Tasmanian Forest Carbon Study Full Report 2012

TASMANIAN FOREST CARBON STUDY FULL REPORT

CO2 AUSTRALIA LIMITED 31 JULY 2012

Page 1 of 372

TASMANIAN FOREST CARBON STUDY

Authored by: Barrie May, James Bulinski, Adrian Goodwin & Stuart Macleod

CO2 Australia Limited

31 July 2012

This report was commissioned by the State Government through the Tasmanian Climate Change Office.

Steering Committee: Jim Reid, Brendan Mackey & Cris Brack

Page 2 of 372

ACKNOWLEDGEMENTS

This report has benefited greatly from data provided by a number of independent sources. This includes a very significant set of data and information provided by Forestry Tasmania. Major data contributions were also made by the Wilderness Society of Tasmania, Ian Riley, the University of Tasmania, the Tasmanian Department of Primary Industries Parks Water and Environment, the Department of Climate Change and Energy Efficiency and CSIRO. In some cases, these parties also contributed through providing technical and interpretive support.

Many individual researchers provided useful advice on the modelling approach and interpretation and analysis of data sets. The authors are particularly grateful to Martin Stone, David Mannes, Mike McLarin, Martin Moroni, Tom Kelley and Rochelle Richards from Forestry Tasmania, Christopher Dean, David Bowman, Lynda Prior, Mark Hovenden, Sam Wood and Jon Marsden Smedley from the University of Tasmania, Vica Bayley from the Tasmanian Wilderness Society, Jim Gould, Keryn Paul, Leanne Webb and Craig Heady from CSIRO, Tom Fisk and Arthur Lyons from Private Forests Tasmania, Angus McNeil for the Forest Practices Authority, Don Aurick from Timberland Pacific, Jamie Norris and Rob Law from DSE, Matt Searson and Rob de Ligt from the DCCEE, Steve Candy from the Australian Antarctic Division, David Van Geytenbeek from the Tasmanian Fire Service, and Guy Barrett and Stuart Davey from ABARES.

This study has also benefited greatly from the expert advice, input and support received from the Steering Committee, comprising Distinguished Professor Jim Reid, Professor Brendan Mackey and Associate Professor Cris Brack, as well as staff at the Department of Premier and Cabinet.

The authors and CO2 Australia would like to express their deep thanks for the contributions made by all of these organisations and individuals, without whom the detailed analyses and outcomes presented within this report would not have been possible. We also thank our co-contributors on this report, Ryan Campbell, Patrick Smith, Ryan Warburton, Darren Grant and Chris Mitchell.

Page 3 of 372

IMPORTANT NOTICE The following disclaimer applies to this report and the information contained within it. Please read this before making any use of this report. This report has been prepared by CO2 Australia Limited for the Tasmanian Department of Premier and Cabinet (DPaC). This report does not purport to contain all information that DPaC, or any other party, may require or consider material in respect of evaluating any preferred forestry management regimes, or carbon abatement and commercialisation strategies (Strategies), and has been prepared by CO2 Australia strictly in accordance with the structure and limits of the project scope of works as set by DPaC. CO2 Australia has not verified, or conducted a detailed independent appraisal, of information and third- party data that has been used or referenced in this report. In evaluating a Strategy, DPaC and any other party, must rely on their own enquiries and due diligence. The report is not, and does not purport to be, or contain any, financial recommendations, or investment advice. Any projected financial information, carbon abatement estimates, Strategies, projections or forecasts of future events, or other forward-looking statements contained in this report (Forward Looking Statements) are based on at least some level of subjective estimate and assumption and relate to circumstances and events that have not yet taken place and which may not take place. No representations are made by CO2 Australia as to the accuracy, or reasonableness, of such information, or the assumptions upon which they are based. Forward Looking Statements are subject to significant natural, economic, regulatory and competitive uncertainties, most of which are beyond the control of CO2 Australia and DPaC. Accordingly, CO2 Australia provides no assurance that Forward Looking Statements will occur, or be realised, and variation with actual results may be material. You are cautioned not to place any undue reliance on any Forward Looking Statement. CO2 Australia, its shareholders and their respective subsidiaries and associated companies, and their respective officers, employees, advisers and agents, expressly disclaims any and all liability which may be based on this report, and any errors or emissions therein. In particular, no representation, or warranty, express or implied, is given as to the accuracy or completeness, or the likely achievement, or reasonableness of the Strategies, Forward Looking Statements, forecasts, estimates, prospects, or returns and nothing in this report should be relied on as a promise, or representation as to the future.

Page 4 of 372

TABLE OF CONTENTS

GLOSSARY ...... 1 SUMMARY...... 10 1 BACKGROUND ...... 30 1.1 SCOPE & OBJECTIVES ...... 32 1.2 SCENARIOS ...... 33 1.3 KEY ACTIVITIES & OUTCOMES ...... 40 2 CARBON MODELLING OVERVIEW ...... 42 2.1 VEGETATION CLASSIFICATION ...... 42 2.2 MODEL DESCRIPTION ...... 45 2.3 FOREST CARBON POOLS ...... 45 2.4 PRODUCT CARBON POOLS ...... 47 2.5 FIRE ...... 48 2.6 HARVESTING...... 49 2.7 NATURAL VARIATION AND UNCERTAINTY IN PARAMETERS ...... 51 3 RESULTS ...... 52 3.1 OVERVIEW ...... 52 3.2 CURRENT CARBON STOCKS ...... 53 3.3 TRENDS IN PAST AND FUTURE CARBON STOCKS ...... 54 3.4 STUDY SCENARIOS: NATIVE FOREST MANAGEMENT ...... 71 3.5 STUDY SCENARIO: VARYING PLANTATION MANAGEMENT ...... 97 3.6 CHANGING LAND USE ...... 106 3.7 BASELINE SENSITIVITY SCENARIOS: FIRE REGIMES ...... 119 3.8 BASELINE SENSITIVITY SCENARIOS: CLIMATE CHANGE ...... 123 3.9 CARBON CARRYING CAPACITY ...... 128

Page 5 of 372

4 CARBON MARKETS ...... 134 4.1 TYPES OF MARKETS & CARBON PROJECTS ...... 135 4.2 KEY CARBON PROJECT CONCEPTS ...... 139 4.3 MANDATORY MARKET OVERVIEW ...... 144 4.4 VOLUNTARY MARKET OVERVIEW ...... 163 5 OPPORTUNITY ASSESSMENT ...... 178 5.1 CHOICE OF MARKETS & STANDARDS ...... 179 5.2 WEIGHTING THE OPPORTUNITIES ...... 180 5.3 THE OPPORTUNITIES ...... 186 5.4 KEY RISKS ...... 198 5.5 POLICY INFLUENCES ...... 199 6 FOREST CARBON MODELLING FRAMEWORK...... 202 6.1 CLASSIFICATION OF TASMANIA’S FORESTS ...... 202 6.2 DETAILED DESCRIPTION OF THE MODELLING FRAMEWORK ...... 209 6.3 GROWTH MODEL FOR NATIVE EUCALYPT FORESTS ...... 214 6.4 GROWTH MODEL FOR RAINFOREST & UNDERSTORY SPECIES ...... 223 6.5 GROWTH MODELS FOR PLANTATIONS ...... 224 6.6 GROWTH MODEL FOR NON-FOREST >2M...... 225 6.7 BIOMASS ALLOCATION AND TURNOVER ...... 226 6.8 STAG TURNOVER MODEL ...... 229 6.9 DEBRIS DECOMPOSITION ...... 234 6.10 SOIL CARBON ...... 237 6.11 FIRE ...... 238 6.12 TIMBER HARVESTING ...... 244 6.13 EMISSIONS FROM FOREST MANAGEMENT & HARVESTING ...... 248 6.14 WOOD PRODUCTS ...... 249 6.15 INPUT DATASET ...... 251

Page 6 of 372

6.16 ESTIMATING CARBON CARRYING CAPACITY ...... 251 6.17 MODELLING THE EFFECTS OF CLIMATE CHANGE ...... 252 6.18 UNCERTAINTY ANALYSES...... 254 6.19 KEY ASSUMPTIONS AND DATA GAPS ...... 255 7 BACKGROUND TO MODELLING APPROACH ...... 260 7.1 OVERVIEW OF FOREST CARBON ACCOUNTING ...... 260 7.2 METHODOLOGY USED FOR THE NATIONAL GHG ACCOUNTS ...... 261 7.3 CARBON ESTIMATES FOR AUSTRALIAN FORESTS ...... 270 7.4 ASSESSMENT OF CARBON STOCKS IN TASMANIA’S FORESTS ...... 279 7.5 TASMANIAN FOREST TYPES AND CLASSIFICATION ...... 293 7.6 FOREST MANAGEMENT IN TASMANIA ...... 297 7.7 EFFECTS OF FOREST MANAGEMENT ON FOREST CARBON ...... 305 7.8 EFFECTS OF FIRE ON FOREST CARBON ...... 307 7.9 EFFECTS OF CLIMATE CHANGE ON FOREST CARBON ...... 313 8 COMPARISONS OF MIXED EUCALYPT FOREST AND RAINFOREST .... 318 8.1 COMPARISON OF PAIRED EUCLYPT FOREST/RAINFOREST PLOTS ...... 318 8.2 ANALYSIS OF WILDERNESS SOCIETY DATA ...... 329 8.3 STUDY BY RILEY (2012), UNIVERSITY OF TASMANIA ...... 331 8.4 CONCLUSION ...... 334 BIBLIOGRAPHY ...... 335 APPENDICES ...... 349 APPENDIX 1: KEY ASSUMPTIONS FOR LEAKAGE ASSESSMENT...... 350 APPENDIX 2: FIRE AND HARVESTING PARAMETERS ...... 351 APPENDIX 3: BASE PARAMETERS USED IN FCMF ...... 356

Page 7 of 372

GLOSSARY Commonly used terms

Activity Shifting Occurs where the Project Proponent implements a change in forest Leakage management that results in increased carbon sequestration within the Project Boundary, but then alters management regimes in other forests under their control in a way that reduces carbon stocks in those forests.

Additional The activity would not have occurred under a business as usual scenario.

Administrator The governing body of a verification program (mandatory or voluntary).

Afforestation The establishment of Trees, through direct human-induced methods on land that is naturally clear of forest.

Allometric Regression functions fitted to a scatter of data-points that relate predictor Functions measures collected through a non-destructive measurement process, to a measure of the weight of Biomass within a Tree.

Approved A set of procedures for establishing Carbon Project eligibility and estimating Methodology related emissions abatement which have been approved by the Administrator of a verification program.

Avoided A term used in verification programs to describe an activity that protects Deforestation carbon stocks of an existing forest by implementing change management that protects the forest from permanent removal.

Bark The outermost layers of stems and roots of woody plants and may contain both dead and live tissue.

Basal Area Defines the area of a given section of land occupied by the cross-section of tree stems at a specified height, typically 1.3m above ground level.

Baseline A concept applied under various emissions reduction programs that Scenario describes the dominant land use and land management practices that existed for a defined period preceding Project commencement as well as the emissions and abatement profile associated with that land use.

Biomass Sections of Trees that are divided on the basis of structure and/or form. In Components this Report, Biomass Components that are referenced include Stemwood, Bark, Branches, Leaves, Coarse Roots and Fine Roots.

Page 1 of 365

Big Trees A proxy for fire intensity used to determine the proportion of trees in a stand killed by fire and combustion of debris and soil carbon. The size of a Big Tree can be varied by forest type to simulate different fire sensitivities of different species.

Business as A term describing the management activity prior to Carbon Project Usual implementation and considers abatement and emissions from that activity.

CABALA Process-based tree growth model developed by CSIRO and used to model forest growth.

Carbon Carrying Average total carbon stocks of a forest subject to prevailing environmental Capacity disturbance regimes but excluding human disturbance.

Carbon Carrying Carbon Carrying Capacity Scenario that has been applied under the Capacity Carbon Study, in which fire regimes are assumed to be the same as those Baseline for the Study Baseline.

Carbon Credit A generic term applied in this report to refer, collectively, to a range of verified emissions reduction units under a range of verification programs (mandatory and voluntary) where one carbon credit is equivalent to one tonne of carbon dioxide equivalent (1 t CO2e) abated.

Carbon Dioxide Carbon dioxide mass equivalent, calculated as carbon mass x . Equivalent

Carbon Project An emissions reduction activity registered, or intended for registration, under a formal emissions reduction verification program (mandatory or voluntary) for the purposes of creating carbon credits.

Carbon Stocks The quantity of carbon, expressed as Carbon Dioxide Equivalent, held within Forest Biomass.

CFI Act Carbon Credits (Carbon Farming Initiative) Act 2011.

CFI Regulations Carbon Credits (Carbon Farming Initiative) Regulation 2012.

Coarse Woody Consists of the tree components of the debris carbon pool including fallen Debris stems and branches and finer material such as dead leaves and bark.

Crown Cover The amount of land covered by the outer limits of the Crown (viewed in horizontal cross-section) of a Tree, or collection of Trees. Can be expressed in a variety of ways, including absolute coverage (m2/ha), or

Page 2 of 365

proportional coverage of a defined land area (%).

Dead Material Dead, Tree derived material (e.g. dead branches, hanging bark, and dead leaves) that remain attached to the Tree and suspended above the ground (as opposed to Litter and Dead Fallen Wood, which occur at ground-level).

Debris The dead material of the various tree components, stemwood, branchwood, bark, leaves and roots

Forest Carbon A spatial forest model that models forest carbon stocks, and changes to Modelling these stocks, across all forest types with consideration of management, Framework forest growth and fire.

Forest Class Defined by Forest Tasmania based on site productivity, stand age, cover and number of cohorts, developed for estimating wood production across its forest estate.

Forest Group Grouping of broad forest types by stand structure. There are seven major forest groups in Tasmania.

Hardwood A generic term to describe wood derived from certain flowering trees such as eucalypts.

Improved Forest A term used in verification programs to describe an activity that increases Management carbon stocks of an existing forest by implementing change management.

Kyoto compliant Abatement activities that meet the eligibility requirements of the Kyoto Protocol of the United Nations Framework Convention on Climate Change.

Kyoto Unit(s) Has the same meaning as the Australian National Registry of Emissions Units Act 2011, being: An assigned amount unit, a certified emission reduction, an emission reduction unit, a removal unit or a prescribed unit issued in accordance with the Kyoto rules.

Leakage Net changes of emissions by GHG sources that occur outside the Project Boundary, but are attributed to the Carbon Project.

Litter Dead, Tree-derived material occurring at ground-level, typically including fallen leaves, twigs, bark and small woody stems in various states of decomposition.

Low Native Non-forest native vegetation < 2 m identified from TasVeg mapping. Vegetation

Page 3 of 365

Low Eucalypt Forest dominated by eucalypt species with a potential height at maturity Forest (>110 years) of <34 m. Also known as dry sclerophyll forest.

Mandatory A carbon market that is established through government legislation and/or markets regulation.

Market Leakage A change in forest management activity reduces production of forest- derived commodities in one forest area, only to result in that commodity being extracted from other forests outside of the control of the Project Proponent.

Market Leakage The proportion of carbon abatement occurring within the Project Boundary Factor as a result of reduced timber product extraction that is anticipated to be offset by leakage emissions occurring from a compensatory increase in extraction rates in other mainland Australian forests.

National GHG The reporting of GHG emissions and sequestration from Australian forests Accounts by the Department of Climate Change and Energy Efficiency

Negative List Activities defined under the Carbon Farming Initiative that have been precluded as eligible abatement activities.

Net Present The difference between the value, expressed in terms of current dollars, of Value total cash inflows and outflows, used to analysis the potential profitability of an investment or project.

Non-Kyoto Unit Has the same meaning as the term ‘non-Kyoto Australian carbon credit unit’ defined under the CFI Act, being: an Australian carbon credit unit other than a Kyoto Unit.

Other Native Native forest (> 2 m) dominated by species other than eucalypts and Forest Rainforest species.

Permanence Retention of carbon stocks over a defined period of time, this requirement varying between verification schemes.

Permanent A defined area of land being circular or rectangular in shape, from within Sample Plot which various field-based measurements are taken in order to estimate Carbon Stocks.

Plantation Tree species planted for their end use product. In Australia plantations can be broadly grouped into hardwood and softwood plantations.

Page 4 of 365

Pool (s) A forest or forest products component that stores carbon such as live vegetation, debris, soil, log products and processed products.

Positive List Activities defined under the Carbon Farming Initiative that have been approved as additional abatement activities under that initiative.

Project Takes on the form of several dimensions, including the geographic extent Boundary of the forest, or land tenure boundaries and the range of greenhouse gas sources and sinks included within the carbon accounting framework.

Project The party that is undertaking a Carbon Project. Proponent

Rainforest Temperate Rainforest dominated by species such as Northofagus cunninghammii, Atherosperma moschatum and Lagarostrobos franklinii.

Reforestation The establishment of Trees, through direct human-induced methods on land that has been subject to an historic forest-clearing event.

Risk of Reversal An amount of carbon credits retained by the Administrator of an emission Buffer reduction program, so as to self-insure against carbon stock reversals.

Softwood A generic term to describe wood derived from non-flowering trees (.e.g. Radiata pine).

Soil C The carbon pool captured in the soil regolith and consisting of decomposable plant material, resistant plant material, humus and microbial biomass.

Stags Dead, standing trees that persist within the forest.

Stand A portion of forest in a contiguous area of similar average productivity and structure but, potentially, consisting of a range of species or tree ages

Study Baseline The Baseline Scenario that has been applied under the Carbon Study, which assumes that forest growth, forest management practices, fire frequency and climate are consistent with recent experience.

Tall Eucalypt Forest dominated by eucalypt species with a potential height at maturity Forest (>110 years) of > 34 m. Also known as wet sclerophyll forest.

Tall Native Non-forest native vegetation > 2 m identified from TasVeg mapping. Vegetation

Voluntary A carbon market that has no government regulation and participation is on market a voluntary basis.

Page 5 of 365

Units, abbreviations & notation

ABARES Australian Bureau of Agricultural and Resource Economics and Sciences

ACCU (s) Australian Carbon Credit Unit (s)

ACPM Australian Carbon Pricing Mechanism

AFOLU Agriculture, Forestry and Land Use

APAR Photo-synthetically Active Radiation

C Carbon

CBS Clearfell, Burn and Sow

CCB Climate, Community & Biodiversity Standard

CCC Carbon Carrying Capacity

CCX Chicago climate exchange

CDM Clean Development Mechanism

CFI Carbon Farming Initiative

CH4 Methane cm centimetre

CO2 Carbon dioxide

CO2e Carbon Dioxide Equivalent

CSIRO Commonwealth Scientific and Industrial Research Organisation

CWD Coarse Woody Debris

DBH Diameter at Breast Height

DCCEE Department of Climate Change and Energy Efficiency

DPIPWE Tasmanian Department of Primary Industry, Parks, Water and the Environment.

DOIC Domestic Offsets Integrity Committee

Page 6 of 365

DPaC Department of Premier and Cabinet

DPIPWE Department of Primary Industries, Parks, Water and Environment

ENGO Environmental Non-Government Organisations

ESV Entire Stem Volume

EU ETS European Union Emissions Trading Scheme

FCMF Forest Carbon Modelling Framework

FPA Forest Practices Authority

FPI Forest Productivity Index

FT Forestry Tasmania

FullCAM Full Carbon Accounting Model

FWPA P191

GHG Greenhouse Gas

GIS Geographic Information System

GPP Gross Primary Productivity

Gt Gigatonne ha hectare

IBRA Interim Biogeographic Regionalisation Zones for Australia

IPART Independent Pricing and Regulatory Tribunal

IPCC Intergovernmental Panel on Climate Change kg kilogram km2 square kilometre

LAI Leaf Area Index

LULUCF Land Use, Land Use Change & Forestry m metre

Page 7 of 365

MAI Mean Annual Increment

MIS Managed Investment Scheme mm millimetre

Mt Million tonnes m2 Square metre m3 metres cubed

NCAS National Carbon Accounting System

NCAT National Carbon Accounting Toolbox

NDVI Normalized Difference Vegetation Index

NFI National Forest Inventory

NMVOCs Non-methane volatile organic compounds

NO2 nitrogen dioxide

NPP Net Primary Productivity

NPV Net present value

NSW GGAS New South Wales Greenhouse Gas Reduction Scheme

NZ ETS New Zealand Emissions Trading Scheme

N2O Nitrous oxide pa per annum

PDF Probability Distribution Frequency

Plots A collective term for Permanent Sample Plots and Temporary Sample Plots ppm parts per million

PSP(s) Permanent Sample Plot(s)

REDD Reduced Emissions from Deforestation and Degradation

RGGI Regional Greenhouse Gas Initiative

Page 8 of 365

SOC Soil Organic Carbon

SRES Special Report Emissions Scenarios (IPCC 2000) t tonne

TFIGA Tasmanian Forests Inter-Governmental Agreement

TSP(s) Temporary Sample Plot(s)

TVMMP Tasmanian Vegetation Monitoring & Mapping Program

UNFCCC United Nations Framework Convention on Climate Change y year

VCS Verified Carbon Standard

VCU Verified Carbon Unit

3PG Physiological Principles in Predicting Growth oC Degrees Celsius

% Percentage

Page 9 of 365

SUMMARY

Scope and objectives The Tasmanian Government engaged CO2 Australia Limited to undertake a Forest Carbon Study (Carbon Study) from November 2011 - July 2012. The over-arching aim of the Carbon Study has been to develop a comprehensive picture of carbon stocks in Tasmania’s forests, including all significant carbon stored in living and dead biomass, soils and harvested wood products, as accurately as possible and across all forest types, land tenures and management regimes.

The Carbon Study references historic carbon stocks (to 1990) and provides estimates of future carbon stocks and emissions under a wide range of forest management and land use scenarios, allowing for a comparison of the emissions, or carbon storage profile associated with differing patterns of use. Additionally, the Carbon Study provides an assessment of the potential to monetise carbon sequestration and avoided emissions under various domestic and international carbon trading programs, including possible pathways to the realisation of verified and tradable carbon credits.

The scope and objectives of this Carbon Study have been to:

• Estimate current carbon stocks across all of Tasmania’s forest estate, taking into account forest type, age-class and disturbance, land use histories and management regimes. • Provide an assessment of the stocks and emissions in decadal time periods for the previous two decades (1990-2010). • Provide an assessment of likely future stocks and emissions (in decadal periods to 2050) under a range of scenarios determined in consultation with the Steering Committee. • Develop a clear accounting framework for carbon stocks and flows in Tasmania’s forests. • Identify areas of significant uncertainty with respect to estimates of carbon stocks and assess the magnitude of uncertainty and error in estimates. • Provide an analysis of opportunities to monetise avoided emissions, emission reductions and increased carbon sequestration, with specific reference to the assessment of current and baseline carbon stocks established as part of the Carbon Study.

Scenarios

To examine the possible impacts of changes in forest and land-use practices on Tasmania’s forest carbons stocks, and estimate the potential Carbon Carrying Capacity, a series of test scenarios was developed in consultation with the Steering Committee. These included a Study Baseline which sought to represent contemporary forest and land management patterns, a range of altered forest management and land-use scenarios, referred to as the Study Scenarios,

Page 10 of 365

and a series of fire and climate change scenarios, referred to as the Baseline Sensitivity Scenarios. In addition, the long term Carbon Carrying Capacity of Tasmania’s forests in the absence of harvesting were modelled over a period of 240 years using a range of assumed fire frequencies.

Study Baseline: Business as usual

• Harvest rates and intensities based on current silvicultural regimes and log type recovery rates with an assumed sawlog quota of 300,000 m3/y for public native forests and hardwood plantations and 55,000 m3/y for private native forests. • Fire regimes based on historic fire frequencies and intensities for different forest types. • Areas of forests and plantations based on those in 2010 assuming no future reforestation, deforestation or conversion of native forest to plantations, or expansion or contraction in areas available for harvesting.

Study Scenarios: Altering management regimes in native forests

N1. Reserve 572,000 hectares (ha) of public forest; reduce the annual sawlog cut from public native forests to 155,000 cubic metres per year (m3/y) and the annual peeler log cut from public native forests to 265,000 m3/y. N2. Phase out harvesting in old growth forest by 2020 on both public and private land, but increase harvesting intensity in regrowth forests so as to maintain annual sawlog production from public forest at or near to 300,000 m3/y. N2A. Phase out harvesting of old growth forests by 2020 on both public and private land, and vary harvest intensity in regrowth forests so as to maintain annual sawlog production from public forest at 155,000 m3/y. N3. Vary rotation length in regrowth forests from 60-120 years, with reference to the current rotation lengths of approximately 80 years (as indicated in the literature). N4. Immediate cessation of the harvest of old growth forests on both public and private land with no increase in harvesting of regrowth forests. N5. Immediate cessation of all native forest harvesting on both public and private land.

Study Scenarios: Altering management in plantations

P1. Increase the rotation length of all existing plantations by 10%. P2. Convert all existing pulplog plantations to sawlog plantations by increasing the rotation length to 25 years. P3. Increase the growth rates of existing plantations by 25% through improved management.

Page 11 of 365

Study Scenarios: Altering land use

A1. Conversion of private native forest to plantation, or non-forest uses, continues at the recent rates indicated by Forest Practices Authority and as permitted under the Permanent Native Forest Estate Policy 2009 up until 2015, after which it ceases. A2. Existing un-stocked areas in native forests are regenerated back to native forest (non- harvested). A3. The plantation forest estate expands into cleared agricultural lands at the average annual rate of expansion that occurred from 2000-2010. A4. The plantation forest estate expands into cleared agricultural lands at double the historic average (2000-2010) to a maximum point where 10% of existing cleared agricultural lands are converted to plantations by 2050. A5. Twenty-five percent (25%) of existing pulpwood plantations are not replanted after their first harvest.

Baseline Sensitivity Scenarios: Altering fire frequency

F1. Fire frequency decreases by 10% (against assumed current frequency) for all native forest types. F2. Fire frequency increases by 10% for all native forest types. F3. Fire frequency decreases by 20% for all native forest types. F4. Fire frequency increases by 20% for all native forest types.

Baseline Sensitivity Scenarios: Effects of climate change

C1. Based on SRES A2, using climate projections from the CSIRO Mk 3.0 climate model which predicts the Tasmanian climate in 2050 will be similar to that of 2010 with a 0.6 oC increase in average temperature and 2% decrease in annual rainfall. C2. Based on SRES A2, using climate projections from the CSIRO Mk 3.5 climate model, which predicts the Tasmanian climate in 2050 will be hotter and drier than in 2010 with a 1.6 oC increase in average temperature and an 8% decrease in annual rainfall. C3. Based on SRES A1b, using climate projections from the CSIRO Mk 3.5 climate model which predicts the Tasmanian climate in 2050 will be warmer and drier with a 1.4 oC increase in average temperature and a 7% decrease in annual rainfall.

Page 12 of 365

Carbon Carrying Capacity Scenarios

L1. Modelled carbon stocks at 2250 in the absence of timber harvesting under Study Baseline fire regimes (Carbon Carrying Capacity Baseline) L3. Modelled carbon stocks at 2250 in the absence of timber harvesting assuming a 20% reduction in fire frequency. L4. Modelled carbon stocks at 2250 in the absence of timber harvesting assuming a 20% increase in fire frequency.

Modelling carbon stocks (Forestry Carbon Modelling Framework)

One of the major areas of activity and outcomes of the Carbon Study has been the development of a unique, built-for-purpose Forest Carbon Modelling Framework (FCMF) that models forest carbon stocks, and changes to these stocks, across all forest types within Tasmania. Given the scale of the modelling task, the FCMF has been designed in a way that allows for the rapid generation of large numbers of simulations, representing a significant efficiency improvement over alternate, publicly available approaches.

The FCMF utilises a similar approach for modelling carbon pools as that included within the Full Carbon Accounting Model (FullCAM), but extends its functionality and application by linking modelling processes to detailed input information for all forest types and carbon pools, referencing forest growth, fire and harvesting regime data and incorporating spatial data for climate, soil and forest type, productivity, management and tenure across Tasmania at a one square kilometre level. The FCMF also links forest growth to amounts of harvested wood products produced, allowing full modelling of the impact of changes in timber harvest volumes on carbon stocks, this being a key requirement of the Carbon Study.

Critically, the FCMF incorporates a significant body of actual growth data for native forest and plantations across Tasmania, access to most of which was negotiated under the Carbon Study. Key data providers have included Forestry Tasmania, DCCEE, the Wilderness Society and the University of Tasmania. These data inputs have been incredibly valuable to the outcomes of the Carbon Study and the authors express their thanks and appreciation to all contributors.

Classifying and mapping forests

The Carbon Study included all vegetation across Tasmania classified as woody vegetation, with height >2 m and >20% canopy cover (this being consistent with definitions applied under the Kyoto protocol). To do this, 10 vegetation groups were developed. These groups are based on information from Forestry Tasmania (Forest Classes and Forest Groups) and the TasVeg spatial layer developed by the Department of Primary Industries, Parks, Water and Environment (DPIPWE).

Page 13 of 365

Distribution of forest and native vegetation types across Tasmania in 2010 (Derived from Forest Group and TasVeg maps). Tasmania is covered by 3.4 million ha of native forests and plantations, making it Australia’s most forested state. 2.3 million ha are on public land with 1.1 million ha on private land. In addition, around 1.0 million ha of native vegetation is classed as non-forest, but is accountable in terms of changes in carbon stocks (i.e. > 2 m tall). Of the total forested area, 1.7 million ha is located in formal or informal reserves and a further 0.2 million ha is classified as non-harvestable for economic or other reasons. The total area of native forest available for wood production is around 1.2 million ha.

Page 14 of 365

Filling in data gaps

Two of the key data gaps identified during the course of the Carbon Study are: the lack of data for carbon stocks in pure temperate Rainforest in Australia and the lack of information on the change in carbon stocks across the succession from Tall Eucalypt Forest to Rainforest. For this reason, a series of field measurements were undertaken as part of the Carbon Study.

Sixteen paired measurement plots were selected at locations ranging from the Styx River valley to near Lake Gordon, in adjacent areas of Rainforest and mature (> 110 year old) Tall Eucalypt Forest. Measurements were collected from trees within the plots and allometric functions were used to convert these to estimates of biomass. Analyses of these data indicated that total live and dead biomass was significantly (P<0.01) lower on the Rainforest plots than eucalypt plots. On eucalypt plots, average biomass was 1170 tonnes per hectare (t/ha), equivalent to 2,145 tonnes carbon dioxide equivalent per hectare (t CO2e/ha) and on Rainforest plots average biomass was 590 t/ha (1,080 t CO2e/ha).

The Tasmanian Wilderness Society also provided relevant data from the Tarkine Wilderness area (four measurement plots in Rainforest) and the Styx Valley (five plots in mature Tall Eucalypt Forest), which were analysed as part of the Carbon Study together with published data from a third Tall Eucalypt Forest site in the Florentine Valley (Dean et al., 2012). Total biomass in live trees was found to vary from 1360 t/ha (3,180 t CO2e/ha) at Styx, to 422 t/ha (1,040 t CO2e/ha) at Tarkine (the Rainforest only site) and, as with the measurement data collected under the Carbon Study, indicated that biomass at the Rainforest site was lower than at the two Tall Eucalypt Forest sites. These results were supported by additional data collected as part of a PhD project by Ian Riley from University of Tasmania from paired eucalypt and Rainforest plots.

These results suggest that, following senescence of the eucalypt stand during the transition from Tall Eucalypt Forest to Rainforest, total biomass and carbon stocks decrease and that the carbon content of live vegetation in Rainforest may be less than 50% of the preceding eucalypt forest. Both this relative difference and estimates of Rainforest biomass need to be confirmed with further measurements. However, for the purpose of modelling growth and decline of Tall Eucalypt Forest, the Carbon Study assumed the average biomass of Rainforest (as mapped by Forestry Tasmania and the Department of Primary Industries, Parks, Water and Environment) was around 50% that of mature Tall Eucalypt Forest.

Interpreting the results – Key assumptions & data gaps

The following results provide an indication of the potential effects of changes in management, land use, fire, and climate on carbon stocks in Tasmania’s forests. While these represent the first detailed assessment of carbon stocks and dynamics across all forest types in Tasmania, they should not be considered definitive. Despite an extensive review of relevant literature and rigorous design and testing of the modelling framework, there are important uncertainties and

Page 15 of 365

assumptions in the underlying data that could have substantial impacts on the results. These are covered in detail within the main body of the report, and it is strongly recommended that these are read in conjunction with this summary. For quick reference, some of the key issues identified during data gap and sensitivity analyses are as follows: • Growth rates for native forests - based on plot data from Forestry Tasmania and field measurements (current average standing volumes of existing mature forest with >40% eucalypt cover, assumed to represent the 95% maximum stand volume). • Growth rates of plantations – based on a relationship between DCCEE’s Forest Productivity Index (FPI) and plantation site index using data from Forestry Tasmania. • Native forest harvesting rates - based on forward estimates from Forestry Tasmania and historic harvesting volumes and set to deliver sawlog quotas of 300,000 m3/y from public native forests and hardwood plantations and 55,000 m3/y from privately owned native forests, assuming an average rotation length of 80 years for regrowth native forests. • Plantation harvest rates – based on data on age class distributions and assumed average rotation lengths of 13 years for pulpwood plantations, 25 years for hardwood sawlog plantation and 30 years for softwood plantations. • Proportion of sawlogs produced from logs harvested and removed from site - based on data from Forestry Tasmania and ABARES and assumed to be: o 12-13% for public native forests, o 5-6% for private native forests, o 38% for hardwood sawlog plantations, and o 47% for softwood plantations. • Frequency of wildfire - based on published estimates for different forest types, and modelled as likelihood of fire in any year as follows: o Tall Eucalypt Forest: 0.4% – 0.7%/y (assumed to be greater for regrowth than mature stands) with an average 50% of 80 year old trees killed; o Low Eucalypt Forest: 3.0 – 3.8%/y (assumed to be greater for regrowth than mature stands) with an average 20% of 40 year old trees killed; o Rainforest: 0.05%/y, with an average 75% of 80 year old trees killed; o Other Non-Forest: 4.0%/y, with an average 20% of 80 year old trees killed; and o Plantations: 1.0%/y, with all trees killed. • Carbon losses from debris and soil due to wildfire and regeneration burns - based on limited published data and assumed to vary with fire intensity, • Stem mortality rates - based on a new stem mortality model developed specifically for the Carbon Study from modelled height, diameter and volume growth for different forest types and published and unpublished studies of changes in stocking over time. • Debris production and decomposition rates – based on plot data from Forestry Tasmania and default figures from DCCEE.

Page 16 of 365

Accounting for uncertainty

'Natural' variation in harvesting and fire events and uncertainty in parameter values were taken into account using a probabilistic approach. The likelihood of an individual stand (i.e. grid point) being harvested or burnt in any year was determined using input probability distributions.

A set of primary (or central) estimates of carbon stocks and emissions was first generated. To do this, all parameters were set at values considered to be most likely to represent their true average. To enable valid comparison between the Study Baseline and Study Scenarios, the same set of simulation conditions were applied between the two, so as to remove potential effects of the randomisation process. As an example, if an individual stand was burnt or harvested under the Study Baseline, the same stand was assumed to be burnt or harvested under the Study Scenario, unless the latter was specifically modelling a change in fire or harvest parameters.

The effect of uncertainty in model parameters on the primary estimates of carbon stocks and emissions was estimated by allowing parameter values to vary within a predefined range and examining the effect of this on the results. To do this, each input parameter for the model was assigned a mean and an uncertainty value (coefficient of variation) based on the degree of confidence in its estimated value and reported variation. The Study Baseline and each Study Scenario were then run over a series of iterations with each parameter allowed to deviate randomly within an assumed normal distribution. Coefficients of variation were calculated for each model output and these were used as the basis for estimating variation in primary outputs.

The effects of these uncertainty analyses are reflected in the presentation of results through this report, which present carbon stock and emissions estimates as ranges, rather than absolute values. These ranges represent one standard deviation from the primary estimate, as calculated from the outputs of the iterative simulation process described above.

Study Baseline: Current carbon stocks

The current total carbon stock in live vegetation in Tasmania’s forests is estimated to be 1,400 – 1,900 million tonnes (Mt) of carbon dioxide equivalent (CO2e). A further 1,000-1,500 Mt CO2e is contained in forest debris and dead trees, with 600-900 Mt CO2e contained in soil (< 30cm). The total amount of carbon stored in vegetation, debris and soil of native forests and plantations is estimated to be 3,000 – 4,400 Mt CO2e. An estimated 97% of the total carbon is contained in native forests, with hardwood and softwood plantations containing 3%.

Page 17 of 365

Modelled distribution of carbon stocks in live vegetation (>2m and >20% cover) across Tasmania in 2010. The highest density carbon stocks are located in highly productive forests in the north-west, southern and the north eastern regions.

Average modelled carbon density in vegetation (amount per ha) is greatest for Tall Eucalypt Forests (averaging 560-730 t CO2e /ha) and least for Other Native Forests (90-120 t CO2e/ha). Estimated current carbon density in plantations averaged 130-150 t CO2e/ha CO2e. Across Tasmania, carbon densities were highest in the more productive forests in the southern, central, north west and north east regions of the state (> 600 t CO2e/ha) with lower carbon densities in the drier forests of the south east and alpine vegetation covering much of the south west.

Page 18 of 365

Study Baseline: Changes in stocks and emissions over time

Carbon stocks have tended to remain relatively stable over time, with the total amount of carbon in vegetation and debris of native forest varying from 2,500-3,100 Mt CO2e in 1990 to 2,500- 3,300 Mt CO2e in 2010. Over the same time period, carbon stocks in plantations are estimated to have increased from 14-17 Mt CO2e to 53-65 Mt CO2e.

Under current harvesting patterns, fire regimes and environmental conditions, the total amount of carbon in vegetation and debris in native forests is expected to increase by around 1.0-1.7% (21-61 Mt CO2e) to 2050. Over the same period, carbon stocks in vegetation and debris in plantations are expected to increase by about 56-65 Mt CO2e, reaching 109-131 Mt CO2e in 2050. Carbon stocks in soil are expected to increase by 42 to 73 Mt CO2e in native forests, but could decrease slightly in plantations by 5-8 Mt CO2e.

3500 e)

2 3000 Live vegetation

2500 PSW PHW RFT ETF 2000 ELF ONF

1500 Debris PSW PHW 1000 RFT ETF 500 ELF ONF Total C in live vegetation (Mt CO 0 1990 2000 2010 2020 2030 2040 2050 Year

Modelled carbon stocks in live vegetation and debris in different forest types from 1990 to 2050 under the Study Baseline. Currently 2,500 - 3,400 Mt CO2e are estimated to be stored in live vegetation and debris. This amount is expected to increase slightly over time to 2,600-3,500 by 2050. PSW: Plantation Softwood, PHW: Plantation Hardwood, RFT: Rainforest, ETF: Tall Eucalypt Forest, ELF: Low Eucalypt Forest, and ONF: Other Native Forest.

Based on the modelled increases in carbon stocks in native forests and plantations over time for the Study Baseline, Tasmania’s forests are expected to sequester an average 3-4 Mt CO2e/y between 2010 and 2050. Emissions from fossil fuel use and non-greenhouse gas emissions from regeneration burns in native forests and plantations are expected to average 0.2-0.3 Mt CO2e/y and thus have a relatively small impact on total carbon fluxes as compared with the total amount of carbon sequestered. CO2 emissions from regeneration burns are included in the net change in carbon stocks for harvested native forests and plantations.

Page 19 of 365

Wood products harvested between 2010 and 2050 and processed and used domestically are estimated to contribute a total of 10-12 Mt CO2e to carbon stocks (including losses through decomposition). The reason for the relatively low contribution from wood products is that a large proportion of forest products are exported, either in the form of woodchips, or as finished products such as veneer. Under current accounting rules used by the Department of Climate Change and Energy Efficiency, carbon in these exported products is assumed to be emitted as CO2. However, in reality, a proportion of this carbon is stored in wood products processed and used overseas (or imported back into Australia).

Study Scenarios: Effect of varying management of native forests

Results from the native forest management Study Scenarios indicate there is potential to increase carbon stocks by reducing harvesting rates, or increasing rotation length. Results for Study Scenario N1, under which 572,000 hectares (ha) of publicly owned forest are reserved and the rate of sawlog harvest is reduced from 300,000 m3/y to 155,000 m3/y, suggest that between 98 and 114 Mt CO2e may be sequestered in live vegetation and debris (here termed forest carbon stocks) by 2050 relative to the Study Baseline. Reducing the rate of harvesting in this way could reduce the total pool of carbon in wood products by an estimated 2 Mt CO2e.

Results for Study Scenarios N2 and N2A differed markedly. Under the Study Scenario N2, (which includes phasing out old growth harvesting while maintaining sawlog production at 3 300,000 m /y) carbon stocks in public forest decreased by around 3 Mt CO2e despite the reservation of old growth forest in 2020. This effect resulted from the increase in harvest intensity required in unreserved forest in order to meet the 300,000 m3 sawlog quota. In contrast, when the sawlog quota was reduced to 155,000 m3/y (N2A) for public land, forest carbon stocks increased by 88-113 Mt CO2e due to the reduction in harvesting intensity across the forest estate. Similarly, in private forests, where harvesting was assumed to decrease in line with the reduction in harvestable area, modelled forest carbon stocks also increased slightly (4-6 Mt CO2e).

Results for Study Scenario N3 suggest that carbon stocks are also sensitive to changes in forest rotation lengths. Extending the average rotation length of regrowth forests from 80 years to 120 years increased forest carbon stocks by 46-59 Mt CO2e. In contrast, reducing rotation lengths to 60 years decreased carbon stocks by 37-46 Mt CO2e.

Results for Study Scenarios N4 and N5 indicate the potential effect of cessation of timber harvest in old growth forest (N4) and across all native forest (N5). Under N4, where old-growth forest was reserved while harvesting rates in unreserved forest were maintained at the same rates as the Study Baseline, forest carbon stocks increased by 36-46 Mt CO2e by 2050. Where all native forest harvesting ceased across public and privately owned forests (N5), forest carbon stocks increased by between 210 and 270 Mt CO2e.

Page 20 of 365

300

250

200 e) 2

150 Vegetation 100 Debris Baseline (Mt CO 50

0 N1 N2 N2A N3_120Y N4 N5 Change in carbon stocks relative to Study to Study relative stocks carbon in Change

-50

Scenario Modelled changes in carbon stocks live vegetation and debris by 2050 for the native forest management Study Scenarios relative to the Study Baseline. Carbon stocks increased for most Study Scenarios with the greatest increases associated with reductions in harvested volumes. Error bars are standard deviations of means.

Under both Study Scenarios N4 and N5, the increase in carbon stocks was greatest in the first 20 years. After 2030, the change in stocks plateaued for the Study Scenario N4 and slowed for Study Scenario N5.

Study Scenarios: Effect of altering plantation management

The plantation management related Study Scenarios include two that increase rotation length (P1 & P2) and one that increases growth rates (P3). A 10% increase in rotation length of all plantations (P1) resulted in an 8-11 Mt CO2e increase in forest carbon stocks by 2050, while conversion of hardwood pulpwood plantations to longer rotation sawlog plantations (P2) increased forest carbon stocks by 44-59 Mt CO2e. Increasing growth rates of all plantations (P3) resulted in a 32-40 Mt CO2e increase in forest carbon stocks.

Page 21 of 365

50 e)

2 40 Vegetation 30 Debris 20

10 Baseline (Mt CO 0

-10 Change in carbon stocks relative to Study to Study relative stocks carbon in Change -20 P1 P2 P3 A1 A2 A3 A4 A5

Scenario

Modelled changes in carbon stocks in live vegetation and debris by 2050 for the plantation management (P1-P3) and landuse change (A1-A5) Study Scenarios relative to the Study Baseline. Large increases in carbon stocks were associated with conversion of pulpwood plantations to sawlog production, increasing plantation growth rates, or establishment of new plantations on cleared grassland. However, conversion of existing plantations to grassland resulted in a large decrease in carbon stocks. Error bars are standard deviations of means.

Study Scenarios: Effect of altering land use patterns

Five Study Scenarios investigated the effects of altered land use patterns, including native forest conversion (A1), replanting unstocked areas of native forests (A2) and either increasing (A3 & A4), or reducing, the total area of hardwood plantations (A5).

Under A1, continued conversion of native forest, up to the current limit of around 10,000 ha by 2015, initially decreased forest carbon stocks by 1-2 Mt CO2e. However, carbon stocks gradually recovered to be 1-3 Mt CO2e above the Study Baseline by 2050. This recovery was a result of reduced harvesting of remaining native forest combined with faster growth and increased sawlog yield from newly established plantations (assumed to be planted on 82% of the area cleared) compared with the relatively low productivity of the original eucalypt forest. Reforestation of cleared land with plantations boosted carbon stocks by reducing the need to harvest the remaining native forest. In contrast, permanent removal of trees across 18% of the total area resulted in a total loss of 1 Mt CO2e over the 40 years.

Page 22 of 365

Under A2, replanting land areas currently classified as ‘unstocked native forest’ (41,000 ha) with native eucalypt forest resulted in no net carbon benefit over the 40 years. This was due to an initial reduction in carbon stocks through removal of existing vegetation and debris in the ‘unstocked’ areas and the relatively slow growth of the newly established forest. Over a longer time frame, it is expected that carbon stocks would increase to exceed current carbon stocks.

Establishment of plantations across 10% of existing cleared land (A3 and A4) resulted in a substantial increase in forest carbon stocks (24-33 Mt CO2e at current planting rates, or 25-33 Mt CO2e for double the current rate). In contrast, removal of 25% of existing hardwood plantations after harvest (A5) resulted in a reduction in forest carbon stocks of 12-15 Mt CO2e. Baseline Sensitivity Scenarios Given how critical assumptions around the Study Baseline are to estimating net abatement or emissions amounts, a series of Baseline Sensitivity Scenarios were used to test the influence of varying rotation length (tested in the native forest management Study Scenarios), varying fire frequency and varying climate.

Estimates were found to be highly sensitive to fire frequency, with a 10% increase in fire frequency resulting in a 25-30 Mt CO2e reduction in forest carbon stocks over 40 years for the Study Baseline. Similarly, decreasing fire intensity by 10% resulted in a 22-31 Mt CO2e increase in forest carbon stocks.

The effect of climate change was tested under three alternative climate projections using published modelled changes in growth rates from ABARES and associated downscaled climate projection data from CSIRO. These analyses indicated that changes in temperature and rainfall could have major impacts on carbon stocks. Increases in growth rates of plantations and native forests under Baseline Sensitivity Scenarios C1 (warmer with similar rainfall to that currently experienced) and C2 (hotter and drier), based on results from ABARES modelling, resulted in large increases in carbon stocks in vegetation (50-240 Mt CO2e over 40 years), while a decrease in tree growth in Baseline Sensitivity Scenario C3 (warmer and drier) reduced vegetation carbon stocks by 30-40 Mt CO2e.

The effect of projected climate change on soil carbon differed from that for vegetation. Warmer climates which tended to increase forest growth also tended to increase decomposition rates while drier climates tended to reduce decomposition rates. As a result, soil carbon stocks decreased by 5-7 Mt CO2e for Baseline Sensitivity Scenario C2, but increased by 12-17 Mt CO2e for C3. The effect of climate change on debris was influenced by both growth rate and decomposition rate with an increase in debris carbon for C1, but decreases for C2 and C3.

Page 23 of 365

80 Total

60 Vegetation Debris e)

2 40 Soil

0 -20% -10% 0 10% 20% -20 Study Baseline (Mt CO (Mt Study Baseline -40

Change in forest carbon stocks relative to relative stocks carbon forest in Change -60

-80 Change in fire frequency (%)

Effect of changes in fire frequency on carbon stocks in vegetation and debris by 2050 relative to Study Baseline. There was a linear relationship between the fire frequency and change in carbon stocks relative to the Study Baseline with a 20% increase in fire frequency resulting in a 50-60 Mt CO2e decrease in carbon stocks. Error bars are standard deviations of means.

400

300 e) 2

200 Live Vegetation

100 Debris

Soil 0 Study Baseline (Mt CO (Mt Study Baseline

-100 Change in forest carbon stocks relative to relative stocks carbon forest in Change

-200 C1 C2 C3 Climate Scenario

Changes in carbon stocks in live vegetation and debris by 2050 under alternative climate change Baseline Sensitivity Scenarios relative to the Study Baseline. Different Climate Change projections resulted in future carbon stocks either increasing or decreasing compared with the Study Baseline which was based on current climate. Error bars are standard deviations of means.

Page 24 of 365

Carbon Carrying Capacity Carbon Carrying Capacity is defined as the potential future carbon stock under prevailing environmental conditions in the absence of human disturbance. The potential Carbon Carrying Capacity of Tasmania’s native forests was estimated by modelling forest growth over 240 years (2010 to 2200) in the absence of timber harvesting and any land clearing, or forest conversion. All native forest types across public and private land were included.

The Carbon Carrying Capacity under current fire regimes was estimated to be 1,700-2,000 Mt CO2e for live vegetation, 1,100-1,800 Mt CO2e for debris and 700-1,000 Mt CO2e for soil, with the total potential carbon content of vegetation and debris estimated to be 2,900-3,900 Mt CO2e. This figure represents an increase of 300-800 Mt CO2e (12-26%) over current carbon stocks in vegetation and debris.

Varying fire frequency by +/- 20% resulted in only a 3% change in the estimated Carbon Carrying Capacity of vegetation and a 4-5% change in the estimated Carbon Carrying Capacity of debris and soil. This represented a 20% difference in the overall increase in carbon stocks live vegetation, debris and soil over the 240 years. However, even with a 20% increase in fire frequency, carbon stocks continued to increase over time in the absence of timber harvesting; only plateauing after 2200.

5000 e) 2 4000

Live 3000 Debris Soil 2000 Low Fire High Fire 1000 Total forest carbon (Mt CO (Mt carbon forest Total

0 2010 2050 2100 2150 2200 2250 Year

Modelled potential increase in forest carbon stocks over time in the absence of timber harvesting under current fire regimes and with a 20% increase (red line) and 20% decrease (blue line) in fire frequency. The results indicate that total carbon stocks in live vegetation, debris and soil could increase by 300-800 Mt CO2e over 240 years. Increasing fire frequency resulted in a smaller increase in vegetation carbon stocks while decreasing fire frequency resulted in a larger increase over time.

Page 25 of 365

Carbon markets

The introduction of the Australian Carbon Pricing Mechanism (ACPM) will create a major domestic mandatory market for eligible carbon abatement. The linked Carbon Farming Initiative (CFI) allows for recognition of a specific set of eligible forest management activities that generate abatement. For eligible forest management activities, it is possible to generate one of two types of credit: Kyoto-compliant ACCU’s (Kyoto ACCU’s), or Non-Kyoto compliant ACCU’s (Non-Kyoto ACCU’s). Critically, only Kyoto ACCU’s can be used by emitters to offset their mandatory emissions compliance obligations under the ACPM.

This has important implications for abatement occurring in Tasmanian forests, since, by virtue of their recognition under a mandatory market (the ACPM); there will be a substantially stronger market demand for Kyoto ACCU’s. In comparison, the main market for Non-Kyoto ACCU’s is likely to be the much smaller, lower value, voluntary market.

Key challenges identified with participation under the CFI included:

• lack of Approved Methodologies for almost all of the activities contemplated by the Study Scenarios; • most activities are presently not listed on the CFI Positive List; • additionality may be challenging to demonstrate for some of the Study Scenarios that contemplate reduced log extraction rates if there is an existing business as usual trend for this reduction, or where legislative instruments are the underlying driver of changes to forest management regimes; • there are significant barriers to entry for plantations (single-species) established in areas where average rainfall is ≥600mm per annum; • activities contemplated under Study Scenario N3 may be excluded by virtue of Section 27(j) of the CFI Act; and • any abatement occurring in plantations established under Managed Investment Scheme (MIS) structures is excluded from participating.

In addition to the CFI, key features of the international voluntary marketplace and the Verified Carbon Standard were reviewed. A number of Approved Methodologies exist under the VCS that have application to the Study Scenarios. Key issues associated with recognition under the VCS include:

• it is presently not possible to recognise Kyoto eligible abatement, at least for reforestation activities, by virtue of this abatement being referenced in Australia’s international reporting (creates a double-counting issue); • as for the CFI, additionality may be challenging to establish in some specific instances.

Page 26 of 365

Projected abatement for 2013-2050 by tenure, with leakage factor and risk buffer netted. Tenure/scenario Public Private Total (x 1 million t CO2e) (x 1 million t CO2e) (x 1 million t CO2e) N1. Reserve 572,000 ha, reduce cut from public native forests to 155,000m3 pa sawlog, 265,000 m3 pa peeler log. 81 – 93 N/A 81 – 93

N2. Phase out old growth harvesting by 2020, increase harvest in No net abatement regrowth forest to maintain 300,000m3 pa sawlog. 3 – 4 3 – 4

N2A. Phase out old growth harvesting by 2020, vary harvest in regrowth forest to maintain 155,000 m3 pa sawlog. 74 – 97 3 – 4 77 – 101

N3. Vary rotation length for harvest in regrowth from the Study Baseline (80 years) up to 120 years. 23 – 29 12 – 15 35 – 44

N4. Immediate cessation of old growth harvest (public & private), no increase in harvesting of regrowth. 19 – 22 9 – 11 28 – 33

N5. Immediate cessation of all native forest harvesting on public & private land. 96 – 111 77 – 88 173 – 199

Tenure/scenario Kyoto <600mm Kyoto ≥600mm Non-Kyoto Total (x1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) Public

P1. Increase plantation ~0 ~0 2 – 3 2 – 3 rotation length by 10%. P2. Convert pulplog ~0 ~0 ~1 ~1 plantations to sawlog. P3. Increase plantation growth ~0 ~0 9 – 11 9 – 11 rate by 25%. Private

P1. Increase plantation ~0 1 – 2 ~3 4 – 5 rotation length by 10%. P2. Convert pulplog ~0 9 – 12 19 – 25 28 – 37 plantations to sawlog. P3. Increase plantation growth ~0 4 – 5 9 – 11 13 – 16 rate by 25%.

Tenure/scenario Kyoto <600mm Kyoto ≥600mm Non-Kyoto Total (x 1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) Public

A2. Unstocked areas in native Uncertain Uncertain Uncertain <0 forests are regenerated. Private

Page 27 of 365

More broadly, key challenges in participating in voluntary markets include the relatively low market size, low value being realised for carbon credits ($6 per credit, global weighted average for 2011) and general oversupply of carbon credits in the market. The introduction of the Non- Kyoto Fund will establish the Australian government as a large voluntary purchaser (Non-Kyoto ACCU’s created under CFI), which should substantially boost the domestic voluntary market ($250 million over six years).

Accessing the opportunity

On balance, while no option would be without substantial challenges, it was considered that there were realistic options available for most of the abatement generating Study Scenarios to participate in either mandatory or voluntary markets, or both in some instances. Exceptions to this were Study Scenarios P1 & P3, which reference a modest increase in plantation rotation length and increased growth through management (respectively). In both cases, it was considered the case for additionality was not strong and that this may limit the successful structuring of Carbon Projects around these activities.

The CFI Regulations indicate Reforestation and Avoided Deforestation are activities that are potentially eligible as Kyoto offsets projects, meaning they are potentially eligible for the creation of Kyoto ACCU’s. Study Scenarios that reference plantation expansion (A3 & A4), and to a lesser extent altered management of existing plantations (P2) have potential as Reforestation projects. If avoided, the deforestation pattern contemplated by Study Scenario (A1) could be structured as an Avoided Deforestation project.

The Study Scenarios that contemplate changes to native forest management (N1 - N5) are likely to be considered a form of Improved Forest Management project. Under current policy settings and structuring of the CFI, abatement arising from this type of activity would presently not qualify for the creation of Kyoto ACCU’s, so that the target market would effectively be limited to voluntary markets. In these cases, proponents effectively have the option of pursuing registration and Non-Kyoto ACCU generation under the CFI, or registration and VCU generation under the Verified Carbon Standard.

Valuing the opportunity

Abatement associated with the Study Scenarios was separated into estimates of potentially Kyoto eligible and non-eligible, split by tenure (private verse public). To consider that proportion of abatement that might be available for the creation of carbon credits, the amount of abatement occurring from 2013-2050 for each Study Scenario was estimated (with consideration of uncertainty in modelling) and allowances for risk buffer (5 - 7.5%) and market leakage (8% for native forest, 17% for plantations) factors were netted.

Net Present Value (NPV) estimates were generated in relation to abatement arising from the Study Scenarios (where a commercial Carbon Project was successfully established, and various

Page 28 of 365

commercial carbon project structuring options. A range of carbon pricing scenarios were referenced with relevance to mandatory and voluntary markets. It is recommended that the assumptions and approach that are fully detailed in the main body of the report be carefully reviewed. It is also recommended that further sensitivity analyses, including around assumed discount rates, be conducted in future commercial assessments.

Critically, in estimating NPV, no consideration was made of opportunity costs arising from, for example, reduced extraction of timber products, as these were beyond the scope of the Carbon Study. In effect, this means that while it is reasonable to use these NPV estimates as a comparison against ‘doing nothing’ following a prior decision to reduce extraction, it is not reasonable to infer the NPV represents the net value that actually arises due to making the decision. For the latter, opportunity costs associated with foregone revenue would need to be considered.

Based on the approach to estimation of NPV that was applied under the Carbon Study, positive returns are associated with establishing Carbon Projects around all of the net abatement generating native forest management Study Scenarios considered, with highest NPV’s associated with N1 (Public: $80-240 million), N2A (Public: $80-250 million) and N5 (Public: $80- 280 million; Private: $80-230 million).

Positive NPV estimates were associated with Kyoto eligible abatement occurring at <600mm, arising from Study Scenarios that referenced expansion of the plantation estate, being A3 & A4 ($60-170 million). For Kyoto eligible abatement occurring at ≥600mm rainfall, there are some considerable eligibility barriers to achieving the issue of Kyoto ACCU’s under the CFI. However, if these could be overcome, the rewards could be substantial where these plantation expansion scenarios arise, with NPV estimates ranging from $270-760 million.

For the plantation management related Study Scenario P2, NPV’s were positive where a Carbon Project was structured around privately owned, non-Kyoto eligible abatement (Private: $10-60 million). In the case of potential Avoided Deforestation (subset of Study Scenario A1) projects, which would likely be Kyoto-eligible, NPV estimates are also positive ($10-30 million).

Page 29 of 365

1 BACKGROUND

Among the many environmental, social and economic benefits that forests deliver, their role in the global carbon cycle and the part that they may play in reducing greenhouse emissions and mitigating the effects of climate change is now more sharply in focus than ever before. Forests are an important carbon store and, globally, it has been estimated that living vegetation stores the greenhouse gas equivalent of close to two trillion tonnes of carbon dioxide, representing over 60 times the total annual global emissions from fossil fuels (IPCC 2001). On an annual basis, the world’s forests and forest soils are estimated to sequester the equivalent of around seven billion tonnes of carbon dioxide each year, representing up to one third of total annual anthropomorphic carbon dioxide emissions (Houghton 2003a, 2003b).

In Australia, forests are also recognised as important carbon stores and contributors to greenhouse emissions abatement. Our forests and forest soils have been estimated to store the equivalent of over 44 billion tonnes of carbon dioxide (Montreal Process Implementation Group for Australia 2008), representing over 70 times the current annual emissions by Australia (DCCEE 2011a). In 2009, net sequestration by Australian forests was estimated to be 55 million tonnes of carbon dioxide (excluding carbon stored in harvested wood products), which is the equivalent of 9% of net total emissions for the year (DCCEE 2011b).

Just as forests can be important carbon stores, they also have the potential to be large sources of greenhouse emissions. Carbon sequestration by forests varies greatly over time and location, with the rates of forest removal, forest establishment, forest harvesting patterns, forest growth rates, forest management practices, natural disturbance events and changes in climate all being important influences. During the 1990’s, land-use change and the removal, or degradation, of forests from around the world are estimated to have resulted in the emission of 1 over eight billion tonnes of carbon dioxide equivalent (t CO2e) , mainly as a result of tropical deforestation (Houghton 2003b). In 2003, by virtue of extensive wildfires in the alpine regions of Victoria and New South Wales, Australia’s forests actually represented a net emissions source of over 150 million tonnes CO2e (DCCEE, 2011b).

After decades of developing and testing policy options around dealing with greenhouse emissions and climate change, there is now a global trend toward the adoption of greenhouse gas reduction targets, or caps, and the use of market-based mechanisms to drive the behavioural change required to achieve these. A number of governments, including Australia, have already implemented mandatory greenhouse gas reduction compliance programs that attach a price to greenhouse emissions. With the introduction of these programs, a large global market-place for the trade of emissions units, often referred to as carbon credits2, has emerged.

1 Throughout this report, greenhouse emissions and sequestration amounts are reported as tonnes of carbon dioxide equivalents (t CO2e). 2 Carbon credit is used throughout this report as a simplified, generic means of referring to the various verified greenhouse gas reduction units, emissions allocation units, or allowances, under various mandatory and voluntary compliance programs. One carbon credit is equivalent to an amount of greenhouse gas that has a global warming potential equivalent to one tonne of carbon dioxide (1 t CO2e). In formal emission

Page 30 of 365

The importance of forests and their role in the global carbon cycle is increasingly being recognised within this market place, with many existing and emerging programs allowing for the issue of carbon credits against carbon sequestered by eligible forest management activities.

With over three million hectares of existing forest assets, Tasmania is Australia’s most heavily forested State, making it potentially well placed to deliver greenhouse gas abatement, to help Australia meet its emissions reduction targets and to participate in carbon markets. However, the forest system is diverse, with varying forest types, each with their own carbon storage and natural disturbance profile, varying land ownership arrangements, and differing patterns of forest use, from permanent reserves to harvested production forest and commercially managed plantations. Attempting to accurately track carbon flows across all of these forest elements is a complex and challenging task, but is nevertheless critical for understanding how large-scale changes to forest management and land use practices may affect Tasmania’s forest carbon stores and potential for greenhouse emissions, or abatement.

The Tasmanian Government engaged CO2 Australia Limited to undertake a Forest Carbon Study (Carbon Study) from November 2011 - July 2012. The over-arching aim of the Carbon Study has been to develop a comprehensive picture of carbon stocks in Tasmania’s forests, including all significant carbon stored in living and dead biomass, soils and harvested wood products, as accurately as possible and across all forest types, land tenures and management regimes.

The Carbon Study references both historic carbon stocks (to 1990) and provides estimates of future carbon stocks and emissions under a wide range of forest management and land use scenarios, allowing for a comparison of the emissions, or carbon storage, profile associated with differing patterns of use. Additionally, the Carbon Study provides an assessment of the potential to monetise carbon sequestration and avoided emissions under various domestic and international carbon trading programs, including possible pathways to the realisation of verified and tradable carbon credits.

This report (the Report) represents the final report for the Carbon Study and follows several interim documents, which were provided for review to DPaC and the independent panel of experts comprising the Forest Carbon Study Steering Committee (Steering Committee). Members of the Steering Committee included Distinguished Professor Jim Reid, Associate Professor Cris Brack and Professor Brendan Mackey.

trading programs, often known as ‘cap and trade’ schemes, the tradeable units are more properly referred to, or thought of, as ‘permits’, or ‘allowances’, that provide emitters with a ‘right’ or ‘permit’ to emit 1 tonne CO2e to the atmosphere. In many programs, there is also a process that recognises either activities undertaken ‘elsewhere’; that is, in other geographies, or within sectors, not covered by the program. Such activities can ‘offset’ emissions and the unit of activity creates a ‘credit’.

Page 31 of 365

1.1 SCOPE & OBJECTIVES

As per the tender specification, the scope and objectives of this Carbon Study have been to:

• Estimate current carbon stocks across all of Tasmania’s forest estate, taking into account forest age-class structures, disturbance, land use histories and management regimes. • Provide an assessment of the stocks and emissions in decadal time periods for the previous two decades (1990-2010). • Provide an assessment of likely future stocks and emissions (in decadal periods to 2050) under a range of scenarios determined in consultation with the Steering Committee. • Develop a clear accounting framework for carbon stocks and flows in Tasmania’s forests. • Identify areas of significant uncertainty with respect to estimates of carbon stocks and assess the magnitude of uncertainty and error in estimates. • Provide an analysis of opportunities to monetise avoided emissions, emission reductions and increased carbon sequestration, with specific reference to the assessment of current and baseline carbon stocks established as part of the Carbon Study.

The tender specifications list a number of items as being specifically out of the scope of the Carbon Study, including:

• The Carbon Study is confined to analyses of carbon stocks and flows, with other economic, social, employment and environmental values of the forest estate being excluded from the scope. • The Carbon Study is confined to analyses of carbon emissions from Land Use, Land Use Change and Forestry Sources, with emissions from other sectors, such as the energy sector, being out of scope. • Management costs of different management regimes analysed as part of the Carbon Study are out of scope. • While the Carbon Study is intended to provide scientific analyses of the carbon implications of different forest management options, it will not directly provide advice on policy matters.

Page 32 of 365

1.2 SCENARIOS

To assess the effects of altered forest use and land use patterns on Tasmania’s forest carbon stocks, a range of forward looking native forest, plantation forest and land use scenarios were developed in consultation with the Steering Committee. Additionally, a baseline forest management scenario (Study Baseline) was agreed in order that the abatement, or emissions, forecast to arise from any change to management practices could be compared against those associated with existing, contemporary forest management practices. These are summarised below, with further rationale and background provided at Sections 6 and 7.

The Study Baseline

For any assessment of changes in carbon stocks arising from changed forest, or land use, patterns, it is critical that a ‘Baseline Scenario’ is defined that describes contemporary patterns of forest and land use, as well as the emissions and abatement associated with that pattern. By understanding the carbon stocks associated with the Baseline Scenario, it is then possible to estimate the net abatement, or emissions, that arise from altered forest and land use patterns (i.e. the change in carbon stocks relative to stocks under the baseline practice). With respect to recognising carbon stocks under various formalised emissions reduction programs, defining the Baseline Scenario is critical to establishing the additionality of any abatement activity and for calculating the carbon stocks that may be converted to carbon credits (Section 4).

For the purposes of the Carbon Study, the Baseline Scenario that has been developed in consultation with the Steering Committee, referred to here as the Study Baseline, is defined as follows (see also Section 7). • Growth rates for native forests and plantations - based on plot data from Forestry Tasmania and field measurements of different forest classes for native forests and a relationship between FPI and site index for plantations. • Future fire frequency is assumed to be equivalent to the regime currently experienced for specific forest types and is set by the annual probability of a Stand being burnt. Fire intensity varies with forest type and is determined by the proportion of Big Trees killed ( ). The age of a Big Tree was assumed to be 40 years for Low Eucalypt Forest and 80 years for all other forest types to simulate the observed lower sensitivity of the former to fire (e.g. Florence, 2004). The various fire regimes were set as follows: o Tall Eucalypt Forest: annual probability 0.3-0.7% with peak risk at age 50 years (equivalent to a frequency of 136-250 years) and an average 50% of Big Trees killed; o Low Eucalypt Forest: annual probability 3.0-3.8% with peak risk at age 25 years (equivalent to a frequency of 26-33 years) and an average 20% of Big Trees killed;

Page 33 of 365

o Rainforest: annual probability 0.05% (equivalent to a frequency of 2000 years) with an average 75% of Big Trees killed; o Other Non-eucalypt Forest: annual probability 4.0% (equivalent to a frequency of 25 years) with 20% of Big Trees killed; and o Plantations: annual probability: 1.0% (equivalent to a frequency of 100 years) with 100% of Big Trees killed. • Tree mortality from fire is assumed to decrease with tree age and increase with fire intensity (defined by the proportion of Big Trees killed) with 100% of trees aged 0 years killed. Thus, if the age of trees is greater than the assumed age of Big Trees (e.g. 80 years for Tall Eucalypt forest), then the average proportion of trees killed will be less than 50%. Similarly, if the stand age is less than 80 years, then the average proportion killed will be greater than 50%. • Proportions of debris combusted as a result of wildfire and regeneration burns were based on studies by Bormann (2008), Hurteau (2011), Lindenmayer (Pers. Comm.) and Norris (2010). Proportions of soil carbon combusted were based on work by Dyrness et al. (1989), Ellis and Graley (1983), Hatch (1960), Pennington et al. (2001) and Tomkins et al. (1991). Losses from the various pools were related to fire intensity (defined by the proportion of Big Trees killed or ) as follows:

o % Dead trees burnt = 1 x ; o % Dead stems = 2 x ; o % Branch debris = 2 x ; o % Bark debris = 3 x ; o % Leaf debris = 5 x ; o % Soil humus = 0.2 x ; and o % Root debris = 0. • It is assumed that, of the proportion burnt, 95% is volatilised and 5% is converted to inert soil C (i.e. charcoal) for all Debris except Stags (standing dead trees) for which 80% of carbon is assumed to be volatilised. • Stem mortality rates were based on published changes in stocking of Tall Eucalypt Forests over time, stems are assumed to remain standing as Stags for varying times depending on tree diameter and cause for death. Maximum half lives for Stags of 200 cm DBH are based on data from Forestry Tasmania (Grove, 2002, 2009) and assumed to be:

o 10 years for stags produced through natural mortality; and o 30 years for stags produced from fire. • Debris wood density, production and decomposition rates – based on data from Forestry Tasmania (Grove, 2009) and default figures from DCCEE.

Page 34 of 365

• Soil carbon – based on the RothC model from FullCAM used for soil carbon cycling which has not yet been validated for Australian native forests. • Effect of temperature and rainfall on carbon in debris and soil – parameters for FullCAM climate modifier based on one study by CSIRO in plantations only (Paul and Polglase, 2004). • There is assumed to be no conversion of native forests to grassland or plantations and no further expansion of plantations, although it is also assumed that Tall Eucalypt Forest eventually succeeds to Rainforest in the absence of fire and that Rainforest in mixed eucalypt forest is replaced by eucalypt regeneration after fire or harvesting. • Climatic conditions are assumed to be the same as those experienced over the past 40 years (i.e. based on the average data for 1968-2004 from the Department of Climate Change & Energy Efficiency, DCCEE). • Rates and intensities of timber harvesting for native forests are based on historic rotation lengths and harvest regimes for different forest types. These were set to give average rotation length of 80 years for regrowth forest and were assumed to be:

o Tall Eucalypt Forest, unthinned 8-15 years: 50% removal (annual 7% chance of harvest); o Tall Eucalypt Forest, unthinned <150 years: clearfell >70 years (annual 5% chance of harvest); o Tall Eucalypt Forest, thinned <150 years: clearfell >60 years (annual 5% chance of harvest); o Tall Eucalypt Forest, > 150 years: clearfell (annual 5% chance of harvest); o Low Eucalypt Forest, < 150 years: 50% removal > 40 years (annual 5% chance of harvest); o Low Eucalypt Forest, > 150 years: 50% removal (annual 5% chance of harvest); o RFT: clearfell (annual 0.5% chance of harvest); and o Other Non-Forest: clearfell >60 years (annual 5% chance of harvest). • Following harvest (excluding thinning), it is assumed that stands are regenerated using high intensity burns (after clearfelling), top disposal burns (lower intensity burns used after partial harvesting), or mechanical disturbance. Based on information from Forestry Tasmania and private plantations the proportions of harvested stands burnt were assumed to vary by forest type as follows:

o Tall Eucalypt Forest: 80%; o Low Eucalypt Forest: 70%; o Rainforest: 75%; o Other Non-Forest: 75%; and o Plantations: 25%. Page 35 of 365

• The intensities of regeneration and top disposal burns were defined as the percentage of Big Trees killed ( ) which determined the proportions of debris burnt (using the same factors as for wildfire). It was assumed that fewer trees would be killed in top disposal burns (predominantly in Low Eucalypt Forests) compared with high intensity burns (predominantly in other forest types). Thus the age of Big Trees was set at 40 years for the former and 80 years for the latter. The effect of post-harvest burns on debris and soil carbon was assumed to vary for different harvest types as follows: o Post clearfell (high intensity regeneration burn): = 80%, and o Post partial harvest (top disposal burn): = 10%. • Rates and intensities of harvest of plantations are based on current age class distributions (ABARES 2012a) and harvest regimes (Forestry Tasmania, 2010, Don Aurick, Timberlands Pacific, Pers. Comm.) and sawlog quotas (Public hardwood Plantations only). Harvest regimes for plantations were assumed to be: o Hardwood pulpwood: clearfelled at 12-16 years; o Harwood sawlog: thinned at 8-12 years and clearfelled at 23-32 years; and o Softwood: thinned at 12-16 years and clearfelled at 28-33 years. It was assumed that 75% of publicly owned hardwood plantations and 5% of privately owned plantations are harvested on sawlog regimes (David Mannes, Forestry Tasmania, Pers. Comm.). • The amount of forest harvested is set by the available resource (i.e. volume of wood in stands of harvestable age), the probability those stands being harvested and the sawlog quota required to be filled (where applicable). Sawlog quotas are based on:

o Legislated minimum amount of sawlogs required to be produced from publicly owned native forest and hardwood plantations (300,000 m3/y; Forestry Tasmania 2012); and o Average sawlog production from privately owned native forest from 2006-2010 (55,000 m3/y; ABARES 2012a). • Volumes of sawlogs, peeler logs and pulplogs harvested are based on the total volume harvested and average recovery rates of different log types from different forest types from Forestry Tasmania, (2006, 2008, and 2011) and ABARES (2012a). Sawlog recovery rates varied with tree size as well as forest type and average:

o 13% for public native forests; o 5% for private native forests; o 38% for public hardwood plantations; o 1.5% for private hardwood plantations; and o 47% for softwood plantations. • Exported wood products are assumed to be converted to CO2. Proportions of wood products exported are based on data from ABARES (2012a) and are assumed to be:

Page 36 of 365

o Sawlogs: 19% exported; o Peeler logs: 19% exported; o Pulplogs and wood chips: 72% exported; o Poles: 13% exported; o Panels: 4% exported; o Veneer: 76% exported; and o Paper: 26% exported. • Recovery rates for wood products processed in Australia are based on data from Tucker et al. (2009). These are assumed to be:

o Domestic sawlogs: 40% sawn timber, 35% woodchips, 25% biofuel; o Domestic peeler logs: 45% veneer, 10% posts; 25% woodchip, 20% biofuel; and o Domestic pulplogs and wood chips: 69% paper, 1% panels, 30% biofuel. • Lifespan of wood products are based on default half lives from UNFCCC (2011) and are assumed to be:

o Sawn timber: 35 years; o Veneer, Plywood and Panels: 25 years; and o Paper: 2 years.

Study Scenarios:

Altering management regimes in native forests A number of the Study Scenarios contemplate a shift in the forest management regime that is applied to areas of native forest, these being:

N1. Reserve 572,000 hectares (ha) of public forest; reduce the annual sawlog cut from public native forests and Hardwood Plantations to 155,000 cubic metres per year (m3/y) and the annual peeler log cut from public native forests to 265,000 m3/y. N2. Phase out harvesting in old growth forest by 2020 on both public and private land, but increase harvesting intensity in regrowth forests so as to maintain annual sawlog production from public native forests and Hardwood Plantations at, or near to, 300,000 m3/y. N2A. Phase out harvesting of old growth forests by 2020 on both public and private land, and vary harvest intensity in regrowth forests so as to maintain annual sawlog production from public native forests and Hardwood Plantations at 155,000 m3/y. N3. Vary rotation length in regrowth forests from 60-120 years, with reference to the current rotation lengths of approximately 80 years (as indicated in the literature).

Page 37 of 365

N4. Immediate cessation of the harvest of old growth forests on both public and private land with no increase in harvesting of regrowth forests. N5. Immediate cessation of all native forest harvesting on both public and private land.

Altering management regimes in plantations Some Study Scenarios contemplate a shift in the forest management regime that is applied to areas of existing plantation forest; these consist of:

P1. Increase the rotation length of all existing plantations by 10%. P2. Convert all existing pulpwood plantations to sawlog plantations by increasing the rotation length to 25 years. P3. Increase the growth rates of existing plantations by 25% through improved management.

Altering land use A set of five Study Scenarios consider the effect that changing land use patterns may have on carbon stocks, these being:

A1. Conversion of private native forest for plantation, or non-forest uses, continues at the recent rates indicated by Forest Practices Authority and as permitted under the Permanent Native Forest Estate Policy from 2009 up until 2015, after which it ceases. A2. Existing areas of native forest classified as “unstocked” are regenerated back to native forest (non-harvested). Although considered unstocked, these areas contain existing non-eucalypt vegetation. A3. The plantation forest estate expands into cleared agricultural lands at the average annual rate of expansion that occurred from 2000-2010. A4. The plantation forest estate expands into cleared agricultural lands at double the historic average (2000-2010) to a maximum point where 10% of existing cleared agricultural land is converted to plantations by 2050. A5. Twenty-five percent (25%) of existing pulpwood plantations are not replanted after their first harvest.

Baseline Sensitivity Scenarios

Given how critical assumptions around the Study Baseline are to estimating net abatement, or emissions amounts, an additional set of Baseline Sensitivity Scenarios were developed to test the influence of fire events and the possible influences of climate change.

Page 38 of 365

Four alternative fire scenarios were tested:

The influence of three alternative future climate projections, selected from the Second Report Emissions Scenarios (SRES) of the Intergovernmental Panel on Climate Change (IPCC, 2000), were tested: C1. Based on SRES A2, using climate projections from the CSIRO Mk 3.0 climate model which predicts the Tasmanian climate in 2050 will be similar to that of 2010. C2. Based on SRES A2, using climate projections from the CSIRO Mk 3.5 climate model, which predicts the Tasmanian climate in 2050, will be hotter and drier than in 2010. C3. Based on SRES A1b, using climate projections from the CSIRO Mk 3.5 climate model which predicts the Tasmanian climate in 2050 will be warmer and drier than that in 2010.

Carbon Carrying Capacity

Carbon Carrying Capacity is defined as the average total carbon stocks of a forest subject to prevailing environmental disturbance regimes but excluding human disturbance. Carbon carrying capacity includes effects of fire and other natural disturbance regimes, but excludes the effects of timber harvesting, or land conversion. Carbon Carrying Capacity was estimated for Tasmania’s native forests by running the Forest Carbon Modelling Framework over a period of 240 years to 2250 with all timber harvesting excluded but with prevailing climate and fire regimes (Carbon Carrying Capacity Baseline). This process was repeated with alternative fire regimes (20% increase and 20% decrease in fire frequency) to test the influence of fire on Carbon Carrying Capacity (Carbon Carrying Capacity Scenarios).

L1. Carbon Carrying Capacity Baseline using the same fire frequencies for different native forest types as for the Study Baseline. L3. Carbon Carrying Capacity Scenario in which fire frequencies were decreased by 20%. L4. Carbon Carrying Capacity Scenario in which fire frequencies were increased by 20%.

Page 39 of 365

1.3 KEY ACTIVITIES & OUTCOMES

One of the major areas of activity and outcomes of the Carbon Study has been the development of a unique, built-for-purpose carbon accounting framework, referred to as the Forest Carbon Modelling Framework (FCMF) that models forest carbon stocks, and changes to these stocks, across all forest types within Tasmania. Given the scale of the modelling task, the FCMF has been designed to allow for the rapid generation of large numbers of simulations, representing a significant efficiency improvement over alternate, publicly available approaches.

The FCMF utilises the base modelling framework and approach included within the Full Carbon Accounting Model, or ‘FullCAM’ (Richards, 2001; Richards et al., 2005), but extends FullCAM functionality and application by linking modelling processes to detailed input information for all forest types and carbon pools, explicitly modelling multi-cohort (i.e. mixed aged) forests and referencing forest growth, fire and harvesting regime data represented spatially across Tasmania at a 1 km2 level. The FCMF also links forest growth with amounts of harvested wood products produced, allowing full modelling of the impact of changes in timber harvest volumes on carbon stocks, this being a key requirement of the Carbon Study.

Critically, the FCMF incorporates a significant body of actual growth data collected from native forest and plantations within Tasmania, access to most of which was negotiated under the Carbon Study. Key data providers have included Forestry Tasmania, DCCEE, the Wilderness Society, the University of Tasmania and Ian Riley. These data inputs have been incredibly valuable to the outcomes of the Carbon Study and the authors express their thanks and appreciation to all contributors. In addition to these data sources, some field measurement was undertaken as part of the Carbon Study and this has generated valuable information addressing key knowledge gaps around the relative change in carbon stocks associated with the transition of mature Tall Eucalypt Forest to Rainforest.

On balance, the authors consider that the Carbon Study represents one of the most detailed, field-data-informed assessments of state-wide carbon stocks yet undertaken within Australia. It is hoped that the FCMF will continue to provide a valuable resource into the future that will allow forest managers, researchers and policy makers to continue to develop the science around carbon accounting in complex forest systems and to evaluate a wide range of forest and land management options that may emerge into the future.

Other key outcomes from the Carbon Study include:

• Detailed estimates of historic, current and future (to 2050) carbon stocks, across all forests within Tasmania, including consideration of the net abatement, or emissions, arising from changes in forest and land management practices. • An assessment of the effects of climate change and fire frequency on future carbon stocks.

Page 40 of 365

• Uncertainty and sensitivity modelling around carbon estimates, as well as identification and documentation of key knowledge gaps to be addressed through future research. • A detailed assessment of current, and near-term, opportunities for monetising carbon sequestration arising from the carbon abatement options contemplated by the Study Scenarios, including a detailed review of relevant programs, key eligibility criteria, Approved Methodologies and features of both mandatory and voluntary carbon markets. • Detailed and consolidated geographical information system (GIS) files and hardcopy maps that include whole-of-state spatial coverage of various forest types, age classes, management regimes, climate scenarios and carbon stock estimates.

Page 41 of 365

2 CARBON MODELLING OVERVIEW

This section provides a brief overview of the carbon modelling approach applied under the Carbon Study, including a description of the Forest Carbon Modelling Framework (FCMF). A more detailed discussion, including of background information sources, previous studies and basis for key assumptions is provided at Sections 6 & 7.

2.1 VEGETATION CLASSIFICATION

Tasmania is covered by 3.4 million ha of native forests and plantations (Table 1). Of this area, 2.3 million ha are classed as public forest and 1.1 million ha are classed as private forest. The most abundant forest type is Low Eucalypt Forest (also known as dry sclerophyll forest) which covers 1.5 million ha. Tall eucalypt forest (otherwise known as wet sclerophyll forest) covers 0.9 million ha, Rainforest covers 0.6 million ha and Other Native Forest covers just 0.1 million ha. Plantations, cover a total 0.3 million ha with 0.1 million ha on public land and 0.2 million ha on private land. Around 3.4 million ha are classed as non-forest. However, this includes a large area of native vegetation, some of which is > 2 m tall and which therefore should be classified as woody vegetation. However, no figures exist for the area of this vegetation class and there is some uncertainty over its extent (e.g. Montreal Process Implementation Group for Australia, 2008).

The Carbon Study included all vegetation across Tasmania classified as woody vegetation, with height >2m and >20% canopy cover (this being consistent with Kyoto definitions). To do this, 10 vegetation groups were developed, based on information from Forestry Tasmania (Forest Classes and Forest Groups) and the TasVeg spatial layer developed by the Department of Primary Industries, Parks, Water and Environment (DPIPWE, TVMP, 2004). These comprised seven woody vegetation classes, two non-woody vegetation classes and one non-vegetation class:

• ETF: Tall Eucalypt Forest (eucalypt dominated forests > 34 m mature height) • ELF: Low Eucalypt Forest (eucalypt dominated forests < 34 m mature height) • RF: Rainforest • ONF: Other Native Forest (> ~10m height) • PHW: Hardwood Plantations • PSW: Softwood Plantations • TNV: Tall Native Vegetation (>2m height ) classed as non-forest • LNV: Low Native Vegetation (<2m height ) classed as non-forest • GRS: Cleared agricultural grassland. • NON: Non-vegetated areas.

