<<

RSPO No consultancy: high cover countries Consultancy report on definitions and recommendations to the RSPO

June 2018

RSPO No deforestation consultancy: High Forest Cover Countries

About The Proforest Initiative The Proforest Initiative supports governments, companies, civil society organisations and other stakeholders with responsible production and sourcing of forest products and agricultural commodities such as timber, , soy, beef and sugar.

The Proforest Initiative is part of the not-for-profit Proforest Group, whose mission is to help people produce and source natural resources sustainably. We work with others to transform commodity supply chains and sectors through developing awareness about sustainability, helping to generate commitment to better practice, supporting implementation of these commitments in practice and working across sectors and scales to increase the positive impact.

The Proforest Initiative was established to use our expertise and experience to contribute to this goal more effectively through long-term programmes. We support multistakeholder initiatives, build local capacity to deliver improved practices and develop mechanisms to accelerate the practical implementation of sustainability in agricultural and forest landscapes. We develop tools, guidance and practices that can be adopted by producers, supply chain organisations, governments and civil society, and build long-term programmes in partnership with other organisations.

The Proforest Initiative team is international and multilingual and comes from a wide variety of backgrounds, including industry, academia and civil society. This allows us to work comfortably with diverse organisations in a range of cultures. We have in-house knowledge of more than 15 languages, including English, Bahasa Indonesia, French, Mandarin, Portuguese and Spanish.

The Proforest Initiative is a registered charity (non-profit organisation) and is overseen by an independent board of trustees.

For this report, your contact person is: Mike Senior [email protected]

The Proforest Initiative South Suite, Frewin Chambers, Frewin Court, Oxford OX1 3HZ United Kingdom

E: [email protected] T: +44 (0) 1865 243439 W: www.proforest.net

The Proforest Initiative is a registered charity in England and Wales (Charity no. 1137523) and a company registered in England and Wales (Company no. 7293440).

2

RSPO No deforestation consultancy: High Forest Cover Countries

Table of contents 1 Introduction & background ------6 1.1 Consultancy objectives ------6 2 Methodology ------7 3 Framing RSPO’s vision on deforestation ------10 3.1 RSPO’s vision and goals------10 3.2 No deforestation and palm oil’s image ------12 3.3 Implementing ND in practice ------13 3.4 The high forest cover debate ------13 3.5 RSPO’s role in tackling deforestation ------19 4 Definitions------20 4.1 Forest ------20 4.2 Carbon and GHGs ------25 4.3 High forest cover ------27 4.4 Spatial scale ------35 4.5 Development status and poverty ------39 4.6 Statistics on potential high forest cover countries ------41 5 Approaches and interventions ------44 5.1 Lessons from Forest Stewardship Council ------44 5.2 Defining which areas to protect ------45 5.3 Assessment requirements ------47 5.4 Analysis of options discussed by RSPO P&C TF ------47 6 Scenarios/options ------48 6.1 Guiding principles ------48 6.2 Defining HFC countries ------49 6.3 HFC approaches ------51 Annex 1: RSPO TF5 HFC approach justification ------54 Annex 2. Profiles of selected HFC countries ------61 Annex 3: High Forest Cover case studies synthesis ------63

3

RSPO No deforestation consultancy: High Forest Cover Countries

List of Acronyms BAU Business-As-Usual CDM CLEAN DEVELOPMENT MECHANISM CGF Consumer Goods Forum

CPO Crude Palm Oil CSPO Certified Sustainable Palm Oil ESIA Environmental & Social Impact Assessment

FMU Unit FPIC Free, prior and informed consent FSC Forest Stewardship Council

GHG Greenhouse Gas HCSA High Carbon Stock Approach HCS High Carbon Stock

HCV High Conservation Value HCVRN HCV Resource Network HCVNI HCV National Interpretation

HFC High Forest Cover HFCC High Forest Cover Country HFCL High Forest Cover Landscape

HFLD High-Forest, Low-Deforestation IFL Intact Forest Landscape LCA Life Cycle Analysis

LDF Low Density Forest LULUCF Land Use, Land Use Change and MDF Medium Density Forest

ND No Deforestation NGO Non-governmental organisation NDPE No deforestation, no Peat, no Exploitation

NPP RSPO’s New Planting Procedure

P&C RSPO’s Principles & Criteria for the production of sustainable palm oil

4

RSPO No deforestation consultancy: High Forest Cover Countries

PNG Papua New Guinea

REDD+ Reducing Emissions from Deforestation and RSPO Roundtable on Sustainable Palm Oil RSPO ND RSPO P&C Review Task Force’s “No deforestation” sub-group

SiHA Simple Historical Approach SpHA Spatial Historical Approach TF RSPO’s P&C Review Task Force

UNFCCC United Nations Framework Convention on Climate Change VCS Voluntary Carbon Standard WIA Welfare Impact Assessment

WRI World Resources Institute YRF Young Regenerating Forest ZND Zero Net Deforestation

5

RSPO No deforestation consultancy: High Forest Cover Countries

1 Introduction & background The Roundtable on Sustainable Palm Oil (RSPO) Principles and Criteria Review Task Force is currently reviewing the P&C with the aim of finalising the new P&C by November 2018. A critical issue for RSPO and its stakeholders in the P&C review is the issue of “No deforestation”, and the P&C Review Task Force has agreed to include requirements for “No deforestation” under Criterion 7.3.

