhow archaeologists rank scientific significance andother codes of significance. Notably, sitesconsidered by archaeologists to be to highlysignificant for their rarity, such as stonearrangements, would invariably be ranked ashighly significant on other grounds, for example,public, and Aboriginal. On a more general level, itis worth noting that given archaeology isconcerned with defining empirical phenomena, itprovides a good basis for assessing other forms ofsignificance that derive from physical remains.

In assessing scientific significance, archaeologistsgive primacy to research or information potential.This is largely based on our understanding of howthe archaeological record manifests behaviouralmeaning, both spatially and on a site by site basis.

5.2 BEHAVIOURAL MEANING OFSITES

The archaeological record is often perceived bycultural heritage practitioners in terms of theethnographic present, that is occurrences ofarchaeological debris are often referred to as‘campsites’ or ‘occupation’ sites.

The archaeological record is often conceived of asa large number of discrete localities of behaviouralrelevance (‘sites’), separated by spaces that haveno archaeological content. Obviously, in relation

to the forest archaeological resource, this view isat odds with what we now know to be theempirical nature of the resource. That is, it islargely spatially continuous, and its contemporarysurface manifestation is the result of a range ofnon-behavioural landscape and taphonomicprocess operating at a range of spatial scales.Trying to manage for this spatial aspect of theresource on the basis of disparate ‘sites’ presents anumber of practical problems.

The need to challenge the ontological status ofsites in relation to forest archaeological resource isnot new (Byrne 1991, Gollan 1992). It is generallyappreciated, even by the most strongly committedto the site notion that the archaeological record is acontemporary phenomena (Binford 1968:271).Consequently, the continuous distributions ofarchaeological debris that are taken to demark sitesare contemporary patterns and are not a prioribehaviourally relevant units. This is evidenced bythe results of formation studies (Schiffer 1987) andtaphonomic studies (Foley 1981) generally thatmake it clear that sites, as they are observed byarchaeologists, are created by the act ofobservation at a particular point in time. Inaddition, materials are added, removed andrearranged continuously in the archaeologicalrecord.

The danger in the uncritical use of the site conceptin resource management leads to systematicexclusions of segments of the archaeologicalrecord and limited understanding of the context ofthe included segment (Byrne 1991). At worst, itleads to a highly skewed managed record (Dunnell1992:33).

The challenge, therefore, is not to identify sites formanagement, but to define and identifyarchaeologically relevant spatial aggregates. Thekey to defining and identifying these units, must


20

take into account the characteristics of theresource, in particular the scale of its variability,and to appreciate its formation as a largelysedimentary process (cf. Schiffer 1987).

5.3 SCIENTIFIC SIGNIFICANCE

Historically, significance assessment for researchand management purposes has focused on theindividual site, particularly large and complex sites(see papers in Sullivan and Bowdler 1984).

In relation to open sites, the wisdom that largerand/or complex sites have inherently moreinformation potential is rarely questioned.However, in reality density or size does not relatedirectly to intensity of behaviour, or site function,but may reflect the stability of landforms, spatialconstraints on human behaviour and theconnectivity of the landscape. For example, siteslocated in xeric forests on narrow ridge linetoposequences are the densest most extensive siteslocated thus far in forests, however, in terms ofhuman behaviour, they do not reflect a differentlevel or type of occupation but merely representthe long term accumulation of low level discardevents. In terms of their information potential, theyare time averaged palimpsests with no temporalpatterning. In short, they may be bigger, but theydo not necessarily have more information content.

Generally speaking, however, the resource ischaracterised by low density surface scatters ofmaterial, which according to the generally appliedprinciples of significance assessment for scientificvalues, are individually of low significance. On anindividual basis, apart from their locations, theremay not be a lot more potential information thatcan be derived from these sites in isolation (Byrne1983b).

However, collectively these sites have the abilityto provide information on the human response tothe forested environment in terms of the movementof people and materials about the landscape. Theirinformation potential lies largely in theirinterrelationship with other sites and landscapecharacteristics at a range of spatial levels.Therefore, the problem becomes how to identifyand manage areas ‘representative’ of thispatterning (Hall & Lomax 1996b).

This is not to say that we need to know the preciselocation of sites to define such areas. Due to theirclose association with landscape characteristics,the problem becomes one of choosing

representative components of the landscape (Byrne1991; Gollan 1992).

While there is some reluctance for archaeologiststo conceive variability in terms of environmentalcharacteristics, lest they be accused of‘environmental determinism’, it is pertinent to notetwo points. Firstly, recognising the opportunitiesand constraints presented by landscape variabilitydo not negate the process of socially defineddecision making, it is relevant to establish whatthese constraints are (Ingold 1981). Secondly,regardless of the behaviour that in part formed thecontemporary surface archaeological record, it islargely a product of past and ongoing landscapeprocesses, so, in order to establish behaviouralvariability, these other formation processes mustfirst be understood.

Stratified open sites with a degree of temporalresolution would automatically be accorded a highdegree of scientific significance. However, theseare likely to be very rare in forests and are mostlikely to occur near water ways/swamps or inrockshelters. It can generally be assumed that theprotection of such sites is catered for by theexisting reserve system and measures applied tomanage a range of natural environmental values.