Page 42 of 365

A geographic information system (GIS) was used to map vegetation with reference to spatial layers from Forestry Tasmania for 1996, 2001 and 2010 (Forest Classes and Forest Groups layer) and from the Department of Primary Industry, Parks, Water and Environment (DEPIPWE) for 2010 (TasVeg layer). The Forest Class layers were used as the primary sources because these are related to actual field data on tree and debris volumes and include information on stand age, structure, growth and history. However, complete coverage for all of Tasmania was available for 1996 only, with public forest only available for subsequent years. For this reason, changes in forest types over time and new plantings were derived from the Forest Group layers which cover all of Tasmania for 2001 and 2010. Both the Forest Group and Forest Class layers classify most vegetation < ~10 m as non-forest. Consequently, the TasVeg layer was used to identify native vegetation > 2 m and < 2m as well as grassland that could potentially be converted to forest. The Department of Primary Industries, Parks, Water and Environment was consulted for advice on which specific vegetation types were most likely to fit the criteria for woody vegetation (> 2 m and > 20% cover, A. Kitchener, DEPIPWE, Personal Communication) The vegetation maps that were derived for 1996, 2001 and 2010 were then combined with spatial data layers showing forest tenure (public or private), potential for harvest, climate and soil type. These three layers were used as the basis for tracking carbon stocks and emissions for 1990, 2000 and 2010. No vegetation data was available for periods prior to 1996. However, the 1996 layer was considered to provide a reasonable indication of vegetation distribution in 1990 as it is compiled from aerial photography taken prior to 1996, much of this being captured during the late 1980’s.

For the purpose of modelling carbon stocks and flows spatially, the vegetation maps developed under the Carbon Study were overlayed with a 1x1 kilometre (km) grid. The vegetation, tenure, potential for harvest, soil and climate at the centre of each grid was assumed to represent the average for each grid cell. This gridded data set was then used to provide the key spatial inputs to the FCMF.

While a 1km x 1km grid is fairly coarse for spatial analyses, with a total of almost 68,000 grid points across Tasmania and nearby islands, it was felt that this approach provided for a reasonable breakdown of the distribution and proportion of different forest types across the State. Cross-checking the total area for each forest type estimated from the Carbon Study vegetation layer against the areas identified from the original, finer-resolution source data layers indicated an error of <1% for the test sample.

Page 43 of 365

Table 1: Land tenure by forest type (including non-forested areas) for Tasmania as at June 2011. Areas are in 1000 hectares. Forestry Tasmania (2008, 2011a).

Forest Type State forest Other public forest Private 1 Total area Multiple-use forest Total Other Total land Formal Cons. Public public Productio Informal Other 5 6 7 Total reserves reserves reserves 8 n areas2 reserves3 forest4 land Native Forest Tall native eucalypt forest9 342 87 53 482 50 532 178 5 12 195 161 887 Short native eucalypt forest10 128 89 106 323 71 394 404 12 45 461 647 1,501 Rainforest11 68 62 7 136 67 203 330 0 5 335 17 555 Other native forest12 7 16 34 56 12 68 49 1 4 54 31 153 Total 544 254 199 997 199 1,196 961 18 66 1,045 856 3,096 Plantation Softwood plantation 53 0 0 53 0 53 0 1 0 1 23 77 Hardwood plantation 55 0 0 55 0 55 0 0 1 1 181 237 Total 108 1 0 108 0 108 0 1 1 2 204 314 Total Forest 652 254 199 1,105 199 1,304 961 19 67 1,047 1,060 3,410 Non-forest13 0 45 120 165 23 188 1,375 23 155 1,553 1,660 3,400 Total 652 299 319 1,270 222 1,492 2,336 42 222 2,600 2,720 6,810

1 Private property forest areas are as advised by Private Forests Tasmania, as of December 2009. 8 Includes land owned by HEC & Commonwealth; municipal reserves; private & municipal lands managed 2 From Forestry Tasmania (2008). Includes special timbers zone by Wellington Park Management Trust. 3 Areas managed as a Protection Zone under the Management Decision Classification System. 9 Eucalypt forest with current, or potential, height of 34 m or more. 4 Areas not part of the wood resource due to economic, topographic or accessibility considerations. 10 Eucalypt forest with current, or potential, height of less than 34 m

5 11 Page Areas reserved from harvest which require action by the Tasmanian Parliament for revocation. Cool temperate Rainforest with no significant eucalypt or acacia species. 6 Nature Conservation Act 2002 (http://www.austlii.edu.au/au/legis/tas/consol_act/nca2002237/) 12 Includes huon pine, King Billy pine, acacia and melaleuca species. 44 7 Crown Lands Act 1976 (http://www.austlii.edu.au/au/legis/tas/consol_act/cla1976134/) 13 Includes grassland, scrub, moorland, farmland, rock and lakes. of of 365

2.2 MODEL DESCRIPTION

The FCMF uses the base models within FullCAM (Richards, 2001; Richards et al. 2005), but extends its functionality and tailors the modelling to Tasmanian forest types, fire regimes and management. These modifications were considered necessary for meeting the Carbon Study’s aim to provide a comprehensive assessment of Tasmanian forest carbon stocks, and critically, to handle the large volume of data inputs, data outputs and simulations required to address key Carbon Study objectives. The capabilities and functionality of the FCMF include:

• use of growth models based on an extensive set of field measurement data collected from Tasmanian native forests and plantations; • inclusion of vegetation-specific fire and harvesting regimes; • explicit modelling of multi-cohort (mixed age) stands resulting from fire or partial harvesting events, • probabilistic modelling of fire and harvest based on user-defined probability distributions; • capacity to set quotas and limits on sawlog volumes, harvesting, and conversion; • efficient calculation of initial values for different carbon pools; • rapid input and export of extensive datasets; • a user interface that allows key parameters and uncertainties to be readily modified; • capacity to develop and batch multiple scenarios; and • capacity to run uncertainty and sensitivity analyses via Monte-Carlo type simulation. A detailed description of the model, its parameters and assumptions is provided in Sections 6 with a review of underlying studies used to develop and a parameterise the model in Sections 7 and 8 of the report. The following provides a brief overview of the FCMF structure and some of the main assumptions.

2.3 FOREST CARBON POOLS

Like FullCAM, the FCMF divides forest carbon into several pools, including a live vegetation pool, a debris pool, and a soil pool. Carbon stocks are transferred between these pools as shown in Figure 1. The key driver of carbon stocks and flows is the rate of growth (i.e. carbon sequestration) of the vegetation. Growth rates used in the FCMF for different vegetation types were based on data provided by Forestry Tasmania and the Tasmanian Wilderness Society, as well as measurements collected by CO2 Australia during the Carbon Study. Altogether, growth models for 88 native forest classes and two plantation classes were developed.

Page 45 of 365

Figure 1: Process overview of the Forest Carbon Modelling Framework (FCMF), showing the interaction between various forest carbon, soil carbon and forest product carbon modelling elements.

The carbon sequestered during tree growth is allocated to different tree components according to the general allocation parameters used in FullCAM for the seven woody vegetation classes modelled. These components contribute to debris pools when they die. The different components modelled for live vegetation and debris include:

• stemwood • leaves • bark • coarse roots • branches • fine roots

Page 46 of 365

Unlike FullCAM, the FCMF has the capacity to model stem mortality using a self thinning function. Dead stems move to a Stag (standing dead tree) pool where they undergo decomposition until they eventually fall, becoming stemwood debris. Decomposition of debris results in the loss of some carbon to the atmosphere with the remainder being transferred to the soil. As with FullCAM, the FCMF uses the Roth-C model (version 6.3, Jenkinson, 1990; Coleman & Jenkinson, 2008) to calculate changes in the various soil carbon pools.

Under constant growing conditions, the three pools generally tend to an equilibrium or steady state, in which inputs are balanced by outputs. For FullCAM, it is normal practice to estimate equilibrium values for each carbon pool by running the model with constant inputs and outputs for an extended period of time (e.g. 500-1,000 years). The FCMF does this more efficiently by calculating the equilibrium amounts for each carbon pools directly from the input and output rates for each pool at the start of each simulation run.

2.4 PRODUCT CARBON POOLS

Wood products produced as a result of timber harvesting were modelled within the FCMF based on average product yields from ABARES (2012a) and Forestry Tasmania (2005a, 2006, 2007, 2011a) and conversion rates from Tucker et al. (2009). When trees are harvested, they are broken down into different log types which are then further broken down by conversion processes into finished products and by-products. The FCMF tracks these processes and allows relevant information for each conversion process to be used. Log types modelled included:

• peeler logs • sawlogs • pulplogs

These may be either exported (in which case their contained carbon is assumed to be emitted) or sent to further processing. These products are then accumulated in product pools that gradually decay over time. Assumed decay rates were based on default figures from the UNFCCC (2011): Products from final processing and assumed decay rates included:

• sawn timber: 35 years • panels: 25 years • paper: 2 years

Woodchips produced from pulplogs, or as by-products from wood processing, were apportioned as export chips (whose carbon was assumed to be emitted) and chips used in domestic paper or wood panel manufacture (and thus contributing to the paper or panel carbon pools).

Page 47 of 365

2.5 FIRE

Two key event types are included within the FCMF: fire and harvesting. Fire influences forest carbon stocks in a number of ways:

• it volatilises some carbon from vegetation, debris and soil; • it moves carbon from live vegetation to debris; • it produces inert carbon (i.e. charcoal) which is assumed to have an infinite life span, • it allows the forest to regenerate, and • it promotes the development of multi-cohort (mixed age) structures when only a proportion of trees are killed. Frequency and intensity of fire are crucial in determining its effect on vegetation growth and carbon stocks. The average frequency of fire events for different forest types was based on studies (reviewed in detail at Section 7) of fire regimes in Tasmanian forests and other similar forest types (e.g. Bowman and Jackson 1981, Gould and Cheney 2007, Jackson 1968, McCarthy et al. 1999; Marsden-Smedley 1998, 2007; Mount 1979; Turner et al. 2009 and Wood et al. 2010).

The FCMF allows fire frequency, intensity and associated tree mortality and changes in carbon pools to be specified for each vegetation type. The frequency of fire events is based on user- defined probability distributions. Under the Carbon Study, the probability distributions used for each of the nine vegetation classes were set as follows:

• Tall Eucalypt Forest: Fire frequency assumed to vary with stand age, starting at 0.5% probability of a fire event occurring at forest age zero, rising to a peak probability of 0.83% per year at age 50, then decreasing to 0.5% per year beyond age 600. On average, 50% (+/- 25%) of Big Trees are assumed to be killed by a fire event. • Low Eucalypt Forest: Fire frequency assumed to vary with stand age, starting at 2% probability of a fire event occurring at forest age zero, rising to a peak probability of 2.7% per year at age 25, and decreasing to 2% per year beyond age 300. On average, 20% (+/- 10%) of Big Trees are assumed to be killed by a fire event. • Rainforest: Fire frequency assumed to be constant over time at 0.1% probability of occurring in any given year. On average, 75% (+/- 19%) of Big Trees are assumed to be killed by a fire event. • Hardwood Plantations & Softwood Plantations: Fire frequency assumed to be constant over time at 0.1% probability of occurring in any given year. All trees are assumed to be killed. • Other Native Forest, Tall Native Vegetation, Low Native Vegetation: Fire frequency assumed to be constant over time at 5% per year. 20% (+/- 10%) of Big Trees are assumed to be killed. • Grassland: Fire frequency assumed to be constant over time at 5% per year.

Page 48 of 365

Because any given fire event may only kill a proportion of trees in a forest area, the FCMF was designed to allow the development of mixed-aged stands comprised of different cohorts (groups of trees of the same age and species type). Multi-cohort stands typically comprise a significant proportion of both tall and Low Eucalypt Forest across Tasmania (Turner et al., 2009). Each fire event that removes trees results in the regeneration of a new cohort. To determine the probability of fire in multi-aged stands, the age of the largest cohort (by area) was assumed to represent the stand age.

Fire intensity is modelled in the FCMF using the proportion of Big Trees killed as a parameter. For the purpose of the model Big Trees were defined as trees aged 80 years in all forest types except Low Eucalypt Forest. For the latter, the age of Big Trees was defined as 40 years to allow for the lower fire sensitivity of species comprising this forest type. It was assumed that younger (i.e. smaller) trees are more susceptible to fire than older larger trees and that the proportion of trees killed by a fire event decreases with the age of the cohort. A simple relationship between fire intensity and tree mortality was developed to estimate this based on the proportion and age of Big Trees killed.

Both the frequency and intensity of fire are based on probability distributions, which allow the user to set the amount of variation around the mean fire events for each species. This allows the random nature of fire to be simulated and also allows development of a large number of different stand types and age class distributions over time.

Like FullCAM, the FCMF models the effects of fire on carbon from live vegetation and debris. However, unlike FullCAM, the FCMF also allows the loss of soil carbon as a result of fire to be modelled. Few published studies on carbon losses from soil or debris associated with wildfires exist. However, an indication of potential losses was obtained from work by Bormann et al. (2008), Pennington et al. (2001), Rab (1996) and Slijepcevic (2001) on carbon losses associated with post-harvest regeneration burns. The proportion lost from debris components was assumed to be greatest for fine material (e.g. bark and leaves) and least for coarse material (e.g. stemwood). It was assumed the amount of carbon lost from humus was equivalent to 20% of proportion of Big Trees killed.

2.6 HARVESTING

Timber harvesting was modelled in a similar manner to fire using a probabilistic approach based on vegetation type and stand age. The variables included in each harvesting event included:

• age at which a stand could be harvested; • relative frequency of harvesting of eligible stands; • proportion (%) of trees felled; • proportion (%) of trees removed; and • likelihood of harvest slash on coupes being burnt to promote regeneration.

Page 49 of 365

These parameters determine the effect of harvesting on stand structure, the amount of wood removed and the average rotation length. Parameters used for the different forest types are explained in detail at Section 7. They can be summarised for the different vegetation types as:

• Tall Eucalypt Forest: Clearfelled (~90% of stems harvested with 85% of harvested stemwood volume removed) after age 70 years at an average rate of 5% per year with 75% of coupes burnt. • Low Eucalypt Forest: Partially harvested (~50% of stems harvested with 85% of harvested stemwood volume removed) after age 50 years at an average rate of 5% per year with 25% of coupes burnt. • Rainforest: Clearfelled (~90% of stems harvested with 85% of harvested stemwood volume removed) at an average rate of 1% of eligible stands harvested with 75% of coupes burnt. • Other Non-Forest: Clearfelled (~90% of stems harvested with 80% of harvested stemwood volume removed) after age 50 years at average rate of 5% of eligible stand harvested with 75% of coupes burnt. • Hardwood Sawlog Plantations: Thinned (50% of stems harvested with 85% of harvested stemwood volume removed) at age 8-12 years; clearfelled (100% of stems harvested with 90% of harvested stemwood volume removed) after age 25 years (25% of eligible stands cut each year) with 25% of coupes burnt. • Hardwood Pulpwood Plantations: Clearfelled (100% of stems harvested with 90% of harvested stemwood volume removed) after age 13 years (35% of eligible stands cut each year) with 25% of coupes burnt. • Softwood Plantations: Thinned (50% of stems harvested with 85% of harvested stemwood volume removed) at age 12-16 years; clearfelled (100% of stems harvested with 90% of harvested stemwood volume removed) after age 28 years (10% of eligible stand cut each year) with 25% of coupes burnt. For native forests it was assumed that between 80% and 85% of the wood volume in harvested stems was removed from the site with 15-20% of stemwood and all branch, bark, leaf and root biomass retained on site or burnt.

The yield of sawlogs from harvested stands was based on average recovery rates for difference forest types and was estimated using a model which related sawlog percentage to average DBH (Farm Forestry Toolbox, A. Goodwin Pers. Comm.). To account for variation in stem quality, a variable scale factor was applied to different tenures (private, public) and forest types. This allows the proportion of sawlog produced to be varied according to the type of forest and its ownership, based on relevant available information.

Page 50 of 365

2.7 NATURAL VARIATION AND UNCERTAINTY IN PARAMETERS

'Natural' variation in harvesting and fire events and uncertainty in parameter values were taken into account using a probabilistic approach. This was used to provide indications of variation in model outputs. This variation was expressed by presenting all results in terms of ranges representing the effect of underlying uncertainty in base parameters.

The likelihood of fire and harvesting were modelled using probability distributions for each forest type. These distributions were used to assess the chance of a particular stand being harvested, burnt (or both) each year. This was done by comparing the chance of an event taking place that year (e.g. a 0.3% chance of a stand being burnt) with a set of random numbers ranging from 0 to 1. If the random number was less than the assumed probability, then the event was assumed to take place. More detail is provided on the actual distributions used in Section 6 of this report. To enable valid comparison between the Study Baseline and Study Scenarios, the same set of random numbers was used for each. This ensured that, if an event happened in a particular year in a particular Stand in one Scenario (e.g. the Study Baseline), then it would happen in the same Stand in the same year in another Scenario (e.g. a Study Scenario) unless the actual probability or associated parameters were changed (e.g. in Study Scenarios where the frequency of fire or harvesting was increased).

To take into account uncertainty in estimates for parameters, each of the 200 or so parameters used in the model was given a mean (assumed most likely value or the default used by DCCEE) and a coefficient of variation. The variance was based on the estimated confidence of the mean estimate or reported variation in the particular parameter from other studies. A set of primary results for the Study Baseline and Study Scenarios was first obtained by fixing each parameter at its assumed mean. Each parameter was then allowed to vary randomly within the range dictated by its coefficient of variation (within an assumed normal distribution) over a series of iterations for each Study Scenario. The results from these iterations were then used to generate means, standard deviations and coefficients of variation for each Study Scenario. The coefficients of variation were then used to estimate variation in the primary outputs. The outputs were presented as ranges with the maximum and minimum values representing one standard deviation from the primary estimate. By this process, the effect of uncertainty in the base parameters on the final results was able to be taken into account.

Page 51 of 365

3 RESULTS

3.1 OVERVIEW

The various scenarios covered in this section are grouped as follows: • Study Baseline; • varying harvesting rates in native forests; • varying growth and management of Softwood Plantations; • deforestation, reforestation and afforestation; • sensitivity to fire frequency; • sensitivity to climate; and • Carbon Carrying Capacity. The results from the analyses are broken down by forest type, tenure, management and potential for Kyoto eligibility. Forest types are divided by Forest Group. Codes used in figures and tables are: • PSW: Softwood Plantations • PHW: Hardwood Plantations • RFT: Rainforest • ETF: Tall Eucalypt Forest (> 34 m mature height) • ELF: Low Eucalypt Forest (< 34 m mature height) • Other Native Forest: Other native forest and woody vegetation > 2 m Tenures include publicly and privately owned forest, management includes harvestable and non-harvestable forest and Kyoto compliance is indicated as either being compliant (i.e. likely to have been grassland in 1990) or non-compliant (likely to have been woody vegetation in 1990). Carbon pools covered in the modelling include: • live vegetation, • debris, • soil, • wood products, and

• non-CO2 emissions associated with regeneration burns and CO2 emissions associated with fossil fuel use.

Non-CO2 emissions associated with wildfire were modelled but were excluded from the results as these are likely to be excluded from Kyoto-based accounting of changes in forest carbon stocks (UNFCCC, 2012).

Page 52 of 365

Presentation of estimates

All results are presented as ranges to reflect uncertainty in base parameters. These ranges represent one standard deviation from the primary estimate, as calculated from the outputs of the iterative simulation process described in Sections 2.7 and 6.18.

3.2 CURRENT CARBON STOCKS

The current total area of native forest and native woody vegetation in Tasmania was estimated to be approximately 4.1 million ha (Table 2), comprising 1.6 million ha of Low Eucalypt Forest, 1.0 million ha of Other Native Forest and woody vegetation > 2 m tall, and 1.0 million ha of Tall Eucalypt Forest. Rainforest covers around 0.6 million ha, almost all of which is reserved from timber harvesting on public land. The predominant forest type available for harvesting on public land is Tall Eucalypt Forest (0.4 million ha) while that on private land is Low Eucalypt Forest (0.6 million ha). Altogether, a total of 1.9 million ha of native forest is available for timber harvesting consisting of 0.8 million ha on public land and 1.1 million ha on private land.

In addition, there are 340,000 ha of plantations consisting of 250,000 ha of hardwood plantations and 80,000 ha softwood plantations. Around two thirds of the plantations are located on private land.

Almost all the carbon is held in native forests (97%) with tall and Low Eucalypt Forests each containing 38% and Rainforests 18% (Table 3). Other native forests (including woody vegetation >2m) contain 6% and hardwood and softwood plantations contain 3% of carbon stocks.

Table 2: Estimated total areas for different forest types across different land tenures and management types for Tasmania’s native forests and plantations in 2010. Note Other Non-eucalypt Forest (ONF) includes native woody vegetation > 2 m.

Tenure and Forest Group Average management ETF ELF RFT ONF PHW PSW Native Plant. Total Public Harvestable 435 229 72 81 59 59 816 118 934 Reserved 372 612 460 520 0 2 1,964 2 1,966 Total 807 841 532 601 59 61 2,780 120 2,900 Private Harvestable 133 634 29 295 189 25 1,090 215 1,305 Reserved 15 151 1 71 1 1 237 2 239 Total 147 784 30 366 191 26 1,327 217 1,544 Total Harvestable 567 863 100 376 249 84 1,907 333 2,239 Reserved 387 762 461 591 1 2 2,201 4 2,205 Total 954 1,625 562 967 250 86 4,108 336 4,444

Page 53 of 365

The total stock of carbon in live vegetation in Tasmania’s forests is estimated to currently be around 1,400-1,900 Mt CO2e (Table 3). Most carbon is in public forests (1,100-1,400 Mt CO2e) with private forests containing 400-500 Mt CO2e. The total amount of carbon in debris is estimated to be 1,000-1,500 Mt CO2e, while that in soil is around 600-900 Mt CO2e. Combined, the total amount of carbon stored in live and dead vegetation and soil in Tasmania’s native forests is estimated to be 3,000-4,300 Mt CO2e with a further 100-130 Mt CO2e stored in plantations.

The results indicate that average carbon densities (i.e. in terms of tonnes per hectare) are greatest for Tall Eucalypt Forests and Rainforests. The amount of carbon in live vegetation in Tall Eucalypt Forests averaged 560-730 t CO2e/ha while the that for Rainforests averaged around 450-620 t CO2e/ha (Table 4a). Low Eucalypt Forests are estimated to have around 330- 430 t CO2e/ha while Other Native Forests and plantations have 90-120 t CO2e/ha.

Carbon stocks in reserved areas of native forests (760-1000 Mt CO2e) tend to be greater than those in areas available for timber harvesting (650-850 Mt CO2e). However, carbon densities are similar for both management types (350-450 t CO2e/ha). Carbon stocks on private land are estimated to be less than half those on public land as a result of the smaller area of private native forests (32% of the total) and its lower productivity (mainly Low Eucalypt Forest compared with mainly Tall Eucalypt Forest for public forests). Thus, average carbon densities tend to be lower for forest on private land. The highest carbon densities are in reserved Tall Eucalypt Forest on public land (650-840 t CO2e/ha).

Carbon densities in debris follow similar trends to those in vegetation, with similar carbon densities in harvestable and reserved forest, and greater carbon densities on public compared with private land (Table 4b). However, estimated soil carbon densities tend to be greater in harvestable areas (Table 4c). As with vegetation, carbon densities for both soil and debris were estimated to be greatest in Tall Eucalypt Forest (debris: 400-560 t CO2e/ha; soil: 250-350 t CO2e/ha) and least in Other Native Forest (debris: 80-110 t CO2e/ha; soil: 60-80 t CO2e/ha).

3.3 TRENDS IN PAST AND FUTURE CARBON STOCKS

Modelled trends in carbon stocks in vegetation and soil indicate that they are reasonably stable under the Study Baseline (Figures 2, 3 & 4). There was a slight increase in vegetation carbon stocks from 1990 to 2010, largely as a result of growth in Low Eucalypt Forests. However, over the coming decades these stocks are expected to decline slightly, resulting in a net emission of 20-60 Mt CO2e over 40 years (Table 5). Carbon stocks in plantations are expected to increase by 10-11 Mt CO2e over the next 40 years, partly offsetting the decline in native forests (Table 6).

Carbon stocks in native forest debris are expected to increase over time from 1060-1370 Mt CO2e in 2010 to 1130-1530 Mt CO2e in 2050. Similarly, carbon stocks in plantation debris are also expected to increase from 10-14 Mt CO2e to 55-70 Mt CO2e. Soil carbon stocks are expected to increase in line with the increased inputs from debris.

Page 54 of 365

Table 3: Estimated carbon stocks (Mt CO2e) in (a) live vegetation, (b) debris and (c) soil in native forests and plantations in 2010. a) Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 208 272 71 93 33 46 15 19 9 11 13 16 327 429 22 27 349 456 Reserved 240 314 214 278 205 284 43 57 0 0 0 0 702 932 0 0 703 933 Total 449 585 285 371 238 330 58 76 9 11 13 16 1,030 1,362 22 27 1,051 1,389 Private Harvestable 76 98 212 276 13 18 23 30 16 18 5 6 325 423 21 24 346 447 Reserved 7 10 37 48 1 1 7 9 0 0 0 0 52 67 0 0 52 68 Total 84 108 249 324 14 19 30 39 16 18 6 6 377 490 21 25 398 515 Total Harvestable 286 369 284 369 46 64 38 50 24 28 19 22 653 851 43 51 696 902 Reserved 249 322 250 326 206 285 50 66 0 0 0 1 755 998 1 1 756 999 Total 535 691 534 694 252 349 88 115 24 29 19 23 1,409 1,850 43 51 1,452 1,901 b) Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 157 218 53 75 24 36 13 17 2 3 3 4 247 346 5 7 252 353 Reserved 169 234 153 216 152 230 40 53 0 0 0 0 513 733 0 0 514 733 Total 326 452 206 291 176 266 53 70 2 3 3 4 760 1,079 5 7 765 1,086 Private Harvestable 52 70 151 215 10 15 20 26 4 5 1 1 233 326 5 6 238 332 Reserved 5 7 27 38 0 1 6 8 0 0 0 0 38 54 0 0 39 54 Total 57 77 178 253 10 16 26 34 4 5 1 1 271 379 5 7 276 386 Total Harvestable 211 286 204 290 34 51 32 43 6 8 4 5 481 670 10 14 491 684 Reserved 176 239 179 254 153 231 46 61 0 0 0 0 554 785 0 0 554 785 Total 386 525 384 544 186 282 78 104 6 8 4 5 1,035 1,455 10 14 1,045 1,469 c) Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 95 138 16 31 16 25 10 13 9 13 7 9 138 207 16 22 154 230 Reserved 98 142 47 90 96 146 27 35 0 0 0 0 267 413 0 0 267 413 Total 193 280 63 121 112 171 37 48 9 13 7 10 405 620 16 22 421 642 Private Harvestable 35 50 59 115 7 11 17 22 30 42 3 4 118 197 33 46 151 243 Reserved 3 5 11 20 0 0 6 8 0 0 0 0 20 33 0 0 20 34 Total 38 54 70 135 7 11 23 30 31 42 3 4 138 230 33 46 171 277 Total Harvestable 131 187 76 146 23 35 27 35 39 55 9 13 257 404 49 68 305 472 Reserved 102 146 57 110 96 147 33 43 0 0 0 0 288 445 0 1 288 446 Total 233 333 133 256 119 182 60 78 40 55 10 13 544 849 49 69 593 918

Page 55 of 365

Table 4: Estimated carbon densities (t/ha CO2e) in live vegetation (a), debris (b) and soil (c) in native forests and plantations in 2010. a) Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 479 625 311 405 460 637 181 237 145 179 221 274 406 520 183 226 380 481 Reserved 646 842 349 454 446 618 83 109 5 6 147 182 365 467 131 162 367 464 Total 557 724 338 442 451 617 94 129 151 172 229 261 377 483 190 217 375 466 Private Harvestable 577 741 335 435 462 641 78 103 82 95 213 246 303 383 98 113 269 338 Reserved 508 652 245 319 462 642 98 129 90 104 383 442 222 280 188 216 223 279 Total 566 736 317 414 466 637 80 110 83 94 219 249 286 367 99 113 264 328 Total Harvestable 504 650 329 427 460 638 100 132 97 115 222 262 347 442 128 153 316 398 Reserved 644 831 329 427 446 618 85 111 77 91 213 252 351 446 159 190 353 444 Total 559 726 328 428 452 618 88 122 99 113 226 258 348 445 132 150 337 418 b) Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 362 501 231 328 333 503 158 210 34 52 45 70 308 418 40 61 275 373 Reserved 453 628 250 354 331 500 77 102 3 5 21 31 269 366 19 28 269 365 Total 403 560 237 354 333 498 88 116 38 48 51 63 280 381 45 55 277 361 Private Harvestable 390 527 239 339 349 527 66 88 22 27 43 53 223 289 24 30 190 246 Reserved 348 470 178 253 350 529 86 115 35 43 68 85 169 219 46 57 168 218 Total 380 527 220 329 351 525 70 93 22 27 43 54 208 282 24 30 186 243 Total Harvestable 371 504 237 336 337 510 86 114 24 33 46 64 259 344 30 41 226 298 Reserved 454 617 235 334 331 500 78 104 28 40 33 46 261 347 32 43 262 346 Total 400 555 229 342 334 499 81 108 26 32 48 60 257 349 32 39 245 320 c) Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 219 318 71 136 227 347 124 163 154 216 114 159 161 262 134 188 163 248 Reserved 263 381 76 146 208 318 51 67 81 113 93 130 132 214 92 128 137 209 Total 243 344 78 141 211 321 61 80 154 216 113 159 143 225 133 187 148 219 Private Harvestable 263 374 94 181 242 371 57 74 160 222 106 147 120 169 154 213 126 176 Reserved 221 315 70 135 255 390 85 112 152 211 69 95 93 132 124 172 94 132 Total 260 367 93 168 244 371 62 82 159 223 104 146 108 170 153 214 117 173 Total Harvestable 231 330 88 169 231 354 71 93 159 221 112 156 138 209 147 204 142 205 Reserved 263 377 75 144 208 318 56 73 141 197 86 121 132 201 108 150 136 197 Total 245 347 85 154 213 324 62 81 158 221 111 155 132 207 146 204 137 203

Page 56 of 365

1800

e) 1600 2 1400

1200 PSW 1000 PHW 800 RFT 600 ETF 400 ELF 200 ONF Total C in live vegetation (Mt CO 0 1990 2000 2010 2020 2030 2040 2050 Year

Figure 2: Trends in carbon stocks in live vegetation for all native forests and plantations from 1990-2050 under the Study Baseline. Codes for forest types are provided in text.

1600

1400

e) 1200 2 PSW 1000 PHW 800 RFT 600 ETF

400 ELF

Total C in debris (Mt CO ONF 200

0 1990 2000 2010 2020 2030 2040 2050 Year

Figure 3: Trends in carbon stocks in debris for all native forests and plantations from 1990-2050 for the Study Baseline.

Page 57 of 365

900 800 700 e) 2 600 PSW 500 PHW 400 RFT 300 ETF ELF 200 Total C in soil (Mt CO ONF 100 0 1990 2000 2010 2020 2030 2040 2050 Year

Figure 4: Trends in carbon stocks in soil for all native forests and plantations from 1990-2050 for the Study Baseline.

An important driver of changes in forest carbon stocks is timber harvesting. The rate of sawlog production from native forests has declined over the past 10 years while that from softwood plantations has increased (Figure 5). In coming years, hardwood plantations are expected to supply an increasing proportion of total sawlogs, which is expected to offset future forecast declines in wood production from native forests (Forestry Tasmania, 2008). For the purpose of the Carbon Study, future sawlog production from publicly owned native forests and hardwood plantations was limited to 300,000 m3 in line with Forestry Tasmania’s estimates (2011), with hardwood plantations being harvested in preference to native forests. Sawlog production from private native forests was assumed to be equal to the average for 2006-11 (55,000 m3/y, ABARES 2012a).

Carbon stocks in wood products harvested from 2010 onwards are expected to increase over time to around 10-12 Mt CO2e by 2050 (Figure 6). These stocks will gradually stabilize as emissions from decomposition are balanced by inputs from harvesting and wood processing. Normally, newly harvested wood products are added to existing pools of wood products to account for any variation in the overall pool over time. However, for the purpose of the Carbon Study we assumed that the any changes in the existing pool of wood products would be constant across Study Scenarios, with only changes in future additions influenced by the Study Scenario in question.

Page 58 of 365

Table 5: Estimated native forest carbon stocks (Mt CO2e) in live vegetation, debris, soil and wood products from 1990-2050 for the Study Baseline.

Tenure Year Live vegetation Debris Soil Wood products Total

Min Max Min Max Min Max Min Max Min Max Public 1990 1023 1291 776 1006 410 580 2208 2877 2000 1059 1337 799 1036 427 604 2285 2977 2010 1048 1343 779 1060 391 634 0.3 0.5 2219 3038 2020 1036 1326 825 1090 402 648 1.7 2.4 2264 3066 2030 1037 1327 830 1106 412 655 1.9 2.6 2280 3091 2040 1058 1354 829 1127 420 662 1.6 2.2 2309 3146 2050 1061 1354 828 1149 429 669 2.1 2.9 2319 3174 Private 1990 379 478 282 366 148 210 809 1054 2000 390 492 288 373 153 217 831 1082 2010 383 484 283 367 153 216 0.2 0.2 819 1067 2020 372 467 304 388 160 225 0.7 0.9 836 1081 2030 355 450 307 388 165 232 0.9 1.3 828 1071 2040 341 437 303 384 169 236 1.1 1.4 814 1059 2050 333 428 298 381 172 239 1.1 1.5 803 1049 Total 1990 1401 1770 1058 1372 558 790 3017 3931 2000 1449 1829 1087 1409 580 821 3115 4060 2010 1432 1827 1063 1427 543 850 0.5 0.7 3038 4105 2020 1408 1793 1129 1478 562 873 2.4 3.3 3101 4147 2030 1391 1777 1137 1494 577 887 2.8 3.8 3108 4162 2040 1399 1792 1132 1512 589 898 2.7 3.6 3123 4204 2050 1394 1781 1125 1530 600 908 3.2 4.4 3123 4223

Emissions of non CO2 GHGs from regeneration burning and CO2 from fossil fuel used in management activities are closely linked to total wood production. Total emissions from these sources are estimated to average 0.2-0.3 Mt CO2e/y for all forests in Tasmania (Figure 7). Thus they represent only a small proportion of the total change in carbon stocks of the forest. The large reduction in emissions between 2000 and 2010 is a result of a reduction in both the rate of harvesting and rate of conversion of native forest to plantation and grassland over the past 10 years. Over time, emissions from native forest activities are expected to continue decline, but those from plantations are expected to increase (Table 6).

Estimated net emissions from live vegetation, debris, soil and wood products harvested post 2010 and associated emissions from fossil fuel use and regeneration burns are presented in Table 7 for native forests and Table 8 for plantations for each decade from 1990 to 2050.

Page 59 of 365

1200 /y) 3 1000

800 PSW PHW 600 RFT

400 ETF ELF Total sawlog volume ('000 m 200 ONF

0 1990 2000 2010 2020 2030 2040 2050 Year

Figure 5: Trends in production of sawlogs from all forest native forests and plantations from 1990-2050 for the Study Baseline.

12 e) 2 10

Paper 6 Panels 4 Solid 2 wood Total C in C in Total wood products (Mt CO

0 2010 2020 2030 2040 2050 Year

Figure 6: Trends in carbon stocks in wood products produced from native forests for the Study Baseline.

Page 60 of 365

400

350

300 PSW

and fossil fuel 250

2 PHW CO

- 200 RFT 150 ETF emissions (t/y) 100 ELF ONF 50

Average annual Average annual non 0 1990 2000 2010 2020 2030 2040 2050 Year

Figure 7: Estimated emissions of CO2 from fossil fuel use and non-CO2 GHGs from regeneration burns from 1990-2050 for the Study Baseline.

Table 6: Estimated plantation carbon stocks (Mt CO2e) in live vegetation, debris, soil and wood products from 1990-2050 for the Study Baseline. Figures in Mt CO2e.

Tenure Year Live vegetation Debris Soil Wood products Total

Min Max Min Max Min Max Min Max Min Max Public 1990 9 10 2 2 9 12 19 24 2000 13 14 3 3 11 15 26 32 2010 22 27 5 7 16 22 0.2 0.3 43 57 2020 28 33 15 18 16 22 1.4 1.9 60 75 2030 26 30 21 25 16 22 2.5 3.4 65 80 2040 22 26 21 27 16 22 3.6 4.8 63 79 2050 25 28 22 28 15 22 3.9 5.3 66 84 Private 1990 3 3 1 1 6 8 9 12 2000 10 11 3 3 18 25 30 40 2010 21 25 5 7 33 46 0.2 0.3 60 77 2020 32 37 19 24 31 43 1.3 1.8 84 106 2030 29 36 26 34 30 41 1.9 2.5 87 114 2040 28 33 30 38 29 40 2.3 3.1 89 114 2050 28 34 32 41 29 39 2.5 3.4 91 118 Total 1990 12 14 2 3 14 20 29 36 2000 22 26 5 6 29 40 56 72 2010 43 52 10 14 49 68 0.4 0.6 103 134 2020 60 70 34 43 47 65 2.7 3.7 143 181 2030 54 66 47 59 46 63 4.4 5.9 151 194 2040 50 59 51 65 45 62 5.8 7.9 152 193 2050 53 63 55 70 44 60 6.4 8.7 158 201

Page 61 of 365

Table 7: Estimated net decadal GHG emissions (positive values) and sequestration (negative values) from vegetation, debris, soil, fossil fuel use and regeneration burns (non-CO2 GHGs only) in native forests from 1990 and 2050 for the Study Baseline. All values in Mt CO2e.

Fossil fuel and Tenure Year Live vegetation Debris Soil Wood products Total non-CO2 Min Max Min Max Min Max Min Max Min Max Min Max Public 1990 2000 -46 -37 -30 -23 -24 -17 0.0 0.0 1.9 2.9 -98 -74 2010 -6 11 -24 19 -30 36 -0.5 -0.3 0.9 1.4 -59 67 2020 13 17 -45 -30 -13 -12 -1.9 -1.4 0.8 1.2 -47 -25 2030 -1 -1 -15 -5 -9 -7 -0.2 -0.2 0.4 0.7 -26 -13 2040 -27 -21 -21 1 -9 -7 0.3 0.4 0.6 0.9 -56 -26 2050 -3 1 -22 2 -8 -7 -0.7 -0.5 0.5 0.8 -33 -5 Private 1990 2000 -13 -11 -7 -6 -7 -5 0.0 0.0 0.3 0.4 -28 -21 2010 6 8 5 6 1 1 -0.2 -0.2 0.6 0.9 13 16 2020 11 17 -21 -21 -9 -8 -0.7 -0.5 0.6 0.9 -19 -11 2030 17 17 -3 0 -6 -5 -0.4 -0.3 0.7 1.1 8 12 2040 13 13 4 4 -4 -4 -0.2 -0.1 0.6 0.9 13 15 2050 8 10 4 5 -3 -3 -0.1 -0.1 0.6 0.9 9 13 Total 1990 2000 -60 -47 -37 -29 -31 -22 0.0 0.0 2.2 3.4 -126 -95 2010 3 17 -18 24 -29 37 -0.7 -0.5 1.6 2.3 -43 80 2020 24 34 -66 -51 -23 -19 -2.6 -1.9 1.4 2.1 -66 -36 2030 16 16 -16 -8 -15 -14 -0.6 -0.4 1.2 1.8 -14 -4 2040 -15 -8 -17 5 -12 -11 0.2 0.2 1.2 1.8 -43 -12 2050 5 11 -18 7 -11 -11 -0.8 -0.6 1.1 1.6 -24 8 Table 8: Estimated net decadal GHG emissions (positive values) and sequestration (negative values) from vegetation, debris, soil, fossil fuel use and regeneration burns (non-CO2 GHGs only) in plantations from 1990 and 2050 for the Study Baseline. All values in Mt CO2e.

Fossil fuel and Tenure Year Live vegetation Debris Soil Wood products Total non-CO2 Min Max Min Max Min Max Min Max Min Max Min Max Public 1990 2000 -4 -3 -1 -1 -3 -2 0.0 0.0 0.1 0.2 -8 -6 2010 -13 -9 -4 -2 -8 -5 -0.3 -0.2 0.2 0.4 -24 -17 2020 -6 -6 -11 -10 0 1 -1.6 -1.2 0.2 0.3 -18 -16 2030 2 3 -7 -6 0 0 -1.5 -1.1 0.4 0.5 -6 -3 2040 3 4 -1 -1 0 0 -1.4 -1.0 0.3 0.4 1 3 2050 -3 -2 -2 -1 0 0 -0.5 -0.4 0.2 0.4 -5 -3 Private 1990 2000 -8 -7 -2 -2 -17 -12 0.0 0.0 0.3 0.4 -27 -21 2010 -13 -11 -3 -3 -21 -15 -0.3 -0.2 0.3 0.5 -38 -29 2020 -13 -11 -18 -14 2 3 -1.5 -1.1 0.5 0.8 -29 -22 2030 1 4 -10 -7 1 2 -0.7 -0.5 0.6 0.8 -8 -1 2040 1 3 -4 -3 1 1 -0.6 -0.4 0.5 0.7 -2 1 2050 -1 -1 -3 -3 1 1 -0.3 -0.2 0.6 0.9 -3 -2 Total 1990 2000 -12 -10 -3 -3 -20 -15 0.0 0.0 0.4 0.6 -35 -27 2010 -26 -21 -7 -5 -29 -21 -0.6 -0.4 0.6 0.8 -62 -46 2020 -19 -17 -29 -24 3 4 -3.1 -2.3 0.7 1.1 -47 -38 2030 4 6 -17 -13 1 2 -2.3 -1.7 0.9 1.4 -13 -6 2040 4 7 -5 -4 1 1 -2.0 -1.4 0.8 1.2 -1 4 2050 -4 -3 -5 -4 1 1 -0.8 -0.6 0.8 1.2 -8 -4

Page 62 of 365

Cumulative change in carbon stocks over time for the Study Baseline

The cumulative changes in carbon pools in vegetation, debris and soil from 2010 to 2050 for the Study Baseline are summarised in Tables 9, 10 & 11. Total carbon stocks in native forests and plantations are expected to decrease in live vegetation (29-56 Mt CO2e), but to increase in debris (113-158 Mt CO2e) and soil (34-67 Mt CO2e). Thus, Tasmania’s forests are expected to be net carbon sinks for the period, sequestering a total of 111-194 Mt CO2e (equivalent to 2.8- 4.7 Mt CO2e/y).

Modelled, net losses of carbon from native forest vegetation occur in harvestable native forests, with the majority from privately owned native forest (51-67 Mt CO2e, Table 9). The net loss from vegetation in harvestable publicly owned native forest is estimated to be 5-9 Mt CO2e. In contrast vegetation carbon stocks in unharvested native forests are estimated to increase by 21- 28 Mt CO2e over the 40 years.

Carbon stocks in debris and soil in native forests are expected to increase in both harvested and unharvested forests between 2010 and 2040 with modelled debris carbon stocks increasing by 66-99 Mt (Table 10) and modelled soil carbon increasing by 42-73 Mt CO2e (Table 11). Thus, total carbon stocks in vegetation plus debris are expected to increase by a total of 21-61 Mt CO2e or 0.5-1.5 Mt CO2e/y.

Carbon stocks in plantations are expected to increase in live vegetation (9-11 Mt CO2e) and debris (45-56 Mt CO2e), but may decrease in soil (-5 to -8 Mt CO2e). Most of the change is expected to occur in hardwood plantations. The modelled increase in carbon stocks in vegetation and debris is a result of the hardwood sawlog and pulpwood plantations established over the past 20 years reaching maturity, while the decrease in soil carbon is a result of the lower input rate of debris from these young plantations compared with the previous forest or grassland. These results are consistent with those from Polglase et al. (2000).

Table 9: Change in carbon stocks in live vegetation from 2010 to 2050 for the Study Baseline for different forest types, tenures and management. All values in Mt CO2e.

Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 3 3 -3 -2 -2 -2 -5 -4 3 3 -2 -2 -9 -5 0 1 -8 -4 Reserved 0 1 8 10 3 4 5 7 0 0 1 1 16 21 1 1 17 22 Total 3 4 4 8 0 2 -1 3 3 3 -1 -1 7 16 2 2 9 19 Private Harvestable -16 -12 -42 -32 -1 -1 -9 -6 7 9 -1 -1 -67 -51 6 8 -61 -43 Reserved 1 1 3 4 0 0 1 2 0 1 0 1 5 7 1 1 6 8 Total -15 -11 -39 -28 -1 -1 -7 -5 8 9 0 0 -62 -44 7 9 -55 -35 Total Harvestable -12 -10 -45 -34 -4 -3 -14 -10 10 12 -3 -3 -75 -57 7 9 -68 -48 Reserved 1 2 11 14 3 4 6 8 1 1 2 2 21 28 2 2 23 30 Total -11 -8 -34 -20 -1 1 -8 -2 11 12 -1 -1 -54 -29 9 11 -45 -18

Page 63 of 365

Table 10: Change in carbon stocks in debris from 2010 to 2050 under the Study Baseline for different forest types, tenures and management. All values in Mt CO2e.

Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 4 6 6 9 -1 -1 -2 -1 11 13 6 7 7 13 16 21 23 33 Reserved 20 28 23 35 2 3 3 4 0 0 1 1 48 70 1 1 48 71 Total 24 34 28 44 1 2 1 2 11 13 6 8 55 83 17 22 72 104 Private Harvestable 0 0 7 11 0 0 -2 -2 24 31 2 3 5 10 26 34 31 43 Reserved 1 1 4 5 0 0 1 1 1 1 0 0 6 8 1 1 6 9 Total 1 2 11 17 0 0 -1 0 25 32 2 3 10 18 27 35 37 52 Total Harvestable 4 6 13 20 -1 -1 -4 -3 35 44 8 10 12 22 43 54 55 76 Reserved 21 29 27 40 2 3 4 5 1 1 1 1 54 77 2 2 55 79 Total 25 35 40 60 1 2 0 2 36 45 9 11 66 99 45 56 111 155 Table 11: Change in carbon stocks in soil from 2010 to 2050 for the Study Baseline for different forest types, tenures and management. All values in Mt CO2e.

Tenure and Forest Group Total management ETF ELF RFT ONF PHW PSW Native Plantation Total Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Min Max Public Harvestable 4 6 4 8 0 0 0 0 -1 -1 0 0 8 13 -1 -1 7 13 Reserved 5 7 13 23 1 1 0 1 0 0 0 0 19 32 0 0 19 32 Total 9 13 17 30 1 1 0 0 -1 -1 0 0 27 45 -1 -1 27 45 Private Harvestable 1 2 12 20 0 0 -1 -1 -7 -5 0 0 12 22 -7 -5 5 17 Reserved 0 1 3 5 0 0 0 0 0 0 0 0 3 6 0 0 3 6 Total 2 2 14 25 0 0 -1 0 -7 -5 0 0 15 27 -7 -5 8 22 Total Harvestable 5 8 16 29 0 0 -1 -1 -8 -6 0 0 20 35 -8 -6 12 30 Reserved 6 8 15 28 1 1 1 1 0 0 0 0 22 38 0 0 22 38 Total 11 15 31 56 1 1 -1 0 -8 -6 0 0 42 73 -8 -5 34 68

Page 64 of 365

Spatial variation in carbon stocks

The modelled distributions of carbon stocks in live trees, debris and soil for 2010 are illustrated in Figures 8, 9 and 10. Carbon stocks in live vegetation vary markedly across Tasmania with the highest carbon densities associated with the highly productive Tall Eucalypt Forests and Rainforests in the southern, central, north-western and north-eastern regions of the State. Although there are some extensive areas with >600 t/ha CO2e in live trees, much of the State is covered by forests and other vegetation with less than 100 t/ha CO2e. Carbon stocks in vegetation on cleared agricultural land were assumed to be zero.

Figure 8: Distribution of carbon stocks in live vegetation across Tasmanian forests in 2010.

Page 65 of 365

Debris carbon stocks show a similar pattern of distribution across the state to that of live vegetation. Highest carbon densities for debris (> 600 t CO2e/ha) are located in the southern and central regions and tend to be associated with Tall Eucalypt Forests.

Figure 9: Distribution of carbon stocks in debris across Tasmanian forests in 2010.

Page 66 of 365

Soil carbon stocks tend to be far lower than those in vegetation and debris. Highest soil carbon densities (> 300 t/ha CO2e) were associated with the productive Tall Eucalypt Forests in the Florentine and Styx valleys in Central Tasmania. Many areas were modelled as having < 200 t CO2e/ha. These values are consistent with the figures for pre- or post clearing carbon amounts (which tend to range from 100-500 t/ha CO2e, Webb 2001).

Figure 10: Distribution of carbon stocks in soil across Tasmania in 2010.

Page 67 of 365

Spatial variation in estimated changes in carbon pools in vegetation, debris and soil from 2010 to 2050 are shown in Figures 11, 12 and 13. These show the expected distribution in net carbon sequestration and emissions across the State. Areas of emissions (red dots) are associated with timber harvesting or wildfire events. The highest rate of sequestration in vegetation, debris and soil is in the Tall Eucalypt Forests in southern and central Tasmania. High rates of carbon sequestration in vegetation and debris are also apparent in the north of the state and are associated with the growth of hardwood plantations. Soil carbon is expected to remain relatively constant under forest, but may decline on cleared agricultural land.

Figure 11: Net change in carbon stocks in vegetation from 2010 to 2050 under Study Baseline.

Page 68 of 365

Figure 12: Distribution in net change in carbon stocks in debris from 2010 to 2050 under the Study Baseline.

Page 69 of 365

Figure 13: Distribution in net change in carbon stocks in soils from 2010 to 2050 for the Study Baseline.

Page 70 of 365

3.4 STUDY SCENARIOS: NATIVE FOREST MANAGEMENT

This first set of Study Scenarios explores the impact of changing the amount of native forest available for harvesting as well as the harvesting rate. They include three scenarios that increase the area reserved from timber harvesting while reducing the sawlog cutting quota (N1, N2A and N4), one scenario which attempts to maintain the quota while increasing the area reserved (N2), and one scenario in which timber harvesting ceases altogether (N5). An additional scenario (N3) explores the effect of varying the rotation length (time between successive harvests) on carbon stocks in regrowth forests.

N1: Reservation of 572,000 ha of forest and reducing total sawlog quota from 300,000 m3/y to 155,000 m3/y

Under Study Scenario N1, 572,000 ha of forest, corresponding with the proposed ENGO reserve system, were placed into formal reserves. All harvestable public native forests and plantations were included in the modelling. Private native forests and plantations were excluded and were assumed to continue to be harvested at the Study Baseline rate.

An analysis of the overlay of the ENGO reserve spatial data with current forest types indicated that a large proportion of this area (~250,000 ha) was already allocated to informal reserves, or non-productive forest. Hence, the net reduction in forest available for timber harvesting used in the modelling was only around 333,000 ha (based on the 1 km2 grid). This result is broadly consistent with modelling by Forestry Tasmania which indicates the net reduction in the area available for wood production under the ENGO proposal is 352,000 ha (Forestry Tasmania, 2011b).

The reduction in both the sawlog quota and area available for timber harvesting increased carbon stocks in vegetation, debris and soils over time relative to the Study Baseline (Figures 14, 15 and 16). Over the 40 years, carbon stocks in live vegetation of native forests are predicted to increase by 64-73 Mt CO2e, those in debris are expected to increase by 28-35 Mt CO2e and those in soil are expected to increase by 5 Mt CO2e (Table 12).

Due to the reduced sawlog quota, there was some reduction in harvest intensity in hardwood plantations (Table 13). Carbon stocks in these were modelled as increasing by 6-7 Mt CO2e. However, it is unlikely that this would actually be the case in the real world as the timing of harvesting of these plantations would likely be based on set rotation length and market demand, rather than sawlog quotas.

Page 71 of 365

80 e)

2 70

60 PSW 50 PHW

40 RFT ETF 30 ELF 20 ONF

Total C in live vegetation (Mt CO 10 Total 0 2010 2020 2030 2040 2050 Year

Figure 14: Cumulative increase in carbon stocks in vegetation for Study Scenario N1, for public forest.

30 e) 2

25 PSW PHW 20 RFT 15 ETF ELF 10 ONF Total C in live debris (Mt CO 5 Total

0 2010 2020 2030 2040 2050 Year

Figure 15: Cumulative increase in carbon stocks over time in debris for Study Scenario N1, for public forest.

Page 72 of 365

5 e) 2 PSW 4 PHW RFT 3 ETF 2 ELF ONF Total C in live soil (Mt CO 1 Total

0 2010 2020 2030 2040 2050 Year

Figure 16: Cumulative increase in carbon stocks over time in soil for Study Scenario N1, for public forest.

Table 12: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario N1 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 22 25 4 5 0 0 -1 -1 -1 0 26 30 2030 40 45 12 16 1 2 -1 -1 -1 -1 54 63 2040 43 49 20 26 3 3 -1 -1 -2 -2 67 79 2050 64 73 28 35 5 5 -2 -2 -3 -2 97 114

Table 13: Differences in carbon stocks (Mt CO2e) in plantations for Study Scenario N2 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 0.4 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.3 0.4 2030 1.3 1.5 0.1 0.1 0.0 0.0 -0.1 -0.1 0.0 0.0 1.3 1.6 2040 4.9 6.1 1.2 1.6 0.1 0.1 -0.4 -0.3 0.0 0.0 5.8 7.4 2050 4.1 4.9 1.8 2.3 0.3 0.3 -0.3 -0.2 0.0 0.0 5.9 7.4

Page 73 of 365

As a result of the reduced rate of harvesting in native forests, there was a 2 Mt CO2e reduction in carbon stocks in wood products over the 40 years. However, this change in carbon stocks in wood products represented only around 2% of the total increase in carbon stocks in native forests associated with the reduction in the rate of harvest. Further, it was balanced by a 2-3 Mt CO2e decrease in CO2 emissions from fossil fuel use and non-CO2 emissions from regeneration burns over the same time period.

Reduction in native forest harvesting resulted in a long-term, steady increase in carbon stocks in vegetation, debris and soil. There was no sign of any decrease in the rate of carbon sequestration by 2050, so the total sequestration potential of the scenario is expected to increase further, beyond 2050. The main forest type influenced by this Study Scenario was Tall Eucalypt Forest, which contributed around 75% of total abatement. This forest type is the main forest type targeted by the proposed reserves and is also the primary source of harvested wood products, so its large influence on the carbon stocks is expected.

The spatial distribution of the cumulative change in carbon stocks in vegetation and debris for Study Scenario N1 relative to the Study Baseline is shown in Figure 17. A large proportion of forests across the area included in Study Scenario N1 increased their carbon their carbon stock as a result of being reserved from timber harvesting or the reduced cutting rate. However, carbon stocks in some forests decreased (shown in red), probably as a result of changes in harvesting patterns associated with the reservation of some forests.

The estimated net carbon sequestration in native vegetation and debris for the Study Scenario was 97-114 Mt CO2e over 40 years. This estimate includes reductions in carbon sequestered in wood products and non-CO2 plus fossil CO2 emissions associated with harvesting native forests, but excludes any potential change in carbon stocks in plantations. This figure is around 75% of that estimated by Macintosh (2012a) for a shorter time period (~8 Mt CO2e /y between 2013 and 2030). This is the result of a number of factors including the assumed timber harvest rate for the reference scenario (assumed to be the same as that in 2008 in the study by Macintosh, but based on forward estimates by Forestry Tasmania in this Carbon Study) and the effect of wildfire which was considered in this Carbon Study, but was not included in the study by Macintosh.

Wildfire limits the potential increase in forest carbon stocks over time by volatilising a proportion of carbon and moving carbon in live vegetation pools to debris which gradually decompose. The default growth rates for harvested native forest used in the study by Macintosh assume that carbon stocks in live vegetation increase up to an age of 200 years (DCCEE 2011b, Table 7.B5). If fire is excluded from the modelling, these default growth rates would result in live vegetation in all Tall Eucalypt Forest eventually having a carbon stock of 1,400 t CO2e/ha – almost double the estimated average current carbon stocks in Tasmanian Tall Eucalypt Forest forests reserved from harvest (640-830 t CO2e/ha). However, if 50% of trees are burnt once every 140-250 years, as assumed in the Carbon Study, then the average carbon stock will be considerably lower. This reduction in average carbon stocks in unharvested forests results in a smaller increase in carbon stocks for the N1 Study Scenario as modelled here.

Page 74 of 365

Figure 17: Distribution in cumulative changes in carbon stocks in live vegetation plus debris for the N1 Study Scenario relative to the Study Baseline.

Page 75 of 365

N2: Cease harvesting of old-growth by 2020 while maintain sawlog production at 300,000 m3/y

Under Study Scenario N2, all old growth forest (as defined by the the RFA 5 Yearly Review mapping of old-growth plus maturity class) was reserved from timber harvesting after 2020, but the hardwood sawlog quota was maintained (as far as possible) at 300,000 m3/y. The harvesting rate of private native forest was assumed to decrease from 55,000 m3/y to 51,400 m3/y consistent with the reduction in area available for harvest.

The results indicated that a net reduction in forest carbon stocks in public native forests could be expected as a consequence of this Study Scenario. Carbon stocks in vegetation decreased by 2-3 Mt CO2e, while those in debris remained fairly constant (Table 14). In contrast, carbon stocks increased slightly (4-5 Mt CO2e) in private native forest vegetation and debris over the same period as a result of the reduced rate of private forest harvesting. There was little change in carbon stocks in soil, little change in emissions from fossil fuel use and regeneration burns and no change in carbon stocks in harvested wood products.

The reduction in total carbon stocks for Study Scenario N2 was unexpected, because it was assumed that the reservation of high carbon content, old-growth forests would more than offset any increase in carbon losses associated with increased harvesting of regrowth forest. However, this may not to be the case. To maintain the sawlog quota, the harvesting intensity was assumed to increase in the remaining forest both, before and after, its implementation in 2020. This was through a combination of increased thinning and reduced rotation lengths in the regrowth forest available for harvesting as suggested by Forestry Tasmania (2008).

Table 14: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario N2 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 -0.6 -0.5 0.7 0.9 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.4 2030 -2.1 -1.6 1.4 1.7 0.1 0.1 0.0 0.0 0.0 0.0 -0.6 0.2 2040 -3.0 -2.4 0.9 1.1 0.0 0.0 0.0 0.0 0.0 0.0 -2.1 -1.2 2050 -3.0 -2.3 -0.2 -0.2 0.0 0.0 0.0 0.0 0.2 0.2 -3.5 -2.7 Private 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 1.2 1.5 0.4 0.6 0.0 0.0 0.0 0.0 0.0 0.0 1.6 2.1 2030 1.7 2.1 0.9 1.2 0.1 0.1 0.0 0.0 -0.1 -0.1 2.7 3.6 2040 2.3 2.9 1.0 1.4 0.2 0.2 0.0 0.0 -0.2 -0.1 3.6 4.7 2050 2.2 2.9 1.4 2.0 0.2 0.3 0.0 0.0 -0.2 -0.2 4.0 5.4 Total 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 0.7 0.8 1.2 1.5 0.0 0.1 0.0 0.0 0.0 0.0 1.9 2.4 2030 0.0 0.1 2.3 3.0 0.1 0.2 0.0 0.0 -0.1 -0.1 2.5 3.3 2040 -0.1 -0.1 1.9 2.5 0.2 0.3 0.0 0.0 -0.2 -0.1 2.1 2.9 2050 -0.1 -0.1 1.3 1.7 0.2 0.3 0.0 0.0 0.0 0.0 1.3 1.9

Page 76 of 365

The modelled effect of the Study Scenario on carbon pools in vegetation, debris and soil for different forest types over time is illustrated in Figures 18, 19 and 20. An increase in carbon stocks in vegetation in Low Eucalypt Forest (the predominant forest type on private land) was offset by a decline in carbon stocks in Tall Eucalypt Forest (the predominant forest type on public land) after the cessation of old-growth harvesting in 2020. There was a small initial increase in carbon stocks in debris as a result of the greater harvesting intensity in regrowth forest. However, these stocks started to decrease after 2030. There was little change in soil carbon stocks. The overall change in forest carbon stocks over 40 years for the Study Scenario is estimated to be a marginal gain of 1-2 Mt CO2e (Table 14).

The distribution across the State in the change in carbon stocks in vegetation and debris relative to the Study Baseline is shown in Figure 21. There is some small increase in carbon stocks in reserved forests indicated as green or yellow. However, many stands (indicated as red dots) show reductions in carbon in live vegetation.

4 e) 2 3

PSW 2 PHW

1 RFT ETF 0 ELF 2010 2020 2030 2040 2050 ONF -1

Total C in live vegetation (Mt CO Total -2

Year

Figure 18: Cumulative change in total carbon stocks in vegetation for Study Scenario N2, for all forest types.

Page 77 of 365

e) 2.5 2 PSW 2 PHW

1.5 RFT ETF 1 ELF ONF

Total C in live debris (Mt CO 0.5 Total

0 2010 2020 2030 2040 2050 Year

Figure 19: Cumulative change in total carbon stocks in vegetation for Study Scenario N2, for all forest types.

0.8 e) 2 PSW PHW 0.6 RFT ETF 0.4 ELF ONF

Total C in live soil (Mt CO 0.2 Total

0 2010 2020 2030 2040 2050 Year

Figure 20: Cumulative change in total carbon stocks in soil for Study Scenario N2, for all forest types.

Page 78 of 365

Figure 21: Distribution in cumulative changes in carbon stocks in live vegetation plus debris for Study Scenario N2 relative to the Study Baseline.

Page 79 of 365

N2A: Cease harvesting of old-growth by 2020 while maintain sawlog production at 155,000 m3/y

Assumptions for Study Scenario N2A were similar to N2 except that the sawlog quota was decreased from 300,000 m3/y to 155,000 m3/y for public native forests and plantations and there was no increase in harvesting intensity of unreserved forest. Harvesting of private native forest was assumed to decrease from 55,000 m3/y to 51,400 m3/y in line with the reduction in harvestable area as for Study Scenario N2.

In contrast to N2, there were substantial increases in carbon stocks in vegetation, debris and soil associated with Study Scenario N2A relative to the Study Baseline (Figures 22, 23 and 24). Altogether, an additional 63 - 81 Mt CO2e was estimated to be sequestered by live vegetation, 29-36 Mt CO2e in debris and 4-6 Mt CO2e in soil pools in native forests over 40 years (Table 15). In addition, there was a small increase in the amount of carbon in hardwood plantations as a result of reduced harvesting intensity associated with the reduced sawlog quota. However, it was assumed that any reduction in the quota would have no impact on the actual management of these plantations, so this Carbon store amount was excluded from consideration in the final result. Under N2A, carbon stocks in all forest pools continued to increase past 2050. There was a reduction in the amount of carbon sequestered in harvested wood products (2 Mt CO2e over the 40 years), representing around 2% of the total amount of carbon sequestered by the forests. However, as with Study Scenario N1 this decrease was balanced by a similar reduction in fossil CO2 and non-CO2 emissions associated with harvesting and regeneration burns.

80 e)

2 70

60 PSW 50 PHW

40 RFT ETF 30 ELF 20 ONF

Total C in live vegetation (Mt CO 10 Total 0 2010 2020 2030 2040 2050 Year

Figure 22: Cumulative increase in carbon stocks in vegetation for Study Scenario N2A, for all forest types.

Page 80 of 365

35 e) 2 30 PSW 25 PHW

20 RFT ETF 15 ELF 10 ONF Total C in live debris (Mt CO 5 Total

0 2010 2020 2030 2040 2050 Year

Figure 23: Cumulative increase in total carbon stocks in debris for Study Scenario N2A, for all forest types.

5 e) 2 PSW 4 PHW RFT 3 ETF 2 ELF ONF Total C in live soil (Mt CO 1 Total

0 2010 2020 2030 2040 2050 Year

Figure 24: Cumulative increase in total carbon stocks over time in soil for Study Scenario N2A, for all forest types.

Page 81 of 365

Table 15: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario N2A relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and Tenure Year Live vegetation Debris Soil Wood products Total non-CO2 Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 21 26 5 6 0 0 -1 -1 -1 0 25 32 2030 38 49 13 16 1 2 -1 -1 -1 -1 52 67 2040 41 53 21 26 3 4 -1 -1 -2 -2 65 84 2050 60 77 28 36 4 6 -2 -2 -3 -2 92 120 Private 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 1 1 0 0 0 0 0 0 0 0 2 2 2030 1 2 1 1 0 0 0 0 0 0 2 3 2040 3 3 1 1 0 0 0 0 0 0 4 5 2050 3 3 2 2 0 0 0 0 0 0 5 6 Total 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 22 28 5 7 0 0 -1 -1 -1 -1 27 34 2030 40 51 14 18 1 2 -2 -1 -1 -1 54 71 2040 44 56 21 28 3 4 -1 -1 -2 -2 68 89 2050 63 81 29 38 4 6 -2 -2 -3 -2 96 127 The spatial distribution of the modelled change in carbon stocks in vegetation plus debris by 2050, relative to the Study Baseline, is shown in Figure 25. Carbon stocks tended to increase most in publicly owned forests, with some increasing by up to 300 t CO2/ha. Larger increases tended to be associated with the more productive forests in the central, north western and north eastern regions.

N3: Varying rotation length from 80 years to 60 or 120 years

For Study Scenario N3, rotation lengths of regrowth native forests were varied around the assumed Study Baseline of 80 years, from 60 through to 120 years. The sawlog quota (representing the maximum amount of sawlog harvested annually) was maintained at 300,000 m3/y but, unlike other Scenarios, harvesting activity was constrained to occur only in regrowth forests. Because of this change, the Study Baseline was also constrained to regrowth forest only, with the rotation lengths set at 80 years. Both public and privately owned native forests were included.

Reducing rotation length to 60 years resulted in a substantial decrease in carbon stocks in vegetation and debris and a small decrease in soil carbon relative to the Study Baseline (Figures 26, 27 and 28, Table 16). In contrast, increasing rotation length to 100 or 120 years increased carbon stocks in all forest pools (Figures 29, 30 and 31, Table 16).

Page 82 of 365

Figure 25: Distribution in cumulative changes in carbon stocks in live vegetation plus debris for Study Scenario N2A relative to the Study Baseline.

Page 83 of 365

e) 5 2 0 2010 2020 2030 2040 2050 -5 ETF

-10 ELF -15

-20 ONF -25 Total

Total C in live vegetation (Mt CO -30

-35

Year

Figure 26: Cumulative change in total carbon stocks in vegetation for Study Scenario N3 in which rotation length was reduced to 60 years relative to the Study Baseline of 80 years .

4 2 e) 2 0 2010 2020 2030 2040 2050 -2 ETF -4 -6 ELF -8 ONF -10 -12 Total C in live debris (Mt CO Total -14 -16 Year

Figure 27: Cumulative change in total carbon stocks in debris for Study Scenario N3 in which rotation length was reduced to 60 years relative to the Study Baseline of 80 years .

Page 84 of 365

0.5

e) 2010 2020 2030 2040 2050 2 -0.5 ETF

-1 ELF -1.5 ONF -2 Total C in live soil (Mt CO -2.5 Total

-3 Year

Figure 28: Cumulative change in total carbon stocks in soil for Study Scenario N3 in which rotation length was reduced to 60 years relative to the Study Baseline of 80 years .

Table 16: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario N3 for rotation lengths of 60, 80 and 120 years relative to the Study Baseline of 80 years showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Scenario Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max N3_60Y 2020 -10 -8 -2 -1 0 0 0 0 0 0 -11 -9 2030 -25 -20 -7 -5 -1 -1 1 1 0 0 -32 -26 2040 -34 -27 -10 -8 -2 -1 1 1 1 1 -47 -37 2050 -31 -25 -15 -12 -3 -2 0 0 1 1 -50 -39 N3_100Y 2020 0 1 1 1 0 0 0 0 0 0 1 2 2030 1 2 1 1 0 0 0 0 0 0 3 4 2040 9 11 3 4 0 1 0 0 0 0 13 17 2050 30 37 8 10 1 1 -1 -1 -1 -1 38 49 N3_120Y 2020 15 18 3 3 0 0 -1 0 0 0 17 22 2030 23 29 7 9 1 1 -1 -1 -1 -1 32 41 2040 20 25 12 15 2 3 0 0 -1 -1 34 43 2050 32 41 14 18 2 3 -1 -1 -1 -1 49 62

Page 85 of 365

e) 35 2 30

25 ETF

20 ELF 15

10 ONF 5 Total

Total C in live vegetation (Mt CO 0 2010 2020 2030 2040 2050 -5

Year

Figure 29: Cumulative change in total carbon stocks in vegetation for Study Scenario N3 in which rotation length was increased to 120 years relative to the Study Baseline of 80 years .

16 e) 2 14

12 ETF

10 ELF 8

6 ONF

4 Total C in live debris (Mt CO Total 2

0 2010 2020 2030 2040 2050 Year

Figure 30: Cumulative change in total carbon stocks in debris for Study Scenario N3 in which rotation length was increased to 120 years relative to the Study Baseline of 80 years .

Page 86 of 365

3.5

3 e) 2 2.5 ETF

2 ELF 1.5 ONF 1 Total C in live soil (Mt CO 0.5 Total

0 2010 2020 2030 2040 2050 Year

Figure 31: Cumulative change in total carbon stocks in soil for Study Scenario N3 in which rotation length was increased to 120 years relative to the Study Baseline of 80 years . Changing rotation length from 80 years to 60 years reduced carbon stocks in vegetation by 25- 31 Mt CO2e between 2010-2050 while those in debris were decreased by 12-15 Mt CO2e (Table 16). Most of the decrease in carbon in live vegetation occurred in the first 20 years, with vegetation carbon stocks recovering slightly between 2040 and 2050. There was only a small effect of the reduction in rotation length on soil carbon and no lasting effect on carbon in harvested wood products or non-CO2 and fossil GHG emissions.

In comparison, increasing rotation length to 120 years increased carbon stocks in live vegetation by 32-41 Mt CO2e, while those in debris increased by 14-18 Mt CO2e. Soil carbon stocks increased by just 2-3 Mt CO2e by 2050. Again there was little effect on carbon stocks in wood products or emissions from fossil fuel use or regeneration burns. However, the average annual volume of sawlog harvested was 60,000 m3 less than that for the Study Baseline.

The estimated total difference in carbon stocks over the 40 years relative to the study Baseline was a 39-50 Mt CO2e decrease for the 60 year rotation length compared with a 49-62 Mt CO2e increase for the 120 year rotation length.

In terms of changes in carbon density, reducing rotation length to 60 years reduced carbon density in live vegetation by 32-40 t CO2e/ha while increasing rotation length to 120 years increased it by 42-52 t CO2e/ha . Carbon density in debris and soil followed similar patterns to that in vegetation, with a 15-19 t CO2e/ha reduction in total carbon stocks for the 60 year rotation length and an 18-23 t CO2e/ha increase in carbon stocks for the 120 year rotation length compared with the Study Baseline.

Page 87 of 365

The results for Study Scenario N3 indicate that rotation length has a major influence on forest carbon stocks. Increasing rotation length increases carbon pools by allowing the stand to grow for longer, sequestering more carbon, and by increasing the time between harvesting events. These results are consistent with those from Grove et al. (2011) who showed that forest debris mass could be expected to be greater for longer rotations (200 years) than shorter rotations (100 years).

This result has important implications for forest mangers wishing to maximise carbon stocks in forests. Increasing rotation length can increase carbon stocks in vegetation and debris. However, it may also decrease the volume of wood harvested. Thus, the net impact on total carbon stocks will depend on the recovery, use and lifespan of final wood products as well as the change in total forest carbon.

N4: Cease harvesting of old-growth native forests immediately

Under Study Scenario N4, all forest areas identified as old growth (according to the 5 Yearly RFA old-growth classification) whether public or private, were assumed to be reserved from timber harvesting. Harvesting across the remainder of the estate was assumed to be the same as for the Study Baseline (i.e. resulting in an effective reduction in the overall harvest rate).

For this Study Scenario, there were large increases in carbon stocks in all forest carbon pools (Figures 32, 33 & 34). The additional amount of carbon sequestered, relative to the Study Baseline was 23-29 Mt CO2e for vegetation, 13-17 Mt CO2e for debris and 2-3 Mt CO2e for soil carbon (Table 17). There was an estimated 0-1 Mt CO2e reduction in carbon in harvested wood products as a result of the reduced rate of harvesting and a 1 Mt CO2e reduction in emissions from fossil fuel use and non-CO2 emissions from regeneration burns. Thus, the net change in carbon stocks was estimated to be 38-50 Mt CO2e.

Unlike N1 and N2A, the rate of increase in carbon in live vegetation for Study Scenario N4 slowed over time; roughly stabilising from 2030. This difference in the long term rate of carbon sequestration is probably due to the modelled old growth stands approaching maximum standing biomass around this time. In contrast, modelled carbon stocks in debris and soil continued to increase over time, although there was some indication that the rate of increase in debris carbon may slow after 2030.

Page 88 of 365

30 e) 2 25

RFT 20 ETF 15 ELF 10 ONF 5 Total Total C in live vegetation (Mt CO

0 2010 2020 2030 2040 2050 Year

Figure 32: Cumulative change in total carbon stocks in vegetation for Study Scenario N4 relative to the Study Baseline.

14 e) 2 12 RFT 10 ETF 8 ELF 6 ONF 4 Total C in live debris (Mt CO 2 Total

0 2010 2020 2030 2040 2050 Year

Figure 33: Cumulative change in total carbon stocks in debris for Study Scenario N4 relative to the Study Baseline.

Page 89 of 365

2.5 e) 2 RFT 2 ETF 1.5 ELF 1 ONF

Total C in live soil (Mt CO 0.5 Total

0 2010 2020 2030 2040 2050 Year

Figure 34: Cumulative change in total carbon stocks in soil for Study Scenario N4 relative to the Study Baseline.

Table 17: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario N4 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and Tenure Year Live vegetation Debris Soil Wood products Total non-CO2 Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 11 14 3 3 0 0 -1 0 0 0 13 17 2030 14 18 7 9 1 1 -1 0 0 0 22 28 2040 13 17 9 11 1 2 0 0 0 0 24 30 2050 15 19 10 13 2 2 0 0 0 0 27 35 Private 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 3 4 0 1 0 0 0 0 0 0 3 4 2030 5 6 1 2 0 0 0 0 0 0 6 8 2040 6 8 2 3 0 0 0 0 0 0 9 12 2050 8 10 3 4 1 1 0 0 0 0 12 15 Total 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 14 17 3 4 0 0 -1 -1 0 0 17 21 2030 19 24 8 11 1 1 -1 -1 -1 0 28 36 2040 20 25 11 15 1 2 -1 0 -1 -1 33 43 2050 23 29 13 17 2 3 -1 0 -1 -1 38 50

Page 90 of 365

Most of the increase in carbon stocks was associated with the Tall Eucalypt forests in Central Tasmania with small declines in some Low Eucalypt Forest along the East Coast (Figure 35). Many Stands showed large increases in carbon (e.g. central Tasmania). However, a number of locations do show decreases (e.g. eastern Tasmania). The forests in which carbon stocks decrease are most likely Low Eucalypt Forest, which are common in the east of the state. This difference may arise from modelled changes in fire regimes in these forests resulting from the exclusion of harvesting from old-growth forests.

Figure 35: Distribution in cumulative changes in carbon stocks in live vegetation plus debris for the N4 Study Scenario relative to the Study Baseline.

Page 91 of 365

N5: Cease all harvesting of public and private native forests immediately

Under this Study Scenario, harvesting was assumed to cease across both public and private native forests immediately. In effect, fire was the only disturbance factor modelled.

As with N4, carbon stocks increased in forest carbon pools (Figures 36, 37 & 38). However, unlike N4, stocks in live vegetation continued to increase beyond 2030. The total increase over the 40 years was estimated to be 142-178 Mt CO2e for live vegetation, 66-85 Mt CO2e for debris and 11-15 Mt CO2e for soil (Table 18). Cessation of harvesting reduced carbon pools in wood products by 3-4 Mt CO2e, but also reduced cumulative emissions from fossil fuel use and non- CO2 emissions from regeneration burns by 5-7 Mt CO2e. Thus, the overall change calculated across all carbon stocks relative to the Study Baseline was 218-286 Mt CO2e.

The spatial distribution of the forest stands reserved in this Study Scenario and the associated changes in carbon stocks relative to the Study Baseline is shown in Figure 39. The distribution of the change in carbon indicates that the largest increases are likely to be located in the productive Tall Eucalypt Forests in central, north western and north eastern Tasmania. Some areas showed decreases in carbon stocks associated with wildfire events (areas coloured red or orange) such as in forests in the south-east of the state. In the absence of harvesting, forest carbon stocks increased, resulting in larger potential losses in the event of fire. However, despite the increased carbon stocks in most forests, there was little evidence of wildfire having a major impact, with only a relatively small proportion of stands showing major declines.

180

e) 160 2 140 RFT 120

100 ETF

80 ELF 60 ONF 40 Total Total C in live vegetation (Mt CO 20

0 2010 2020 2030 2040 2050 Year

Figure 36: Cumulative change in total carbon stocks in vegetation for Study Scenario N5 relative to the Study Baseline.

Page 92 of 365

70 e) 2 60 RFT 50 ETF 40 ELF 30 ONF 20 Total C in live debris (Mt CO 10 Total

0 2010 2020 2030 2040 2050 Year

Figure 37: Cumulative change in total carbon stocks in debris for Study Scenario N5 relative to the Study Baseline.

12 e) 2 10 RFT

8 ETF

6 ELF

4 ONF Total C in live soil (Mt CO 2 Total

0 2010 2020 2030 2040 2050 Year

Figure 38: Cumulative change in total carbon stocks in soil for Study Scenario N5 relative to the Study Baseline.

Page 93 of 365

Table 18: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario N5 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and Year Live vegetation Debris Soil Wood products Total non-CO2 Min Max Min Max Min Max Min Max Min Max Min Max

2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 40 50 8 10 1 1 -2 -1 -1 -1 47 60 2030 60 77 23 29 2 3 -2 -2 -2 -1 85 109 2040 59 75 33 42 5 6 -2 -1 -3 -2 97 125 2050 76 97 40 51 7 9 -2 -2 -3 -3 122 159

2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 20 26 2 3 0 0 -1 -1 -1 -1 23 29 2030 38 48 10 13 1 2 -1 -1 -2 -1 49 64 2040 53 68 18 24 2 4 -1 -1 -3 -2 74 97 2050 65 83 25 35 4 6 -1 -1 -3 -3 96 127

2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 60 76 10 13 1 1 -3 -2 -2 -2 70 90 2030 98 125 32 42 3 5 -3 -2 -3 -3 133 173 2040 111 143 51 67 7 10 -3 -2 -5 -4 170 222 2050 141 180 65 87 11 15 -4 -3 -7 -5 218 286

It was assumed that the risk of fire was related to stand age, as suggested by Lindenmayer et al. (2009). Thus, as the proportion of older stands increased in the absence of timber harvesting, the average frequency of fire was assumed to decrease. This assumption appears reasonable, given the different structure and micro-climate in younger stands compared with older stands. However, it needs to be tested and validated with further research.

Page 94 of 365

Figure 39: Distribution in cumulative changes in carbon stocks in live vegetation plus debris for Study Scenario N5 relative to the Study Baseline.

Page 95 of 365

Summary of native forest management Study Scenarios

Outputs from the FCMF for the native forest management Study Scenarios suggest there is potential to increase carbon stocks in Tasmania’s native forests by reducing the rate of wood extraction, either by reducing the sawlog quota and reserving areas from forest, or by increasing rotation lengths (Table 19). On the other hand, reserving areas of forest without reducing the sawlog quota may actually have a negative impact on carbon stocks.

The Study Scenario with the most sequestration potential was N5 (all harvesting ceases, Figure 40). N1 and N2A also resulted in substantial increases in forest carbon stocks. In contrast, increasing harvest intensity by reducing rotation length or maintaining the current sawlog quota across a smaller harvestable area is likely to reduce carbon stocks. The results indicate that reducing harvesting rates decreases carbon stocks in wood product pools. However, this effect is small relative to the increase in carbon stocks in live vegetation and debris.

Table 19: Summary of estimated cumulative differences in carbon stocks (Mt CO2e) in live vegetation plus debris in native forest relative to the Study Baseline for alternative Study Scenarios N1-N5 (2010 to 2050).

Scenario and Tenure Area Change in carbon stocks in vegetation and debris (Mt CO2e) Rotation length 2020 2030 2040 2050 000 ha Min Max Min Max Min Max Min Max N1 Public 843 27 31 54 62 70 82 98 114 Private 0 0 0 0 0 0 0 0 0 N2 Public 843 0 0 -1 0 -2 -1 -3 -3 Private 1038 2 2 3 3 3 4 4 5 N2A Public 843 25 32 51 65 62 79 88 113 Private 1038 1 2 2 3 4 5 4 6 N3 60Y Public 432 -10 -8 -26 -21 -35 -28 -40 -32 60Y Private 345 -1 -1 -6 -5 -10 -8 -6 -5 120Y Public 432 13 16 21 26 22 28 30 39 120Y Private 345 4 5 10 13 10 12 16 20 N4 Public 432 14 17 22 27 23 28 25 32 Private 345 3 4 6 8 9 11 11 14 N5 Public 133 48 60 84 105 92 117 116 147 Private 104 23 29 48 60 72 91 92 116

Page 96 of 365

300

250

200 e) 2

150 Vegetation 100 Debris Baseline (Mt CO 50

0 N1 N2 N2A N3_120Y N4 N5 Change in carbon stocks relative to Study to Study relative stocks carbon in Change

-50

Scenario

Figure 40: Summary of modelled cumulative differences in carbon stocks in vegetation and debris for Study Scenarios (N1-N5) relative to the Study Baseline between 2010 and 2050.

3.5 STUDY SCENARIO: VARYING PLANTATION MANAGEMENT

The Study Scenarios considered in this section focus on the effects of varying plantation management on carbon stocks in Tasmania’s softwood and hardwood plantations. These forests were estimated to cover a total of 340,000 ha and could have potential to sequester more carbon under alternative harvesting regimes, or if growth rates were improved. The Study Scenarios investigated here include an increase in rotation length, conversion of hardwood pulplog plantations to sawlog plantations and increasing growth rates by 25%.

P1: Increase plantation rotation length by 10%

Increasing plantation rotation lengths by 10% increased carbon stocks in vegetation, debris and soil over time (Figures 41, 42 & 43). The modelled total increase in carbon stocks was 5-6 Mt CO2e for vegetation 3-4 Mt CO2e for debris and 1 Mt CO2e for soil (Table 20). This increase stabilised after 2030 for vegetation and after 2040 for debris, indicating that carbon stocks had achieved a new equilibrium with the longer rotation lengths.

Page 97 of 365

7 e) 2 6

4 PSW

3 PHW Total 2

1 Total C in live vegetation (Mt CO

0 2010 2020 2030 2040 2050 Year

Figure 41: Cumulative change in total carbon stocks in live vegetation for Study Scenario P1 relative to the Study Baseline.