In the draft of the revised P&C for the first public consultation in August 2017, it was stated that “there is clear intent from the RSPO P&C Review Task Force to include in addition, requirements for “no deforestation” potentially using the HCSA [High Carbon Stock Approach] toolkit and methodology as a reference point.”

At present the P&C Review Task Force has not reached consensus on the methodology or wording that will be included in the new standard to implement “No deforestation”. In particular, the Task Force recognises the following:

- The current HCSA methodology and toolkit applies to fragmented tropical moist forest landscapes, and could be adopted for these contexts in the RSPO P&C, - The need to look at how inclusion of a “no deforestation” requirement would be applied in “High Forest Cover Landscapes or Nations, that are at high risk of deforestation”, and - “Preventing responsible oil palm development in such areas may result in the development of other crops instead, defeating the purpose of RSPO.”

To try and address the above challenges the RSPO released a call for tender for two consultancy projects on “Input to development of No Deforestation element of RSPO P&C Criterion 7.3”. The proposed objective of these consultancy projects was to provide technical support to the RSPO Principles and Criteria Review Task Force on the No Deforestation (ND) element of Criterion 7.3. The call for tender was divided into two parts: A) High Forest Cover Countries, and B) Application to small and medium growers. This report is the output for Part A: High Forest Cover Countries. 1.1 Consultancy objectives The Terms of Reference for the consultancy specified that the work will include the following:

• Desk study, to include review of relevant peer-reviewed scientific literature around High Forest Cover (HFC) countries and landscapes; • Collate information, case studies and experience relevant to HFC countries incorporating information from HCSA studies, (e.g. Olam, New Britain Palm Oil, and other companies that have conducted HCSA studies); • Propose definitions of High Forest Cover and identify the countries where HFC process might be applied based on different methodology approaches (carbon thresholds, % forest cover, % forest cover vs arable land, poverty index); • Map out a possible country level process; and • Suggest how this links to support the development of NI processes.

6

RSPO No deforestation consultancy: High Forest Cover Countries

2 Methodology This work was conducted in the following four phases. The methodology was developed based on the ToR prepared by RSPO and the consultants’ experiences from other similar processes. Throughout the work the consultants had regular contact with the RSPO P&C Review Task Force’s (TF) “No deforestation” Interim Group (NODIG) through calls and via email to support the process and ensure the work remained focused on the required deliverables.

2.1.1 Phase 1: Document review and targeted consultation The first phase collated an overview of the current status of knowledge on HFCL definitions, social dimensions, case studies, implementation challenges and policies, based on:

- A desk-based literature review of relevant peer-reviewed literature, - Targeted consultation with a limited number of selected experts, e.g. HCSA members, companies operating in HFCLs, scientists, social and environmental NGOs and policy makers, - Collation of relevant case studies of proposed palm developments and HCS assessments in HFCLs, e.g. Olam, New Britain Palm Oil, Sime Darby, Golden Veroleum, Sipef, and - Existing policies or approaches for avoiding or limiting deforestation in HFCLs, e.g. HCS+, FSC experience on IFLs (motions 12 and 65)

The following stakeholders were directly consulted during this phase: Name Organisation Sector

Reuben Blackie IDH NGO

Alwi Hafiz GVL Grower

Sander van den Ende NBPOL Grower

Philippa Atkinson Independent Social and economic researcher and consultant

Jan Pierre Jarrin Peters Oleana Miller and refiner

Anders Lindhe HCVRN NGO/technical organisation

Mike Zrust Daemeter Consulting Technical organisation/consultancy

Grant Rosoman was also contacted, as a representative, to request a consultation call, but the call did not proceed due to scheduling issues.

7

RSPO No deforestation consultancy: High Forest Cover Countries

In addition, the HCSA High Forest Cover Landscape WG co-chairs were contacted to request relevant information for the literature review. The HCSA co-chairs and Executive Director (Grant Rosoman, Perpetua George and Judy Rodrigues) were also contacted at the start of the consultancy to notify them about the work.