Other rare, obtrusive kinds of sites, such as thosementioned above, are generally considered to havehigh scientific significance and would need to bespecially catered for in any management strategy.

21

6.CURRENTAPPROACHES TOMANAGEMENT ANDIDENTIFICATION

This section of the document will examine some ofthe approaches that are currently advocated for‘modelling’ the forest archaeological resource, inaddition to current forest management practices.This is undertaken as a basis for providingrecommendations for a more strategicidentification and management of the resource inthe final section of the report.

In particular, this chapter will examine the recentuse of ‘predictive modelling’ and discuss itslimitations for determining land use decisions forAboriginal archaeological sites. In particular,whether more resources and the application ofmore sophisticated mapping technology willgreatly advance our understanding of the resourcepast first principles.

Expanding on this discussion, current forestmanagement and identification practices withreference to Forestry Tasmania and SFNSW andthe recently instigated South East Forests RegionalAdvisory Committee (SEFRAC) archaeologicalproject in New South Wales will be examined.

6.1 PREDICTIVE OR CORRELATIONMODELLING

Predictive or correlation modelling in archaeologyis an inductive process based on the recognition ofcomposite patterns of uniformities that aredetected by empirical observations. Patterns aregenerally described in a way that can provideexpectations concerning the archaeologicalcharacteristics of unknown areas (Warren1990:91).

All recent regional scale EIA forest archaeologicalsurveys involve some form of predictive modellingwith recommendations for further survey to‘test/refine’ the ‘preliminary’ models. However,the question of what level of generality orspecificity is adequate is rarely addressed.Perhaps, more importantly, is the recognition thatcorrelation models, because of their very nature,are probably untestable in any formal sense(Norton & Williams in Lewis 1996:19).

All stakeholders need to have a realisticappreciation of archaeological data and itslimitations for predictive modelling. The NorthAmerican literature is worth examining in thisregard. Such models have been used much longerthere, and have gone through similar stages ofdevelopment as in Australia, so there are importantlessons. In particular, the North American exampledescribed below indicates the problems inherent inseeing predictive modelling as an end in itself.

There is a long history of predictive modelling andmore recently the application of GeographicInformation Systems (GIS) to modelling sitelocations in cultural resource management inNorth America. Great fervour and extensiveresources have been devoted to probabilistic-basedsurveys aimed at elucidating patterns of settlement(Altschul 1990:227).

Faced with the problem of how to manage largenumbers of often poorly documented resourcesand resources whose locations were not evenknown, North American Federal archaeologistsand land managers of the 1970s seized upon the


22

potential of these studies. However, by the 1980sthe popularity of these models had significantlywaned. Why their popularity failed was not to dowith their failure necessarily to predict sitelocations but was related to the changing face ofcultural resource management (Altschul1990:227).

In the late 1970s, the site data bases of many NorthAmerican regions was very meagre, however,regional scale surveys undertaken to rectify thissituation had produced hundreds, if not thousandsof sites. Generally speaking, the survey effort hadproduced enough data to predict the broad patternsof site locations (Altschul 1990: 227).

Apparently the models had done their job, but hadthey? While they provided a broad indication ofarchaeological sensitivity, they did not provide anycontext for assessing the relative significance ofsites located as a result of compliance surveys. Asone American Federal archaeologist remarked -she did not need another model that told her wheresites were but rather a framework for assessing therelative values of some forty seemingly identicallithic scatters (Altschul 1990:227). This scenario isundoubtably familiar to our own forest plannerswho need in addition to models that predict theunknown, frameworks that bring some order anddirection to the increasingly large data bases thatare being amassed.

To explicate some of the limitations and problemswith predictive modelling, recent approaches tomodelling archaeological site locations in forestsin Australia will be briefly discussed.

When considering the merits of variousapproaches, it is important to keep in mind someof the constraints imposed by the characteristics ofthe forest archaeological record. In particular,attention needs to be drawn to the circularity ofcorrelation modelling resulting from the inherentlyskewed sample data produced by surface survey inforests. In addition, correlation models (asgenerally devised by cultural resourcemanagement archaeologists) fail to recognise thatlandscape configuration, scale and processes affectsite location at a range of different spatial scales.

6.2 GIS MODELS (EAST GIPPSLANDEXAMPLE)

It is often assumed that GIS models of site locationwill have greater accuracy than non - GIS models.While GIS offers superior mapping capabilitiesand quantitative and qualitative analysis, it is only

as good as the data that is entered into it andpossibly sometimes worse. Collectingarchaeological data with sufficient rigour for GISmodelling is a very expensive and time consumingtask. Because of the vagaries influencingarchaeological observations, we need to thinkcritically whether GIS is likely to produce a levelof accuracy significantly greater than currentmodelling strategies, or in fact leads to increasedlevels of unacceptable error.

A pilot assessment was undertaken to examine theutility of GIS for modelling archaeological sitelocations in the forests of East Gippsland, Victoria(Lewis 1996).The project was commissioned bythe Australian Heritage Commission as part of theNational Estate cultural heritage assessment of theEast Gippsland RFA.

A number of lessons can be learned from thisproject. Perhaps most important is a recognition ofthe gulf between data requirements for GISmodelling and the sample biases inherent in datacollected from surface surveys in forests.