4.5

4 e) 2 3.5

3 PSW 2.5

2 PHW

1.5 Total 1 Total C in live debris (Mt CO 0.5

0 2010 2020 2030 2040 2050 Year

Figure 42: Cumulative change in total carbon stocks in debris for Study Scenario P1 relative to the Study Baseline.

Page 98 of 365

1.2

1 e) 2 0.8 PSW 0.6 PHW

0.4 Total

Total C in live soil (Mt CO 0.2

0 2010 2020 2030 2040 2050 Year

Figure 43: Cumulative change in total carbon stocks in soil for Study Scenario P1 relative to the Study Baseline.

Table 20: Differences in carbon stocks (Mt CO2e) in plantations for Study Scenario P1 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 2.9 3.7 1.3 1.7 0.0 0.1 0.0 0.0 0.1 0.1 4.2 5.3 2030 2.6 3.0 1.1 1.3 0.2 0.3 0.2 0.2 0.0 0.0 4.0 4.9 2040 1.5 1.8 1.8 2.2 0.4 0.5 0.4 0.5 0.0 0.0 3.9 5.0 2050 2.4 3.0 1.0 1.2 0.4 0.5 0.3 0.3 0.0 0.0 4.0 5.1 Private 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 0.3 0.3 0.9 1.2 0.1 0.1 0.1 0.1 0.1 0.1 1.3 1.7 2030 2.9 3.8 0.9 1.2 0.1 0.2 0.0 0.0 0.1 0.1 3.9 5.1 2040 2.6 3.3 1.8 2.2 0.4 0.5 0.2 0.3 0.2 0.3 4.7 6.1 2050 2.7 3.5 2.4 3.1 0.5 0.7 0.2 0.3 0.3 0.3 5.5 7.3 Total 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 3.2 4.0 2.2 2.9 0.1 0.2 0.1 0.1 0.1 0.2 5.4 7.0 2030 5.5 6.9 2.0 2.5 0.4 0.5 0.1 0.1 0.0 0.1 7.9 10.0 2040 4.0 5.1 3.6 4.5 0.7 1.0 0.6 0.7 0.2 0.3 8.6 11.1 2050 5.1 6.4 3.4 4.3 0.9 1.2 0.5 0.6 0.2 0.4 9.6 12.3

Page 99 of 365

Increasing rotation lengths by 10% increased carbon stocks in wood products by around 0.5 Mt CO2e as a result of an increase in the proportion of sawlogs recovered from harvested trees. This result contrasts with that from native forests (Study Scenario N3) where increasing rotation length reduced carbon stocks in wood products (Table 20).

P2: Converting all hardwood pulpwood plantations to sawlog

Conversion of hardwood pulpwood plantations to sawlog plantations resulted in a substantial increase in carbon stocks in vegetation, debris and soil relative to the Study Baseline (Figures 44, 45 & 46). After peaking above 50 Mt CO2e, the increase in carbon stocks in live vegetation stabilised at 26-36 Mt CO2e, while the increase in debris carbon stabilised at 17-23 Mt CO2e. Carbon stocks in soil continued to increase over time reaching 5-7 Mt CO2e by 2050.

60 e) 2 50

40 PHW 30 Total 20

10 Total C in live vegetation (Mt CO

0 2010 2020 2030 2040 2050 Year

Figure 44: Cumulative change in total carbon stocks in live vegetation for Study Scenario P2 relative to the Study Baseline.

Page 100 of 365

e) 25 2

20 PHW 15 Total

Total C in live debris (Mt CO 5

0 2010 2020 2030 2040 2050 Year

Figure 45: Cumulative change in total carbon stocks in debris for Study Scenario P2 relative to the Study Baseline.

6 e) 2 5

4 PHW

3 Total

2 Total C in live soil (Mt CO 1

0 2010 2020 2030 2040 2050 Year

Figure 46: Cumulative change in total carbon stocks in soil for Study Scenario P2 relative to the Study Baseline.

Page 101 of 365

The increase in carbon stocks in live vegetation and debris as a result of conversion of plantations from pulplog to sawlog regimes under Study Scenario P2 is consistent with results from a previous study of potential management options for carbon plantings in Tasmania (Beadle et al. 2011). This study showed that increasing rotation length of a hardwood plantation from 12 years to 24 years could increase total volume production by 67% boosting the rate of carbon sequestration above-ground by 3.6-7.8 t CO2e/ha/y. In comparison, for the Carbon Study, the average increase in the net rate of carbon sequestration in total vegetation and debris (above plus below ground) from 2010 to 2050 was equivalent to 6-8 t CO2e/ha/y when average rotation length was increased from around 14 to 27 years.

The conversion of plantations from pulpwood to sawlog production increased carbon stocks in harvested wood products by 2-3 Mt CO2e (Table 21). Initially, carbon stocks decreased slightly as a result of a reduction in the amount of paper produced from short rotation pulpwood plantations. However, by 2030, when the sawlog plantations reached harvestable age, wood product carbon stocks increased. This increase is a result of the longer half life of sawn timber and veneer compared with paper and the lower proportion of wood exported (and thus assumed to be converted to CO2) from sawlog plantations compared with pulpwood plantations. There was no change in emissions from fossil fuel use or non-CO2 emissions from regeneration burns.

The net increase in carbon stocks for this scenario was 50-70 Mt CO2e. Most of this increase occurred on private land as most publicly owned plantations are already managed for sawlog production.

Table 21: Estimated differences in carbon stocks (Mt CO2e) in plantations for Study Scenario P2 relative to the Study Baseline showing decadal changes in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 1 1 0 0 0 0 0 0 0 0 1 1 2030 1 1 0 0 0 0 0 0 0 0 1 1 2040 1 1 0 1 0 0 0 0 0 0 1 2 2050 1 1 1 1 0 0 0 0 0 0 2 2 Private 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 28 38 3 4 1 1 -1 -1 -1 0 31 43 2030 46 63 17 23 2 3 0 0 0 0 66 89 2040 16 21 22 30 4 6 3 4 0 0 45 61 2050 25 34 17 22 4 7 2 3 0 0 49 67 Total 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 28 39 3 4 1 1 -1 -1 -1 0 32 44 2030 47 64 17 23 2 3 0 0 0 0 67 91 2040 17 22 22 31 4 6 3 4 0 0 46 63 2050 26 36 17 23 5 7 2 3 0 0 50 70

Page 102 of 365

P3: Increasing plantation growth rate by 25%

Increasing growth of all softwood and hardwood plantations by 25% resulted in a moderate increase in modelled carbon stocks in live vegetation for the first 10 years relative to the Study Baseline (Figures 47, 48 & 49). However, the effect peaked in 2040 at 16-19 Mt CO2e and, by 2050, carbon stocks in vegetation appeared to have stabilised around 15-18 Mt CO2e above the Study Baseline (Table 22).

Carbon stocks in debris increased by a similar amount to those in vegetation as a result of increased input from the larger trees. By 2050, debris Carbon was estimated to reach 17-22 Mt CO2e. The increase in total carbon stocks in vegetation and debris was estimated to be 32-40 Mt CO2e by 2050.

Soil carbon increased after 2020, reaching 4-5 Mt CO2e by 2050. Carbon stocks in wood products increased by 2-3 Mt CO2e as a result of the combined effects of increased rate of wood production and larger tree size increasing the yield of sawlogs and thus sawn timber. Emissions from fossil fuel use and non-CO2 emissions from regeneration burns were 1 Mt CO2e above the Study Baseline.

20 e) 2

PSW 10 PHW

Total 5 Total C in live vegetation (Mt CO

0 2010 2020 2030 2040 2050 Year

Figure 47: Cumulative change in total carbon stocks in live vegetation for Study Scenario P3 relative to the Study Baseline.

Page 103 of 365

25 e) 2 20

15 PSW

PHW 10 Total

5 Total C in live debris (Mt CO

0 2010 2020 2030 2040 2050 Year

Figure 48: Cumulative change in total carbon stocks in debris for Study Scenario P3 relative to the Study Baseline.

4.5

e) 3.5 2 3 PSW 2.5 PHW 2 Total 1.5

1 Total C in live soil (Mt CO 0.5

0 2010 2020 2030 2040 2050 Year

Figure 49: Cumulative change in total carbon stocks in soil for Study Scenario P3 relative to the Study Baseline.

Page 104 of 365

Table 22: Differences in carbon stocks (Mt CO2e) in plantations for Study Scenario P3 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 4 5 2 3 0 0 0 0 0 0 6 8 2030 6 7 4 5 0 1 1 1 0 0 11 13 2040 8 9 6 7 1 1 1 1 0 0 15 18 2050 6 8 7 8 1 2 1 2 0 0 15 19 Private 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 7 9 4 5 0 0 0 0 0 0 12 15 2030 8 10 7 10 1 1 1 1 0 0 17 22 2040 8 10 9 12 2 2 1 1 0 1 19 25 2050 8 10 11 14 2 3 1 1 1 1 21 28 Total 2010 0 0 0 0 0 0 0 0 0 0 0 0 2020 11 13 6 8 0 1 1 1 0 0 18 23 2030 14 17 11 15 1 2 1 2 0 1 28 35 2040 16 19 15 19 3 3 2 2 1 1 34 43 2050 15 18 17 22 4 5 2 3 1 1 37 47

Summary of plantation management Study Scenarios

Changing plantation management and growth rates has the potential to increase carbon stocks by up to 90-130 Mt CO2e over 40 years (Table 23). This figure excludes any increases in carbon in soil or wood product pools. Increasing rotation length by 10% resulted in a net gain of 8-11 Mt CO2e, while improving growth rates increased carbon stocks by around 32-40 Mt CO2e. However, the largest increase was through converting existing hardwood pulpwood plantations to sawlog regimes, resulting in a potential increase in carbon stocks of 44-59 Mt CO2e relative to the Study Baseline. This increase was a result of the increase in assumed average rotation lengths from 13 years for pulpwood plantations to 28 years for sawlog plantations.

Table 23: Summary of estimated differences in carbon stocks (Mt CO2e) in live vegetation plus debris in plantations relative to the Study Baseline for alternative Study Scenarios P1-P3 (2010 to 2050). Scenario Tenure Area 2020 2030 2040 2050 000 ha Min Max Min Max Min Max Min Max P1 Public 120 4 5 4 4 3 4 3 4 Private 217 1 2 4 5 4 6 5 7 P2 Public 59 1 1 1 1 1 1 2 2 Private 191 31 42 63 85 38 51 42 57 P3 Public 120 6 7 10 12 13 16 13 16 Private 217 11 14 16 20 18 22 19 24

Page 105 of 365

3.6 CHANGING LAND USE

Alternative land uses considered in the Carbon Study included continued conversion of native forest to plantations and grassland until 2015, regenerating unstocked areas of native forest, expanding the area of plantations and reducing the area of plantations. These Study Scenarios provide indications of the potential to increase carbon stocks, or reduce emissions, though reforestation of existing cleared land or prevention of deforestation across Tasmania.

A1: Conversion of private native forest for plantation or non-forest uses will continue at recent rates

Study Scenario A1 assumes that private native forests will continue to be converted to plantations and grassland at the average rate experienced over the past 10 years until the allowable limit (~10,000 ha) is achieved by 2015 (when further conversion of native forest will cease). Under this scenario, 82% of converted forests were assumed to be reforested with hardwood or softwood plantations with the remaining 18% converted to grassland.

The results indicate that, while conversion of native forest results in an initial reduction in carbon stocks in live vegetation, the faster growth of new hardwood plantations more than makes up the difference by 2050, even though not all the land was replanted (Figure 50).

This increase was due to the combined effects of an increase in the rate of carbon sequestration in vegetation and reduced harvesting pressure on the remaining native forest as a result of the greater rate of sawlog production in hardwood plantations. Stocks of carbon in debris and soil fluctuated as a result of conversion of native forest to plantations and grassland but, by 2050, had stabilised with little net change over the period (Figures 51 & 52).

The total change in carbon stocks in vegetation, debris and soil for Study Scenario A1 relative to the Study Baseline, is estimated to range from -2 to -3 Mt CO2e for native forests and from 3 to 5 Mt CO2e for plantations. A breakdown of the difference in carbon stocks in vegetation and debris resulting from deforestation (i.e. forests converted permanently to grassland) is shown in Table 24. This indicates that total losses from deforestation are around -1 Mt CO2e or 370-450 t CO2e/ ha. However, where plantations are established on ex-native forest, stocks increased relative to the Study Baseline. Thus, avoiding permanent deforestation could increase total carbon stocks in native forests, but avoiding conversion of native forests to plantations may not.

Page 106 of 365

1.5 e) 2 1

0.5 PSW PHW 0 RFT -0.5 ETF

-1 ELF ONF -1.5 Total C in live vegetation (Mt CO Total -2 2010 2020 2030 2040 2050 Year

Figure 50: Cumulative change in total carbon stocks in vegetation for Study Scenario A1 relative to the Study Baseline. Solid black line indicates total net change in carbon stocks.

0.5

e) 0 2 PSW -0.5 PHW

-1 RFT ETF -1.5 ELF ONF

Total C in live debris (Mt CO -2 Total

-2.5 2010 2020 2030 2040 2050 Year

Figure 51: Cumulative change in total carbon stocks in debris for Study Scenario A1 relative to the Study Baseline.

Page 107 of 365

0 -0.2 -0.4 e) 2 PSW -0.6 PHW -0.8 RFT -1 ETF -1.2 ELF -1.4 ONF

Total C in live soil (Mt CO -1.6 Total -1.8 -2 2010 2020 2030 2040 2050 Year

Figure 52: Cumulative change in total carbon stocks in soil for Study Scenario A1 relative to the Study Baseline. Solid black line indicates total net change in carbon stocks.

Table 24: Estimated change in carbon stocks (Mt CO2e) in native forests cleared permanently for Study Scenario A1 relative to the Study Baseline, showing the total and per hectare changes in carbon stocks in live vegetation, debris and soil. Year Live vegetation Debris Soil Total Live+Dead Min Max Min Max Min Max Min Max Total (Mt CO2e) 2020 -0.5 -0.6 -0.3 -0.4 0.0 0.0 -0.8 -1.0 2030 -0.5 -0.6 -0.4 -0.5 0.0 0.0 -0.9 -1.1 2040 -0.6 -0.7 -0.4 -0.5 0.0 0.0 -1.0 -1.2 2050 -0.6 -0.7 -0.4 -0.5 0.0 0.0 -1.0 -1.3 Per ha change (t CO2e/ha) 2020 -171 -209 -115 -141 5 6 -286 -350 2030 -187 -229 -139 -170 10 12 -326 -399 2040 -202 -247 -150 -183 9 11 -352 -430 2050 -210 -257 -158 -193 0 0 -368 -450

Page 108 of 365

A2: Existing unstocked areas in native forests are regenerated back to native forest

Around 41,000 ha of native forest were identified as being unstocked. These areas are most likely associated with failed regeneration or repeated burning and generally carry existing scrub or non-eucalypt forest that has some carbon storage capacity. Data from Forestry Tasmania indicated that some of these Stands could have significant amounts of biomass. Thus this vegetation was assumed to have half the carbon content of a eucalypt forest of a similar productivity and age.

The results of the modelling indicated that reforestation of these areas could be expected to result in a net reduction in carbon stocks. Carbon stocks would likely fall in the short-term as a result of removal of existing vegetation in unstocked areas, but could rise in the longer term as a result of faster growth of the eucalypt forest (Figure 53). By 50 years, carbon in live vegetation was estimated to be 0.8-0.9 Mt CO2e above the Study Baseline (Table 25). However, debris Carbon increased more slowly and, by 2050, was still around 1 Mt CO2e less than the initial carbon stocks (Figure 54). Carbon stocks in soil decreased by 0.4-0.5 Mt CO2e (Figure 55).

The net effect of reforestation of unstocked land was to reduce carbon stocks by 0.3-0.8 Mt CO2e over 40 years (Table 26). However, over the longer term (~50-60 years) carbon stocks are expected to increase above those of the original vegetation and debris. Potential areas that could be replanted are mapped in Figure 56.

1.5 1 e) 2 0.5

0 ETF -0.52010 2020 2030 2040 2050 -1 ELF

-1.5 ONF -2 -2.5 Total -3

Total C in live vegetation (Mt CO -3.5 -4

Year

Figure 53: Cumulative change in total carbon stocks in vegetation for Study Scenario A2 relative to the Study Baseline. Solid black line indicates total net change in carbon stocks.

Page 109 of 365

0 2010 2020 2030 2040 2050 -0.5 e) 2 -1

-1.5 ETF

-2 ELF

-2.5 ONF

-3 Total

-3.5 Total C in live debris (Mt CO -4

-4.5 Year

Figure 54: Cumulative change in total carbon stocks in debris for Study Scenario A2 relative to the Study Baseline. Solid black line indicates total net change in carbon stocks.

0 2010 2020 2030 2040 2050 -0.5 e) 2 -1

-1.5 ETF

-2 ELF

-2.5 ONF

-3 Total Total C in live soil (Mt CO -3.5

-4 Year

Figure 55: Cumulative change in total carbon stocks in vegetation for Study Scenario A2 relative to the Study Baseline. Solid black line indicates total net change in carbon stocks.

Page 110 of 365

Table 25: Differences in carbon stocks (Mt CO2e) in native forests for Study Scenario A2 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050. Fossil fuel and non- Tenure Year Live vegetation Debris Soil Wood products Total CO2 emissions Min Max Min Max Min Max Min Max Min Max Min Max Public 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 -0.8 -0.6 -0.5 -0.3 -0.1 0.0 0.0 0.0 0.0 0.1 -1.3 -1.0 2030 -0.8 -0.6 -0.6 -0.5 -0.1 -0.1 0.0 0.0 0.2 0.2 -1.7 -1.3 2040 -0.4 -0.3 -0.6 -0.5 -0.1 -0.1 0.0 0.0 0.2 0.2 -1.4 -1.1 2050 0.3 0.3 -0.5 -0.4 -0.2 -0.1 0.0 0.0 0.1 0.2 -0.5 -0.3 Private 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 -1.4 -1.1 -0.8 -0.5 -0.1 -0.1 0.0 0.0 0.0 0.1 -2.3 -1.7 2030 -1.3 -1.0 -1.3 -1.0 -0.3 -0.2 0.0 0.0 0.1 0.1 -2.9 -2.2 2040 -0.5 -0.4 -1.1 -0.8 -0.4 -0.2 0.0 0.0 0.1 0.1 -2.0 -1.5 2050 0.5 0.6 -0.6 -0.5 -0.4 -0.3 0.0 0.0 0.1 0.1 -0.6 -0.2 Total 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 -2.2 -1.7 -1.2 -0.9 -0.2 -0.1 0.0 0.0 0.1 0.1 -3.7 -2.7 2030 -2.0 -1.6 -1.9 -1.4 -0.4 -0.2 0.0 0.0 0.2 0.3 -4.6 -3.5 2040 -0.9 -0.8 -1.7 -1.3 -0.5 -0.3 0.0 0.0 0.2 0.3 -3.4 -2.6 2050 0.8 0.9 -1.1 -0.8 -0.5 -0.4 0.0 0.0 0.2 0.3 -1.1 -0.6

A3 & A4: Expansion of plantation estate on cleared land

Study Scenarios A3 and A4 consider the effect on carbon stocks of continued expansion of plantations on cleared land. The average rate of conversion from 2000 to 2010 was around 0.5% per year (~5,000 ha/year, FPA 2011). For A3, it was assumed that planting of cleared land would continue at this rate, while for Study Scenario A4 it was assumed that the rate of planting would double to 1% per year (ie.10,000 ha per year). An upper limit of 10% of the current area of current cleared land was set for both Study Scenarios.

The results indicated that continued planting at either the historic rate, or the faster rate could be expected to sequester an additional 12-17 Mt CO2e Mt CO2e in live vegetation relative to the Study Baseline (Figure 57). The main difference between the Study Scenarios being that the maximum carbon stock in live vegetation was achieved earlier for A4 than for A3.

Changes in carbon stocks in debris showed a similar pattern for the two Study Scenarios reaching 11-15 Mt CO2e for A3 and 13-18 Mt CO2e for A4 by 2050 (Figure 58). Soil carbon stocks initially decreased for both Study Scenarios as a result of the reduced rates of input of dead material from the young stands compared with the previous grassland (Figure 59). This result is consistent with findings from Polglase et al. (2000) and Paul et al. (2003), indicating that soil carbon stocks usually decline in the first 10 years after afforestation of cleared land before subsequently recovering. The average loss in the first 10 years reported by Polglase et al. (2000) was equivalent to 5 t CO2e/ha which was close to our estimate of 7-9 t CO2e/ha. By 2050, soil carbon had largely recovered for Study Scenario A3 and was positive for Study Scenario A4 as a result of the elevated rate of debris production in the older stands.

Page 111 of 365

Figure 56: Cumulative changes in carbon stocks in live vegetation for Study Scenario A2 relative to the Study Baseline.

Page 112 of 365

a) b) 18 18 e) 16 e) 16 2 2 14 14

12 12 PSW 10 10 PSW PHW 8 8 PHW 6 6 Total Total 4 4 Total C in live vegetation (Mt CO 2 Total C in live vegetation (Mt CO 2

0 0 2010 2020 2030 2040 2050 2010 2020 2030 2040 2050 Year Year Figure 57: Cumulative change in total carbon stocks in live vegetation for Study Scenario A3 relative to the Study Baseline. a) b) 14 18

12 16 e) e) 2 2 14 10 12 PSW 8 10 PSW PHW 6 8 PHW

Total 6 4 Total 4 Total C in live debris (Mt CO 2 Total C in live debris (Mt CO 2

0 0 2010 2020 2030 2040 2050 2010 2020 2030 2040 2050 Year Year Figure 58: Cumulative change in total carbon stocks in debris for Study Scenarios A3 (a) and A4 (b) relative to the Study Baseline. a) b) 0 0.4 2010 2020 2030 2040 2050 -0.1 0.2 e) e) 2 2 -0.2 0 2010 2020 2030 2040 2050 -0.3 PSW PSW -0.2 -0.4 PHW PHW -0.4 -0.5 Total Total -0.6 -0.6 Total C in live soil (Mt CO Total C in live soil (Mt CO -0.7 -0.8

-0.8 -1 Year Year

Figure 59: Cumulative change in total carbon stocks in soils for Study Scenarios A3 (a) and A4 (b) relative to the Study Baseline.

Page 113 of 365

A comparison of the total net change in carbon stocks for the two scenarios indicates that Study Scenario A4 is likely to sequester carbon more quickly than A3 (Table 26). However, by 2050, there was only a small difference in the total carbon gain s of the two scenarios because of the assumption that only 107,000 ha could be planted. There was little change in carbon stocks in wood products for either of the scenarios with a cumulative 0.5-0.9 Mt CO2e increase balanced by a similar increase in cumulative emissions from fossil fuel use and regeneration burns.

The distribution of new plantings under Study Scenario A3 is shown in Figure 60. As a result of the random nature of the spatial modelling, these tend to be scattered across cleared agricultural land. In reality, they could be expected to be clumped to reduce management costs. Nonetheless, these results provide an indication of the variation in carbon stocks across the landscape associated with reforestation.

Table 26: Estimated differences in carbon stocks (Mt CO2e) in plantations for Study Scenarios A3 and A4 relative to the Study Baseline showing cumulative changes in carbon stocks in vegetation, debris, soil and wood products and cumulative non-CO2 and fossil fuel GHG emissions for the period 2010-2050.

Area Fossil fuel and Scenario Year Live vegetation Debris Soil Wood products Total planted non-CO2 000 ha Min Max Min Max Min Max Min Max Min Max Min Max A3 2020 5 2 3 0 1 -0.5 -0.4 0.0 0.0 0.0 0.0 2 3 2030 50 7 9 4 5 -0.7 -0.7 0.3 0.4 0.2 0.2 10 14 2040 95 13 17 8 11 -0.7 -0.7 0.5 0.7 0.5 0.6 21 28 2050 107 12 17 11 15 -0.3 -0.2 0.7 0.9 0.7 0.9 23 32 A4 2020 10 5 7 1 1 -0.8 -0.8 0.0 0.0 0.0 0.0 5 7 2030 99 11 16 7 10 -0.8 -0.7 0.5 0.7 0.3 0.5 17 25 2040 107 13 19 10 15 -0.4 -0.3 0.5 0.8 0.6 0.9 22 34 2050 107 13 16 13 18 0.2 0.2 0.8 1.1 0.9 1.3 25 35

Page 114 of 365

Figure 60: Cumulative changes in carbon stocks in live vegetation for Study Scenario A3 relative to the Study Baseline, showing areas of grassland (white) planted with trees (coloured dots).

Page 115 of 365

A5: 25% of existing pulpwood plantations not replanted after 1st harvest

Study Scenario A5 considered the potential impact of converting 25% of hardwood plantations to pasture following harvest. This could be a result of poorer than expected growth economic uncertainty or unfavourable market conditions. It is considered a potentially real risk as a result of the recent collapse of a number of MIS (Managed Investment) schemes (Chris Lafferty, FWPA, Pers. Comm.).

The results indicate that removal of pulpwood plantations could have a major negative impact on carbon stocks. An estimated 5-7 Mt CO2e could be lost from vegetation and 6-8 Mt CO2e from debris with only a small potential increase (0.1-0.2 Mt CO2e) in soil carbon (Figures 61, 62 & 63). The total reduction in carbon stocks could be 12-15 Mt CO2e (Table 27).

Measures to discourage conversion of existing plantations to grassland could limit this potential reduction in forest carbon stocks and therefore reduce potential emissions. In contrast, removal of existing plantations would reduce total forest carbon stocks and increase the net carbon emissions of Tasmania.

0 2010 2020 2030 2040 2050 e) 2 -1

-2

-3 PHW -4 Total -5

-6 Total C in live vegetation (Mt CO

-7

Year

Figure 61: Cumulative change in total carbon stocks in live vegetation for Study Scenario A5 relative to the Study Baseline.

Page 116 of 365

0 2010 2020 2030 2040 2050 -1 e) 2 -2

-3 PHW -4

-5 Total -6 Total C in live debris (Mt CO -7

-8 Year

Figure 62: Cumulative change in total carbon stocks in debris for Study Scenario A5 relative to the Study Baseline.

0.5 0.45 0.4 e) 2 0.35 0.3 PHW 0.25 0.2 Total 0.15 0.1 Total C in live soil (Mt CO 0.05 0 2010 2020 2030 2040 2050 Year

Figure 63: Cumulative change in total carbon stocks in soil for Study Scenario A5 relative to the Study Baseline.

Page 117 of 365

Table 27: Estimated differences in carbon stocks (Mt CO2e) in plantations for Study Scenario A5 relative to the Study Baseline showing decadal changes in vegetation, debris, soil, wood products and cumulative non-CO2 and fossil fuel GHG emissions across different tenures for the period 2010-2050.

Tenure and Fossil fuel and Live vegetation Debris Soil Wood products Total Year non-CO2 Min Max Min Max Min Max Min Max Min Max Min Max Total 2010 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2020 -1.1 -0.9 -2.2 -1.8 0.2 0.2 -0.2 -0.2 -0.1 -0.1 -3.3 -2.5 2030 -5.2 -4.0 -6.2 -4.8 0.4 0.5 -0.4 -0.3 -0.2 -0.2 -11.2 -8.5 2040 -7.1 -5.5 -7.4 -5.8 0.3 0.3 -0.4 -0.3 -0.2 -0.1 -14.4 -11.0 2050 -6.8 -5.4 -8.2 -6.4 0.1 0.2 -0.4 -0.3 -0.1 -0.1 -15.2 -11.8

Summary of results for alternative land use Study Scenarios

Change in landuse could have either positive, or negative, effects on Tasmania’s carbon stocks. Planting up to 10% of cleared agricultural land could increase total carbon stocks in live vegetation and debris by 23 and 34 Mt CO2e (Table 28, Figure 64). However, removal of existing plantations could reduce carbon stocks by 11-15 Mt CO2e. Conversion of existing native forest ,or replanting areas that are currently unstocked, is expected to have a minor negative impact as a result of the relatively small areas involved and offsetting reductions on potential forest harvesting rates (for A1) and carbon stocks in existing vegetation (A2).

Table 28: Summary of estimated cumulative differences in carbon stocks (Mt CO2e) in live vegetation, debris and soil relative to the Study Baseline for alternative Study Scenarios A1-A5 (2010 to 2050).

Scenario Forest type and Live vegetation Debris Soil Total tenure Min Max Min Max Min Max Min Max A1 Plantation 1 2 1 1 1 2 3 5 Native forest 1 1 -1 -1 -2 -1 -3 -2 Total 2 3 -1 0 -1 0 0 3 A2 Public 0 0 0 0 0 0 0 0 Private 0 1 -1 0 0 0 0 0 Total 1 1 -1 -1 -1 0 -1 0 A3 Private plantation 12 17 11 15 0 0 23 32 A4 Private plantation 13 16 13 18 0 0 26 34 A5 Private plantation -7 -5 -8 -6 0 0 -15 -12

Page 118 of 365

30 e) 2 20

10 Vegetation Debris 0 A1 A2 A3 A4 A5

to Baseline (Mt CO -10 Change in carbon stocks relative relative stocks carbon in Change -20 Scenario

Figure 64: Summary of cumulative changes differences in carbon stocks in vegetation and debris for the Landuse Change Study Scenarios relative to the Study Baseline between 2010 and 2050.

3.7 BASELINE SENSITIVITY SCENARIOS: FIRE REGIMES

The potential effect of varying fire frequency on carbon stocks in native forests was investigated in a series of Baseline Sensitivity Scenarios (F1-F4). In these scenarios, the prevailing average fire frequency for each forest type was varied by up to +/- 20%. All native forests were included in the analyses. Non-forest vegetation > 2 m was excluded due to inadequate data on growth, or fire regimes. Plantations were also excluded.

The results of these analyses show that, as expected, increasing fire frequency tends to reduce carbon stocks. The effect was greatest for live vegetation, in which carbon pools decreased by 29-37 Mt CO2e for a 20% increase in fire frequency, and increased by 28-35 Mt CO2e for a 20% reduction in frequency (Table 29). Debris pools were the next most affected, while soil pools were only marginally affected by varying fire frequency.

The change in carbon stocks in vegetation and debris was linearly related to fire frequency for all forest types. Increasing or decreasing frequency by 20% resulted in double the change in carbon stocks relative to the Study Baseline compared with a 10% change in fire frequency (Figure 65).

Page 119 of 365

Table 29: Change in carbon stocks (Mt CO2e) live vegetation, debris and soil associated with changes in average fire frequency for the period 2010-2050. F1: 10% decrease, F2: 10% increase, F3: 20% decrease, F4 20% increase. Scenario Year Live vegetation Debris Soil Total Min Max Min Max Min Max Min Max F1 2020 4 5 2 2 0 0 6 7 2030 8 10 4 5 1 1 12 16 2040 11 14 6 8 1 2 18 23 2050 14 18 8 11 2 2 23 31 F2 2020 -5 -4 -2 -1 0 0 -7 -6 2030 -10 -8 -4 -3 -1 -1 -15 -12 2040 -14 -11 -7 -5 -2 -1 -23 -17 2050 -20 -15 -11 -8 -3 -2 -33 -25 F3 2020 8 10 3 4 0 1 11 15 2030 16 20 8 10 1 2 25 32 2040 22 28 12 16 2 3 36 48 2050 28 35 17 23 3 5 48 63 F4 2020 -10 -8 -3 -2 -1 0 -13 -10 2030 -19 -15 -8 -6 -2 -1 -29 -22 2040 -28 -22 -14 -10 -3 -2 -45 -34 2050 -37 -29 -20 -15 -5 -3 -62 -47 The change in carbon stocks in vegetation increased almost linearly with time for the different Baseline Sensitivity Scenarios (Figure 66). This indicates that carbon stocks will continue to change for well over 40 years after a new fire regime is in place. Change in debris carbon over time also increased over time, indicating that a small relative change in fire frequency is likely to have a large, long-term effect on debris pools (Figure 67). The effects of changes in fire frequency were similar for all forest types modelled.

Other studies that have investigated the effect of potential changes in fire regimes have reported similar results to those presented here. Grove et al. (2011) showed that Coarse Woody Debris mass in high productivity Eucalyptus obliqua forests could be expected to increase with a halving of fire frequency from 100 years to 200 years. Equilibrium mass of Coarse Woody Debris for the 100 year scenario was around 210 t/ha (equivalent to 385 t CO2e/ha) compared with ~290 t/ha (equivalent to 530 t CO2e/ha). However, modelled debris mass decreased when fire frequency was extended beyond 200 years.

Page 120 of 365

80 Total 60 Vegetation Debris e) 2 40 Soil

0 -20% -10% 0 10% 20%

Study Baseline (Mt CO (Mt Study Baseline -20

-40 Change in forest carbon stocks relative to relative stocks carbon forest in Change

-60 Change in fire frequency (%)

Figure 65: Relationship between changes in fire frequency and changes in carbon stocks in vegetation and debris by 2050 relative to the Study Baseline. Error bars indicate standard deviations of means. a) c) 35 35 e) e) 2 30 2 30

25 PSW 25 PSW PHW PHW 20 20 RFT RFT 15 ETF 15 ETF

10 ELF 10 ELF ONF ONF 5 5 Total C in live vegetation (Mt CO Total Total C in live vegetation (Mt CO Total 0 0 2010 2020 2030 2040 2050 2010 2020 2030 2040 2050 Year Year b) d) 0 0 2010 2020 2030 2040 2050 2010 2020 2030 2040 2050 e) e) 2 -5 2 -5

-10 PSW -10 PSW PHW PHW -15 -15 RFT RFT -20 ETF -20 ETF

-25 ELF -25 ELF ONF ONF -30 -30 Total C in live vegetation (Mt CO Total Total C in live vegetation (Mt CO Total -35 -35

Year Year Figure 66: Cumulative change in carbon stocks in vegetation for the fire related Baseline Sensitivity Scenarios F1 (a), F2 (b), F3 (c), F4 (d), relative to the Study Baseline.

Page 121 of 365

a) b) 25 25 e) e) 2 20 2 20 PSW PSW

15 PHW 15 PHW RFT RFT

10 ETF 10 ETF ELF ELF

5 ONF 5 ONF Total C in live debris (Mt CO Total C in live debris (Mt CO Total Total

0 0 2010 2020 2030 2040 2050 2010 2020 2030 2040 2050 Year Year b) d) 0 0 2010 2020 2030 2040 2050 2010 2020 2030 2040 2050 e) e) 2 2 -5 -5 PSW PSW PHW -10 PHW -10 RFT RFT

-15 ETF -15 ETF ELF ELF

-20 ONF -20 ONF Total C in live debris (Mt CO Total C in live debris (Mt CO Total Total

-25 -25 Year Year

Figure 67: Cumulative change in carbon stocks in debris for the fire related Baseline Sensitivity Scenarios F1 (a), F2 (b), F3 (c), F4 (d), relative to the Study Baseline.

In another study, King et al. (2011) predicted that carbon stocks and sequestration rates in dry eucalypt forest near Canberra could be expected to be lower as a result of increased fire frequency and intensity under projected climate change scenarios (SRES B1 and A1F1, IPCC 2000) compared with the current climate. These results are consistent with those from the Baseline Sensitivity Scenarios presented here which indicate that carbon stocks may decrease, or at least increase at a slower rate if fire frequency increases substantially as a result of climate change.

Page 122 of 365

3.8 BASELINE SENSITIVITY SCENARIOS: CLIMATE CHANGE

The potential impact of climate change on carbon stocks in Tasmania’s forests assessed three climate related Baseline Sensitivity Scenarios C1, C2 and C3 which reference climate projections based on two SRES emissions scenarios. These were:

• C1: similar rainfall, warmer climate projection from the CSIRO Mk 3.0 model, based on the SRES A2 emissions scenario, • C2: drier, hotter climate projection from the CSIRO Mk 3.5 model, based on the SRES A2 emissions scenario, and • C3: drier, warmer climate projection from the CSIRO Mk 3.5 model, based on the emissions scenario SRES A1b.

The effects of these projections on growth rates of Tasmania plantations and native forests have been modelled by ABARES (2012b) using data from the climate models downscaled to a resolution of 25 x 25 km by CSIRO. The average percentage change in Mean Annual Increment for hardwood and softwood plantations and native forests estimated by ABARES for each projection were used to modify the tree growth in the FCMF. Associated climate data sets, used by ABARES, were provided to CO2 Australia by CSIRO and were incorporated in the FCMF to assess the impact on carbon pools in litter and soil. In addition to the original climate data sets used by ABARES, which included average temperature and rainfall, CSIRO developed an evaporation dataset (required for the Roth-C soil carbon model) specifically for the Carbon Study.

The three climate projections used in the Carbon Study are shown together with their effect on average temperatures and rainfall in Table 30. The average percentage changes in growth rates for different forest types estimated by ABARES (2012b) based on these projections are shown in Table 31. In ABARES modelling, the relative change in growth rates of eucalypt sawlog plantations were used as a proxy for modelling growth of native forests. This approach was adopted for the Carbon Study.

Table 30: Climate models used by ABARES (2012b) to model impacts of climate change on growth rates and in the Carbon Study to model the effects of climate change on debris and soil.

Scenario Climate Model Name SRES Projected Climate Scenario Max Temp Rainfall oC oC mm mm 2030 2050 2030 2050 C1 BOM-CSIRO_MK3_0.SRES_A2 A1b 16 16 1133 1066 C2 BOM-CSIRO_MK3_5.SRES_A2 A1b 16 17 723 1128 C3 BOM-CSIRO_MK3_5.SRES_A1B A2 16 16 1105 1030

Page 123 of 365

Table 31: Estimated changes in growth rates for hardwood and softwood plantations in 2030 and 2050 from ABARES (2012b) used to model the effect of climate change projections on average forest growth rates in the Carbon Study. Parameter Bluegum pulpwood Bluegum sawlog P. radiata 2030 2050 Avg 2030 2050 Avg 2030 2050 Avg Change in MAI (%) C1 25% 16% 20% 14% 17% 15% 3% 1% 2% C2 9% 1% 5% 2% 1% 1% -4% -9% -7% C3 -6% 9% 2% -5% -3% -4% -7% -12% -10% The results of the climate related Baseline Sensitivity Scenarios indicate that climate change can be expected to have a major impact on carbon pools in vegetation, debris and soil. The cumulative change in carbon stocks in vegetation relative to the Study Baseline varied from a 187-240 Mt CO2e increase for C1 to a 34-44 Mt CO2e decrease for C3 (Figure 68, Table 32). Changes in carbon stocks in debris followed the same pattern ranging from a 92-123 Mt CO2e increase for C1 to a 46-62 Mt CO2e decrease for C3 (Figure 69).

In contrast, soil carbon stocks followed a different trend for the three Baseline Sensitivity Scenarios, decreasing by 5-7 Mt CO2e for C2 and increasing by 12 to 17 Mt CO2e for C3 (Figure 70).

The differing effects between the climate related Baseline Sensitivity Scenarios on carbon stocks in soil compared with those in vegetation and debris is a result of the effect of changes in rainfall and temperature on decomposition rates. The Baseline Sensitivity Scenario that increased stand growth rates the most (C1), increased in the rate of decomposition and turnover of soil carbon pools, decreasing soil carbon stocks. In contrast, the climate projection that decreased growth, as a result of lower rainfall, also decreased decomposition rates, elevating soil carbon stocks. This change was greater than the effect of increased inputs from debris resulting from the more rapid growth associated with this scenario.

The reduction in soil carbon associated with C1 is consistent with results from Jenkinson et al. (1991) and Dean and Wardell-Johnson (2010) who showed that soil carbon stocks could be expected to decrease in response to higher temperatures, provided that the annual input of plant debris to soil was kept constant. However, the increase in soil carbon in C3 in spite of the reduction in plant debris indicates that the climate projections for Tasmania for this Baseline Sensitivity Scenario result in a lower rate of decomposition and soil carbon turnover than those predicted from these previous studies.

Page 124 of 365

a) 250 e) 2 200 PSW

150 PHW RFT

100 ETF ELF

50 ONF

Total C in live vegetation (Mt CO Total 0 2010 2020 2030 2040 2050 Year

b) 70 e) 2 60

50 PSW PHW 40 RFT 30 ETF

20 ELF ONF 10 Total C in live vegetation (Mt CO Total 0 2010 2020 2030 2040 2050 Year

c) 0 2010 2020 2030 2040 2050

e) -5 2 -10 PSW -15 PHW -20 RFT -25 ETF -30 ELF -35 ONF

Total C in live vegetation (Mt CO -40 Total -45

Year

Figure 68: Cumulative change in vegetation carbon stocks in for the three climate related Baseline Sensitivity Scenarios: C1 (a), C2 (b) and C3 (c).

Page 125 of 365

a) 140

120 e) 2

100 PSW PHW 80 RFT 60 ETF ELF 40 ONF Total C in live debris (Mt CO 20 Total

0 2010 2020 2030 2040 2050 Year

b) 0 2010 2020 2030 2040 2050

e) -5 2 PSW -10 PHW

-15 RFT ETF -20 ELF ONF

Total C in live debris (Mt CO -25 Total

-30 Year

c) 0 2010 2020 2030 2040 2050 -10 e) 2

-20 PSW PHW -30 RFT -40 ETF ELF -50 ONF Total C in live debris (Mt CO -60 Total

-70 Year

Figure 69: Cumulative change in debris carbon stocks for the three climate related Baseline Sensitivity Scenarios: C1 (a), C2 (b) and C3 (c).

Page 126 of 365

a) 16 14 12 e) 2 PSW 10 PHW 8 RFT 6 ETF 4 ELF 2 ONF

Total C in live soil (Mt CO 0 2010 2020 2030 2040 2050 Total -2 -4 Year

b) 10

5 e) 2 PSW 0 PHW 2010 2020 2030 2040 2050 RFT -5 ETF -10 ELF ONF Total C in live soil (Mt CO -15 Total

-20 Year

c) 25

20 e) 2 PSW PHW 15 RFT ETF 10 ELF ONF

Total C in live soil (Mt CO 5 Total

0 2010 2020 2030 2040 2050 Year

Figure 70: Cumulative change in soil carbon stocks for the three climate related Baseline Sensitivity Scenarios: C1 (a), C2 (b) and C3 (c).

Page 127 of 365

Table 32: Changes in carbon stocks in vegetation, debris and soil over time relative to the Study Baseline for the three climate related Baseline Sensitivity Scenarios C1, C2 and C3. Scenario Year Live vegetation Debris Soil Total Min Max Min Max Min Max Min Max C1 2020 49 62 17 22 0 0 66 84 2030 100 127 48 63 -3 -2 145 188 2040 142 182 75 98 1 1 218 282 2050 187 240 92 123 7 11 287 374 C2 2020 15 19 -13 -10 3 4 4 13 2030 29 36 -23 -18 -8 -5 -3 13 2040 41 53 -29 -22 -17 -12 -5 19 2050 53 68 -25 -19 -7 -5 21 44 C3 2020 -11 -9 -12 -9 5 7 -19 -12 2030 -23 -18 -23 -18 9 14 -36 -22 2040 -33 -26 -44 -33 14 21 -63 -38 2050 -44 -34 -62 -46 12 17 -94 -64

3.9 CARBON CARRYING CAPACITY

Carbon Carrying Capacity refers to the potential amount of carbon that can be contained in a natural ecosystem subject to natural disturbance but with the exclusion of human interference. The potential Carbon Carrying Capacity of Tasmania’s native forests was estimated by modelling their growth over of 240 years from 2010 to 2250 in the absence of timber harvesting and land clearing or conversion. All native forest types and vegetation > 2 m across public and private land were included. Three fire regimes were included:

• L1: a Carbon Carrying Capacity Baseline in which fire frequency was assumed to be the same as that used in the Study Baseline; • L3: a Carbon Carrying Capacity Sensitivity Scenario in which there was assumed to be a 20% increase in fire frequency across all forest types; and • L4: a Carbon Carrying Capacity Sensitivity Scenario in which there was assumed to be a 20% decrease in fire frequency across all forest types.

L1: Carbon Carrying Capacity under current fire regimes The Carbon Carrying Capacity under current fire regimes was estimated to be 1,700-2,000 Mt CO2e for live vegetation, 1,100-1,800 Mt CO2e for debris and 700-1000 Mt CO2e for soil (Table 33). The total carbon content of vegetation and debris was 2,900-3,900 Mt CO2e, representing an increase of 300 to 800 Mt CO2e over current carbon stocks.

Page 128 of 365

Table 33: Estimated total carbon stocks in vegetation, debris and soil for all Tasmanian public and private native forests from 2010 to 2200 in the absence of timber harvesting and land conversion.

Soil C excluding Total excluding Tenure Year Live vegetation Debris Soil Inert C Inert C Inert soil C Min Max Min Max Min Max Min Max Min Max Public 2010 1109 1285 769 984 405 526 36 54 2283 2796 2050 1183 1402 841 1104 363 634 45 68 2387 3139 2100 1241 1469 854 1213 393 675 57 85 2487 3357 2150 1280 1510 823 1296 420 706 69 103 2523 3512 2200 1295 1522 802 1363 443 727 81 121 2540 3613 2250 1289 1514 773 1401 460 738 93 139 2522 3653 Private 2010 399 462 269 345 145 188 12 19 813 995 2050 426 493 306 391 163 211 17 26 894 1096 2100 445 517 324 408 178 231 24 36 947 1157 2150 454 528 330 416 190 246 30 46 974 1191 2200 460 538 337 427 200 257 37 55 997 1222 2250 458 535 335 429 206 264 44 65 999 1228 Total 2010 1508 1747 1039 1329 550 715 48 73 3096 3791 2050 1609 1896 1147 1495 526 845 62 94 3281 4236 2100 1685 1987 1178 1621 571 907 80 121 3435 4514 2150 1734 2039 1153 1713 610 952 99 148 3497 4703 2200 1755 2060 1139 1790 643 984 118 176 3537 4834 2250 1747 2049 1108 1830 666 1001 136 205 3522 4881

The trends in carbon and debris indicate that carbon contents appear to stabilize after year 2200 (Figure 71). Soil inert carbon (i.e. charcoal) continued to increase throughout the time period. However, the dynamics of this carbon pool are highly uncertain. In FullCAM and the current version of the FCMF, inert soil carbon is assumed not to decompose and thus accumulates forever as a result of inputs from fire. This is unlikely to be the case, so, in the absence of any data, this carbon pool was excluded for the purpose of estimating Carbon Carrying Capacity. Most of the increase in carbon pools over time occurred in Tall Eucalypt Forests. Carbon stocks in live vegetation and debris for this forest type increased by 24% and 35% respectively above estimated current amounts. There was also a 30% increase in carbon stocks in Other Native Forest. However, this forest type comprised only 5% of total forest carbon across Tasmania so had a smaller impact on total carbon stocks.

Page 129 of 365

a) 2500 e) 2 2000

1500 RFT ETF 1000 ELF ONF 500 Total C in live vegetation (Mt CO 0 2010 2050 2100 2150 2200 2250 Year

b) 1600

1400

e) 1200 2

1000 RFT 800 ETF 600 ELF

400 ONF

Total C in debris (Mt CO 200

0 2010 2050 2100 2150 2200 2250 Year

c) 900 e

2 800 700 600 500 RFT 400 ETF ELF 300 ONF 200

Soil C excludning C excludning Soil InertC Mt CO 100 0 2010 2050 2100 2150 2200 2250 Year

Figure 71: Estimated long term trends for total carbon stocks in vegetation (a), debris (b) and soil (c) for Tasmanian native forests under current fire regimes in the absence of timber harvesting, land clearing or conversion to plantations.

Page 130 of 365

L2 and L3: Carbon Carrying Capacity Sensitivity Scenarios: Fire Regimes

Varying the frequency of fire across all forest types by + or –20% had only a small relative impact on total carbon stocks in vegetation, debris and soil. However, it had a relatively large impact on the estimated changes in total carbon stocks over time (Figure 72).

A 20% reduction in fire frequency, compared with the assumed current fire regimes in different forest types, resulted in a 3% increase in vegetation, 4% increase in debris and a 5% increase in soil carbon stocks in 2250 relative to the Carbon Carrying Capacity Baseline (L1). Likewise, a 20% increase in fire frequency resulted in a 3% decrease in vegetation, a 4% increase in debris and a 5% decrease in soil carbon stocks relative to the Carbon Carrying Capacity Baseline.

In terms of the relative increase in carbon stocks over time compared with current amounts, a 20% reduction in fire frequency resulted in a 19% increase in the growth of vegetation carbon stocks (between 2010 and 2250), and a 20% increase in the change in debris and soil carbon stocks relative to the Carbon Carrying Capacity Baseline (Table 34). Conversely, a 20% increase in fire frequency reduced the growth in vegetation carbon by 23%, and reduced the change in debris and soil carbon by 20%. Importantly, neither fire scenario resulted in a decrease in carbon stocks compared with current estimates (Figure 73).

Table 34: Estimated differences in total carbon stocks in vegetation, debris and soil for all Tasmanian public and private native forests from 2010 to 2250 in the absence of timber harvesting and land conversion, with a 20% decrease (L3) or increase (L4) in fire frequency relative to the Carbon Carrying Capacity Baseline (L1). Soil excluding Scenario Year Live vegetation Debris Soil inert C Total inert C Min Max Min Max Min Max Min Max Min Max L3 2050 32 38 20 27 7 9 -4 -3 60 73 2100 46 54 37 50 17 20 -8 -6 99 125 2150 54 63 42 62 25 31 -13 -9 121 155 2200 50 59 43 66 31 38 -17 -11 125 163 2250 46 54 45 71 35 43 -21 -14 126 169 L4 2050 -37 -31 -24 -18 -7 -7 3 3 -68 -57 2100 -53 -45 -50 -37 -18 -18 6 7 -121 -100 2150 -68 -58 -61 -42 -28 -28 9 11 -157 -127 2200 -64 -55 -66 -43 -35 -35 12 15 -165 -132 2250 -67 -57 -70 -43 -40 -40 16 19 -176 -140

Page 131 of 365

a) 70 0 2010 2050 2100 2150 2200 2250 e) e) 2 60 2 -10

50 -20 PSW 40 RFT -30 PHW ETF 30 -40 RFT ELF ETF 20 -50 ONF ELF 10 -60 ONF Total C in live vegetation (Mt CO Total C in live vegetation (Mt CO 0 -70 2010 2050 2100 2150 2200 2250 Year Year b) 70 0 2010 2050 2100 2150 2200 2250 60 -10 e) e) 2 2 50 -20 PSW 40 RFT PHW -30 ETF RFT 30 ETF ELF -40 20 ONF ELF Total C in debris (Mt CO Total C in debris (Mt CO 10 -50 ONF

0 -60 2010 2050 2100 2150 2200 2250 Year Year c) 45 0

e 2010 2050 2100 2150 2200 2250 e

2 40 2 -2 35 -4 30 PSW PSW -6 25 PHW PHW -8 20 RFT RFT ETF -10 15 ETF ELF ELF 10 -12 Soil C excludning C excludning Soil InertC Mt CO ONF C excludning Soil InertC Mt CO ONF 5 -14 0 -16 2010 2050 2100 2150 2200 2250 Year Year

Figure 72: Estimated long term trends for total carbon stocks in vegetation (a), debris (b) and soil (c) for Tasmanian native forests with a 20% increase in fire frequency in the absence of timber harvesting, land clearing or conversion to plantations.

Page 132 of 365

5000 e) 2 4000

Live 3000 Debris Soil 2000 Low Fire High Fire 1000 Total forest carbon (Mt CO (Mt carbon forest Total

0 2010 2050 2100 2150 2200 2250 Year Figure 73: Estimated long term trends for total carbon stocks in vegetation, debris and soil for Tasmanian native forests under assumed current fire regimes and with a 20% increase (red line) or decrease (blue line) in fire frequency.

Page 133 of 365

4 CARBON MARKETS

Part of the scope of the Carbon Study is to provide an analysis of opportunities to monetise avoided emissions, emission reductions and increased carbon sequestration. This section of the report details key carbon markets, concepts, processes and terminology related to the realisation of commercial benefits from forest sequestered carbon. These concepts are then applied in Section 5, where the Study Scenarios are reviewed in detail with respect to their potential for commercialisation.

With the introduction of formalised emissions reduction programs, a large global market-place for the trade of emissions units, often referred to as carbon-credits, has emerged. In 2005, the global carbon market was valued at around US$11 billion3. Six years later (2011), the market was valued at close to US$176 billion (Kossoy & Guigon, 2012), making it one of the fastest growing commodity markets in the world. This rapid increase is all the more notable when it is considered that around 85% of the total market value is driven by just one program, the European Union Emissions Trading Scheme (EU ETS), and that some of the world’s largest emitters are yet to formalise nation-wide compliance programs (Kossoy & Guigon, 2012).

While the global carbon market is substantial, the market for forest carbon remains a relatively small and complex niche and realising commercial opportunities from forest carbon projects remains challenging. The larger ‘mandatory’, or compliance, market is dependent on the policies imposed by multiple governments around the world, from federal to state level, often with little consistency and cross-recognition between jurisdictions. History has shown these are subject to change on short-notice, which makes this a dynamic market within which to operate.

The smaller ‘voluntary market’ is still in relatively early stages of development. Transparency within this market is low. Understanding who the buyers are, what influences their purchasing decisions, what the scale of demand is, what the likely supply pipeline is and what price point the market can bear, is difficult. Finally, the process for realising the ‘product’ that these markets trade, being carbon credits, is technically complex, with a range of legislation, standards and verification processes to be navigated before the first credit is created.

While these challenges cannot be ignored in any serious commercial assessment, the emerging market is creating some exciting commercial possibilities and this section of the report provides a background to these in the context of Tasmanian forests and the Study Scenarios.

3 All currency amounts are expressed as Australian dollars, unless otherwise indicated.

Page 134 of 365

4.1 TYPES OF MARKETS & CARBON PROJECTS

Carbon markets can be broadly categorised as being either ‘mandatory’ markets, or ‘voluntary’ markets. A mandatory market is created when greenhouse emissions are capped. Emitters are compelled, typically through government legislation, or regulation, to comply with greenhouse gas reduction targets and/or to pay for greenhouse gas emissions incurred in the course of their operations. Emissions targets are defined for emitters by a government body and a penalty (usually financial) is applied where these targets are exceeded. The global trade in carbon credits is dominated by exchanges taking place within, and around, mandatory markets. In 2010, these represented 97.7% of the total carbon market (Linacre et al., 2011).

As the term suggests, parties participating under a voluntary market do so on a voluntary basis, rather than in response to a legal, or regulatory obligation. Globally, the voluntary market remains very small relative to the mandatory market, representing just 0.3% of all trades by volume (Linacre et al., 2011). It is also a volatile and difficult-to-navigate market-place, with competing standards, protocols, programs and registries, and buyer demand waxing and waning depending on financial conditions and developments in government climate policy. During 2010-11, this volatility saw the then largest voluntary carbon market, the Chicago Climate Exchange, fold. Voluntary markets are, nevertheless, worthy of consideration as an opportunity to realise value from forest ecosystem services and, indeed, can offer some advantage over mandatory markets for commercialising certain forest carbon management approaches.

Creating the product – Carbon Credits

Within each of these broad market categories, mandatory and voluntary, there are a number of discrete market places that trade different types of carbon credits. The pathway for forest carbon owners to participate in these varies on the particular legislation, regulation, or independent standard favoured within the market-place of interest. Broadly, however, the process for realising carbon credits is as described in Figure 74.

Project Project Proponent registered with develops Project program Independent Carbon credits audit and issued Measure net verification emissions abatement

Figure 74: Key processes in realising carbon credits under formalised emissions reduction programs.

Page 135 of 365

A party that owns, or controls a forest4, referred to in this report as a Project Proponent, seeks to register an eligible forest carbon project (Carbon Project) under a formalised verification program. In registering, some form of independent verification of eligibility and net emissions reduction claims are made. This can be conducted by the body administering the program (the Administrator), or third-party audit firms recognised by the Administrator (Auditors).

Typically, the Project Proponent will be required to continue to assess net carbon abatement over the life of the Carbon Project, to report the outcomes of these assessments, and to have these assessments independently verified. Carbon Project life is long, typically extending over many decades, and the Project Proponent may be required to commit to reinstating carbon stocks, or surrendering carbon credits, whenever there is a reversal in carbon stocks (e.g. through fire). Carbon credits may be issued to the Project Proponent variously at the time of achieving registration of the Carbon Project and/or at the time that independently verified carbon abatement claims are submitted to the Administrator. In the literature, arrangements that enable carbon credits to be issued at the time of registration are referred to as ex ante programs and those that only enable credits to be issued after the abatement has occurred are known as ex post programs.

The types of forest management activity that are considered eligible for creation of carbon credits vary from program to program, as does the terminology that is applied to described these eligible activities. Broadly, these can be categorised as follows:

• Reforestation/afforestation; • Improved forest management; • Avoided deforestation.

These options and associated opportunities and challenges are discussed below and the types of Carbon Projects that are likely to apply against the Study Scenarios are provided in Table 35.

4 This is a highly simplified description used here for the purposes of illustrating underlying concepts across various programs. The actual treatment of this concept varies significantly from program to program and, in the case of the Carbon Farming Initiative, the term Project Proponent makes specific reference to ownership structures around carbon rights and land titles.

Page 136 of 365

Table 35: Carbon Project types relating to Study Scenarios and assessment of potential to realise credits under Carbon Farming Initiative (CFI, Section 4.5) and Verified Carbon Standard (VCS, Section 4.17). Scenario CFI Non-Kyoto CFI Kyoto VCS Project type ACCU’s ACCU’s (VCU’s) Altered management regimes - native forests N1. Reserve 572,000 ha of public forest, Potentially Not Potentially IFM* reduce sawlog & peeler log cut from public currently native forests to 155,000 m3 and 265,000 m3 per year (respectively). N2. Phase out old growth harvesting by Potentially Not Potentially IFM 2020 (public and private), but increase currently harvesting in re-growth forest, aiming to maintain sawlog production at 300,000 m3/y. N2A. Phase out old growth harvesting by Potentially Not Potentially IFM 2020 (public and private), and vary currently harvesting intensity in re-growth to maintain annual saw-log production at 155,000 m3/y. N3. Vary rotation length from 60-120 years. Potentially** Not Potentially IFM currently N4. Immediate cessation of harvest of old Potentially Not Potentially IFM growth on both public and private land with currently no increase in harvesting of regrowth forest. N5. Immediate cessation of all native forest Potentially Not Potentially IFM harvesting on public and private land. currently Altered management regimes – existing plantations P1. Increase rotation length by 10%. Very difficult Very difficult Very difficult IFM P2. Convert pulplog plantations to sawlog Potentially Potentially Potentially IFM plantations (increased rotation length). P3. Increase plantation growth rate by 25% Very difficult Very difficult Very difficult IFM through improved management. Altered land use A1. Conversion of private forest for plantation No No No None or non-forest, continues at recent rates. A2. Existing unstocked areas in native Potentially Potentially Potentially Reforestation forests are regenerated to native forest. or IFM A3. Expansion of plantation estate Potentially Potentially Potentially Reforestation continues at average rate for 2000-2010. A4. Rate of future planting doubles to plant Potentially Potentially Potentially Reforestation up to 10% of existing cleared land by 2050. A5. 25% of existing pulpwood plantations No No No None not replanted after first harvest.

* IFM: Improved Forest Management **Note Section 4.6, which discusses CFI eligibility criteria relating to this scenario.

Page 137 of 365

Reforestation/afforestation

With this type of Carbon Project, new forest is established into lands that have previously been cleared of forest through human activities (Reforestation), or which are naturally clear of forest (Afforestation). Depending on the requirements of the carbon program of interest, forest establishment may include direct methods, such as sewing seed or planting seedlings, or indirect methods such as facilitated regeneration (e.g. through exclusion of grazing, fire, or herbicide applications). Usually, carbon programs seek to avoid the perverse outcome of encouraging forest removal, for the purposes of later reinstating it and realising carbon credit benefits. For this reason, Carbon Projects of this type typically need to be able to demonstrate that the land established to the forest was clear of pre-existing forest for an extended period of time (e.g. clear for greater than seven years in the case of the Carbon Farming Initiative).

A key commercial challenge in delivering these types of Carbon Projects is that a comparatively large, up-front capital expenditure is typically required to establish the forest, while carbon sequestration and delivery of saleable carbon credits is a gradual process extending across many years from the time of planting.

Improved Forest Management (IFM)

Under this approach, the forest management regime being applied to existing forested areas is altered, so as to increase the amount of carbon sequestered over time as compared against a business as usual baseline. This could include, for example, reducing harvest frequency so as to allow for an increase in average stand age, increasing forest growth rates through management inputs, or increasing stocking levels in under-stocked forest areas.

Key technical challenges in delivering these types of Carbon Projects include establishing a reliable, business as usual, baseline, particularly as there is usually considerable temporal variation in forest management regimes, and proving that changes in carbon stocks relate to a long-term change in forest management regime that was not already a business as usual process. Additionally, it can be difficult to separate, or account, for natural variations in the forest system as opposed to carbon stock changes arising from human activity. A key commercial challenge, at least for those Carbon Projects that relate to reducing average harvesting frequency, is that the forest manager typically must accept reduced revenues from wood product harvesting in order to realise carbon revenues.

Page 138 of 365

Avoided deforestation (or avoided conversion)

For this type of Carbon Project, areas of forest subject to threatening processes are protected, resulting in avoiding an emissions release. In this context ‘threatening processes’ are generally intended to refer to a significant, permanent, change in land use (e.g. permanent conversion of forest to non-forest land use). Often, this requires the establishment of some legally enforceable instrument that provides for tangible protection of a forest area (e.g. land covenants, contracts, other formal agreements) and/or extinguishment of some pre-existing right to clear the forest. One of the key eligibility criteria is that the Project Proponent must be able to establish the forest is subject to a threatening process and, in the absence of the Carbon Project, the forest will be permanently converted to a non-forest land use. Another eligibility criterion relates to leakage (Section 4.2), wherein the Project Proponent must demonstrate that preventing forest conversion in one location does not result in conversion elsewhere.

Establishing points of proof demonstrating these requirements can be difficult and, in some jurisdictions at least, preventing the threatening process itself over extended periods can be extremely challenging. From a commercial perspective, a highly appealing feature of Avoided Deforestation Carbon Projects is that they require a comparatively small amount of capital to establish, while delivering relative large volumes of carbon credits over a short-time period.

4.2 KEY CARBON PROJECT CONCEPTS

While there are a wide diversity of mechanisms and programs that occur across carbon markets for verifying carbon abatement amounts, many of these share similar features and are founded on common guiding principles, some of which are detailed below.

The baseline

Accurately defining the Baseline Scenario for a Carbon Project is critically important as it is linked to (i) establishing the case for additionality (see following section) and (ii) calculating the subsequent changes in carbon stocks. The approach that is required to define the Baseline Scenario varies between programs, but typically involves describing the historical trend for a forest or land area and then evidencing this trend through various records. In the case of large, complex forest systems, where harvesting and disturbance regimes can exhibit a high level of spatial and temporal variation, identifying a correct and defensible Baseline Scenario can be very challenging and highly influential on carbon abatement recognition.

Additionality & the business as usual case

It must be possible to demonstrate that carbon sequestered through the forest management activity is ‘additional’, meaning that the forest management activity was not already required by law, or regulation, and would not have occurred under some existing business as usual process.

Page 139 of 365

While a range of approaches have been prescribed for demonstrating additionality, fundamentally these involve establishing the baseline management regime, detailing the change in management regime, and then demonstrating that the change in management regime does not represent a business as usual process. On the latter, this is often tested through comparing the management activity under the Carbon Project, against the historic baseline management activity (e.g. as determined from long-term forest management records) and/or a common practice baseline that references the typical management activity applied in the region.

In practice, experience under a range of emissions reduction programs has proven the concept of additionality to be difficult to apply since it can be challenging to evidence, or audit, the business as usual scenario and regional common practice.

Permanence

Project Proponents must be able to demonstrate that carbon stocks will be retained over the long-term. Durations vary, ranging from 30 years in the case of the Verified Carbon Standard (Section 4.17) to 100 years in the case of the Carbon Farming Initiative (Section 4.6). In most cases, where carbon stock losses occur (e.g. through fire, or deliberate removal), Project Proponents are required to surrender carbon credits equivalent to the loss, or reinstate the carbon stock (e.g. through replanting). In some cases, programs have sought to self-insure against carbon stock losses across registered Carbon Projects by retaining a proportion of carbon credits in a buffer account, or credit pool, not accessible by Project Proponents.

Leakage

Under the leakage concept, it must be possible to demonstrate that sequestering carbon through applying the forest management activity in one area does not result in an equivalent, or greater, reduction in sequestration by forests elsewhere. If this were not the case, there would be no real carbon abatement. As an example, reserving one area of forest from harvesting activities would not result in a net emissions reduction if this simply shifted harvesting activities so that an equivalent area was harvested in the same forest type elsewhere. Typically, verification programs require Project Proponents to perform a leakage assessment around the forest management activity and to net leakage emissions from calculations of net change in forest carbon stocks.

Two types of leakage are often referenced (e.g. Verified Carbon Standard 2012): • Activity Shifting Leakage – occurs where the Project Proponent implements a change in forest management that results in increased carbon sequestration within the Project Boundary (see following Section), but then alters management regimes in other forests under their control in a way that reduces carbon stocks in those forests. • Market Leakage – occurs where the change in forest management activity reduces production of forest-derived commodities (e.g. sawn-timber) in one forest area, only to

Page 140 of 365

result in that commodity being extracted from other forests outside of the control of the Project Proponent.

Leakage is an important concept in the context of the Study Scenarios, some of which involve reduction, or cessation, in harvesting and some of which explicitly refer to shifting harvesting effort from one type of forest to another.

Noting the discussion on the Project Boundary concept in the next Section, activity shifting leakage is effectively accounted for under the FCMF for many Study Scenarios and, for others, is assumed not to apply in the context of treatment under carbon verification programs. Market leakage does, however, need to be considered for those Study Scenarios that contemplate reduced timber extraction from existing forest assets.

Market leakage is challenging to quantify empirically and there are, globally, very few examples where this has been attempted. Instead, market leakage has been estimated using model- based approaches, or more typically in the context of Carbon Projects, through using qualitative frameworks. One published study by Gan & McCarl (2007), which utilised a model-based approach, concluded that for each unit of reduced production that occurs within Australasia, as much as 89% of that reduction will be replaced through an increase in the rate of extraction from forests outside Australasia (with 21-75% of this replacement occurring in developing, mainly tropical timber producing countries).

While international market leakage is an important consideration with respect to carbon flows and forest management activities occurring outside of Australia, carbon verification programs typically only extend the boundary of market leakage assessments to the national boundaries of the country hosting the Carbon Project (i.e. only domestic market leakage is considered). For this reason, domestic market leakage is the focus for the remainder of this report.

There are a range of approaches to performing domestic market leakage assessments evident within carbon markets, many of which combine a qualitative assessment framework with quantitative timber production and market data. Of these, the authors favour a conceptually appealing approach detailed within an Approved Methodology published under the VCS (Approved VCS Methodology VM0012; 3Greentree Ecosystem Services 2012). In brief, this approach requires the Project Proponent to:

i. estimate the total market leakage that is anticipated (this being the reduction in timber volume anticipated under the Carbon Project); ii. estimate the proportion of leakage calculated at (i) expected to be replaced from domestic sources; iii. determine whether domestic replacement sources are likely to be more/less emissions intensive per unit of extraction than the forests within the Project Boundary; and iv. apply default, VCS defined, leakage factors to the statistics generated at (ii) and (iii) to calculate expected domestic market leakage.

Page 141 of 365

Based on statistics provided by ABARES (2012a), roughly 69% of the timber products extracted from Tasmania are exported to international markets, inferring the remaining 31% are consumed domestically. This suggests there is scope for domestic market leakage to occur where timber product extraction from Tasmanian forests is reduced. For this reason, the leakage assessment procedure described above (3Greentree Ecosystem Services, 2012) was applied in relation to the Study Scenarios and the following domestic Market Leakage Factors were estimated:

• Reduced production from native forests: Market Leakage Factor = 8% • Reduced production from plantation forests: Market Leakage Factor = 17%

A key assumption in both cases is that the replacement source is plantation grown timber located in mainland Australia (assumptions and metrics are further detailed in Appendix 1).

The authors acknowledge that other approaches to leakage assessment exist and alternate estimates can be generated depending on the approach, market statistics and assumptions that are applied. Estimates presented here should be considered preliminary.

The Project Boundary

It is a requirement of carbon verification programs that Project Proponents clearly define the boundaries of the activity proposed for recognition. With respect to forest management activities, the ‘Project Boundary’ has several dimensions, including the geographic extent of the forest (typically determined through detailed mapping, or land tenure boundaries) and the range of greenhouse gas sources and sinks included within the carbon accounting framework.

The particular spatial boundary that is referenced has significant bearing on the approach that is taken to determine the business as usual forest use scenario, to determine the carbon stock baseline, and to perform leakage assessments. A stylised example of two approaches to setting the Project Boundary is presented in Figure 75, which shows two patches of forest subject to differing management regimes.

Clearly the business as usual scenario for both forest patches differs (harvesting verse non- harvested), as do carbon stocks, so the definition of these varies depending on whether forests are treated separately, or combined under the one Carbon Project. With respect to leakage, assuming the two forests were the sum of all forests owned by the Project Proponent, then there would be no requirement to account for activity-based leakage under boundary scenario 2, while activity leakage assessment would need to be considered under boundary scenario 2.

Page 142 of 365

Forest A: No harvesting has Forest B: Regular harvesting, occurred historically with proposed cessation

Boundary scenario 1: The two forests are treated as distinct units, with the Project Boundary (shown in red) only referencing the limits of Forest B.

Forest A: No harvesting has Forest B: Regular harvesting, occurred historically with proposed cessation

Boundary scenario 2: The two forests are combined under a single Project Boundary (shown in red).

Figure 75: Stylised representation of the Project Boundary concept and different approaches to inclusion and exclusion of forest areas.

The Study Scenarios contemplated by the Carbon Study and the modelling approach that has been adopted essentially have a whole-of-Tasmania spatial reference. For this reason, the assessment of commercial opportunities provided under this Carbon Study largely considers the Project Boundary as being inclusive of all of the forest types referenced by each particular Scenario (in some cases forests occurring in different tenure types are further separated as having separate Project Boundaries).

This is a highly simplified view and, in reality, recognising the carbon sequestration occurring under the Study Scenarios under a verification program would require the establishment of a number of separate Carbon Projects, each with their own discrete Project Boundary.

Page 143 of 365

4.3 MANDATORY MARKET OVERVIEW

There is presently no mandatory, global compliance framework in place. Instead, a series of regionally applicable programs have emerged, each with their own particular approach to the establishment of emissions compliance obligations, the generation of carbon credits, and the recognition of eligible carbon credit generating activities. To a large extent, these programs operate autonomously, although some limited level of linkage occurs between some regional programs and cross-program trading in carbon credits occurs in some instances.

A summary of major mandatory markets is provided in Insert 1. Of these, the Australian Carbon Pricing Mechanism and the linked Carbon Farming Initiative offer the most immediate opportunities for commercialising carbon stored in Tasmanian forests.

Australian Carbon Pricing Mechanism

Australia ratified the Kyoto Protocol to the United Nations Framework Convention on Climate Change (UNFCCC) during 2007. In doing so, the Australian government committed to limiting annual carbon pollution to an average of 108 per cent of 1990 levels during the Kyoto period (2008-2012). Subsequently, the Australian government has also committed to targets of reducing Australia’s emissions by at least 5% to 2020 and by 80% to 2050, as compared with emissions levels in 2000.

A key policy instrument that seeks to drive these reductions is the Australian Carbon Pricing Mechanism. Legislation underpinning the ACPM, referred to as the Clean Energy Act, passed into law during 2011. Under the Clean Energy Act, certain greenhouse gas emitters will be required to surrender carbon credits in line with the volume of their greenhouse gas emissions, or face strict financial penalties applied by the Australian government. The first compliance year under the ACPM commences on 1 July 2012.

Page 144 of 365

Insert 1. Summary of major mandatory markets

Australian Carbon Pricing Mechanism (ACP) – broad coverage of emitting sectors, anticipated to capture around 500 liable parties from commencement (Australian Government, 2011) and in excess of 350 million tonnes of greenhouse gas emissions. Implementation is underway and the first compliance period commenced on 1st July 2012. Includes a ‘fixed price’ period from 2012-2014, during which liable emitters may offset up to 5% of their emissions profile using eligible carbon credits. Beyond the fixed price period, there are no limits on the amount of eligible carbon credits that may be applied. Eligible carbon credits can be generated through the Carbon Farming Initiative (CFI), which establishes eligibility and governance requirements. Certain Australian-based forest management activities are eligible for participation under the CFI.

Californian Cap and Trade5 – broad coverage of emitting sectors within California (USA). Implementation anticipated during 2012 and forecast to become the second largest carbon market in the world with annual trading volumes of close to 400 million carbon credits and over $4 billion in value anticipated6 from 2015. Recognises domestic (USA) forestry Carbon Projects, including Reforestation, Avoided Deforestation and Improved Forest Management. Limited recognition of international forest Carbon Projects within jurisdictions (developing countries) that have entered direct bilateral agreements with California. Allows for inclusion of for-harvest forest but does not allow for the crediting of carbon stored in wood products.

European Union Emissions Trading Scheme (EU ETS) – operational from 2005 and the largest market by volume and value, accounting for over 80% of the global market (Linacre et al. 2011). Has the broadest regional coverage of any mandatory program, now including over 30 participating countries (the 27 European Union Member States and Iceland, Lichtenstein and Norway). Captured emitters include industrial facilities, power plants and oil refineries. Emissions and removals of greenhouse gases from land use change and forestry management activities are presently not included, meaning it is not possible to recognise forest carbon under this program.

New South Wales Greenhouse Gas Reduction Scheme (NSW GGAS) – operational from 2003, making it one of the oldest mandatory markets. Captures electricity retailers within New South Wales (NSW). Historically, one of the largest operating markets by volume, with around 40 million carbon credits surrendered under the program annually (IPART 2011). Recognises Reforestation Carbon Projects that occur within the borders of (NSW) and, to date, in excess of 4 million carbon credits have been issued against these project types7. Does not allow for crediting of carbon stored in wood products, allows for inclusion of for-

5 http://arb.ca.gov/cc/capandtrade/capandtrade.htm 6 Based on allowance reserve prices and data sourced from Californian Office of Administrative Law (2011). 7 Based on data maintained in the NSWGGAS Registry: https://www.ggas-registry.nsw.gov.au/

Page 145 of 365

harvest forest, but does not allow crediting of carbon retained in harvest residue (harvest events treated as a 100% release of carbon).

New Zealand Emissions Trading Scheme (NZ ETS) – partially operational from 2008 and now fully operational, the NZ ETS will have captured all major greenhouse emitting sectors by 2015 (agriculture proposed for inclusion at that time). Forestry was the first sector to enter the NZ ETS. Two classes of forest are recognised: Pre-1990 forest, being planted forest that existed at 31 December 1989 and which remained in predominantly exotic forest on 31 December 2007; and Post-1989 forest, being planted exotic, or indigenous, forest established after 31 December 1989 on land that was clear at 31 December 1989. Pre-1990 forest is automatically included under the NZETS, while forest owners may voluntarily register Post-1989 forest. Old growth, remnant indigenous forest is not eligible for NZU’s. Does not allow for the crediting of carbon stored in wood products, allows for inclusion of for- harvest forest, and allows for crediting of carbon retained in harvest residues.

Regional Greenhouse Gas Initiative (RGGI) - operated by ten states within the United States of America. Operational from 2009, RGGI captures the electricity generating sector. Annual trading volumes of 125-150 million carbon credits per annum for a value of around $400 million. Recognises domestic (USA-based) forestry Carbon Projects, including Reforestation and Improved Forest Management, but low carbon credit prices and high compliance hurdles for Project Proponents has meant no forest carbon has traded within this scheme to date. Does not allow for the crediting of carbon stored in wood products, allows for inclusion of for- harvest forest, and allows for crediting of carbon retained in harvest residues.

Page 146 of 365

Key features of the Australian Carbon Pricing Mechanism (ACPM) include:

• During the first three years (2012-2015), often referred to as the ‘fixed price’ period, the Clean Energy Regulator will issue carbon credits at a charge of $23/credit (2012), $24.15/credit (2013) and $25.40/credit (2014). These credits are automatically surrendered upon purchase. The Clean energy Regulatory will also issues credits for free under the Jobs and competitiveness Program and under administrative arrangements under the Energy Security Program

• After 1st July 2015, referred to as the ‘flexible charge’ period, open-market emissions trading commences. A price ceiling and floor will be established in early transitional stages so as to manage price volatility. After the first three years of this period, the price will be set by the market alone. The government has regulated a floor price of $15 for 2015-2016, rising by 4% per annum (in real terms) and a price ceiling set at $20 above the expected international price, rising by 5% per annum (in real terms). This form of price regulation is known as a ‘collar’.

• Three types of eligible carbon credits can be surrendered by emitters: units issued by the Clean Energy Regulator, Australian Carbon Credit Units (ACCUs) issued under a program referred to as the Carbon Farming Initiative (CFI), and certain international carbon credits that are recognised under the second commitment period of the Kyoto Protocol (or successor).

• During the fixed price period, emitters may surrender eligible ACCU’s as a means of meeting up to 5% of their annual emissions liability. Beyond the fixed price period, there is no limit on the amount of eligible ACCU’s that may be surrendered. To be eligible, ACCUs must meet Kyoto Protocol related eligibility criteria (i.e. must be ‘Kyoto ACCUs’).

• Emitters cannot use international credits to avoid emissions liabilities during the fixed price period. Beyond the fixed price period, international credits can be surrendered as a means of meeting up to 50% of an emitters total annual emissions liability.

The ACPM links to the Carbon Farming Initiative (CFI) by virtue of the ACPM recognizing certain types of ACCUs created from eligible Carbon Projects registered under the CFI. In simple terms, the ACPM creates the compliance obligation for emitters and, consequently, the market for carbon credits (including ACCU’s), while the CFI is the system by which various emissions abatement activities generate ACCU’s. A summary of this inter-relationship is shown in Figure 76 and further detail on the CFI is presented in Section 4.5.

Page 147 of 365

Australian Carbon Pricing Mechanism Carbon Farming Initiative • Clean Energy Act 2011 • Carbon Credits (Carbon Farming Initiative) Act • Clean Energy Regulations 2011 2011 • Carbon Credits (Carbon Farming Initiative) Regulations 2011

• Liable emitters are participants • Project Proponents are participants

• Sets out emissions compliance rules for • Sets out rules for participants to register Carbon participants and creates structures for Projects and create carbon credits (ACCU’s), as administering and enforcing these well as establishing administrative structures

• Creates demand for eligible ACCU’s • Creates supply of eligible ACCU’s

Figure 76: Interrelationship between the Australian Carbon Pricing Mechanism and the Carbon Farming Initiative.

Carbon Farming Initiative

Legislation underpinning the CFI, referred to as the Carbon Credits (Carbon Farming Initiative) Act 2011 (CFI Act), was passed by the Australian parliament on 23 August 2011. The CFI Act (2011) sets out the broad framework and processes by which the CFI will operate. Some of the finer detail around the operation of the CFI, and interpretation of the legislation, is detailed in a set of approved regulations referred to as the Carbon Credits (Carbon Farming Initiative) Regulations 2011 (CFI Regulations).

The Department of Climate Change and Energy Efficiency (DCCEE) has led the development of relevant legislation and regulation. The Clean Energy Regulator8 (the Administrator), established in April 2012, is responsible for administration of the CFI. The Administrator’s responsibilities include approving abatement projects, issuing carbon credits and managing them within the Australian National Registry of Emissions Units (ANREU).

The CFI allows for the creation of carbon credits, referred to as Australian Carbon Credit Unit’s (ACCU’s), from eligible projects. It is possible to generate two types of credit, being Kyoto- compliant ACCU’s (Kyoto ACCU’s) and Non-Kyoto compliant ACCU’s (Non-Kyoto ACCU’s). Critically, only Kyoto ACCU’s can be used by emitters to offset their mandatory emissions compliance obligations under the Clean Energy Act 2011.

This has important implications for abatement occurring in Tasmanian forests, as well as any prospective Carbon Projects, since, by virtue of their recognition under a mandatory market (the ACPM), there will be a substantially stronger market demand for Kyoto ACCU’s. In comparison, the main market for Non-Kyoto ACCU’s is likely to be the much smaller, lower value, voluntary

8 http://www.cleanenergyregulator.gov.au/Carbon-Farming-Initiative/Pages/default.aspx

Page 148 of 365

market. From a commercial perspective, the generation of Kyoto ACCU’s offers a considerably greater opportunity than the generation of Non-Kyoto ACCU’s.

The process for Project Proponents to participate under the CFI and realise ACCU’s is described in Figure 77 and in the passages that follow.

Registering as a Project Proponent

In order for a Project Proponent to be able to register forests and to receive ACCU’s, they first need to apply to the Administrator to become a ‘Recognised Offsets Entity’. In considering such applications, the Administrator applies a “fit and proper person” test to the applicant. Broadly, the intention of this test is to ensure that parties can be excluded from the CFI for any of the following reasons: convictions for dishonest conduct; previous breaches of emissions scheme related legislation, regulation or processes; and insolvency.

Registering a Carbon Project & realising ACCU’s

In order to register an eligible Carbon Project with the CFI, a Project Proponent, who has already successfully applied to become a Recognised Offsets Entity, submits an ‘Application for certificate of entitlement’ (Application). The project proponent must demonstrate that they own the Carbon Right/s related to the project, or demonstrate that they otherwise have the right to carry out the project. The Application should include an ‘offsets report’ detailing the features of the forest and the amount of sequestration that has occurred. The Project Proponent is also required to nominate the approved CFI Methodology that will be applied to the Carbon Project. All of these details are subject to audit and the legislation indicates that an auditor’s report should be included with the Application.

Using an Approved Methodology

Under the CFI, the government does not prescribe that a particular set of technical procedures must be applied in order to calculate carbon stocks within forests. Instead, interested parties are able to submit these procedures, referred to as methodologies9, for consideration and approval by an independent body of experts referred to as the Domestic Offsets Integrity Committee (the DOIC). Critically, a Project Proponent can only nominate one Approved Methodology to be applied to a forest at a given time (though Methodology switching is allowed over time) and, once nominated, the Project Proponent must demonstrate that they have fully complied with the processes outlined in the methodology in order to be allocated ACCU’s.

9 More details around Methodologies, including a list of those that have been drafted or approved, can be found at: http://www.climatechange.gov.au/government/initiatives/carbon-farming-initiative/methodology-development.aspx

Page 149 of 365

Page

150 of of Figure 77: Registering Eligible Offsets Projects under the CFI and realising Australian Carbon Credit Units (ACCU’s). 365

The availability of an appropriate, Approved Methodology is a critical first-step in registering a particular forest carbon management activity, and subsequently realising ACCU’s from that activity. A review of existing methodologies is provided at Section 6.9.

CFI eligibility criteria

For forest plantings to be successfully registered under the CFI, they must meet a range of eligibility criteria, some of which are detailed below.

Additionality

The CFI has sought to address additionality through publishing a ‘Positive List’ that describes activities automatically considered to be additional.

Leakage

The DCCEE’s ‘Guidelines for submitting methodologies’10 describes leakage as “emissions caused by the activity [project] that are outside of the control of the project proponent and cannot be reasonably estimated or measured by the project proponent” (p. 11).