As well as direct consultations, the consultants were able to draw on previous conversations with a range of private sector, NGO and government stakeholders through initiatives such as the HCSA, Africa Palm Oil Initiative, The Forest Dialogue’s “Understanding Deforestation Free” Gabon chapter amongst others. 2.1.2 Phase 2: Development of scenarios Based on the results of Phase 1, a series of potential scenarios were developed, with a strength-weakness analysis. These scenarios cover the following:

- Definitions for High Forest Cover based on various parameters, including carbon threshold, % forest cover, forest quality, forest area/patch size, - Definitions of landscape and/or scale, - Definitions and thresholds for HFC countries/jurisdictions based on income or development status, - Identification of potential HFC countries, jurisdictions or landscapes, - Possible options for an RSPO HFC approach, to include: o Pre-conditions for development (including social requirements), o Scale of implementation (e.g. concession/management unit, landscape or jurisdiction), o Roles of/links to RSPO National Interpretation processes, o Minimum outputs/requirements, e.g. area or % to be protected, social outcomes, o Due diligence and assessment process, e.g. use of HCV and HCS assessments - Gaps and outstanding questions - Proposed preliminary timeline and process for finalising RSPO’s HFC approach 2.1.3 Phase 3: Presentation & consultation on scenarios and definitions Scenarios developed in Phase 2 were presented to the RSPO P&C Review TF and RSPO NODIG during the TF5 meetings in on 14-16th May 2018 (as well as during pre-meetings with the NODIG). Draft findings and recommendations were shared with the NODIG twice prior to the TF5 meetings during update calls.

The objective of Phase 3 was to receive TF and specifically ND sub-group member feedback on potential definitions and scenarios to feed into Phase 4.

A summary of calls and meetings held with the RSPO during this consultancy is provided here: Date (2018) Meeting

29th March Kick-off call with NODIG

19th April Update call with NODIG

8

RSPO No deforestation consultancy: High Forest Cover Countries

4th May Update call with NODIG to present initial findings & recommendations prior to meeting with HCSA on 11th May

13th May Preparatory meeting with NODIG to present results of Phase 2

14-16th May Presentation of Phase 2 results to RSPO P&C TF & technical support to NODIG

2.1.4 Phase 4: Finalisation of scenarios and proposed process for formalizing an RSPO HFC approach The final phase incorporated feedback received from the TF and ND sub-group, to revise proposed definitions and scenarios. The objective of this was to refine definitions and scenarios into a smaller set that can practically be taken forward by the RSPO P&C TF for wider consultation and finalisation.

The key outputs of this phase were the refined definitions and scenarios, as well as a proposed process to be followed by the RSPO after May for formalising and finalising a RSPO HFC approach.

Note: the above methodology was developed at the start of the project as a proposed approach. In practice, the work required a more iterative discussion with the NODIG throughout, rather than a strict 4 phase process. As a result, the final outputs produced included separate outputs not covered in the original scope and omitted some areas deemed less relevant by the NODIG.

9

RSPO No deforestation consultancy: High Forest Cover Countries

3 Framing RSPO’s vision on deforestation Before diving into the technical details of definitions or discussing potential approaches to tackling deforestation, it is important to take stock of RSPO’s wider vision and goals and how these may inform the RSPO’s vision on tackling deforestation. In turn this vision must be informed by market signals, wider commitments and initiatives in the global palm sector. 3.1 RSPO’s vision and goals As a voluntary private sector initiative, the RSPO was established in response to consumer and market demands for sustainability in the sector, to be a tool that allowed consumers to reliably source sustainable palm oil. Although RSPO is a multi-stakeholder, membership organisation, ultimately as a voluntary initiative it must respond to consumer demands on sustainability, including on deforestation.

However, through its vision and mission the RSPO seeks to do more than simply respond to consumer demands, with the vision to: “transform markets to make sustainable palm oil the norm”.

This bold vision is also supported by the RSPO’s new Theory of Change (ToC) which specifies as goals broadly focused around People, Planet and Prosperity of “Resilient & Healthy Landscapes & Communities”, and “Green & Inclusive Growth”. The ToC also specifies desired impacts as shown in Figure 1, as well as more tangible desired long-term outcomes. The outcomes of greatest relevance to the concept of “No deforestation” are as follows: - protected, - services enhanced, - Land degradation neutrality, and - Sustainable land use planning.

10

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 1. RSPO’s Theory of Change, vision, goals and the intended impacts.

The essence of RSPO’s vision was also captured in the revised P&C draft for public consultation in 2017 in wording on No deforestation for Criterion 7.3, which stated that RSPO wants to “transform all markets worldwide” by adopting a “highly inclusive approach that is accessible to less developed high forest cover areas; and … the free, prior, informed decisions of customary land rights holders”. The draft text for 7.3 stated that RSPO’s aim was “to achieve forest landscape conservation and manage conservation areas according to RSPO requirements”.

In summary and most critically, RSPO needs to ensure that it not only avoids deforestation caused by oil palm, but needs to actively contribute to reducing and ultimately stopping deforestation for oil palm production. In practice, to transform global markets means taking into account:

1. Future, global CPO consumption trends as well as trends in consumption of RSPO-certified CPO (CSPO). By continent, the main CPO consumers are Asia (53%, of which 17% in India and 9% in China), Europe (24%), Africa (16%) and North America (5%).1 However, by far the main markets for CSPO are Europe and North America (30% of global production). Future growth in CPO demand is set to continue increasingly dramatically with

1 Imports by value, taken from https://atlas.media.mit.edu/en/profile/hs92/1511/ 11

RSPO No deforestation consultancy: High Forest Cover Countries

global trade in palm oil to increase by 24% by 2026,2 and demand is expected to be highest in Africa and Asia3, and 2. Future, global CPO production trends. Numerous palm producer and high forest countries have ambitious growth plans for the sector, e.g. the PNG government plans to increase production from 0.5 to 1.5 million tonnes of CPO by 2030 and Gabon to develop a further 300,000 ha of palm (under the “Gabon Emergent” agenda). Whether these projections are met is another question, but what is undeniable is that large areas have been identified or allocated for palm development (540,000 ha in Liberia4, at least 500,000 ha across the Congo Basin5).