As has been discussed previously, the locationaldata generated by archaeological surface surveyreflects a range of formational and site detectionlimiting factors. Variable ground surfaceconditions, problems associated with assessingarchaeological visibility and difficulties of accessintroduce a range of factors which greatly reducethe accuracy and the nature of behaviouralinferences that can be made from the data.

This above point is similarly noted by Lewis(1996:34) who raises the likelihood that thepatterns revealed by modelling based onarchaeological point location data sets are onlyartefacts of their sampling methodology (Lewis1996: 34). In concluding he suggests the results ofhis study highlighted those environmentalvariables not that influence site location, but whichinfluence survey accessibility: topographicattributes and values related to slope and wetness(Lewis 1996: 35).

In recording data for GIS it is important thataccurate data is recorded for both site locationsand null observations. However, because of therange of factors determining and obscuring theobservation of archaeological surface materials,determining the status of null observations isinherently problematic. For example, althoughformulaic approaches have been developed toestimate archaeological visibility conditions, theunderlying factors affecting visibility conditions


23

are often poorly understood and a significantmargin of error can be assumed.

Even if full sets of data including null observationscould be systematically recorded, the presence ofbias in the sampling method remains a potentialproblem. Solutions to these problems, such asformal probabilistic sampling strategies, are wellknown (Shennan 1988), however, in practice theyare impractical and often fail to avoid the problemsthey were designed to overcome. This is becausein forested environments observation will alwaysbe limited to a sub-set of topographic features thatprovide ground surface visibility and access,regardless of survey transect location.

Furthermore, interpolation from current GISderived variables also raises questions oflocational accuracy and scale. This is particularlyrelevant for some of the fine scaled variableswhich are fundamental to specifyingarchaeological site location. Variables such aslocal scale terrain strongly correlate with sitelocations, however small errors in this variable cantake the observer from the lowest to highest valuesobserved (ie from a ridge crest to a steeply slopingridge side slope). Obviously errors of this naturewill have a profound affect on the accuracy of sitelocation predictions.

6.3 TASMANIAN MANAGEMENTPRACTICES FORIDENTIFICATION

Management for cultural heritage values isoutlined in Forestry Tasmania's Forest PracticesCode (1993). The code formalises procedures forcultural heritage management in terms of aprogram for preoperational site identification,recording and assessment in addition toconsultation with special interest groups.

More recently Forestry Tasmania has devised astatewide predictive model of archaeologicalsensitivity known as the Archaeological PotentialZoning (APZ) system (McConnell 1995). Thezoning system is designed to improve site capturerates for archaeological sites, with an eye toenhancing the protection of significantarchaeological sites. It is predicted that the systemwill operate for the next two years, after which thesystem will be reviewed (Gaughwin, D., ForestryTasmania pers. comm., 1997).

Under the zoning system, the landscape is dividedinto zones of archaeological sensitivity each with

specific requirements in terms of intensity ofsurvey coverage and survey strategy.

Preoperational and in some cases postoperationalsurveys are required. This year alone (June, 1997)300 archaeological surveys were undertaken as aresult of the zoning system.

The model is formulated on the basis of a range ofdisparate environmental variables selected on thebasis of ‘expert opinion’. The model is primarilybased on the notion that there is a directrelationship between past human behaviour andthe contemporary archaeological record and doesnot address the overarching complexities of siteformation and detection.

Unlike recent work in Victoria and New SouthWales, Tasmanian field recording methods havenot systematically recorded data for factorsaffecting site detection and negative site location.These factors are further compounded by the poorground surface visibility conditions experiencedby surveyors in Tasmanian forests. In terms ofexamining aspects of site formation much of thedata has limited potential. Interpretativeframeworks have also focused on definingAboriginal settlements patterns.

A program of surface survey was recentlyundertaken for Forestry Tasmania by Sim (1996)to test the predictive potential of the model. Giventhe absence of a formal sampling framework totest the model, the results were as would beexpected, inconclusive. While more sites werelocated in areas predicted to have higharchaeological potential, the model was notconsistent and in some survey areas no sites werelocated in areas of high sensitivity. In addition,more sites were located in areas of low sensitivitythan those of medium sensitivity.

In wet forest types ground surface visibilityconditions restricted the utility of surface surveyand it was concluded that the zoning system couldonly be applied to dry sclerophyll forests whereground surface visibility conditions permitted thedetection of archaeological surface materials.

The results of the test not only indicated thelimited utility of surface survey for locatingarchaeological surface materials in Tasmanianforests, in particular wetter forest types, it alsoindicated a conceptual weakness with the zoningsystem. That is, the model does not recognise theextent to which the archaeological record reflects arange of formational and site detection limitingfactors.


24

The model continues to be used in Tasmania,where its success is seen largely as a formal toolfor compliance. While the system will be effectivein mitigating impacts to a narrow range of sites,that is those sites which are easily located bysurface survey, it offers no management for siteswhich are not predicted to occur, or for those areaswhere surface survey is a particularly ineffectivemethod for specifying site location. Furthermore,the absence of a coherent environmentalframework means that it is very difficult todetermine the broad parameters of the datacaptured through the survey process.