While avoidance of leakage is listed as a key integrity criterion under the CFI, the specifics of its application to Carbon Projects remains unclear. It appears that leakage may be intended to be dealt with primarily through the methodology development and approvals process. The Methodology Guidelines state the DCEE “will consult with stakeholders on options to address leakage and will provide further guidance on the issue. The approach proposed in the guidance may result in reduction in abatement estimates to take account of leakage risks” (p.11).

Permanence

The CFI specifies that carbon must be retained within the forest for at least 100 years after the most recent date that a tree planting was added to a registered Carbon Project. Permanence is addressed in some detail and complexity within the legislation.

In the event of a significant temporary carbon sequestration reversal occurring through natural processes (e.g. fire), the Project Proponent is not necessarily required to immediately relinquish ACCU’s to make good the reversal, provided that the Administrator is satisfied that the Project Proponent has taken reasonable steps, within a reasonable time frame, to mitigate the effects of the reversal (CFI Act, Section 91f).

10 http://www.climatechange.gov.au/government/initiatives/carbon-farming-initiative/methodology-development/methodology-guidelines.aspx

Page 151 of 365

Reasonableness is not defined, so it remains unclear to what extent the Project Proponent must act in relation to such events. It might be inferred that, for forest systems that rapidly regenerate naturally after fire, the main requirement of the Project Proponent is to monitor and report carbon stocks. Where forests do not naturally regenerate, some direct management action, such as replanting, may be required.

Native forest harvesting within the Project Boundary

The CFI Act includes a number of provisions that are specific to the treatment of native forests. One of these, detailed at item 27(j) of the Act, has particular significance to those Study Scenarios that contemplate native forest management as it refers to a situation under which the Administrator must not declare an offsets project as being eligible under the CFI as follows:

(j) the project does not involve: (i) the clearing of native forest; or (ii) using material obtained as a result of the clearing and harvesting of native forest.

One interpretation of this is that, where harvesting of native forests is included within the Project Boundary, the activity would not be eligible for consideration under the CFI, which would effectively rule out Improved Forest Management projects in areas of native forest that are subject to some level of harvesting activity.

The authors understand, however, that the original intent of this section of the legislation was to preclude activities that change the fundamental nature of native forest areas, such as conversion to higher carbon storing plantations. The Explanatory Memorandum that accompanies the CFI Act includes the following statements:

1.16 Projects may not involve the clearing of native forests or using material obtained as a result of the clearing or harvesting of native forests [Part 3, Division 2, clause 27(4)(j)]. This would prevent clearing of low-density native forest to establish a higher-density carbon sink plantation, or biochar projects that make use of materials from native forests.

1.13 Land management practices that may enhance sequestration and are covered by the CFI include, but are not limited to: • reforestation; • revegetation; • native forest protection; • avoided de-vegetation; • improved management of forests; • reduced forest degradation; • forest restoration;

Page 152 of 365

• rangeland restoration; • improved vegetation management; • enhanced or managed regrowth; and • enhanced soil carbon. The Carbon Farming Initiative handbook (DCCEE, 2012), as well as related web-resources, also list Improved Forest Management activities as potentially eligible for Non-Kyoto ACCU’s.

Positive list

The Positive List identifies activities that are considered to be additional under the CFI, by virtue of those activities being deemed to ‘go beyond common practice’11. Details regarding the Positive List are provided in a set of guidelines (DCCEE, 2011d, 2011e) and CFI Regulations. In order to successfully register a Carbon Project under the CFI, Project Proponents must be able to demonstrate that the project activity is on the Positive List. Where an activity is not listed, the Project Proponent has recourse to request an update to the Positive List.

Presently, the following activities appear on the Positive List: 1. Establishment of permanent, not-for-harvest, tree plantings in areas where average annual rainfall is <600 mm, and where the planting was established after 1st July 2007. 2. Tree plantings of the type described at 1, where establishment took place prior to 1st July 2007, provided that there is adequate documentary evidence demonstrating that the primary purpose of the tree plantings was to generate carbon offsets. 3. Where rainfall is >600mm, tree plantings of the type described at 1 and 2 can be considered additional provided that: - they meet the definition of ‘Environmental Planting’ detailed in CFI Regulations; and/or - it can be demonstrated that the tree planting contributes to the mitigation of dryland salinity in accordance with Salinity Guidelines (DCCEE, 2011f); and/or - the planting is in a region for which the National Water Commission has determined the relevant state agency can adequately manage interception by plantations; and/or - the Project Proponent holds a water access entitlement for the land area into which the tree plantings have been established and which is sufficiently large to offset 90% of the water interception by the tree plantings. 4. Human-induced regeneration of native vegetation on land that is not conservation land by excluding livestock, grazing management, feral animal management, management of non-native plants, or cessation of re-growth suppression activities.

11 http://www.climatechange.gov.au/government/initiatives/carbon-farming-initiative/activities-eligible-excluded/additional-activities-positive- list.aspx

Page 153 of 365

In addition, forest Carbon Projects previously registered under the Greenhouse Friendly Program, or NSW Greenhouse Gas Abatement Scheme, are on the positive list, but neither of these situations appears to occur within Tasmania.

CFI Regulations define an Environmental Planting as “…a planting that consists of species that: (a) are native to the local area of planting; and (b) are sourced from seeds: (i) from within the natural distribution of the species; and (ii) that are appropriate to the biophysical characteristics of the area of the planting; and (c) may be a mix of trees, shrubs, and under-storey species where the mix reflects the structure and composition of the local native vegetation community.” Critically, a number of the Study Scenarios are not presently listed on the Positive List, including those that relate to cessation, or reduction, of harvesting in native forests and a range of plantation management activities. A further review of Study Scenarios against the Positive List is provided in Section 5.

Negative list

The Negative List identifies activities that are excluded from participation under the CFI on the grounds that they are associated with a ‘material risk’ of having an adverse impact water availability, biodiversity conservation, employment, local communities, or agricultural production12. As for the Positive List, details of the Negative List are provided in a set of guidelines (DCCEE, 2011d, 2011g) and CFI Regulations. Application can be made to modify the Negative List, including through adding, modifying, or removing activities.

Presently, the following forestry related activities appear on the Negative List: 1. The activity was required under a law (e.g. as is the case for vegetative offsets under some state legislation, or mine-site rehabilitation required as part of operating approvals). 2. Planting of tree species that are known weeds. 3. Establishing forest under a managed investment scheme (MIS). 4. Cessation, or avoidance, of the harvest of a plantation (where CFI Regulations define a plantation as a ‘forest established for harvest’).

12 http://www.climatechange.gov.au/government/initiatives/carbon-farming-initiative/activities-eligible-excluded/excluded-activities-negative- list.aspx

Page 154 of 365

5. The trees are planted into land that was illegally cleared of native forest, or which was legally cleared of native forest within 5-7 years13 prior to lodging the application to register the tree planting under the CFI.

The Negative List has important implications for some of the Study Scenarios, since items 3, 4 and 5 potentially limit CFI participation of a number of forest management activities that occur within the Study Scenario set. With respect to item 1, the forest protection mechanisms heralded within the Tasmanian Forests Intergovernmental Agreement (Tasmanian Government, 2011), which appear to contemplate legislating for an expansion of forest reserve areas, may have the potential to create a situation where changes to forest management activities are mandated by law. In this case, the case for eligibility under CFI may be weakened for some activities occurring under some Study Scenarios. A further review of Study Scenarios against the Negative List is provided in Section 5.

Kyoto forests

The CFI Regulations indicate that Reforestation and Avoided Deforestation are activities that are potentially eligible as Kyoto offsets projects, meaning they are potentially eligible for the creation of Kyoto ACCU’s.

In the case of Avoided Deforestation, the CFI Regulations (Sec. 1.3) define deforestation as: “The direct human-induced conversion of forest to a non-forest land use if: a) the conversion occurred on or after 1 January 1990; and b) the land on which the conversion occurred was forest on 31 December 1989.” In the case of Reforestation, in order to be potentially eligible for the generation of Kyoto ACCU’s the planting must be established: 1. on cleared land that was clear of forest at 31st December 1989; and 2. by direct human-induced means, such as planting of tree seeds or tree seedlings.

In the case of both Reforestation and Avoided Deforestation, the forest must have the following characteristics: 3. be at least 0.2 ha in size; 4. include trees capable of reaching at least 2 metres in height at maturity; and 5. include sufficient trees that, at maturity, the forest canopy will cover of at least 20% of the land area occupied by the forest.

13 The exact duration (5 or 7 years) is dependent on timing of land ownership changes.

Page 155 of 365

With respect to Tasmanian forests and the scope of this Carbon Study, these are critical eligibility considerations as they potentially exclude much of the carbon sequestered under the set of Study Scenarios from generating Kyoto ACCU’s. Based on an analysis of historic forest spatial coverage (Table 36), it is estimated that, of the cleared lands presently occurring in Tasmania, close to one million hectares was cleared prior to 31 December 1989, making this land potentially eligible for the establishment of Kyoto compliant forest. Of this, close to 200,000 hectares occurs in areas receiving less than 600mm average rainfall per annum. Tree plantings established in this area have the potential to meet Kyoto compliance requirements and would also not be required to meet the definition of an Environmental Planting in order to participate under the CFI.

Table 36: Areas of Kyoto compliant and non-Kyoto compliant land summarised by tenure and rainfall.

Rainfall

Tenure ≥600mm <600mm Total

Area (ha) Area (ha) Area (ha)

Kyoto compliant Public 9,700 1,500 11,200 Private 788,400 196,500 984,900

Total 798,100 198,000 996,100

Non-compliant Public 500 0 500 Private 30,500 6,700 37,200

Total 31,000 6,700 37,700

Total Public 10,200 1,500 11,700 Private 818,900 203,200 1,022,100

Total 829,100 204,700 1,033,800

In relation to Kyoto ACCUs arising from Avoided Deforestation, this is likely to be a relatively rare type of Carbon Project in the Tasmanian context, since there now appears to limited permanent conversion of forest to non-forest land-use and, in order to establish an Avoided Deforestation Carbon Project, it will likely be necessary to demonstrate eligibility and additionality through extinguishment, or forfeit, of an existing land clearing permit (Section 5.3). With respect to logging of native forests within Tasmania, in the main, contemporary practice does not result in permanent conversion to a non-forest land-use, by virtue of post-harvest regeneration and replanting programs that are established ‘common practice’. In this case, it is difficult to establish that even complete cessation of harvest constitutes Avoided Deforestation. Instead, this is likely to be considered to represent a form of Improved Forest Management activity and these are, currently, specifically excluded from recognition as Kyoto offsets projects.

Page 156 of 365

A further review of Study Scenarios against Kyoto criteria is provided at Section 5.

CFI Methodologies

In order for a Carbon Project to be registered under the CFI, the Project Proponent must be able to nominate an Approved Methodology that is applicable to the forest management activity. At this time, only one Approved Methodology exists that has relevance to forest Carbon Projects, with a further three drafted and presently at various stages of review14. Features of relevant existing and proposed methodologies are summarised at Insert 2, including the Study Scenarios against which each methodology may potentially be applied (assuming methodology approval in current form), with expanded detail in sections that follow.

Environmental plantings methodology

This methodology has been approved by the DOIC. It can be applied to forest management activities that have the following features (not exhaustive): • the Project activity is the reforestation of land that has been clear for at least five years. • Reforestation must involve establishing Environmental Plantings. • Large-scale harvesting15 is not carried out within the reforested area (Project Boundary).

These applicability criteria preclude application to most forest management situations in Tasmania, including nearly all Study Scenarios contemplated for the Carbon Study. The Environmental Plantings Methodology is potentially applicable to Study Scenario A2, being regeneration of un-stocked native forest areas (provided no harvesting applied).

The Environmental Plantings Methodology is considered particularly useful for smaller-scale, mixed species plantings owned by private landholders and appears to have been developed with this audience very much in mind. Based on contemporary practice, this activity is likely to represent a small carbon pool and it is not specifically represented in the Study Scenarios.

Insert 2. Existing and proposed Methodologies under the Carbon Farming Initiative. Title: Methodology for quantifying carbon sequestration by permanent environmental plantings of native species using the CFI reforestation modelling tool Short title (for this report): Environmental Plantings Methodology Developer: Department of Climate Change and Energy Efficiency

14 http://www.climatechange.gov.au/government/initiatives/carbon-farming-initiative/methodology-development.aspx 15 Limited firewood collection for personal use is allowed.

Page 157 of 365

Status: Approved Applicability: Reforestation of cleared lands where a mix of species native to the local area is planted and there is no broad-scale tree harvesting. Basis for accounting: In-office modelling (Reforestation Modelling Tool). Potentially applicable to • A2: Un-stocked areas in native forests are regenerated. Study Scenarios:

Title: Reforestation and afforestation Developer: CO2 Australia Limited Status: Draft (closed for public comment 13th May 2012) Applicability: Reforestation, or afforestation, of historically cleared lands that were utilised for agricultural production prior to forest establishment and where the forest is not harvested. Allows for mixed or single-species plantings. Basis for accounting: In-field forest measurement with application of allometric functions. Potentially applicable to • A2: Un-stocked areas in native forests are regenerated. Study Scenarios: • A3 & A4: Expansion of planted forest into currently cleared lands.

Title: Methodology for native forest protection projects Short title (for this report): Native Forest Protection Methodology Developer: Redd Forests Status: Draft (closed for public comment 13th May 2012) Applicability: Avoidance, or cessation, of harvesting in areas of native forest, where it can be demonstrated that a plan to harvest existed prior to the Carbon Project commencing. Basis for accounting: A combination of in-field measurement (volume based inventory) and application of a forest growth model (FullCAM). Potentially applicable to • N1: Reservation of production forest Study Scenarios: • N4: Cessation of old growth harvest with no increase in regrowth harvest. • N5: Immediate cessation of all native forest harvesting

Title: Measurement based methodology for farm forestry projects Short title (for this report): Farm Forestry Methodology Developer: Future Farm Industries CRC Limited Status: Draft (closes for public comment 18th June 2012) Applicability: Reforestation of historically cleared lands where reforestation is not already occurring through natural processes. Can be applied to for-harvest plantings and large-scale plantation systems. Allows for mixed- or single-

Page 158 of 365

species plantings. Basis for accounting: In-field forest measurement with application of allometric functions. Potentially applicable to • P1: Increase rotation length of all plantations by 10% Study Scenarios: • P2: Convert pulplog plantations to sawlog by increasing rotation length. • P3: Increase plantation growth rate by 25% through management. • A2: Existing un-stocked areas in native forests are regenerated. • A3 & A4: Expansion of planted forest into currently cleared lands.

This methodology allows for the use of an online mapping tool to delineate forest areas, as well as the use of Geographic Information System (GIS) data-files. The carbon accounting process relies on using the Reforestation Modelling Tool, which has been developed by the DCCEE, to make projections of forest carbon stores over time. There is no requirement to conduct in-field measurements, and record keeping requirements are low compared with other draft Methodologies, so this is likely a relatively low-cost methodology to apply. A potential disadvantage, however, is that this approach may result in an overly conservative estimate of sequestered carbon, as evidenced by comparisons presented in this and other previously reported studies (Roxburgh et al., 2010, DCCEE 2011c, p. 218).

Reforestation & afforestation methodology

This methodology is presently part-way through the DOIC review process, with the public review phase having recently closed (13th May 2012). As with all draft Methodologies that are under review, it is possible that modifications may be required in order to achieve DOIC approval. Since these are not certain, this report contemplates drafts in their current form and does not make any assessment of the likelihood of approval.

The Reforestation & Afforestation Methodology can be applied to forest management activities that have the following features (not exhaustive): • Reforestation/afforestation of land that has been clear of forest for at least the past five years and, prior to the project activity, was managed for agricultural production. • Forest must be established by direct human-induced means (does not allow for purely ‘natural’ re-growth, but regeneration through the application of seed is allowable). • Tree harvesting is excluded at all times.

In the Tasmanian context, these criteria are likely to preclude application to most existing forest assets, including native forests and the commercial, for-harvest plantation estate. With respect to the Study Scenarios, the Methodology is potentially applicable to A2, being regeneration of un-stocked native forest areas (provided no harvesting was applied) and to Study Scenarios A3

Page 159 of 365

& A4, being expansion of plantations, provided that an element of this expansion includes not- for-harvest tree plantings.

The Reforestation and Afforestation Methodology allows for the use of GIS data-files for forest boundary mapping, including through ground and aerial image based mapping approaches. The carbon accounting process relies on the use of in-field forest measurements (forest inventory) coupled with species-specific allometric functions to convert forest dimensions into estimates of carbon sequestration. This is likely a more costly methodology to apply than the Environmental Plantings Methodology, but does allow for a reduced reliance on modelling carbon stocks and application to a much wider variety of forest types.

Native forest protection methodology

This methodology is presently part-way through the DOIC review process, with the public review phase having recently closed (13th May 2012). It can be applied to forest management activities that have the following features (not exhaustive): • Project area contains native forests (not applicable to plantations). • The project activity results in the permanent avoidance, or cessation, of harvesting. • It can be demonstrated that a plan to harvest existed prior to the project activity.

If approved in its present form, this methodology is potentially applicable to a number of the Study Scenarios, including N1 reservation of production forest; N4 cessation of harvest of old growth with no increase in harvest of regrowth; and N5 immediate cessation of all native forest harvesting. Notably, the methodology developer (Redd Forests) has successfully applied a similar Verified Carbon Standard Approved Methodology16 to several Tasmanian-based Carbon Projects that have collectively realised some 179,223 carbon credits17.

The Native Forest Protection Methodology allows for a range of forest mapping approaches based on geo-referenced spatial data. The carbon accounting process relies on a combination of in-field measures of merchantable volume coupled with wood density and carbon conversion factors. The FullCAM model is also utilised to forecast future forest growth. In terms of costs of application, the Native Forest Protection Methodology is likely intermediate with respect to the other Methodologies reviewed here.

Farm forestry projects methodology

16 VM0010 Methodology for Improved Forest Management: Conversion from logged to protected forest. 17 http://www.vcsprojectdatabase.org/

Page 160 of 365

This methodology is presently part-way through the DOIC review process, with the public review phase closing 18th June 2012. It can be applied to forest management activities that have the following features (not exhaustive): • Land to be Reforested must have been clear of forest for at least the past five years and forest regeneration would not have occurred in the absence of the project activity. • Forest must be established through direct seeding or planting of seedlings/propagules. • Native scrub or woody biomass is not removed prior to reforestation (except in the case of weed species where removal mandated by law).

While this methodology has a focus on farm forestry systems, it is also intended to be applicable for use in “broad-scale plantation forests and other reforestation projects not part of a farm enterprise” (p.13). Critically, the methodology allows for harvesting activity, so that, if approved in present form, it is potentially applicable to a comparatively wide number of Study Scenarios, including P1 & P2, which see plantation rotation lengths increased; P3 increasing plantation growth rates; A2 the regeneration of un-stocked areas in native forests; and A3 & A4 the expansion of planted forest into currently cleared lands.

If approved, the Farm Forestry Methodology would allow for harvesting, but Project Proponents must nevertheless manage the project area so as to ensure that claimed carbon stocks are retained for at least 100 years. This means that forest must be re-established (e.g. through replanting), or allowed to grow back (e.g. through coppicing), following harvest events. The methodology seeks recognition of an averaging approach to carbon stock calculation, which combines in-field measurement with modelling based approaches to predict the average carbon stock over the life of the project (100 years), so as to allow the Project Proponent to claim the average carbon stock amount without subsequently surrendering ACCU’s post-harvest. The current draft is silent on the specifics of forest mapping.

CFI Risk of Reversal Buffer

For any participating forest, the CFI requires that a proportion of the forests carbon sequestration be retained by the Administrator in a ‘Risk of Reversal Buffer’. The intent of the Risk of Reversal Buffer is to at least partially insure the CFI against temporary, or even permanent, losses of sequestered carbon through disturbance events, such as fire. Presently, the Risk of Reversal Buffer is proposed as 5% of the eligible carbon sequestration for a forest. Importantly, with respect to potential commercial opportunities for Tasmanian forest owners, the Risk of Reversal Buffer acts as a discount to the total carbon that can be claimed from forests.

The Risk of Reversal Buffer does not insure Project Proponents in the event of a significant temporary carbon sequestration reversal occurring through natural processes (e.g. fire).

Supply, demand & pricing under the ACPM

Page 161 of 365

The DCCEE has provided some forward estimates of abatement anticipated for a range of activities that will potentially participate under the CFI (Table 37, DCCEE, 2011h). In 2020, it is estimated that 5-15 Mt CO2e will be delivered by Kyoto compliant activities, with up to a further ca. 5 Mt CO2e delivered by non-Kyoto compliant activities. Of this, reforestation, avoided deforestation and managed regrowth is expected to deliver from 2.5 to 8 Mt CO2e, representing roughly half of total Kyoto compliant abatement. In separate reporting, the Australian Treasury (2011) estimates that 7 million ACCU’s will be generated under the CFI in 2020.

Based on estimates by the Australian Treasury (2011), in the absence of a carbon price, Australia’s emissions are expected to reach 679 Mt CO2e in 2020. This means that, in order to meet the 5% reduction target that the government has committed to for 2020, abatement in excess of 150 Mt CO2e must be realised in 2020. This demand far exceeds the forecasts for abatement detailed in Table 37, as well as those provided by the Australian Treasury (2011), so it is highly unlikely that there will be a surplus of Kyoto ACCU’s in the market.

By virtue of demand well outstripping supply, the price of ACCU’s during the first three years of operation of the ACPM are likely to track close to the issue price for units issued by the Clean Energy Regulator, being $23/credit (2012-23), $24.15/credit (2013-14) and $25.40/credit (2014- 15). Beyond this, the pricing of ACCU’s is far from certain and there is wide divergence around estimates provided by market commentators and analysts. The reports of Garnaut (2008, 2011) and the ‘Strong growth, low pollution’ report by the Australian Treasury (2011), are among the more widely referenced forecast sources (particularly the latter in recent times).

The Australian Treasury (2011) report includes two carbon pricing scenarios related to different global targets for atmospheric CO2 stabilisation and domestic targets for emissions cuts: i. Core Policy Scenario – assumes a global stabilisation target of 550ppm and an Australian target of a 5% cut on 2000 levels by 2020 and an 80% cut by 2050. ii. High Price Scenario – assumes a global stabilisation target of 450ppm and an Australian target of a 25% cut on 2000 levels by 2020 and an 80% cut by 2050.

To reduce market volatility, at least during the early stages of transition to an open market mechanism, the Australian government has proposed introducing a floor price of $15, which will rise by 4% per annum in real terms.

Page 162 of 365

Table 37: Indicative estimates of abatement in 2020 attributable to the Carbon Farming Initiative (data source: DCCEE 2011h).

Activity Abatement in 2020 (Mt CO2e) Low High Reforestation 1 2 Avoided deforestation & managed regrowth 1.5 6 Reduced enteric fermentation <0.5 1.3 Nitrous oxide from agricultural soils <0.1 <0.5 Manure management <0.1 1.1 Reduced emissions from rice cultivation 0 <0.1 Reduced emissions from field burning residues 0 <0.1 Savannah fire management <0.5 <1 Legacy waste management <1 3.5 Total <5 <15

4.4 VOLUNTARY MARKET OVERVIEW

An appealing aspect of the voluntary market is that it is global and sales of carbon credits are not necessarily constrained by the location of the purchaser, or the carbon credit origination activity. Trades in the voluntary market also take place through a wide variety of mechanisms: from small-scale retail (e.g. web-based sales) to large-scale wholesale; from spot-market purchases to forward supply contracts; and from independently verified carbon credits to a ‘promise’ that abatement activities will be undertaken. With respect to this latter form of carbon trade, however, it is worth noting that increasing customer sophistication appears to be now limiting this activity and there is a broad trend toward independent verification (as evidenced by the emergence of a range of formalised verification standards).

The flexibility within the global voluntary market means that, as compared with existing and emerging mandatory compliance programs, there are a broader range of avenues by which carbon sequestered within Tasmanian forests could be sold to both domestic and international customers. Balanced against this, however, is the relatively small size of voluntary markets, the low values typically achieved from voluntary carbon sales, and an apparent over-supply of carbon credits feeding into voluntary programs.

Over the three years 2009 -2011, some 324 million carbon credits were traded globally, with sales during 2011 of 95 million credits (Peters-Stanley et al. 2011; Peters-Stanley & Hamilton 2012). The weighted average price for on-market sales during this period was around US$6.00-

Page 163 of 365

$6.50 per carbon credit (Peters-Stanley et al. 2011), with the weighted average price for 2011 estimated as US$6.20. There has been wide variation around this average sale price. In the US, for example, on-market sales during 2009 ranged from as little as US$0.15 per tonne, up to US$8.30 (Point Carbon, 2010).

Taking the figures reported by Peters-Stanley et al. (2011) and Peters-Stanley & Hamilton (2012) the implied value of global voluntary carbon trades during the 2009-2011 period was US$576-700 million per annum. This figure is consistent with an estimate by Kossoy & Guigon (2012) of US$569 million for 2011). This means that, as a market for an Australian forest carbon product, the global voluntary carbon market is small by volume and very small by value. Putting this in context, Australia exported some 5.3 million tonnes of woodchip alone during 2008-2009, for a total value approaching $1 billion (ABARE, 2009).

Competing within this relatively small market-place are a number of already well established project developers, aggregators and trading intermediaries with the capacity to pull large portfolios of carbon credits together from a range of sources. Some of these operators are based in developing countries, where the costs of producing verified carbon credits is low due to cheap labour and the availability of large tracts of forests subject to threatening processes that potentially make them eligible as Avoided Deforestation Carbon Projects.

For this reason, we consider that Australian-based Project Proponents, including those based in Tasmania, will find it difficult compete on price in the global wholesale market. Within the retail market, however, there is considerably less transparency and consumer understanding around carbon pricing and it may be possible for Tasmanian Project Proponents to carve themselves a relatively high-value niche in this area.

Achieving any significant market penetration and sale volumes within the voluntary market would, no doubt, require a significant marketing effort aimed at raising the profile of Tasmanian Carbon Projects and positively differentiating Tasmanian originated carbon credits (or unverified offsets) from the wide range of alternate carbon credit sources (both forest and non-forest). Potential points of differentiation include promotion of co-benefits associated with Tasmanian Carbon Projects (e.g. biodiversity values), as well as relatively transparent forest/carbon ownership structures and the relatively low risk of sequestration reversals arising from illegal logging, land ownership changes, or sovereign risk.

Independent verification standards

During the early days of the development of the voluntary carbon market, it was not uncommon for parties wishing to abate, or offset, greenhouse gas emissions to effectively purchase a commitment from a Project Developer, or third-party promoter, to take some action that would result in future abatement (e.g. plant a forest now, for abatement into the future).

Page 164 of 365

Over time, however, customers in voluntary markets have become increasingly sophisticated in purchasing decisions and there is now a strong trend towards only purchasing carbon credits that have been verified through some independent, formalised standard.

A diverse set of sometimes competing standards now exist within the voluntary market-place, some with a regionally specific focus and others intended for international use. Several of the latter are described at Insert 3 and the most dominant of these, the Verified Carbon Standard (VCS) is dealt with in further detail at Section 4.17. With respect to locally relevant standards for Australia, the Carbon Farming Initiative effectively serves as the standard for generation of both mandatory compliance grade carbon credits (Kyoto ACCU’s) and voluntary market carbon credits (Non-Kyoto ACCU’s).

Insert 3. Verification standards featuring in voluntary markets.

Verified Carbon Standard (VCS)18 • Founded in 2005 and headquartered in Washington, with regional nodes including South Korea and South America. • Arguably now the dominant, most widely recognised, voluntary standard globally, having registered some 805 Carbon Projects around the world. • Allows for a very wide range of Carbon Projects, including Avoided Deforestation, Improved Forest Management and Reforestation. • Potential application to a number of forest management scenarios within Tasmania, with three Tasmanian-based Carbon Projects having already been registered. • A key limitation is that it is presently not possible to apply the VCS to areas of Kyoto compliant forest, due to the potential to ‘double-count’ abatement that is already being reported by the Australian government under its current international emissions reporting framework.

Carbon Fix Standard19 • Launched in 2007, administered from headquarters in Germany. • Differentiated from a number of other standards through its exclusive focus on afforestation and reforestation activities, as well as its prescriptive approach to methodology application - only allowing for the use of a single Approved Methodology developed by Carbon Fix. • Has gained comparatively little traction, with 11 Carbon Projects registered and 200,100 verified carbon credits issued to date20. • Potential application to a number of forest management scenarios within Tasmania, but no Australian-based are yet registered.

18 http://www.v-c-s.org 19 http://www.carbonfix.info/CarbonFix-Standard.html?PHPSESSID=e9g0f02ta0bfp4j34dr2iq5le7 20 http://www.carbonfix.info/Project/Carbon-Registry.html

Page 165 of 365

• As with VCS, difficult to achieve application to areas of Kyoto compliant forest.

Plan Vivo Standard21 • Formulated in 1994 and now administered from the United Kingdom. • Focuses on ‘community based’ projects, predominantly in under-developed countries and includes forest management Carbon Projects. • A unique feature is its direct linking of payments to abatement ‘producers’ via Payment for Ecosystem Services Agreements. • Over 8,000 small-holders and community groups presently participating, with close to 1.2 million carbon credits issued to date. • Due to its focus on projects in developing countries and community based projects, this standard is not suited to application within Tasmania.

Climate, Community & Biodiversity Standard (CCB)22 • Established in 2003, administered as a partnership between several conservation organisations, including Conservation International, CARE, the Nature Conservancy, Rainforest Alliance and the Wildlife Conservation Society. • Not designed as a tool for independent verification, and recognition, of carbon credits. Instead, the focus is on confirming that Carbon Projects deliver a net greenhouse gas reduction and that co-benefits, such as socio-economic outcomes and environmental benefits are delivered. • Typically, Project Proponents interested in achieving both CCB standard recognition and verified carbon credits will undertake a dual verification process and include a second standard for carbon credit recognition (e.g. VCS). • To date, 37 projects have been validated, including 32 related to forest management, one of which is based in Tasmania.

Chicago climate exchange (CCX)23 • Established in 2003 and, to early 2010, the rapidly established itself as the dominant platform for generating and trading carbon credits from forest Carbon Projects. • During mid-2010, was bought out by Intercontinental Exchange, after which carbon credit trading activities were rapidly wound back to the point that they have now effectively ceased. • In its seven years of operation, CCX registered 336 Carbon Projects (including 32 forestry- related) generating some 11.6 million carbon credits.

21 http://www.planvivo.org/ 22 http://www.climate-standards.org/index.html 23 http://www.chicagoclimatex.com/

Page 166 of 365

• Uniquely, the CCX included emitters who committed themselves to binding emissions reduction targets, including through carbon credit purchases, effectively combining, within a single platform, carbon verification processes and trading mechanisms. • This verification option and market-place is no longer available to forest managers.

Supply, demand & pricing in the voluntary market

Trends in the voluntary market-place have been comprehensively reviewed and reported annually by Ecosystem Marketplace and Bloomberg New Energy Finance since 2007 (Hamilton et al. 2007; Hamilton et al. 2008; Hamilton et al. 2009; Hamilton et al. 2010; Peters-Stanley et al. 2011, Peters-Stanley & Hamilton 2012).

Based on their analyses (years 2002-2011), the trend has been an increase in the volume of carbon credits traded in the marketplace from 2002. However, a plateau is evident from 2008- 2011 (Figure 78), where total annual volume traded ranged from 95-132 million credits. Total market value shows a sharp rise to 2008 (US$ 126.6 million), followed by a sharp drop in 2009 and a gradual recovery to 2011, when the weighted average price per carbon credit was $6.20 (Figure 79). This pattern is most likely linked to the global economic recession of 2008 and gradual recovery of global markets (Hamilton et al. 2010). For the most recently reported year, 2011, total market value was estimated at US$576 million.

The supply of carbon credits into voluntary markets has far exceeded demand for at least the past several years. Between the VCS, Gold Standard and Plan Vivo programs, in excess of 102 million verified carbon credits have been issued, predominantly within the last five years (Table 38). Of these, only 24% have so far been retired (representing consumption), so that around 78 million credits remain live and potentially available within the market-place.

Compared with other standards, the VCS is clearly the dominant program with respect to credit production, accounting for over 90.5 million credits, followed by Gold Standard and Plan Vivo at around 1.9 million and 1.2 million credits respectively. Energy production related projects (e.g. renewable, biogas) are by far the dominant credit source, accounting for over 62 million credits. Agriculture, Forestry and Land Use (AFOLU) projects account for around 5.9 million credits (3.8 million still live), which ranks as the third largest production source.

The credit retirement figures suggest that, on a proportional basis, credits derived from AFOLU projects may tend to be retired more frequently than average (AFOLU 35%, verse average of 24%) and are certainly retired more frequently than credits arising from the second largest source, emissions reductions arising from mining and mineral production projects. Nevertheless, in absolute terms, credits derived from energy production projects are retired more frequently than all other credit sources combined and credits from AFOLU projects are retired less often than credits arising from waste handling & disposal projects, arguable a source with more marketing challenges than exist for AFOLU.

Page 167 of 365

800 Volume (no. of credits) 700 Value ($US)

600

500

Volume/ 400 value (x 1million) 300

200

100

- 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year

Figure 78: The annual volume and value of credits traded within voluntary markets (as reported by Ecosystem Marketplace and Bloomberg New Energy Finance).

$5 Weighted average $4 price ($US/credit) $3

$- 2006 2007 2008 2009 2010 2011 Year

Figure 79: The weighted average price paid for carbon credits in the voluntary market-place for 2006-2011 (as reported by Ecosystem Marketplace and Bloomberg New Energy Finance).

Page 168 of 365

Table 38: Voluntary carbon credit statistics for VCS, Plan Vivo and Gold Standard programs, including credits issued to date (Issued), credits retired, or extinguished (Retired) and credits still available (Live).

Project type Issued Live Retired Proportion retired Energy 62,939,203 47,641,364 15,297,839 24% Mining/Mineral production 9,134,469 8,723,938 410,531 4% Ag, Forestry, Land Use 5,914,539 3,858,409 2,056,130 35% Waste handling & disposal 5,887,356 3,172,545 2,714,811 46% Fugitive emissions (gas) 4,552,905 4,546,854 6,051 0% Energy demand 4,474,308 3,880,563 593,745 13% Fugitive emissions (fuel) 4,050,503 2,890,314 1,160,189 29% Manufacturing industries 2,586,723 1,928,455 658,268 25% Other 1,973,112 588,860 1,384,252 70% Chemical industry 683,929 392,995 290,934 43% Metal production 286,468 286,468 - 0% Transport 90,004 29,433 60,571 67% Energy distribution 41,543 39,044 2,499 6% Livestock & manure 37,157 34,622 2,535 7% Total 102,652,219 78,013,864 24,638,355 24% Data sourced from the Markit24, Gold Standard25 and Verified Carbon Standard 26 registries.

A review27 of two leading registries, Markit and Carbon Trade Xchange, shows over 2.2 million credits offered for sale, for a volume weighted average price of $3.19, representing combined value of $7.1 million (Table 39). No bids were current at the time of the review, meaning there were no buyers in the market place with unfilled orders. In comparison with other standards, Plan Vivo and Gold Standard credits are offered at premiums, although relatively small volumes are available to assess pricing (particularly for Gold Standard). Credits generated from the VCS represented 72% of the volume of credits for sale, for a weighted average price of $2.82.

Looking at the contribution by various project types (Table 40), we see a similar pattern in the volume spread of credits for sale as that for credit creation, with credits arising from energy production related projects representing the bulk of credits in the market, followed by waste handling and forestry related projects (AFOLU). The offer price for forestry related projects

24 http://www.markit.com/en/products/registry/environmental.page 25 https://gs2.apx.com/mymodule/mypage.asp 26 http://www.vcsprojectdatabase.org/ 27 Search performed 3pm, 19th May 2012.

Page 169 of 365

represents a premium over all other credit types. Although the volumes are small for such analyses, this suggests that either the cost of production of these credits is higher than for other project types, or that buyers will preferentially pay a premium for forestry-derived credits.

Table 39: Volume and value of credits listed for sale on the Markit28 and Carbon Trade Xchange29 registries. Program Credits for Average Total value sale price (volume weighted) Verified Carbon Standard 1,608,642 $2.82 $4,544,276 ISO 14064 – 2* 516,720 $3.36 $1,737,509 Plan Vivo 86,235 $8.87 $764,799 American Carbon Registry 10,285 $2.26 $23,213 Gold Standard 4,139 $8.67 $35,881 2,226,021 $3.19 $7,105,677 * This is a linked standard, but no reference given as to which compatible standard these credits are linked to.

Table 40: Volume and value of credits listed for sale on the Markit30 and Carbon Trade Xchange31 registries for a range of project types (forestry related highlighted in bold). Project type Credits for Average Total value sale price (volume weighted) Energy industries 1,591,010 $2.40 $3,817,551 Waste handling & disposal 288,990 $4.67 $1,348,412 Renewable energy 171,030 $5.08 $868,453 Improved forest management 45,400 $8.88 $403,200 Afforestation/Reforestation 40,964 $8.83 $361,599 Mining & mineral production 33,725 $2.21 $74,534 Manufacturing industries 27,457 $2.50 $68,643 Landfill gas 10,285 $2.26 $23,213 Avoided deforestation (REDD) 10,000 $10.70 $107,000 Energy demand 4,560 $4.00 $18,240 Fugitive emissions from fuels 2,600 $5.71 $14,848 2,226,021 $3.19 $7,105,677

28 http://www.markit.com/en/products/registry/environmental.page 29 http://www.carbontradexchange.com/ 30 http://www.markit.com/en/products/registry/environmental.page 31 http://www.carbontradexchange.com/

Page 170 of 365

* This is a linked standard, but no reference given as to which compatible standard these credits are linked to.

The latter seems unlikely based on an alternate set of pricing trends reported by Diaz et al. (2011), which show forestry-derived credits trading for an average of $5.50 in 2010, which is slightly lower than the average for all credits sources during that year ($6).

The supply of forestry-derived carbon credits appears to be well exceeding consumption. This observation is evidenced by the data presented in Figure 80 which shows the current status of credits derived from AFOLU projects, including forestry projects, under the VCS. Carbon credits generated from 2002-2006 are now largely retired, however, from 2007 onwards, the majority of credits remain live. Overall, only 27% of the entire volume of credits generated have been retired to date and, averaged over a ten year period (2002-2012), only 129,000 credits have been retired.

At the time of writing, over 3.4 million AFOLU derived credits remain live on the VCS registry.

2,500,000 Retired Live

2,000,000

1,500,000

Carbon credits 1,000,000

500,000

- 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year

Figure 80: Carbon credits generated from AFOLU projects under the VCS program, showing the proportions of credits that have been retired, or which remain live in the market-place. The figures for 2002-2004, which do not appear here due to scale, are 7134, 7134 and 7975 respectively (all of which are retired).

Page 171 of 365

Market segmentation

Any marketing strategy seeking to affect the sale of large volumes of Tasmanian forest carbon into the voluntary market will need to carefully consider the purchasing motivations of various market segments. Even a cursory review of these suggests that mandatory markets exert considerable influence on the voluntary markets, and that purchases, by traders, aggregators and project developers, introduce distortion into sales statistics for these markets.

Detailed analyses provided by Peters Stanley et al. (2011) and Point Carbon (2010) indicates there are a range of parties purchasing carbon credits within voluntary carbon markets, including large emitters, traders, financial institutions, project developers, project aggregators and retail buyers, such as small organisations and individuals (Figure 81). Of these, private sector business is the dominant purchaser of carbon credits. Sales to traders, project developers and project aggregators (as opposed to parties seeking to offset their emissions) account for a surprisingly significant proportion of annual sales.

Point Carbon (2010) report that, with respect to voluntary purchases of carbon credits within the USA, as many as half arise from so-called “pre-compliance” speculation, wherein market players buy carbon credits in the voluntary market on the expectation that these may be traded into developing compliance markets.

Global (2009) USA (2010)

Figure 81: Buyer segments (over-the counter trades) globally during 2010 (A, data source: Peters-Stanley et al. 2011), and for the US during 2009 (B, data source: Point Carbon, 2010).

Page 172 of 365

Future trends for voluntary markets

Attempting to forecast future market trends for any commodity traded under open-market conditions involves a considerable amount of speculation and assumption. This can give rise to a wide divergence in views between market commentators and this is certainly evident within the voluntary carbon market.

From 2008, Ecosystem Marketplace has conducted annual surveys of market participants to gauge their views on likely growth trajectories for market development. On average, participants have forecast a growth in market volume, but the forecast rate of growth has varied widely from year to year. In 2010, survey participants were forecasting that market-size, as measured by volume of credits traded, would exceed 1.6 billion tonnes CO2e in the year 2020. Just one year later, however, participants were forecasting a 2020 market size of 600 million, representing a drop in forecast volume of over 60%.

Assuming the latter prediction were correct, this implies an average year on year growth in the volume of credits traded from 2011-2020 of 63% per annum. However, analyses by Peters- Stanley & Hamilton (2012) suggest that there is a tendency for survey participants to over- predict future market trends and, with each successive year of market activity, participants are applying increasingly conservative estimates for future market growth.

With respect to the market-place for forest-derived credits, Diaz et al. (2011) forecast that over 350 million forestry-derived credits will enter the voluntary market over the next 3-4 years, the majority arising from Avoided Deforestation Carbon Projects. This aligns with anecdotal observations reported by Bulinski (2012), which suggest that larger voluntary-market players are generally anticipating significant volumes of credits to be generated from south-American and south-east Asian based Avoided Deforestation Carbon Projects. Finally, a number of Avoided Deforestation related methodologies have now been approved under the VCS program, potentially paving the way for an increasing number of Carbon Projects to enter the program.

All of this suggests that, without a major shift in demand, voluntary markets are highly likely to continue to be over-supplied for the next several years at least. As a market segment, large volumes of forestry-derived carbon credits already exist and it is anticipated that even greater volumes will enter the market-place to at least 2015. The general over-supply in the market is likely to apply downwards pressure on carbon credit prices broadly, and forest-derived credits specifically, in coming years.

On balance, the authors consider that there is a low likelihood that prices in the voluntary market will exhibit strong upwards movement. The prospects for increases in traded volumes appears more promising. However, the authors consider it likely that growth in volume will be considerably less than that suggested by survey participants in the latest reporting from Ecosystem Marketplace (Peters-Stanley & Hamilton 2012).

Page 173 of 365

Verified carbon standard

The VCS has now established itself as the dominant standard in the global voluntary market place. Peters-Stanley & Hamilton (2012) report that credits generated through the VCS now account for 58% of all voluntary carbon trades (up from 34% in 2010; Peters-Stanley et al. 2011). Certain Tasmanian forest management activities have the potential to participate under the VCS and the authors consider that, by virtue of its market-share and wide recognition, Tasmanian forest owners should consider the VCS as the international standard of choice.

With respect to Australian-based forestry activities, a key restriction arises from the Australian government reporting carbon sequestration from Australian forests as part of its international obligations under the Kyoto protocol. For many forest management activities that would otherwise be eligible under the VCS, the government’s approach to international reporting creates a ‘double-counting’ issue, whereby Kyoto-eligible sequestration is already ‘claimed’ by the Australian government. Effectively, this means that only those forest activities that do not meet Kyoto eligibility requirements, such as reforestation of lands cleared after 1st January 1990, or Improved Forest Management projects, have the potential to participate under the VCS.

The VCS requires Project Proponents to meet the requirements of the latest version of the standard. Additionally, forest carbon Project Proponents must adhere to the latest version of VCS’ ‘Agriculture, Forestry and Other Land Use Requirements’ and have their methodology approved through a double verification process, or adopt a methodology previously approved.

Additionality

The VCS prescribes that additionality is demonstrated according to the procedures outlined in the Approved Methodology selected by the Project Proponent. These vary, but in broad terms, involve establishing the baseline management regime, detailing the change in management regime, and then demonstrating that the change in management regime does not represent a business as usual process. The latter point is often tested through comparing the management activity under the Carbon Project against the historic baseline management activity (e.g. as determined from long-term forest management records) and/or a common practice baseline that references the typical management activity applied in the region.

Permanence

The VCS deals with permanence through requiring Project Proponents to undertake a Non- Permanence Risk Assessment (Verified Carbon Standard, 2011) in relation to the proposed Carbon Project. The assessment involves rating a range of risk factors, including: • Internal risks - management team experience, project structuring, financial viability, opportunity costs, project longevity; • External risks – land tenure arrangements, community engagement, political risk; and

Page 174 of 365

• Natural risks – fire, pest/disease, extreme weather, geological.

Based on this assessment, a project risk rating factor is calculated. Where this exceeds a critical threshold, the project is not eligible for participation under VCS. Notably, situations where retention of the carbon forest cannot be adequately demonstrated for at least 30 years will automatically fail against this threshold. Where the risk rating falls below the critical threshold for ineligibility, a proportion of the carbon credits issued to the Carbon Project will be retained in a ‘pooled buffer account’ (referred to here as the Risk of Reversal Buffer). The higher the risk rating, the greater the proportion of credits retained in the Risk of Reversal Buffer account. Project Proponents have the option of undertaking regular verification audits for their Carbon Project, which can be as frequent as every five years. Where verification is undertaken, the proportion of credits retained in the buffer account is reduced over time.

Based on a preliminary assessment, it may be possible to reduce the VCS project risk rating for many Study Scenarios to 10%, being the minimum allowable. Key assumptions include: • significant capital is available to support the development and ongoing management of the Carbon Project; • highly experienced personnel will establish and manage the Carbon Project; • ownership and protection of forest areas will be established through formal legal instruments extending beyond 70 years; and • verification is performed no less frequently than once every five years.

The authors acknowledge that other risk rating outcomes are possible and that there is scope for the above estimate, which represents the minimum allowable under VCS, to increase depending on the particular features of the Carbon Project being contemplated. Opportunity costs associated with, for example foregone revenue from harvesting or agricultural land uses, can exert considerable upward influence in the context of Tasmanian forests.

Leakage

The VCS requires proponents to account for leakage (activity shifting and market) and to net leakage amounts from carbon stock claims. The particular procedure that is applied to assess leakage is detailed in the Approved Methodology applied by the Project Proponent and these vary. Broadly, leakage assessments can involve direct monitoring (usually a mandatory requirement for activity shifting leakage), or indirect assessment where ‘scientific knowledge provides credible estimates of likely impacts’ (Verified Carbon Standard 2012).

Approved Methodologies with relevance to Tasmanian forests do not allow for any activity shifting leakage, but market leakage is allowed provided it is netted from carbon stock estimates. Importantly, leakage is only considered within the boundaries of the host country, so that timber extraction displacement to international sources is not required to be accounted for.

Page 175 of 365

Insert 4. Selected Methodologies approved under the Verified Carbon Standard. Title: VM003: Methodology for improved forest management through extension of rotation age Developer: Ecotrust Applicability32: Extension of rotation ages for areas of forest subject to clear-cutting, or patch-cutting practices. Forests must be certified by the Forest Stewardship Council within one year of Carbon Project commencement (or at commencement). There is no leakage through activity shifting to other lands owned/managed by the Project Proponent. Basis for accounting: In-field measurement from sample plots coupled with allometric functions/expansion factors. Potentially applicable to: • N3: Increase rotation length from 80-120 years. • P1 & P2: Increase plantation rotation length.

Title: VM0010: Methodology for Improved Forest Management: Conversion of logged to protected forest Developer: Green Collar Solutions Applicability: Permanent cessation of planned timber harvest activities in non-plantation forest (i.e. native forest in the Australian context). A legal right to harvest must exist, prior to implementation of the Carbon Project, that includes details of timber resource extraction (forest size, timber volume). Basis for accounting: Forest inventory data referenced by harvest plans, followed by regular in- field forest measurements from sample plots. Potentially applicable to: • N1: Reservation of production forest, reduced cut volumes. • N2/N2A: Phase out old growth harvest, increase regrowth harvest. • N4: Immediate cessation of old growth harvest. • N4: Immediate cessation of all native forest harvest.

Title: CDM methodology AR-AM0011: Afforestation and reforestation of land subject to polyculture farming Developer: Clean Development Mechanism Applicability: Afforestation and reforestation activities on land previously used as a mixed agricultural production system. Cannot be applied where forest management occurs on natural grasslands. Basis for accounting: In-field forest measurement from sample plots, with application of allometric functions/expansion factors. Potentially applicable to: • A3 & A4: Expansion of plantation estate.

32 Listed applicability criteria are a brief summary only and are not intended to be exhaustive.

Page 176 of 365

Approved methodologies

Presently, nine methodologies have been approved that focus on improved forest management, or avoided deforestation, with three of these reviewed at Insert 4. No VCS methodology is yet available for afforestation/reforestation. However, the VCS also recognises methodologies approved under the Clean Development Mechanism and several are published that have potential application to afforestation/reforestation activities within Australia

Domestic market & the non-Kyoto fund

The Australian domestic voluntary market is small by global standards, with less than 0.5 million credits transacted during 2011 (Peters-Stanley & Hamilton, 2012), representing less than 2% of the north American market (30 million credits) and 0.5% of the global market (95 million credits).

Purchasing patterns within the Australian market are not well documented, but in the authors’ experience, purchasers include private individuals, government agencies, local councils, small to medium size enterprises, events, banks and airlines. The latter two sectors appear to be the largest purchasers and a web-based review of their carbon credit (or offset) purchasing policies suggest purchases are made from a portfolio of international and domestic sources. Notably, QANTAS state33 that they are utilising credits from Tasmanian native forest protection Carbon Projects, these having been developed and registered under the VCS by REDD Forests34.

The small size of the Australian market and its established pattern of purchasing, at least part of its requirements, from international sources make it a very challenging target for selling large volumes of carbon credits from Tasmanian Carbon Projects. However, the domestic market is likely to receive a significant boost from the Australian government having established, in connection with the CFI, a non-Kyoto Carbon Fund that will commence from July 2013.

Under this fund, the government will purchase non-Kyoto ACCU’s with a view to encouraging abatement and offsetting activities that deliver verifiable greenhouse gas reductions, but which don’t meet the requirements for Kyoto eligibility. Currently, the fund has been announced as having a six year life, over which time the government has committed to spend $250 million (ca. $41.6 million per annum) on the purchase of non-Kyoto ACCU’s. This will effectively establish the Australian government as the largest purchaser of carbon credits in the domestic market.

Purchases will take place via a ‘competitive tender’ process, possibly suggesting that pricing will be the main criterion for purchasing. The authors are unaware of any forward estimates or market commentary around the likely price range within which purchases will be made. A price cap is applied, in that the government has indicated they will not pay a higher price than the price of Kyoto ACCUs in the compliance market.

33 http://www.qantas.com.au/travel/airlines/abatement-projects/global/en#native-forest-protection 34 http://www.reddforests.com/projects-and-media

Page 177 of 365

5 OPPORTUNITY ASSESSMENT

Based on the review of carbon markets and eligibility requirements presented at Section 4, there would appear to be some immediate opportunities for abatement realised under various Study Scenarios to trade into either voluntary, or mandatory, markets (and both in some cases). This Section provides a detailed assessment of the potential scale of these opportunities, the possible pathways for realising them and the key challenges involved.

The focus for this assessment is existing market mechanisms under current policy settings and the more immediate opportunities that appear to exist within these. Government policy is the fundamental driver of opportunities within carbon markets, with the capacity to influence these either positively, or negatively. Changes to policy settings have the potential to greatly alter the outcomes of the assessment provided here and, for this reason, some commentary around the influence that future policy pathways may have, is provided at Section 5.5.

This assessment also focuses on monetisation of carbon abatement through abatement verification, carbon credit creation and, subsequently, the trade of carbon credits. Critically, participants operating within the Carbon Farming Initiative (CFI), or parties intending to trade carbon units within the framework of the Australian Carbon Pricing Mechanism (ACPM) need to be aware that ACCU’s and other carbon units are considered to be financial products under the legislation. Trading of carbon units is considered to be a trade in financial derivatives. It is beyond the scope of the Carbon Study to provide a detailed review of this issue, but in broad terms, this means that depending on the exact role a participant intends to play within the market, there may be a need to operate under an Australian Financial Services Licence. For this reason, anyone contemplating these activities should seek relevant regulatory guidance from ASIC and other relevant agencies.

The authors acknowledge there are valid arguments that achieving emissions abatement in one sector of the economy (e.g. the forest sector) provides for economic benefits to other sectors of the economy, even where carbon credits are not created and traded. In Australia, this could occur through, for example, the government passing on the benefit of increased carbon sequestration by Tasmanian forests through reducing the abatement burden for those sectors captured by the Australian Carbon Pricing Mechanism (ACPM).

Modelling the net-effect of these macro-economic influences is very complex, since it requires a detailed consideration of factors such as: comparative marginal abatement costs (i.e. is abatement through forest management less expensive than abatement elsewhere in the economy; Lewis et al. 2008), foregone revenue generating opportunities associated with the emissions reduction activity, and the capacity of affected sectors to absorb the costs associated with driving emissions reductions. While these are important issues worthy of further consideration, they are beyond the scope of the Carbon Study.

Page 178 of 365

In order to conduct this opportunity assessment, the authors have had to make a number of assumptions around future trends within the carbon market, and the scale of opportunities that may exist within them. In so far as is possible, these are informed by expert sources, independent reporting and the authors’ own commercial operating experience. Inevitably, however, the analysis includes some level of subjective assessment. The authors acknowledge it is entirely valid to apply alternate sets of assumptions around key metrics and that alternate approaches, assumptions and forecasts exist elsewhere in the market-place.

Finally, by virtue of the scale contemplated by the Study Scenarios (i.e. whole-of-state), which well exceed the likely Project Boundary for any Carbon Project, as well as the timeframes and information available for the Carbon Study, the assessment provided here should be considered as preliminary and indicative in nature. A more detailed financial and technical assessment would need to be performed in advance of making any commercial decision to pursue the opportunities outlined here, or in working up detailed structures around specific real-world activities (as opposed to possible future, state-level scenarios) proposed as Carbon Projects.

5.1 CHOICE OF MARKETS & STANDARDS

Due to the demand and relatively high value created by the Australian Carbon Pricing Mechanism (ACPM), it will be in the Project Proponents interests to seek to realise all Kyoto eligible abatement as Kyoto ACCU’s under the CFI and then trade this to liable parties captured by the ACPM. Study Scenarios that contemplate plantation management (P1-3), or expansion activities (A3-4), include some level of Kyoto eligible abatement that may have potential to be brought to account under the CFI. Additionally, if the deforestation contemplated by Study Scenario A1 could be avoided, then this would also be potentially a Kyoto compliant activity for the purposes of the CFI (albeit that the scale of future deforestation is likely to be relatively limited within Tasmania).

The remaining abatement generating Study Scenario’s reference forest types and/or forest management activities that are presently not considered Kyoto eligible. This includes all of the native forest related Study Scenarios (N1-N4) and plantation scenarios (P1-3, A2-4) wherever plantations have been (or will be) established on land that was not clear of forest at 31 December 1989. In these cases, Project Proponents effectively have the option of pursuing registration and Non-Kyoto ACCU generation under the CFI, or registration and VCU generation under the VCS.

For this latter set of Study Scenarios, wherever there is potential to generate very large volumes of carbon credits, it is recommended that Project Proponents consider registering discrete Carbon Projects under both CFI and VCS. Creating two types of carbon credits from separate Carbon Projects would allow for participation in both domestic and international markets, thereby widening the potential customer base and reducing exposure to individual markets. Notably, the VCS already appears to have a high level of recognition and acceptance by

Page 179 of 365

Australian purchasers, so that the main driver for pursuing CFI registration is to realise sales to the Australian government under the Non-Kyoto fund ($250 million over six years).

5.2 WEIGHTING THE OPPORTUNITIES

In the sections that follow, a series of net present value estimates are provided around abatement arising from the Study Scenarios and various commercial carbon project structuring options. Net present value (NPV) is a metric that is commonly used to evaluate the potential value delivered against investment into long term projects. Where NPV is positive, this indicates that net cash flows are anticipated to be positive in present value terms. If negative, then the enterprise will be loss generating. Net present value can be used to evaluate the potential value associated with pursuing one opportunity over another, either through direct comparison, or through estimating NPV for one opportunity with incorporation of the opportunity cost associated with foregoing the alternate opportunity. With respect to carbon projects, NPV has been previously used to evaluate potential project value by a number of studies (e.g. Polglase et al. 2008; Polglase et al. 2011; Macintosh 2012a).

The influence of various costs on participating in carbon markets is considered by the NPV estimates provided here (detailed in following Sections). Critically, however, opportunity costs have not been considered. In relation to Study Scenarios that reference changed management practices within native forests, there are opportunity costs associated with, for example, reduced timber extraction. These costs are not considered here (noting that economic considerations beyond carbon flows have been specifically identified as out of scope for the Carbon Study).

In effect, this means that while it is reasonable to use the NPV’s presented in this report as a comparison against ‘doing nothing’ following a decision to reduce extraction rates, it is not reasonable to infer that the NPV represents the net value that actually arises due to making the decision. For the latter, opportunity costs associated with foregone revenue would need to be considered and this would reduce the NPV estimates presented here.

In relation to the Study Scenarios that reference altered, or expanded, plantation forest activity, land values, the costs associated with establishing and managing plantation forests, and the opportunity costs associated with converting agricultural lands to forest, are not considered. In effect, the plantation related NPV’s should be considered as the additional, incremental, value associated with converting a timber producing plantation into a Carbon Project and realising carbon revenue on top of timber production related revenue (which is assumed to result in at least a neutral NPV in the absence of the Carbon Project).

Estimates of NPV’s are presented as ranges that take into account the combined effect of uncertainty in carbon stock estimates (upper and lower bounds based on 95% confidence intervals, Section 3), discounts applied for leakage and risk buffer factors, carbon pricing scenarios and market penetration (or carbon sales) scenarios. Further detail on these and key assumptions applied in calculating NPV estimates are provided in the sections that follow.

Page 180 of 365

Key assumptions

• The Project Boundary is set as being all inclusive of the forest referenced by each Study Scenario, split by tenure type and, in the case of plantation forests, by Kyoto compliance category and rainfall zone. In reality, converting the abatement generated under the Study Scenarios to carbon credits would likely require the establishment of multiple Carbon Projects and this would reduce NPV’s through increasing the cost of carbon credit production (as a result of economy of scale effects). • For the purpose of estimating carbon credit production, the Project Boundary is assumed to reference only that proportion of carbon stocks included within live trees, dead trees and litter. Soil carbon stocks are not considered due to the difficulties in cost-effectively accounting these under Carbon Projects, the low confidence associated with modelled soil carbon stock outcomes for the Carbon Study, and the lack of an Approved Methodology for accounting soil carbon under the CFI. Wood product carbon effects are not included, as these presently are not considered for crediting under the CFI. Wood product replacement is considered under the leakage assessment provided here. • Revenues relate only to Carbon Project activities, with no consideration of revenues arising from extracted wood products. Estimates of future revenues are based on annualised carbon abatement figures, net of leakage and Risk Buffer allowances for the period 2013-2050. Future revenue estimates reference the carbon pricing scenarios described at Section 4. • Activity based leakage is captured by the carbon modelling process applied under the FCMF, as well as the assumption around Project Boundary described above, so no discount for activity based leakage is applied. • Market based leakage occurs for all Study Scenarios contemplating a reduction in extraction rates as compared with the Study Baseline (Section 4). Market Leakage Factors netted from total abatement arising from Study Scenarios:

o Reduced extraction in native forests: 8%. o Reduced extraction in plantation forests: 17%. • The Risk Buffer Factor netted for each Study Scenario has been varied depending on the recommended target market for carbon credits: o For Kyoto ACCU’s created under the CFI: 5%. o Where carbon credits will be traded into a combination of voluntary markets: 7.5%, being the average of CFI and estimated VCS requirements. o Where carbon credits will only be created under VCS: 10%. • In the case of Study Scenarios that involve cyclical forest harvesting, temporary increases in abatement against baseline are sometimes followed by decreases. In these cases, it is assumed that carbon credits would be generated only to the limit of the average carbon

Page 181 of 365

stock retained over the period 2013-2050, rather than to the peak stock. This approach aligns with the ‘average stocks’ concept reported by Baalman & Obrien (2006), and detailed within the CFI Farm Forestry draft Methodology. • It is assumed there is no constraint to sale of Kyoto ACCU’s (i.e. sales occur immediately after creation) under the ACPM. In the case of voluntary carbon credits, constraints are applied (refer ‘Price & market penetration – Voluntary markets). • A discount rate of 3% is applied. In the authors’ experience’, this is at the low end of rates typically applied for commercial Carbon Project investment decisions. However, the discount rate assumed here was felt justified due to the nature of the Study Scenarios (many of which reference crown land) and long-term nature of the abatement activities. In the published literature, discount rates of 1-10% have been previously applied around carbon forest projects (e.g. Polglase et al. 2008; Polglase et al. 2011; Macintosh 2012a). In limited testing applied under the Carbon Study, we found discount rate settings considerably less influential to NPV outcomes than carbon pricing assumptions and uncertainty around carbon stock estimates. It is recommended that the influence of discount rate be given a more detailed treatment in any future related studies. • Costs considered relate only to Carbon Project activities (see following section). Forest establishment, forest management and any harvest activities (where specified in a scenario) are not included. Carbon accounting, record keeping and administration costs are calculated on a per hectare basis, with a lower limit of $50,000 and upper limit of $450,000 applied (annual). • All figures are pre-tax.

Costs of market participation

There are a range of costs associated with seeking to participate within carbon markets. These include (but are not limited to) costs associated with: • developing an Approved Methodology with relevance to the forest management activity (where these do not already exist); • audits for the registration and verification of carbon abatement claims; • supporting the technical program and expertise required to conduct carbon accounting; • developing and maintaining forest and carbon accounting record keeping systems, noting record keeping requirements extend across decades; • creating carbon credits; • administration of program compliance and reporting obligations; and • sales, marketing and carbon credit trading activities.

An estimate of indicative costs against some of these items is presented in Table 41.

Page 182 of 365

Table 41 Schedule of costs (not exhaustive) associated with participating in carbon markets. Estimates are based on the authors’ experience.

Item Cost Frequency Audit (VCS registration) $200,000/audit At initial registration Audit (VCS verification) $50,000/audit Up to 5 yearly Audit CFI registration/report $25,000/audit Up to 5 yearly Brokerage fee on credit sales $0.1/credit At sale Carbon accounting & related technical support $6/ha/y* Annual Credit creation fees (VCS) $0.1/VCU created At creation Sales & marketing (voluntary market) 10% of sales revenue Annual Record keeping & project administration $3/ha/y** Annual For Study Scenarios: * Minimum $30,000, maximum $300,000. ** Minimum $20,000, maximum $150,000. Key assumptions: Reference Carbon Project is large (>10,000 ha) and carbon accounting involves a combination of in- office modelling and limited field based sampling. Costs can be expected to increase where these are not met.

There are economies of scale with Carbon Projects, so the larger the Carbon Project (by carbon volume), the less important these cost considerations become. Converting the abatement arising under the Study Scenarios to carbon credits would necessitate creation of a range of discrete Carbon Projects and Project Boundaries so that the costs detailed above will likely exert a stronger, downwards, influence on projected investment outcomes than represented in the simplified, whole-of-state by tenure estimates calculated here. Indeed, for small Carbon Projects, these costs can represent a significant proportion of total revenue and, in some cases, can mean small-scale abatement activities are not commercially viable to pursue.

Price & market penetration – Mandatory market (Australian Carbon Pricing Mechanism)

To gauge the potential scale and value of commercial opportunities around mandatory markets some estimate of future carbon pricing is required. The domestic targets referenced by the Australian Treasury’s (2011) Core Policy Scenario (Section 4) are consistent with those announced under the Australian Governments’ Clean Energy Future initiatives. For this reason, we reference carbon price forecasts (2015 and beyond) for the Core Policy Scenario as a possible Kyoto ACCU pricing trend: • Core Policy Scenario: $23/credit (2012-13), $24.15/credit (2013-14) and $25.40/credit (2014-15). Beyond 2014-15, the price paid for ACCU’s tracks as per Australian Treasury (2011) forecast (Figure 82).

Anecdotally, the authors’ experience is that there is an expectation amongst at least some market participants that prices will drop significantly following the fixed price period and may track close to any government imposed floor-price. For this reason, a second pricing scenario, referred to here as the Floor Price Scenario is referenced:

Page 183 of 365

• Floor Price Scenario: Beyond 2014-15, the price of ACCU’s drops to $15 in 2015-2016, thereafter increasing to 2050 by 4% per annum in real terms.

A very similar pricing scenario to the above has been previously reported by Macintosh (2012a).

With respect to the demand for ACCU’s, based on the review provided at Section 4, which shows demand for abatement well outstripping supply, it is assumed that there will be no demand-imposed constraint to selling the amount of Kyoto ACCU’s that could be produced from eligible forest management activities contemplated by the Study Scenarios.

140

Core Policy 120 Floor Price

100

80 Kyoto ACCU price ($) 60

0 2013 2023 2033 2043 Year Figure 82: The two pricing scenarios, Core Policy and Floor Price, applied to sales of Kyoto ACCU’s.

Price & market penetration – Voluntary market (CFI & VCS)

Assumptions relating to the development of voluntary markets, the price for carbon credits and the potential limits to sales within those markets are detailed below.

Market size

• The volume of carbon credits traded within the voluntary market will increase from 2011 levels (95 million) at 5% per annum to 2020 (noting this well exceeds the rate of change from 2010-2011, which was -28%, but falls well below the forward estimate of 63% reported in Peters-Stanley & Hamilton (2012)). Thereafter, increases in volume are assumed as 2% per annum, on the expectation that growth in the voluntary sector will slow as mandatory compliance mechanisms are increasingly implemented.

Page 184 of 365

Carbon Credit pricing

• Stable Price Scenario: Volume weighted average price is $5 per carbon credit in 2013, thereafter increasing in line with inflation (i.e. zero increase in real terms) to 2050 (noting this is lower than reported by Peters-Stanley & Hamilton (2012), but in line with the author’s experience of current market expectation). • Reduced Price Scenario: Volume weighted average price is $3 in 2013, thereafter increasing in line with inflation (i.e. zero increase in real terms) to 2050.

In relation to the Reduced Price Scenario, the authors consider that there is potential for the large volumes of carbon credits potentially arising from some of the Study Scenarios to exert downwards pressure on current reported global averages. In the authors’ experience, $3 is in line with current wholesale pricing for large volume transactions of voluntary carbon credits.

Market penetration (or sales potential)

Based on the review detailed at Section 4.15, the authors consider it highly likely that the supply of carbon credits into the voluntary market will continue to outstrip demand for many years to come and that this dynamic will act to constrain the volume of voluntary carbon credits that might be sold from Tasmanian forest Carbon Projects. Based on this assessment, the authors consider that it is unrealistic to expect that very large volumes of carbon credits could immediately be traded into the voluntary market at the time of creation. Instead, the following two market penetration scenarios are used to describe the assumed upper bounds to sales. • High Market Penetration scenario: A 0.2% share of the global market (by volume) during 2013, thereafter market share grows at 0.2% pa to a maintainable upper limit of 2%. • Low Market Penetration scenario: A 0.1% share of the global market (by volume) during 2013, thereafter market share grows at 0.1% pa to a maintainable upper limit of 1%.

While these figures may appear conservative on first read, in the authors’ experience, which includes an extended history of trading in voluntary markets, the challenges associated with actually selling large volumes of voluntary carbon credits are not to be underestimated. Achieving even the Reduced Market Penetration scenario would require a strategic and well funded marketing strategy. In context, achieving this level of sales would place the Project Proponent amongst the top operators globally. Both scenarios assume the Non-Kyoto Fund will allow for significant sales of Non-Kyoto ACCU’s extending beyond the proposed six year term.

Page 185 of 365

5.3 THE OPPORTUNITIES

Many of the Study Scenarios generate significant amounts of abatement. There are a range of options currently available, or potentially available over the short-term, for realising carbon credits and commercial opportunities from this abatement. A detailed review of these is provided through the remainder of this section.

Altered management regimes – Native forests

Of all of the Study Scenarios, some of those that reference reductions, or complete cessation, of harvesting in native forests result in the largest delivery of abatement (Table 42). In the case of N2A, for example, as much as 70-100 million tonnes of abatement (net of Market Leakage Factor and Risk Buffer Factor) may be available for the generation of carbon credits.

Table 42: Projected abatement for 2013-2050 by tenure, with leakage factor and risk buffer netted. Tenure/scenario Public Private Total (x 1 million t CO2e) (x 1 million t CO2e) (x 1 million t CO2e) N1. Reserve 572,000 ha, reduce cut from public native forests to 155,000m3 pa sawlog, 265,000 m3 pa peeler log. 81 – 93 N/A 81 – 93

N2. Phase out old growth harvesting by 2020, increase harvest in No net abatement regrowth forest to maintain 300,000m3 pa sawlog. 3 – 4 3 – 4

N2A. Phase out old growth harvesting by 2020, vary harvest in regrowth forest to maintain 155,000 m3 pa sawlog. 74 – 97 3 – 4 77 – 101

N3. Vary rotation length for harvest in regrowth from the Study Baseline (80 years) up to 120 years. 23 – 29 12 – 15 35 – 44

N4. Immediate cessation of old growth harvest (public & private), no increase in harvesting of regrowth. 19 – 22 9 – 11 28 – 33

N5. Immediate cessation of all native forest harvesting on public & private land. 96 – 111 77 – 88 173 – 199

Unfortunately, however, current policy settings and the structuring of the CFI (noting Section 4.9), mean that this abatement would presently not qualify for the creation of Kyoto ACCU’s, so that the target market for carbon credits is effectively limited, in this case, to voluntary markets.

Based on the assessment of eligibility criteria and methodologies detailed at Section 4, there does appear to be potential to verify abatement derived from these activities as Non-Kyoto ACCU’s under the CFI, or as VCU’s under the Verified Carbon Standard. With respect to the CFI, one of the key challenges to address is that no Approved Methodology presently exists that relates to the activity contemplated by this set of Study Scenarios. For VCS, this is less of an issue since several relevant Approved Methodologies are available.

Another key issue, that applies to consideration under both CFI and VCS, is demonstrating the case for additionality. None of the native forest related Study Scenarios is presently listed on

Page 186 of 365

the CFI Positive List, meaning that the case for additionality still needs to be proved under the CFI and that successful application for additions to the Positive List would likely need to occur prior to Carbon Projects being developed around these activities.

In demonstrating additionality, it will be necessary to show that the change in forest management practice represents a departure from the business as usual, or common practice, scenario. This should be readily demonstrated for Study Scenarios N4 & N5, which reference major shifts in forest management.

For N1 & N2A, demonstrating additionality will likely require a detailed comparison of the changed management practice against historic and contemporary trends and may be complicated if there is a pre-existing trend toward reduced extraction. In this regard, the nature of policy initiatives such as the Tasmanian Forests Intergovernmental Agreement (Tasmanian Government, 2011), which proposes the introduction of legislation to expand forest reserve areas and set minimum harvest volumes, would likely be reviewed carefully by independent verifiers to ensure they did not impact the case for additionality.

With specific reference to the CFI, provisions are included within the CFI Act (Sec. 41(1b)) that deal with situations where ‘offsets projects’ are proposed around activities that are legislated, stating that projects have the potential to pass the additionality test where: “the project is not required to be carried out by or under a law of the Commonwealth, a State or a Territory” (p.58) Additionally, the CFI Regulations state (Sec. 3.36(a)) that the “following kinds of project are excluded offsets projects: (a) a project that: (i) was mandated under a law of the Commonwealth or a State or Territory. (p. 46) This is potentially a significant issue for recognition under the CFI of any abatement arises from reserves being established through legislative instruments, and this would need to be successfully dealt with where recognition under the CFI was to be contemplated. This might be easier where it could be successfully demonstrated that relative legislative instruments were used to ensure the permanence of changes in management activities and carbon stocks arising from these, rather than creating a mandatory requirement for the management activity to change. It is recommended that expert legal advice be sought around these issues and the potential for structures proposed under the Tasmanian Forests Intergovernmental Agreement (Tasmanian Government, 2011) to impact the case for CFI recognition.

For N3, it is likely that a detailed review of historic management records would be required to evidence the baseline practice and to demonstrate the case for additionality. While these have not been reviewed under the Carbon Study, it is assumed that relevant records would be readily available and, therefore, that demonstration of additionality would be relatively straight-forward.

Page 187 of 365

One issue that is of particular importance to Study Scenario N3 is that the wording of the CFI Act (Sec. 27 (j)) may exclude activities where harvesting of native forests is conducted within the proposed Project Boundary. As previously detailed under Section 4.6, the authors understand that it was not the original intent of the legislation to exclude Improved Forest Management activities from the CFI. Nevertheless, this is noted here as a potential barrier to participation under the CFI for Study Scenario N3. In any event, abatement occurring under this scenario could be brought to market via the VCS.

It should also be noted that this issue (Sec. 27(j)) would affect Project Boundary setting for N1 & N2A (so as to exclude harvesting), in which case the activity based leakage assumptions applied here would not hold (i.e. activity based leakage would likely need to be accounted).

Since abatement under a Carbon Project around this set of Study Scenarios is likely to be estimated relative to a Baseline Scenario, accurately defining the Baseline Scenario, including the carbon stocks that occur under it, is critical. Given the complexity of the forest systems considered, this is a technically challenging process and interpretation of the Baseline Scenario may be challenged by independent verifiers. The authors’ understand that the DCCEE is presently considering how baselines might be developed in native forest situations and it should be noted that the Study Baseline applied here may not necessarily be that which would apply under the CFI, or for situations where a Carbon Project was established at the less than whole- of-state by tenure level contemplated by the Study Scenarios.

On balance, we consider that, while a number of important issues would need to be addressed and this would require time and resources, it is likely that Carbon Projects could be successfully developed around those native forest related Study Scenarios that deliver net abatement. A breakdown of market options, project structuring, eligibility considerations and potential value range for a successful Carbon Project is provided at Insert 5.

Based on the NPV estimates presented here, positive returns are associated with establishing Carbon Projects around most of the native forest management Study Scenarios considered, with the highest NPV’s associated with N1, N2A and N5. Again, these NPV estimates should be viewed in the context that they do not consider opportunity costs associated with reduced timber product extraction and, therefore, should be interpreted as being associated with establishing a Carbon Project, rather than doing nothing, following a pre-existing decision to reduce extraction rates (see Section 5.2). Net present value estimates are negative to marginal in one instance (private forest holdings under N2A), and positive but marginal in one other instance (private forest holdings under N2) particularly where considered on a per hectare basis.

Page 188 of 365

Insert 5. Assessment of market opportunities for native forest related Study Scenarios (N1-N5), including key considerations for Carbon Project establishment and potential value range where carbon credits are successfully created and traded.

N1: Reservation of 572,000 ha of public forest, reduction of annual sawlog and peeler log cut.

Markets (current policy settings) Mandatory, CFI Kyoto ACCU’s: No. Domestic voluntary, CFI Non-Kyoto ACCU’s: Feasible, but note eligibility considerations. International voluntary, VCS: Likely. Proposed project structuring • Seek registration and credit realisation under VCS (targeting international markets) and CFI (targeting domestic market, including Non-Kyoto Fund). Leakage Reduction in extraction volumes. Market leakage factor: 8%. Risk buffer applied 7.5% (average of CFI and VCS requirements) Eligibility considerations • Activity not listed on CFI Positive List. • No methodology approved under CFI (one draft is proposed). • May be difficult to prove that change in forest management regime represents a shift from the business as usual, or common practice, if there is an existing trend towards reserving and reduced extraction. NPV range Project ($ million) Per hectare ($/ha) Public 84 – 243 107 - 310 [132 – 262]* [168 – 334] * Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained.

N2: Phase out old growth harvesting by 2020 on public and private land, vary harvesting in re-growth aiming to maintain annual sawlog production from public forest at 300,000 m3/year.

Markets & project structuring As for N1. (current policy settings) Leakage No reduction in extraction volumes, zero market leakage assumed. Risk buffer applied 7.5% (average of CFI & VCS buffer requirements) Eligibility considerations As for N1. NPV range Project ($ million) Per hectare ($/ha) Private (3)** – 5 (3) – 6 [(3) – 5] (3) – 6 * Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained.

Page 189 of 365

N2A: Phasing out old growth harvest by 2020 on public and private land, harvesting re-growth forest aiming to maintain annual sawlog production from public forest at 155,000 m3/year.

Markets & project structuring As for N1. (current policy settings) Leakage Reduction in extraction volumes. Market leakage factor: 8%. Risk buffer applied 7.5% (average of CFI & VCS buffer requirements) Eligibility considerations As for N1. NPV range Project ($ million) Per hectare ($/ha) Public 84 – 249 107 – 318 [120 – 271]* [153 – 346] Private (3)** – 5 (3) – 6 [(3) – 5] (3) – 6 * Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained. ** Values shown in parentheses are negative.

N3: Varying rotation length in regrowth forest from 80-120 years.

Markets (current policy settings) Mandatory, CFI Kyoto ACCU’s: No Domestic voluntary, CFI Non-Kyoto ACCU’s: Feasible, but significant challenges arising from harvesting of native forests. International voluntary, VCS: Likely. Proposed project structuring Seek registration and credit realisation under VCS, targeting both international and domestic markets. Leakage Reduction in extraction volumes. Market leakage factor: 8%. Risk buffer 10% (VCS requirement) Eligibility considerations • The wording of the CFI Act may currently preclude this type of project structuring (IFM where harvesting continues within the Project Boundary). See Section 4.6 for further detail. • Otherwise, considerations are as for N1. NPV range Project ($ million) Per hectare ($/ha) Public 30 – 77 70 – 177 [32 – 78]* [75 – 182] Private 10 – 32 30 – 94 [10 – 32] [30 – 94] * Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained. ** Values shown in parentheses are negative.

Page 190 of 365

N4: Immediate cessation of old growth harvesting (public/private), with maintenance of regrowth harvest.

Markets & project structuring As for N1. (current policy settings) Leakage Reduction in extraction volumes. Market leakage factor: 8%. Risk buffer 7.5% (average of CFI & VCS requirements) Eligibility considerations As for N1. NPV range Project ($ million) Per hectare ($/ha) Public 28 – 68 36 – 86 [30 – 70]* [39 – 90] Private 6 – 23 7 – 27 [6 – 23] [7 – 27] * Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained.

N5: Immediate cessation of all native forest harvesting (public and private).

Markets & project structuring As for N1. (current policy settings) Leakage Reduction in extraction volumes. Market leakage factor: 8%. Risk buffer applied 7.5% (average of CFI & VCS requirements) Eligibility considerations As for N1. NPV range Project ($ million) Per hectare ($/ha) Public 84 – 281 107 – 358 [173 – 352] [221 – 449] Private 84 – 234 99 – 276 [121 – 251] [143 – 295] * Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained.

Altered management regimes - Plantations

While the FCMF outputs indicate that large net-abatement amounts are possible through altering the management regimes applied to plantations (Table 43), and that a significant proportion of this abatement is Kyoto compliant, there are a number of significant challenges associated with bringing this carbon to account under the CFI, or VCS.

Page 191 of 365

Table 43: Projected abatement for period 2013-2050, net of leakage and risk buffer allowances, by tenure, Kyoto eligibility (current policy settings) & rainfall (Kyoto compliant abatement only).

Tenure/scenario Kyoto <600mm Kyoto ≥600mm Non-Kyoto Total (x1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) Public

One challenge common to all of the altered plantation management scenarios is demonstrating that the change in management regime is additional. In general terms, the case for additionality is strongest where management practices are substantially altered from business as usual practice and it can be demonstrated that, in the absence of expected carbon revenues, the change in practice is commercially marginal (i.e. the change is not already highly financially attractive in the absence of carbon related revenue). With specific reference to the CFI, none of the activities contemplated by Study Scenarios P1 - P3 are presently listed on the Positive List, meaning that any prospective Project Proponent would likely first need to seek to establish additionality through applying for an amendment to the Positive List.

On balance, our opinion is that the case for additionality is relatively weak for Study Scenarios P1 & P3, which in the case of P1 contemplate a small change in practice (10% increase in rotation length), or in the case of P3, already align with commercial drivers under business as usual practice (i.e. increasing plantation growth rates is commercially attractive even where carbon is not considered). For this reason, these are not considered further here as possible carbon-related commercial opportunities. We wish to stress, however, that this is a preliminary opinion not informed by a detailed review of forest management records, or commercial financial models retained by plantation owners, so that the case for additionality may be considerably improved where these are referenced. We also note that, reasonably significant volumes of potentially high-value Kyoto compliant abatement could be generated through these two Study Scenarios, so that a more detailed assessment of carbon-related commercial options would be warranted where these practices were being seriously contemplated by forest managers.

Page 192 of 365

In relation to Study Scenario P2, we consider this represents a significant change to business as usual practice and, therefore, that the case for additionality is relatively strong. Arguments for additionality would be further supported where it could be demonstrated that the commercial viability of the proposed management change is dependent on anticipated carbon revenues. With respect to CFI, application would need to be made to include this on the CFI Positive List.

The CFI specifically excludes participation of plantations established under a Managed Investment Scheme (MIS) structure. Given that MIS projects have been a major driver of plantation establishment within Tasmania, particularly for the period 1995 – 2008 (e.g. Board of Private Forests Tasmania, 2005), this likely precludes a significant proportion of the abatement arising from this set of Study Scenarios from being brought to account under the CFI. This issue has not been considered in the NPV calculations presented at Insert 6, since detailed statistics around MIS plantation have not been available to the Carbon Study, so it has not been possible to net abatement derived from plantations developed under MIS structures. Where this was considered, project-level NPV’s would be reduced.

An issue common to all plantation management and expansion scenarios is that there are barriers to realising carbon credits from abatement occurring in plantations located in areas where annual rainfall is equivalent, or greater than, 600mm (≥600mm). The CFI prescribes that tree plantings occurring in these higher rainfall areas must either meet the definition of an Environmental Planting (a mix of endemic species), meet salinity mitigation criteria, be deemed to be in an area where the relevant state agency is adequately managing water interception by plantations, or where the Project Proponent already owns water rights close to equivalent to predicted water interception. In the authors’ experience, gained from operating in other landscapes, these requirements can, at best, complicate and delay Carbon Project development or, at worse, render proposed projects unviable.

For Study Scenario P2, Kyoto compliant plantations located in the <600mm rainfall zone would likely represent the least challenging option for potential creation of Kyoto ACCU’s under the CFI, but these appear very limited within the existing estate. For Kyoto compliant plantations located at ≥600mm, a situation which again appears limited within the existing estate, the path to recognition under CFI would be considerably more difficult and may ultimately prove unachievable. In both cases, abatement arising from plantations established under MIS structures would be excluded from the CFI.

Based on the NPV estimates presented at Insert 6, positive returns are associated with establishing Carbon Projects around non-Kyoto compliant abatement arising from the privately owned plantation estate. Negative returns are associated with structuring projects around the relatively small amount of non-Kyoto compliant abatement arising in publicly owned forests.

Page 193 of 365

Insert 6. Assessment of market opportunities for plantation management related Study Scenarios (P1-P3), including key considerations for Carbon Project establishment and potential value range where carbon credits are successfully created and traded.

P1: Increase rotation length of all plantations by 10%

Markets (current policy settings) None considered here due to potential additionality issues. Leakage Reduction in rate of extraction. Market leakage factor: 17%. Eligibility considerations • Additionality, baseline and change from business as usual likely highly challenging to satisfactorily evidence given the limited shift in management practice. • CFI considerations as for P2.