If RSPO is to have global influence and transform global markets it needs to find a balance that:

- Satisfies sustainability demands in key current markets, such as Europe and North America, - Ensures greater uptake in Asian and African markets, and - Supports sustainable growth in new production areas. 3.2 No deforestation and palm oil’s image “Brand” palm oil continues to struggle in European and North American markets as a result of NGO campaigns as well as lobbying by European food and agri- business groups. Whatever the motives for the campaigns, criticism has focused on claims of deforestation, land-grabbing and worker exploitation in the industry. This is wrapped up in criticism, by international social and environmental NGOs, of palm’s ‘large-scale model’ as one that contributes little to improving the livelihoods of rural populations.

This has led to a large number of companies in the palm supply chain making commitments to No Deforestation, Peat or Exploitation (NDPE) over the past 5 years. In practice these mean committing to avoiding conversion of, and in some cases protecting, High Conservation Value (HCV) and High Carbon Stock (HCS) areas. Furthermore, the EU Parliament has voted to ban palm oil in biofuels by 2020.

Whilst the demand for ‘no deforestation’ may have originated from markets and consumers in Europe and North America, the scope of the commitments made is now much wider. For example, the Consumer Goods Forum (CGF) has also been an important initiative for change and represents 400 companies across 70 countries and CGF has committed to ‘zero net deforestation’ by 2020, and in its

2 USDA Agricultural Projections to 2026

3 Pacheco P, Gnych S, Dermawan A, Komarudin H and Okarda B. 2017. The palm oil global value chain: Implications for economic growth and social and environmental sustainability. Working Paper 220. Bogor, Indonesia: CIFOR.

4 HCS+ Consulting Study 17: Palm oil in Liberia

5 Earthsight. 2018. The Coming Storm 12

RSPO No deforestation consultancy: High Forest Cover Countries

Sustainable Palm Oil Sourcing Guidelines it expects members to protect HCV areas and encourages them to protect HCS areas.

In the palm sector in Indonesia and , company groups with NDPE policies cover 74 percent of the total refinery capacity in these countries and 85% of global refining capacity.6 The upshot of this is that these companies should be aiming for all of this production to be deforestation-free by 2020. 3.3 Implementing ND in practice To date many companies have struggled to fully implement their ND commitments. The High Carbon Stock Approach (HCSA) was developed as a tool for supporting companies to implement ND in practice and has been used in allowing large companies to identify HCS forest to avoid (and in theory protect) in fragmented moist tropical forest landscapes. HCSA decided not to adapt its toolkit for High Forest Cover Landscapes (HFCLs) – recognising that it is a toolkit for ‘no deforestation’, and so could not sanction deforestation. HCSA will only allow some limited development for ‘legacy cases’ and will focus on alternative development options in HFCLs. Therefore, HCSA still faces several key implementation gaps:

1. How do you implement ND in smallholder and high forest cover contexts? 2. How do you avoid ‘leakage’ and ensure that are not only avoided by companies, but are actually protected from deforestation by other actors?

It is important to note that most NDPE commitments also include commitments to improving rural livelihoods and ensuring smallholder inclusion – two factors often in conflict with ND, especially in high forest cover contexts. This conflict has not been well communicated in consumer markets, with deforestation messaging dominant over social messaging, and there are as yet no widely accepted tools for how to implement these different commitments on the ground. Although some pilot projects are now underway to grapple with this at a landscape-level, few consumers are aware of the challenges, resources and time involved in simultaneously and quickly implementing ND, smallholder inclusion and rural livelihood development on the ground. 3.4 The high forest cover debate Advocates for sustainable palm oil are broadly in agreement about the need for ND in fragmented landscapes, but there remains significant debate about high forest cover (HFC) contexts. The debate has raged for a number of years through the HCS convergence process and more recently in the HCSA’s HFC Landscapes Working Group.

On the surface, the argument is polarised between:

6 http://www.aidenvironment.org/publication/4927/ 13

RSPO No deforestation consultancy: High Forest Cover Countries

- On one hand the urgent need to stop deforestation globally and protect the world’s last remaining large blocks of forest along with their biodiversity and the livelihoods of peoples dependent on them, and - On the other, the view that some deforestation is required to support economic and rural development in what are typically some of the world’s poorest countries or areas.