6.4 SOUTH EAST FORESTARCHAEOLOGICAL RESEARCHPROJECT

The South East Forests Archaeological ResearchProject was developed by NPWS in 1994 and isfunded by SEFRAC. The project is designed toformulate an archaeological zoning system for theSouth East Forests along the lines of thatinstigated by Forestry Tasmania.

The aims of the project, as described by thecurrent project brief, are as follows. The overallobjective of the project is to ensure that the mostsignificant sites in the South East Forests studyarea are recognised and appropriately protected,while developing processes for salvage orprotection of Aboriginal cultural heritage sites andlandscapes within timber production forests.Prescriptions for archaeological sites andarchaeologically sensitive land units are to be usedto determine requirements for pre-logging surveysand survey techniques to be employed. Processesare to be developed for the determination of whenSection 90 Consents (National Parks and WildlifeAct 1974) may be required to allow timberharvesting to proceed.

Work on the SEFRAC archaeological projectcommenced in 1997. The initial stage of theproject has been concerned with the collation ofexisting archaeological site location data. A highdegree of error in recorded sites has meant that theproject archaeologist's time has been spent on datacorrection. The data has been entered into NPWSWindows version Environmental ResourceMapping System (WinERMs) data base with theintention of using data layers to provideenvironmental correlates for site locations(Heffernen, K., NSW NPWS, pers. comm., 1997).However, the low level of spatial resolution forenvironmental data layers suggest that this may

further increase uncertainty in determiningenvironmental correlates for site locations.

Personal computer based ERMs programs used sofar in Australia for cultural resource managementgenerally do not have fine enough spatialresolution to map most individual landforms orlandscape features. As site location appears fromall previous research to be related to landforms,these must be identifiable in the systems to be usedfor predictive modelling (Sullivan & Hiscock1992:85). Given the absence of this level ofenvironmental information, it seems likely thatWinERMS will be extremely limited as a tool foridentifying archaeological sensitivity and probablyless reliable than that which could be establishedon the basis of first principles.

The second stage of the project, which willcommence in the near future, is a program ofarchaeological survey. The survey will beundertaken to produce systematic data to formulatea ‘predictive model’.

6.5 SFNSW (NORTHERN REGION)

This section of the paper will outline SFNSW(Northern Region) system of management forAboriginal archaeological values. Northern Regionhave formalised their management systems forcultural heritage in the document ‘SFNSWNorthern Region Policies and Procedures forCultural Heritage Management (draft)’. Theseguidelines were developed specifically forNorthern Region in consultation with NPWS,however some of the practices in them have beenadopted by SFNSW Central and Southern Regions(note: this refers to SFNSW regional structure -1992-97).

The draft guidelines articulate policies andprocedures for managing cultural values includingarchaeological values. The document outlinesguidelines for the identification, assessment andmanagement of both the known and unknownresource. The underlying principal for managingarchaeological values is the need to reserve arepresentative sample of the resource. In addition,Aboriginal heritage values are identified andmaintained principally through on goingconsultation with Aboriginal communities atvarious stages of the planning process in additionto field assessment.

Known sites are managed in accordance withlegislation. They are either protected from impacts,or if they are sites of low value, a Section 90


25

Consent may be sought from NPWS if supportedby the relevant Local Aboriginal Land Council/s.

The strategy for managing the unknown resourceis to develop and implement a strategy foridentifying gaps in the reserve system (all tenures)for maintaining areas representative of thearchaeological resource. The method for doing thisis currently being trialed in the Glen InnesManagement Area (Lomax & Fife in prep.).

For sites other than stone artefact scatters,preharvesting surveys are undertaken by SFNSWstaff and/or local Aboriginal communityrepresentatives to ensure that sites are identifiedand protected. SFNSW staff are trained in siteidentification, consultation and site managementprocedures for this purpose.

The document also outlines guidelines forundertaking pre-roading surveys which wereprescribed under some EIS determinations.However, the preferred position of SFNSW(Northern Region) is to undertake further surveywithin the context of those areas not adequatelyrepresented within the reserve system (as yet to bedetermined).

The procedures were developed largely to meet therequirements of Department of Urban Affairs &Planning (DUAP) EIS determinations andlegislation (SFNSW Northern Region nd).

6.6 CONCLUSION

Correlation modelling has inherent limitations interms of producing sound empirical results. Moresignificant even than this weakness is that thesystems described above are not explicitly linkedto a conservation strategy with an effectivemethodology for reserve selection. While all of theabove approaches to predictive modelling will,with varying degrees of success, indicate likelyareas to locate sites, there is no broader frame ofreference for assessment (other than the sitesidentified by the survey process itself) and thesample captured will be largely unspecified. Inaddition, there is no consideration given to thespatial dimension of the resource, or recognitionthat much of its analytical potential is derived fromits examination at a range of spatial scales. Asargued previously, site based management for thesurface record fails not only to recognise the‘systematic’ nature of the resource, but potentiallyleads to a highly skewed, albeit managed, record inthat it protects only that small proportion of the

archaeological record that can be detected oncurrent landsurfaces.