P3: Increase plantation growth rate by 25% through improved management.

Markets (current policy settings) None considered here due to potential additionality issues. Leakage None. Eligibility considerations • Likely to be significant challenges in establishing the case for additionality since increasing plantation growth rates may be considered established common practice for industrial plantations. • CFI considerations as for P2.

P2: Convert all pulplog plantations to sawlog plantations (increase rotation length to 25 years).

Markets (current policy settings) Domestic, CFI: Feasible for forest occurring in areas where rainfall <600mm, but highly challenging in higher rainfall areas. International voluntary, VCS: Likely for non-Kyoto abatement. Proposed project structuring • Seek registration and credit realisation of Non-Kyoto abatement under VCS (targeting international markets). • Kyoto abatement occurring at ≥600mm not considered here due to current CFI eligibility considerations (see below). Leakage factor applied Reduction in extraction volumes. Market leakage factor: 17%. Risk buffer applied Non-Kyoto abatement: 10% (VCS requirements) Kyoto abatement: 5% (CFI requirements) Eligibility considerations • Not CFI eligible wherever plantations have been established under MIS structures. • High level of difficulty registering plantings of this type (historically established industrial plantations intended for harvest) under CFI, particularly where annual rainfall is ≥600mm. • Not listed on CFI Positive List. • No methodology approved under CFI (one draft is proposed).

Page 194 of 365

• Case for additionality will be strongest where it can be demonstrated change in management practice would not have resulted in a more profitable outcome in the absence of carbon revenues. NPV range Project ($ million) Per hectare ($/ha) Public, Non-Kyoto (8) – (6) (139) – (97) Private, Non-Kyoto 13 – 64 105 – 508 [32 – 84]** [255 – 665] * Values shown in parentheses are negative. ** Values in square brackets are NPV calculated where sales into voluntary markets are unconstrained.

Altered land use

Based on the FCMF outputs (Table 44), there is significant potential to generate abatement, including Kyoto compliant abatement, through expanding the plantation estate (Study Scenarios A3 & A4), with net abatement exceeding 20 million t CO2e. Additionally, around 1 million t CO2e of avoided emissions, which would likely be Kyoto eligible, could be realised if the deforestation contemplated by Study Scenario A1 (ca. 1,800 ha to 2015) was avoided through Carbon Projects.

Table 44: Projected abatement for period 2013-2050, net of risk buffer allowances, by tenure, Kyoto eligibility (current policy settings) and rainfall (Kyoto compliant abatement only).

Tenure/scenario Kyoto <600 Kyoto ≥600 Non-Kyoto Total (x 1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) (x1 million t CO2e) Public

A2. Unstocked areas in native Uncertain Uncertain Uncertain <0 forests are regenerated. Private

A1. Conversion of native forest ~1* to non-forest. A2. Unstocked areas in native Uncertain Uncertain Uncertain ~0 forests are regenerated. A3. Expansion of plantations 4 – 5 18 – 24 ~0 22 – 29 at average rate. A4. Expansion of plantations at 4 – 6 19 – 25 ~0 23 – 30 double rate. * Study Scenario A1 includes emissions arising from deforestation (ca. 1,800 ha to 2015). The values indicated here are the abatement estimated to arise if Avoided Deforestation Carbon Projects could be established so as to avoid this deforestation. In this case, abatement would by Kyoto eligible and assumed likely to occur in privately owned forest.

Many of the challenges identified in realising carbon credits from the altered plantation management scenarios (P1 - P3) also apply to the altered land use scenarios, including additionality, MIS and rainfall considerations. However, under plantation expansion scenarios, these may be mitigated to some extent through designing future plantings with a view to aligning with key CFI structures and eligibility requirements.

Page 195 of 365

For A1 (where Avoided Deforestation Carbon Projects were contemplated), the case for additionality and eligibility would be critically dependent on being able to demonstrate that, in the absence of a Carbon Project, a specific forest area was likely be subject to deforestation (permanent) and that this would not occur under the project scenario. Within the Australian context, evidencing this has typically required extinguishment of some pre-existing right to deforest, such as land clearing permits. In Tasmania, the Forest Practices Authority is responsible for the issue of clearing permits. While a detailed review has not been performed, the authors understand that a number of permits have been issued in recent times.

For A3 and A4, demonstrating additionality may be complicated by the underlying trend for plantation expansion within Tasmania35, since periods of expansion have occurred historically and A3 is based on the historical average trend (Forest Practices Authority data). The case for additionality will be strengthened where it can be demonstrated that carbon abatement considerations, including associated revenue, contributes strongly to any future expansion trend.

With specific reference to the CFI, none of the activities contemplated by this set of Study Scenarios is presently listed on the Positive List. However, a draft Methodology is presently under review by the DOIC that is applicable to for-harvest plantation systems and it is likely that, in parallel, a review will be undertaken of the potential for this activity to be included on the Positive List. If both Methodology and application for inclusion to the Positive List were accepted, this would pave the way for CFI participation for Study Scenarios A3 and A4.

As for the plantation management scenarios, there are barriers to realising carbon credits under A3 and A4 from Kyoto compliant abatement occurring in areas where annual rainfall is ≥600mm. Again, these would need to be successfully dealt with before abatement from plantations established in this higher rainfall zone could be recognised under the CFI. Additionally, since plantings established under MIS projects are excluded from the CFI, the abatement arising from Study Scenarios A3 and A4 could only be fully recognised under the CFI where expansion was not driven by MIS.

On balance, the authors consider that, while a number of important issues would need to be addressed, it is likely that Carbon Projects could be successfully developed around abatement generating Study Scenarios A3 - A4. Additionally, if the deforestation occurring under A1 could be avoided, then it would be feasible to establish an Avoided Deforestation Carbon Project around this activity. A detailed breakdown of market options, project structuring, eligibility considerations and potential value range for successful Carbon Projects is provided at Insert 7.

Based on the NPV estimates presented at Insert 7, positive returns are associated with establishing Carbon Projects around the Kyoto-compliant abatement arising from Avoided

35 These comments are made with reference to plantations established for the dual purpose of timber and carbon credit creation. The case for additionality is considerably less complicated for not-for-harvest, ‘carbon-only’ plantations. Establishing plantings of this type in the <600mm rainfall zone is an activity that is already included on the Positive List.

Page 196 of 365

Deforestation (sub-set of A1). Positive returns are also associated with establishing Carbon Projects around Kyoto abatement arising from the privately owned plantation estate under plantation expansion scenarios A3 & A4. On a per unit area basis, these represent the largest projected returns for any of the Study Scenarios and, while the challenges associated with realising Kyoto ACCU’s for plantation areas occurring at ≥600mm rainfall are substantial, structuring Carbon Projects around these activities would need to be given serious consideration due to the significant potential for upside.

Insert 7. Assessment of market opportunities for plantation management related Study Scenarios (A1-A4), including key considerations for Carbon Project establishment and potential value range where carbon credits are successfully created and traded. A1: Conversion of native forest to non-forest uses: Where an Avoided Deforestation Carbon Project is established around halting the deforestation activity.

Markets (current policy settings) Mandatory, CFI Kyoto ACCU’s: Possible. International voluntary, VCS: Likely. Proposed project structuring • Seek registration and Kyoto ACCU realisation under CFI, targeting domestic mandatory market. Risk buffer applied Kyoto compliant: 5% (CFI requirement) Eligibility considerations • Would likely require extinguishment/surrender of land-clearing permits. • Presently not listed on the CFI Positive list. • No CFI Approved Methodology. NPV range* Project ($ million) Per hectare ($/ha) Private (Kyoto compliant)** 14 – 31 7,489 – 16,651 * In this case, carbon credit realisation associated with avoided emissions amounts are assumed to be realised over the first 20 years. ** It is assumed deforestation activities will largely be confined to privately owned forest areas.

A3: Expansion of plantation estate at average rate.

Markets (current policy settings) Mandatory, CFI Kyoto ACCU’s: Feasible for Kyoto eligible forest occurring at <600mm, but note eligibility considerations. Likely highly challenging for forest occurring at ≥600mm, but may be feasible where designed with consideration of additionality and features of the CFI. International voluntary, VCS: Appears unlikely in short-term, no methodology yet available. Not possible to realise Kyoto compliant abatement due to international reporting and obligations under Kyoto protocol. Proposed project structuring • Seek registration & Kyoto ACCU realisation under CFI for Kyoto abatement, targeting domestic mandatory market.

Page 197 of 365

Risk buffer applied Kyoto compliant: 5% (CFI requirement) Eligibility considerations • Where plantations are for-harvest, case for additionality likely strongly dependent on financial additionality. This most readily demonstrated where high-value Kyoto ACCU’s can be produced from otherwise commercially marginal plantations. • Case for additionality may be weakened by virtue of rate of expansion tracking at historical regional average (i.e. the baseline expansion rate prior to introduction of ACPM). • CFI eligibility considerations as for P2. Under a plantation expansion scenario (as opposed to management of existing plantations), it may be possible to mitigate some of the CFI eligibility challenges identified for plantings located in ≥600mm rainfall zone through careful planting design and project structuring. • No relevant methodologies approved under VCS/CDM. NPV range Project ($ million) Per hectare ($/ha) Private, Kyoto <600mm 60 – 171 3,457 – 9,881 Private, Kyoto ≥600mm 269 – 759 3,526 – 9,949

A4: Expansion of plantation estate at double the average rate (to maximum 10% cleared land by 2050).

Markets & project structuring As for A3. Risk buffer As for A3. Eligibility considerations • As for A3, except case for additionality may be stronger where above regional average expansion rates can be linked to carbon revenues as driver of expansion. NPV range ($) Project ($ million) Per hectare ($/ha) Private, Kyoto <600mm 63 – 156 3,218 – 7,966 Private, Kyoto ≥600mm 285 – 695 3,304 – 8,052 * Values shown in parentheses are negative.

5.4 KEY RISKS

There is technical and financial risk associated with structuring Carbon Projects around any of the Study Scenarios. These include (but are not limited to) the following.

• Failure to meet key emissions program eligibility criteria and to realise carbon credits from the forest management activity. • Failure to meet standards applied by independent verifiers, potentially resulting in a slowing, reduction, or halt to carbon credit creation and/or increased audit costs.

Page 198 of 365

• Overestimation of claimable abatement amounts, carbon prices and sales capacity. • Underestimation of costs associated with developing and managing the Carbon Project. • Incorrect, or invalid, interpretation of the Baseline Scenario, so as to over-estimate claimable abatement. • Significant reversal events (e.g. wildfires) impacting the Project Proponents’ capacity to generate carbon credits or, in a worst case situation, having to ‘buy-back’ and acquit carbon credits. In this regard, the authors note that a number of the forest types considered by the Carbon Study are fire sensitive and may require active management intervention to restore carbon stocks (see Section 4.6 for CFI considerations). • Changes to emissions reduction related policy (Section 5.5) and, in particular, changes to compliance obligations, pricing mechanisms, emissions/carbon accounting requirements, and Carbon Project eligibility requirements.

In considering investing in any Carbon Project, these risks should be carefully considered and, where possible, management plans should be implemented that seek to mitigate them. In the case of reversal events, the authors note that insurance products now exist within the marketplace that can be used to limit exposure in some circumstances.

5.5 POLICY INFLUENCES

Changes to government policy have the potential to greatly alter the outcomes of the opportunity assessment presented here, both positively and negatively. Historically, the development and implementation of emissions reduction policies around the world, including those that related to verification and carbon credit realisation programs, has followed a highly unpredictable pathway.

Policy setting in Australia has been dynamic, with the first discussion papers relating to introduction of a national emissions trading program released in 1999 (AGO, 1999a-c); introduction of a state based emissions trading program in 2004 (NSWGGAS) which ceased in July 2012 leaving many Project Proponents with untradeable carbon credits; two failed attempts to introduce a national carbon trading program from 2006-2009; wind-up on short notice of the Greenhouse Friendly program in 2009, leaving a number of large abatement projects with no option for carbon credit realisation before legislation underpinning the CFI and ACPM were finally successfully introduced in 2011, an outcome that still attracts a high level of political and public debate.

Given the high rate of policy evolution in this area, as well as the intense, ongoing domestic and international debate around the most appropriate structures for dealing with emissions reductions, any predictions around future policy development are speculative and, at best, can only be represented as informed opinion. For this reason, the Carbon Study has focused on opportunities associated with existing policies and programs, and has not attempted to provide a detailed review of the implications of the many, equally possible future policy pathways that

Page 199 of 365

could be inferred at this pivotal point in time. However, two key policy influences are dealt with below that warrant close monitoring over the next 12-18 months since they have the potential to significantly alter the outcomes of the assessment provided here.

Kyoto Protocol & Article 3.4

At the Durban Climate Conference (UNFCCC, 2011), new rules were agreed to govern the way that emissions associated with forest management will be accounted and internationally reported. After the completion of the first Kyoto agreement period in 2012, a second commitment period covering the years 2013-2020 may be entered by countries willing to participate. Unlike the current commitment period, under the new agreement, it will likely be compulsory to fully account for emissions associated with management of existing native forests and plantations.

The Australian government has not yet announced its position on participating under the second commitment period. If it were to participate, the specifics of the approach to accounting carbon stocks in native forests would need to be agreed internationally and then implemented. It is likely that a considerably more detailed accounting approach would be required, potentially including spatially explicit modelling of emissions associated with forest management, harvesting and fire in much the same way that the FCMF has been applied to the Carbon Study. However, it remains uncertain exactly which emissions associated with fire and other activities will need to be reported and the authors understand the accounting methodology is still under development by the DCCEE.

Macintosh (2011) detailed one possible scenario for native forest carbon accounting under a second Kyoto commitment period, assuming that Australia actually elects to participate (Mackintosh 2012)36, which would essentially see some forms of abatement occurring in native forests being recognised as Kyoto compliant units and potentially include the issue of Kyoto ACCU’s under the CFI. If this policy scenario were to play out in this way, it would be hugely favourable to commercial opportunities around the set of native forest related Study Scenarios as it would greatly increase the value that could be derived from the abatement. However, beyond the uncertainty around whether Australia will actually participate in the second commitment period, there are a number of other issues that make this a highly uncertain outcome, some of which have been detailed by Macintosh (2011, 2012b), including:

• the specific approach to carbon accounting in native forests under the second commitment period is yet to be developed, or agreed to by Australia;

36 http://www.crikey.com.au/2012/06/12/kyoto-protocol-will-gillard-stay-or-go/

Page 200 of 365

• it is possible that caps may be applied to the amount of forest abatement that could be claimed internationally, thereby strongly limiting the capacity to issue Kyoto units in relation to native forest management; and • the Australian government may not to pass on the benefit of abatement in native forests to forest owners through allowing the issue of Kyoto ACCU’s via the CFI. In relation to the latter point, it is worth noting that significant volumes of Kyoto compliant forest sequestered carbon has already been reported internationally by the Australian government, without any benefit flowing to the forest owners and that this situation continues under the CFI (noting, for example, barriers to entry for many Kyoto compliant plantation forest activities and the lack of retrospective crediting for sequestration occurring prior to 1st July 2010).

Nevertheless, it is recommended that developments in this area be monitored closely over the coming 18 months and that the assessment provided here be reviewed should an alternate treatment of native forest abatement emerge.

Direct action plan

The Coalition party has released an environment and climate change policy statement referred to as the ‘direct action plan’ (Australian Liberal Party, 2010). Under the direct action plan, the Coalition commits to meeting an emissions reduction target of 5% against 1990 levels by 2020. However, unlike the current governments approach, the direct action plan advocates establishment of an ‘emissions reduction fund’ in place of a ‘carbon tax’ (and presumably by inference, a market-based carbon pricing mechanism). The fund, which is envisaged to be valued at an average of $1.2 billion per annum to 2020, will be used to fund activities that seek to deliver emissions abatement, including sequestration of carbon in soils (this being a key focus of the direct action plan), tree planting and renewable/clean energy programs. Projects would be selected for funding through an open tender process.

Should a policy platform, along the lines of the direct action plan, be introduced, the nature of the opportunities detailed under the Carbon Study would change significantly. It may be that creation and sale of carbon credits into a mandatory market would no longer be possible and, instead of seeking to trade into voluntary markets, the larger commercial opportunity might lie in targeting projects at the direct action fund through the tender process. What options would exist for Tasmanian forests to participate are unknown, but references to tree planting activities within the direct action plan are at least a positive signal for reforestation activities and, perhaps, forest management more broadly.

What the real specifics of direct action would look like, should it come to pass, can presently only be speculated. Again, it is recommended that developments in this area are monitored closely over the coming 18 months and that the assessment provided here be reviewed in light of any move to alter the features of the ACPM and/or to introduce a funds-based program along the lines of the direct action plan.

Page 201 of 365

6 FOREST CARBON MODELLING FRAMEWORK

This section provides a detailed explanation of the process used to estimate carbon stocks, including coverage of: • the system used to classify the vegetation types included in the Carbon Study, • the approach to modelling carbon stocks and stock changes spatially, • a description of the overall modelling framework, • details of specific processes used to model: stand growth, mortality, turnover and decomposition, soil carbon, wood products, fire, and timber harvesting, • the approach used to estimate uncertainty in results, and • key assumptions and knowledge gaps

6.1 CLASSIFICATION OF TASMANIA’S FORESTS

Forest group classification

The approach used to classify Tasmania’s forests was based on the Forest Group System used by Forestry Tasmania and DPIPWE, which references seven different vegetation types: • Rainforest (RFT), • Tall Eucalypt Forest (ETF), • Low Eucalypt Forest (ELF), • Other Native Forest (ONF), • Hardwood Plantations (PHW), • Softwood Plantations (PSW), and • Non-forest (NON).

These are described in Section 2.1. Because the Carbon Study included all vegetation >2m height and >20% cover, the list was expanded to include non-forest vegetation >2m and was also modified to include agricultural land that could potentially be planted (Grassland).

Forest groups were mapped across Tasmania for 1996, 2001 and 2010 using layers provided by Forestry Tasmania (example shown at Figure 83). These data were supplemented with data from TasVeg to identify areas of agricultural grassland, non-forest native vegetation >2m as well as non-forest native vegetation <2m.

Page 202 of 365

Figure 83: Map of forest groups across Tasmania. Data provided by Forestry Tasmania and processed by CO2 Australia.

Page 203 of 365

A large proportion of vegetation across Tasmania is classed as non-forest according to the Forest Group and Forest Class systems. Although much of this vegetation includes mooreland, alpine areas and buttongrass plains, a large extent includes vegetation >2m. Since the objective of the Carbon Study was to model carbon content in all vegetation >2m, it was important to identify and assign carbon contents to this land.

To identify areas of vegetation > 2 m not included in the forest layers, the Forest Class and Forest Group classifications were overlaid by the TasVeg layer. Areas classed as non-forest under the Forest Class system (FC81 and FC91) that included vegetation types typically >2m were included in a new layer called Tall Native Vegetation. Based on information from DPIPWE, vegetation types classed as non-forest but >2m tall with >20% cover in this layer were assumed to include: • Highland treeless vegetation o Alpine coniferous heathland • Non-eucalypt forest and woodland o Bursaria - Acacia woodland and scrub o Subalpine Leptospermum nitidum woodland • Rainforest and related scrub o Rainforest fernland o Northofagus gunnii Rainforest and scrub • Scrub, heathland and coastal complexes o Acacia longifolia coastal scrub o Broad-leaf scrub o Coastal scrub on alkaline sands o Coastal scrub o Dry scrub o Leptospermum scrub o Melaleuca squarrosa scrub o Melaleuca pustulata scrub o Riparian scrub o Queenstown regrowth mosaic o Western wet scrub o Western subalpine scrub In addition, any non-forest areas from the Forest Group or Forest Class classification, but classified forest in the TasVeg Layer were added to the Tall Native Vegetation class. These TasVeg classes included: • Dry eucalypt forest and woodland • Wet eucalypt forest and woodland

Page 204 of 365

• Non-eucalypt forest and woodland • Rainforest and related scrub

Forest Classes

While the Forest Group classification system is useful for distinguishing broad forest types, it provides little information on stand age, growth, structure, or size. Since these are critical for modelling forest carbon, a more detailed classification system was required.

Forestry Tasmania’s Forest Class System provided a suitable classification for this purpose. Developed originally for quantifying the volume of wood available for harvest across Forestry Tasmania’s estate, it contains much of the information needed to be able to estimate total carbon stocks with some degree of accuracy. It also sits within the Forest Group system with each Group consisting of one or more forest classes. Thus, the Forest Class system was used as the primary means for modelling forest carbon stocks spatially for the Carbon Study.

The system consists of 90 Forest Classes (Table 45). It is primarily focussed on eucalypt forest with only four classes covering non-eucalypt native vegetation. Each eucalypt forest class is linked to inventory plots randomly located across Forestry Tasmania’s estate which provide detailed information on tree growth and debris. In addition, each class provides information on stand structure, productivity, current height and, for some classes, age.

Unfortunately, the only state-wide classification of Tasmania’s forests using the Forest Class System was undertaken in 1996. Subsequent classifications in 2001 and 2010 covered public forest only. For the Carbon Study, changes in cover of forests on private land were derived by combining the Forest Group layers (which cover all of Tasmania) with the 1996 Forest Class layer. This allowed identification of areas associated with clearing, or conversion of native forest to plantations as well as establishment of new plantings on grassland. Since Rainforest and Other Native Forest each include just two forest classes, changes in these could be readily derived from the Forest Group layer. Where there was no change in forest group over time, it was assumed that the forest class remained constant.

The resultant layers for 1996, 2001 and 2010 included all 90 forest classes across the whole of Tasmania. An additional eight classes were generated to cover the following: • Non-forest native vegetation >2m, • Non-forest native vegetation <2m, • Agricultural grassland, • Water, • Mud, rock and sand, • Urban areas, and • Other non-plantable, non-forest areas.

Page 205 of 365

Table 45: Forest classes used by Forestry Tasmania to classify native forests and plantations.

Forest class Mature trees Regrowth Forest class Mature trees Regrowth Potential Potential No. Code Height Cover Height Year No. Code Height Cover Height Year height height m m m m m m Mature eucalypts without regrowth Regrowth eucalypt with silvicultural regeneration 1 E1a&b >55 >40% - - - 46 ER/1&2 - - - >41 - 2 E1c&d >55 5-40% - - - 47 ER/+3 - - - 34-41 - 3 E2a&b 41-55 >40% - - - 48 ER/-3 - - - 27-34 - 4 E2c&d 41-55 5-40% - - - 49 ER/4&5 - - - <27 - 5 E+3a&b 34-41 >40% - - - Even-aged silvicultural eucalypt regeneration 6 E+3c&d 34-41 5-40% - - - 50 <=1959/1&2 - - - >41 <=1959 7 E-3a&b 27-34 >40% - - - 51 <=1959/+3&X - - - 34-41 <=1959 8 E-3c&d 27-34 5-40% - - - 52 <=1959/-3 - - - 27-34 <=1959 9 E4a,b&c 15-27 >20% - - - 53 <=1959/4 - - - 15-27 <=1959 10 E4d 15-27 5-20% - - - 54 1960s/1&2 - - - >41 1960s 11 E5a&b <15 >40% - - - 55 1960s/+3&X - - - 34-41 1960s Mature eucalypts with unheighted regrowth 56 1960s/-3 - - - 27-34 1960s 12 E1a&b >55 >40% - - - 57 1960s/4 - - - 15-27 1960s 13 E2a&b 41-55 >40% - - - 58 1970s/1&2 - - - >41 1970s 14 E+3a&b 34-41 >40% - - - 59 1970s/+3&X - - - 34-41 1970s 15 E-3a&b 27-34 >40% - - - 60 1970s/-3 - - - 27-34 1970s 16 E4a,b&c 15-27 >40% - - - 61 1970s/4 - - - 15-27 1970s 17 E5a&b <15 >40% - - - 62 1980s/1&2 - - - >41 1980s Mature eucalypts with silvicultural regeneration 63 1980s/+3&X - - - 34-41 1980s 18 E1&2 >41 - - - - 64 1980s/-3 - - - 27-34 1980s 19 E+3 34-41 - - - - 65 1980s/4 - - - 15-27 1980s 20 E-3 27-34 - - - - 66 1990s/1&2 - - - >41 1990s 21 E4&5 <27 - - - - 67 1990s/+3&X - - - 34-41 1990s Unaged eucalypt regrowth with mature 68 1990s/-3 - - - 27-34 1990s 22 ER4-6 + E1 >55 - >37 - - 69 1990s/4 - - - 15-27 1990s 23 ER3 + E1 >55 - 27-37 - - 70 2000s/1&2 - - - >41 2000s 24 ER1&2 + E1 >55 - <27 - - 71 2000s/+3&X - - - 34-41 2000s 25 ER4-6 + E2 41-55 - >37 - - 72 2000s/-3 - - - 27-34 2000s 26 ER3 + E2 41-55 - 27-37 - - 73 2000s/4 - - - 15-27 2000s 27 ER1&2 + E2 41-55 - <27 - - 74 All regen/5 - - - <15 All 28 ER3&4 + E+3 34-41 - 27-44 - - Unstocked eucalypt 29 ER1&2 + E+3 34-41 - <27 - - 75 Some/1&2 - - - >41 - 30 ER3 + E-3 27-34 - 27-44 - - 76 Some/+3&X - - - 34-41 - 31 ER1&2 + E-3 27-34 - <27 - - 77 Some/-3 - - - 27-34 - 32 ER1&2 + E4 15-27 - <27 - - 78 Some/4 - - - 15-27 - 33 ER1 + E5 <15 - <15 - - 79 Some/5 - - - <15 - Pure eucalypt unaged regrowth 80 None/1&2 - - - >41 - 34 ER4-6/1 - - >37 >55 - 81 None/+3&X - - - 34-41 - 35 ER3/1 - - 27-37 >55 - 82 None/-3 - - - 27-34 - 36 ER1&2/1 - - <27 >55 - 83 None/4 - - - 15-27 - 37 ER4-6/2 - - >37 41-55 - 84 None/5 - - - <15 - 38 ER3/2 - - 27-37 41-55 - Rainforest 39 ER1&2/2 - - <27 41-55 - 85 M+ >25 - - - - 40 ER3&4/+3 - - 27-44 34-41 - 86 M- <25 - - - - 41 ER1&2/+3 - - <27 34-41 - Other native forest 42 ER3/-3 - - 27-37 27-34 - 87 Secondary species - - - - - 43 ER1&2 /-3 - - <27 27-34 - 88 Wattle - - - - - 44 ER1&2/4 - - <27 15-27 - Plantation 45 ER1/5 - - <15 <15 - 89 Hardwood - - - - - 90 Softwood - - - - - Non-forest 91 Scrub, waste, water - - - - - 92 Unclassified - - - - -

Page 206 of 365

In addition to providing primary inputs to the FCMF to model carbon stocks across Tasmania over time, these layers provided other useful information on changes in vegetation. For example, to identify potentially Kyoto compliant forest and non-forest, areas classified as grassland in the original 1996 Forest Class layer were assumed to be unforested in 1990. This assumption was considered reasonable since much of the 1996 classification was based on aerial photography taken several years earlier.

A key limitation of the Forest Class system is its treatment of Rainforest which comprises around 563,000 ha or 18% of the total area of native forest in Tasmania (Hickey et al. 1993). This is subdivided into only two types: • M+: > 25 m (primarily callidendrous and thamnic Rainforest) • M-: < 25 m (consisting of thamnic and implicate Rainforest).

Further, there are few if any inventory plots located in areas of Rainforest, so that there is very limited data available on wood volumes in trees, or debris, from which carbon stocks can be estimated. The only published data for Tasmanian Rainforest include one transect by Gilbert (1959) and a study of potential wood volume yields from red myrtle in north-eastern Tasmania (Mesibov 2002). The data from Gilbert did not provide diameter measurements for larger trees (> 8 ft girth or 0.75 m DBH) while the data from Mesibov excluded non-merchantable trees. Thus, this data was considered too unreliable and, potentially, unrepresentative to base estimates of average stem volume or biomass on.

This deficiency was partly addressed through CO2 Australia undertaking field measurements under the Carbon Study and incorporating data collected by the Tasmanian Wilderness Society, and Ian Riley of University of Tasmania (see Section 8).

Tenure & management

Classification of forests by tenure and management is important for modelling future carbon stocks and assessing commercial opportunities in relation to these. The likelihood and method of timber harvesting, products harvested and management regime may differ for public verses private forests, while the available area for harvest (and therefore harvest likelihood) is dictated by the reserve status and management of the land. For the purpose of the Carbon Study, Tasmanian forests were divided into four tenure by management types: • public forest reserved from timber harvesting; • public forest available for timber harvesting; • private forest reserved from timber harvesting; and • private forest available for timber harvesting;

Spatial layers (current for 2010) for forest tenure and management were provided by DPIPWE. It was assumed that forests reserved from timber harvesting included:

Page 207 of 365

• national parks; • World Heritage Reserves and other conservation reserves; • formal and informal reserves on state forest, • forest where harvesting is restricted as a result of code of forest practice regulations; • non-commercial forest; and • forest covered by conservation covenants.

Public forest was assumed to include: • forest managed by Forestry Tasmania; • forest managed by Department of Parks and Wildlife; and • forest owned by the Hydro-electric Commission, or the Department of Defence.

In addition, DPaC and NGOs provided spatial layers for old-growth forest and proposed ENGO reserves. These layers were included with other layers to provide a range of inputs to the model showing old-growth status, potential for harvest, ownership and potential new reserves.

A breakdown of forest tenure and management is shown in Table 1 in Section 2.1. In 2010, the total area of public native forest (i.e. forest >10m) where wood production was permitted was 540,000 ha, compared with 1,700,000 ha managed for other purposes (primarily conservation, Forestry Tasmania 2011a). Of the latter 1,180,000 ha were in formal reserves, conservation reserves or public reserves and 250,000 ha were in informal reserves, 200,000 ha were classed as other state forest and 70,000 ha were classed as other public land.

In 2011, publicly owned plantations covered 108,000 ha. These consisted of 53,000 ha of softwood plantations (P. radiata), owned in a joint venture with GMO Renewable Resources which were recently purchased by New Forests and 55,000 ha of hardwood plantations (E. globulus and E. nitens), primarily owned by Forestry Tasmania, but including joint ventures.

As of 2010, the total area of private native forest was 860,000 ha most of which is potentially harvestable. According to DEPIPWE, 80,400 ha of private native forest are currently covered by conservation covenants (http://www.dpiw.tas.gov.au/inter.nsf/WebPages/DRAR-7TR9ZR?open). The total area of privately owned plantations was 204,000 ha excluding those owned in joint ventures with Forestry Tasmania. These consisted of 23,000 ha of softwood plantations and 181,000 ha of hardwood plantations.

Private native forests are managed for a range of purposes including wood production, grazing and conservation. Private softwood plantations (P. radiata) are managed primarily for sawlog production, while private hardwood plantations (E. globulus and E. nitens) are mostly managed for pulpwood production. In the past, many private plantations were established on areas of ex- native forest. However, a growing proportion has been established on ex-pasture (i.e. potentially Kyoto compliant) sites.

Page 208 of 365

6.2 DETAILED DESCRIPTION OF THE MODELLING FRAMEWORK

An introduction and schematic overview of the FCMF has already been provided at Section 2. This section provides more in-depth detail of the structure behind the FCMF and key modelling processes within the system.

The FCM, like FullCAM, establishes a chain of carbon pools (stand, debris, soil, log products, and processed products) and models the flow of carbon from the stand pool into downstream pools. Outside the stand, processes that determine carbon flow are either simple (e.g. the proportion of sawlogs converted to sawn timber), or established (in the case of the Rothamsted Carbon model RothC).

Forestry Tasmania’s Forest Class data has had an important bearing on the Carbon Study, not just in terms of providing data for a native eucalypt growth model which is discussed below, but on the structure of the FCMF. Forest Classes describe the height and crown coverage of cohorts of different ages which can be modelled independently. It made sense, therefore, that the FCMF allow the modelling of growth of one or more cohorts in the same stand, and that events such as harvesting and fire affect each cohort differently.

The FCMF operates through referencing “patches” of forest located at grid points for which Forest Class, climate, Forest Productivity Index, soil clay and carbon content are known; plantation species and age may also be supplied as stochastic variables. And for each point, a list of conditional events is also created. Stand cohorts are initiated from Forest Class data, grown with an appropriate model, and manipulated by events.

Within the growing forest, carbon flows in monthly steps through the debris and soil pools. Events invariably affect one or more carbon pools within the forest, but only commercial harvesting sends a pulse of carbon out of the forest into the processing pools.

Stand growth & debris production

The primary driver of the model is the growth of the Stand, which is defined as a portion of forest in a contiguous area of similar average productivity and structure but, potentially, consisting of a range of species or tree ages. The growth model is based on data from Forestry Tasmania for native forests and models developed by Candy et al. (1989b, 1997a) for plantations.

The outputs of these models are in stem volumes. These are converted to masses based on assumed basic densities for different species. Amounts of other stem components are calculated from stem mass using species-specific allocation percentages. Components include: • Stemwood • Leaves, • Branchwood • Fine roots, and • Bark • Coarse roots

Page 209 of 365

These bio-component pools contribute material to corresponding debris (dead material) pools (Figure 84). The amount of material transferred from live to dead pools is calculated from: i. a relationship between Stand age, productivity and tree mortality for Stags (dead trees); ii. a set half life based on stem diameter (estimated from site productivity and age of mortality) for Stags moving to stemwood debris; and iii. fixed monthly percentage losses for other components.

It is important to note that, like FullCAM, debris production is not dynamic (i.e. there is no corresponding reduction in the amount of live biomass as a result of debris production). This is because the Stand growth model estimates the net live tree biomass after loss of material to debris.

Various events can influence the growth of the Stand, debris production and debris decomposition. The main events modelled in the Carbon Study were Harvest and Fire, the effects of which are described in a later section.

Figure 84: Diagram of the flows of carbon to Stand components and debris in the FCMF.

Page 210 of 365

Debris decomposition & soil carbon cycling

Debris contribute to soil carbon through decomposition. Rates of decomposition are influenced by temperature and rainfall as well as soil and vegetation type. To allow more flexible modelling of decomposition rates, each debris component is split into decomposable and resistant pools (Figure 85).

The rate of decomposition of debris is based on average percent monthly decomposition rate for decomposable and resistant portions of each debris pool. A portion of the amount decomposed is lost to the atmosphere, and the remainder moves to the soil. Resistant debris move to Resistant Plant Material and the soil and decomposable debris move to Decomposable Plant Material in the soil layer. These two components then drive carbon input into the soil carbon cycle. This cycle is modelled using the RothC soil carbon model (26.3, Coleman & Jenkinson 2008). Soil carbon pools consist of:

• Decomposable Plant Material (DPM) • Resistant Plant Material (RPM) • Humus (Hum) • Microbial Biomass (Bio)

Figure 85: Diagram of the flow of carbon from debris to soil in the Forest Carbon Model.

Page 211 of 365

Decomposition of carbon in DPM and RPM layers contributes material to the Hum, Bio pools as well as atmospheric CO2. Decomposition of the Bio and Hum pools results in further losses of atmospheric CO2 with the residual transferred between each of the pools. This cycle continues on monthly time steps with the soil carbon pools eventually reaching an equilibrium based on the amount of monthly input of resistant and decomposable material from debris, the rate of cycling and the percentage loss of CO2 from the system. The rate of cycling is influenced by temperature and rainfall in a similar manner to debris decomposition.

Harvesting and wood products

Product pools arise as a result of forest harvesting. A harvesting event gives rise to different log types and debris, with the proportion of each log type produced based on cutting rules specific to each forest type. The various log types modelled include:

• Veneer logs (peeler logs) • Sawlogs • Pulplogs

In addition to moving carbon from stemwood to wood product pools, harvesting results in the direct emission of CO2 as a result of consumption of fossil fuels (Figure 86). These emissions are related to the total volume of logs harvested. Following harvesting, regeneration burns result in further emissions in the form of CO2 and non-CO2 greenhouse gasses. CO2 emissions are accounted for in reductions in carbon stocks in vegetation, debris and soil. However, non- CO2 emissions must be accounted for separately as explained Section 6.11.

Figure 86: Diagram of the flow of carbon from the forest to wood products in the Forest Carbon Model.

Page 212 of 365

The three log types are processed into different products according to a set of processing rules. Currently these are non-species specific, so that it is assumed that the proportion of sawlogs, or pulplogs exported, or converted to, sawn timber is the same for softwoods as hardwoods.

Wood products from processing modelled include:

• Biofuel • Domestic solid wood • Domestic paper (i.e. sawn timber)

During processing, a portion of each log type is assumed to be burnt for biofuel and is thus converted to CO2. In addition, a portion of logs from each log type and product type are assumed to be exported. Exported products are assumed to be converted directly into CO2 and, thus, that they do not add to product pools.

The products from log processing are decayed over time according to product half-life rules and it is assumed a proportion of the total pool is lost each year. Where the rate of input of new products is constant, the wood product pool will gradually increase until it reaches an equilibrium where inputs are equal to losses through decomposition.

Fire

Fire results in the movement of carbon from vegetation, debris and soil to the atmosphere as well as movement of carbon between pools (Figure 87). It also results in the production of inert soil C (charcoal) that is assumed not to be cycled and therefore, accumulates over time. The movement of carbon depends on the frequency and intensity of fire, which, in turn, is assumed to depend on forest type and time since last burn.

Two types of fire are modelled: • Wildfire • Post-harvest burns

Wildfire is assumed to result in the following movements between pools:

• from live vegetation to debris, inert soil C, and atmospheric CO2;

• from debris to inert soil C and atmospheric CO2; and

• from soil to inert C and atmospheric CO2.

Post-harvest burns result in similar movements of carbon between live vegetation, debris and soil pools as a result of mortality of retained trees and combustion of carbon in debris and soil.

In addition to moving of carbon between pools and to the atmosphere, biomass burning results in the production of non-CO2 Greenhouse gas emissions including CH4 and N2O. These emissions are modelled based on IPCC default values for extra-tropical forests as explained in Section 6.11 (IPCC 2006a).

Page 213 of 365

Fire intensity rules

Live CO2 Live vegetation and non- Vegetation combustion CO2 rules emissions

Stags Stags Debris and Debris stag combustion Debris rules Inert soil carbon Soil Soil combustion rules

Figure 87: Diagram of the effect of fire on carbon vegetation, debris and soil within the FCMF.

6.3 GROWTH MODEL FOR NATIVE EUCALYPT FORESTS

Estimating maximum Stand volume

Growth of native eucalypt forests is based on a model derived from volume data for Forest Classes provided by Forestry Tasmania. Forest Classes are based on site productivity (as indicated by current or potential height of mature trees), Stand age (i.e. mature, regrowth, unaged regeneration and aged regeneration), cover (>40% or 5-40%) and number of cohorts (e.g. single aged, mature plus regrowth, mature plus regeneration, or regrowth plus regeneration). Site productivity is divided into six mature height classes (known as Strata). These Strata indicate the potential maximum height and hence productivity of a particular Stand.

• E1: > 55 m • E3-: 27-34 m • E2: 41-55 m • E4: 15-27 m • E3+: 34-41 m • E5: < 15 m

Using data from permanent sample plots from Forestry Tasmania, Mature Height (assumed to be the midpoint for each stratum) was related to average Entire Stem Volume (ESV). Only data for Forest Classes consisting of pure mature forest of >40% cover were used. The resulting relationship explained 98% of variation in average stem volume for fully stocked mature forest (Figure 88).

Page 214 of 365

1000

800

600 /ha) 3

400 ESV (m

y = 2.2873x1.4339 200 R² = 0.978

0 0 20 40 60 80 Mean dominant height (m)

Figure 88: Relationship between ESV and height of mature Stands with >40% cover.

The form of the relationship was:

Equation 1

Modelling volume growth & decline of over-storey

The Forest Class system assigns ages to silvicultural regeneration in which the decade of establishment is known. Other Forest Classes are classified as being mature, regrowth or regeneration. Mature Forest Classes are considered to be >110 years old while regrowth is considered to be < 110 years old (Moroni, 2011). This broad age classification system creates difficulties when attempting to model growth because measured volumes from mature stands span a very broad range of ages (i.e. 110-500 years). Although some mature stands may be at or close to their maximum volume, others are likely to be much younger and still growing. Thus, the average volume of mature trees of a particular Stratum (i.e. site productivity) is likely to be less than the potential maximum volume of the stand. Further, assuming that Forest Classes comprised purely of mature trees were already at their maximum volume, would result in the modelled growth of these forest classes being zero or even negative.

Page 215 of 365

To avoid underestimating maximum stand volumes and growth rates of mature Forest Classes, those Stands classed as mature were given a nominal age that was less than the assumed age of maximum Stand volume. Average age of Stands classified as consisting purely of mature trees (FCs 1-9) was assumed to be 150 years, while the age of maximum stem volume was assumed to be 200 years. Other Forest Classes with crown cover greater than 40% and comprising just one age class (i.e. regeneration, regrowth, or mature forest) were assigned an average age based on the following assumptions: • age of unheighted regrowth: 70 years; • age of heighted regrowth: 20-100 years based on a relationship between current average height of the forest class, potential mature height and age; and • age of regeneration from a particular decade: 2010 – midpoint for the decade of regeneration.

Using these ages and the ESV data for each forest class, a general growth model was developed relating current ESV to age and potential maximum ESV. The form of this relationship was:

Equation 2 where: is current volume for each Forest Class, is the maximum volume of the Stratum, is the assumed age (weighted by cover for multi-aged Forest Classes), and 0.95 is the ratio between modelled ESV at age 150 years (assumed average age of mature stands) and ESV at 200 years (assumed age of maximum volume).

This model explained 93% of variation in ESVs across the forest classes (Figure 89).

In the absence of fire, or other disturbance, Tall Eucalypt Forests (wet sclerophyll) generally cannot regenerate and therefore may succeed to Rainforest which contains shade tolerant species. During this period, it is possible that there is a decline in the volume of live stems as the taller eucalypts die and are replaced by shorter Rainforest species. This potential decline is suggested by the lower volume of Forest Classes with mixed forest (<40% dominant eucalypt cover, with an assumed age of 350 years) compared with pure mature eucalypt stands 9Figure 89).

Page 216 of 365

1200

E1 1000 E2 E3+

/ha) 800 E3- 3 E4

600 E5 E1 E2 400 E3+ Stand volume (m volume Stand E3- 200 E4 E5 0 0 100 200 300 400 Stand age (years)

Figure 89: Modelled growth curves for different productivity classes (strata) based on assumed maximum ESV and average ESV for each Forest Class for single aged Stands of different assumed ages.

To allow the transition from pure eucalypt forest to mixed forest and eventually pure Rainforest to be modelled, a decline parameter was added to Strata corresponding to Tall Eucalypt Forests (E1, E2 and E3+). It was assumed there would be no decline in other forest types (i.e. Low Eucalypt forests, Rainforests and Other Native Forests).

The form of this relationship was assumed to be:

Equation 3 where is a decline or senescence factor. This was modelled as:

Equation 4 where & Equation 5

is the age at which 95% of the total maximum eucalypt volume remains alive and is the age at which 5% of the total maximum eucalypt volume remains alive.

Page 217 of 365

Thus, can be parameterised dynamically by specifying the ages at which 95% and 5% of the eucalypt Stand remains alive. There is currently little data on the timing or change in biomass associated with eucalypt forest senescence. Recent work indicates that dominant eucalypts can reach ages in excess of 500 years in wet sclerophyll forests (Wood et al. 2010). It was assumed that this represented the age at which 5% of the original eucalypt volume remained and that the age at which the decline commenced ( ) was assumed to be half this (i.e. 250 years).

Modelling volume growth of Rainforest understorey

As the eucalypt forest declines, Rainforest understorey develops and eventually becomes the dominant forest type. There is currently little data on the timing or change in biomass associated with this transition. Further, there is no published data on the volume or biomass content of temperate Rainforest Stands or comparisons of volumes in Rainforest and adjacent eucalypt forest.

To help address this issue, CO2 Australia, surveyed Rainforest and old-growth eucalypt Stands in the Styx valley/Florentine valley area. These consisted of 16 paired plots randomly located in mature, high productivity wet sclerophyll forest with >40% cover (Forest Classes 1 or 3) and adjacent Rainforest (Forest Classes 85 or 86) with few if any eucalypts present. The two plots for each pair were located on similar soil, aspect and topography within 200 m of each other. Thus, it was assumed the difference in vegetation types reflected the different time periods since disturbance by fire and that the stands of mature eucalypt forest and Rainforest represented different stages in the successional sequence of the transition from Tall Eucalypt Forest to Rainforest.

The plots consisted of two concentric circles, one with a 20m radius, within which all live and dead trees >40cm were measured and another with a 7m radius, within which all live and dead trees >15cm DBH were measured. The results from these measurements indicated that the volume of live trees on the Rainforest plots was 45% of that on the eucalypt plots. The results of this comparison are discussed more fully in Section 8.

Based on these results and comparisons with data from the Tasmanian Wilderness Society and the University of Tasmania, a model for succession of eucalypt forest to Rainforest was developed. For this model it was assumed that: • volume of Rainforest species increases in tandem with decline in the volume of eucalypts; • Rainforest reaches maximum volume at the point when eucalypt volume reaches zero (i.e. after 500+ years); • the maximum volume of Rainforest is 50% that of the maximum eucalypt volume (based on the results from the field measurements); and • mature wet sclerophyll forest with <40% cover represents senescing Stands aged around 350 years.

Page 218 of 365

The resulting Rainforest model is essentially the inverse of the eucalypt senescence factor , but with an asymptote set to be a proportion of the maximum eucalypt volume (i.e. 0.5). With the assumption of Rainforest growth inversely matching eucalypt decline, the Rainforest growth model can also be dynamically parameterised to test a range of growth possibilities including: • timing of the start of eucalypt senescence and increase in Rainforest volume; • timing of the completion of eucalypt senescence, and • the maximum volume of Rainforest species.

The form of the relationship for estimating the stem volume of Rainforest trees ( ) was:

...... Equation 6

Where is the proportion of eucalypt asymptotic volume and is assumed to be 0.5 for Rainforest based on the plot measurements.

While we have explicitly modelled the maximum volume achieved by Rainforest trees, total volume from permanent sample plots includes volumes of Rainforest species. No breakdown was available for apportioning volumes to overstorey and understorey species. As a result, the current model tends to overestimate the contribution of eucalypt species and to underestimate the contribution from Rainforest species in younger Stands (i.e. aged ~50-250 years). However, the total volume is correct (at least, as far as possible given the limitations in the model and natural variation in the original data).

It should also be noted that the volume of Rainforest trees may continue to increase over time after all eucalypts have died. By setting the age of 95% loss of eucalypt volume as 500 years, we assumed that stem volumes of pure Rainforest (assumed to be aged 500 years) will continue to increase slowly over the time period modelled (i.e. 40 years for most scenarios). This assumption is less conservative than that of the default model for harvested native forests used by the Department of Climate Change in which Rainforest is assumed to cease growing after age 200 years (DCCEE, 2011, Table 7.B.6).

Combined eucalypt & Rainforest growth models

The resultant growth model for a Stand within the E1 strata is illustrated in Figure 90 together with data points for average ESVs for different Forest Classes within the strata. This shows the Stand passing through four distinct growth stages which are consistent with current understanding of the ecology of these forests: • a stage of rapid growth (a), during which the Stand is developing – Regrowth forest; • a stage where maximum Stand volume is reached (b) in which volume growth is equal to volume loss as result of mortality and decay – Mature Forest;

Page 219 of 365

• a declining stage (c) where losses due to mortality of larger trees exceed volume increment of surviving eucalypts and the developing Rainforest understorey – Mixed Forest; and • a steady state stage (d) in which most eucalypts have died and Rainforest species regenerate naturally in gaps left as trees die.

1000 Eucalypts b) Rainforest c) Total 800 Plot data /ha) 3 600 a) d)

400 Stem volume (m

200

0 0 100 200 300 400 500 600 Stand age (years)

Figure 90: Example of the growth model for an E1 stratum (mature height 41-55 m). Points show actual average volumes for plots in the Forest Classes within the stratum.

Modelling multi-cohort Stands

In most cases, fire will partially or completely destroy a eucalypt Stand before it succeeds to Rainforest. Fire may kill all trees, resulting in even aged regeneration, or more commonly, may remove only of a proportion of trees allowing development of mixed aged Stands (see Section 6.13). In some cases, a Stand may be burnt one or more times per century by fires removing only a proportion of trees, resulting in the development of multiple cohorts.

In low open forests (dry sclerophyll forests), fires may occur once every decade or so, but generally kill only a small proportion of mature trees. This prevents succession to Rainforest and results in a complex mixture of cohorts that effectively reach a steady state in which younger trees fill the gaps provided by the eventual death of older trees after numerous fires.

Page 220 of 365

To model multi-cohort Stands, a proportion of the total area was allocated to each cohort. Each cohort was then given a nominal age based on its current height, the productivity class of the site, where this was known, and the total Stand volume (based on the average for the forest class). For example, for Forest Class 23, which consists of mature forest of E3+ productivity with regrowth between 27 and 37m tall, it was assumed that each cohort occupied 50% of the site and that the age of the mature trees was 200 years while that of the regrowth was 30 years (as inferred from a relationship between height and age for that productivity class). The assumed initial ages and proportions of younger cohorts in other multi-cohort forest classes were:

• FC 12-17 (mature forest with unheighted regrowth): 90 years, 20% occupancy; • FC 18-21 (mature forest with silvicultural regeneration.): 20 years, 90% occupancy; and • FC 22- 33 (mature forest with heighted regrowth): 10-100 years, depending on height, 50% occupancy.

This multi-cohort approach allows a maximum of 10 cohorts of different ages to be modelled simultaneously for an individual Stand. This allows the development of complex, multi-aged Stands that develop as a result of numerous fire events, each removing a proportion of trees from each cohort. This can also be used to model the effects of partial harvesting, in which only a proportion of trees are removed from each cohort in a Stand.

Performance of native forest growth model

The final model was tested by relating measured stem volumes of all forest classes with volume data, to predicted volume based on Stand productivity, and proportion and age of each cohort. The resulting model explained 93% of variation across all forest classes with close to a 1:1 relationship between actual and predicted volumes (Figure 91).

Page 221 of 365

1000 y = 0.9692x + 42.002 900 R² = 0.9314

800

700 /ha)

3 600

500

400

Actual ESVActual (m 300

200

100

0 0 200 400 600 800 1000 Predicted ESV (m3/ha)

Figure 91: Relationship between actual entire stem volumes (ESVs) and predicted volumes based on the native forest growth model across all forest classes.

Scaling volume growth

To allow for variation around mean predicted growth rate, a scaling factor was used. This has two purposes: i. to adjust growth of the model to be consistent with average known volumes of individual Forest Classes; and ii. to allow for natural variability in growth of individual Stands within each forest class using a probability function.

After the growth curve of each Forest Strata was calculated, growth of each individual Forest Class was adjusted up or down based on the average measured volume of the Forest Class and the assumed age of each cohort. This procedure ensured that the estimated volume of each Forest Class matched its actual volume.

The final form of the model was:

Equation 7 where is the volume at age , is the number of cohorts , is the scaling factor, is the proportion of the Stand occupied by cohort , is the decline factor and is the modelled maximum volume of the forest class.

Page 222 of 365

6.4 GROWTH MODEL FOR RAINFOREST & UNDERSTORY SPECIES

In the FCM, Rainforest, wattle and secondary species (i.e. non-Rainforest and non acacia understorey species such as banksia and tea tree) are treated in three distinct ways: i. as an understorey to eucalypt forests, ii. as a dominant forest type in forest classes where all eucalypts have died due to: a. senescence (as in the case for tall open forests in the absence of fire); or b. frequent fires removing the eucalypt seed source (in which case the forest was assumed to revert to secondary species); or c. failed regeneration following harvest; or iii. as a dominant forest type in non-eucalypt forest classes (i.e. 85-88).

In the first two cases, Rainforest was assumed to occur under Tall Eucalypt Forests (Strata 1, 2 and 3+) while secondary species (e.g. Pomaderris, Olearia and Pittosporum) were assumed to occur under Low Eucalypt Forests (Strata: 3-, 4, 5 and 6). All understorey types were assumed to follow the model already described – developing as the eucalypt volume starts to decline and plateauing at ~50% of the maximum eucalypt volume, when most eucalypts had died.

In the third case, the Rainforest, wattle or secondary species were assumed to develop as unique cohorts in the Stand, following similar, but lower, growth trajectories to those of the equivalent eucalypt forest. The growth of each forest type was assumed to be 50% of the eucalypt equivalent for the following productivity classes for each type: • M+ Rainforest (FC85): E2; • M- Rainforest (FC 86): E3+; • secondary species (FC 87): E4; and • wattle (FC 88): E4. It was also assumed that growth rates of Rainforest and secondary species would reach a steady state at around 300 years with no subsequent decline. Additionally, the following assumptions were made: • stem volume of understorey species were assumed to be included in the ESV figures provided by Forestry Tasmania for all forest classes with values; • mature Low Eucalypt Forest with sparse canopy cover (FC8 and FC10) was assumed to be in a steady state with an average age of 200 years and to consist of 50% understorey and 50% eucalypt species; • understocked areas (canopy <10%) were assumed to consist of 30% eucalypts and 70% secondary species with an average age of 50 years; and

Page 223 of 365

• unstocked areas were assumed to consist of 100% secondary species with an average age of 50 years.

The latter two assumptions were made to give understocked and unstocked areas ESVs consistent with those provided by Forestry Tasmania.

6.5 GROWTH MODELS FOR PLANTATIONS

Hardwood and softwood plantation growth was modelled using published growth functions for E. nitens (Candy, 1997a) and P. radiata (Candy, 1989a and 1989b, and Candy, 1997b) respectively. These models are driven by basal area, age and site index (or peak MAI).

There was no available site index data from private growers. However, using DCCEE’s Forest Productivity Index layer (Kesteven et al., 2004), relationships between productivity class of Forestry Tasmania’s plantations and average FPI from 1970 to 1995 were developed for both hardwood and softwood plantations (Figure 92).

Forest productivity index explained around 37% of variation of each plantation type. The main outliers in this relationship were for FPI 10 which corresponded with a sharp divide in FPI across the state based on a change in soil type – assumed to correspond with higher N status (Figure 93). This relationship was used to vary plantation growth across the state. a) b)

28 30

26 25

20 Site Index Site Index 22

15 20 y = 0.43x + 16.50 y = 0.15x + 21.99 R² = 0.38 R² = 0.37 18 10 5 7 9 11 13 15 5 7 9 11 13 15 FPI FPI Figure 92: Relationship between Site Indices of a) hardwood and b) softwood plantations and Forest Productivity Index based on data from Forestry Tasmania. Error bars indicate standard deviation of means for each FPI class.

Page 224 of 365

Figure 93: Average FPI (Forest Productivity Index) across Tasmania for period 1990-2006. Source: DCCEE, data processed and presented by CO2 Australia.

6.6 GROWTH MODEL FOR NON-FOREST >2M

For non-forest woody vegetation >2 m, the following assumptions were made: • growth rate was assumed to be 10% of that of eucalypts for mature forest, and • average age (i.e. time since fire) was assumed to be 20 years Vegetation <2m is classed as non-forest under Australia’s agreed definition under the Kyoto protocol. Therefore carbon stocks and emissions in this vegetation were not included in the Carbon Study. However, since some of the scenarios included planting of cleared agricultural land, it was important to model soil carbon cycling for this land prior to planting. The method used for this is explained in the Section 6.10.

Page 225 of 365

6.7 BIOMASS ALLOCATION AND TURNOVER

Allocation refers to the apportioning of biomass to different tree components (stems, branches, foliage, bark, coarse roots and fine roots). In FullCAM, this is modelled by assuming a fixed or variable allocation rate for apportioning production of total biomass to the various components. The FCMF uses a similar approach allocating biomass to tree components based estimated stem biomass.

Modelled stem volume was converted to mass assuming the following densities: • native hardwood: 550 kg/m3 • plantation hardwood: 540 kg/m3 • plantation softwood: 450 kg/m3

Allocation percentages for Tall Eucalypt Forests and Low Eucalypt Forests in Tasmania were based default figures for harvested tall dense eucalypt forests from DCCEE (2011b, Table 46). Allocation parameters for Rainforest bio-components were adjusted slightly from the default figures which were developed from tropical and subtropical species (Keith et al., 2000). Allocation parameters for hardwood and softwood plantations were based on averages for selected Tasmanian plantation regimes downloaded through FullCAM’s databuilder. Turnover refers to the rate of conversion of live biomass to dead biomass or Debris. Debris include above-ground dead biomass such as Stags, Coarse Woody Debris (CWD, fallen stems and branches), and finer material such as dead leaves and bark. They also include below- ground dead coarse and fine roots. Understanding carbon cycling in debris is important because these form an intermediary between vegetation and soil carbon and can hold significant amount of carbon themselves. They are also important sources of emissions during fire.

In FullCAM, six debris pools are modelled: • dead wood, • leaf litter • chopped wood • coarse dead roots • bark litter • fine dead roots

These pools correspond to their respective biomass components with dead wood corresponding with stem (or Stag) and branch debris. Chopped wood consists of stem and branch material left on site as harvest residues. For the FCMF, the chopped wood pool was split into stem debris and branch debris. Turnover rates for bio-components other than stem debris were based on default figures from DCCEE (2011b, Table 46).

Page 226 of 365

Table 46: Parameters used in the FCMF to estimate carbon content and turnover rates for different tree components and forest types from stem volume. Non-bold figures are defaults from DCCEE (2011b, Tables 7.B4, 7.B1, 7.B3 and 7.D2). Parameter Bio- Forest Type component Tall Low Other Softwood Hardwood Rainforest eucalypt eucalypt native plantation plantation forest forest forest Allocation Stems 50% 55% 55% 50% 67% 67% Branches 15% 12% 12% 15% 9% 9% Bark 10% 10% 10% 10% 10% 10% Leaves 4% 3% 3% 5% 3% 3% Coarse roots 17% 17% 17% 17% 8% 8% Fine roots 4% 3% 3% 3% 3% 3% Carbon Fraction (%) Stems 52% 52% 52% 52% 50.0% 51.0% Branches 47% 47% 47% 47% 46.8% 51.4% Bark 49% 49% 49% 49% 48.7% 53.3% Leaves 52% 52% 52% 52% 52.9% 51.1% Coarse roots 49% 49% 49% 49% 49.2% 50.4% Fine roots 46% 46% 46% 46% 46.1% 48.4% Turnover Stems 1.5% 1.5% 1.5% 1.5% 1.5% 1.5% Branches 6.5% 6.5% 6.5% 6.5% 3% 5.0% Bark 9.5% 9.5% 9.5% 9.5% 5% 7.0% Leaves 40% 40% 40% 40% 30% 50% Coarse roots 6.5% 6.5% 6.5% 6.5% 7% 10.0% Fine roots 73.3% 73.3% 73.3% 73.3% 80% 85.0% In FullCAM, stem turnover is not modelled (except as a result of assumed mortality due to harvesting or fire). However, to obtain amounts of CWD consistent with available data, it was necessary to include stem mortality in the FCMF. This was done by developing a separate tree mortality model (see Section 6.8).

Forestry Tasmania has an extensive database of Stag and CWD volumes that was made available to the Carbon Study. This data was used to determine the average mass of these carbon pools. The data covered most forest classes and had been collected during routine measurements of permanent sample plots. Data were highly variable with volumes ranging from 3 - 1200 m3/ha CWD and 0 - 320 m3/ha for Stags, indicating that it was not appropriate to use averages for different forest classes to initialise the model. Consequently, the data were further analysed to determine if there were any relationships between these carbon pools and Stand productivity or age that could be used to derive averages for different Forest Classes.

Page 227 of 365

There was a significant relationship between stand productivity (as indicated by potential old growth height of each Forest Class) and CWD volume (P < 0.025, R2 = 0.666, Figure 94):

Equation 8 where is the volume of CWD (in m3/ha) and is the potential old growth height (in m).

This relationship is independent of stand age which is surprising. However, it is consistent with findings from a study by Thauvin et al. (2010) which indicated that the total volume of CWD is similar between old growth forest and older silvicultural regeneration. In that study, the main factor shown to influence the volume of CWD was disturbance history. Those sites with a history of greater disturbance, due to either timber harvesting or wildfire, had significantly less CWD. Silvicultural regeneration on plots that had previously been old-growth forest tended to have relatively large amounts of CWD while those on sites that had previously been mature forest had relatively small amounts of CWD.

700

600

500 /ha) 3

400

300

200 CWD volume (m 100

0 0 10 20 30 40 50 60 70

Old growth height (m)

Figure 94: Relationship between total volume of Coarse Woody Debris and stand productivity (as indicated by potential height of old growth). Data points are averages for Forest Classes.

Page 228 of 365

There was a significant relationship between average Stag volume (from data provided by Forestry Tasmania) and stand age and productivity across forest classes (P < 0.025, R2 = 0.346)

Equation 9 where is the volume of stags (in m3/ha), is the age of the main cohort in the stand (in years) and is the potential old growth height (in m).

Coarse Woody Debris volume was converted to mass assuming a weighted average density of 335 kg/m3 based on data from Grove (2009). Stag volume was converted to mass by assuming that the initial density of Stags was the same as that of live trees. The resultant estimates for forest classes were averaged for Tall Eucalypt Forest (old growth height > 34 m) and Low Eucalypt Forest (old growth height < 34 m) for comparison with data Woldendorp et al. (2002). Average masses of CWD were 73 t/ha in tall eucalypt and 46 t/ha in Low Eucalypt Forest, while the average mass of Stags was 22 t/ha in tall eucalypt and 8 t/ha in Low Eucalypt Forest. If it is assumed that Tall Eucalypt Forest is equivalent to wet sclerophyll forest and Low Eucalypt Forest is equivalent to dry sclerophyll forest, then the values are within the reported ranges (Table 47).

Table 47: Average figures for amounts of coarse woody debris (CWD), stags from Forestry Tasmania data compared with those from Woldendorp et al. (2002). Debris Average Based on data from Forestry Tasmania Woldendorp et al. (2002) Volume Mass Mass Avg. SD Avg. SD Avg. SD CWD Tall eucalypt 235.0 129.4 72.9 40.1 108.7 241.8 Low eucalypt 149.4 72.3 46.3 22.4 50.9 64.4 Stags Tall eucalypt 40.3 29.3 22.2 16.1 26.8 35.3 Low eucalypt 14.3 11.5 7.9 6.3 9.0 10.0

6.8 STAG TURNOVER MODEL

In the absence of fire and disease, eucalypts die in the normal course of Stand growth due to competition, but also due to senescence. Since we were modelling the long-term growth and eventual decline and senescence of Stands, it was important to be able to account for the carbon stocks removed from live trees and deposited in the debris pools. However, FullCAM does not model stem mortality in undisturbed forests and no parameters were available to estimate rates of transfer of stemwood carbon to debris pools. Thus, a new model was developed and parameterised to simulate the transfer of carbon from live stems to debris through both natural mortality and disturbance events such as fire.

Page 229 of 365

The approach used to model Stag pools was similar to that developed by Grove et al. (2011). In that study, Stags were assumed to be the first stage of the process of stemwood debris production, forming in an intermediary carbon pool between live stems and debris. The rate of Stag production was estimated from a tree mortality model with Stag life span (i.e. time until falling) varying with origin (fire or natural mortality) and stem diameter, with larger trees standing for longer. In addition, Stags were assumed to decay while standing, but at a slower rate than stem debris on the ground.

To estimate the rate of Stag production through natural mortality, it could simply be assumed that a proportion of trees die each year. However, this method would result in an underestimation of Stag production in senescing Stands when the volume of live trees is declining, but the amount of Stags being produced may be increasing rapidly. Further, it is likely that larger Stags will persist for longer than small Stags. Thus, it is important to be to estimate the average diameter of dying trees as well as their volume. The following sections describe the development of the mortality model for the FCMF.

Modelling Mean Diameter of Live Trees Mean diameter increases with tree age and stand volume and decreases with stocking. Observed average diameters of trees in senescent Stands of productive eucalypt forests may around 200 cm (Dean 2011). If it is assumed that these trees are from the most productive eucalypt forests (i.e. E1 Stratum) and are aged around 350 years, then average maximum diameter can be related to Old Growth Height for each Stratum as follows:

Equation 10 It is assumed that these surviving trees were originally the dominant (i.e. largest diameter) trees in the stand. The increase in diameter of these trees over time was modelled as:

Equation 11

To estimate mean diameter, it was assumed that ratio between mean tree diameter and maximum tree diameter was initially 0.6 (based on data from Forestry Tasmania). It was also assumed that this ratio increased over time to reach a maximum of 1 by 350 years. The model was assumed to have similar characteristics to the model for Rainforest growth:

Equation 12 where & as used to model stand senescence and Rainforest growth (Equations 3 and 4).

Page 230 of 365

By combining Equations 10, 11 and 12, average tree diameter could be estimated from Mature Height and current Age.

Modelling Stocking Stocking is related to average diameter in the following relationship:

Equation 13 where is the Stand Basal Area (m2/ha), is Stocking (stems/ha) and is the average tree diameter (cm). By manipulating this relationship, stocking can be derived from stand Basal Area and average diameter:

...... Equation 14

Stand Basal Area is related to Stand Volume, Height and Age. As a general rule, total stem volume for a Stand ( , m3/ha) can be estimated from Stand Basal Area and mean dominant height ( , m) as follows (Goodwin, Farm Forestry Toolbox, Version 5; 2011):

...... Equation 15 Therefore, can be expressed as:

...... Equation 16

Mean dominant height can be estimated from Stand Age ( ) and the Mature Height ( ) of the Forest Class Stratum:

...... Equation 17 where and and are forest specific parameters derived from field measurements. By combining equations 14, 15, 16 and 17, Stocking can be estimated from average diameter (estimated in previous section), age, volume and mature height as follows:

...... Equation 18

This model produced a change in stocking over time which was close to those for E. regnans Tall Eucalypt Forests in Victoria and Tasmania (Ashton, 1981; Gilbert, 1959; and unpublished data from Britten, 1955 compiled by Griffin et al., 2008, Figure 95).

Page 231 of 365

3000

2500

2000

1500

1000 Stocking (stems/ha)

500

0 0 50 100 150 200 250 300 350 Stand age (years)

Figure 95: Change in stocking with time for a Tall Eucalypt Forest as implied by Equations 7 and 12 (blue line), overlaid over measured changes in stocking of Eucalyptus regnans in Victorian and Tasmania as reported by Gilbert (1959), Ashton (1981) and Britten (1955) and compiled by Griffin et al. (2008, unpublished). Figure reproduced with permission from author.

Modelling Stag Production In senescent stands, volume can be estimated from the decline in total stand volume from one year to the next. In contrast, in growing Stands, the volume of dead trees entering the Stag Pool each year in growing stands must be estimated from the average diameter and height of dying trees and the change in stocking. Thus the model for Stag production had to include two functions: one for estimating Stag production in growing Stands and another for estimating Stag production in senescent Stands. In young Stands, the mean basal area of trees dying is about 30% of the mean basal area of live trees (Garcia, 2010; Goodwin, 2012). However, this proportion necessarily tends towards one, late in the senescent phase, when mean and dominant diameters coincide (i.e. when only one tree remains standing). We assumed that:

– Equation 19

Page 232 of 365

where: is the volume of dying trees expressed as a proportion of average stem volume for the Stand Cohort in m3/ha, is the age of the Cohort in years, and and are parameters determining the rate of increase in with stand age. In this study we set and as 250 years and as 500 years.

To allow there be a smooth transition between the rate of stag production in growing and senescing stands, the final equation for estimating the volume of stems dying had the following form:

...... Equation 20

This equation estimates the volume of stems dying as the maximum of either:

a. the annual rate of self thinning, estimated from change in stocking (from Equation 17), average stem volume and the relative volume of dying trees , or, b. The annual decline in total stem volume due to senescence (Equation 3).

This approach allows the input of carbon from dead trees into the Stag Pool to be consistent with the loss of stemwood carbon in live trees as the stand declines as a result of senescence.

Annual Stag input volume is shown in Figure 96. The striking thing about Stag pool input is the strong pulse between ages 250 and 400 years. The pulse is clearly driven by modelled senescence. However, there is little evidence of such a pulse in the data. Mean Stag volumes rarely exceed 20 m3/ha in older forest classes. There is some evidence that CWD volumes increase with age, and this may lend support to idea that Stags remaining standing for only a short time, unless produced from the death of healthy trees as a result of fire. Fire killed Stags are likely to persist for much longer (Grove 2002, 2011).

In the Carbon Study, we assumed that the half life of stags arising from natural mortality was 10 years, while that for stags from trees killed in wildfire was 30 years. In addition, we assumed that 1.5% of Stag mass was lost each year as a result of decomposition.

Page 233 of 365

Figure 96: Annual Stag pool input. The blue line is based on modelled Dq, and the orange line based on the linear diameter model.

6.9 DEBRIS DECOMPOSITION

Debris decomposition rates were calibrated to maintain debris pools at levels that were relatively stable over time with amounts consistent with observed estimates (Woldendorp et al. 2002). Inputs to this calibration process included default figures for FullCAM (DCCEE 2011b), CWD and Stag volumes, input rates from the native forest growth and Stag production models and parameters for a temperature/rainfall modifier included in the FullCAM model.

FullCAM divides debris pools for each bio-component into decomposable and resistant fractions. This allows the rate of decomposition to change over time for each bio-component to reflect the changing nature of the material as it decomposes. It also allows the decomposing debris to feed in to the Roth-C soil carbon model for which the primary inputs are decomposable plant material (fed from the decomposable debris fractions) and resistant plant material (fed from the resistant debris fractions, see Section 6.11). The proportions of decomposable and resistant material for each debris bio-component were based on default figures from DCCEE (2011b, Table 48).

Turnover and decomposition rates used in FullCAM have been tested and validated for a range of forest types by Mackensen & Bauhus (1999), Paul et al. (2003), Paul & Polglase (2004) and Roxburgh et al. (2010), so that they are currently considered the most suitable values for Australian forests (DCCEE 20011b p. 97). However, these rates assume that the climate modifier that is included in the model is turned off. In the Carbon Study, one objective was to evaluate the potential effects of climate change on forest carbon stocks. Thus, it was necessary to include this modifier and recalibrate debris decomposition rates in order to model the effects of projected changes in temperature and rainfall on carbon stocks in debris and soil.

Page 234 of 365

In FullCAM, the amount of debris decomposed is calculated as follows:

Equation 21 where: = is the amount of debris component i decomposed during the month, = is the monthly decomposition rate for debris component I, = is the amount of debris component i present at the start of the month, = a modifier to account for the effect of climate.

In the default version of FullCAM, and for the national carbon accounts, C is set to 1 (i.e. switched off). If the climate modifier is switched on, the rate of decomposition of each debris pool is modified based on variations in temperature and rainfall using the following function:

– – – – Equation 22 where: = the sensitivity of breakdown to temperature (s >= 0) = the average air temperature (in degrees Celsius, negative temperatures treated as zero for this purpose) during the time period (T >= 0) = the sensitivity of breakdown to water (v >= 0) = the total rainfall and irrigation (in mm) during the time period (W >= 0)

It is important to note that this modifier varies between 0 (when either temperature or rainfall are very low) to 1 (when both temperature and rainfall are high). An intermediate temperature or rainfall will result in a value of C < 1. Thus, inclusion of the climate modifier effectively reduces the rate of decomposition for each pool from its default value.

The values for and were based on work by Paul and Polglase (2004) who calibrated FullCAM using data from field data for estimating changes in debris and soil carbon associated with afforestation or deforestation. In that study, the parameter s was set to 0.06 and v was set to 0.1 (K. Paul, CSIRO, Pers. Comm.).

Based on these settings, the average decomposition rates for Tasmanian climate would effectively halve the default rates (i.e. with the climate modifier switched off).

Decomposition rates were adjusted after the inclusion of the climate modifier to obtain stable amounts for debris pools that were consistent with data from Forestry Tasmania and published figures from Woldendorp et al. (2002). In general, the default rates for most debris pools were suitable (Table 48). However, decay rates for stem and branch debris were decreased slightly for all forest types to maintain stable pools over time. In addition, the decay rate of resistant coarse roots for softwood plantations was decreased from the default figure of 50% to 20% to be consistent with rates for other forest types.

Page 235 of 365

Initial amounts of coarse woody debris and stags in eucalypt forest were estimated for each forest class based on stand productivity (for CWD, Equation 7) and stand productivity and age (for stags, Equation 8). Initial values for other debris pools were calculated from initial mass, carbon content, turnover rate and decomposition rate of each bio-component for each forest class, assuming they were in a state of equilibrium. An equation for estimating the equilibrium size of each debris pool was developed:

Equation 23 where: = the equilibrium amount of debris for component = the annual input for component = the annual decay rate for component These equilibrium amounts represent a simplification of the dynamic nature of these forests. However, it provided stable initial figures for debris from which initial soil carbon amounts could be estimated (see following section).

Table 48: Parameters used to estimate decomposition rates of debris pools for different forest types in the FCMF. Non-bold figures are defaults from DCCEE (2011b, Tables 7.B2 and 7.D3). Rainforest and Other Native Forest were assumed to have the same decomposition rates as Tall Eucalypt Forest. Low eucalypt Other native Hardwood Softwood Parameter Bio- forest forest Plantation Plantation component Decomp. Resist. Decomp. Resist. Decomp. Resist. Decomp. Resist. Fraction of total Stems 10% 90% 10% 90% 10% 90% 10% 90% Branches 15% 85% 15% 85% 15% 85% 15% 85% Bark 10% 90% 10% 90% 10% 90% 10% 90% Leaves 70% 30% 70% 30% 70% 30% 70% 30% Coarse roots 55% 45% 55% 45% 55% 45% 55% 45% Fine roots 70% 30% 70% 30% 70% 30% 70% 30% Decomposition Rate Stems 5% 5% 7% 7% 7% 7% 12% 12% Branches 7% 7% 10% 10% 10% 10% 15% 15% Bark 35% 35% 50% 50% 50% 50% 50% 50% Leaves 56% 56% 80% 80% 80% 80% 100% 100% Coarse roots 28% 7% 40% 10% 40% 10% 50% 20% Fine roots 70% 70% 100% 100% 100% 100% 100% 100%

Page 236 of 365

6.10 SOIL CARBON

Estimates of amounts of soil carbon include pre-clearing soil carbon (used for estimating initial carbon stocks under relatively undisturbed) native vegetation and post-clearing soil carbon (used in modelling initial carbon pools under agricultural land, Webb 2000, ASRIS, 2012). Both carbon layers were made available to the Carbon Study by DPIPWE and CSIRO. Total carbon mass (T, t/ha) for the post-clearing layer was calculated from carbon concentration (C, %), bulk density (BD, kg/m3) and fraction of coarse material (R) as follows:

– Equation 24

The post-clearing layer was used for modelling carbon under native forest, while the post- clearing layer was used for modelling carbon under plantations and on grassland prior to conversion to plantation.

In the data provided by CSIRO from the ASRIS system (2012, Figure 97), a large proportion of Tasmania has no values for soil carbon. This is because: • there is no data for some soil types under native vegetation for the pre-clearing layer and • the post-clearing soil carbon sampling only included soil types from cleared land.

Further, these layers do not disaggregate soil carbon into the separate pools modelled in Roth- C. Thus, they cannot be used to initialise the model directly. Instead, they were used in the calibration process (described below) to ensure that calculated total soil carbon values were not unrealistically low or high.

To model soil carbon cycling, the Roth-C model was used (Version 6.3, Coleman & Jenkinson, 2008, see Section 7.2). To initialise the various carbon pools in this model, it is necessary to run it for an extended period of time (e.g. 200-1000 years). To do this, without having to model the growth of the entire Stand for this period, average debris decomposition rates were calculated based on the equilibrium mass of each debris pool, and estimated decomposition rates (see Equation 23 in Section 6.9).

For modelling grassland, there was no data on debris input rates. Thus, an algorithm was developed to calculate the required rate of input to maintain soil carbon at its post-clearing level for a given temperature and rainfall. This debris input rate was maintained until the grassland was converted to plantation.

Page 237 of 365

a) b)

Soil Carbon (t/ha) Soil Carbon (t/ha)

Figure 97: Pre-clearing (a) and post-clearing (b) soil carbon layers used to model initial soil carbon across Tasmania. Source: Darren Kidd (DPIPWE); data processed and represented by CO2 Australia.

6.11 FIRE

Fire frequency

Fire frequency was modelled using information from published studies by Jackson, (1968), Marsden-Smedley (1998, 2007), Lindenmayer (2009), McCarthy et al. (1999), Mount (1979), Wood et al. (2011) and Wood and Bowman, (2011). Past research on fire frequency and its effects on Tasmania’s forests are reviewed in detail Section 7.8. Here we explain the approach used to model fire and its effects on stand growth and carbon pools. Initially, it was assumed that fire would be modelled using a simple average frequency for different forest types for high intensity and low intensity fires with a set variance based on a normal distribution. However, a number of issues were identified with this approach including: • it left little scope for modelling variation in fire intensity in a probabilistic manner; • it did not accurately model the potential for fire in very old Stands (where the frequency was effectively modelled as 0); • it did not allow for complex probability distributions as suggested by Jackson et al. (1968) and McCarthy et al. (1999) in which likelihood of fire may vary with Stand age.

Page 238 of 365

Thus, it was considered that a simple normal distribution of fire frequency would not accurately represent variation in frequency, or intensity for native forests. Instead, it was assumed that fire risk could be modelled using a probability distribution. For Tall, and Low, Eucalypt Forests, it was assumed that the probability of fire increased over time in young stands up to a maximum, after which it declined. A triangular distribution was selected as the simplest way to model this variation in fire probability over time (Figure 98).

Based on discussions with researchers (D. Bowman, and J. Marsden Smedley University of Tasmania, Pers. Comm.), the mode (peak probability) was set at age 50 years for Tall Eucalypt Forests and 25 years for Low Eucalypt Forests. The maximum age for the distribution was based on the maximum known age of Stands of Tall Eucalypt Forests and was assumed to be 600 years for these based on data from Wood et al. (Wood et al. 2010) and 250 years for Low Eucalypt Forests.

Because the area under the probability distribution must equal 1 for a Probability Distribution Frequency, the maximum likelihood of fire can be readily calculated from the maximum Stand age. The value of Px is such that the area under the triangle is 1. Therefore:

Equation 25

Equation 26

0.004 1.2

PDF 0.0035 CDF 1 0.003 0.8 0.0025

0.002 0.6

0.0015 0.4 0.001

Annual probability of fire (PDF) fire of probability Annual 0.2

0.0005 Cumulative probability of fire (CDF)

0 0 0 100 200 300 400 500 600 Stand Age

Figure 98: Example of a triangle probability distribution of fire risk with stand age in a Tall Eucalypt Forest.

Page 239 of 365

where: = the age at which fire risk is greatest, = the maximum risk of fire, and = the age at which the probability of fire falls to zero. The equation for the Probability Distribution Frequency is in two parts:

Equation 27 where is the probability of fire occurring at Age .

This distribution can be used to estimate the chance of fire at any given age. With the addition of a random number generator, fire events can be modelled for a given Stand in a probabilistic manner. However, this approach has a problem, in that when a Stand age exceeds the maximum assumed age, the function assumes that it cannot be burnt. This is unrealistic as even Rainforests are sometimes burnt by wildfire. Thus, it was assumed that there was a base fire frequency across all forest types that could be varied from 0 to 1 (100% chance of fire at all ages). Using this approach, a set of distributions of probabilities of fire by age was developed for each forest types (Figure 99).

0.06

0.05

0.04

0.03 Other Non-eucalypt Forest 0.02 Low eucalypt forest Tall eucalypt forest

Annual probability of fire (P) fire of probability Annual Rainforest 0.01

0 0 100 200 300 400 500 600 Stand Age (years)

Figure 99: Assumed distributions of probabilities of fire for different forest types and ages used for the Study Baseline.

Page 240 of 365

For scenarios in which the frequency of fire was raised or lowered (F1-F4 and L3-L4) the probability distributions were scaled up or down accordingly. For example, to model a 10% increase in fire frequency in Low Eucalypt Forest, the base was increased from 0.02 to 0.022 while the peak fire risk (at age 50) was raised from 0.024 to 0.0264. The full set of frequencies used for the Study Baseline and Study Scenarios is provided in Appendix 2.

Fire intensity

Fire intensity affects both the proportion of trees killed in a given fire as well as the amount of carbon that is lost from forest carbon pools. Thus, it effects carbon stocks directly, by moving carbon to debris pools and the atmosphere, and indirectly, by changing the structure and subsequent growth of the stand. Crucially, fires may kill only a proportion of trees, resulting in multi-aged stands that are made up of old-growth and regenerating trees of varying ages. Such stands have been shown to be relatively common, not just in low eucalypt (i.e. dry sclerophyll) forests, but also in tall eucalypt (wet sclerophyll forests) dominated by Eucalyptus regnans (e.g. Turner et al. 2009, Ough and Ross 1992; Lindenmayer et al. 1999). Thus, the FCMF was designed to model the effects of varying fire intensity on tree mortality and carbon stocks as well as the growth of the subsequent cohorts of different ages.

Fire intensity was modelled on the assumption that the probability of a tree being killed by a fire decreases with its age. For example, a moderate intensity fire was assumed to be more likely to kill younger, smaller trees than older, larger trees in a given Stand. To model this effect, relationships between tree age and fire risk were developed for each forest type (Figure 100). The key parameter in these relationships is the proportion of Big Trees killed.

The age of a “Big Tree” may be varied for a given forest type. Since Tall Eucalypt Forests are typically more sensitive to fire than Low Eucalypt Forests, the age of Big Trees Killed was set at 80 years for the former and 40 years for the latter. This meant that a fire of a given intensity was less likely to kill trees in a Low Eucalypt Forest compared with a Tall Eucalypt Forest of the same age.

We assumed the average proportion of big trees killed by fire was 50% for Tall Eucalypt Forests based on work by McCarthy et al. (1999), 20% for dry eucalypt forests, 75% for Rainforest and 100% for hardwood and softwood plantations. We also assumed fire intensity could vary by +/- 50% (coefficient of variation) for all forest types except Rainforest for which the variation was set at +/- 25%. The resulting model provided a wide range in potential tree mortality and fire intensity for a given fire event in each forest type (e.g. Figure 101). The resulting multi-cohort stands were modelled as described in Section 6.4.

Page 241 of 365

1 Rainforest Big Tree: RFT Tall eucalypt forest Big Tree: ETF Low eucalypt forest Big Tree: ELF 0.8 Other Non-Forest Big Tree: ONF

0.6

0.4

0.2 Average proportion of trees killed by fire killed trees of proportion Average 0 0 100 200 300 400 500 Tree age (years)

Figure 100: Assumed average proportions of trees of different ages killed in a given fire event for different forest types. Losses of debris and soil due to fire were based on the average intensity as set by the proportion of Big Trees killed. Little data was available for estimating carbon losses from debris and live pools associated with fire. However, based on work by Norris (2010) and Lindenmayer (Pers. Comm.) in Australia and Bormann (2008) and Hurteau (2011) in the USA, we estimated losses from each debris pools. Losses of soil carbon were based on research of losses after regeneration burns by Dyrness et al. (1989), Ellis and Graley (1983), Hatch (1960), Pennington et al. (2001) and Tomkins et al. (1991).