In reality, there are some more nuanced concerns on both sides of the debate that have arguably not been objectively or impartially considered – and as a result have not allowed the debate to advance. These include:

- A slippery slope: Many NGOs fear that allowing some limited deforestation in HFCLs will set a precedent or open a backdoor to allow further deforestation, e.g. once mills are established then opportunistic clearance will follow, - Business as usual & unintended consequences: Many feel that a strict ‘gross zero deforestation’ requirement for palm oil in HFCLs will have unintended consequences of even higher deforestation resulting from leakage to other crops or actors. It is felt that such a hard-line approach ignores the business-as-usual trajectory in most HFC countries. Unintended consequences may result from: o Pushing responsible companies away from these high-risk areas, leading to concessions being given to less reputable companies o Ignoring other primary deforestation drivers in HFCLs. For example, annual crops cause 66% of deforestation in Central Africa7 and in PNG 48% of deforestation is caused by commercial and 46% by subsistence agriculture (compared to 1.6% for agriculture and ). In most cases, these threats are only increasing with logging ongoing and subsistence agriculture increasing its impact as rural populations grow. - Community expectations: There has been a tendency to oversimplify the social aspects of deforestation in HFCLs. For example, on one side portrayals of all local communities objecting to oil palm development, versus claims that communities are desperate for jobs. The reality is often highly variable both within individual communities and across geographies: o In instances where growers already have some footprint in the area, it is not uncommon for communities (or at least some community members) to expect some level of palm development due to desires for jobs and the absence of other immediate alternatives. In the context of PNG in particular, where 97% of land is customarily owned, many clans actively want companies to develop oil palm on their land because they receive a land rent and share of the profits from the company.

7 Mosnier, et al. 2017. CoForTips Congo basin forests: tipping points for biodiversity conservation and resilience. Final Report (La modélisation des changements d’utilisation des terres dans les pays d’frique Centrale 2000-2030). 14

RSPO No deforestation consultancy: High Forest Cover Countries

o Ideas that all communities want to or even are able to live exclusively off traditional forest livelihoods and/or subsistence agriculture does not reflect the reality of all situations. In fact, in some cases such ideas may even be a threat to communities’ health or food security given that palm growers are often only allocated land by government after forest has been logged and degraded, and with growing local populations putting increasing pressure on forest resources, “traditional” livelihoods may no longer be able to provide reliably healthy or diverse diets. It could be argued that the presence of a responsible company in such situations could help to support a) a transition to a cash income, b) help communities to secure their land tenure of forests and gardens, and c) improve productivity of food crop areas or protect forest from illegal encroachment. - Rural livelihoods: Oil palm is often touted by companies as one of the best options for rural development where rural populations are unskilled as it provides jobs and other social infrastructure. Conversely, international NGOs often argue that the ‘large-scale plantation model’ does not delivery adequate social benefits to rural populations, with evidence of declining food security and low or inadequate wages. - Messaging: A huge amount of attention has been given to ND commitments in international media and by NGO campaigns, and many companies are bought into this. As a result, there is an apparent reluctance to adjust or nuance this message to communicate the challenge of HFCLs because of fears it may undermine credibility.

The above concerns & unknowns can be partially answered by an analysis of existing evidence (where available). This is synthesised in the table below:

Concern Evidence available or justification

“Slippery slope” • Studies have shown that development in the form of roads or palm oil mill construction frequently trigger further deforestation8 – suggesting that establishment of palm operations in HFCLs (with accompanying mills and roads) may lead to some associated deforestation. However, much of this seems highly likely, if not more likely, to happen whether or not RSPO- certified operations are present. RSPO can be a tool to reduce deforestation compared to business-as-usual (BAU), as shown by a recent study9.

• Other initiatives to limit deforestation are far more advanced than when palm development first started in Kalimantan and Sumatra 40+ years ago,

8 e.g. Laurance, W. F., Albernaz, A. K., Schroth, G. , Fearnside, P. M., Bergen, S. , Venticinque, E. M. and Da Costa, C. 2002. Predictors of deforestation in the Brazilian Amazon. Journal of Biogeography, 29: 737-748. doi:10.1046/j.1365-2699.2002.00721.x ; Tomita, Atsushi.2017. "Land Change History of Oil Palm in Northern Bengkulu Province, Sumatra Island, Reconstructed from Landsat Satellite Archives" . CUNY Academic Works. http://academicworks.cuny.edu/gc_etds/1857

9 Carlson et al. 2017. PNAS. Effect of oil palm sustainability certification on deforestation and fire in Indonesia 15

RSPO No deforestation consultancy: High Forest Cover Countries

for example, with improved government regulation and also government/jurisdictional recognition of RSPO.

• Tropical countries that have managed to somewhat stabilise deforestation have a balance of approximately 50% forest in their territories, and have developed diversified economies. For example, oft-sighted best in class example Costa Rica in fact only has 50% forest cover. Malaysia’s efforts to stabilise forest cover at 50% of the territory have been supported by fast economic growth driven by oil & gas, oil palm and technology leading to a more educated and urbanised population.