Within the current SFNSW (Northern Region)guidelines, the need to reassess management forstone artefact scatters has been recognised.Establishing the ability of the reserve system (alltenures) to maintain areas representative of thesurface archaeological record has been identifiedas a priority by SFNSW and has been raised anumber of times by NPWS as a possible strategy.The current CRA process provides an opportunityto explore and develop strategic managementoptions that give protection to its significantelements both in respect to its landscaperepresentativeness and highly significant sites,wherever they occur.

Recommendations concerning the key elements inany future strategy for identification andmanagement are presented in Chapter 7.

27

7.RECOMMENDATIONS

7.1 REQUIREMENTS FOR FURTHERSURVEY

Requirements for further regional scale forestarchaeological surveys need to be criticallyexamined in relation to the issues discussed in thispaper. In particular, the extent to which furtherCRA survey work will add significantly to ourability to provide improved management advicefor the RFA. One issue that needs to be criticallyexamined is the extent to which further survey datawill significantly add to the existing corpus ofarchaeological forest data, especially given theinherent limitations of surface survey as a meansto comprehensively define the resource.

7.2 CONSERVATION STRATEGY

Objectives for management that give properconsideration to the spatial aspect of the resourceand how to protect its significant elements, both inrespect to its landscape representativeness andsignificant sites, need to be clearly articulated byan overarching conservation strategy. The strategyneeds to clearly define those values that needmanagement and to set criteria for conservationoutcomes.

Less attention should be given to futuremanagement of identified open sites in workingforests or future identification, especially in highlydisturbed contexts, and instead ensuring adequaterepresentation of such sites within the reservesystem. This will necessitate a paralleldevelopment of a more strategic approach to thegranting of Consents for previously identifiedsites.

A system of management is required to ensure thatthe spatial aspect of the forest archaeologicalresource (all tenures) is maintained. Given thefundamental effects of landscape characteristicsand processes, identification and managementshould be within the context of ‘representative’landscape areas.

The Nationally Agreed Criteria for theEstablishment of a Comprehensive, Adequate andRepresentative (CAR) Reserve System for forestsin Australia (Commonwealth of Australia 1997),state that the reserve system should sample the fullrange of forest communities and ecosystems acrossthe landscape. The National Forest PolicyStatement also defines the need for the CARreserve system to protect old growth forest andforested values in the reserve system(Commonwealth of Australia 1992). The ScopingAgreement for NSW RFAs acknowledges thatheritage issues are to be addressed.

Given that the CAR reserve system is to beestablished spatially across environmental units offorests, as well as in least disturbed areas (oldgrowth and wilderness), it would seem valid toexamine if such a system will allow the reservationof a ‘comprehensive, adequate and representative’sample of the archaeological record acrossforested landscapes. This would be supplementedby off-reserve management techniques for sitetypes other than stone artefact scatters and thoselandscape areas not adequately represented withinthe CAR system.

The methodology for undertaking this task wouldentail an examination of the reserve systemincluding defacto reserve areas (defined bymanagement prescription and zoning controls).The analysis would take into account disturbancehistory and the inherent stability of differentlandscape systems. Data requirements for thisanalysis are likely to be met by data layers createdas a result of the current CRA assessment process.The scale of areas for selection and theenvironmental parameters utilised will needfurther consideration, however, the most relevantparameters will be those that characterise variationin landscape configuration, scale and form. Suchas, in descending order of scale; regionallandsystem, major geological group and dominantlandform pattern. This should at least enable thesetting of rational statewide and regional priorities


28

for further survey work/other managementpriorities.

Because the objective is to capture variability andspatial patterning rather than identifying ‘sites’,environmental data is considered to be the moreefficient for this task than archaeological data.Rather considerations such as disturbance history,in particular the existence of minimally disturbedforest areas, are likely to be a major considerationin the identification of reserve adequacy.

7.3 PRE-HARVESTING SURVEYSAND MANAGEMENT FORDIFFERENT SITE TYPES

Requirements for pre-harvesting survey forarchaeological sites need to be critically assessed.The need for further work should also be assessedon the basis of a realistic appreciation of futurepotential threats taking into account the overallconservation strategy. This will need to becritically assessed in terms of both past cultural(prehistoric, historic and contemporary)disturbance and landform stability.

Given that site types other than stone artefactscatters are rare and therefore of relatively highregional significance, their identification andmanagement need to be site specific orientated andthis is best achieved as part of ongoing operationalplanning. The method adopted should entail a mixof measures depending on the degree of risk tosignificant elements and broader considerations ofAboriginal heritage (ie survey by operational staff,and/or Aboriginal community survey/ consultation,including consideration of post-contact sites andplaces). An example of such a strategy is putforward in the SFNSW (Northern Region) draftguidelines.

29

REFERENCESAltschul, J. H. 1990, Red flag models: the use of

modelling in management contexts. In K. M.S. Allen, S. W. Green & E. B. W. Zubrow.Interpreting Space GIS and Archaeology.Taylor & Francis, London. pp. 229-238.

Binford, L. R. 1968, Some Comments onHistorical Versus Processual Archaeology,Southwestern Journal of Anthropology 24:267-275.

Binford, L. R. 1980, Willow Smoke and Dog'sTails: Hunter-gatherer Settlement Systems andArchaeological Site Formation, AmericanAntiquity 45:4-20.