It was assumed that losses increased with increasing fire intensity (as set by the proportion of big trees killed). It was also assumed that a larger proportion of smaller material (i.e. bark and leaves) would be lost compared with larger material (stem debris). The proportion of each debris pool combusted was set using a multiplier of the proportion of Big Trees killed as follows: • Stags: 1 • Stem debris: 2 • Branch debris: 2 • Bark debris: 3 • Leaf debris: 5 • Humus: 0.2

Page 242 of 365

Since the average proportion of Big Trees killed was assumed to be 50% for Tall Eucalypt Forests and 20% for Low Eucalypt Forests, debris losses tended to be 2.5 times greater for the former. This is consistent with the tendency for less frequent but higher intensity fires in Tall Eucalypt Forests compared with Low Eucalypt Forests.

Of the proportion of each component combusted, 95% was assumed to be volatilised with 5% being converted to inert carbon (i.e. charcoal) except for Stags, where it was assumed that 80% was volatilised. It was also assumed that no roots or live tree components were combusted in fire. However tree mortality from fire resulted in live tree components being shifted to debris pools.

0.8

0.6

0.4

Upper limit 0.2 Average Lower limit Proportion of killed trees by fire Assumed average age and proportion of big trees killed 0 0 50 100 150 200 Tree Age (years)

Figure 101: Example of modelled variation in proportion of trees killed in a given fire event for a Tall Eucalypt Forest. Upper and lower limits are based on +/- one standard deviation from the mean.

Page 243 of 365

6.12 TIMBER HARVESTING

Harvest regimes

Harvesting rates and intensity were assumed to vary by forest type, age and harvesting history. Only harvestable Stands (those grid points not included in formal or informal reserves) were assumed to be able to be cut. The assumed rates and intensities of harvest for the Study Baseline were set as follows:

Tall eucalypt forests • Thinning: age 10-15, 7% total area harvested each year, 50% stemwood removed • Clearfell: o if thinned - age 60-150 years, 5% of eligible Stands cut each year; o if not thinned - age 80-150 years, 5% of eligible Stands cut each year; o for both thinned and unthinned harvested Stands - 90% of stem volume harvested, 85% of harvested stemwood removed, 80% chance of coupe being burnt, o if >150 years - 5% of Stands cut each year, 90% of total volume harvested, 85% stemwood removed, 80% of coupes burnt.

Low eucalypt forests • Cohorts 60-100 years: partial cut, 5% of total area harvested each year, 50% of total volume harvested, 85% of harvested stemwood removed, 70% coupes burnt. • Cohorts >100 years: partial cut, 5% of total area harvested each year, 90% of total volume harvested, 85% of harvested stemwood removed, 70% coupes burnt.

Other Native Forests • 1-1000 years: 5% cut each year, 50% of total volume cut, 80% of stemwood removed, 75% of coupes burnt.

Rainforests • 1-1000 years: 0.5% cut each year, 90% of total volume cut, 85% of stemwood removed, 75% of coupes burnt.

Softwood plantations • Thinning: age 14-18 years, 20% of total area harvested each year, 50% of stemwood cut, 85% harvested stems removed. • Clearfell: >23 years if unthinned or >28 years if thinned, 100% of volume cut, 90% stemwood removed, 25% of coupes burnt.

Page 244 of 365

Hardwood plantations • Thinning: age 6-8 years (sawlog plantation only), 40% of total area harvested per year, 50% of stem volume cut, 80% of harvested stemwood removed. • Clearfell: o if thinned: age 28 years, 100% of volume cut, 90% of harvested stemwood, 25% of coupes burnt, o if unthinned: >13 years, 22% harvested each year, 100% of volume cut, 90% of harvested stemwood removed, 25% of coupes burnt. Stemwood left on site was assumed to be 10% of the total stem mass for softwood and hardwood plantations and 15% for native forests based on studies by Raison and Squire (2007) and Ximenes et al. (2008a) and information from Forestry Tasmania (D. Mannes, Pers. Comm.).

Log quotas

To ensure that the rate of harvesting matched the forecast rates set by Forestry Tasmania (2011) or historic rates for private native forests, sawlog harvesting quotas were used. These were based on the rate of sawlog removal. For the Study Baseline, sawlog quotas were set at 300,000 m3/y for public native forests and hardwood plantations and at 55,000 m3/y for private native forests for the period 2010-2050, based on the average rate of sawlog production from private forests from 2006-2011 (Forestry Tasmania 2008, 2010 and ABARES 2012a). For the period 1990-2010, sawlog quotas were set at 400,000 m3/y for public native forests and hardwood plantations and at 110,000 m3/y for private native forests, based on the average rate of sawlog production from private forests from 1998-2005 (Forestry Tasmania 2005a, 2007 and ABARE 2001).

No log quotas were used for private hardwood plantations (primarily pulpwood) or softwood plantations.

The proportions of harvested wood going to sawlog were based on harvested log volumes from Forestry Tasmania annual reports (2008, 2010) and data from ABARES (2012a). These indicated that the average proportion for the period 2001-2011 ranged from 10-15% for public native forests and from 4-6% for private native forests. Thus, it was assumed that the average sawlog percentage was 13% for the former and 6% for the latter (which is slightly less than the 15% assumed by Dean and Wardell Johnson 2010). However, rather than simply setting this proportion as a fixed value, a function was added that allows the sawlog proportion to be decreased based on stem diameter. This allows the sawlog proportion to vary line with the size of the trees harvested. The actual average proportion of sawlogs produced through this process varied 12-13% for public native forests and 5-6% for private native forests under the Study Baseline.

Page 245 of 365

Assumed breakdown rates of stems into different log types were based on the production rates from Forestry Tasmania’s annual reports (Forestry Tasmania 2008, 2011a). More detailed information on the conversion efficiency to sawlogs for stems from Stands of different Stand ages was obtained from Forestry Tasmania’s wood flow report (Forestry Tasmania 2007). Assumed percentages of other log types were called by the proportion of sawlogs with averages set for different forest types as follows: • Public native forests: o Peeler logs: 8% o Chips: 78% • Private native forests: o Peeler logs: 8% o Chips: 86%% • Rainforest: o Sawlogs: 14% o Chiplogs: 86% • Other native forests: o Sawlogs: 10% o Chiplogs: 90% • Hardwood plantations: o Sawlogs: 38% o Chiplogs 62% • Softwood Plantations: o Poles: 1% o Sawlogs: 47% o Peeler logs: 5% o Chiplogs: 47%

Carbon losses from regeneration burns

Harvested biomass not removed from the site (including a proportion of stemwood and all branches bark, leaves and roots of harvested trees), was assumed to be harvesting slash and was moved to debris pools. If the coupe was burnt following harvest (randomly determined from the assumed likelihood of burning) then a proportion of this material was assumed to be combusted together with existing debris. The remaining material was assumed to decay at the same rate as debris arising from Stand Turnover and was modelled as such.

Page 246 of 365

The proportion of harvesting slash combusted (and mortality of retained trees) in regeneration burns was assumed to increase with fire intensity. Fire intensity was modelled in the same way as for wildfires, through the proportion of Big Trees killed. This was assumed to vary with harvesting system as follows:

• Post clearfell: 80% killed with the age of Big Trees set at 80 years and • Post partial harvest: 10% killed with the age of Big Trees set at 40 years.

The proportion of carbon from debris and soil pools combusted in regeneration burns was calculated from the proportions of Big Trees killed using the same multipliers as for wildfire (see previous section). This approach assumed that the proportion of debris and soil humus combusted was lower for partial harvests than for clearfell. Since the likelihood of clearfell was assumed to be lower for Low Eucalypt Forests than other forest types, this tended resulted in lower losses from post harvest burns in these forest types. This is consistent with the current harvesting and regeneration regimes used in different forest types (e.g. Elliot et al. 2007). However, it may be conservative (i.e. result in higher than actual carbon losses) in terms of estimated emissions from Old Growth harvesting if clearfell systems are phased out in these forests (e.g. Forestry Tasmania 2005b).

Importantly, this approach allows losses of soil carbon from regeneration burns to be modelled. As mentioned, FullCAM lacks this capacity even though soil carbon losses have been shown to be an important factor in carbon fluxes associated with timber harvesting (Dean and Wardwell Johnson 2010). We assumed that the average loss of humus carbon (comprising around 90% of total soil carbon) from the surface 0.3 m was 16% for post-clearfell regeneration burns and 4% for post-partial-harvest burns. Assumed losses were considerably greater than those assumed by Dean and Wardell Johnson (2.5% of total soil carbon, 2010).

Unlike CO2 emissions from regeneration burns which are accounted for in the changes in carbon stocks in vegetation, debris and soil pools, non-CO2 emissions associated with burning biomass must be accounted for as separate emissions (IPCC 2006b). These were estimated from the assumed amount of material combusted using specific emission factors for each non- CO2 greenhouse gas as recommended by the IPCC. The factors used were the same as those applied by the DCCEE for estimating the National Greenhouse Gas Accounts and are shown in Table 49.

Page 247 of 365

Table 49: Assumed emission factors, for GHGs covered under biomass burning, and global warming potentials based on 100-year global warming potentials reported (IPCC, 1996). Source: DCCEE (2011b). Gas Element Emission factor (mass Elemental to Global Warming of element in gas/mass molecular mass Potential element in fuel burnt) conversion factor

CO2 C 1 3.67 1

CH4 C 0.0054 1.33 21

N2O N 0.0077 1.57 310

NOx N 0.15 3.29 NA CO C 0.091 2.33 NA NMVOC C 0.022 1.17 NA

6.13 EMISSIONS FROM FOREST MANAGEMENT & HARVESTING

In addition to emissions resulting from changes in carbon stocks in vegetation, debris and soil and non-CO2 emissions associated with regeneration burns, forest management activities produce a range of other CO2 and non-CO2 greenhouse gas emissions. Potentially accountable emissions from these activities include CO2 and non-CO2 emissions associated with fossil fuel use, and fertiliser application. Non-CO2 emissions associated with wildfire were not included in the results.

Greenhouse gas emissions associated with forest management and harvesting were estimated from a study reported by Tucker et al. (2009). This study included three native hardwood case studies (including Tasmania) and four plantation softwood studies and covered all aspects of forest management, burning and harvest, including transport of logs to sawmills. The average production of greenhouse gasses per unit log extracted was 40 kg/m3 for softwood plantations and 85 kg/m3 for native forests. A large proportion of emissions were associated with non- greenhouse gas emission from burning of harvest slash. Since these emissions are modelled separately in the FCMF, only those emissions from sources other than regeneration burns were included from the estimates from Tucker et al. 2009. These were:

• native hardwood: 50 kg/m3 • plantation softwood: 25 kg/m3 There is no data for emissions from fossil fuel and chemical use in hardwood plantations. Thus, it was assumed that these were mid-way between those for native hardwood and plantations softwood (i.e. 35 kg/m3).

Page 248 of 365

These emissions were associated with wood production and were thus related to the total merchantable volume of wood harvested each year. Emissions associated with management of National Parks and other non-harvestable forest were not included in the Carbon Study. The Study also did not consider emissions associated with processing of logs and sawn timber or other downstream production. Results from Tucker et al. (2009) indicate that emissions associated with sawmilling and other downstream processing are relatively small, in absolute terms) compared with fossil fuel and non-greenhouse gas emissions associated with harvesting and log transport.

6.14 WOOD PRODUCTS

Wood products were modelled according to the general methodology used in FullCAM, but with more explicit product types and processes to allow increased transparency and ease of data input. Estimates for proportions of products exported and converted to different end products (e.g. sawn timber, veneer, plywood and paper) were based on data from ABARES (2012a, average for period 2006-11) and published studies by ABARE (Burns et al. 2009) and data from Tucker et al. (2009). Exported wood products were assumed to be immediately converted to CO2 based on DCCEE guidelines (2011b). This approach was used for both raw material (e.g. woodchips) and finished products (e.g. veneer and sawn timber). Thus, estimates of carbon storage in wood products are considered very conservative.

The assumed outputs and decomposition rates for the various products are shown in Figure 102.

Finished products retained within Australia (i.e. not exported) were added to wood product pools. These pools were assumed to decay over time at default rates recommended by the UNFCCC (2010). The half lives used were:

• Sawn timber: 35 years • Panels and plywood: 25 years • Paper: 2 years.

Page 249 of 365

Sawlog processed domestically Chips: 35%

Sawn timber: 40% Biofuel and garden products: 25% Emitted as CO2

Sawn Timber

Used domestically: 100% Exported: 0%

Peeler log processed domestically

Chips: 40% Veneer: 35% Posts: 40%

Biofuel and garden products: 25%

Posts

Domestic use: 87% Export: 13%

Veneer

Domestic plywood: 24% Product Pools

Export: 76% Sawn Timber: 35 year half life Domestic panel production Panels: 25 year half life

Panels: 80%

Biofuel: 20% Paper: 2 year half life

Pulplogs and chips

Domestic panel production: 1%

Domestic paper production: 27% Export: 72%

Domestic paper production Paper: 70%

Biofuel: 30%

Figure 102: Diagram showing the various product pools modelled and processing steps, assumed proportions exported or recovered in products and by-products, and half live for finished products.

Page 250 of 365

6.15 INPUT DATASET

The input data for the model was based on a series of data layers holding the required spatial information required for it to run. These layers cover the whole of Tasmania, including Flinders and King Islands and included: • Stand growth inputs: (from which species, age, and productivity were derived): o Forest Class from 2010; o Forest Class from 2001; and o Forest Class from 1996. • Forest Group: including the 6 Forest Groups plus additional classes covering non-forest native vegetation > 2 m, native vegetation < 2 m, agricultural grassland and non-plantable areas. • Climate data: average monthly temperature, rainfall, evaporation and soil moisture deficit. • Forest Productivity Index (to derive plantation growth rate). • Soil data: percentage clay and t/ha carbon. • Tenure: private or public. • Management: harvestable or non harvestable (including both formal and informal reserves). • Fire history (time of last wildfire).

The land surface of Tasmania and surrounding islands was divided into a 1km2 grid, which was then used to extract data from the various polygon and raster layers. It was assumed, the portion of each layer that intersected the centroid of a particular grid cell was representative of conditions at that location. It was assumed that, over an area of the size of Tasmania this method would provide a representative sample of each major forest type.

6.16 ESTIMATING CARBON CARRYING CAPACITY

Carbon Carrying Capacity of Tasmania’s forests was estimated based on the definition used by Roxburgh et al. (2006) as: “the mass of carbon able to be stored in a forest ecosystem under prevailing environmental conditions and natural disturbance regimes, but excluding anthropogenic disturbance”.

Under this definition, the effects of prevailing fire regimes and other disturbance events on forest carbon stocks must be taken into account, but the effects of timber harvesting and other disturbances from humans are excluded.

Page 251 of 365

Carbon Carrying Capacity was estimated by running the model from 2010 to 2250 with no timber harvesting and assessing changes in carbon stocks in vegetation, debris and soil. It was assumed fire frequency would be a major influence on Carbon Carrying Capacity, so the model was run with three fire regimes: • L1: the Study Baseline fire regime (assumed to be consistent with current fire regimes for different forest types); • L3: 20% decrease in fire frequency across all forest types; and • L4: 20% increase in fire frequency across all forest types. Since plantations are, by definition, man-made forest, it is not possible to readily estimate their carbon stocks in the absence of human inputs, or model the effects of prevailing natural disturbance regimes (e.g. a softwood plantation burnt by wildfire would be unlikely to regenerate). Consequently, plantations were excluded from estimated Carbon Carrying Capacity.

6.17 MODELLING THE EFFECTS OF CLIMATE CHANGE

Climate change is expected to result in increased temperatures and changes in rainfall across much of Australia. These changes will influence, not just the growth rates of trees, but also the rates of turnover and decomposition of carbon in debris and soil. Like FullCAM, the FCMF does not have the capacity to model the effects of changes in climate on forest growth rates directly. However, predicted relative changes in growth rates from other studies can be used to adjust future growth of different forest types. However, the FCMF models the effects of variations in temperature, rainfall and evaporation on decomposition rates of debris and soil dynamically using the climate modifiers from Roth-C for soil carbon and FullCAM for debris. Three climate projections were selected from a study by ABARES (2012b) which modelled the effect of climate change on growth of eucalypt and pine plantations as a well as native forests. This study used projections from 23 global climate models under two emission scenarios (SRES A1b and A2, IPCC 2000) downscaled to a resolution of 10 km, to model future growth with the model 3PG2. This study was selected because: • data for the climate projections and emission scenarios used in the study was available at the scale used in the study from CSIRO; • data for changes in growth associated with climate projections from individual climate models for each emission scenarios were available; and • the study included projected changes in growth for both native forests and plantations. A major consideration in this selection was the joint availability of modelled growth predictions together with the climate projections on which they were based. Advice from CSIRO was that it

Page 252 of 365

was important to avoid complications associated with using climate and growth datasets that were poorly matched (L. Webb, CSIRO CMAR, Pers. Comm.). This was also the reason that the CSIRO downscaled climate data was used in preference to that potentially available from the Climate Futures for Tasmania project (AEC CRC 2010). Three climate projections used in the ABARES were selected for the Baseline Sensitivity Scenarios modelled in the Carbon Study. These were: • CSIRO Mk 3.5 SRES A1b, • CSIRO Mk 3.5 SRES A2, and • CSIRO Mk 3.0 SRES A2. Spatial data on the relative changes growth for each forest type modelled in the study across Tasmania is held by ABARES. However, this could not be provided in time to be included in the analysis. As a result, only average changes in growth for each forest type across Tasmania for selected climate projections were used. These were calculated from average changes in MAI and the baseline MAI for each combination of forest type and climate projection reported by ABARES (2012b, Table 50). The percent change in native forest growth was based on the change in growth bluegum sawlog regimes as described by ABARES (2012b).

Table 50: Estimated changes in growth rates for hardwood and softwood plantations in 2030 and 2050 from ABARES (2012b) used to model the effect of climate change projections on average forest growth rates in the Carbon Study. Parameter Bluegum pulpwood Bluegum sawlog P. radiata 2030 2050 Avg 2030 2050 Avg 2030 2050 Avg Change in MAI (m3/ha/y) C1 6.48 4.17 4.03 4.71 0.78 0.38 C2 2.48 0.3 0.45 0.29 -1.13 -2.61 C3 -1.49 2.46 -1.55 -0.78 -2.09 -3.43 Average Current Growth Rate (m3/ha/y) 26.3 26.3 28.5 28.5 28.5 28.5 Change in MAI (%) C1 25% 16% 20% 14% 17% 15% 3% 1% 2% C2 9% 1% 5% 2% 1% 1% -4% -9% -7% C3 -6% 9% 2% -5% -3% -4% -7% -12% -10% CSIRO provided climate projections for each of the climate Baseline Sensitivity Scenarios at the same resolution as that used by ABARES to predict growth for the period 2000-2050. Evaporation data was not required for the original growth modelling by ABARES, but is required to model soil carbon in Roth-C. As a result, CSIRO developed new spatial layers for potential areal evaporation downscaled to a resolution of 25 km specifically for the Carbon Study (C. Heady, CMAR CSIRO, Pers. Comm.).

Page 253 of 365

The downscaled climate data for each climate scenario was incorporated into the FCMF together with the respective average changes in growth for each forest type. Only climate data for 2010, 2030 and 2050 was used into the FCMF with intervening data interpolated. This was necessary to allow the model to run efficiently (i.e. < 4 hours per scenario).

6.18 UNCERTAINTY ANALYSES

As detailed in preceding sections and Section 6.6, the modelling approach applied under the Carbon Study includes reference to a number of assumptions, incomplete data-sets, and information of varying accuracy. To assess the effect of uncertainty of parameter values on the results, each parameter was allocated with a coefficient of variation (CV) that was used to randomly vary the value for each parameter for an entire model run. These CVs were based on using known variation in the raw data, or were estimated based on expert opinion. They ranged from 5-50%, depending on the parameter, and averaged 20%. Altogether, it was possible to vary up to 200 parameters each run. To allow meaningful comparisons between outputs from the Study Baseline and Study Scenarios, in the primary runs, all parameters were fixed at their mean value with the exception of fire and harvesting, which were set a fixed random probability for each grid point and year.

The variation of each parameter was based on a normal distribution with values permitted to deviate according to the resulting probability distribution. A variety of systems were tested with the final one based on each parameter being fixed for a certain value across all grid points for an entire run. The model was rerun several times, each time with a new random setting for each parameter until the amount of variation in the results stabilised. This approach allowed the standard deviation and coefficient of variation (SD/n) in the results to be estimated.

Because results for the Study Scenarios are expressed relative to the Study Baseline, a preferred approach would be to vary the same parameter settings for multiple runs of both the Study Baseline and each Study Scenario and to calculate the variation around the difference between these each time. However, the FCMF does not yet have the capacity to handle this type of analysis. Instead, the parameters in the Study Scenarios were allowed to vary, while those for the Study Baseline were kept constant. This allowed the standard deviations and CVs of the difference between each Study Scenario and the Study Baseline to be estimated.

It is important to note that, while this method provides an indication of variation in estimates, it should not be considered a definitive analysis of uncertainty. This would require many more runs of the model with uncertainty for the Study Baseline and Study Scenarios synchronised.

In the uncertainty analysis applied here, a distinction was made between the effects of natural variability (e.g. in fire frequency and intensity) and uncertainty in regard to estimates. This

Page 254 of 365

distinction is important because the former concerns variation which is an intrinsic component of the forest ecology, while the latter represents a gap in current knowledge and understanding.

Our analysis separated the effect of uncertainty in data from the natural variation in bushfire and random effect of harvesting. These latter parameters were permitted to vary randomly each model run and thus provide a degree of spatial uncertainty to the results. However, since most runs were across >10,000 grid points, the effect of this variation on overall totals was minimal because the variation in results for individual grid points tends to cancel out with large numbers of points. This issue should be reviewed for any future studies.

Variables around which errors occur that are likely to influence estimation outcomes include: • harvesting rates based on assumed harvesting quotas and sawlog recovery rates; • growth rates of trees of different forest types and age classes; • expansion factors used to estimate under-storey and non-merchantable components in native forests; • long term carbon trajectories for mature forest classes; • areas of different forest classes within various forest tenures; • amounts of coarse woody debris in different forest types; • decomposition rates of coarse woody debris and soil humus; • losses associated with fire, harvesting and site preparation; • proportions of sawlog recovery; • wood product destination and half life and; • emissions associated with management, harvest and processing; • potential carbon prices; and • eligibility of changes in various carbon stocks.

6.19 KEY ASSUMPTIONS AND DATA GAPS

Forest growth rates

Forest growth is a key driver of total carbon sequestration and flows. Thus, the model used and associated assumptions have a major bearing on the results. In this study, actual plot data from eucalypt forest was used to develop a growth model for estimating stem volume for all native Forest Classes, while published growth models were used to estimate growth of plantations.

The native eucalypt model derived from Forest Class data assumes that Stand volume can be represented by a smooth monotonically increasing asymptotic curve until the Stand reaches

Page 255 of 365

senescence. Growth rates of individual Stands probably fluctuate over time, but we consider these smooth curves to be reasonable representations of average growth.

These curves were derived using assumed average ages, some of which we know to be fairly accurate (e.g. aged regrowth). Regrowth forest classes, which have the greatest effect on curve slope, were estimated using height growth equations given known regrowth height and potential height. Mature (forest > 110 years) was assumed to have an average age of 150 years and the average stem volume for each mature Forest Class was assumed to be 95% of the maximum potential volume that could be achieved for given stand productivity class.

While it could be argued that this approach is overly conservative, the growth curves obtained were consistent with average stem volumes from inventory plots. Further, this approach is far less conservative compared to FullCAM growth projections derived from the Forest Productivity Index, which have at times been found to be around 50% lower than actual measurements for more productive forests (e.g. Roxburgh et al. 2010).

In contrast to the extensive database on growth rates of Tasmanian eucalypt forest, there is little data on stem volumes in non-eucalypt native forest. Although some field sampling in Rainforest and adjacent mature eucalypt forest was undertaken as part of the Carbon Study, estimation of carbon stocks in non-eucalypt forest and low vegetation remains a significant knowledge gap that should be addressed in any future studies.

Conversion of stem volume into total biomass

We estimated standing biomass from entire stem volume using the default figures that are used by the DCCEE (2011b) for basic densities for each forest type and proportions of oven-dry weight of stem, branch, bark, leaf, coarse roots and fine roots.

There is little published data on the biomass allocation for many of Tasmania’s forest types. In particular, there are no published figures for Northofagus Cunninghammii and other Rainforest and scrub species.

Both wood basic density and allocation of biomass to tree components vary with age. However, we assumed these remained constant. Since most of the carbon in mature trees is held in the stem, this assumption is not expected to be a major influence on estimates. However, it would be useful to develop age related relationships for biomass allocation.

Soil & debris

Coarse Woody Debris (CWD) volume is estimated on many Forestry Tasmania plots, and is summarised for many of the Forest Classes. While these data represent the most comprehensive survey of CWD in Tasmania, it does not reference non-eucalypt forests. Limited field data was obtained as part of the project for Rainforest and mature eucalypt forest.

Page 256 of 365

However, more extensive field surveys would improve the level of confidence in estimates. In particular, the dynamics of stem mortality, Stag fall, coarse woody debris production and debris decay rates in Stands of different ages and productivities need to be measured and the assumptions used in the modelling tested.

Harvesting & regeneration of native forests

In general, publicly owned native forests in Tasmania are successfully regenerated after harvest. Figures provided by Forestry Tasmania to the NFI indicate that between 93% and 99% of the total area of native forest harvested each year from 1998-99 to 2005-06 achieved required stocking levels for regeneration.

Forestry Tasmania classifies potential growth of regeneration into four mature height classes: >41 m; 34-41 m; 27-34 m; and 15-27 m. Harvest regeneration is classified according to the decade harvested (<1959, 1960s, 1970s, 1980s, 1990s and 2000s) and its potential mature height. Growth curves for each of these height classes have been estimated from sample plots randomly located in each forest class for each regeneration period.

We assumed that when existing mature, or regrowth, forest is harvested, the height potential of the subsequent regeneration will be similar to that of the harvested forest. Thus, regeneration on a site that previously held tall mature forest will have a greater height potential (and thus faster volume growth) than regeneration from shorter forest. It should be noted current height classes of mature forest differ from those of the regeneration. The actual growth of regenerating stands needs to be more fully quantified to determine its long- term growth trajectory.

Effects of fire

Fire regimes were based on published studies of fire frequency across different forest types. However, these are based on little field data. In particular, the instance of large, stand-replacing wildfires is very difficult to predict, with past events associated with relatively rare weather conditions. For example, the 1967 fires that burnt much of the forest around Hobart were associated with a once in a 100 year or more weather event. Very rare, but highly significant, fire events such as these are almost impossible to predict with any certainty (D. Bowman, University of Tasmania, Pers. Comm.).

In the Carbon Study, we effectively assumed that large wildfire events occur over a small proportion (0.5% for Tall Eucalypt Forest and 2% for Low Eucalypt Forest) of the total forest estate each year. While this method is convenient and gives a useful indication of the effect of fire on long term carbon stocks, it does not replicate the effect of extensive wildfires that may burn large areas of forest in a single year. These large fire events could have major implications for carbon stocks which could fluctuate sharply over time, rather than being in smooth steady state as reflected by the modelling approach used here.

Page 257 of 365

The effects of fire on carbon stocks in vegetation, debris and soil are poorly understood. Figures used in the Carbon Study were largely based on limited studies of carbon losses associated with regeneration burns. A number of studies on the effects of fire on carbon stocks are currently underway, but no results have yet been published (D. Lindenmayer Pers. Comm. J. Norris Pers. Comm.). This data will improve understanding of changes of carbon pools associated with fire and coarse woody debris decomposition.

Debris input and decomposition

There is little information on rate of input and decomposition of debris in native forests since most published studies focus on litter-fall and decomposition in plantations. In particular, there are few studies of coarse woody debris and rates of stem mortality in native forests. Stem mortality due to self thinning and senescence is assumed to be zero in the FullCAM model. Since coarse woody debris comprise the largest debris pool and have a relatively slow decomposition rate, quantifying the amount of coarse woody debris and rates of input and decomposition for different forest types is crucial for better quantifying carbon stocks in native forest debris pools.

Influence of climate on decomposition

Although the publicly available version of FullCAM includes optional parameters for including the effect of temperature and rainfall on decomposition rates, these are normally turned off and there is little supporting material on appropriate settings. In the Carbon Study, the settings suggested by Paul et al. (2006) were used. However, these settings are based on a calibration for softwood plantations and it is not known if they are appropriate for native forests. Better understanding the effect of changes in temperature and rainfall on debris decomposition rates in native forests is crucial for being able to quantify the potential effect of climate change on debris carbon pools as well as for improving modelling of debris pools in general.

Soil carbon

Soil carbon stocks are modelled in FullCAM and the FCMF using the Roth C model (version 6.3 for the FCMF and 6.5 for the latest version of FullCAM). The parameters for this model are based on the initial parameterisation by its developers (Jenkinson, 1990; Jenkinson et al. 1987, 1999; Coleman & Jenkinson, 2008). Testing by Paul et al. (2002, 2003) demonstrated these parameters produced realistic results for soil carbon in grassland converted to plantation. However, no studies have investigated whether they are suitable for native forests.

A key issue for soil carbon modelling using the Roth C model is that it is difficult to obtain amounts of soil carbon that consistent with actual soil carbon amounts in high carbon soils. For example, across Tasmania, many soils have in excess of 150 t C/ha carbon. However,

Page 258 of 365

equilibrium carbon stocks based on known forest growth rates, and default turnover and decomposition rates tend to be far lower (~20 t C/ha).

In the Carbon Study, soil carbon stocks were increased by including stem mortality in the model and reducing the loss of CO2 during decomposition (assumed to 97.5% in the default settings for FullCAM but reduced to 80% in this study). This approach raised soil carbon stocks to levels close to those indicated on maps provided by CSIRO and DPIPWE (this data itself subject to uncertainty). However, further work is required to better calibrate the model to field data for native forests.

Page 259 of 365

7 BACKGROUND TO MODELLING APPROACH

This section of the report provides extended coverage of the background to selection of the Study Scenarios and the Baseline Sensitivity Scenarios, the classification system applied to forest types within Tasmania, development of the carbon modelling framework, selection of key values applied within the modelling framework and the basis for key assumptions. A review of previous studies of carbon stocks in Tasmania’s forest is also provided.

7.1 OVERVIEW OF FOREST CARBON ACCOUNTING

Role of forests in Australia’s carbon balance

Australia’s forests are an important carbon store and are estimated to contain over 12,000 Mt C (Montreal Process Implementation Group for Australia, 2008). This carbon stock represents over 70 times current annual emissions from fossil fuel use and land-use change (599.8 Mt CO2e in 2009, DCCEE 2011a).

In 2009, net sequestration by forested land in Australia was estimated to be 55 Mt CO2e (excluding carbon stored in harvested wood products), or 9% of net total emissions (DCCEE 2011b). Between 1990 and 2009, GHG emissions associated with conversion of forests to cropland, or grassland decreased by 87 Mt CO2e (66%) while carbon sequestration by lands converted to forest increased by 15 Mt CO2e (a 75% increase, DCCEE 2011b).

Carbon sequestration by forests varies greatly over time and in some years Australia’s forests have been a net source for emissions. For example, extensive wildfires in the Alpine regions of Victoria and NSW in 2003 associated with one of the worst droughts in Australia’s history resulted in net emission from all forest land exceeding 150 Mt CO2e. Fires and droughts are an integral component of the Australian landscape and forest ecology. Understanding their impact on forest carbon sequestration potential and the interaction between climate change and fire frequency is crucial to be able to model future emission scenarios for Australia.

Influence of forest management on carbon

Understanding and quantifying the various carbon inputs and emissions associated with forests is crucial in order to know how to maximise their carbon sequestration potential. Forests can sequester large amounts of CO2 which is stored in live biomass as well as debris (dead biomass including roots, bark, leaves, branches and stems) and soil. When harvested, burnt or degraded, forests release a proportion of the sequestered carbon. However, disturbances such as fire, or harvesting, can also add carbon to other pools including charcoal and wood products

Page 260 of 365

(which store carbon for varying time periods depending on their useful lifespan, recycling and final disposal).

Forest management can influence carbon stocks in a number of ways: • planting cleared land increases the amount of carbon stored in vegetation; • harvesting and removing trees from a forest reduces total carbon stocks; • improving the productive capacity of a site by applying fertiliser or other inputs can increase the rate of carbon sequestration of a site; • minimising the risk of damage to the forest from pests or fire can increase carbon sequestration potential of a forest; and • management of the forest to maximise storage of carbon in debris and soil can increase the total amount of carbon stored at the site over time.

7.2 METHODOLOGY USED FOR THE NATIONAL GHG ACCOUNTS

In Australia, GHG emissions and sequestration by forests is reported by the DCEEE in the National GHG Accounts (DCCEE 2011b). Since this is the current approach used to estimate carbon stocks, emissions from Australia’s forests for Australia’s National Carbon Accounts, it is useful to understand the methodology and assumptions used and identify aspects that apply to estimating carbon stocks and emissions for Tasmanian Forests.

Net emissions from forests are modelled in the National Carbon Accounting System (NCAS) using the carbon modelling tool FullCAM (Full Carbon Accounting Model) and a series of spatial and non spatial datasets that contain information on climate and soil, as well as forest cover, productivity, species, management and products. This evaluates changes in carbon stocks over time associated with tree growth, land use, management and fire.

Currently, full spatial accounting for Australia’s forests in the NCAS has been implemented only for post 1990 forests (known as Kyoto forests). These are covered under Article 3.3 of the Kyoto Protocol, under which it is mandatory to account for emissions associated with forests established on land cleared prior to 1990 and all previously forested land cleared after 1990. Management of existing forests, or forests established prior to 1990, is covered under the optional Article 3.4, to which Australia is not a signatory. Emissions from these forests are accounted for at a national level and reported in the National GHG Accounts, but are not yet accounted for spatially. Work is now underway to include pre-1990 plantations in the spatial NCAS modelling system (DCCEE 2011c).

For the National GHG Accounts, forest land is defined as all lands with a tree height of at least 2 m, a crown canopy cover of 20% or more and a minimum area of 0.2 ha (DCCEE 2011b). This definition is consistent with that used for Australia’s National Forest Inventory. For the purpose of modelling net carbon emissions, Australia’s forests are divided into a number of categories:

Page 261 of 365

• Harvested native forests: forests of endemic species arising from natural regrowth for which harvest records are available or which may be harvested in the future. • Other native forests: other forests with endemic species that have remained as forests • Pre-1990 plantations: established prior to 1990 which have remained as forest (although they may have been harvested and replanted one or more times). • Post-1990 plantations: established post 1990 on land that was grassland in 1990. • Forest land converted to cropland: areas that were forested in 1990 and subsequently converted to cropland. • Forest land converted to grassland: areas that were forested in 1990 and subsequently converted to grassland.

The approach used to estimate carbon and emissions from forests varies for each category. Specific methodologies used for the various forest types are outlined below.

Harvested native forests

Growth rates of harvested native forests are based on generic aged-based growth data for broad forest types from work undertaken by Lucas et al. (1997) under the National Forest Inventor. These forests are modelled at a national (i.e. non-spatially explicit) scale with FullCAM using aggregate estimates for age class-distributions; rates and methods of harvest; rates of debris accumulation and decomposition; and emissions associated with slash burning. Biomass is allocated to different tree components according to estimated amounts for each forest class. These figures are converted to total carbon based on work by Gifford (2000). Rates of litter fall (turnover) and decomposition of debris are based on default parameter sets. It is assumed there is no change in soil carbon over time, regardless of forest age or management.

Native forests converted to plantations since 1990 are modelled as harvested native forests, because these cannot currently be differentiated from native forests using Landsat Imagery. Native forests converted to plantations prior to 1990 are modelled as pre-1990 plantations based on data provided by the National Plantation Inventory.

Other native forests

Changes in carbon stocks in Other Native Forests are based on observed changes in forest extent (based on Landsat data) and changes in leaf area index estimated from satellite data. Changes in carbon stocks resulting from changes in forest extent that are not related due to land-use change are estimated using assumed rates based on the type of change: • Ephemeral, intermittent changes in forest cover: where forest cover (as assessed from Landsat imagery) fluctuates over time, foliage mass is assumed to change by 0.15 t/ha pa and carbon stocks by 0.057 t C/ha pa for both losses and gains.

Page 262 of 365

• Changes in Leaf Area of Continuous Forest Cover: where there is a change in foliage mass (as estimated from NDVI) the change in the amount of carbon is based on the change in leaf area index assuming a difference of 0.2 t/ha between the maximum and minimum values for average LAI across Australia (equivalent to a change in foliage mass of 0.24 t/ha for every unit change LAI). • Permanent loss, or gain, of forest cover: where there is a long term change in forest extent not associated with land use change, carbon stocks are increased or decreased at a rate of 0.5 t C/ha pa for a maximum of 20 years. This assumes most changes in forest cover due to natural causes occur on drier, less productive regions. • Woody thickening: some studies indicate carbon stocks may be increasing due to woody thickening (e.g. Burrows et al., 2002), while those of others may be decreasing due to dieback and degradation (e.g. Fensham, 2008). It is assumed net area of forest subject to woody thickening is 9.55 million ha and annual increase in carbon stocks is 1.22 t CO2e/ha. Each year, 12.2 Mt CO2e is added to the total for Other Native Forests.

Pre-1990 plantations

Plantations established prior to 1990 are modelled in FullCAM using growth increment tables, species specific conversion factors, regionally specific silvicultural regimes and average rates of change in soil carbon. Carbon pools include above and below-ground biomass, debris and soil carbon. Currently, estimates are non-spatial (model inputs and outputs are at a regional level) for different species and management regimes.

Growth rates and harvest regimes are based on yield curves developed by Turner & James (2002) for different regions, species and management regimes. Stem volumes are converted to masses using published wood density figures (Ilic et al. 1997). Masses of other tree components are calculated from stem mass based on species, age and management- dependent partitioning rates estimated using data from Keith et al. (2000), Eamus et al. (2000) and Snowdon et al. (2000). These masses are converted to carbon content using average figures for different tree components from Gifford (2000). Rate of debris turnover and decomposition are estimated using averages for different plant components.

Soil carbon is estimated using models developed by Polglase et al. (2000) showing the annual change in soil carbon for plantations from establishment to age 100 years. It was assumed that the initial average soil carbon under pre-1990 plantations was 37.81 t C/ha.

Post 1990 plantations

Plantations established in 1990 or later are modelled spatially in FullCAM. These plantings are identified using Landsat imagery which allows location, plantation type and time of establishment to be determined with a high degree of accuracy.

Page 263 of 365

Modelling carbon in vegetation

Vegetation growth rates are estimated using a Forest Productivity Index (P), based on a simplified version of 3PG as described by Kesteven et al. (2004) combined with region and species specific growth modifiers. The Forest Productivity Index is based on the amount of photo-synthetically Active Radiation (APAR) for a given location, modified by climatic, soil and topographic factors to account for various growth constraints at the site. Values for P have been calculated for all of Australia at a 1 km grid scale on an annual basis since 1970. This provides reasonably accurate growth rates for different regions that reflect actual spatial and temporal variability in growth arising from variation in site productivity, climate and management.

Average maximum biomass production M of native forest at a given site can be estimated from P in the following relationship:

0.5 2 M = (6.011 Pav – 5.291) Equation 28 where Pav is the long term average for P for a given site.

The formula used to estimate above-ground biomass production for a particular site and species and management regime is:

(-2G/d) (-2G/ (d-1)) Ia = M r (ya e – ya-1 e ) P/Pav Equation 29 where:

Ia is the annual biomass increment at age a, M is the maximum potential above ground biomass r is a non-endemic species modifier

ya is an adjustment to account for effects of management inputs on long term site productivity at age a, G is the age of maximum growth, and d is the effective age of the Stand after accounting for short term responses to management inputs.

The derivation of the formula and associated parameters used to adjust growth is explained in the National Carbon Accounts (DCCEE 2011b). No growth modifiers are used to adjust M for the growth of environmental plantings which are assumed to grow at the same rate as native vegetation.

Page 264 of 365

Modelling carbon in debris

The growth formula (Equation 29) is used by FullCAM to estimate Stand growth rates and subsequent production of debris and soil carbon of post-1990 plantations. Allocation of carbon to different vegetation components is varied by species, management and age based on work by Keith et al. (2000) and Polglase et al. (2004). Carbon contents in different plant components for hardwoods and softwoods are based on values from Gifford et al. (2000). Rates of debris production and decomposition are estimated from default rates for different plant components based on work by Paul et al. (2004) and Mackensen and Bauhus (1999, Table 51). These rates are multiplied by the total amount of carbon present in each pool at the end of each period.

Table 51: Turnover rates used to estimate debris production from standing biomass in post-1990 plantations Tree Turnover rates Decomposition component Hardwood Softwood rates %/y %/y %/y Stems 0 0 10 Branches 3 5 10 Bark 5 7 50 Leaves 30 50 100 Coarse roots 7 10 40 Fine roots 100 80 85 Few estimates are available of the rates of turnover of different components. The focus of the modelling was therefore to obtain “rates that coincided with measures of litter mass when the models were run over extended periods so that a reasonable balance between turnover (input to litter pools) and decomposition of litter pools was achieved” (DCCEE, 2011b, p. 140).

Modelling soil carbon

To model soil carbon under post-1990 plantations, the Roth-C component of FullCAM is used (Jenkinson et al. 1987, 1999). Calibration of this model referenced long term field trial data by Skjemstad & Spouncer (2002) plus an additional 75 soil pairs from areas of major forest conversion activity. Paul et al. (2002 and 2003 and 2004) further calibrated and validated the Roth-C model for use in afforestation and reforestation situations (Table 52). Paul et al. (2004) showed that Roth-C was able to accurately predict soil carbon in afforestation and reforestation situations using the default parameters from Jenkinson et al. (1999). The main change was proportion of carbon lost as CO2 from decomposing plant debris in which optimum values for losses from Decomposable and Resistant Plant debris were 0.77 and 0.39 respectively.

Page 265 of 365

Table 52: Default parameters used in the Roth C model in FullCAM to estimate changes in soil carbon. From Paul et al. (2002). Soil Carbon Inputs from debris pools Decomposition Proportion of decomposed soil pool Pool rate moving to:

Decomp. Resistent BIO-F BIO-S HUM % % %/y % % % DPM 100 0 1000 46 0 54 RPM 0 100 30 46 0 54 BIO-F 0 0 66 46 0 54 BIO-S 0 0 66 46 0 54 HUM 0 0 2 0 46 54 These parameters differ from the defaults in the FullCAM model which assume that a portion (10%) of decomposable and resistant debris moves directly to soil humus (by-passing the DPM and RPM pools). They also assume that the 87.5% of the carbon in decomposing debris is lost as CO2 (as opposed to the fractions of 39% for resistant debris and 77% for decomposable plant debris used by Paul et al. 2004).

Initial values for soil carbon at the time of plantation establishment are derived from a map of pre-clearing soil carbon across Australia (Webb 2001). These estimates are modified to account for effects of clearing prior to plantation establishment using FullCAM and it is assumed that all areas were cleared 20 years prior to the establishment of post-1990 plantations. These new values for soil carbon are used to initialize the model for post-1990 plantations (DCCEE 2011b).

Forest converted to grassland or crops

Emissions from forests converted to grassland, or cropland, are estimated spatially using FullCAM to model changes in vegetation, debris and soil carbon pools. As with post-1990 plantations, areas of conversion are identified using Landsat Imagery. However, unlike post- 1990 plantations, all conversions that can be identified back to 1972 are modelled.

Where there is no history of prior disturbance, the original vegetation pre-clearing is assumed to be at its maximum biomass (M) as estimated from Pav (Equation 28). Where there is evidence of prior disturbance, the biomass at time of clearing is estimated from the tree yield formula:

Y = M . e (-2G/a) Equation 30 where G is the age of maximum growth (assumed to be 14 years for all native vegetation) and a is the estimated current age of the vegetation.

Carbon sequestration and emissions from vegetation (i.e. crops or grasses), debris and soil following clearing are modelled in FullCAM based on estimated production for different regions of Australia, crop specific expansion factors, rates of harvest and management inputs.

Page 266 of 365

Emissions from fire

Emissions from fire in native forests are estimated at a national level for prescribed burning and wildfires and at a site level for post-harvest regeneration burning. Emissions from prescribed burning and wildfire are calculated from total area burnt each year, and assumed fuel load, burning efficiency, mass of carbon (50% dry weight) and nitrogen (0.55% dry weight) in the material burnt and emission factors for each greenhouse gas produced (DCCEE 2011b, Table 53).

Table 53: Assumed fuel loads and amounts of fuel burnt (t/ha dry matter) in prescribed burning and wildfires in native forests. Source: DCCEE (2011b, Tables 7.12, 7.13 and 7.14). Fire Type State ACT NSW NT Qld SA Tas Vic WA Prescribed Burning Fuel Load 17.6 18.2 4.1 9.7 9.6 20 17.9 12 Material Burnt 7.4 7.6 1.7 4.1 4.0 8.4 7.5 5.0 Wildfire Fuel Load 35.2 36.4 7.2 19.4 19.2 40 35.8 33.4 Material Burnt 25.3 26.2 5.2 14.0 13.8 28.8 25.8 24.0

Note: assumed burning efficiency is 42% for prescribed burns and 72% for wildfire

Greenhouse gasses covered include CO2, methane (CH4), nitrous oxide (N2O), nitrogen dioxide (NO2) and non-methane volatile organic compounds (NMVOCs). These gasses are reported both as actual masses and in terms of CO2 equivalents (CO2e) where available, based on estimated 100-year global warming potentials reported by IPCC (1 for CO2, 21 for CH4 and 310 for N2O, 1996). CO2 emissions from forest fires are reported under Other Native Forests while non-CO2 emissions are reported under biomass burning.

Emissions from harvest regeneration burns in native forests and plantations are modelled in FullCAM. This estimates fuel loads and inputs forest type and harvest-regime-specific factors for burning efficiency. For native forests, the modelling is at a national level while, for plantations, it is at a Stand (0.25 ha) level. Greenhouse gas emissions are estimated using the same emission factors and carbon and nitrogen contents as for prescribed burning and wildfire.

Wood products

Carbon stocks and emissions associated with harvested wood products are reported in Australia’s National GHG Accounts under land use, land use change and forestry where they arise during the service life of products and under waste where they arise from landfill. Wood removals from forests and plantations are counted as emissions from the harvested native

Page 267 of 365

forest, pre-1990 plantation and post-1990 plantation categories, but those wood products used domestically are counted as additions to the category other - harvested wood products.

Only emissions and inputs from wood products that are processed and used in Australia are included in the National Accounts. Thus, exported wood chips and logs are excluded (i.e. the carbon in harvested wood is assumed to be emitted at the point of harvest) while imported wood products are included. Data for sawlog and pulpwood harvest and the manufacture, import and export of wood products are obtained from the Forest and Wood Product Statistics collected by ABARES (e.g. ABARES 2012a). Although FullCAM has the capacity to estimate emissions from wood products, this aspect of the model has not been implemented for the purpose of Australia’s National GHG Accounts. Development of the model used to estimate these emissions is described by Jaakko Pöyry Consulting (1999, 2000) and Richards et al. (2007).

Because log harvest and production of sawn timber, veneer and panels are reported on a volume basis, it is necessary to assume a basic wood density to estimate total biomass flows. The assumed basic densities and carbon contents of different wood products are based on values from Ilic et al. (1997).

Recovery rates of sawn timber from sawlogs are based on industry estimates. The recovery rates of pulp and paper from pulp logs and woodchips is assumed to be 70% (based on an average for Kraft pulp production with a recovery of 50% and thermo-mechanical pulp production which has a recovery of around 95%).

An important component in accounting for carbon stocks and emissions in wood products is the assumed life span and end use of each product, as well as the initial stock. Products are divided into six pools, ranging from very short-term (e.g. paper and paper products), to very long term (e.g. framing and flooring timber). The total estimated amount of material in each pool is divided into three age classes (young, medium and old) and the current age, rate of loss and maximum possible loss is estimated for each.

Total stock of carbon in wood products in service is estimated to have increased from 77 Mt C (283 Mt CO2e) in 1990 to 105 Mt C (384 Mt CO2e) in 2009, DCCEE, 2011b). Thus, the estimated annual rate of increase in carbon stocks in wood products has averaged 4-5 Mt C per annum (16-17 Mt CO2e) over the past decade. In comparison, total emissions from forest land have ranged from 150 Mt CO2e per annum to -110 Mt CO2e per annum, while emissions from forest converted to grassland have averaged 50-150 Mt CO2e pa. Thus, the impact of harvested wood products on carbon emissions from LULUCF is fairly minor.

Application of DCCEE methodologies to the Carbon Study

The methodologies and models used to estimate emissions from forests developed by DCCEE (Richards et al. 2001, 2005, 2007, Waterworth et al. 2007, Waterworth and Richards, 2008), have a number of aspects that were adapted for the current Carbon Study. These include:

Page 268 of 365

• a general model (FullCAM) for estimating and tracking carbon stocks in vegetation, debris, soil and wood product pools; • forest extent surfaces showing the change in cover of broad forest types (hardwood plantation, softwood plantation and native forest back to 1990); • climate, soil and forest productivity surfaces that can be used for running the model; • tabulated parameters for modelling carbon allocation, turnover and decomposition in for different forest types and stocks and emissions from wood products; • estimates of current stocks in soil and debris for different locations and forests; and • parameters and management regimes for predicting growth and carbon sequestration in different plantation types across Tasmania.

However, the DCCEE approach has a number of limitations with respect to the objectives and scope of the Carbon Study. Growth and carbon sequestration of native forests are based on: • general growth estimates for different growth stages of only a few different forest types based on national averages, or • a general Forest Productivity Index that tends to underestimate growth of intensively managed plantations and highly productive native forests.

The underlying growth model used in FullCAM to estimate growth based on the Forest Productivity Index utilises only two primary parameters and is not well suited to estimating growth of native forests with complex structures and dynamics. Additionally, some of the key processes and assumptions used in the FullCAM model may be inappropriate for modelling carbon emissions and sequestration in native forests, including its inability to readily model: • movement of stemwood to debris except as a result of harvest or fire (i.e. no self thinning or mortality); • losses of soil carbon from wildfire or regeneration burns; • natural regeneration in partially burnt, or harvested, Stands; or • growth of multi-aged stands.

Further, FullCAM lacks a suitable interface for the efficient input of large sets of growth or other data and, although the FullCAM model is designed to run spatially, this capability is not possible outside of the NCAS. These limitations mean that accurate modelling of carbon stocks using actual spatial data for complex native forests is very difficult, if not impossible in the current version of the model.

We therefore developed a new modelling framework (the Forest Carbon Modelling Framework) based on the general functions and processes within FullCAM, but with added flexibility to allow the growth and dynamics of Tasmania’s forests to be more accurately simulated.

Page 269 of 365

7.3 CARBON ESTIMATES FOR AUSTRALIAN FORESTS

National forest inventory state of the forests report

The National Forest Inventory estimates the amount of carbon emitted through deforestation and sequestered by existing and new forests and plantations. These estimates are mainly based on data provided by the DCCEE, disaggregated at a finer level than in the National GHG Accounts. The most recent estimates are in the 2008 State of Forests Report (Montreal Process Implementation Group for Australia, 2008).

A key difference between the GHG accounts of NCAS and estimates from the NFI is the estimated total area of native forest in Australia. According to DCCEE (formerly the AGO), woody biomass in plantations, commercial native forests and conservation forests covered a total of 108 million ha in 2005, while NFI estimated that the total area of forests was 149 million ha (Montreal Process Implementation Group for Australia 2008). This difference arises from the different methods to determine forest cover. Estimates of forest extent used in the NFI are based on aerial photography provided by state agencies while the NCAS uses Landsat imagery. Most of the variation is a result of differences in the extent of drier, open forests and woodlands across northern Australia. In contrast, there is a 90% correlation between coverages of dense forests such as Rainforest and eucalypt tall open forest and eucalypt tall woodland. Across Tasmania, the area of forest estimated for the NCAS is larger than that for the NFI, primarily because large areas of woody vegetation across the southwest and central west of the state are classified as forest for NCAS and non-forest for the NFI (e.g. Montreal Process Implementation Group for Australia, 2008, p. 14).

Native forests

In 2005, total stocks of carbon in native vegetation classed as forest (i.e. > 2 m tall and > 40% crown cover) was estimated to be 44 Gt CO2e. This is equivalent to around 40 years of Australia’s total GHG emissions (599.8 Gt CO2e, DCCEE 2011a). Of the total carbon stock, 54% was contained in vegetation biomass (including roots) with the remainder in soil (including debris). Vegetation types with the largest total stores of carbon included eucalypt woodlands (12 Mt CO2e) and eucalypt medium open forests (13 Mt CO2e). Altogether, woodlands contained around 50% of total carbon in native vegetation.

Total gains and losses of carbon from different forest types have been estimated based on changes in area and canopy cover assessed using remotely sensed imagery. Between 1989 and 2004 there was a reduction in carbon in biomass of around 166 Mt CO2e or 0.7%. This is equivalent to an annual rate of around 11 Mt CO2e per annum. This reduction excludes changes in soil carbon, so total losses are likely to be even greater. The largest reductions were in medium open forests (101 Mt CO2e) and eucalypt woodlands (65 Mt CO2e). Carbon contents in some forests increased over the period with that of Acacia shrublands rising by 34 Mt CO2e.

Page 270 of 365

The main causes of declines in carbon stocks in vegetation include broad-scale clearing for agriculture, firewood collection, and mortality due to a variety of causes including dieback and bushfire. Causes of increases in carbon stocks of some vegetation classes include increases in extent, and increase in tree density, probably due to reduced fire frequency.

According to the State of the Forests report, carbon stocks in harvested native forest increased by 43.5 Mt CO2e in 2005. This increase was in spite of the fact that total wood removals over the period 2001-04 averaged 14.1 Mt CO2e per annum.

Plantations

Total carbon stocks in plantations rose from 264 Mt CO2e in 2000 to 312 Mt CO2e in 2004 (Montreal Process Implementation Group for Australia, 2008). This increase was due to the large increase in carbon stocks in hardwood plantations resulting from new plantings as well as the ongoing growth of young Stands. In contrast, carbon stocks in softwood plantations remained constant at around 200 Mt CO2e. Total wood removals from plantations over the period were equivalent to 22 Mt CO2e.

Wood products in use & in landfill

Other major pools for carbon reported in the State of the Forests Report include wood products and landfill. Figures for carbon in wood products are obtained directly from the National GHG Accounts reported by the DCCEE. Total carbon stored in wood products in service in 2005 was estimated to be 97 Mt C (310 Mt CO2e) and increased at a rate equivalent to around 14 Mt CO2e pa between 2001 and 2005. The bulk was stored in very long term products.

Total stock of wood products in landfill in 2004 was estimated to be equivalent to 525 Mt CO2e, being 50% greater than the amount of carbon in wood products in service. Between 2000 and 2005 the total stock of carbon in wood products in landfill increased at a rate of around 12 Mt CO2e pas. This was largely due to an increase in the stock of the largest pool - very short-term products (i.e. paper) - which increased at a rate of 9 Mt CO2e pa. There was little change in most other pools with structural timber remaining around 60 Mt CO2e for the period.

The State of the Forests Report commented that future emissions from wood products in landfill could exceed additions. A key issue is the potential production of methane (which is 20 times more potent a greenhouse gas than CO2) from organic material in landfills. On the other hand, work by Ximenes et al. (2008b) indicates that the lifespan of wood products in landfill may be far greater than currently assumed under IPCC default guidelines (2003) with only 17-18% of the original carbon contained in either softwood or hardwood lost over 46 years.

Page 271 of 365

Fire & biomass burning

Few data are available on emissions from wildfires and prescribed burns in Australia. The State of the Forests Report cited an estimate from the National Carbon Accounts indicating that planned and unplanned fires in harvested native forests released 1.3 Mt CO2e in 2005 (AGO 2007). The report also estimated that the alpine fires in NSW and Victoria in 2002-03 released the equivalent of 40 Mt CO2e. In addition, 9-11 Mt CO2e were estimated to have been emitted nationally from the burning of firewood (compared with 43 Mt CO2e sequestered by native forests managed for wood production). Although firewood production and use reduces emissions from fossil fuels, 72% of firewood was sourced from low rainfall, low biomass forests which could contribute to depletion of forest carbon stocks in some areas if the rate of harvest exceeds the rate of growth.

Emissions from fossil fuel use

The State of the Forests Report also provided estimates of emissions associated with fossil fuel use in forest management, harvest and log transport. In the National Carbon Accounts, these emissions are aggregated with those of other sectors (e.g. transport, fertiliser use in agriculture).

Based on the information provided, estimates of total carbon emissions per m3 of log were:

• ~12 kg CO2e for production and harvest of native forest in Tasmania;

• 0.44 kg CO2e for seedling production and establishment of softwood plantations;

• ~10 kg CO2e for management and harvest of a softwood sawlog plantation; and

• ~15 kg CO2e for log transport (assuming an average distance of 500 km).

Lifecycle of wood products

When accounting for the carbon stored in wood products, it is arguable that emissions associated with wood processing be included (Mackey et al. 2008). However it is also arguable that the emissions associated with manufacture of alternatives such as steel or concrete also be included in this comparison. In reality, the full benefits (or costs) of a particular material or process can only be judged by comparing full life cycle carbon emissions associated with its manufacture, use and disposal with those of potential substitutes.

The State of the Forests Report includes a comparison of life cycle emissions of wood products with those of alternatives. As a result of the relatively low carbon intensity of wood processing, combined with the fact that wood contains sequestered carbon, wood-used in house construction can be a net carbon store. The State of the Forests Report cites work by 3 Buchanan and Honey (1994) indicating that treated timber stores a net 228 kg CO2e/m while 3 glue-laminated timber stores 168 kg CO2e/m . In comparison, structural steel emits 8,117 kg 3 3 CO2e/m and reinforced concrete emits 182 kg/m . Work by Turner (1990) indicates that the net carbon stored in a timber framed house is 7.5 kg/m3 compared with net emissions of 2.9 kg/m3

Page 272 of 365

from steel framed housing. Thus, the report argues that increasing the amount of timber-framed houses compared with steel framed houses will reduce total carbon emissions. Total carbon sequestration associated with forestry The State of the Forests Report estimates that the total amount of carbon sequestered in native forests, plantations and wood products in 2005 was 56.5 Mt CO2e. In comparison, deforestation through the conversion of forests to agricultural land emitted 53.3 Mt CO2e (95% of Australia’s total emissions for the year). As a result, a net 3.2 Mt CO2e was sequestered. The sector 'Land use, land use change and forestry' was the only sector in the Australian economy to store more GHGs than it emitted in spite of the broad scale clearing of some forest types.

Evaluation

The report provides useful information on fossil fuel emissions associated with wood production as well as comparisons between emissions associated with wood products and other materials. It also highlights a key area of uncertainty relating to Australia’s forest carbon stocks which is that actual area of vegetation > 2 m tall and > 20% cover.

Green carbon assessment

Mackey et al. (2008) assessed the potential carbon storage in native forests across south-east Australia in the absence of human disturbance. This assessment included an estimation of the Carbon Carrying Capacity (the average amount of carbon stored in a forest not subject to human disturbance) as well as potential increase in carbon stocks in the absence of timber harvesting.

The results of the study, which are summarised in Mackey et al. (2008) and reported in more detail by Keith et al. (2009), indicate the average Carbon Carrying Capacity of eucalypt forests across SE Australia was 640 t C/ha in living and dead biomass and soil with 360 t C/ha contained in biomass plus debris. It was estimated that allowing previously harvested forests to reach their full Carbon Carrying Capacity could result in the sequestration of a total of 7.5 Gt CO2e, argued to be equivalent to avoiding 136 Mt CO2 emissions each year for 100 years.

Potential Carbon Carrying Capacity of native forests was estimated by measuring total carbon in living and dead biomass and soil across a series of plots located in Stands with little or no evidence of previous human disturbance. A relationship was developed between total forest carbon and Net Primary Productivity (NPP) soil, climate and topography. NPP was estimated across the landscape using an estimated average gross primary productivity referencing satellite data and a relationship between NPP and GPP and environmental conditions based on published estimates from 27 sites across the world (Keith et al. 2010).

Relationships were then developed between measured total forest carbon and modelled NPP and environmental variables measured at each site, based on a sample of 240 plots across

Page 273 of 365

eucalypt forests in south-eastern Australia. The relationship for predicting total biomass carbon was based on the following model:

Ln (Biomass) = a + b1NPP + b2W + b3P + b4SdAB + b5T Equation 31 where: Biomass is the total living biomass in t C ha-1, NPP (net primary productivity) is estimated from gross primary productivity and the relationship between NPP, GPP and environmental variables, W is the water availability index calculated from annual precipitation, solar radiation, and the latent heat of vaporization of water – see Keith et al. 2010), P is annual precipitation,

SdAB is the total depth of soil A and B horizons, T is the mean annual surface temperature, and

a, b1, b2, b3, b4 and b5 are constants that vary with vegetation, geology and topography. Total soil carbon was estimated from spatial data layers provided by the Australian Soil Resource Information System (McKenzie et al. 2005). From these, a model of variation in soil carbon across the landscape was developed.

These models accounted for 51.4% of variation in living biomass and 47.7% of variation in total biomass across the plots. The models were used to estimate of Carbon Carrying Capacity across eucalypt forests in south eastern Australia, for forest with a canopy cover of 30-70% and tree height > 10 m. Rainforest and riverine forests within agricultural zones were excluded.

The weighted average amount of carbon stored in eucalypt forest across the study area (assumed to represent all eucalypt forests in south-eastern Australia) was:

• live vegetation: 289 +/- 161 t C/ha; • debris: ~71 t C/ha; • soil: 280 +/- 161 t C/ha; and • total: 640 +/- 383 t C/ha.

Based on the total area of forest included in the study (14.5 million ha), it was estimated that the potential Carbon Carrying Capacity was around 9 Gt C (33 Gt CO2e)

Mackey et al. (2008) estimated reductions in carbon associated with timber harvesting and potential sequestration if harvesting was discontinued although these calculations were not included in the published articles by Keith et al. (2009, 2010). Based on work by Roxburgh et al. (2006), timber harvesting was assumed to deplete total carbon stocks (including those in live vegetation, debris and soil) by 40% compared with undisturbed forests (Mackey et al., 2008).

Page 274 of 365

Thus, the average total carbon stock in previously harvested forest was assumed to be around 383 t C/ha. Mackey et al. (2008) also estimated that around 56% of the 14.5 million ha of eucalypt forests in the study area had been harvested previously.

Based on these assumptions, the total current stock of carbon in harvested forest was estimated to be around 3 Mt C compared with a potential 5 Gt C, if they were at their assumed Carbon Carrying Capacity. Thus, it was concluded that cessation of harvesting in eucalypt forests across Australia could increase carbon sequestration by up to 2 Gt C (7.5 Gt CO2e). This is 12 times greater than Australia’s total emissions for 2008 (DCCEE 2011a).

An analysis of differences between the estimates from the study compared with IPCC estimates for natural forests and potential sources of error was reported by Keith et al. (2010).

Evaluation

The Green Carbon study provides useful reference to a methodology by which carbon stocks can be estimated over a large area for forests not subject to either human disturbance or recent (i.e. <100 year) Stand replacing fires.

Some assumptions and conclusions have been debated in the literature (e.g. Moroni et al. 2010; Reid 2010), including interpretation of the Carbon Carrying Capacity concept (e.g. Moroni et al. 2010, Dean 2011; Moroni et al. 2012). This is significant because estimation of Carbon Carrying Capacity is a key objective of the Carbon Study.

Under the definition used by the authors, Carbon Carrying Capacity is “the mass of carbon able to be stored in a forest ecosystem under prevailing environmental conditions and natural disturbance regimes, but excluding anthropogenic disturbance” (Keith et al. 2009, p. 11635, from Gupta and Rao 1994). This contrasts with the interpretation of Moroni (2010) and Reid (2010) of the reported Carbon Carrying Capacity representing “the maximum carbon stock that can be held at the point of maximum forest development”. However, it is uncertain whether the measurement plots used to estimate Carbon Carrying Capacity (under the first definition) provide a truly representative sample of eucalypt forest subject to natural disturbance regimes.

Stand replacing fires, in particular, are important components in the ecology of many eucalypt forest systems (e.g. Jackson et al., 1968). Thus, a representative sample of a forest type would include all age classes normally found in its natural growth cycle, with appropriate weighting for each class according to its frequency across space and time. A forest, not subject to human disturbance could be expected to consist of a wide range of growth stages. However, if a sampling strategy is biased towards older forests with higher carbon stocks, then average carbon stocks under natural conditions will be overestimated.

This study correctly points out that, by basing estimates of Carbon Carrying Capacity on Stands subject to human disturbance, we risk underestimating the potential contribution of forests and management decisions on global carbon cycles. It follows that any assessment of life cycle

Page 275 of 365

emissions associated with wood products should consider emissions associated with conversion of forest subject to natural disturbance regimes only, to those from a forest that is periodically harvested.

Carbon stores in Victorian forests

Grierson et al. (1992) conducted one of the first carbon assessments in Australia, covering both native forests and plantations across Victoria. The assessment was based on inventory records for different forests, with stratification by forest type, age and region. Expansion factors for different species were used to convert stem volume to above-ground biomass. Relationships between total above-ground biomass across all productivity classes and Stand age were then developed for each species. These relationships were then used to estimate total above ground biomass and carbon (assuming 50% carbon content).

The total amount of carbon stored in Victoria’s forests was estimated to be 1.0 Gt C (3.4 Gt CO2e, Table 54). Average carbon density for native forests was estimated to be 157 t/ha (1035 t CO2e/ha). Highest carbon densities were in alpine forests (E. delegatensis, 250 t C/ha) and mountain ash forests (E. regnans, 246 t C/ha), while the lowest densities were in mallee (18 t C/ha) and Callitrus woodland (20 t C/ha). Carbon density in softwood plantations was 90 t C/ha.

Table 54: Estimated areas, carbon densities and total carbon stocks for Victorian native forests and plantations. Source: Grierson et al. (1992). Mean carbon Forest type Area Total Carbon density -1 ('000 ha) (t ha ) (Mt C) (Mt CO2e) Alpine ash 312 197 61.6 205 Mountain ash 182 246 44.8 149 Shining gum 13 235 3.1 10.2 Mountain mixed species 465 212 98.5 328 Foothills mixed species 2,640 239 629.8 2,099 Alpine mixed species 203 250 50.8 169 Coastal mixed species 404 190 76.6 255.2 Box-ironbark 236 75 17.6 58.8 River red gum 113 24 2.7 8.8 Closed rainforest 0.1 209 0.0 0.1 Callitris 5.3 20 0.1 0.3 Black box woodland 53 25 1.3 4.4 Scrub species 195 60 11.6 38.7 Mallee 1,696 18 30.5 102 P. radiata 61 91 5.5 18.5 Total 6,578 157 1035 3,448

Page 276 of 365

The study did not estimate rates of sequestration. However, based on growth data, the rate of carbon uptake was expected to be greatest for younger forests (5-10 years) and to decline with age. Average rates of carbon sequestration were estimated to range from 6-7 t C/ha/y for newly established forests, and to vary in older forests from a maximum of 8-10 t C/ha/y for mountain ash forests growing on good sites to less than 3 t C/ha/y for mallee.

Victorian land carbon assessment

Norris et al. (2010) estimated carbon stocks and emissions from Victoria’s publicly managed land (forests and grassland), covering the period 1930-2009 and including all publicly owned native forests across the state (~7.5 million ha). The authors used default growth rates, allocations and decomposition rates, climate and soil data from NCAS to model carbon stocks and emissions using FullCAM. Information on past fire and harvesting events, wood product conversion rates and changes in pools resulting from fire was obtained from literature. The study included carbon in above- and below-ground biomass (including both overstorey and understorey vegetation), debris, soil and emissions from wood products.

The various events (i.e. natural or man-made factors influencing carbon stocks) modelled and associated assumptions included: • timber harvesting was assumed to be of one intensity, based on an average between thinning and clear cutting; • prescribed burns were assumed to be low intensity; • fire intensity for post 2002 wildfires was based on actual data; • fire intensity for pre-2002 wildfires was assumed to be of a single intensity based on data from Cheney (1981); • all Stands were assumed to be burnt by wildfire in either 1855 or 1905 to generate an ecologically appropriate state of disturbance prior to the reporting period (1930-2010); • the timing of events was condensed into 10 year periods and all events for a particular decade were assumed to take place in the middle of that decade; and • allocations of biomass to tree components were based on those for low open forests.

These inputs were allocated spatially to a series of layers to develop scenarios for the whole forest estate and forests with the same scenarios were grouped by IBRA region (Interim Biogeographic Regionalisation Zones for Australia). Areas of public land <100 ha were removed, reducing the total area to be modelled by only 4%, but reducing the number of records by 97%. To obtain productivity, climatic, and soil data, a centroid was located for each IBRA region x scenario combination. Each IBRA region by scenario was run through FullCAM using the centroid coordinates and data output into Microsoft Access. To stabilize soil carbon stocks prior to running the model, each scenario was run for 400 years prior to the first major

Page 277 of 365

disturbance event (i.e. wildfire in either 1855 or 1905). The resultant outputs from FullCAM were used to produce a spatial dataset of total carbon stocks in vegetation, debris soil and products for public native forests across the whole of Victoria spanning the period 1930 to 2009.

The results indicate carbon stocks in Victoria’s forests are stable, at around 2,750 Mt CO2e. Based on a total area of 7.5 million ha, this indicates average carbon stock in vegetation, debris, soil and products was 370 t CO2e/ha. Stocks exceeded 920 t CO2e/ha in some highly productive forests in East Gippsland, while carbon stocks in less productive forests in the northern, western and central parts of the state were less than 350 t CO2e/ha.

The most influential factor on carbon stocks was wildfire, with 16 million t carbon (2% of total stock) emitted as a result of wildfires between 2000 and 2009. However the effect of fire was shown to be largely transient, especially for low intensity fires (i.e. non-Stand replacing). Harvesting had a greater impact on carbon stocks at a particular location than either wildfire or prescribed burning, but affected only a small proportion of the total estate in any year. Forested lands were shown to be resilient to the effects of either harvesting or fire with total carbon mass (including harvested wood products) returning to pre-disturbance levels within 100 years.

Key uncertainties that had a major effect on carbon stocks and flows in the study included: • age of forest at beginning of the simulation; • number of years to run model prior to commencement of the simulation to allow carbon in soil and dead biomass to reach an equilibrium (400 years); • Stand growth rate (based on the FPI which may underestimate growth of more productive native forest Stands); • allocations of carbon to different pools in living vegetation; • rates of turnover and decomposition of forest debris; • rates of turnover and decomposition of soil carbon; and • proportion of carbon moving from vegetation, debris and soil pools into the inert soil carbon pool (i.e. charcoal) as a result of fire.

In particular, over long timeframes, total carbon stocks can be very sensitive to the proportion of material moving into the inert carbon pool. In FullCAM, carbon in this pool is assumed to never decompose and thus accumulates over time.

The estimated total amount of carbon stored in Victoria’s public lands (750 Mt) was ~27% lower than the estimate by Grierson et al. (1992). Further, the estimate of Norris et al. included above- ground, below-ground and soil carbon and covered all vegetation on public land while that from Grierson et al. included above-ground carbon in overstorey trees only. Thus, the actual difference is probably much greater. The main reason for this difference is likely the lower assumed default productivity for FullCAM compared with actual growth used by Grierson et al. (1992).

Page 278 of 365

A review of the work by Norris et al. (2010) by CSIRO pointed to weaknesses in the approach (Polglase 2010). Chief among these was that climate and soil data for only a relatively small number of points were used to model growth. Furthermore, as they were based on centroids (the average centre of a number of polygons) the points did not necessarily coincide with the actual locations of the forest type or disturbance regime.

Following these suggestions, DSE changed the methodology and downloaded climate and soils data from the DCCEE server, at a much finer spatial resolution, for each individual forest polygon covering an area with similar disturbance history and forest type. This process created around 750,000 plot files and required the development of an automated procedure to download the data, run the plot files, export the results and convert them into spatial layers. Although this far more intensive process reduced the amount of uncertainty associated with the outputs, the estimate of total carbon stocks was altered by just 2% (Polglase, 2010).

Evaluation

The approach used by Norris et al. (2010) provides a useful basis for spatial estimation of carbon stocks in native forests. However, a limitation with respect to the objectives of the Carbon Study is the reliance on FullCAM default growth rates rather than using extensive growth date from forests. The default growth estimates of FullCAM tend to underestimate growth of more productive forests (Roxburgh et al. 2010). Since growth is the driver of sequestration, this has a major influence on carbon pools in live vegetation, debris, soil and products. In addition, the study did not attempt to model future forest growth or assess the effects of alternative management scenarios as required in the current Carbon Study. Without an efficient system to allow the rapid input of alternative growth rates, harvesting regimes or fire frequencies, this would have been very difficult given the approach used.

7.4 ASSESSMENT OF CARBON STOCKS IN TASMANIA’S FORESTS

There have been several assessments of carbon stocks in Tasmania’s State forests over the past 15 years. However, these have focussed on State Forests and no studies have assessed stocks in National Parks, or private forests. Estimates of stocks in above-ground vegetation in State Forests range from 366 Mt CO2e (MBAC, 2007) to 843 Mt CO2e (Lucas et al., 1997).

Three of the estimates (Table 55, studies 1, 2, and 4) are based on stem volume data from Forestry Tasmania’s sample plots and one estimate (3) was based on biomass growth from FullCAM. There was a wide range between estimates, even for those based on measured stem volumes. These differences were likely a result of differences in assumed wood density, expansion factors (used to covert stem mass into whole tree mass), as well as the areas covered by different forest classes.

Page 279 of 365

Table 55: Estimates of total carbon stocks in Tasmania’s State Forests.

No. Source Year Forest Type Carbon Stock (Mt CO2e) Vegetation Debris Soil Total Above- Below- Total ground ground 1 Lucas et al. (1997) 1995 Commercial 269 Non-Commercial 574 Total 843 2 MBAC (2007, FT data) 2007 Commercial 167 42 209 Non-Commercial 264 66 330 Total 431 108 539 3 MBAC (2007, FullCAM)1 2007 Commercial 158 18 176 139 114 429 Non-Commercial 211 23 235 205 169 609 Total 366 41 407 345 282 1034 4 Moroni et al. (2010) 2009 Commercial 296 74 370 Non-Commercial 182 45 227 Total 478 120 598 1Includes hardwood plantations

All studies except that by Lucas et al. (1997) estimated below-ground carbon stocks. Studies 2 and 4 assumed root biomass comprised 20% of the total biomass. In contrast, study 3 assumed that root biomass comprised10.8% of above ground biomass for native forests. As a result, the estimated carbon content of this latter study was less than half that of studies 2 and 4. The only estimate of carbon content of debris and soil is from the MBAC (2007), where it was estimated that debris contained 345 t/ha carbon and soil contained 282 t/ha carbon.

Only two studies have estimated rates of carbon sequestration by Tasmania’s native forests and results differ markedly (Table 56). Lucas et al. (1997) predicted native forests would sequester 8 Mt CO2e per annum (above-ground) between 1995 and 2025, while MBAC (2007) predicted they would sequester 2 Mt CO2e per annum (total) between 2007 and 2050. The difference is probably due to different assumed growth rates of old-growth forests (assumed to continue growing through the modelled timeframe by Lucas et al. (1997), but ceased growing after 20 years in the MBAC, 2007 study).

Lucas et al. (1997) modelled a single alternative scenario, in which there was no further harvesting, under which carbon stored in live trees increased by 2 Mt CO2e per annum. MBAC (2007) modelled sequestration using FullCAM and found losses due to fire, decomposition of debris and soil carbon, exceeded uptake by trees resulting in emissions of 0.8 Mt CO2e pa.

In the following sections the individual studies and their results are described in detail, as well as other methodologies used to estimate carbon stocks in individual stands by Dean et al. (2011).

Page 280 of 365

Table 56: Estimates of total carbon sequestration in Tasmania’s State Forests.

Source Year Forest Type Carbon Sink (Mt CO2e/y) Vegetation Debris Soil Fire Loss Total Above- Total ground Lucas et al. (1997) 1995- Commercial 3.10 Harvest 2025 Non-Commercial 5.04 Total 8.14 Lucas et al. (1997) 1995- Commercial 5.22 No Harvest 2025 Non-Commercial 5.04 Total 10.26 MBAC (2007) 2007- Commercial 0.60 FT data 2050 Non-Commercial 1.36 Total 1.96 MBAC (2007) 2007- Commercial 0.43 0.17 -0.68 -0.19 -0.27 FullCAM 2050 Non-Commercial -0.17 0.09 0.00 -0.29 -0.55 Total 0.34 0.26 -0.77 -0.49 -0.82

Tasmanian case study report to DEST by Lucas et al. (1997)

This report by Lucas et al. (1997) provided the first estimate of carbon stocks in Tasmania forests. Actual growth data from plots and modelled forward estimates provided by Forestry Tasmania were used in the study instead of default growth rates estimated from FullCAM. Growth data was based on current and predicted bole volume (i.e. the portion of the trunk excluding stump and the upper section in the crown for all trees <10 cm DBH) for 1995, 2005, 2015 and 2025.

Forest types consisted of 75 forest classes, ranging from regeneration to mature eucalypt forest, across 25 inventory areas. These were each split into areas available and unavailable for harvest based on existing formal and informal reserves, areas excluded from harvest for other reasons and area classed as production forest. Both overstorey and understorey species were included. Volumes were converted to total above-ground biomass with the aid of published figures for density and expansion factors for each species. Only above-ground biomass was included. Rates of carbon sequestration were estimated from the change in total carbon stocks over time.

The estimated total stock of carbon across publicly owned native forests in 1995 was 843 Mt CO2e (Table 57). The largest stock was in mature forests (398 Mt CO2e). Forests available for harvest contained a total of 269 Mt CO2e while mature forests unavailable for harvest contained 574 Mt CO2e. Rainforests available for harvest contained just 3% of total Carbon stocks.

Page 281 of 365

Table 57: Estimated carbon stocks in 1995 and annual sequestration (carbon sink) from 1995-2005 in public native forests in Tasmania. Source: Lucas et al. (1997). Harvest Availability Forest type Mature Regrowth Regen. Rainforest Other Total

Carbon Stocks (Mt CO2e) Available 129 90 14 26 9 269 Unavailable 269 87 1 189 28 574 Total 398 177 16 216 36 843

Carbon Sink (Mt CO2e) Available 0.8 1.1 2.4 0.1 0.1 4.6 Unavailable 2.2 1.3 0.2 1.1 0.3 5.1 Total 3.0 2.5 2.6 1.2 0.4 9.7

The annual carbon sink across the estate for the period 1995-2005 was estimated to be 9.7 Mt CO2e per annum (Table 57). The largest total sink was mature forests (3.0 Mt CO2e per annum). The largest sink in forests available for harvest was younger regeneration (forests harvested between 1950 and 1990) which was estimated to sequester 2.4 Mt CO2e per annum.

Cessation of harvest of publicly owned native forests was estimated to potentially increase total carbon stocks by 63.5 Mt CO2e between 1995 and 2025 (equivalent to 2.1 Mt CO2e per annum). However, the rate of carbon sequestration of growing forests was estimated to be 2.5 Mt CO2e per annum greater with continued harvesting compared with no further harvesting.

The report did not provide areas of Tasmanian forest types, but did estimate per hectare rates of carbon sequestration. These estimates ranged from 0.6 t C/ha pa for Rainforest to 4.1 t/ha pa for young regeneration from tall dense eucalypt forest. In general, rates of carbon sequestration were highest in young forest classes and lowest in mature forest classes.

Extrapolation of the results across the whole of Tasmania indicated the total carbon sink in public and private forests was 10.9 Mt CO2e per annum for 1995-2025. This figure represented 30% of the estimated total rate of sequestration across Australia (37.2 Mt CO2e pa).

Evaluation

This study provides a useful framework for considering large-scale modelling of carbon stocks in Tasmania, and it adopts a more field-data-informed approach compared with that used by Norris et al. (2010). The estimated growth rates from the study form the basis for DCCEE’s (2011b) default estimates for harvested native forests (see Section 7.2). Some limitations with respect to the objectives of the Carbon Study are that no attempt was made to include carbon stocks in debris or wood products, no detailed information was provided on the growth models used, or their accuracy, and there was no consideration of the effects of fire on carbon stocks.

Page 282 of 365

Projected carbon stocks in Tasmania’s State Forests (MBAC, 2007)

MBAC (2007) modelled current and future carbon stocks in commercial and non-commercial forests and plantations managed by Forestry Tasmania using a number of approaches, each of which is detailed below.

Woodstock merchantable volume estimates

The approach based on the Woodstock model is similar to that employed by Lucas et al. (1997) in that it used outputs of total and merchantable wood volume from the State Forest estate and converted these to tonnes of carbon using a series of multiplication factors. Woodstock outputs were based on predicted volumes for different forest classes from Forestry Tasmania’s Forest Class growth model for the period 2005 to 2050. Estimates were then converted into bole carbon content assuming a wood density of 500 kg/m3 and a carbon content of 50%.

A series of multiplication factors were then applied to account for non-harvested areas of coupes, non-merchantable trees and other tree components (Table 58). Two similar methods were used to convert merchantable bole volume into total biomass: one based on expansion factors provided by Forestry Tasmania and another based on MBAC’s own expansion factors. The study did not consider carbon in dead standing trees, coarse woody debris or soil.

Table 58: Multiplication factors used to estimate total carbon from merchantable and total stem volumes provided by Forestry Tasmania for different forest types. Two sets of expansion factors were use: one provided by Forestry Tasmania (FT) and another based on MBACs own estimates (MBAC).

Component Multiplication factor Description/reference

Native forest Hardwood Softwood Commercial Non-commercial plantation plantation FT MBAC FT MBAC FT MBAC FT MBAC Convert MSV to total underbark Assumes 20% of stemwood 1.25 1.25 1 1 1 1 1 1 stemwood volume unmerchantable. Adjust harvested volume for area Assumes, 22.5% of stand 1.29 1.29 1 1 1 1 1 1 of coupe unharvested volume retained on site. NCAS Technical Report No. Density (t/m3) 0.5 0.5 0.5 0.5 0.5 0.5 0.4 0.4 18 Expansion factor to convert stem FT figures from NCAS mass into total aboveground tree 1.46 1.4 1.46 1.4 1.34 1.4 1.37 1.4 Technical Report No. 17, p mass 21 FT figures from NCAS Root to Above ground (FT) and 0.25 0.4 0.25 0.4 0.25 0.4 0.25 0.4 Technical Report No. 17, p Root to Stem (MBAC) ratios 63 NCAS Technical Report No. Carbon mass (% dry weight) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 17, p 615 Total 0.736 0.73 0.456 0.45 0.419 0.45 0.343 0.36

Page 283 of 365

There was only a small difference in the estimated total standing stock of carbon from the two methods. Estimated carbon stocks held in commercial and non-commercial native forests, hardwood plantations and softwood plantations varied from 150 Mt C to 151 Mt C (Table 59).

Table 59: Merchantable volumes (for commercial native forest only) and total volumes (for other forest types) output from the Woodstock model and total standing carbon stock estimated, using two alternative approaches, for Tasmania’s State Forests as at June 30th 2007. Factor Native forest Total Hardwood Softwood Com- Non- plantation plantation mercial commercial

Total volume (million m3) 77.1 197.8 3.1 8.9 286.9 Total carbon (Mt) - FT 56.7 90.2 1.3 3.0 151.3 Total carbon (Mt) - MBAC 55.9 89.0 1.4 3.2 149.6 Forward estimates of carbon for Tasmania’s State Forest estate were based on grown-on per hectare volumes for each forest class. These estimates indicated total carbon stocks across all forest types could be expected to increase from 151 Mt C in 2007 to 177 Mt C in 2050. After 2030, total carbon stocks were forecast to increase in all forest types, reaching 64 Mt in commercial native forests, 106 Mt in non-commercial native forests, 4 Mt in hardwood plantations and 3 Mt in softwood plantations by 2050.