Unintended • Major development is planned in many HFCC/Ls for forestry or agriculture, consequences & e.g. 3 million ha of land allocated to concessions in West Papua/Papua10, leakage approximately 25% of Congo Basin is under logging concessions11 –much of which is subject to destructive, poorly regulated logging.12 Non-RSPO members/non-NDPE committed companies already clearing in Papua (especially in Merauke Regency, Papua where a huge chunk has been allocated for logging, pulp plantations or oil palm; Neville Kemp pers. comm.).

• Although NDPE-committed companies may touch up to 80% of global palm production13, the major current and projected future demand14 for palm oil (including biofuels) from Indonesia), Malaysia, India, China and also sub-Saharan Africa means that market drivers will likely see new actors emerging or efforts to vertically integrate in order to avoid trading through NDPE companies – potentially leading to a split market.15 Failing this, it’s likely that growth in production of other oil crops such as soya oil will meet the demand (demand for which continues to increase)14 – leading to even more pressure on land.

• Other commodities or drivers of deforestation and degradation are likely to fill the void at a production level as companies or communities look for income sources, e.g. pulp plantations, subsistence agriculture (with growing populations), etc. Indeed other deforestation drivers like

10 https://chainreactionresearch.com/wp-content/uploads/2017/04/palm-oil-stranded-land-size- equals-ten-million-football-fields-crr-170407.pdf

11 WRI Congo Basin Forest Atlas; and https://globalforestatlas.yale.edu/congo/forests-and- logging/logging

12 For example: https://www.reuters.com/article/us-congodemocratic-environment- idUSKBN0OJ00E20150603

13 https://chainreactionresearch.com/report/unsustainable-palm-oil-faces-increasing-market- access-risks-ndpe-sourcing-policies-cover-74-percent-of-southeast-asias-refining-capacity/

14 USDA 2017. USDA Agricultural Projections to 2026

15 Earth Innovation Institute. 2016. Making Corporate Deforestation Pledges Work. earthinnovation.org 16

RSPO No deforestation consultancy: High Forest Cover Countries

subsistence agriculture are currently primary drivers in Africa and Papua New Guinea.16

Community • Community support for or resistance to oil palm is highly context expectations dependent. 17 Much coverage is polarised on the issue suggesting either that palm has universally negative social impacts or that it is a silver bullet for rural development. The reality is that 40% of palm is produced by smallholders, many of whom have significantly improved their quality of life as a result, however, in many cases establishment of large plantations without FPIC has led to negative impacts on community rights holders.

• It is often stated that local communities are the best forest protectors,18 and this is undoubtedly the case for certain indigenous groups where land rights are clear and livelihoods are still derived from the forest. However, in many cases communities (sometimes local, sometimes migrants) are driving deforestation either for subsistence agriculture or smallholder commodity agriculture.19

• HCV-HCS assessments often show ‘traditional livelihoods’ less reliable due to logging and population growth

Rural livelihoods • It is often argued that palm oil production does not support rural poverty alleviation20, rather leading to food insecurity, loss of land rights etc. However, there are social benefits from palm development if carefully managed, e.g. with adequate wages and social infrastructure for workers, provisions to ensure food security, through smallholder programmes, etc.21 Once again results are context specific and overgeneralising is unhelpful.

16 Kissinger, G., M. Herold, V. De Sy. 2012. Drivers of Deforestation and Forest Degradation: A Synthesis Report for REDD+ Policymakers. Lexeme Consulting, Vancouver Canada, August 2012.

17 E.g. Li TM. 2015. Social impacts of oil palm in Indonesia: A gendered perspective from West Kalimantan. Occasional Paper 124. Bogor, Indonesia: CIFOR; Obidzinski, et al. 2012. Environmental and social impacts of oil palm plantations and their implications for biofuel production in Indonesia. and Society 17(1): 25.

18 RRI and WRI 2014. Securing Rights, Combatting Climate Change. How Strengthening Community Forest Rights Mitigates Climate Change

19 Ravikumar et al. (CIFOR). 2017. Is small-scale agriculture really the main driver of deforestation in the Peruvian Amazon? Moving beyond the prevailing narrative. Conservation Letters 10(2): 170-177; Forest Trends. 2014. Consumer Goods and Deforestation: An Analysis of the Extent and Nature of Illegality in Forest Conversion for Agriculture and Timber Plantations.

20 Rhein (RRI). 2015. Industrial Oil Palm Development. Liberia’s Path to Sustained Economic Development and Shared Prosperity? Lessons from the East.

21 E.g. Atkinson. 2015. Palm oil in Liberia: Missed opportunities and second chances. HCS+ Consulting Study 17; Zen et al. 2015. High Carbon Stock (HCS) and the socioeconomics of palm oil: Towards improving the sustainability of the palm oil sector in Indonesia. HCS+ Consulting Study 14. 17

RSPO No deforestation consultancy: High Forest Cover Countries

• Smallholders have benefitted significantly from palm in Indonesia, although there are negative impacts from lack of FPIC, unequal benefit sharing & poor land rights22.