Bird, C. 1993, Desktop study of archaeology:Central Highlands. A report to the AustralianHeritage Commission, Canberra.

Boot, P. 1993, Pleistocene ArchaeologicalExcavations from the Hinterland of the NewSouth Wales Coast. Australian Archaeology.37. p. 59.

Bowdler, S. 1983, Aboriginal Sites on the Crown-timber lands of New South Wales. Report tothe Forestry Commission of New South Wales.

Byrne, D. 1983a, Archaeological Survey of theWandella-Dampier State Forests. In ProposedFoothills Logging Operations, Wandella-Dampier, Narooma District. ForestryCommission of New South Wales.

Byrne, D. 1983b, The Five Forests: anArchaeological and AnthropologicalInvestigation. National Parks and WildlifeService New South Wales, Sydney.

Byrne, D. 1983c, Aboriginal sites in DeuaNational Park. Unpublished report to theNational Parks and Wildlife Service.

Byrne, D. 1983d, Aboriginal sites in WadbilligaNational Park. Unpublished report to theNational Parks and Wildlife Service.

Byrne, D. 1984, A Survey Strategy for a CoastalForest. In S. Sullivan & S. Bowdler (eds.) SiteSurveys and Significance Assessment inArchaeology. Research School of PacificStudies, Australian University, Canberra. pp.61-70.

Byrne, D. 1987, The Aboriginal andArchaeological Significance of the New SouthWales Rainforests. Forestry Commission ofNew South Wales, Sydney.

Byrne, D. 1990, Archaeological Survey ofProposed Forest Roads in Eden and BombalaDistricts. A report to the Forestry Commissionof New South Wales.

Byrne, D. 1991, Aboriginal archaeology in forests- circles around the past. In D. Lunney (ed.)Conservation of Australia's Forest Fauna.Royal Zoological Society of New SouthWales, Sydney. pp. -385-92.

Byrne, D. 1992, An investigation of theAboriginal archaeological record in theGloucester-Chichester EIS Area. Gloucesterand Chichester Management AreasEnvironmental Impact Statement. SupportingDocument No. 8. State Forests of New SouthWales, Sydney.

Byrne, D. 1995, An investigation of theAboriginal archaeological record in theTenterfield EIS Area. Tenterfield ManagementArea EIS, Supporting Document No. 5


30

Archaeology. State Forests of New SouthWales, Sydney.

Byrne, D. & L. J Smith. 1987, Survey forAboriginal archaeological sites in the StateForests of the Eden Area. A report to theForestry Commission of New South Wales.

Byrne, D. & L. J Smith. 1988, Cultural resourceoverview: Aboriginal sites in the SoutheastRegion. Unpublished report to the NationalParks and Wildlife Service and the AustralianHeritage Commission.

Collins, J. 1991, Report on the archaeologicalsurvey at the Duck Creek environmentalimpact statement area, northern New SouthWales. A report to the Forestry Commission ofNew South Wales.

Collins, J. & M. J. Morwood. 1991, Anarchaeological assessment of the forests of theWingham Management Areas, New SouthWales. Environmental Impact StatementWingham Management Areas Proposed ForestManagement Supporting Report 5 -Archaeology Survey. Truyard. Pty. Ltd.

Comber, J. 1991, Archaeological assessment ofthe Dorrigo Management Area. A report toSinclair Knight Pty. Ltd.

Commonwealth of Australia. 1992, NationalForest Policy Statement. A New Focus forAustralia's Forests.

Commonwealth of Australia. 1997, TheNationally Agreed Criteria for theEstablishment of a Comprehensive, Adequateand Representative (CAR) Reserve System forforests in Australia. A Report by the jointANZECC/MCFFA National Forest PolicyImplementation Sub-committee.

Creamer, H. 1980, The Aboriginal Heritage inNew South Wales. In C. Haigh & W.Goldstein (eds.) The Aborigines of New SouthWales. National Parks and Wildlife Service,Sydney. pp. 88-93.

Curby, P. 1993, European Disturbance History ofMount Marsh & Billilimbra/Washpool StateForests . A report to State Forests of NewSouth Wales.

Davies, S. & A. Stewart-Zerba. 1995, AnArchaeological Assessment of State Forests ofNew South Wales' Coffs Harbour - UrungaManagement Areas. A report to GutteridgeHaskins and Davey Pty. Ltd.

Davies, S. with A. Stewart-Zerba, S. Crangle, L.Lamb, J. Prangnell and M. Salmon. 1995, AnArchaeological Assessment of the State Forestof New South Wales Walcha-Nundle and StyxRiver Management Areas. A report to MitchellMcCotter.

Dunnell, R. C. 1992, The Notion Site. In J.Rossignol & L. Wandsnider (eds.) Space, Timeand Archaeological Landscapes. Plenum, NewYork.

Egloff, B. J. 1979, Mumbulla mountain: ananthropological and archaeologicalinvestigation. National Parks and WildlifeService, Sydney.

Flannery, T. 1995, Future Eaters: An ecologicalhistory of the Australasian lands and people.Reed, Melbourne

Flood, J. 1980, The Moth Hunters: AboriginalPrehistory of the Australian Alps. AustralianInstitute of Aboriginal Studies, Canberra.