FullCAM Modelling

MBAC used FullCAM to model carbon stocks in vegetation, debris and soil based on two methods: one in which FullCAM was used to model total standing carbon in live trees as well as carbon in litter and soil, and a second where FullCAM was used to model carbon in litter and soil only with these estimates added to Woodstock-based estimates of carbon in standing trees.

The first method modelled growth of different forest types based on inputs to the tree yield formula from Forestry Tasmania rather than the default parameters. Non-endemic species modifiers ranged from 1-1.2 to the tree yield formula for native forests.

The commercial forest estate was stratified into 14 different forest types or management systems and the non-commercial estate was stratified into 6 different types. Carbon in hardwood plantations was included. Around 100,000 ha of non-commercial forest estate had not been allocated to a forest type at the time of MBAC’s report and no parameters for softwood plantations were provided.

Estimated carbon in commercial and non-commercial native forests and hardwood plantations for June 2007 was: • commercial native forests and hardwood plantations: 48 Mt; • non-commercial native forests: 71 Mt; • softwood plantations: 3 Mt.

Page 284 of 365

These estimates were substantially less than those derived from Woodstock. The authors speculated this difference was a result of the smaller root: shoot ratio assumed by Forestry Tasmania in the FullCAM model (0.09 - 0.11) compared with the expansion factors applied to the Woodstock outputs (equivalent to 0.25). It is also likely that inputs to the tree yield formula may need to be varied together with the allocation of biomass to roots in order for FullCAM to better approximate the growth of trees on State Forest. Alternatively, increments taken from Forestry Tasmania’s growth model can be input directly into FullCAM.

Forward estimates of total carbon from the FullCAM model indicated the amount of carbon stored in State Forests could be expected to increase from 120 Mt C in 2010 to 129 Mt C in 2050. As with the Woodstock modelling, there was a decrease in carbon stocks in commercial native forests between 2010 and 2020, although this included the increase in carbon stocks in the maturing hardwood plantations. However, the increase in non-commercial native forests 72 Mt C in 2010 to 74 Mt C in 2050 was far less than that estimated from the Woodstock model (92 Mt C in 2010 increasing to 106 Mt C in 2050). This difference is because the tree yield formula in FullCAM assumes growth of Stands tends to flatten off after 100 years, while Forestry Tasmania’s growth model assumes that stem volume of mature forests continues to increase.

Carbon emissions from wood products

To estimate emissions from wood products, three product types were considered – long cycle (5% loss every 50 years with 10% being recycled), short cycle (20% loss every 3 years with 50% being recycled) and waste (0 years). It was assumed carbon loss over time was linear. Sawn timber and panels were assumed to be long cycle, woodchips short cycle and mill residues were assumed to be waste. The various products produced from native forest, hardwood plantations and softwood plantations were based on log removals in 2006.

This analysis provided an indication of total emissions from different forest products over time (600,000-800,000 t C per annum for woodchips, 0-80,000 t C pa for sawn wood /panels and 60,000 t pa for waste). However, the study did not estimate the change in the pool of total carbon in wood products over time, nor were emissions from wood products in service prior to 2006 included.

Evaluation

The approaches used by MBAC to estimate carbon stocks in State Forests provide a useful indication of current and potential future carbon stocks under the current harvesting scenario. Limitations with respect to its direct application to the objectives of the Carbon Study are:

• modelled growth projections rely on outputs from Forestry Tasmania’s growth model, without reference to underlying data; • absence of detailed uncertainty estimates; • restricted to just one forest management scenario;

Page 285 of 365

• lack of information for the basis of the above scenario (e.g. harvesting rates, fire); • reliance on default parameters (especially for FullCAM); • lack of information on emissions associated with wood products; and • lack of emissions associated with post harvest regeneration burning.

A specific issue raised in the report relates to the forward estimates of growth of mature Stands. These were considered to be unrealistically high. Thus, growth was constrained after 20 years. There is no discussion about the reason for the potential over-prediction of growth.

Assessment of carbon stocks in State Forests (Moroni et al., 2010)

This study used forest growth models developed by Forestry Tasmania to estimate current carbon stocks in State Forests managed by Forestry Tasmania. A central focus of the paper was that the concept of Carbon Carrying Capacity as used by Mackey et al. (2008) should actually have been termed “Theoretical C Saturation” because the plots used to estimate carbon content in unharvested forest were located in old-growth forest and thus did not provide a representative sample covering all age classes arising after natural disturbance. The paper considered carbon in live and dead standing trees (including roots) and excluded carbon in other pools (coarse woody debris, litter and soil). It covered native forest managed by Forestry Tasmania, but did not include hardwood or softwood plantations, private native forest, or native forest managed by other agencies.

Total carbon in live and dead standing trees was estimated from Stand volume with the addition of expansion factors. Current standing volume in 2009 was estimated using measurements of standing volume from ~3500 plots in 2007 grown on for two years using stratum level volume growth curves derived from an empirical growth model developed by Forestry Tasmania. Total standing volume for each forest class was calculated as the weighted average volume of all strata comprising each Forest Class.

Silvicultural regeneration aged <20 years was not of sufficient size for routine measurement, so regeneration from 1990-2000 was assumed to have half of the stem volume of that from 1980- 90 while regeneration from coupes harvested since 2000 was assumed to have zero volume. Little data was available for volumes of Rainforest, other forest, or unstocked forest so volumes for these Forest Classes were estimated from expert opinion.

Standing bole volumes were converted into standing biomass by assuming a dry density of 500 kg/m3 then multiplying this by a series of expansion factors: • 1.46: to estimate total above ground mass; and • 1.25: to include root biomass.

Dead trees were assumed to consist of stem and roots only and carbon content of the trees was assumed to be 50% of the total biomass.

Page 286 of 365

This study found the carbon in standing live and dead trees in Tasmanian state forest was 163 Mt C, with 60% in wet eucalypt forest. Average carbon density (the amount of carbon per hectare of land) ranged from 0 to 470 t C/ha across all Forest types and ages. Average carbon density for Rainforest was 67-100 t C/ha while that in Other Native Forest was 72-83 t C/ha. Forest that had never been commercially harvested had an average carbon stock of 155 t C/ha. Carbon stocks in forests regenerating after harvest ranged from 19-170 t C/ha depending on age and predicted height at maturity. Mature wet and dry eucalypt forests contained 121 t C/ha and 232 t C/ha respectively.

Average carbon density of areas of production forest scheduled for future harvest (couped production, 124 t C/ha) was similar to that in informal and formal reserves (~120 t C/ha) as well as that in areas of production forest unlikely to be harvested (uncouped production 102 t C/ha).

Moroni et al. (2010) predicted that, if all areas of State Forest were able to reach theoretical C saturation (the maximum Carbon Carrying Capacity in any point of time for the respective forest type), an additional 65 Mt C could be sequestered. However, the authors also pointed out, that, if the Rainforest was converted to moderate productivity eucalypt forest, an additional 28 Mt C could be sequestered in the standing trees by the time they attained maturity.

The authors argued theoretical carbon saturation is impossible to achieve across the total forest estate simultaneously because of the role of natural disturbance and the change in forest structure over time. The main form of natural disturbance is fire and, in the absence of this, older eucalypt trees will not be replaced by younger regeneration resulting in a change in species composition to Rainforest vegetation. Since the carbon content of Rainforest plots measured in the study was lower than that of the mature eucalypt plots, the authors concluded that this would result in a reduction in carbon stocks over time in the absence of disturbance.

In reference to previous studies by Mackey et al. (2008) and Keith et al. (2009, 2010), Moroni argued that the estimates of Carbon Carrying Capacity of forests in SE Australia were: • too high relative to the average carbon density of tall forests in Tasmania aged >110 years; • based on old carbon dense forest which were not the best reference point for evaluating landscape carbon dynamics; and • assumed that all Stands could achieve maximum carbon content simultaneously without taking into the account the effects of fire or the ecology of fire dependent ecosystems such as mountain ash forests.

Moroni et al. (2010) estimated that 79% of the eucalypt component of State forest and 50% of all forest had regenerated after wildfire and that the total eucalypt component of state forests would have been likely to have been burnt sometime in the past 400 years. Thus, natural disturbance prevents theoretical carbon saturation being achieved across the whole estate simultaneously. Further, the authors argued that, since a) wet eucalypt forest eventually succeeds to Rainforest

Page 287 of 365

in the absence of fire and b) the carbon density of Rainforest is less than wet eucalypt forest, carbon stocks were likely to decline in the long term in the absence of any form of disturbance.

Critique of Moroni (Dean, 2011)

Following the publication of the analysis by Moroni et al. (2010), a discussion ensured in the literature regarding the article’s criticism of some of the proceeding work (including the report by Mackey et al. 2008) and treatment of the concept of Carbon Carrying Capacity.

Moroni et al. (2010) had argued that, although the definition of CCC was “the average carbon stock for an ecosystem, taking into account the effects of periodic natural disturbance such as fire”, what had actually been reported by Mackey et al. (2008) and Keith et al. (2010) as CCC was actually closer to Theoretical C saturation. The authors argued that this meant that it was invalid to assume that the whole forest ecosystem could achieve this estimated CCC at a landscape scale.

In reply, Dean (2011) claimed that earlier studies had attempted to estimate CCC as originally defined and that Moroni et al. (2010) were wrong to indicate that the term CCC required redefining. Dean also used the data provided by Moroni et al. (2010) to estimate the C deficit in standing trees resulting from previous timber harvesting (29 Mt C) based on the area weighted difference in carbon density between carbon stocks in unharvested forest >110 years and harvest regeneration.

A key weakness noted by Dean (2011) in the original article was the lack of plot data for Rainforests. These comprise a large portion of the total estate managed by Forestry Tasmania and thus represent a major carbon stock. However, values for carbon stocks in these forests were based on expert opinion with no indication of error or confidence levels. This meant that any claims that forest carbon stocks declined as eucalypt forest progressed to Rainforest in the absence of fire were not backed by any hard data.

Moroni (2012) subsequently responded to the critique by stating that, even though Dean (2011) agreed that CCC should include the effects of natural disturbance, he had then gone on to estimate it using only the older Stands (>110 years) without including younger regeneration from wildfire. Dean (2011) had argued that, since E. regnans can reach a maximum age of at least 400 years, using forest aged 110 years and above to estimate CCC was reasonable. Dean (2012) then responded with a follow-up reply in which he stated that a range of stand ages including regrowth forest had been used to estimate CCC in the previous reply.

Evaluation

This study provides another useful framework for contemplating estimation of carbon stocks in Tasmanian forests and, critically, makes an important contribution to published figures on carbon stocks in native forest by using actual plot data collected by Forestry Tasmania. The various reviews and opinions presented around the study are also very useful because they

Page 288 of 365

provide a clear indication of key points of uncertainty surrounding estimates of the carbon content of forests with, and without, timber harvesting. A key point of contention appears to be disagreement about the appropriate way to estimate the true Carbon Carrying Capacity (in the original sense of the term) of a forest in the absence of harvesting including: • accounting for the effects of natural disturbance on the carbon content of stands; • estimating the long term change in carbon content of tall open forest in the absence of timber harvesting and wildfire; • obtaining a true area weighted estimate of CCC at a landscape level; and • estimating carbon stocks in Tasmanian Rainforest. On balance, with respect to estimating Carbon Carrying Capacity in the current Carbon Study, it is considered important to account for the impact of fires occurring in different forest types.

Estimation of carbon stocks in Tasmanian old-growth forest (Dean et al., 2011)

Dean et al. (2011) estimated carbon stocks in an old-growth forest. The estimate was based on 30 plots across 7.7 ha in one planned logging coupe. Estimates for Stags were based on DBH measurements from ten 30 m diameter plots located in a logging coupe prior to harvest. All material > 0.2 m diameter was measured and classified according to degree of decay (hard: 100% undecayed mass present; medium: 67% undecayed mass present; and soft: 33% undecayed mass present). It was assumed that the wood density of live trees was 524 kg/m3 (based on data from Ilic, 1997). Total biomass of live trees and Stags was estimated using three eucalypt allometric equations: • E. obliqua stem mass from Keith et al. (2000); • E. regnans stem mass from Dean et al. (2003) and Dean and Roxburgh (2006); and • E. delegatensis merchantable stem volume from Wang & Hamilton (2003) multiplied by a ratio of total stem volume and merchantable stem volume for E. regnans from Wang & Hamilton (2003) and an estimated wood density of 524 kg/m3 from Ilic et al. (1997).

Estimates from the three equations were averaged for each Stag and the calculated above- ground mass multiplied by 0.67 (to account for loss of branches and bark and the presence of internal voids) and the proportion of undecayed mass remaining based on assessed decay. This value was then multiplied by the proportion of original tree height remaining. Original height was estimated from a height: DBH relationship for E. regnans from Dean et al. (2003) adjusted to account for the shorter stature of E. delegatensis.

Total above-ground biomass was estimated from DBH measurements in live overstorey (i.e. E. delegatensis) trees, live understorey trees, Stags (standing dead trees) and coarse woody debris using DBH measurements (for live trees), DBH and height (for Stags) and diameter and length (for coarse woody debris), using a range of allometric equations and assumed wood densities for different components and degrees of decomposition of dead material.

Page 289 of 365

Total average carbon in above-ground live and dead biomass was estimated to be 622 t/ha (+/- 108 t/ha) based on the averages of the various allometric equations used for overstorey and understorey species and assuming a loss of 25% of total estimated carbon in overstorey trees as a result of senescence. Overstorey trees contributed 55% to total carbon stocks above ground, understorey trees 21% and dead biomass 25%.

Evaluation

This study provides valuable information for carbon stocks in mixed old-growth eucalypt forest. However, it is limited to one site with no indication of the productivity of the site or its representativeness of mixed eucalypt forest in general. This makes extrapolation to other sites is difficult. Nevertheless, this is a useful reference for potential carbon stocks in mature eucalypt forest.

Assessment of carbon stocks in relation to inter-governmental agreement (Macintosh 2012a)

A study by Macintosh (2012a) examined the potential impact of new reserves planned by Environmental Non-Government Organisations (ENGO reserves) under the Tasmanian Forests Inter-Governmental Agreement (TFIGA) on Australia’s total emissions and estimated potential commercial value associated with sequestered carbon under a scenario where Kyoto eligible units were issued in relation to abatement arising through altered forest management.

At the Durban Climate Conference, new rules were agreed governing the way that emissions associated with forest management (FM) will be accounted in future (UNFCCC, 2011). Under the terms of the second commitment period, it is compulsory to account for emissions associated with management of existing native forests and plantations. Although Australia has not yet decided to participate in the second Kyoto agreement, the author considered it likely that, even if Australia opted for an alternative framework, this would likely include emissions associated with forest management. In this case, it might be possible to realise Kyoto ACCU’s under the CFI in relation to reserving of areas of native forest, although the author noted: “If Australia does not account for FM, or there is a cap on FM credits that blocks the impact of the TFIGA, the creation of the ENGO reserves will have no impact on Australia’s mitigation commitments” (Macintosh 2012b, p. 14).

The study estimated the potential impact of increasing the area of forest reserves as proposed by ENGOs on total emissions from Tasmanian native forests. It considered both emissions associated with forest management and harvesting and emissions associated with wood products arising from these forests. Two scenarios were considered: • reference scenario in which no additional reserves were created, and • ENGO scenario where reserves created as proposed under ENGO and TFIGA proposals.

Page 290 of 365

Two time periods were covered 2013-20 and 2021-30. Under the reference scenario: • annual rate of harvesting was equal to the average for Tasmanian publicly owned native forests for the period 2002-09. • flow of wood products from Tasmanian multiple use forests was equal to that for 2008.

Under the ENGO scenario it was assumed:

• annual area of forest harvested was reduced on a pro-rata basis across forest with specialty timbers and forest with non-specialty timbers according to the proposed reductions in harvestable areas under the ENGO scenario; and • flow of wood products was equal to the projections under the ENGO scenario from a report produced by Forestry Tasmania (2011b).

Two approaches were used to accounting for emissions. The first, based on the methodology currently used by the Australian Government to account emissions under its obligations under the Kyoto agreement and the second was based on a possible CFI methodology. The main difference between the approaches related to the assumed rate of future harvesting. Under the Kyoto approach this was assumed to remain constant at 2008 levels while, under the CFI approach, it was based on projected log removals for 2011-30 by Forestry Tasmania (2011a).

Emissions under each scenario were estimated using similar approaches as the Australian Government currently uses to account emissions from native forests in its National Greenhouse Accounts (DCCEE 2011a, 2011b). Thus, the FullCAM carbon accounting model was used together with plot files from the DCCEE database to model emissions from above- and below- ground biomass and debris in Tasmanian native forests. Log volumes and wood product data from ABARES (2012a), Forestry Tasmania (2005a, 2006, 2007, 2008, 2011a) and Ryan et al. (2002) were used to model emissions from harvested wood products.

For modelling emissions from forest management and harvesting, only forests subjected to timber harvesting were considered. It was assumed that emissions from non-harvested forests would be the same under either scenario. It was also assumed emissions associated with wildfire would be the same for each scenario. Thus, effects of potential future fires were not modelled. Out of the original 73 plot files used by DCCEE, 19 were identified as representing harvesting of Tasmanian native forest for the period 2002-09 and these were used to model future emissions.

Modelling emissions associated with harvested wood products included carbon in exported wood products and excluded carbon in imported wood products. This approach differed from that used for the national accounts, which excludes carbon in exported wood products (assumes all carbon is converted to CO2 when products exported), but includes carbon in imported wood products. Adjustments were also made to assumed decay rates, with decay rates of both domestic exported wood products based on defaults set out by the UNFCCC (2010, p. 34)

Page 291 of 365

rather than those used by DCCEE. Assumed decay rates were: Paper: 2 years; Wood panels: 25 years; Sawn wood: 35 years. The study assumed a 15% increase in emissions as a result of leakage (arising from an increase in wood production from forests in other parts of Australia as a result of a potential reduction in wood supply from Tasmanian native forests) and allowed for a 5% deduction as a CFI risk of reversal buffer. In considering leakage, the study referenced domestic sources only.

Under the Kyoto methodology, in which reference case harvest volumes were assumed to be constant at the 2008 level, the study estimated that creation of ENGO reserves would result in an average of 8.3 Mt CO2e per annum reduction in total emissions with no assumed leakage or a 7.0 Mt CO2e pa reduction with 15% leakage from 2010 to 2030.

Under the CFI methodology, future harvest volumes were projected to be 28% and 48% below 2008-09 levels for specialty timbers and non-specialty timbers respectively. Thus, the estimated reductions in emissions associated with creation of the ENGO reserves were much less (2.3 Mt CO2e per annum assuming no leakage and 2.0 Mt CO2e per annum for 15% leakage).

The study acknowledged a proportion of the estimated reduction in emissions associated with the formation of ENGO reserves would probably occur anyway as a result of increased supply of wood from plantations and depressed export markets for woodchips. Thus, the author stated that “the results should not be taken to represent the Kyoto ACCUs (Australian Carbon Credit Units) that might be issued in relation to ENGO reserves” (Macintosh 2012a, p. 24).

Evaluation

This study provides a useful treatment of possible accounting approaches that may arise where Australia takes on full forest emissions accounting under Article 3.4 of the Kyoto Protocol and is a useful reference for considering how carbon abatement might be valued under a possible future policy setting. The study also provides a valuable contribution in demonstrating how the current non-spatial methodology applied by DCCEE for estimating carbon emissions from Australia’s native forests could be adapted for State-wide or regional estimates.

With respect to the specific objectives of the Carbon Study, the approach adopted under this study has limitations. These include assumptions around treatment of wildfire (assumed to have the same effect on emissions from harvested forests compared with non-harvested), and a lack of detail around emissions associated with various forest types and biomass components. No distinction is made between emissions from forests and emissions from wood products, so it is not possible to assess the impact of the different pools on total emissions.

However, the underlying parameters used in the modelling together with assumed harvest rates and intensities provide a basis for future work. Importantly, the results from the study provide a useful reference for at least one of the Study Scenarios considered in the current Carbon Study.

Page 292 of 365

7.5 TASMANIAN FOREST TYPES AND CLASSIFICATION

Tasmania’s forests and plantations cover a wide range of geologies, topographies and climates. They are also highly diverse ecologically. Their current form is largely determined by a long history of natural and human disturbance that continues to change their structure and age. Thus, estimating their past, current and future carbon stocks is a complex task that requires a detailed understanding not just of the ecological processes involved in carbon cycling, but also of the disturbance history and potential future impacts of climate change and fire on their future growth and regeneration. In this section we review the different approaches used to classify Tasmania’s forest types and discuss them in relation to estimating carbon stocks.

Forest groups

Tasmania’s vegetation can be classified into seven major groups. These are: • Rainforest; • Tall Eucalypt Forest; • Low Eucalypt Forest; • Other Native Forest; • hardwood plantations; • softwood plantations; and • non-forest.

A map of forest groups across Tasmania is publicly available from the Department of Primary Industries, Parks, Water and the Environment (DPIPWE).

Rainforest is closed forest without dominant eucalypts and containing typical Rainforest species including myrtle beech (Northofagus cunninghammii), sassafras (Atherosperma moschatum), huon pine (Lagarostrobos franklinii), king billy pine (Atherotaxus seliganoides) or pencil pine (Atherotaxus cupressoides).

Tall eucalypt forest (also known as wet sclerophyll forest) is classified as eucalypt forest with a mature height > 34 m usually dominated by Eucalyptus regnans, E. obliqua or E. delegatensis growing in relatively moist environments.

Low eucalypt forest (also known as dry sclerophyll forest) is classified as eucalypt forest with a mature height < 34 m which may be dominated by a wide range of species including E. obliqua, E. delegatensis, E. amygdalina, E. nitida, E. coccifera, E. sieberi, E. pauciflora, E. globulus, E. viminalis, E. dalrympleana and E. ovata. Non-eucalypt native forest includes forest dominated by Acacia species such as A. dealbata or A. melanoxylon, tea tree (Leptospermum spp.), and paperbark (Melaleuca spp.).

Page 293 of 365

Forest groups do not include non-eucalypt forest < 10 m in height. Thus, like the National Forest Inventory, they exclude large areas of south-west Tasmania that contain vegetation that is taller than 2 m which is classed as forest by DCCEE. In addition, plantations are simply grouped as hardwood (predominantly E. globulus or E. nitens) or softwood (Pinus radiata). Non- forest covers all other areas including shrubland (e.g. buttongrass plains), grassland, farmland, built-up areas and water.

TasVeg classification

Forest groups are useful for providing an indication of extent of different broad forest types. However, they provide little information on species composition, age or structure. A more detailed vegetation classification system is the TasVeg map produced by the Tasmanian Vegetation Monitoring & Mapping Program (TVMMP, 2004) and publicly available from the DPIPWE (http://www.dpiw.tas.gov.au/inter/nsf/WebPages/LJEM-7FAVUU?open). The vegetation communities on which it is based are described in detail in Reid et al. (2005).

The system divides vegetation into 153 distinct communities, aggregated into 11 broad groups:

• agricultural, urban and exotic vegetation: includes hardwood and softwood plantations; • highland treeless vegetation: classed as non-forest under forest groups; • dry eucalypt forest and woodland: roughly equivalent to Low Eucalypt Forest; • wet eucalypt forest and woodland: roughly equivalent to Tall Eucalypt Forest; • native grassland: classed as non-forest under forest groups; • non-eucalypt forest and woodland: includes non-eucalypt forest from forest groups; • other natural environments: classed as non-forest under forest groups; • Rainforest and related scrub: includes areas classed as either Rainforest, non-eucalypt forest and non-forest under the forest groups; • scrub, heathland and coastal complexes: classed as non-forest under the forest groups; • moorland, sedgeland, rushland and peatland: classed as non-forest under forest groups; and • saltmarsh and wetland: classed as non-forest under forest groups.

Each of these is subdivided into different vegetation communities, primarily on the basis of species mix. However hardwood and softwood plantations are aggregated.

Forest Class system

Another approach used to classify forest is the Forest Class system, developed by Forestry Tasmania (Stone, 2000) principally for estimating wood production across its forest estate. More recently, it has been implemented to assess representativeness of the reserve system, identify old-growth forest and estimate carbon (e.g. Lucas et al. 1997; Moroni et al. 2010).

Page 294 of 365

The system divides Tasmania’s forests into 91 Classes (see Table 45). These in turn are made up of over 500 Photo-Interpretation (PI) types which are used to stratify forests from aerial photography. Eucalypt forest types comprise 84 classes (1-84), non-eucalypt forests, 4 classes (85-88) and plantations, 2 classes (89 and 90).

Each Forest Class contains detailed information on forest stand structure age and productivity. It also contains information on past disturbance history and so assists with tracking changes in forest classes over time. The Forest Class System is therefore ideally suited to estimation of standing volumes and carbon across eucalypt forest. However, it is less suited to assessing floristics and has little information on non-eucalypt and non-forest communities.

Although Forest Class does not identify species, it does discriminate native eucalypt forest from plantations, non-forest, Rainforest and Other Native Forest. Forest Class can also be used to estimate forest age and determine past disturbance history. While the system has been primarily used for classifying Tasmania’s multiple use State Forests, in 1996 all native forests were classified in preparation for the 1997 Regional Forest Agreement (Anon, 1997). Since then, coverage of all publicly owned forest has been regularly updated. However, the only available mapping of Forest Classes across privately owned forest is from 1996.

One limitation of the Forest Class system is its treatment of Rainforests. While there is relatively high differentiation of types, ages and structures of eucalypt forest in the Forest Class system, there are only two classes of Rainforest: M+ (FC85 > 25 m height) and M- (FC 86 < 25 m height). In contrast, the classification by Reid et al. (2005) references four Rainforest groups: • callidendrous Rainforest: tall (up to 40 m) to medium height forests dominated by myrtle beech (N. cunninghammii) and/or sassafras (Atherosperma moschatum); • thamnic Rainforest: medium height forest with a distinct shrub layer and a combination of species usually including myrtle beech, king billy pine (Atherotaxus seliganoides) and/or huon pine (Lagarostrobos franklinii); • implicate Rainforest: Low forests (< 20 m) with broken uneven canopies and tangled understories, dominated by a mixture of species including myrtle beech, huon pine, king billy pine, woolly tea tree (Leptospermum lanigerum) and celery top pine (Phyllocladus aspleniifolius); and • open montane Rainforest: high elevation (600-1100 m), low (6-15 m) forests consisting mainly of pencil pine (Atherotaxus cupressoides) with a shrubby or grassy understorey.

These different communities correspond only roughly with the two Rainforest classes used in the Forest Class Systems. Another limitation of the Forest Class system is that, as with Forest Groups, non-eucalypt forest that is < 15 m tall is classified as non-forest compared with the definition used by DCCEE (2011b) of > 2 m. This means carbon stores in these forests would be excluded from estimated total stocks, resulting in forest area (and associated carbon stocks) being underestimated. Additionally, there is no coverage of privately owned plantations established since 1996.

Page 295 of 365

Despite the limitations, the Forest Class classification system is well suited to estimating forest carbon. An important strength is the fact that Forestry Tasmania has around a series of around 3000 permanent and temporary sampling plots located in most forest classes across its estate. These are all mapped to forest classes and provide detailed information, not only of standing tree volumes, but also coarse woody debris and Stag (standing dead tree) volumes.

Greater resolution of different vegetation communities comprising Rainforest types, Other Native Forest types and other woody vegetation of 2-15 m height may be obtained by overlaying maps of Forest classes with the TasVeg map. Similarly, the Forest Group maps provide more up to date information on plantation areas on private land.

Forest tenure

Forest tenure classification is important for modelling past and future potential management across different forest types and locations. Tasmanian forests can be divided into four main tenure types:

• public forest managed primarily for conservation purposes; • public forest managed primarily for wood production; • private forest where wood production is excluded; and • private forest where wood production is permitted.

Forests managed primarily for conservation purposes include National Parks, World Heritage Reserves and other conservation reserves. These also include formal and informal reserves on state forest, the difference being the level of sign off required to change the reserve status. In general, formal reserves require an act of Parliament to change their status, while informal reserves can, in theory at least, be converted to production forest without this requirement.

A further component of the reserve system includes public state forest where wood production is judged to be uneconomic, dangerous or is restricted by code of forest practice requirements (such as stream reserves and buffer strips). Forest managed primarily for conservation purposes includes Other public land, such as that owned by the Hydro-electric Commission or the Department of Defence.

Public forests managed primarily for wood production include all other public forests where timber harvesting is permitted. They include both plantations (almost entirely hardwood) as well as native forests. A range of silvicultural systems and rotation lengths are used in native forests depending on the forest type and production objectives. A small component of forest is included in Special Timbers Production zones which are managed on very long rotations (~ 150 years) for the main objective of supplying high value timber for craftwood and furniture production. In contrast, plantations are managed highly intensively on short rotations ranging from 10 years for pulpwood to 35 years for sawlog production.

Page 296 of 365

Woody vegetation extent

Forest extent layers derived from LANDSAT imagery for the period 1972-2010 are held by DCCEE. These layers are used to track land-use change across the country for modelling changes in carbon stocks. They are primarily designed to track changes since 1990 to allow Australia to account for its obligations under the Kyoto agreement. Thus, the level of detail and disaggregation is greater for forests established or cleared within this period than for prior years.

Forest extent layers from 1990 onwards separate forest cover into three types: • softwood plantations; • hardwood plantations; and • other forest (i.e. native forest).

Although algorithms have been developed to distinguish native forest, hardwood plantations and softwood plantations using Landsat data, these are currently used within the NCAS framework to identify plantings on previously cleared land only. Thus, native forest that is converted to plantations post-1990 is assumed to remain as harvested native forest (DCCEE 2011b, M. Searson DCCEE, Pers. Comm.). Further, although harvesting of plantations after 1990 is captured spatially, thinning of plantations and harvesting of native forests is not. Methodology to map native forest harvesting and conversion to plantations is currently being developed.

Woody vegetation extent based on Landsat data currently classifies button grass plains as > 2 m woody vegetation. These areas cover large portions of southern and central Tasmania and thus represent a large error in area estimates. DCCEE is currently working with DPIPWE to resolve this problem and improve the accuracy of mapping forest extent.

7.6 FOREST MANAGEMENT IN TASMANIA

In order to be able to model past, current and future carbon stocks, it is important know what silvicultural systems are used to harvest and regenerate the forest. The type of system used influences not just the proportion of carbon removed from a particular site, but also the amount of carbon that is moved into debris and soil pools or emitted to the atmosphere and the subsequent growth of the stand. Different silvicultural systems such as clearfelling and burning, or partial harvest followed by mechanical disturbance will have differing effects on the amount and type of material entering each of these pools. Furthermore, the cycling time for the various carbon pools can vary from less than a year to more than a century. Thus, we need to know, not just what types of silvicultural systems are used currently, but also what was used in the past as well as something of the general disturbance history of Tasmania’s forests.

Page 297 of 365

The following section provides a brief history of forest management and disturbance of Tasmania’s native forest, reviews the development of the various silvicultural systems in use today, and describes them and the types of forest where they are used. It also provides a brief outline of the silviculture used for hardwood and softwood plantations.

Native eucalypt forest

Past timber harvesting

Tasmania’s native forests have been harvested for wood production over the past 200 or so years. This management has had a varying impact on the current mixture of forest types and structure as well as the current silvicultural systems used. Some information on very early management of Tasmania’s forests since the 1800’s is provided by Mount (2009). A comprehensive history of forest management in Tasmania since 1900 is provided by Elliot et al. (2008).

Prior to European arrival there is evidence of widespread use of fire by aborigines in Tasmania’s forests, probably for the purpose of producing green feed and hunting animals. Fire frequencies may have been around once every 10 years, promoting the development of an intricate mosaic of different ages in woodlands and dry forests plus large areas of button grass (Mount, 2009). In some areas, this created an open savannah type landscape such as in parts of the Florentine valley, which later reverted to forest (Gilbert 1959). There is less evidence of aboriginal habitation in the wetter forests, so it is more likely that fires in these forests were started as a result of lightning or escaped man-made fibres from the drier forest in the foothills (Mount 2009).

With the arrival of Europeans and the displacement of the aborigines, fire regimes changed, sometimes radically. A study of fire scars by Von Platen et al. (2010) in dry sclerophyll forests in eastern Tasmania indicates that initially, fire frequency decreased from once every 14 years to once every 25 years, possibly as a result of the conflict between aborigines and early settlers. However, with increasing grazing and growing demand for wood, fire frequency increased from the 1850’s. Around 1910, fire frequency in the area peaked at once every seven years, where it remained till 1989. This increase was associated with increased clearing of poorer land for agriculture, associated widespread burning and timber harvesting.

Prior to the 1930’s, eucalypt forests were generally selectively harvested to remove only the highest quality logs (i.e. logs from the tallest, straightest and best formed trees for converting into sawn timber). In addition, Rainforests were selectively harvested for high value wood from huon pine, myrtle beech and blackwood (A. melanoxylon). This practice often resulted in large wildfires in the wetter forests which destroyed much of the unlogged forest as well as the natural regeneration from harvesting in adjacent areas (Elliot et al. 2008).

In the 1880’s, it was recommended that a sustainable yield practice be adopted. Over subsequent years, a process of mapping forest types and measuring growth was initiated and

Page 298 of 365

silvicultural treatments aimed at maximising survival and growth were implemented. This sustainable yield policy was suspended during the Second World War. After 1945, wood production increased further, peaking in the late 1950’s when it was noted that Tasmania “had reached its maximum permissible production” (Forestry Commission 1961, p. 4).

Silvicultural systems in tall open lowland forests

Selective logging of Tall Eucalypt Forests continued until 1941. However, the development of paper mills in the north-west and southern Tasmania resulted in a market for pulpwood and integrated harvesting commenced. Intensive studies were initiated by the likes of Gilbert, and Cunningham in the 1950’s aimed at better understanding the regeneration requirements of wet eucalypt forests. This work showed many dominant wet eucalypt species were highly shade intolerant and, in the absence of disturbance, the eucalypt forest gradually transitioned into Rainforest or other forest types dominated by former understorey species as the older trees gradually died. It also confirmed earlier observations that the trees were well adapted to regenerate from seed following fire or other disturbance that exposed the mineral soil. As a result of this research, the seed tree silvicultural system was developed which aimed to replicate the effect of fire. This system consisted of: • harvesting coupes of 8-20 ha in size; • removing of all merchantable understorey trees; • retaining 13 dominant eucalypts per hectare to provide seed for the regeneration; • burning of logging slash and non-merchantable trees; and • removing of the seed trees after establishment of regeneration.

Subsequently a clearfell and burn approach followed by aerial or hand application of seed (known as clearfell, burn and sow or CBS) was adopted across most of the estate.

In the 1980’s Forestry Tasmania evaluated alternative silvicultural systems to clearfelling. These included conventional CBS, partial logging and group selection (in which groups of trees are harvested to create gaps of varying sizes). However, in an early trial in E. obliqua forest, successful regeneration was obtained only with the CBS system. A follow-up trial in the Arve valley produced similar results. However, this latter study did demonstrate that alternative silvicultural treatments could be practical where timber production was not the primary aim.

In 1995, the Warra silvicultural trial was established to further study the ecological, economic and safety aspects of alternatives to CBS. These alternatives covered a wider range of residual stocking and gap sizes than previous studies, including: • CBS with understorey islands, • stripfells, • patchfells, • dispersed retention (10% of overstorey retained),

Page 299 of 365

• aggregated retention, • single tree selection, and • small group selection.

Initial results from this trial again indicated that the CBS system was the most successful in terms of regeneration establishment. However, at the time of writing, Elliot et al. (2008) noted that one alternative; aggregate retention was being tested operationally in a number of districts.

Silvicultural systems in alpine forests

Unlike the lowland wet forest, alpine forests dominated by E. delegatensis, often responded poorly to the CBS system (Elliot et al., 2008). This was a result of competition with the grassy understorey and the effects of frost. In the mid-1970’s, a retained tree system (shelterwood) followed by a regeneration burn was shown to be successful in producing adequate regeneration. Subsequent trials in the 1980s resulted in the development of a series of prescriptions for alpine forests. These included: • retaining advance growth (established regeneration in the older stands) where possible; • retaining a forest environment as far as possible to minimise grass invasion and browsing; • providing a seedbed that remained viable for as long as possible; and • minimising competition.

As a result of this work, silviculture in alpine forests focussed on maintaining or developing an uneven age structure in the stands by retaining overwood trees and established younger trees during harvesting, while, at the same time, developing a viable seedbed.

Silvicultural systems in dry sclerophyll forests

As a result of widespread selective logging for sawlogs, combined with grazing and excessive burning, many dry sclerophyll forests across Tasmania were in a poor condition up until the 1970s (Elliot et al., 2008). The resulting forest had a low proportion of sawlog quality trees remaining, and grazing and burning had limited regeneration and arrested understorey growth.

With the development of the export woodchip industry in the 1970s, in the Triubunna area it was recommended that the CBS system be employed despite the highly variable multi-aged structure of these forests. At the same time, large areas of dry sclerophyll forests were being converted to pine plantations in the north-east of Tasmania in the Fingal district.

The CBS system was demonstrated to provide excellent regeneration, but resulted in even aged stands, that differed structurally from the original forest. As a result, attention focussed on the development of appropriate alternative silvicultural systems. Systems tested included burning verse no burning, potential sawlog retention, partial harvest and shelterwood.

Page 300 of 365

During the 1970’s alternative silvicultural systems were trialled in dry sclerophyll forests. Initial results for partial harvesting were poor due to many young trees being killed in regeneration burns. However, reducing fire intensity or mechanically disturbing sites improved survival of retained trees. A trial in north-eastern Tasmania demonstrated that trees retained as potential sawlogs reached sawlog size in just 7 years compared with 24 years in the unharvested forest. However, clearfelling was considered to be the best approach in stands with dense understoreys.

Thinning of regrowth

Early trials in the 1940’s, of thinning of dense regrowth to promote diameter growth, demonstrated that this could significantly increase the rate of merchantable volume growth. However, the lack of a suitable market for the thinnings, combined with the enhanced risk of wildfires due to the logging slash, prevented its operational adoption until the 1980s. Thinning trials established in the 1950s were destroyed by subsequent fires.

In the 1980s and 1990s, thinning was seen as an option to maintain harvestable volumes in the face of increasing areas of forests being reserved for conservation purposes. Research showed that thinning markedly increased sawlog to pulplog ratio and maintained total stand volume, provided > 50% of the original trees were retained. Subsequent work demonstrated that one pre-commercial plus one commercial thinning maximised sawlog yield at 65 years. Harvest of thinned regrowth is now incorporated into Forestry Tasmania’s future sawlog production plans and is seen crucial for maintaining sawlog volumes as harvest of old-growth forests is phased out over the next 20 years (Forestry Tasmania, 2008).

Rainforest

Rainforest covers 563,000 ha of Tasmania and represents 18% of the total area of forest (Hickey et al., 1993). Unlike the eucalypt forest, cool temperate Rainforest evolved to regenerate in the absence of catastrophic bushfires. Rainforest species tend to be relatively shade tolerant and are readily killed by fire.

Early exploitation of Rainforests consisted largely of harvesting valuable species such as Huon pine along rivers. Because much of the Rainforest is located in fairly inaccessible regions, exploitation of Rainforest was far less intensive than that of the eucalypt forests. Thus, it is estimated that around 85% of the original Rainforest estate that existed in 1803 at the time of the first European settlement remains today (Hickey, 1990). However, clearing of Rainforest for agriculture occurred from 1830-1930 in the north-west and north-eastern highlands.

Between the 1960’s and 1970s some Rainforest stands were clearfelled and converted to plantation or native eucalypt forest (Hickey, 1990). However, this practice was restricted by a poor market for Rainforest pulpwood and, in 1982, by a moratorium against clearing of

Page 301 of 365

Rainforest on public land. However, the practice of converting Rainforest to eucalypt plantation (mainly E. nitens) continued on private land, particularly in northwest Tasmania.

Probably the most significant impact of European settlement on Rainforest communities was an increase in the frequency and severity of bushfires. Since 1950, around 8% of Tasmania’s Rainforest has been burnt (Working Group for Rainforest Conservation, 1987). The largest recorded fire in Tasmania’s Rainforest occurred near the Savage River in 1982 which burnt 45,000 ha, 15,000 ha of which were Rainforest (Barker, 1990). The study by Barker (1990) provided evidence that once an area of Rainforest had been burnt, the risk of subsequent fires may increase as a result of the increase in proportion of flammable vegetation present.

Succession of eucalypt forest and other vegetation types to Rainforest in the absence of fire has been observed across Tasmania (Hickey, 1990). This includes E. regnans forest in the Florentine Valley (Gilbert, 1959), Poa grassland and E. delegatensis forest the Mount Morrice Plateau (Ellis, 1985) as well as growth on abandoned farmland (Read and Hill, 1983).

Current silvicultural systems in publicly owned native forests

Current silvicultural systems used in different native forest types across Tasmania are summarised by Forestry Tasmania (2010). A variety of silvicultural systems are currently applied including: • clearfelling • thinning • seed tree

• aggregate retention • shelterwood • advance growth retention

• group selection • seed tree • potential sawlog retention

The main variations between these systems are: • coupe size: with clearfelling using large coupes and aggregate retention or group selection using progressively smaller effective coupe sizes; • mature tree retention: where ~10% mature trees may be retained for seed tree systems and ~50% for shelterwood systems which are removed in a subsequent harvest once the regeneration has become established (5-15 years after initial harvest); • regeneration retention: where mature trees are removed and existing younger trees are retained, ranging from seedling to saplings for advanced growth retention, or older trees for potential sawlog retention; and • thinning: where a proportion of trees are removed from even-aged regeneration.

For modelling purposes, these systems can be simplified into two main systems: • clearfell: in which all mature trees are removed; and • partial harvest: in which a proportion of trees are removed.

Page 302 of 365

Clearfell type systems are normally applied in tall open forests regrowth with a dense understorey and some low open forests on steep slopes. Two methods are used, conventional harvest on coupes with flatter slopes and cable harvesting on steeper slopes.

Partial harvesting systems are normally applied to tall open alpine forests and low open forests. These systems aim to create a multi-aged forest structure by retaining a proportion of trees of varying ages. Regeneration after harvest may be encouraged through mechanical disturbance or low-moderate intensity fire. Shelterwood and seed-tree systems can be considered as a form of partial harvest where retained mature trees are removed 5-10 years after initial harvest.

Thinning, which aims to increase sawlog yield or reduce the required rotation length to produce sawlogs, is a form of partial harvest. Two types of thinning are applied: • pre-commercial: stem injection of herbicide of selected cull trees at age 10-25 years to reduce total stocking to 500-600 stems/ha or 50% of the original stocking; and • commercial: suppressed trees are harvested for pulpwood at age 25-40 years to reduce stocking to 150-250 stems/ha, or 50% of original basal area (Forestry Tasmania, 2001).

Post-harvest regeneration

Following harvest, seedbed preparation may be through burning of harvest slash, or mechanical disturbance. Where fire is used, intensity varies according to harvest method. Typical regeneration methods are: • high intensity burn: after clearfelling; • moderate intensity burn: after aggregate retention, or group selection; • low intensity burn: after initial harvest (for shelterwood or seedtree systems) of alpine forests or low open forests; and • mechanical disturbance: after partial harvest of some forest types including multi-aged low open, mature and alpine forests.

The prevalence of different harvesting systems used in Tasmanian State forest has varied over time. Since the 1990’s, the proportion of forests harvested with partial harvesting systems (i.e. systems other than clearfelling) has increased from <50% to an average of around 60%. In 2005 Forestry Tasmania adopted a policy of reducing the use of clearfelling in old growth forest. This has resulted in the increasing use of aggregate retention in these forests (Forestry Tasmania, 2005).

Page 303 of 365

Rotation lengths

Rotation length (the time between germination and harvest of the mature trees) varies with forest type, productivity and ownership. State owned native eucalypt forests are managed on rotations of sufficient length to allow dominant trees to reach sawlog size (60-80 cm). The nominal rotation length is 90 years, but actual rotations required to achieve this vary with site quality and market conditions (Forestry Tasmania, 2007). Based on the reported results of thinning trials, it is expected that the minimum rotation length for thinned stands could be as low as 65 years (Forestry Tasmania, 2001). Because of their slower growth rate, less productive stands will normally require longer rotations than more productive stands.

Whitely (1999) lists typical rotation lengths used in different native forest types across Tasmania:

• eucalypt (extensively managed): 90 years • eucalypt (thinned regrowth): 65 years • blackwood (swamp forest): 70 years • myrtle (Special Timbers Management Zone): 200 years

The annual amount of sawlog production depends on the total area of forest available for harvest, its long term average growth rate, the rotation length and the average sawlog yield (% of total merchantable log volume that can be used for sawlog). Since more intensive management (i.e. thinning) increases sawlog yield relative to similar aged unthinned stands, this can increase the total amount of sawlog production per unit area. However, this increased level of management intensity may impact on other environmental values of multiple-use forests and reduce their carbon storage capacity (e.g. Dean & Wardell Johnson, 2010).

It is important to distinguish the concept of rotation length from that of harvest return interval. In partially harvested stands, retained mature trees may be removed as soon as 5 years after the initial harvest, once the regeneration has become established. However, the overall rotation length of the stand is still 90 years. Similarly, in stands with multi-age class type silviculture such as aggregate retention, or group selection, the return interval may be around 20-30 years, but the average age of trees harvested may still be around 90 years (for eucalypt stands) or 200 years (for trees in special timbers zones). This is particularly important for modelling carbon, where carbon stocks are closely related to the age of trees at time of harvest (e.g. Dean & Wardell Johnson, 2010).

Current silvicultural systems in plantations

Current silvicultural systems for plantations can be grouped as either long rotation (for sawlog production) or short rotation (for pulpwood production). In general, softwood plantations are managed for sawlog on a 28-30 year rotation while hardwood plantations are either managed over long rotations for sawlog production or short rotations for pulpwood production. Virtually all softwood plantations in Tasmania are thinned Pinus radiata.

Page 304 of 365

Short rotation hardwood plantations are predominantly on private land while long rotation hardwood plantations are predominantly on public land (Parsons et al., 2006). Hardwood pulpwood rotations are normally 10-12 years in length depending on the site, while sawlog rotations are typically 20-25 years (Wood et al., 2009). The current thinning regime used for sawlog rotations consists of a commercial thinning at age 8-12 years in which stocking is reduced from 1100 stems/ha to 300-350 stems/ha. In addition, potential sawlog trees are pruned in a series of three lifts between ages 3 and 5 years.

Site preparation following harvest normally consists of mechanical disturbance for both hardwood and softwood plantations (P. Adams, Forestry Tasmania, Pers. Comm., D. Aurick, Timberlands Pacific, Pers. Comm.). Windrowing and burning is used only on steeper sites where it is difficult for machinery to operate.

7.7 EFFECTS OF FOREST MANAGEMENT ON FOREST CARBON

Timber harvesting

Harvesting influences carbon stocks in a number of ways: • it kills trees, stopping growth and carbon sequestration; • it physically removes carbon from the stand in the form of logs; • non-merchantable material is left on site adding to forest debris; • harvesting and site preparation disturbs soil and litter and affects decomposition; and • following harvest, burning of slash combusts a portion of debris and soil humus, changing its composition and converting a proportion to charcoal. Impact of harvesting on carbon pools in Tasmanian old-growth native forests has been modelled by Dean & Wardell-Johnson (2010). These authors showed that average total carbon stocks (including wood products) could decrease by 45-50% following conversion of old-growth eucalypt forest to regrowth forest harvested on an 80 year cycle. Thus, conversion of old- growth forest subject to periodic natural disturbance, to a regrowth forest or plantation harvested on a shorter cycle, is likely to result in an overall reduction in carbon stocks. The shorter the rotation the likely greater impact on carbon stocks as a result of more frequent burning and lower biomass production.

Site preparation methods

The two methods of site preparation used after harvest of native forest (mechanical disturbance and slash burning) have different effects on carbon stocks in debris and soil. Burning combusts a proportion of harvest and litter debris with the amount removed tending to increase with fire intensity and coverage. Some of the uncombusted material is converted to charcoal which is

Page 305 of 365

usually considered to be inert (i.e. does not decompose), while the remainder remains as debris. Particularly intense fire can also remove some carbon from humus and soil layers.

Studies of the effects of regeneration burning on carbon in debris or soil have generally shown decreases in both pools (Dyrness et al., 1989; Ellis and Graley, 1983; Hatch, 1960; Pennington et al., 2001; Tomkins et al., 1991). Pennington et al. (2001) reported a 6% decline in soil carbon to 0.3 m depth (with all the decrease occurring within the surface 0.1 m) after high intensity regeneration burning on other coupes in the same region. Slijepcevic (2001) showed 58-63% of original carbon in logging slash was lost as a result of high intensity regeneration burns on clearfelled coupes, with most of the loss from fuel >7cm in diameter. However there was no evidence in the latter study of loss of soil carbon. Rab (1996) showed that while there was a significant decrease in total carbon concentrations after high intensity fire on one logging coupe, on a second there was no effect and there was no significant reduction in soil carbon after moderate or low intensity regeneration burns.

Few studies have assessed the effect of mechanical disturbance on soil carbon. Rab (1996) assessed the effects of mechanical disturbance and slash burning on soil physical and chemical properties across four logging coupes in the Central Highlands in Victoria disturbed mechanically after harvest. The results showed no change in organic carbon after litter disturbance only, but significant declines in the concentration of organic carbon in the top 10 cm of soil can occur as a result of redistribution of organic matter during mechanical disturbance of the topsoil or subsoil.

Landuse change

Land-use change can have a significant impact on carbon stocks. Land-use change includes: conversion of old-growth forest to periodically harvested native forest or plantation, afforestation of cleared land or unstocked forest, and permanent clearing of previously forested land. Because carbon stocks are generally greatest in old growth forest, conversion of this forest type may result in a relatively large impact on carbon stocks (on a per hectare basis). Further, because biomass accumulation on a given area of land is partly a function of time, conversion of old growth forest to short rotation plantation will usually have a greater impact on carbon stocks than conversion to a longer rotation plantation or a periodically harvested native forest.

Accounting for the effect of land-use change is important, not just for estimating emissions associated with vegetation removal, or accumulation of carbon as result of new plantings, but also for estimating the effects of past changes on current stocks. This is because the past land- use influences the amount of carbon stored in soil and debris, which can take many years to stabilize after the actual conversion has taken place. Thus, total carbon stocks and emissions from areas that were previously old-growth are likely to be greater than that for regeneration on areas that were previously younger regrowth.

Page 306 of 365

Change in land use likely to have a significant effect on forest carbon stocks and emissions in Tasmania include conversion of native forest or plantation to cleared land; cleared land to forest or plantation; native forest to plantation; and old-growth native forest to regrowth forest.

Conversion of native forests

Between 1996 and 2011, there was a decrease in the area of Tasmanian native forest communities of 137,000 ha or 4.7% of the total forest estate (FPA 2011). Since 2001, around 80% of this decrease was a result of conversion to plantations while the remainder was a result of conversion to non-forest (i.e. cleared agricultural land, roads or dams). Conversion of public native forests to plantations ceased in 2007. The Permanent Native Forest Estate Policy states broad-scale conversion of private native forest must end by January 2015 or when a threshold of 95% of the area of native forest in 1996 is reached (whichever is the earlier, FPA 2011).

As at June 2011, 4.7% of the native forest present in 1996 (with an original area of 2.91 million ha) had been cleared. Thus, a further 8,670 ha of private native forest could potentially be converted prior to 2015. Changes in land use since 1996 are provided in the Sustainability Indicators for Tasmanian Forests for 1996-2001 (Anon, 2002) and for 2001-2006 (Anon, 2007). These figures show areas for new plantation establishment, areas of forest burnt, the amount of native forest harvested and regenerated or converted to plantation, the area available for future harvest and the area of reserves in various forest groups for State Forest and Private Native Forest. In addition, the National Plantation Inventory, which has been reported since 1993, provides information on the annual area of different plantation species established across Tasmania and the current age class distribution (in 5 year aggregates (Parsons et al., 2007; Gavran and Parsons, 2011).

Information relating to land-use change prior to 1996 in native forests can be determined from the total area of stands established each decade from 1960-2000 and rates of harvesting from Forestry Tasmania Annual Reports. In addition, DCCEE has provided spatial layers of forest extent from 1990 to 2008. These layers allow changes in plantation and native forest extent over time to be verified and allow any expansion of native vegetation (> 2 m) onto cleared land to be identified.

7.8 EFFECTS OF FIRE ON FOREST CARBON

Fire frequency and intensity are key factors which need to be considered in any assessment of potential carbon sequestration. However, despite many years of research, there remains considerable uncertainty in regard to the longevity of different species and vegetation communities in the absence of fire as well as the “normal” fire regimes for different vegetation communities.

There has been ongoing debate over whether the location and distribution of different vegetation communities is relatively stable over time as a result of relatively stable fire frequency (e.g.

Page 307 of 365

Mount 1979), or can change as a result of chance variation in fire frequency (e.g. Jackson 1968). New research suggests the spatial distribution of many vegetation communities is relatively stable over time as a result of positive feedbacks between vegetation communities, fire and soil fertility (Wood and Bowman, 2011; Wood et al., 2011).

Like harvesting, fire influences forest carbon stocks in a number of ways: • combusting a proportion of vegetation and debris components resulting in direct emissions of CO2 and other GHGs; • removing vegetation components, converting them from living biomass to debris which then decompose over time; • killing entire trees, halting carbon sequestration and converting any components, not combusted, from living biomass to debris; • converting decomposable biomass to charcoal which is relatively inert (i.e. does not decompose over time); and • encouraging regeneration, resulting a multi-aged stand (where only a portion of dominant trees are removed) or a single aged stand (where all dominant trees are killed) and changing the rate of carbon sequestration.

History & frequency of fire

Fire extent across Tasmania from 1945 to 2000 for both wildfires and managed fires reported by Gould and Cheney (2007). This includes data for prescribed burning back to 1970 and silvicultural burning (i.e. CBS burning) back to 1965. The data shows total areas burnt each year and these figures can be used to estimate frequencies of large and small fires general frequencies of years with larger and smaller areas burnt by wildfires.

Years in which large areas are burnt (e.g. > 50,000 ha) are likely to correspond with years with larger, more intensive wildfires. On average, these appear to occur around once every eight years and burn an average of around 75,000 ha, although there appears to have been a decrease in both frequency and extent over the past 40 years. This decrease may correspond with increasing capacity to control wildfires, combined with an intensive program of prescribed burning during the 1980s and early 1990s.

In contrast, smaller fires occur almost every year with an average of 16,000 ha burnt. There is no information provided as to what vegetation types were burnt or whether the wildfire extent relates to both grassland and forest. Further, neither Forestry Tasmania nor DPIPWE collect fire data for fires on private land (except where these spread into forested areas). Thus, the figures include only a proportion of fires on cleared agricultural land.

Marsden-Smedley (1998, 2007) and Johnson and Marsden-Smedley (2002), collected historic data on fires in world heritage listed areas across Tasmania. This focussed primarily on the effects of fire on button grass plains. However, it also includes information relating to other

Page 308 of 365

vegetation types in these regions. In particular, the reports contain important historic information on fires as far back as the 1850s. In large fires in 1850 and 1897, there is evidence that up to 30% of the region was burnt. Other major fires occurred in 1934 and 1967. Fire extent back to 1930 and covering most Parks and Reserves in the state has been digitised and has been provided to the Carbon Study (Marsden-Smedley, Pers. Comm.).

Other work on fire frequency has been conducted by Hickey et al. (1999), Nicholls and Lucas (2007) and Von Platen (2011) based on fire scars in forests in the Warra region in southern Tasmania and dry eucalypt forests in south-eastern Tasmania. More general work on the relationships between vegetation types, topography and fire frequencies in Tasmania has been reported by Jackson (1968), Bowman and Jackson (1981), Pyrke and Marsden-Smedley (1998, 2007) Wood and Bowman (2011) and Wood et al. (2011).

Modelling fire frequency & probability

If we assume that the current mix of vegetation communities in unharvested forest will be maintained across the landscape into the future through periodic natural disturbance, then it is important to determine the appropriate fire regime required for this to be achieved. This requires a detailed understanding of past fire regimes and the way in which they have influenced the development of the vegetation structures that we see today.

Jackson (1968) provided estimated frequency distributions for fire regimes in different forest types, indicating the likely change for a given community under a specific fire frequency. More recent work by Wood et al. (2011) indicates that eucalypts in mixed forest may survive for over 500 years. Thus, it is likely that these frequency distributions could be extended. However, they provide a useful guide for the application of probability-based fire regimes for different forest types. The approach also allows factors such as the variation in fire risk for communities of different ages to be modelled.

Based on this and other work on fire frequency in different forest types in Tasmania (e.g. Brown and Podger 1982, Hickey et al., 1999, McCarthy et al., 1999, Jackson 1968, Pyrke and Marsden-Smedley 1998, 2007) we can provide an indication of the likely return frequencies of high intensity and low intensity wildfires in different forest types together with potential variability (Table 60).

Page 309 of 365

Table 60: Possible frequencies of high intensity (stand replacing) and low intensity (understory replacing) fires in various vegetation communities in Tasmania based on data from Jackson 1968, Pyrke and Marsden-Smedley 2007 and Turner et al., 2009).

Fire Scrub Low Open Tall Open Mixed Forest Rainforest Intensity Forest Forest

Avg. SD. Avg. SD. Avg. SD. Avg. SD. Avg. SD. High 50 30 120 40 240 60 350 80 500 120 Low 10 5 20 10 40 20 70 35 125 50

There are a number of important considerations associated with modelling the potential effect of fire on forest carbon stocks: • fires tend be relatively random in their occurrence and extent and so their effects are better modelled using probability functions rather than set time intervals; • mature forests span a wide potential age distribution (i.e. from 110 to 500 years) making it difficult to nominate a year when a forest will be burnt under a given scenario; • large wildfires are rare, but can affect very large areas of land (e.g. the 1898 wildfire burnt ~2 million ha) making scenarios that assume a set proportion of each forest class burnt each year potentially unrealistic; and • carbon losses and changes for different carbon pools will vary depending on fire intensity, land use and management operations following fire (e.g. salvage logging).

Random fire events can be modelled using a probability distribution for fire frequency (e.g. McCarthy et al., 1999). This distribution can be used to model the effects of the full range of fire frequencies within the distribution and calculating the average carbon stock together with the final variability from the probability of a particular frequency and the associated carbon stock.

McCarthy et al. (1999) also demonstrated how the use of a probabilistic fire model can assist with modelling potential age class distributions in eucalypt forest. This allows the cumulative probability of a stand being burnt in any year to be estimated. Initially, there would be a relatively low probability. However, as the time since fire increases, the chance of the stand being burnt gradually increases until very few unburnt stands remain (e.g. 500 years for E. regnans).

The relatively low frequency of severe wildfires can be partly accounted for by aggregating the figures for each decade. Since severe fires occur around once per decade, decadal variability in fire incidence should be lower than annual variability. Assuming a constant area is burnt each decade may therefore be justified for some forest types. This assumption can be checked by assessing the extant of different forest types burnt each year by overlaying long term data on fire frequency with forest type.

Page 310 of 365

Interactions between fire frequency and forest age & management

The early work conducted by Jackson (1968) and Bowman and Jackson (1981) indicates that fire risk is greater in young (< 50 years) sclerophyll forest, mixed forest and Rainforest compared with older stands (50-100 years). This is consistent with early reports of young regeneration and thinned stands frequently being burnt as a result of the increase in harvest slash (Elliot et al., 2009). Subsequent work by Barker (1990) also showed that fires risk in regenerating Rainforests was greater than in older stands.

Lindenmayer et al. (2009) provided evidence of an increase in risk of wildfire in regeneration from harvested stands of tropical Rainforest and other wet forests compared with unharvested forest. This increase was associated with species composition with a larger proportion of more flammable species present in younger stands, a changed microclimate and altered stand structure. In contrast, there appeared to be a reduced risk of fire in regenerating stands of dry sclerophyll forests after harvest. It was concluded the effects of harvesting on fire risk were contingent on forest practices, forest type and the natural fire regime. This work has major relevance to the Carbon Study as it supports the model proposed by Jackson (1968) in which fire risk can be increased in younger stands.

The potential effect of changes in stand age on fire risk can be modelled using the probabilistic approach already mentioned. Where specific data is available indicating an elevated risk associated with forest management (e.g. in recently thinned regrowth) this can be incorporated.

Effect of fire on forest type

The timing and intensity of fire influences not just the forest type, but also its structure. For example, moderate intensity fires in a single aged stand can kill some trees and allow new regeneration, resulting in a multi-aged stand. An absence of fires may allow the development of a Rainforest understorey, in wet forests, or an understorey comprised of Acacia Spp., Olearia Spp. and Pomaderris Spp. in drier forests.

The potential pathways for the development of different vegetation structures and communities under different fire regimes may be modelled based on the work by Jackson (1968), Mount (1979), Jackson and Bowman (1981) and Wood (2011). An example of the potential pathways for a tall open forest is illustrated in Figure 102. The forest types shown are based on the various forest classes used by Forestry Tasmania. A model such as this potentially allows the effects of timing and intensity of fires on forest type, structure and age class distribution to be estimated.

Page 311 of 365

Time period

0 years Non-eucalypt Eucalypt forest regeneration

50 years

Eucalypt Eucalypt regrowth regrowth with 100 regeneration years

Mature Mature 200 eucalypt forest eucalypts with years regeneration

Mature eucalypt forest 300 with regrowth years

Mixed forest 400 years

Rainforest 600 years

Figure 103: Example of potential paths of development of Tall Eucalypt Forest and Rainforest. Green lines indicate growth, red lines indicate effect of stand replacing fire and yellow indicate the effect of non- stand replacing fire.

Page 312 of 365

Effect of fire intensity on carbon stocks

The effects of fire on carbon stocks are likely to vary with intensity, and the type and amount of material present in each pool. High intensity fires could be expected to remove more carbon from all pools than low intensity fires, while fires in forest with large amounts of dry debris are likely to be more intense than those where there are fewer debris or where moisture contents are greater.

There are few published studies on the effects of fires of different intensities on carbon pools in forests. In Oregon, USA, Bormann et al. (2008) found that high intensity fire resulted in so much soil loss (~127 t/ha) that changes in soil carbon were difficult to assess, but estimated that 60% of the total carbon to 65 cm had been combusted. In a study from California, Hurteau et al. (2011) reported that a wildfire which killed 97% of trees, combusted 25% of the total ecosystem carbon (30% of the carbon in trees and debris), and moved 70% of the remaining carbon (65% of the total carbon in trees and debris prior to burning) to the debris pool.

In Australia, Norris (2010) provided figures that indicate that soil carbon concentrations may be halved in Jarrah forest after either low intensity or high intensity fires, but did not estimate the effect on carbon mass. The article indicated that soil charcoal was significantly reduced by low intensity fires but not by high intensity fires. McIntosh et al. (2005) proposed that greater fire frequency in dry eucalypt forest compared with wet eucalypt forest in Tasmania was the main factor contributing to the far lower carbon (~50% in the surface 13 cm for the two sites sampled) and nutrient concentrations in the former.

7.9 EFFECTS OF CLIMATE CHANGE ON FOREST CARBON

Climate change and elevated atmospheric CO2 levels can be expected to influence future forest growth and carbon stocks. Increasing surface temperatures as a result of climate change can be expected to increase evaporation and change rainfall patterns. Across southern Australia, summer rainfall is expected to increase while total rainfall may decrease or stay the same. Elevated atmospheric CO2 concentrations are expected to increase growth of some species as a result of the CO2 fertilisation effect. However, the potential effect on growth and interaction with other factors such as water and nutrient availability is uncertain.

Dean and Wardell Johnson (2010) summarised work on the potential effects of climate change on forest carbon stocks as follows: • increased decomposition rates and hence C efflux; • more frequent fires; • changes vegetation that could affect soil organic carbon (SOC); • increased moisture stress for some vegetation communities reducing productivity; • reduced frost risk for other vegetation communities;

Page 313 of 365

• increased risk of disease, insect attacks, weeds and storms reducing productivity; and • Increased or decreased plant growth associated with elevated CO2.

A key factor in predicting the effects of climate change on forest carbon stocks is the actual trend in future GHG emissions and their predicted impact on climate. A variety of future GHG emission scenarios have been modelled by the IPCC (2000) in the Special Report Emissions Scenarios (SRES). These provide GHG emission projections under a range of different emission mitigation and technological scenarios and form the basis for modelling future climate change and atmospheric CO2 concentrations.

Current & predicted future climate

Over the past 40 years there has been a trend of increasing average temperature and declining rainfall across south-eastern Australia. This trend is reflected in Tasmania where temperature has increased at a rate of 0.05-0.15oC per decade and rainfall has decreased by 30-50 mm per decade. The change in temperature has been most pronounced in the north-east of the state with little change along the west coast, while the change in rainfall has been most pronounced in the northwest. These observations are consistent with future climate projections for the state.

Future climate projections for Tasmania based on averages of 22 separate climate models for different emission scenarios are available from the Bureau of Meteorology. The average projections from these models for 2050 indicate a 0.6-1.5oC increase in temperature and a 2-5% reduction in rainfall across most of the state under all emission scenarios compared with the average for 1980-1999.

The Climate Futures for Tasmania study has produced a series of reports on predicted changes in Tasmania’s climate based on modelled future projections associated with alternative global emission scenarios (e.g. Corney et al. 2010 Grose et al. 2010; White et al. 2010). Two emissions scenarios from the IPCC (2000, B1: low emissions and A2: high emissions) were selected for the study with climate projections for each Baseline Sensitivity Scenario based on outputs from 6 global climate models. These projections were downscaled from the original output resolution of 200-300 km to a grid area of approximately 10 x10 km for the studies (Corney, et al. 2010).

This study showed that, average temperatures across Tasmania could be expected to increase over the 21st century by 1.6oC under the low emission scenario to 2.9oC under the high emission scenario (ACE CRC, 2010). Although there was predicted to be no major change in total average rainfall across Tasmania, there were significant differences in regional patterns and seasonality. There was predicted to be an increase in winter rainfall and a decrease in summer rainfall on the west coast, a reduction in rainfall for every season on the Central plateau and an increase in autumn and summer rainfall in the northeast. In general, there was a pattern of increasing rainfall over coastal regions and a reduction in rainfall over central Tasmania and in the north-west.

Page 314 of 365

Although the reports covered potential impacts of climate change on agricultural and the frequency of extreme climatic events, they did not consider forestry or predict changes in risk of wildfires (Holz et al. 2010, White et al., 2010).

Effects of climate change on growth of Tasmania’s forests

Two studies have modelled the effect of climate change on forest productivity across Tasmania. Both studies focussed on growth of plantations, although one included predicted effects on growth and wood production from native forests.

The first study, undertaken by the Forestry CRC for Forest and Wood Products Australia (FWPA), used the growth model CABALA in conjunction with climate projections from two global climate models for three emission scenarios (Battaglia et al. 2009). This study found that tree growth was likely to increase across most of Tasmania, both as a result of expected increases in temperature and rainfall as well as in response to elevated CO2 concentrations. The second study used the 3PG2 growth model in conjunction with climate projections from up to 23 global climate models and two emission scenarios (ABARES 2012b). This study found that the growth of radiata pine plantations was likely to decrease by around 10% as a result of reduced rainfall, while the growth of eucalypt plantations and native forests could increase slightly. The interaction between elevated CO2 and climate change on tree growth was not modelled in this study. However, it was considered that the effect of CO2 could potentially partially, or fully, offset the modelled declines.

Predictions for effects of climate change and CO2 (Battaglia et al., 2009)

The study by Battaglia et al. (2009) used climate projections for 2030 and 2070 for three greenhouse gas (GHG) emission scenarios (A1F1: very high emission, A2: high emissions and B1: low emissions) from two climate models: Hadleys Mk2 (with high climatic sensitivity to elevated CO2, used for the A1F1 and B1 emission scenarios) and CSIRO Mk 3 (with low climatic sensitivity to CO2, used for the A2 emission scenario).

Climate data projections were downscaled to a resolution of 0.05 degrees using historic climate data from the Bureau of meteorology’s data drill (http://www.longpaddock.qld.gov.au/silo/).

The study analysed the effects of changes in climate alone (i.e. assuming no CO2 fertilising effect) as well as the effects of both climate change and rising CO2 levels on plantation growth across Australia at a resolution of 5 km2. It included eight Pinus radiata sites, 10 Eucalyptus nitens and seven Eucalyptus globulus sites located in Tasmania (change in growth for these sites shown in Table 61).

Detailed maps of expected changes in plantation growth across Tasmania (excluding regions with native forests) were produced allowing the potential variability in growth across the landscape to be assessed.

Page 315 of 365

Table 61: Estimated average changes in growth for E. globulus (G), E. nitens (N), and P. radiata (P) in 2030 and 2070 with and without CO2 fertiliser effect.

Scenario No CO2 fertilisation effect CO2 fertilisation effect

2030 2070 2030 2070

G N P G N P G N P G N P

1 18% 6% NA 37% 6% NA 36% 10% NA 72% 13% NA

2 22% 8% 10% 29% 12% 14% 40% 14% 14% 69% 20% 17%

3 21% 5% NA 45% 8% NA 46% 8% NA 98% 13% NA

Predictions of effects of climate change (ABARES, 2012b)

The ABARES (2012b) study was based on average climate projections for 2030 and 2050 from 22 climate models for two GHG emission scenarios: a low emission scenario (A1b) and a high emission scenario (A2). The forecast changes in global temperatures for the two scenarios are: • A1B: +0.9 (+0.54 to +1.44) by 2030 +1.5 (+0.92 to +2.45) by 2050 • A2 +0.8 (+0.48 to +1.28) by 2030 +1.4 (+0.84 to +2.24) by 2050

Climate data was downscaled from the original model projections to a resolution of 25 km for the purpose of modelling growth.

Average productivity for 2050 was estimated to increase by 3-8% for E. globulus, but decrease by 10-11% for P. radiata. The ABARES study also predicted effects of climate change on growth of native forests based on modelled changes in growth rates of E. globulus sawlog plantations (ABARES 2012b).

Analysis of climate scenarios for the Carbon Study

The results of Battaglia et al. (2009) and ABARES (2012b) provide general indications of the ranges in future growth rates of Tasmanian plantations under future climate scenarios and CO2 emission scenarios. These studies demonstrate that there is still a large amount of uncertainty regarding both future climate trajectories and the impacts of climate change on tree growth. However, they represent the best information available of the likely future changes.

Results from these studies can be used as the basis for future scenarios that predict the effects of climate change on carbon stocks in plantations and native forests.

Page 316 of 365

Effects of climate change on fire regimes Other studies have assessed effects of climate change on fire risk across south-eastern Australia. These show increased temperatures and reduced rainfall arising due to climate change can be expected to increase the frequency and intensity of forest fires across southern Australia (e.g. Beer & Williams, 1995; Hennessey, 2005; King et al. 2011, Williams et al. 2001). However, few have looked specifically at Tasmania. Hennesey et al. (2005) found that an increase in fire-weather risk was likely at most sites across south-east Australia by 2020 and 2050. Tasmania, however, was expected to be largely unaffected. Across the two sites covered in Tasmania (Hobart and Launceston) projected increases in temperature were largely offset by increases in temperature and humidity. As a result, the Forest Fire Danger Index (FFDI) was expected to increase by only 1-2% at Hobart and by 3-17% at Launceston by 2050. An increase in fire frequency is expected to reduce carbon stocks. A case study of forests near Canberra (King et al. 2011) showed increased fire frequency and intensity associated with a warmer drier future climate could reduce forest carbon stocks by 4-8% over a 180 year period. At the same time, net carbon sequestration rates were reduced as mature vegetation was replaced by younger post-fire regeneration. In addition to the direct impacts of climate change, Noble (1999) suggested an increase in fire frequency in Tasmanian forests could result in an increase in cover of flammable species, further increasing risk of fire.

Effects of climate change on carbon stocks in native forests

Modelling effects of climate change on native forests presents a difficult challenge. On one hand, the complex associations between plants and animals in these ecosystems mean that the effect of major long-term changes in climate on one key species can have unpredictable and catastrophic effects on others (e.g. the effect of bark beetles in forests in western North America (Bentz et al., 2010). On the other hand, the diversity of these ecosystems can increase their resilience and provide improved capacity to adapt to change (Mackey, 2008). Increased CO2 concentrations and higher temperatures could have a variable impact on ecosystems.

Dean & Wardall Johnson (2010) estimated current and future stocks of soil carbon in old-growth forests located in south Western Australia and Tasmania under different climate change scenarios. The scenarios derived from OzClim (2007) were based on four models with low climate sensitivity ECHAM5/MPI-OM, Miroc-M, CSIRO Mk3.5, and ECHO-G.

The effects of climate change on soil organic carbon (SOC) were calculated using a temperature and rainfall dependent formula for SOC (Bird et al. 2002) to show the relative change between 2000 and 2100 under different climate scenarios. Modelling with FullCAM indicated soil organic carbon did not change over time in tall open forests (average change +0.4% to -2.1%) but decreased by 10-12% in Rainforests in Tasmania. The latter loss represent a loss of around 22- 26 t C/ha. However, based on a model of changes across the landscape, potential losses of soil carbon from tall open forests were estimated to range from 5-10% (13-27 t C/ha).

Page 317 of 365

8 COMPARISONS OF MIXED EUCALYPT FOREST AND RAINFOREST

A key characteristic of Tall Eucalypt Forests in Australia is that they require fire to regenerate and, in the absence of fire, or other major disturbance, they eventually change into other forest systems such as Rainforest (Ashton, 1981; Attiwill, 1994; Bowman and Jackson, 1981; Gilbert, 1959; Jackson, 1968). Species such as Eucalyptus regnans, or E. delegatensis, are shade intolerant and often do not regenerate successfully in established forests. In contrast, Rainforest species such as Northofagus cunninghammii and Atherosperma moschatum can and do regenerate in shaded conditions. Rainforests are able to grow and regenerate indefinitely in the absence of fire. Thus, if fire is absent for a period of time beyond which eucalypts produce viable seed (350-500 years) then a tall open forest stand may be replaced by Rainforest.

Vegetation type and structure is important in determining carbon stocks. Typically, stands with large old trees will contain more carbon (in live biomass) than those with smaller or younger trees (e.g. Keith et al. 2009). Eucalyptus regnans is the tallest flowering plant in the world. Thus, mature stands of this species contain some of the largest carbon stocks of any ecosystem. In contrast, Rainforest trees are shorter and tend to be less massive. Thus, it is likely that above ground carbon stocks are smaller in Rainforest than in the eucalypt stand that may have preceded it.

Unfortunately, there is little data available on carbon stocks in pure temperate Rainforest in Australia and no studies have attempted to compare carbon stocks in a time series of stands covering forest successional stages. For this reason, a series of field measurements were undertaken as part of the Carbon Study. Additionally, relevant data-sets, generously provided by the Tasmanian Wilderness Society and Ian Riley of University of Tasmania, were analysed.

8.1 COMPARISON OF PAIRED EUCLYPT FOREST/RAINFOREST PLOTS

Study site

The study site was located in the Styx/Florentine valley in central Tasmania (Figure 103). The site was selected on the basis of having areas of fully stocked highly (FC1), or moderately highly (FC3) productive mature eucalypt forest adjacent to M+ (FC 85) or M- (FC 86) Rainforest. A series of paired plots were established within the mature eucalypt forest and the adjacent Rainforest (as mapped by Forestry Tasmania). Plot locations were selected by first importing a forest class map into Google Earth and selecting points that were in appropriate adjacent forest classes (i.e. Rainforest and mature forest) and which were readily accessible (< 100 m from the nearest road). Using Google Earth, a transect was drawn between the plots comprising each pair to ensure that the difference in elevation was < 50 m. The grid coordinates of each point were recorded to ensure that the plots would be randomly located within the selected area.

Page 318 of 365

Altogether 16 paired plots were selected at locations ranging from the Styx River valley to near Lake Gordon. These included 6 plots within FC1, 2 plots within FC3, 4 plots within FC85 and 4 plots within FC86.

Measurements

Plots consisted of two concentric circles. One with a 20 m radius within which all trees >40 cm DBH ob (over bark) were measured and another of 7m radius within which all trees >15 cm were measured (Figures 104 and 105). The species of each tree was recorded. Heights of the tallest two trees in the eucalypt overstorey and Rainforest understorey were recorded where present, and the heights of all stags were measured. A 20 m transect line was located within each plot and the diameter of all coarse woody debris that it crossed were measured. In addition, the decay class of each piece of wood was identified according to Forestry Tasmania’s decay classifications.

Figure 104 Location of measurement plots in the Styx and Florentine Valleys. Purple and blue shaded regions indicate areas of Tall Eucalypt Forest (Forest Classes 1: dark blue or 3: light blue) and Rainforest (Forest Classes 85: dark purple and 86: light purple) from which plots were selected.

Page 319 of 365

Figure 105: Measurement of a large E. obliqua in one of the eucalypt plots in the Styx valley.

Figure 106: Measurement of a coarse woody debris in a Rainforest plot dominated by N. cunninghammii and A. moschatum in the Styx valley.

Page 320 of 365

Calculations

The Farm Forestry Toolbox was used to estimate the volumes of all eucalypts and an algorithm for Northofagus cunninghammii was used to estimate volumes of Rainforest species. Because the Farm Forestry Toolbox requires height to estimate stem volume, a height diameter relationship was developed to estimate the height of trees whose heights were not recorded (Figure 106).

Volumes of stags were estimated using the Farm Forestry Toolbox by estimating their original height based on the height diameter relationship for live trees. From this original stem volume, the toolbox’s stem apportioning tool was used to select only that portion of the stem remaining (based on its measured height) and assuming that bark thickness was zero.

y = 12.893x0.2649 60 R² = 0.4156

Ht (m) 30

y = 8.3469x0.2793 20 R² = 0.345 Eucalypts

10 Rainforest

Wattle 0 0 50 100 150 200 250 300 DBH (cm)

Figure 107: Relationships between diameter and heights of eucalypts and Rainforest species.

Page 321 of 365

Coarse woody debris volume was estimated using the line transect intersect formula (DeVries 1979):

Equation 32 where is the volume of debris in m3/ha, is the intersect diameter in cm, and is the length of the transect in m. The density of coarse woody debris was based on its estimated decay class based on results from Grove et al. (2009). These were as follows:

• 1: (hard) 488 kg/m3 • 2: (medium) 362 kg/m3 • 3 (medium soft) 284 kg/m3 • 4 (soft) 180 kg/m3 Wood density of Stags was assumed to be 400 kg/m3 and all biomass was assumed to have a carbon content of 50%

Analysis

The difference in total volume, mass and carbon content of live trees, stags and coarse woody debris of paired plots was compared using Student’s T-Test for paired plots. In addition, eucalypt basal area was regressed against each variable to identify significant trends.

Results

Stand descriptions

Trees measured ranged in diameter up to 260 cm DBH over bark for eucalypts and to 168 cm for Rainforest species (N. cunninghammii). The tallest eucalypt measured was 55 m, while the tallest Rainforest tree measured was 35 m. The Rainforest plots were characterised by having a dense canopy and were noticeably darker than the eucalypt plots. Occasionally there were remnant eucalypts in the Rainforest plots and these were measured because they were representative of the mapped Rainforest Forest Class.

Page 322 of 365

Live trees

The average biomass and carbon content in live trees on the plots was 746 t/ha and 373 t C/ha respectively. The total standing volume, biomass and carbon content of live trees was significantly greater (P < 0.01) on eucalypt plots than Rainforest plots. On average, Rainforest plots (462 t/ha) had 44% of the live biomass of eucalypt plots (1,030 t/ha, Figure 107). Rainforest species comprised 15% of the total live volume and biomass on eucalypt plots while remnant eucalypts comprised 12% of the live total on Rainforest plots. There were no Rainforest species present on the FC3 plots with an E. obliqua overstorey. The FC1 plot overstories were mainly E. regnans, but one was E. delegatensis.

There were no significant differences in live biomass between the eucalypt forest classes (FC1 and FC3) or Rainforest forest classes (FC85 and FC86, Figure 108). There was a significant positive relationship (Figure 109) between the basal area of live eucalypts and the total biomass of all trees (R2 = 0.88, P <0.01). However, there was a significant negative relationship between the basal area of eucalypts and the biomass of Rainforest trees (R2 = 0.53, P < 0.025). This relationship indicates that total biomass is strongly dependent on eucalypt basal area. Where no live eucalypts were present, the intercept of the line of best fit indicated that Rainforest biomass may be around 400 to 600 t/ha or around a third of that of eucalypt dominated plots when the last eucalypt has died.

1400 Rainforest 1200 Eucalypts 1000

800

600

400

Total Live Biomass (t/ha) LiveTotal Biomass 200

0 Eucalypt Rainforest Forest Class and Type

Figure 108: Average total volume eucalypt and Rainforest trees across plots.

Page 323 of 365

1400

1200 Rainforest Eucalypts 1000

800

600

400

Total Live Biomass (t/ha) LiveTotal Biomass 200

0 1 3 85 86 Eucalypt Eucalypt Rainforest Rainforest

Forest Class and Type

Figure 109: Average total volume for individual forest classes.

1600

Total 1400 Rainforest 1200

1000 y = 6.536x + 424.46 R² = 0.8767 800

600

Total biomass (t/ha) biomass Total 400 y = -2.8535x + 413.99 R² = 0.5313 200

0 0 50 100 150 Basal area of eucalypts (m2/ha)

Figure 110: Relationships between basal area of live eucalypts and total biomass of all trees and biomass of Rainforest species only.

Page 324 of 365

Stags

Although biomass of dead trees and coarse woody debris tended to be greater on eucalypt dominated plots compared with Rainforest plots, this difference was not significant (P = 0.333). The average mass of stags present was 46 t/ha (Figure 110). There was no relationship between Stag mass and eucalypt basal area.

120 Rainforest 100 Eucalypts

Total Stag (t/ha)Total Biomass 20

0 Eucalypt Rainforest Forest Class and Type

Figure 111: Biomass of stags on eucalypt and Rainforest plots.

Coarse woody debris

Like stags, there appeared to be a greater amount of coarse woody debris on eucalypt plots than Rainforest plots, but this difference was not significant (Figure 111). The average mass of CWD on eucalypt plots was 210 t/ha, while, on Rainforest plots, it was 96 t/ha. There appeared to be less variation in biomass on Rainforest plots than eucalypt plots. Most of the variation between eucalypt plots was due to one plot with the equivalent of 1000 t/ha. Without this plot, the mass of coarse woody debris on eucalypt plots (84 t/ha) was slightly less than that on Rainforest plots.

Combined biomass

Total live and dead biomass was significantly less on the Rainforest plots than eucalypt plots (P < 0.01, Figure 112). On eucalypt plots the average biomass was 1170 t/ha, and on Rainforest plots was 590 t/ha. The average breakdown of biomass consisted of: Live trees: 84%; Stags: 5%; Coarse woody debris: 11%.

Page 325 of 365

600

500

400

300

CWD (t/ha) 200

100

0 Eucalypt Rainforest Forest Class and Type

Figure 112: Biomass of coarse woody debris on eucalypt and Rainforest plots.