• Some communities in Liberia are also now supportive of palm development as a job provider, as companies have improved community engagement following initial mistakes.21 Similarly, in Papua New Guinea palm oil production either by smallholder or large estates where land is leased from customary land owners have led to micro-level economic and social development.23

Messaging: “no • Many campaign groups often cite the rapid rate of climate change and means no” major contribution of deforestation24, and the loss of biodiversity resulting from tropical deforestation25, as reasons for urgent action to stop deforestation for oil palm and other commodities.

• However, recent research has indicated that unsustainable hunting poses the greatest imminent threat to the survival of endangered vertebrates in Southeast Asia,26 which is likely to have knock-on impacts on forest conservation and livelihoods. Evidence suggests that the presence of FSC- certified companies can help to protect biodiversity through community engagement and efforts to reduce hunting pressure on threatened .27 Can RSPO companies play a similar role through effective management of adequately sized HCV or HCS conservation areas? • Given the evidence above, perhaps the most relevant questions are:

• What are the most effective ways of stopping deforestation as quickly as possible and protecting tropical biodiversity (given the risks of leakage etc outlined above)?

• How can positive momentum from corporate commitments be leveraged to actively contribute to stopping deforestation and protecting biodiversity, rather than simply displacing risk by avoiding sourcing from high risk or high forest areas?

22 Rist et al. 2010. The Livelihood Impacts of Oil Palm: Smallholders in Indonesia. Biodiversity and Conservation

23 ITS Global. 2011. The Economic Benefits of Palm Oil in Papua New Guinea

24 http://www.unredd.net/about/what-is-redd-plus.html

25 Alroy, J.2017. Effects of disturbance on tropical forest biodiversity. Proc. Natl. Acad. Sci. USA 114:6056–6061

26 Harrison et al. 2016. Impacts of hunting on tropical forests in Southeast Asia. Conserv. Biol. 30(5): 972-81

27 Christophersen et al. 2010. 6.2 Addressing the bushmeat crisis through certification. ETFRN News 51: September 2010 18

RSPO No deforestation consultancy: High Forest Cover Countries

3.5 RSPO’s role in tackling deforestation Considering the wider context of corporate “NDPE” commitments, the position of HCSA, and RSPO’s vision and Theory of Change, RSPO needs to decide what it role is and vision on tackling deforestation.

Due to its growing market share and its strong role in convening multiple stakeholder, RSPO is well placed to contribute to halting deforestation, although it is important to recognise limitations of RSPO: primarily that it is a voluntary standard for the site-level. This means allowing members to have viable projects at a site-level, where some deforestation may be needed in order to reduce deforestation against the BAU scenario in HFCLs.

Efforts are ongoing to explore jurisdictional RSPO certification where impact could be increased even further, but this report focuses on the potential role RSPO could play through site-level certification.

In this context RSPO could potentially play a pivotal role in stopping deforestation High Forest Cover contexts, but only if it stays engaged and relevant and does not shut the door on viable, sustainable projects in HFCLs. This means building a positive vision for stopping deforestation, protecting forest and supporting livelihoods in HFCLs.

Such a vision could support supply chain companies to meet commitments on: stopping deforestation, smallholder inclusion and improving rural livelihoods through addressing stakeholder concerns (Table 1).

Table 1. Summary of a positive vision and message for RSPO on allowing deforestation in HFC contexts Concerns Possible messages

Slippery slope & messaging • Time bound & maximum cut-off. Option to use 2020 cut-off and 2030, or sooner • Name in P&C: is “No deforestation” appropriate? Or is it more accurate to refer to stopping deforestation and improving rural livelihoods?

Community expectations & • FPIC assured rural livelihoods • Social benefits assured through a new plantation model or smallholder model

Climate change • Carbon neutrality

Biodiversity conservation • All HCVs identified and maintained, • Requirements for corridors or enhanced connectivity, • Reduced hunting pressure, • Require ratios of conserve: develop area or percentages of Management units to be protected?

19

RSPO No deforestation consultancy: High Forest Cover Countries

4 Definitions 4.1 Forest Before going through the different criteria that could be considered to define High Forest Cover, it is worth recalling that there are as many definitions of forest as purposes of defining forest. This is because forest definitions provide the conceptual, institutional, legal, and operational basis for the policies and monitoring systems that drive or enable deforestation, forest degradation, , and .28

Figure 2. Forest definitions adopted by major international environmental and forestry organizations28 International environmental and forestry organizations have adopted various official definitions of forest, even though these are not intended to encompass the totality of forests’ values and uses.28

In more specific contexts, forest definitions vary depending on their purpose: be it value for timber; carbon storage; improving livelihoods of forest dependent people; whether forests are natural or planted; whether forests are pre-existing or newly established; whether forest are continuous or fragmented; whether forests are composed of native or non-native species. 4.1.1 RSPO’s forest definitions The RSPO does not have its own definition of forest in the P&C,29 although a definition of primary forest is given:

- “A primary forest is a forest that has never been logged and has developed following natural disturbances and under natural processes, regardless of its age. Also included as primary, are forests that are used

28 Chazdon et al. 2016. When is a forest a forest? Forest concepts and definitions in the era of forest and landscape restoration. Ambio.