Foley, R. 1981, Off-Site Archaeology and HumanAdaption in Eastern Africa. BAR InternationalSeries 97, Oxford.

Forestry Commission of Tasmania. 1993, ForestPractices Code. Forestry Commission ofTasmania.

Geering, K. 1981, Lower Snowy Riverarchaeological survey. A report to the NationalParks and Wildlife Service of New SouthWales, Sydney.


31

Godwin, L. 1990, Inside Information: Settlementand Alliance in the Late Holocene ofNortheastern New South Wales. UnpublishedPh.D. thesis, University of New England.

Gollan, K. 1987, Archaeological Investigations onNewnes Plateau. A report to New South WalesNational Parks and Wildlife Service.

Gollan, K. 1992, Aboriginal Cultural HeritageStudy: North East Forests New South Wales.National Parks and Wildlife Service.

Grinberg Knight & Associates. 1995, AnArchaeological Survey of the Queanbeyan &Badja Forest Management Areas. Queanbeyan& Badja Management Areas EIS SupportingDocument No: 6 Archaeological Survey, StateForests of New South Wales, Sydney.

Hall, R. 1990, Archaeological survey of theSnowy River Management Unit. A report tothe Department of Conservation andEnvironment, Melbourne.

Hall, R. 1991, Far East Gippsland ArchaeologicalSurvey. A report to the VictoriaArchaeological Survey.

Hall, R. 1992a, A predictive archaeologicalassessment of crown-timber lands in the GlenInnes Management Area. In Mandis RobertsConsultants. Glen Innes Forest ManagementAreas EIS. Forestry Commission of NewSouth Wales, Sydney.

Hall, R. 1992b, Artefact density patterns in areasof high relief: a case study form Far EastGippsland. In B. Scougall (ed.) CulturalHeritage of the Australian Alps. Proceedingsof the symposium held at Jindabyne, NewSouth Wales 16-18 October 1991. AustralianAlps Liaison Committee, 1992.

Hall, R. & K. Lomax. 1993a, GraftonManagement Area EIS Supplementary Report(Archaeology). State Forests of New SouthWales, Sydney.

Hall, R. & K. Lomax. 1993b, Casino District EISSupplementary Report (Archaeology). StateForests of New South Wales, Sydney.

Hall, R. & K. Lomax. 1996a, The formation ofopen sites in forests. Paper presentedAustralian Archaeological AssociationConference 5 -7 December 1996, Normanville,South Australia.

Hall, R. & K. Lomax. 1996b, A RegionalLandscape Approach to the Management ofStone Artefact Sites in Forested Uplands inEastern Australia. Australian Archaeology. 42pp.35-38.

Hughes, P. J. & M. E. Sullivan 1978, Apreliminary archaeological survey of the FiveForests, South Coast New South Wales. Areport to the National Parks and WildlifeService.

Hughes, P. J. & M. E. Sullivan. 1981a, Theimpact of bushfires on soils and soil erosion.In P. Stanbury (ed.) Bushfires: their effect onAustralian life and landscape. The MacleayMuseum, University of Sydney.

Hughes, P. J. & M. E. Sullivan. 1981b, Aboriginalburning and Late Holocene geomorphic eventsin eastern new South Wales. Search 12:277-278.

Hughes, P. J. & M. E. Sullivan. 1986, Aboriginallandscape. In J. S. Russell & R. F. Isbell (eds.)Australian Soils: the human Impact.University of Queensland Press, St. Lucia.

Hunt, P. 1993, Hinterland forests of EastGippsland. An archaeological survey of theEast Gippsland Forest Management Areas.Department of Conservation and NaturalResources, Victoria and the AustralianHeritage Commission, Canberra.

Ingold, T. 1981, The hunter and his spear: noteson the cultural mediation of social andecological systems. In A. Sheridan and G. N.


32

Bailey (eds.) Economic Archaeology. Oxford,B.A.R. International Series 96:119-130.

James. R. & B. Conyers. 1995, NPWS NortheastNSW Aboriginal Archaeological Sites Projectfor NRAC 1994. New South Wales NationalParks and Wildlife Service.

Jenkin, J. J. 1986, Western Civilisation. In J. S.Russell & R. F. Isbell (eds.) Australian Soils:the human Impact. University of QueenslandPress, St. Lucia.

Jones, R. 1969, ‘Fire stick farming’. AustralianNatural History. 16, pp. 224-8.

Kinhill Engineers Pty. Ltd. 1995, Morissetforestry Districts An Assessment of AboriginalArchaeological Sites. Morisset ForestryDistrict Environmental Impact Statement.Supporting Document No. 8. State Forests ofNew South Wales, Sydney.

Knight, T. 1996, The Batemans Bay forestsArchaeological Project. Site DistributionAnalysis. Department of Archaeology &Anthropology, Australian National University,Canberra.

Kuskie, P. 1994, An archaeological assessmentwithin Dorrigo 3 Year EIS study area, northcoast New South Wales. A report to ForestryCommission of New South Wales.

Lamb, D. 1986, Forestry. In J. S. Russell & R. F.Isbell (eds.) Australian Soils: the humanImpact. University of Queensland Press, St.Lucia.