1600 CWD 1400 Stags Rainforest 1200 Stags Eucalypts Live Rainforest 1000 Live Eucalypts 800

600

400 Total Biomass (t/ha) Biomass Total

200

0 Eucalypt Rainforest Forest Class and Type

Figure 113: Total live and dead biomass on eucalypt and Rainforest plots.

Page 326 of 365

There was no significant difference between total biomass in different eucalypt or Rainforest forest classes (Figure 113). There were some interesting trends, with the total mass of eucalypts the same for both eucalypt classes (FC1 and FC3), but with the presence of Rainforest species on FC1 providing additional biomass. Across the two Rainforest Forest Classes (FC 85 and FC 86) there was no difference in the amount of live biomass even though we would normally expect greater biomass on FC 85 (dominant height > 35 m) than on FC 86 (dominant height < 35 m).

1600 CWD 1400 Stags Rainforest 1200 Stags Eucalypts 1000 Live Rainforest Live Eucalypts 800

600

400 Total Biomass (t/ha) Biomass Total 200

0 1 3 85 86 Eucalypt Eucalypt Rainforest Rainforest Forest Class and Type

Figure 114: Total live and dead biomass for individual forest classes.

Comments on assessing forest carbon

Average total carbon stocks varied from 295 t/ha for Rainforest to 585 t/ha for eucalypt forest (Table 62). These values are lower than those reported by Dean et al. (2011) and Keith et al. (2009), but are greater than estimates from Moroni et al. (2010) based on randomly located permanent sample plots within these forest classes. It is important to interpret the results of this study in terms of the context under which our plots were selected and region in which they were located. Although some attempt at randomisation was made, for the purpose of the exercise, it was important to ensure the plots were located in either mature eucalypt forest or Rainforest, and that they were at least 30 m inside the mapped forest class boundary. Thus, our plot selection was not truly random, and did not sample gaps which inevitably occur in mapped forest classes. Thus, the fact that measured biomass was greater than the reported average for each forest class across the whole of Tasmania is not unexpected.

Page 327 of 365

Table 62: Estimated total carbon stocks in Rainforest and eucalypt forest classes.

Type Total Carbon (t C/ha) Forest Live Trees Dead Trees CWD Total Cla ss Eucalypt Rainforest Total Eucalypt Rainforest Total Eucalypt 1 440 97 537 15 7 22 42 601 3 449 0 449 45 0 45 43 537 Avg 442 73 515 22 5 28 42 585 SD 152 81 134 26 9 24 32 52 Rainforest 85 47 187 234 4 5 9 34 277 86 13 215 228 14 9 23 62 312 Avg 30 201 231 9 7 16 48 295 SD 39 70 47 19 6 23 36 45 Avg 236 137 373 16 6 22 48 440 SD 238 99 176 23 8 23 32 47 It is also important to note that the definition of Rainforest used in this study reflects the classification system used by Forestry Tasmania and may differ from ecological classification of Rainforest. For example, no attempt was made to determine the age of the Rainforest species. However, the fact that dead eucalypts were present in most Rainforest plots indicates that the stands were probably < 600 years old. Thus it is possible that total biomass may continue to increase in these Rainforest plots over time. Bearing these limitations in mind, the study provides a useful indication of the potential relative difference in biomass and carbon contents between high productivity mature eucalypt forest and subsequent (young) Rainforest.

Page 328 of 365

8.2 ANALYSIS OF WILDERNESS SOCIETY DATA

The Tasmanian Wilderness Society offered to provide some of its data on sites measured around Tasmania. Two data sets were provided, one from the Tarkine Wilderness area and the other from the Styx Valley. A third data set was analysed by Dean et al. (2011) and results are included here for comparison.

Methodology

The general methodology used to measure the plots and calculate total biomass stocks is described by Dean et al. (2011). Four plots were located in Rainforest in the Tarkine region and five were located in the Styx Valley region. The plots were randomly located in each region by walking a predetermined distance in a set direction that had been randomly chosen. The Tarkine site consisted of almost pure Rainforest while the Styx and Florentine sites included a mix of large mature eucalypts and Rainforest species.

The plots consisted of three concentric circles of 60m, 30 m and 7m radius. In the 60 m circle all trees >100 cm DBH over bark were measured; in the 30m radius circle, all trees between 20 and 100cm were measured; and in the 7m radius circle, all trees < 20cm were measured. Species was recorded wherever possible. Heights and diameters were recorded for dead trees including stumps and stags. In addition, all coarse woody debris with a diameter > 10 cm was measured in the 20m plot. Measurements included log length and the diameters at both ends.

To calculate biomass from the individual tree diameters, five allometric equations developed by Dean et al. (2011) were used. These consisted of one for E. regnans, one for E. obliqua, one for E. delegatensis and two for Rainforest species. For E. delegatensis, the average of the results from the three eucalypt formulae was used. For E. nitida, the allometric relationship for E. obliqua was used. For Rainforest species, the average of two Rainforest allometrics reported in Dean et al. (2011) were used.

Results

The average results from the Tarkine and Styx sites are presented together with those from Dean et al. (2011) for the Florentine site (Figure 114). The total biomass in live trees varied from 1360 t/ha at Styx to 422 t/ha at Tarkine (the Rainforest only site, Table 63). Biomass in stags varied from 98 t/ha at Styx to 30 t/ha at Tarkine. Biomass of coarse woody debris varied from 280 t/ha at Styx to 116 t/ha at Tarkine. The Florentine site was mid-range for all results.

Page 329 of 365

2500

2000 Eucalypts Rainforest 1500 Live Total ground (t/ha)ground - Stags 1000 CWD Dead Total Total 500 Biomass Above

0 Styx Florentine Tarkine Site

Figure 115: Average biomass in various live ad dead components from the three sites. Results for Florentine site from Dean et al. (2011).

Table 63: Summary of results from the three sites Item Site Live Dead Total Eucalypts Rainforest Total Stags CWD Total Total Biomass (t/ha) Styx 1192 166 1358 98 280 378 1736 Florentine 680 258 938 49 257 306 1244 Tarkine 12 411 422 30 116 146 569 Percent breakdown Styx 70% 10% 80% 6% 14% 20% 100% Florentine 55% 21% 75% 4% 21% 25% 100% Tarkine 1% 76% 77% 5% 18% 23% 100%

Page 330 of 365

8.3 STUDY BY RILEY (2012), UNIVERSITY OF TASMANIA

As part of the Carbon Study, a report (Riley, 2012) was prepared by Ian Riley of the University of Tasmania, referencing preliminary results from a PhD study of changes in carbon stocks during forest succession of Tall Eucalypt Forests in the Styx and Florentine Valleys. While not definitive, the report provided useful information on differences in carbon stocks in vegetation, debris and soil of adjacent Rainforest and eucalypt communities.

The objective of this work was to quantify changes in standing biomass, coarse woody debris (CWD; >10cm diameter) and soil carbon with mixed forest-Rainforest succession, in large old growth forests.

Method

Site description

The study area was located in the southern forests of in the upper reaches of the Florentine, Styx, and Tyenna Valley’s (Table 64). The study focused on forests perceived to have large carbon stocks. Sites were paired and constrained by geography, geology and biology. A paired site represented one Mixed-forest site and one Rainforest Site within 500m horizontally and 200m vertically of one another and located on similar geologies.

Biologically, all sites were restricted to mature forms. Mixed forests sites contained eucalypts > 40m in height, from the same single cohort, with no regrowth present. Eucalypt abundance was 20-70% of live crown density and representative of a healthy, mature, mixed forest, with an understory community which matched that of the paired Rainforest site.

Rainforests sites were restricted to the communities found within the Thamnic and Callidendrous groups. Sites containing Anodopetalum biglandulosum (Horizontal) were removed. Rainforest sites had no live standing, and where possible, no dead standing or fallen eucalypts present. It was assumed, in the absence of any evidence of eucalypts at a Rainforest site, that all Rainforest sites had the potential to host eucalypts, and that their absence was due to the absence of fire. Sporadic Rainforest regeneration was acceptable. However, large-scale, Rainforest regeneration was considered to be evidence of fire and excluded.

Page 331 of 365

Table 64: Forest Class and Geographic Location of Identified Sites

Forest Class Site and Forest Type Forest Class No. PI Type Easting Northing F1 MX E1a&b 1 E1b.M- 459108 5270317 RF M+ 85 M+ 457769 5270964 F2 MX E2c&d 4 E2d.M+ 453100 5270650 RF M+ 85 M+ 453600 5271150 F3 MX E2c&d 4 E2d.M - 452500 5282000 RF M+ 85 M+ 452750 5281500 S2 MX E1c&d 2 E1c.M+ 467700 5258100 RF M+ 85 M+ 467850 5258100 T1 MX E2c&d 4 E2c.M - 461116 5261155 RF M+ 85 M+ 461050 5261150

Measurements

To measure standing live and dead biomass, six plots were established at each paired site, three in each forest type. Each plot consisted of two concentric circles, an interior one 10m in radius, in which the diameter and height of all trees greater than 10 cm in diameter at breast height (DBH) where measured, and an exterior one 20 m in radius in which the DBH and height of all trees greater than 40cm DBH where measured. Measurements were incomplete at the time of reporting with all Rainforest plots measured at two sites (F1 and S2), but only one eucalypt plot measured at each site. Stem volume was calculated using species specific allometric functions developed by Forestry Tasmania. Coarse Woody Debris were measured using the Line Intersect Method, measuring all downed dead and hung-up wood > 10 cm in diameter along a 300 m transect. The diameter of each log that intersected the transect was measured, the level of decay and species assessed, and the land slope and piece angle recorded. These measurements were then used in the Van Wagner probability formula to gain a per hectare level estimates of wood volume.

Soil carbon was measured for the organic layer (O horizon) and mineral soil (0-15 cm, 15-30 cm, 30-45 cm) from each plot. Bulk density and total soil carbon (Leco Analyser) were measured for each depth.

Page 332 of 365

Results

Standing biomass

Total standing volume was consistently greater on the eucalypt plots (average 2120 m3/ha) compared with Rainforest plots (average 820 m3/ha, Table 65). Even though the sample size was too small for this to be definitive, it is consistent with the measurements made by CO2 Australia at other similar paired sites in these forest types.

Table 65: Comparison of standing biomass volume among paired sites. Volume (m3/ha) Leather Site and Forest Type Eucalypt Myrtle Sassafras wood Site Total F1 MX 1283 273 220 42 1820 RF 41 750 260 0 1051 S2 MX 2110 122 196 0 2429 RF 0 436 162 0 598

Coarse woody debris and soil carbon

The average amount of coarse woody debris was similar for the eucalypt forest (598 +/- 224 m3/ha) and Rainforest (565 +/- 356 m3/ha) plots. This is an important result as it could be speculated that the mortality of eucalypts could result in increased amounts of coarse woody debris in areas succeeding to Rainforest. However, as with the study by CO2 Australia there was no evidence of this being the case.

Similarly, there was no significant difference in the amounts of soil carbon in the organic soil layer (89 +/- 58 t/ha for eucalypt plots and 84 +/- 91 t/ha for Rainforest plots) or mineral soil (92 +/- 11 t/ha for eucalypt plots and 85 +/- 15 t/ha for Rainforest plots). Thus, there is no evidence that soil carbon stocks are greater in Rainforest that in eucalypt forest. However, it should be noted that soil samples were collected at least 2 m from the nearest tree. Previous work has indicated that organic matter tends to accumulate close to the trunks of temperate Rainforest species and may be relatively shallow further away (Gilbert 1959). Thus, it is possible that these measurements may underestimate the total soil carbon in Rainforest. However, the same sampling protocol was used for both eucalypt and Rainforest plots, so the estimates are assumed to reflect the relative difference (or similarity) in total soil carbon between the two forest types.

Page 333 of 365

8.4 CONCLUSION

These studies suggest that the total carbon stocks in Rainforest biomass are lower than those for adjacent mature eucalypt forest. In contrast, there is likely to be little difference in carbon stocks in coarse woody debris and soil between the two forest types. If it is assumed that a) the plots are representative of mature eucalypt forest and Rainforest b) that eucalypt forest will eventually succeed to Rainforest in the absence of fire, these results indicate that carbon stocks could gradually decrease as a result of this transition.

Overall, the combined results suggest that, during eucalypt stand senescence, the effective live carbon content of Rainforest may be less than 50% of the preceding eucalypt forest. Both this relative difference and the estimate of total Rainforest biomass need to be confirmed with further measurements at other sites around Tasmania. However, for the purpose of modelling growth and decline of eucalypt forest, it appears reasonable to assume that the maximum biomass of the resulting Rainforest is around 50% of the eucalypt forest.

Page 334 of 365

BIBLIOGRAPHY

3GreenTree Ecosystem Services Limited (2012). Approved VCS Methodology VM0012 (Version 1.1, Sectoral Scope 14): Improved forest management in temperate and boreal forests (LtPF). Verified Carbon Standard, Washington. ABARE (2001) Australian forest products statistics: September quarter 2000 Summary. , ABARE, Canberra. ABARE (2009) Australian commodity statistics. Available at: http://www.abare.gov.au/publications_html/data/data/data.html ABARES (2012a) Australian forest and wood products statistics: September/December quarters 2011. 22 May 2012. ABARES (2012b) Potential effects of climate change on forests and forestry: technical report for Tasmania. Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra, January. ACE CRC 2010, Climate Futures for Tasmania general climate impacts: the summary, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania. AGO (1999a) National Emissions Trading: Establishing the Boundaries. Discussion Paper 1. Canberra: Commonwealth of Australia. AGO (1999b). National Emissions Trading: Issuing the Permits. Discussion Paper 2. Canberra: Commonwealth of Australia. AGO (1999c). National Emissions Trading: Designing the Market. Discussion Paper 4. Canberra: Commonwealth of Australia. AGO (2007) National Inventory Report 2005, Vol. 1, AGO, Canberra. Anon (1997) Tasmanian regional forest agreement between the Commonwealth Government and the state of Tasmania November 1997. Anon (2002) Sustainability Indicators for Tasmanian Forests 1996 – 2001. Prepared by the Tasmanian and Commonwealth Governments for the 2002 Review of the Tasmanian Regional Forest Agreement. Anon (2007) Sustainability Indicators for Tasmanian Forests 2001 – 2006. Prepared by the Tasmanian and Australian Governments for the 2007 Ten Year Review of the Tasmanian Regional Forest Agreement. ASRIS (2012) http://www.asris.csiro.au/index_other.html Accessed March 2012. Ashton, D. H. (1981) Fire in tall open-forests (wet sclerophyll forests). In Gill AM, Groves, R. H. and Noble, I. R., eds. Fire and the Australian biota. Australian Academy of Science, Canberra: Australia, 339 – 366. Attiwill, P. M. (1994) The disturbance of forest ecosystems - the ecological basis for conservative management. Forest Ecology and Management 63: 247-300. Australian Government. (2011). Carbon pricing mechanism: Who is liable? Clean Energy Future Fact Sheet 15, Canberra.

Page 335 of 365

Australian Liberal Party (2010). The Coalitions direct action plan: Environment and climate change. Liberal Party, Canberra: Commonwealth Australia Australian Treasury. (2011) Strong growth, low pollution: Modelling a carbon price. Commonwealth of Australia, Canberra. Baalman, P. & O’Brien, N. (2006). Crediting of carbon storage in reforestation projects under timber management. Australian Forestry 69 (2): 90-94. Barker, M. (1990) Effects of fire on the floristic composition and flammability of Rainforest vegetation. Tasforests 117-120. Battaglia, Bruce, J., Brack, C., & Baker, T. (2009) Climate Change and Australia’s plantation estate: Analysis of vulnerability and preliminary investigation of adaptation options. FWPA Project No: PNC068-0708. Beadle, C., Paul, K., Reeson, A., Bruce, J., & Battaglia, M. (2011) Carbon Plantations – A case study into management options. CSIRO, National Research Flagships, Sustainable Agriculture Beer, T., Williams, A. (1995) Estimating Australian forest fire danger under conditions of doubled carbon dioxide concentrations. Climatic Change 29: 169–188. Bentz, B. J., Regniere, J., Fettig, C. E., Hansen, M., Hayes, J. L., Hicke, J. A., Kelsey, R. G., Neron, J. F., & Seybold, S. J. (2010) Climate change and bark beetles of the forests of the western United States and Canada: direct and indirect effects. Bioscience 60: 602- 613. Bird, M.I., Wynn, J.G., Vellen, L., Cowley, J. & Zhou, Y. (2002) Soil carbon and carbon isotopes inventory from Australian sandy soils [unpublished report to the CRC for Greenhouse Accounting, p. 30]. Canberra: Research School of Earth Sciences, Australian National University. Board of Private Forests Tasmania (2005) Australian Government review of taxation treatment of plantation forestry: submission by Board of Private Forests Tasmania. 18 pp. Bormann, B. T., Homann, P. S., Darbyshire, R. L. & Morrissette, B. A. (2008) Intense forest wildfire sharply reduces mineral soil C and N: the first direct evidence. Canadian Journal of Forest Research 38: 2771-2783. Bowman, D. M. J. S. & Jackson, W. D. (1981) Vegetation succession in southwest Tasmania. Search 12: 358-362. Brown, M. & Podger, F. (1982) Floristics and fire regimes of a vegetation sequence from sedgeland- heath to Rainforest at Bathurst Harbour, Tasmania. Australian Journal of Botany 30: 659-676. Buchanan A and Honey B (1994). Energy and carbon dioxide implications of building construction. Energy and Buildings 20:205–217. Burns, K., Sledge, P. and Wicks, S. (2009) ABARE 2007 sawmill survey report ABARE report to the Australian Government Department of Agriculture, Fisheries and Forestry and Forest and Wood Products Australia. 20 pp. http://adl.brs.gov.au/data/warehouse/pe_abarebrs99001634/sawmill_survey_07_report.pdf Burrows, W., Henry, B. K., Back, P. V., Hoffmann, T. B., Tait, L. J., Anderson, E. R., Menke, N., Danaher, T., Carter, J. O. & McKeon, G. M. 2002. Growth and carbon stock change in eucalypt woodlands in northeast Australia: ecological and greenhouse sink implications. Global Change Biology 48, 769-784.

Page 336 of 365

Bulinski, J. (2012). US carbon markets and opportunities for Australian forest managers. Gottstein Fellowship Report, Melbourne. Candy, S.G. (1989a) Compatible tree volume and variable-form stem taper models for Pinus radiata in Tasmania. New Zealand Journal of Forestry Science 19, 97-111. Candy, S. G. (1989b) Growth and yield models for Pinus radiata in Tasmania. New Zealand Journal of Forestry Science 19, 112-133. Candy, S.G. (1997a) Growth and yield models for Eucalyptus nitens plantations in Tasmania and New Zealand Tasforests 9,167-198. Candy, S. G. (1997b) Estimation in forest yield models using composite link functions with random effects. Biometrics 53, 146-160. CFI Act (2011) Carbon Credits (Carbon Farming Initiative) Act 2011. Act No. 101 of 2011. http://www.comlaw.gov.au/Details/C2012C00417. Carbon Credits (Carbon Farming Initiative) Regulations 20111. Select Legislative Instrument 2011 No. 268. http://www.comlaw.gov.au/Details/F2011L02583 Cheney, N.P. (1981) Fire behaviour. In: Gill, A.M., Groves, R.H. and Noble, I.R. (eds.) Fire and the Australian Biota. Australian Academy of Science, Canberra, pp.151–175. Coleman, K., Jenkinson, D.S., 2008. ROTHC-26.3: a Model for the Turnover of Carbon in Soil: Model Description and User's Guide. November 1999 issue (modified August 2008) Rothamsted Research Harpenden, Herts. 47 pp. Corney S.P., Katzfey, J.J., McGregor, J.L., Grose, M.R., Bennett, J.C., White, C.J., Holz, G.K., Gaynor, S.M. & Bindoff, N.L. 2010, Climate Futures for Tasmania: climate modeling technical report, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania. CSIRO (2012). Australian Soil Resource Information System. http://www.asris.csiro.au/index_ie.html. Accessed February 2010. DCCEE (2011a) Australian national greenhouse accounts. National inventory report 2009 Volume 1. The Australian Government Submission to the UN Framework Convention on Climate Change. Department of Climate Change and Energy Efficiency, April 2011. http://www.climatechange.gov.au/publications/greenhouse- acctg/~/media/publications/greenhouse-acctg/national-inventory-report-2009-vol1.pdf DCCEE (2011b) Australian national greenhouse accounts. National inventory report 2009 Volume 2. The Australian Government Submission to the UN Framework Convention on Climate Change. Department of Climate Change and Energy Efficiency, April 2011. http://www.climatechange.gov.au/publications/greenhouse- acctg/~/media/publications/greenhouse-acctg/national-inventory-report-2009-vol2.pdf DCCEE (2011c) Australian national greenhouse accounts. National inventory report 2009 Volume 3. The Australian Government Submission to the UN Framework Convention on Climate Change. Department of Climate Change and Energy Efficiency, April 2011. http://www.climatechange.gov.au/publications/greenhouse- acctg/~/media/publications/greenhouse-acctg/national-inventory-report-2009-vol3.pdf DCCEE (2011d). Methodology for quantifying carbon sequestration by permanent environmental plantings of native species using the CFI Reforestation Modelling Tool. Carbon Farming Initiative

Page 337 of 365

Approved Methodology, Department of Climate Change and Energy Efficiency, Australian Government, Canberra. DCCEE (2011e). Carbon Farming Initiative positive list guidelines – Common practice (October 2011). Department of Climate Change and Energy Efficiency, Australian Government, Canberra. DCCEE (2011f). Carbon Farming Initiative – Salinity guidelines (December 2011). Department of Climate Change and Energy Efficiency, Australian Government, Canberra. DCCEE (2011g). Carbon Farming Initiative negative list guidelines – Management of adverse impacts (October 2011). Department of Climate Change and Energy Efficiency, Australian Government, Canberra. DCCEE (2011h). Carbon Farming Initiative preliminary estimates of abatement. Discussion Paper, Department of Climate Change and Energy Efficiency, Australian Government, Canberra. DCCEE (2012). The Carbon farming initiative handbook. Department of Climate Change and Energy Efficiency, Australian Government, Canberra. http://www.climatechange.gov.au/government/initiatives/carbon-farming- initiative/~/media/government/initiatives/cfi/handbook/CFI-Handbook-20120403-PDF.pdf Dean, C. (2011) Comment on “Carbon in Trees in Tasmanian State Forest”. Letter to the editor. International Journal of Forestry Research 2011, Article ID 212361, 10 pages. Dean, C. (2012) Creative Carbon Accounting—A Reply to “The Wood, the Trees, or the Forest? Carbon in Trees in Tasmanian State Forest: A Response to Comments ”Comment on “Carbon in Trees in Tasmanian State Forest”. Letter to the editor. International Journal of Forestry Research 2012, Article ID 762080, 3 pages. Dean, C. Fitzgerald, N. B., Wardell-Johnson, G. W. (2011). Pre-logging carbon accounts in old- growth forests, via allometry: An example of mixed-forest in Tasmania, Australia. Plant Biosystems, 2011: 1–14. Dean, C., Roxburgh, S.H. (2006). Improving visualisation of mature, high-carbon sequestering forests. Forest Biometry, Modelling and Information Sciences. 1. 48-69. Dean, C., Roxburgh, S. & Mackey, B. G., (2003). Growth modelling of Eucalyptus regnans for carbon accounting at the landscape scale. In: Amaro, A., Reed, D. and Soares, P. (Eds.) Modelling Forest Systems. CABI Publishing, Wallingford, UK. Dean, C. & Wardell-Johnson, G. (2010) Old-growth forests, carbon and climate change: Functions and management for tall open-forests in two hotspots of temperate Australia. Plant Biosystems 144, 180–193 DeVries, P. G. (1979) Line intersect sampling – statistical theory, applications, and suggestions for extended use in ecological inventory. In “Sampling Biological Populations” pp1-70. eds. R. M. Cormack, G. P. Patil and D. S. Robson. International Co-operative Publishing House, Maryland, USA. Drew, T.J. and Flewelling, J.W. (1979) Stand density management: an alternative approach and its application to Douglas-fir plantations. Forest Science 25: 518-532. Dyrness, C. T., K. Van Cleve, and J. Levison. 1989. The effects of wildfire on soil chemistry in four forest types in interior Alaska. Canadian Journal of Forest Research 19:1389-1396.Diaz, D.,

Page 338 of 365

Hamilton, K., & Johnson, E. (2011). State of the forest carbon markets 2011 – From canopy to currency. Ecosystem Marketplace, Washington. Eamus, D., McGuiness, K. & Burrows, W. (2000) Review of Allometric Relationships for Woody Biomass for Queensland, the Northern Territory and Western Australia’. National Carbon Accounting System Technical Report No. 5a. 60 pp. Australian Greenhouse Office, Canberra. Elliott, H.J., Felton, K.C., Jarman, S.J., & Stone, M.G. (2008) A history of innovation: eighty-five years of research and development at Forestry Tasmania. (Forestry Tasmania: Hobart) Ellis, R.C. & Graley, A.M. (1983). Gains and losses in soil nutrients associated with harvesting and burning eucalypt Rainforest. Plant and Soil 74: 437–450. Ellis, R.C. (1985). The relationships among eucalypt forest, grassland and Rainforest in a highland area in north-eastern Tasmania. Australian Journal of Ecology 10: 297-314. Fensham, R.J., Fairfax, R.J., and Ward, D.P. 2008, Drought-induced tree death in savanna. Global Change Biology 15 (2): 380-387. Florence, R. G. (2004) Ecology and silviculture of eucalypt forests. CSIRO. 413 pp. Forestry Commission (1961). Annual Report 1960–61, Forestry Commission, Tasmania. Forestry Tasmania (2001). Thinning regrowth eucalypts. Native Forest Silviculture Technical Bulletin No. 13, Forestry Tasmania. Forestry Tasmania (2005a). Sustainable forest management report 2004-05. Forestry Tasmania (2005b). Towards a new silviculture in Tasmania’s oldgrowth forests. Final advice to the Tasmanian Government. Forestry Tasmania, Hobart. Forestry Tasmania (2006). Annual report 200. Forestry Tasmania (2007). Sustainable high quality eucalypt sawlog supply from Tasmanian State Forest. Forestry Tasmania (2008). Forest Management Plan 2008. Sustainability Charter. http://www.forestrytas.com.au/sfm/sustainability-charter Forestry Tasmania (2010). Silvicultural systems for native eucalypt forests. Native Forest Silviculture Technical Bulletin No. 5, Forestry Tasmania. Forestry Tasmania (2011a). Sustainable Forest Management. Appendix 2: Data tables. http://www.forestrytas.com.au/uploads/File/pdf/pdf2011/appendix2_data_tables_2011.pdf Forestry Tasmania (2011b). Evaluation of Wood Resource Scenarios relevant to the Tasmanian Forests Statement of Principles to lead to an Agreement – Final Report to Signatories. http://www.forestrytas.com.au/uploads/File/pdf/pdf2011/final_report_signatories_resource_modell ing.pdf. FPA (2011). Forest Practices Authority Annual Report 2010-11. Gan, J. & McCarl, B.A. (2007). Measuring transnational leakage of forest conservation. Ecological Economics 64: 423-432. Garnaut, R. (2008) The Garnaut climate change review. Final report. Cambridge University Press. 634 pp.

Page 339 of 365

Garcia, O. (2011) A Parsimonious Dynamic Stand Model for Interior Spruce in British Columbia. Forest Science 57: 265-280. Garnaut, R. (2011) The Garnaut review 2011, Australia in the Global Response to Climate Change. Cambridge University Press. 222 pp. Gavran, M. & Parsons, M. (2011) Australian plantation statistics 2011, Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra, 43. Gifford, R. (2000) Carbon Content of Woody Roots: Revised Analysis and a Comparison with Woody Shoot Components (Revision 1). National Carbon Accounting System Technical Report No. 7 (10pp). Australian Greenhouse Office, Canberra. Gilbert, J. M. (1959). Forest succession in the Florentine Valley, Tasmania. Papers and Proceedings of the Royal Society of Tasmania 93, 129-151. Goodwin, A. (2000) Fire Management in Australian Forests. In R. J. Raison and R. Squire, (eds.) Forest management in Australia: implications for carbon budgets. National Carbon Accounting System Technical Report No. 32, Australian Greenhouse Office, Canberra, Gould, J. & Cheney, P. (2007) Fire Management in Australian Forests. In R. J. Raison and R. Squire, (eds.) Forest management in Australia: implications for carbon budgets. National Carbon Accounting System Technical Report No. 32, Australian Greenhouse Office, Canberra, Australia. Available at: http://www.greenhouse.gov.au/ncas/publications/index.html. Grierson, P.F., Adams M.A., & Attiwill P.M. (1992). Carbon storage in soil and in forest products. 44 pp. (School of Botany, University of Melbourne: Parkville, Vic.). Griffin, R., Hardner, C. Buxton, P. Potts, B. Bell, C. (2008) The breeding system of the forest giant, Eucalyptus regnans. In Old Forests, New Management. Conservation and use of old- growth forests in the 21st century. Conference 17 - 21 February 2008, Hobart, Tasmania, Australia. http://oldforests.com.au/pages/Presentations/Griffin.pdf Grose, M.R., Barnes-Keoghan, I., Corney, S.P., White, C.J., Holz, G.K., Bennett, J.C., Gaynor, S.M. & Bindoff, N.L. (2010) Climate Futures for Tasmania: general climate impacts technical report, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania. Grove, S.J., Meggs, J., Goodwin, A., (2002) A review of biodiversity conservation issues relating to coarse woody debris management in the wet eucalypt production forests of Tasmania. In. Forestry Tasmania, Hobart, p. 72. Grove, S.J., Stamm, L., Barry, C., (2009) Log decomposition rates in Tasmanian Eucalyptus obliqua determined using an indirect chronosequence approach. Forest Ecology and Management 258, 389-397. Grove, .J. & Stamm, L. (2011). Downed woody debris in Tasmanian eucalypt forest: modelling the effects of stand-replacing disturbance dynamics. Division of Forest Research and Development Technical Report 15/2011. Gupta R. K. & Rao, D. L. N. (1994). Potential of wastelands for sequestering carbon by reforestation. Current Science 66: 378–380. Hamilton, K., Bayon, R., Turner, G., & Higgins, D. (2007). State of the voluntary carbon markets 2007: Picking up steam. Ecosystem Marketplace & New Carbon Finance, Washington.

Page 340 of 365

Hamilton, K., Sjardin, M., Marcello, T. & Xu, G. (2008). Forging a frontier: State of the voluntary carbon markets 2008. Ecosystem Marketplace & New Carbon Finance, Washington. Hamilton, K., Sjardin, M., Peters-Stanley, M., & Marcello, T. (2010). Building bridges: State of the voluntary carbon markets 2010. Ecosystem Marketplace & Bloomberg New Energy Finance, Washington. Hamilton, K., Sjardin, M., Shapiro, A. & Marcello, T. (2009). Fortifying the foundation: State of the voluntary carbon markets 2009. Ecosystem Marketplace & New Carbon Finance, Washington. Hatch, A.B. (1960). Ash bed effects in Western Australian forest soils. Forests Department of Western Australia Bulletin No. 64. Hennessy, K. Lucas, C., Nicholls, N., Bathols, J., Suppiah, R. & Ricketts, J. (2006). Climate change impacts on fire-weather in south-east Australia. CSIRO, Australia Hickey, J. (1990). Change in Rainforest vegetation in Tasmania. Tasforests. 2: 143-149. Hickey, J., Davis, S., Wardman, R. & Harris, J. (1993). How much Rainforest is in Tasmania? A better answer to a difficult question. Tasforests 5, 13-24. Hickey, J. E., Su, W. Rowe, P. Brown, M. & Edwards, L. (1999). Fire history of the tall wet eucalypt forests of the Warra ecological research site, Tasmania. http://www.warra.com/documents/publications/Hickey_etal_1999a.pdf Holz, G.K., Grose, M.R., Bennett, J.C., Corney, S.P., White, C.J., Phelan, D., Potter, K., Kriticos, D., Rawnsley, R., Parsons, D., Lisson, S., Gaynor, S.M. & Bindoff, N.L. 2010, Climate Futures for Tasmania: impacts on agriculture technical report, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania. Houghton (2003a) Revised estimates of the annual net flux of carbon to the atmosphere from changes in land use and land management 1850–2000. Tellus 55B: 378–390. Houghton, R. A. (2003b). Why are estimates of the terrestrial carbon balance so different? Global Change Biology 9: 500-509. Hurteau, M., Koch, G. W. & Hungate, B. (2011) Carbon protection and fire risk reduction: toward a full accounting of forest carbon offsets. Frontiers in Ecology and the Environment. 6: 493-498 Ilic J. 1997. Woods of eucalyptus – part1. Distinguishing three species from the ash group (E. regnans, E. delegatensis and E. obliqua). IAWA J. 18: 27–36. IPART. (2011). Compliance and operation of the NSW Greenhouse Gas Reduction Scheme during 2010. Report to minister (July 2011), Independent Pricing and Regulatory Tribunal, Sydney. IPCC 1996, Climate Change 1995: The science of climate change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. University Press, Cambridge. IPCC (2000) IPCC Special report emissions scenarios. Cambridge University Press, UK. 570 pp. http://www.ipcc.ch/ipccreports/sres/emission/index.php?idp=0 IPCC, 2001. Climate Change 2001: The Scientific Basis. Chapter 3: The Carbon Cycle and Atmospheric Carbon Dioxide. http://www.ipcc.ch/ipccreports/tar/wg1/index.htm

Page 341 of 365

IPCC (2003). Good Practice Guidance for Land Use, Land-Use Change and Forestry. IPCC/IGES, Hayama, Japan. Appendix Table 3a.1.3. IPCC (2006a) Chapter 2: Generic methodologies applicable to multiple land-use categories. In IPCC Guidelines for National greenhouse gas inventories. Volume 4: Agriculture, forestry and other land use. Table 2.5, p. 2.47. http://www.ipcc- nggip.iges.or.jp/public/2006gl/pdf/4_Volume4/V4_02_Ch2_Generic.pdf IPCC (2006b) Chapter 5: Land-use change & forestry. In IPCC Revised Guidelines for National greenhouse gas inventories: Reference Manual (Volume 3). http://www.ipcc- nggip.iges.or.jp/public/gl/guidelin/ch5ref1.pdf Jaakko Pöyry Consulting 1999. National Carbon Accounting System – Usage and Lifecycle of Wood Products. NC AS Technical Report No.8. Canberra. Jaakko Pöyry Consulting 2000, Analysis of Wood Product Accounting Options for the National Carbon Accounting System. National Carbon Accounting System Technical Report No. 24 (37pp). Australian Greenhouse Office, Canberra. Jackson, W. D. (1968). Fire, air, water and earth - an elemental ecology of Tasmania. Proceedings of the Ecological Society of Australia 3, 9-16. Japan Lumber Journal (June, 2012). Japan Wood Market Stats. Online resource: http://www.jlj.gr.jp/publications/market.html Jenkinson, D. S. (1990) The turnover of organic carbon and nitrogen in soil. Philos. Trans. R. Soc. London B 329, 361–368. Jenkinson, D.S., Adams, D.E. & Wild, A. (1991). Model estimates of CO2 emissions from soil in response to global warming. Nature 351: 304-306. Jenkinson, D.S., Harris, H.C., Ryan, J., McNeill, A.M., Pilbeam, C.J. ,Coleman, K. (1999) Organic matter turnover in a calcareous clay soil from Syria under a two-course cereal rotation Soil Biology and Biochemistry 31 (1999) 687±693 Jenkinson, D.S., Hart, P.B.S., Rayner, J.H. & Parry, L.C. (1987). Modelling the turnover of organic matter in long-term experiments at Rothamsted. INTERCOL Bulletin 15: 1–8. Johnson, K. A. & Marsden-Smedley J. B. (2002). Fire history of the northern part of Tasmania’s wilderness world heritage area and its associated regions. Papers and Proceedings of the Royal Society of Tasmania, 136: 145-152. Keith, H., Barrett, D. & Keenan, R. (2000). ‘Review Allometric Relationships for Woody Biomass for New South Wales, the Australian Capital Territory, Victoria, Tasmania and South Australia’. National Carbon Accounting System Technical Report No. 5b (119pp). Australian Greenhouse Office. Keith, H., Mackey, B. & Lindenmayer, D. (2009). Re-evaluation of forest biomass carbon stocks and lessons from the world’s most carbon-dense forests. PNAS 106: 11635–11640. Keith, H., Mackey, B., Berry, S., Lindenmayer, D. & Gibbons, P. (2010). Estimating carbon carrying capacity in natural forest ecosystems across heterogeneous landscapes: addressing sources of error. Global Change Biology 16: 2971–2989.

Page 342 of 365

Kesteven, J., Landsberg, J. & URS Consulting (2004). Developing a national forest productivity model. National Carbon Accounting System Technical Report No.23, Australian Greenhouse Office, Canberra. King, K. J, de Ligt, R. M, Cary, G. J. (2011) Fire and carbon dynamics under climate change in south-eastern Australia: Insights from FullCAM and FIRESCAPE modelling', International Journal of Wildland Fire, 20: 563-577. Kossoy, A. & Guigon, P. (2012). State and trends of the carbon market 2012. Carbon Finance at the World Bank, Washington DC. Lewis, A., Gorner, S., Downey, L., Slezak, J., Michael, J. & Wonhas, A. (2008) An Australian cost curve for greenhouse gas reduction. McKinsey & Company report, Sydney. Linacre, N., Kossoy, A., Ambrosi, P., Sahlia, M.A., Bishop, V., Bosquet, B., de Gouvello, C., Matsuki, T., & Randade, M. (2011). State and trends of the carbon market 2011. Report, Carbon Finance at the World Bank. Lindenmayer, D.B., Mackey, B.G., Mullen, I.C., McCarthy, M.A., Gill, A.M., Cunningham, R.B., Donnelly, C.F., 1999. Factors affecting stand structure in forests—are there climatic and topographic determinants? Forest Ecology Management 123: 55–63. Lindenmayer, D. B., Hunter, M. L. Burton, P.J. & Gibbons, P. (2009). Effects of logging on fire regimes in moist forests. Conservation Letters 2: 271-277. Lucas, R.M., King, S., Davis, S. & Whitely, S. (1997). The role of Australia’s native forests in carbon sequestration: Tasmania as a case study. Report to the Department of Environment, Sports and Territories, Canberra, Australia. McCarthy, M. A. Gill, A. M. & Lindenmayer, D. B. (1999). Fire regimes in mountain ash forest evidence from forest age structure, extinction models and wildlife habitat. Forest Ecology and Management 124: 193-203. McIntosh, P. D., Laffan, M.D. & Hewitt, A.E. (2005). The role of fire and nutrient loss in the genesis of the forest soils of Tasmania and southern New Zealand. Forest Ecology and Management 220 (2005) 185–215. Macintosh, A. (2011). Credits from reduced native forest harvesting in Australia. Australian National University Centre for Climate Law & Policy, Working Paper No. 2011/1. ANU, Canberra. Macintosh, A. (2012a). Tasmanian Forests Intergovernmental Agreement: an assessment of its carbon value. ANU Centre for Climate Law and Policy. Macintosh, A. (2012b). Kyoto Protocol: Will Gillard stay or go? Crikey, web article, posted and accessed 12 June 2012. Mackensen, J. & Bauhus, J. (1999). The Decay of Coarse Woody Debris’. National Carbon Accounting System Technical Report No. 6 (41pp). Australian Greenhouse Office, Canberra. McKenzie, N.J., Jacquier, D. W., Maschmedt, D.J., Griffin, E. A. & Brough, D. M. (2005) ASRIS technical specifications version 1.5, October 2005. Available at http://www.asris.csiro.au/downloads/asris%20Tech%20Specs%201.5.pdf

Page 343 of 365

Mackey, B. G., Keith, H., Berry, S. L., & Lindenmayer, D. B. (2008). Green carbon: the role of natural forests in carbon storage. Part 1, a green carbon account of Australia’s south–eastern Eucalypt forest, and policy implications. ANU Press, Canberra, Australia. Marsden-Smedley J. B. (1998). Changes in southwestern Tasmanian fire regimes since the early 1800s. Papers and Proceedings of the Royal Society of Tasmania, 132: 15-29. Marsden-Smedley J. B. (2007). Fire management in the Tasmanian world heritage area: a report to the Tasmanian Parks and Wildlife Service. UTAS Innovation Ltd. 25 pp. MBAC (2007). Forestry Tasmania’s carbon sequestration position. Forestry Tasmania December 2nd 2007. Mesibov, R. (2002). Review of the deep red myrtle resource in Tasmania. Steering Committee for RFA Deep Red Myrtle Resource Review, April 2002. Forestry Tasmania, Hobart, 56 pp. Montreal Process Implementation Group for Australia (2008) Australia’s State of the Forests Report 2008. Bureau of Rural Sciences, Canberra. Moroni, M. T., Kelley, T. H. & McLarin, M. L. (2010). Carbon in trees in Tasmanian State forest. International Journal of Forestry Research. Volume 2012, Article ID 848502, 6 pages Moroni, M. T. Kelley, T. H. McLarin, M. L. & Read, S. M. (2012). The wood, the trees, or the forest? Carbon in Trees in Tasmanian State Forest: A Response to Comments. Volume 2012, Article ID 848502, 6 pages. Mount, A. B. (1979). Natural regeneration processes in Tasmanian Forests. Search 10, 180-186. Mount, A. B. (2009). Managing Tasmania’s fire environment. Forest and Forest Industries Council of Tasmania. http://www.ffic.com.au/publications Nicholls, N. & Lucas, C. (2007). Interannual variations of area burnt in Tasmanian bushfires: relationships with climate and predictability. International Journal of Wildland Fire16: 540-546. Nobel, I. (1999) Impacts of global change on Australian forests: fire. In S, M. Howdon and J. T. Gorman eds. Impacts of global change on Australian temperate forests. Working Paper Series 99/08. CSIRO, Canberra. pp. 89-93. Norris, J., Arnold, S. & Fairman, T. (2010). An indicative estimate of carbon stocks on Victoria’s publicly managed land using the FullCAM carbon accounting model. Australian Forestry 73, 209. Norris, J. (2010). Spatial patterns of soil carbon and nitrogen after eucalypt forest fire. Firenote 57: 1-4. Ough, K., Ross, J., 1992. Floristics, Fire and Clearfelling in Wet Forests of the Central Highlands of Victoria. Silvicultural Systems Project Technical Report no. 11. Department of Conservation and Environment, Melbourne, Australia. OzClim (2007) Exploring climate change scenarios for Australia. http://www.csiro.au/ozclim/home.do Parsons, M., Gavran, M., & Davidson, J. (2006). Australia’s plantations 2006. Bureau of Rural Sciences, Department of Agriculture Fisheries and Forestry. Parsons, M., Frakes, I. & Gavran, M. (2007). Australia's plantation log supply 2005-2049. Canberra, Bureau of Rural Sciences, Department of Agriculture, Fisheries and Forestry

Page 344 of 365

Paul, K., Polglase, P., Coops, N., O’Connell, T., Grove, T., Medlam, D., Carlyle, C., May, B., Smethurst, P. & Baillie, C. (2002) Modelling Change in Soil Carbon Following Afforestation or Reforestation: Preliminary Simulations Using GRC3 and Sensitivity Analysis. National Carbon Accounting System Technical Report No. 29 (106pp), Australian Greenhouse Office, Canberra. Paul, K., Polglase, P. J. & Richards, G. P. (2003). Predicted change in soil carbon following afforestation or reforestation, and analysis of controlling factors by linking a C accounting model (CAMFor) to models of forest growth (3PG), litter decomposition (GENDEC) and soil C turnover (RothC). Forest Ecology and Management 177: 485–501. Paul, K. and Polglase, P. J. (2004) Prediction of decomposition of litter under eucalypts and pines using the FullCAM model. Forest Ecology and Management 191 73–92. Pennington P, Laffan M, Lewis R, & Otahal P. (2001). Assessing the long-term impacts of forest harvesting and high intensity broadcast burning on soil properties at the Warra LTER site. Tasforests 13: 291–301. Peters-Stanley, M., Hamilton, K., Marcello, T. & Sjardin, M. (2011). Back to the future – State of the voluntary carbon markets 2011. Ecosystem Marketplace & Bloomberg New Energy Finance, report, 2nd June 2011. Peters-Stanley, M. & Hamilton, K. (2012). Developing dimension: State of the voluntary carbon markets 2012. Ecosystem Marketplace & Bloomberg New Energy Finance, report, 31st May 2012. Point Carbon. (2010) US offset markets in 2010: The road not yet taken. Carbon Market Analyst North America, Point Carbon, 1st March 2010. Polglase, P.J., Paul, K.I., Khanna, P.K., Nyakuengama, J.G., O’Connell, A.M., Grove, T.S. & Battaglia, M. (2000). Change in Soil Carbon Following Afforestation and Reforestation. National Carbon Accounting System Technical Report No. 20 Australian Greenhouse Office, Canberra. 89 pp. Polglase, P.J., Snowdon, P., Theiveyanathan, T., Paul, K.I., Raison, R.J., Grove, T. & Rance, S.J. (2004). Calibration of the FullCAM model to Eucalyptus globulus and Pinus radiata and uncertainty analysis. National Carbon Accounting System Technical Report No. 41, Australian Greenhouse Office, Canberra. Polglase, P. Paul, K., Hawkins, C., Siggins, A., Turner, J., Booth, T, Crawford, D., Jovanovic, T., Hobbs, Opie, K., Almeida, A. & Carter, J. (2008). Regional Opportunities for Agroforestry Systems in Australia. RIRDC Publication No. 08/176. Polglase, P. Paul, K., Roxburgh, S., Culvenor, D. & Sullivan, A. (2010). Victoria’s LandCarbon project: review and recommendations for further research. CSRIO report. Polglase, P., Reeson, A., Hawkins, C. Paul, K., Siggins, A., Turner, J., Crawford, D., Jovanovic, T., Hobbs, Opie, K., Carwardine, J. & Almeida, A. (2010) Opportunities for carbon forestry in Australia: Economic assessment and constraints to implementation. CSIRO. 24 pp. Pyrke, A. F. & Marsden-Smedley, J. B. (2005). Fire-attributes categories, fire sensitivity, and flammability of Tasmanian vegetation communities. Tasforests 16. Rab, M. A. (1996). Soil physical and hydrological properties following logging and slash burning in E. regnans forest of southeastern Australia. Forest Ecology and Management 84: 159-176

Page 345 of 365

Raison, R. J. And Squire, R. O. (2007) Forest management in Australia implications for carbon budgets. National Carbon Accounting System Technical Report No. 32. Part 1 of 2. November 2007. 391 pp. Read, J. & Hill, R. S. (1983). Rainforest invasions onto Tasmanian old-fields. Australian Journal of Ecology 8: 149-161. Reid, J. B., Hill, R. S., Brown, M. J. & Hovenden, M. J. (2005). Vegetation of Tasmania. Flora of Australia Supplementary Series No. 8. Australian Biological Resources Study. Reid, S. (2010). Fire and carbon in managed and unmanaged forests. Division of Forest Research and Development. Technical Report 14/2010 Reineke, L.H. (1933) Perfecting a stand-density index for even-aged forests. Journal of Agricultural Research 46: 627-638 Richards, G.P. (2001). The FullCAM carbon accounting model: Development, calibration and implementation. IEC Bioenergy Task 38: Workshop paper, Canberra, March 2001. Richards, G., Evans, D., Reddin, A & Leitch, J. (2005). The FullCAM carbon accounting model (version 3.0) user manual: National carbon accounting system. Australian Government Department of Environment & Heritage Australian Greenhouse Office, Canberra. Richards, G.P., Borough, C., Evans, D., Reddin, A., Ximenes, F. & Gardner, D. (2007). Developing a carbon stocks and flows model for Australian wood products. Australian Forestry 2007 70: 108-119. Riley, I. (2012) Effect of late stage succession on Tasmania’s forest carbon stocks and fluxes. Report prepared for CO2 Australia, 15th May 2010. Roxburgh, S., Wood, S., Mackey, B., Woldendorp, G. & Gibbons, P. (2006). Assessing the carbon sequestration potential of managed forests: a case study from temperate Australia. Journal of Applied Ecology 43, 1149-1159. Roxburgh, S., England, J. & Paul, K. (2010). Predicting biomass carbon in biodiverse plantings and native forest ecosystems. Prepared for the Department of Sustainability & Environment (VIC), Environmental Policy & Climate Change Branch. CSIRO, Canberra, ACT Ryan, M. F., Spencer, R. D., & Keenan, R. (2002) Private native forests in Australia: What did we learn from the Regional Forest Agreement program? Australian Forestry. 65: 141-152. Skjemstad, J. & Spouncer, L. (2002). Estimating Changes in Soil Carbon Resulting from Changes in Land Use. National Carbon Accounting System Technical Report No. 36, Australian Greenhouse Office, Canberra. Slijepcevic, A. (2001) Loss of carbon during controlled regeneration burns in Eucalyptus obliqua forest. Tasforests Vol. 13: 281-290. Snowdon, P., Eamus, D., Gibbons, P., Khanna, P., Keith, H., Raison, J. & Kirschbaum, M. (2000). Synthesis of allometrics, review of root biomass and design of future woody biomass sampling strategies. National Carbon Accounting System Technical Report No. 17, September 2000. Stone, M.G. (1998). Forest-type mapping by photo-interpretation: A multi-purpose base for Tasmania’s forest management. Tasforests 10, 15-32.

Page 346 of 365

Tasmanian Government (2011). Tasmanian forests intergovernmental agreement between the Commonwealth of Australia and the State of Tasmania 7 August 2011. http://www.environment.gov.au/land/forests/pubs/tasmanian-forests-intergovernmental- agreement.pdf Tomkins, I.B., Kellas, J.D., Tolhurst K.G. & Oswin, D.A. (1991). Effects of fire intensity on soil chemistry in a eucalypt forest. Australian Journal Soil Research 29: 25–47. Thauvin, T., Libis, E., Grove, S.J. & Wardlaw, T. (2010). Comparison of coarse woody debris volumes between mature and silviculturally regenerated eucalypt forest along a disturbance gradient. Technical Report no. 15/2010. Forestry Tasmania, Hobart. Tucker, S. Tharumarajah, A. May, B. England, J. Paul, K. Raison, J. Hall, M. McPharlain, S. Mitchell, P. Rouwette, R. Seo, & S. Syme, M. (2009). Life cycle inventory of Australian forestry and wood Products. FWPA Report Project No. PNA008-0708. Turner, J. (1990). Forestry, the Timber Industry and the Greenhouse Effect. Miscellaneous paper, Forestry Commission of New South Wales, Sydney. Turner, B. & James, R. (2002). Derivation of indicative yields for major plantation species. In: Richards, G. (Ed.) Biomass Estimation: Approaches for Assessment of Stocks and Change. National Carbon Accounting System Technical Report no. 27, Australian Greenhouse Office, Canberra, pp 71-77. Turner, P. A. M., Balmer, J. & Kirkpatrick, J. B. (2009). Stand-replacing wildfires? The incidence of multicohort and single-cohort Eucalyptus regnans and E. obliqua forests in southern Tasmania. Forest Ecology and Management 258: 366-375. TVMP (2004) Documentation for TASVEG 1_0. http://www.dpiw.tas.gov.au/inter/nsf/Attachments/LJEM- 7FB3N3/$FILE/Documentation%20for%20TASVEG%201.0.pdf UNFCCC (2010). Report of the Ad Hoc Working Group on Further Commitments for Annex I Parties under the Kyoto Protocol on its fifteenth session, held in Cancun from 29 November to 10 December 2010. http://unfccc.int/resource/docs/2010/awg15/eng/18a01.pdf UNFCCC (2011) Decision -/CMP.7, Land use, land-use change and forestry. Advance unedited version. http://unfccc.int/files/meetings/durban_nov_2011/decisions/application/pdf/awgkp_lulucf.pdf UNFCCC (2012) Report of the Conference of the Parties serving as the meeting of the Parties to the Kyoto Protocol on its seventh session, held in Durban from 28 November to 11 December 2011. Addendum, Part Two: Action taken by the Conference of the Parties serving as the meeting of the Parties to the Kyoto Protocol at its seventh session. http://unfccc.int/resource/docs/2011/cmp7/eng/10a01.pdf Vanclay, J.K. and Sands, P.J. (2009) Calibrating the self-thinning frontier. Forest Ecology and Management 259:81-85. Verified Carbon Standard (2011). AFOLU Non-permanence risk tool. VCS Version 3 Procedural Document, 8 March 2011, V3.0 Verified Carbon Standard (2012). Agriculture, forestry and other land use (AFOLU) requirements. Requirements Document, 1 February 2012, V3.2

Page 347 of 365

Von Platen, J., Kirkpatrick, J. B., & Allen, K., J. (2011). Fire frequency variation in south-eastern Tasmanian dry eucalypt forest 1740-2004 from fire scars. Australian Forestry 74: 180-189. Wang Y, & Hamilton F. 2003. Projecting stand attributes of regrowth ash eucalypts sampled in forest inventory. Australia Forestry. 66: 75–86. Waterworth, R.M., Richards, G.P., Brack, C.L., & Evans, D.M.W. (2007). A generalised process- empirical hybrid model for predicting forest growth. Forest Ecology and Management 238, 231– 243. Waterworth, R. M. & Richards, G.P. (2008) Implementing Australian forest management practices into a full carbon accounting model. Forest Ecology and Management 255: 2434-2443. Webb, A. (2001). Pre-clearing Soil Carbon Levels in Australia. National Carbon Accounting System Technical Report No. 12. White, C.J., Sanabria, L.A., Grose, M.R., Corney, S.P., Bennett, J.C., Holz, G.K., McInnes, K.L., Cechet, R.P., Gaynor, S.M. & Bindoff, N.L. 2010, Climate Futures for Tasmania: extreme events technical report, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania. Whiteley, S.B. (1999) Calculating the sustainable yield of Tasmania’s state forests. Tasforests 11, 23–34. Williams, A. A. J., Karoly, D. J., Tapper, N. (2001) The sensitivity of Australian fire danger to climate change. Climatic Change 49: 171–191. Woldendorp, G. Keenan, R.J. Ryan, M.F. (2002). Coarse Woody Debris in Australian Forest Ecosystems. A Report for the National Greenhouse Strategy, Module 6.6 (Criteria and Indicators of Sustainable Forest Management), April 2002. Wood, M. J., McLarin, M., Volker, P & Syme, M. (2009) Management of eucalypt plantations for profitable sawlog production in Tasmania. Tasforests 18: 117-130. Wood, S. W., Hua, Q, Allen, K. J. & Bowman, D. M. J. S. (2010). Age and growth of a fire prone Tasmanian temperate old-growth forest stand dominated by Eucalyptus regnans, the world’s tallest angiosperm. Forest Ecology and Management 260: 438–447. Wood, S. W., Hua, Q, Allen, K. J. & Bowman, D. M. J. S. (2011). Firescape ecology: how topography determines the contrasting distribution of fire and rain forest in the south-west of the Tasmanian Wilderness World Heritage Area. Journal of Biogeography 38: 1807–1820. Wood, S. W. & Bowman, D. M. J. S. (2011). Alternative stable states and the role of fire– vegetation–soil feedbacks in the temperate wilderness of southwest Tasmania. Landscape Ecology DOI 10.1007/s10980-011-9677-0. Working Group for Rainforest Conservation (1987). The Rainforest of Tasmania. Working groups for Rainforest conservation. Hobart. Ximenes, F. A., Gardner, W. D. & Kathuria, A. (2008a). Proportion of above-ground biomass in commercial logs and residues following the harvest of five commercial forest species in Australia. Forest Ecology and Management. 256: 335–346. Ximenes, F.A., Gardner, W.D & Cowie, A. (2008b). The decomposition of wood products in landfills in Sydney, Australia. Waste Management 28: 234-2354.

Page 348 of 365

APPENDICES

Page 349 of 365

APPENDIX 1: KEY ASSUMPTIONS FOR LEAKAGE ASSESSMENT

Metric/assumption Value Proportion of Tasmanian extracted wood products exported 69% Proportion of Tas. extracted wood products consumed domestically 31% Total Australian production of extracted wood products 26,490,000 m3 Total Australian import of wood products 1,328,000 m3 Total Australian export of wood products 12,339,000 m3 Treatment of paper products Not considered Proportion of domestic consumption met by international sources 5% Key market customer for exports Japan Total Japanese wood product imports (all global sources) 16,811,000 m3 Total Australian wood product export to Japan 3,196,000 m3 Japanese imports from non-Australian sources 81% International Leakage Factor (LFInt) 58%

Proportional Biomass Leakage Factor (native forest to plantation sourced) 20%

Proportional Biomass Leakage Factor (plantation to plantation) 40% Key assumptions: Where reduced extraction from Tasmanian forests is replaced domestically, the replacement source is plantation grown timber located on mainland Australia. Plantation grown timber has a substantially (>15%) lower carbon emissions (or biomass removal) profile associated with each unit of extraction than for native forest. Wood import, export and production figures are based on 2011 statistics, which are assumed to be indicative of the long-term average trend.

Calculated Market Leakage Factor (reduced native forest extraction) 8%

Calculated Market Leakage Factor (reduced plantation extraction) 17%

Data sources: ABARES (2012); Japan Lumber Journal (2012).

Page 350 of 365

APPENDIX 2: FIRE AND HARVESTING PARAMETERS

Fire Frequency

Study Baseline

Forest Type Fire frequency Fire intensity Distribution Age Frequency % Big trees Age of Big Max Mode Mode Base killed Tree Tall eucalypt forest Triangle 600 50 0.0033 0.004 50 80 Low eucalypt forest Triangle 250 25 0.008 0.03 20 40

Rainforest Rectangle 0.001 75 80

Other native forest Rectangle 0.04 10 80

Hardwood plantations Rectangle 0.01 100 80

Softwood plantations Rectangle 0.01 100 80

Grassland Rectangle 0.05 NA NA

10% decrease (Scenarios F1, L3)

Forest Type Fire frequency Fire intensity Distribution Age Frequency Age of Big % Big trees Tree Max Mode Mode Base killed Tall eucalypt forest Triangle 600 50 0.003 0.0036 50 80 Low eucalypt forest Triangle 500 25 0.0072 0.027 20 40

Rainforest Rectangle 0.0009 75 80

Other native forest Rectangle 0.036 10 80

Hardwood plantations Rectangle 0.009 100 80

Softwood plantations Rectangle 0.009 100 80

Grassland Rectangle 0.045 NA NA

10% increase (Scenarios F1, L3)

Forest Type Fire frequency Fire intensity

Distribution Age Frequency % Big trees Age of Big Tree Max Mode Mode Base killed Tall eucalypt forest Triangle 600 50 0.0037 0.0044 50 80 Low eucalypt forest Triangle 500 25 0.0088 0.033 20 40

Rainforest Rectangle 0.0011 75 80

Other native forest Rectangle 0.044 10 80

Hardwood plantations Rectangle 0.0011 100 80

Softwood plantations Rectangle 0.0011 100 80

Grassland Rectangle 0.055 NA NA

Page 351 of 365

Timber harvesting

Study Baseline

Forest type Harvest rate Harvest intensity Site preparation and and product conversion Area Trees Stems Slash Type Age Conversion cut cut removed burning years %/y % total % cut % % Tall eucalypt forest

Sawlog Thinning 8-15 7% 50% 85% 0% 0%

Sawlog Clearfell, thinned 50-150 5% 90% 85% 80% 0%

Sawlog Clearfell, unthinned 70-150 5% 90% 85% 80% 0%

Sawlog Clearfell >150 5% 90% 85% 80% 0%

Low eucalypt forest

Sawlog Partial harvest 50-100 5% 50% 85% 70% 0%

Sawlog Partial harvest >100 5% 50% 85% 70% 0%

Rainforest

Sawlog Clearfell >150 0.50% 90% 85% 80% 0%

Other native forest

Sawlog Clearfell >60 5% 90% 80% 75% 0%

Hardwood plantations

Pulplog Clearfell >13 35% 50% 80% 0% 0%

Sawlog Thinning 8-12 40% 50% 80% 0% 0%

Sawlog Clearfell >25 25% 100% 90% 25% 0%

Softwood plantations

Sawlog Thinning 14-18 20% 50% 80% 0% 0%

Sawlog Clearfell >28 10% 100% 90% 25% 0%

Scenario N2 changes from Study Baseline

Sawlog Thinning 8-15 15% 50% 85% 0% 0%

Sawlog Clearfell, thinned 50-150 5% 90% 85% 80% 0%

Sawlog Clearfell, unthinned 70-150 7% 90% 85% 80% 0%

Low eucalypt forest

Sawlog Partial harvest 50-100 6% 50% 85% 70% 0%

Page 352 of 365

Scenario N3-60Y changes from Study Baseline

Sawlog Thinning 8-15 15% 50% 85% 0% 0%

Sawlog Clearfell, thinned 60-150 5% 90% 85% 80% 0%

Sawlog Clearfell, unthinned 60-150 5% 90% 85% 80% 0%

Sawlog Clearfell >110 5% 90% 85% 80% 0%

Low eucalypt forest

Sawlog Partial harvest 30-100 5% 50% 85% 70% 0%

Timber Harvesting: Scenario N3-80Y changes from Study Baseline

Sawlog Thinning 8-15 15% 50% 85% 0% 0%

Sawlog Clearfell, thinned 80-150 5% 90% 85% 80% 0%

Sawlog Clearfell, unthinned 80-150 5% 90% 85% 80% 0%

Sawlog Clearfell >130 5% 90% 85% 80% 0%

Low eucalypt forest

Sawlog Partial harvest 40-120 5% 50% 85% 70% 0%

Scenario N3-120Y changes from Study Baseline

Sawlog Thinning 8-15 15% 50% 85% 0% 0%

Sawlog Clearfell, thinned 120-170 5% 90% 85% 80% 0%

Sawlog Clearfell, unthinned 120-170 5% 90% 85% 80% 0%

Sawlog Clearfell >170 5% 90% 85% 80% 0%

Low eucalypt forest

Sawlog Partial harvest 60-140 5% 50% 85% 70% 0%

Page 353 of 365

Conversion

Study Baseline: No conversion post 2010

Forest Type Harvested Converted to: proportion Native Hardwood Softwood converted Grassland eucalypt plantations plantations % % % % % Tall eucalypt forest 0 NA NA NA NA Low eucalypt forest 0 NA NA NA NA Rainforest 0 NA NA NA NA Other native forest 0 NA NA NA NA Hardwood plantations 0 NA NA NA NA Softwood plantations 0 NA NA NA NA Grassland 0 NA NA NA NA

Conversion: Scenario A1

Forest Type Proportion Converted to: converted Native Hardwood Softwood Grassland eucalypt plantations plantations %/year % % % % Tall eucalypt forest 2.5% (> 70 years) 0 74 8 18 Low eucalypt forest 2.5% (> 50 years) 0 74 8 18 Rainforest 0.4% 0 74 8 18 Other native forest 2.5% (> 60 years) 0 74 8 18

Scenario A2

Forest Type Proportion Converted to: converted Native Hardwood Softwood Grassland eucalypt plantations plantations %/year % % % % Other native forest 5% (> 60 years) 100 0 0 0

Scenario A3

Forest Type Proportion Converted to: converted Native Hardwood Softwood Grassland eucalypt plantations plantations %/year % % % % Grassland 0.52% 0 93 7 0

Page 354 of 365

Scenario A4

Forest Type Proportion Converted to: converted Native Hardwood Softwood Grassland eucalypt plantations plantations %/year % % % % Grassland 1.04% 0 93 7 0

Scenario A5

Forest Type Proportion Converted to: converted Native Hardwood Softwood Grassland eucalypt plantations plantations %/year % % % % Grassland 13% (>13 years) 0 0 0 100

Page 355 of 365

APPENDIX 3: BASE PARAMETERS USED IN FCMF

Growth Parameters Current Age of OG = 150 years Age when 95% max eucalypt volume remains = 250 years Age when5% max eucalypt volume remains = 500 years Biomass Allocation and Turnover Biomass Allocation CategoryID SppGpID BioComponentID Val BioAllocation (%) NativeEuc Stem 55 BioAllocation (%) NativeEuc Branch 12 BioAllocation (%) NativeEuc Bark 10 BioAllocation (%) NativeEuc Leaf 3 BioAllocation (%) NativeEuc RootCoarse 17 BioAllocation (%) NativeEuc RootFine 3 BioAllocation (%) Rainforest Stem 50 BioAllocation (%) Rainforest Branch 15 BioAllocation (%) Rainforest Bark 10 BioAllocation (%) Rainforest Leaf 4 BioAllocation (%) Rainforest RootCoarse 17 BioAllocation (%) Rainforest RootFine 4 BioAllocation (%) HardwoodPlantation Stem 67 BioAllocation (%) HardwoodPlantation Branch 9 BioAllocation (%) HardwoodPlantation Bark 10 BioAllocation (%) HardwoodPlantation Leaf 3 BioAllocation (%) HardwoodPlantation RootCoarse 8 BioAllocation (%) HardwoodPlantation RootFine 3 BioAllocation (%) SoftwoodPlantation Stem 67 BioAllocation (%) SoftwoodPlantation Branch 9 BioAllocation (%) SoftwoodPlantation Bark 10 BioAllocation (%) SoftwoodPlantation Leaf 3 BioAllocation (%) SoftwoodPlantation RootCoarse 8 BioAllocation (%) SoftwoodPlantation RootFine 3 BioAllocation (%) Understorey Stem 50 BioAllocation (%) Understorey Branch 15 BioAllocation (%) Understorey Bark 10 BioAllocation (%) Understorey Leaf 5 BioAllocation (%) Understorey RootCoarse 17 BioAllocation (%) Understorey RootFine 3

Page 356 of 365

Basic density

CategoryID SppGpID BioComponentID Val BasicDensity (t/m3) NativeEuc All 0.55 BasicDensity (t/m3) Rainforest All 0.55 BasicDensity (t/m3) HardwoodPlantation All 0.54 BasicDensity (t/m3) SoftwoodPlantation All 0.45 BasicDensity (t/m3) Understorey All 0.55 BasicDensity (t/m3) Grass All 0.55

Carbon content

CategoryID SppGpID BioComponentID Val CarbonDensity (%t/m3) NativeEuc Stem 52 CarbonDensity (%t/m3) NativeEuc Branch 47 CarbonDensity (%t/m3) NativeEuc Bark 49 CarbonDensity (%t/m3) NativeEuc Leaf 52 CarbonDensity (%t/m3) NativeEuc RootCoarse 49 CarbonDensity (%t/m3) NativeEuc RootFine 46 CarbonDensity (%t/m3) Rainforest Stem 52 CarbonDensity (%t/m3) Rainforest Branch 47 CarbonDensity (%t/m3) Rainforest Bark 49 CarbonDensity (%t/m3) Rainforest Leaf 52 CarbonDensity (%t/m3) Rainforest RootCoarse 49 CarbonDensity (%t/m3) Rainforest RootFine 46 CarbonDensity (%t/m3) HardwoodPlantation Stem 50 CarbonDensity (%t/m3) HardwoodPlantation Branch 46.8 CarbonDensity (%t/m3) HardwoodPlantation Bark 48.7 CarbonDensity (%t/m3) HardwoodPlantation Leaf 52.9 CarbonDensity (%t/m3) HardwoodPlantation RootCoarse 49.2 CarbonDensity (%t/m3) HardwoodPlantation RootFine 46.1 CarbonDensity (%t/m3) SoftwoodPlantation Stem 51 CarbonDensity (%t/m3) SoftwoodPlantation Branch 51.4 CarbonDensity (%t/m3) SoftwoodPlantation Bark 53.3 CarbonDensity (%t/m3) SoftwoodPlantation Leaf 51.1 CarbonDensity (%t/m3) SoftwoodPlantation RootCoarse 50.4 CarbonDensity (%t/m3) SoftwoodPlantation RootFine 48.4 CarbonDensity (%t/m3) Understorey All 50

Page 357 of 365

Stag production

CategoryID Val bDeviate DyingTreePropOfMeanV 0.3 TRUE

Turnover

CategoryID SppGpID BioComponentID Val bDeviate TurnoverToGrnd (%/y) NativeEuc Stem 1 Yes TurnoverToGrnd (%/y) NativeEuc Branch 6.5 Yes TurnoverToGrnd (%/y) NativeEuc Bark 9.5 Yes TurnoverToGrnd (%/y) NativeEuc Leaf 40 Yes TurnoverToGrnd (%/y) NativeEuc RootCoarse 6.5 Yes TurnoverToGrnd (%/y) NativeEuc RootFine 73.3 Yes TurnoverToGrnd (%/y) Rainforest Stem 1 Yes TurnoverToGrnd (%/y) Rainforest Branch 6.5 Yes TurnoverToGrnd (%/y) Rainforest Bark 9.5 Yes TurnoverToGrnd (%/y) Rainforest Leaf 40 Yes TurnoverToGrnd (%/y) Rainforest RootCoarse 6.5 Yes TurnoverToGrnd (%/y) Rainforest RootFine 73.3 Yes TurnoverToGrnd (%/y) HardwoodPlantation Stem 1 Yes TurnoverToGrnd (%/y) HardwoodPlantation Branch 5 Yes TurnoverToGrnd (%/y) HardwoodPlantation Bark 7 Yes TurnoverToGrnd (%/y) HardwoodPlantation Leaf 50 Yes TurnoverToGrnd (%/y) HardwoodPlantation RootCoarse 10 Yes TurnoverToGrnd (%/y) HardwoodPlantation RootFine 85 Yes TurnoverToGrnd (%/y) SoftwoodPlantation Stem 1 Yes TurnoverToGrnd (%/y) SoftwoodPlantation Branch 3 Yes TurnoverToGrnd (%/y) SoftwoodPlantation Bark 5 Yes TurnoverToGrnd (%/y) SoftwoodPlantation Leaf 30 Yes TurnoverToGrnd (%/y) SoftwoodPlantation RootCoarse 7 Yes TurnoverToGrnd (%/y) SoftwoodPlantation RootFine 80 Yes TurnoverToGrnd (%/y) Understorey Stem 1 Yes TurnoverToGrnd (%/y) Understorey Branch 10 Yes TurnoverToGrnd (%/y) Understorey Bark 15 Yes TurnoverToGrnd (%/y) Understorey Leaf 50 Yes TurnoverToGrnd (%/y) Understorey RootCoarse 6.5 Yes TurnoverToGrnd (%/y) Understorey RootFine 73.3 Yes

Page 358 of 365

Debris decomposition

Decomposable fraction

CategoryID SppGpID BioComponentID Val DebrisDecomposable (%) All Stem 10 DebrisDecomposable (%) All Branch 15 DebrisDecomposable (%) All Bark 10 DebrisDecomposable (%) All Leaf 70 DebrisDecomposable (%) All RootCoarse 55 DebrisDecomposable (%) All RootFine 70 DebrisDecomposable (%) Grass All 10 StagDecomposable (%) All Stem 10

Debris decay rates

CategoryID SppGpID BioComponentID bDecomposable Val DebrisDecay (%) NativeEuc Stem Yes 5 DebrisDecay (%) NativeEuc Stem No 5 DebrisDecay (%) NativeEuc Branch Yes 10 DebrisDecay (%) NativeEuc Branch No 10 DebrisDecay (%) NativeEuc Bark Yes 50 DebrisDecay (%) NativeEuc Bark No 50 DebrisDecay (%) NativeEuc Leaf Yes 80 DebrisDecay (%) NativeEuc Leaf No 80 DebrisDecay (%) NativeEuc RootCoarse Yes 40 DebrisDecay (%) NativeEuc RootCoarse No 10 DebrisDecay (%) NativeEuc RootFine Yes 100 DebrisDecay (%) NativeEuc RootFine No 100 DebrisDecay (%) Rainforest Stem Yes 5 DebrisDecay (%) Rainforest Stem No 5 DebrisDecay (%) Rainforest Branch Yes 10 DebrisDecay (%) Rainforest Branch No 10 DebrisDecay (%) Rainforest Bark Yes 50 DebrisDecay (%) Rainforest Bark No 50 DebrisDecay (%) Rainforest Leaf Yes 80 DebrisDecay (%) Rainforest Leaf No 80 DebrisDecay (%) Rainforest RootCoarse Yes 40 DebrisDecay (%) Rainforest RootCoarse No 10 DebrisDecay (%) Rainforest RootFine Yes 100 DebrisDecay (%) Rainforest RootFine No 100

Page 359 of 365

Debris decay rates (cont.)

CategoryID SppGpID BioComponentID bDecomposable Val DebrisDecay (%) HardwoodPlantation Stem Yes 5 DebrisDecay (%) HardwoodPlantation Stem No 5 DebrisDecay (%) HardwoodPlantation Branch Yes 10 DebrisDecay (%) HardwoodPlantation Branch No 10 DebrisDecay (%) HardwoodPlantation Bark Yes 50 DebrisDecay (%) HardwoodPlantation Bark No 50 DebrisDecay (%) HardwoodPlantation Leaf Yes 80 DebrisDecay (%) HardwoodPlantation Leaf No 80 DebrisDecay (%) HardwoodPlantation RootCoarse Yes 40 DebrisDecay (%) HardwoodPlantation RootCoarse No 10 DebrisDecay (%) HardwoodPlantation RootFine Yes 100 DebrisDecay (%) HardwoodPlantation RootFine No 100 DebrisDecay (%) SoftwoodPlantation Stem Yes 10 DebrisDecay (%) SoftwoodPlantation Stem No 10 DebrisDecay (%) SoftwoodPlantation Branch Yes 15 DebrisDecay (%) SoftwoodPlantation Branch No 15 DebrisDecay (%) SoftwoodPlantation Bark Yes 50 DebrisDecay (%) SoftwoodPlantation Bark No 50 DebrisDecay (%) SoftwoodPlantation Leaf Yes 100 DebrisDecay (%) SoftwoodPlantation Leaf No 100 DebrisDecay (%) SoftwoodPlantation RootCoarse Yes 50 DebrisDecay (%) SoftwoodPlantation RootCoarse No 20 DebrisDecay (%) SoftwoodPlantation RootFine Yes 100 DebrisDecay (%) SoftwoodPlantation RootFine No 100 DebrisDecay (%) Understorey Stem Yes 5 DebrisDecay (%) Understorey Stem No 5 DebrisDecay (%) Understorey Branch Yes 10 DebrisDecay (%) Understorey Branch No 10 DebrisDecay (%) Understorey Bark Yes 50 DebrisDecay (%) Understorey Bark No 50 DebrisDecay (%) Understorey Leaf Yes 80 DebrisDecay (%) Understorey Leaf No 80 DebrisDecay (%) Understorey RootCoarse Yes 40 DebrisDecay (%) Understorey RootCoarse No 10 DebrisDecay (%) Understorey RootFine Yes 100 DebrisDecay (%) Understorey RootFine No 100

Page 360 of 365

Relative decay rates of dry sclerophyll (Low Eucalypt Forest) and wet sclerophyll (Tall Eucalypt Forest) debris

CategoryID Val bDeviate Wet_Euc_Decay_multiplier 1 TRUE Dry_Euc_Decay_multiplier 0.7 TRUE

Debris density and losses during decomposition

CategoryID SppGpID BioComponentID Val DebrisDensity ( t/m3) All Stem 0.335 DecayedDecompDebrisToCO2 (%) All All 77 DecayedResistDebrisToCO2 (%) All All 39

Stag decomposition rates

CategoryID SppGpID bFireKilled bDecomposable Val StagDecay (%) All TRUE TRUE 1.5 StagDecay (%) All TRUE FALSE 1.5 StagDecay (%) All FALSE TRUE 1.5 StagDecay (%) All FALSE FALSE 1.5

Stag half life

CategoryID SppGpID bFireKilled maxHalfLifeYrs StagHalfLife (years) All TRUE 30 StagHalfLife (years) All FALSE 10

Losses in fire

Debris combustion in fire

CategoryID SppGpID BioComponentID Val Debris_BIGkilled_multiplier All Stem 2 Debris_BIGkilled_multiplier All Branch 2 Debris_BIGkilled_multiplier All Bark 3 Debris_BIGkilled_multiplier All Leaf 5 Debris_BIGkilled_multiplier All RootCoarse 0 Debris_BIGkilled_multiplier All RootFine 0 PropBurntDebrisVolatilized (%) All All 95

Page 361 of 365

Soil loss in fire

CategoryID Val bDeviate HUMvolatilize_BIGkilled_mult 0.2 TRUE

Debris combustion in fire

CategoryID SppGpID Val PropBurntStagsVolatilized (%) All 80 Stags_BIGkilled_multiplier All 1

Wood products

Recovery of log types from harvested trees

CategoryID ProcessID SppGpID ProductID Val ProcessToProductRates (%) StemBreakDown NativeEuc CWD 0 ProcessToProductRates (%) StemBreakDown NativeEuc Poles 0 ProcessToProductRates (%) StemBreakDown NativeEuc SawLog 14 ProcessToProductRates (%) StemBreakDown NativeEuc VeneerLog 8 ProcessToProductRates (%) StemBreakDown NativeEuc ChipLog 78 ProcessToProductRates (%) StemBreakDown Rainforest CWD 0 ProcessToProductRates (%) StemBreakDown Rainforest Poles 0 ProcessToProductRates (%) StemBreakDown Rainforest SawLog 14 ProcessToProductRates (%) StemBreakDown Rainforest VeneerLog 0 ProcessToProductRates (%) StemBreakDown Rainforest ChipLog 86 ProcessToProductRates (%) StemBreakDown HardwoodPlantation CWD 0 ProcessToProductRates (%) StemBreakDown HardwoodPlantation Poles 0 ProcessToProductRates (%) StemBreakDown HardwoodPlantation SawLog 38 ProcessToProductRates (%) StemBreakDown HardwoodPlantation VeneerLog 0 ProcessToProductRates (%) StemBreakDown HardwoodPlantation ChipLog 62 ProcessToProductRates (%) StemBreakDown SoftwoodPlantation CWD 0 ProcessToProductRates (%) StemBreakDown SoftwoodPlantation Poles 1 ProcessToProductRates (%) StemBreakDown SoftwoodPlantation SawLog 47 ProcessToProductRates (%) StemBreakDown SoftwoodPlantation VeneerLog 5 ProcessToProductRates (%) StemBreakDown SoftwoodPlantation ChipLog 47 ProcessToProductRates (%) StemBreakDown Understorey CWD 0 ProcessToProductRates (%) StemBreakDown Understorey Poles 0 ProcessToProductRates (%) StemBreakDown Understorey SawLog 10 ProcessToProductRates (%) StemBreakDown Understorey VeneerLog 0 ProcessToProductRates (%) StemBreakDown Understorey ChipLog 90

Page 362 of 365

Sawlog recovery rates applied to equation estimating sawlog % from tree diameter

CategoryID SppGpID Val SawlogScaler All 0.26 SawlogScaler HardwoodPlantation 0.5 SawlogScaler SoftwoodPlantation 0.93 SawlogScalerPrivate All 0.2 SawlogScalerPrivate NativeEuc 0.1 SawlogScalerPrivate HardwoodPlantation 0.1 SawlogScalerPrivate SoftwoodPlantation 0.93

Destinations of harvested logs and wood products

CategoryID ProductID ProcessID Val ProductToProcessRates (%) Stem StemBreakDown 100 ProductToProcessRates (%) Poles Export 13 ProductToProcessRates (%) Poles SolidDecay 87 ProductToProcessRates (%) SawLog Saw 81 ProductToProcessRates (%) SawLog Export 19 ProductToProcessRates (%) VeneerLog Peel 81 ProductToProcessRates (%) VeneerLog Export 19 ProductToProcessRates (%) ChipLog Chip 100 ProductToProcessRates (%) SawnTimber SolidDecay 100 ProductToProcessRates (%) Veneer Reconstitute 24 ProductToProcessRates (%) Veneer Export 76 ProductToProcessRates (%) Chips Reconstitute 1 ProductToProcessRates (%) Chips Paper 27 ProductToProcessRates (%) Chips Export 72 ProductToProcessRates (%) Chips PaperDecay 0 ProductToProcessRates (%) Panels Export 4 ProductToProcessRates (%) Panels BoardDecay 96 ProductToProcessRates (%) Paper Export 26 ProductToProcessRates (%) Paper PaperDecay 74 ProductToProcessRates (%) BioFuel Burn 100

Page 363 of 365

Recovery of finished products from log and primary product processing

CategoryID ProcessID SppGpID ProductID Val ProcessToProductRates (%) Saw All SawnTimber 40 ProcessToProductRates (%) Saw All Chips 35 ProcessToProductRates (%) Saw All BioFuel 25 ProcessToProductRates (%) Peel All Poles 10 ProcessToProductRates (%) Peel All Veneer 45 ProcessToProductRates (%) Peel All Chips 25 ProcessToProductRates (%) Peel All BioFuel 20 ProcessToProductRates (%) Chip All Chips 100 ProcessToProductRates (%) Reconstitute All Panels 80 ProcessToProductRates (%) Reconstitute All BioFuel 20 ProcessToProductRates (%) Paper All Paper 70 ProcessToProductRates (%) Paper All BioFuel 30 ProcessToProductRates (%) Export All CO2 100 ProcessToProductRates (%) Burn All CO2 100

Wood product half lives

CategoryID ProductID Val ProductHalfLife (years) SolidDecayC 35 ProductHalfLife (years) PaperDecayC 2 ProductHalfLife (years) BoardDecayC 25

Destinations of losses from product pools

CategoryID ProcessID SppGpID ProductID Val ProcessToProductRates (%) Export All CO2 100 ProcessToProductRates (%) Burn All CO2 100 ProcessToProductRates (%) SolidDecay All SolidDecayC 100 ProcessToProductRates (%) PaperDecay All PaperDecayC 100 ProcessToProductRates (%) BoardDecay All BoardDecayC 100

Page 364 of 365

Miscellaneous Plantation previous land land-use, rotation length and thinning likelihood CategoryID SppGpID Val Probability of 1R All 0.5 Probability of 1R HardwoodPlantation 0.75 Probability of 1R SoftwoodPlantation 0.26 Probability prev crop grass|1R All 0.2 Probability prev crop grass|1R HardwoodPlantation 0.46 Probability prev crop grass|1R SoftwoodPlantation 0.36 Probability of long rotation All 0 Probability of long rotation HardwoodPlantation 0.75 Probability of long rotation SoftwoodPlantation 0.9 Length short rotation (years) All 12 Length long rotation (years) All 28 Length long rotation (years) HardwoodPlantation 25 Length long rotation (years) SoftwoodPlantation 30 PrivateThinningProbMult All 1 PrivateThinningProbMult HardwoodPlantation 0.07 PrivateRotationLenMult All 1 PrivateRotationLenMult HardwoodPlantation 0.07

Growth rate parameters and multipliers CategoryID SppGpID Val PropOfMaxOGVol % All 100 PropOfMaxOGVol % Rainforest 50 PropOfMaxOGVol % Understorey 50 Growth multiplier All 1 Growth multiplier NativeEuc 1 Growth multiplier HardwoodPlantation 1 Growth multiplier SoftwoodPlantation 1

Emissions from fossil fuel use CategoryID SppGpID Val ProcessingFuelCO2 (kg/m3) NativeEuc 33 ProcessingFuelCO2 (kg/m3) Rainforest 33 ProcessingFuelCO2 (kg/m3) HardwoodPlantation 28 ProcessingFuelCO2 (kg/m3) SoftwoodPlantation 23 ProcessingFuelCO2 (kg/m3) Understorey 33

Page 365 of 365