Lewis, A. 1996, Archaeological Survey Strata -Easy Gippsland Forest Management Area.Draft report GIS Unit Department ofConservation and Natural Resources (DCNR),Victoria.

Lomax, K. 1994, Comara Range (NPWS Site No.21-5-0054) Excavation, Nulla Five Day StateForest, mid north coast, New South Wales. Areport to State Forests of New South Wales.

Lomax, K. 1995, Supplementary Analysis ofArchaeological Data for Walcha-Nundle andStyx River Management Areas. Appendix 8,Walcha-Nundle and Styx River ManagementArea EIS. State Forests of New South Wales,Sydney.

Lomax, K & R. Fife. in prep. Glen InnesDetermination Aboriginal ArchaeologicalSurvey (Stage 2), State Forests of New SouthWales, Northern Region.

McBryde, I. 1974, Aboriginal Prehistory in NewEngland. Sydney University Press.

McConnell, A. 1995, Archaeological PotentialZoning: A Strategy for the Protection ofAboriginal Archaeological Sites in TasmanianState Forests. Volume 1 - Background,Methods, Use & Discussion. ForestryTasmania.

McDonald, J. 1988, The Proposed Warre WarrenAboriginal Place McPherson State Forests -Archaeological Investigation. National EstateGrant.

Mitchell, P. B. 1995, A review of thegeomorphology and soil stratigraphy at RSI(45-5-982) Regentville. Implications for thearchaeological significance of the sites, andrecommendation for further investigation. Areport to Transgrid.

Norton, T. W. & J. E. Williams. 1992, Habitatmodelling and simulation for natureconservation: a need to deal systematicallywith uncertainty. Mathematics and Computersin Simulation. 33 pp.374-384.

Morwood, M. J. & J. Collins. 1991, Anarchaeological Assessment of the Yarrahapinniand Way Way State Forests, New SouthWales. A report to the Forestry Commission ofNew South Wales.

Packard, P. 1991, Sites and sampling in thesoutheast forests: research design, sampling


33

methodologies and sample selection for aproposed site survey project. A report toNational Parks and Wildlife Service and theForestry Commission of New South Wales.

Packard, P. 1992, An archaeological assessmentof state forests in the Kempsey and Wauchopeforestry Management Areas. A report to theForestry Commission of New South Wales.

Packard, P. (in prep). Aboriginal sites in thesouth-east forests of NSW: results of a sitesampling and predictive modelling program.Report to Forestry Commission New SouthWales and National Parks and Wildlife.

Ross, A. 1986 (ed.) Planning for Aboriginal SiteManagement. A Handbook for LocalGovernment Planners. National Parks andWildlife, Sydney.

Schiffer, M. B. 1987, Formation Processes of theArchaeological Record. University of NewMexico Press, Albuquerque.

Scoping Agreement for New South WalesRegional Forest Agreement between theCommonwealth of Australia and the State ofNew South Wales.

SFNSW. 1994, Eden Management Areas EIS -Main Report A. State Forests of New SouthWales, Sydney.

SFNSW Northern Region. nd. Policies andProcedures for Cultural Heritage Management(draft).

Shennan, S. 1988, Quantifying Archaeology.Edinburgh University Press. Edinburgh.

Sim, R. 1996, Archaeological Potential Zoningtesting project. Forestry Tasmania.

Smith, A. 1995, Archaeological Survey Methodsfor Forested Environments - An Assessment. Areport to Forestry Tasmania and the AustralianHeritage Commission.

Smith, C. 1993, An archaeological investigationof the Urbenville Management Area, north-eastern New South Wales. UrbenvilleManagement Area EIS. Supporting DocumentsNo. 5. State Forests of New South Wales,Sydney.

Stafford, C. E. & E. R. Hajic. 1992, LandscapeScale Geoenvironmental Approaches toPrehistoric Settlement Strategies. In JRossignol & L Wandsnider (eds.) Space, Timeand Archaeological Landscapes. Plenum, NewYork. pp. 137-161.

Sullivan, M. E. & P. Hiscock. 1992, A suggestedcultural resource GIS for the NorthernTerritory. In C. C. Devonport, S. J. Riley & S.M. Ringrose (eds) Proceedings of the GIS andEnvironmental Rehabilitation Workshop 4-5September 1992 Darwin. AGPS, Canberra.

Sullivan, M. E., P. J. Hughes & W. Shawcross.1996, Northeast Forests ArchaeologyInvestigations in Chaelundi State Forests. Areport to National Parks and Wildlife Serviceand State Forests of New South Wales.

Sullivan. S. & S. Bowdler (eds.) 1984, SiteSurveys and Significance Assessment inArchaeology. Research School of PacificStudies, Australian University, Canberra. pp.61-70.

Warren, R. E. 1990, Predictive modelling inarchaeology a primer. In K. M. S. Allen, S. W.Green & E. B. W. Zubrow. Interpreting SpaceGIS and Archaeology. Taylor & Francis,London. pp. 91-111.

Zone Team Accountabilities and OperatingGuidelines. 1995, National Parks and WildlifeService (Internal Working Document).

Date post:	17-May-2020
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times