29 RSPO 2013. Principles and Criteria for the Production of Sustainable Palm Oil 20

RSPO No deforestation consultancy: High Forest Cover Countries

inconsequentially by indigenous and local communities living traditional lifestyles relevant for the conservation and sustainable use of biological diversity. The present cover is normally relatively close to the natural composition and has arisen (predominantly) through natural regeneration”

RSPO also provides for protection of any forests of High Conservation Value, with Criterion 7.3 requiring that no new plantings have replaced primary forest, or any area required to maintain or enhance one or more HCVs30. 4.1.2 Forest definitions under the HCV approach Forests were the original context for the development of the HCV concept: “All forests are valuable, but some are more valuable than others.” To make this concept operational, there was a need to define “exceptional” or “significant” forest values31. The HCV methodology includes 6 conservation values (HCVs) applicable to forests and any other ecosystem, although it does not provide specific global definitions of what is a forest. Any forests with significant values are considered HCVs and are defined with reference to global Common Guidance for the Identification of HCVs by the HCVRN and its interpretation in different local or national contexts.

Any forest considered to be of global, regional or national significance is considered an HCV. The only definitive global forest definition used in the HCV approach is for HCV 2, which explicitly includes Intact Forest Landscapes (IFLs). However, forests are also frequently considered HCV under the other 5 categories – recognising their importance:

- For supporting threatened species (HCV 1), - As rare or threatened (HCV 3), - For provision of ecosystem services (HCV 4), - As sources of communities’ basic needs (HCV 5), and - For cultural sites and values (HCV 6).

The global HCV Common Guidance requires interpreting in each national context, and is often done through HCV National Interpretations (HCVNIs). Many national interpretations currently focus on “HCV Forests” – a hangover from before the HCV approach was expanded to all ecosystems. HCVNIs are important for two reasons: Firstly, because the generic values include terms like significant, critical and concentration, which need to be qualified according to the local context to determine what really are the “High” Conservation values in that context (e.g. which forest types or species populations). Secondly because appropriate management of an HCV depends on the level of threat to the value, which can vary dramatically between countries.

30 Criteria 7.3.1: There shall be evidence that no new plantings have replaced primary forest, or any area required to maintain or enhance one or more High Conservation Values (HCVs), since November 2005. New plantings shall be planned and managed to best ensure the HCVs identified are maintained and/or enhanced (RSPO 2013)

31 HCVRN 2013. Common guidance for the identification of High Conservation Values 21

RSPO No deforestation consultancy: High Forest Cover Countries

➔ Both the HCV approach, and the RSPO, use national interpretations to recognise national priorities and differences, whilst still being framed by global guidance. In the context of forest definitions, this means that the HCV approach recognises that what may be considered important forest in a forest-poor country like Nigeria is unlikely to be considered important in a highly forested country like Gabon. Consequently, the HCV approach provides a tool for sustainable land development that aims to protect the most important values whilst permitting some socially and environmentally responsible agricultural development. 4.1.3 Forest as ‘defined’ under the HCSA The High Carbon Stock Approach (HCSA) methodology aims to identify and protect viable tropical forest. It was developed as a tool for implementing “No deforestation” in practice in plantation and agricultural developments – in reality it is currently primarily a toolkit for identifying potentially viable forest. The toolkit was developed for use in low and medium forest cover tropical moist forest landscapes, where viable agricultural developments can generally be established at the same time as protecting HCS forest and HCV areas.

HCSA uses field data on levels of , vegetation structure and composition, together with a view from above (satellite or Light Detection and Ranging – LiDAR), to create an HCS classification ranging from high-density forest to degraded former forest areas of scrub and open land32. The HCS vegetation classes are as follows:

• High, Medium and Low Density Forest: Closed to open canopy natural forest ranging from high density to low density forest. Inventory data indicates presence of with diameter >30 cm and dominance of climax species. More detailed parameters for these 3 classes are available in the HCSA toolkit. • Young Regenerating Forest: Highly disturbed forest or forest areas regenerating to their original structure. Diameter distribution dominated by trees 10-30 cm and with higher frequency of pioneer species compared to LDF. This land cover class may contain small areas of smallholder agriculture. • Scrub: Land areas that were once forest but have been cleared in the recent past. Dominated by low scrub with limited canopy closure. Includes areas of tall grass and fern with scattered pioneer tree species. Occasional patches of older forest may be found within this category. • Open Land: Recently cleared land with mostly grass or crops. Few woody plants.

The methodology identifies High Carbon Stock forest (‘viable forest’) to be excluded from development, and non-HCS areas that can be developed. The line

32 High Carbon Stock Approach Steering Group Toolkit v2. 2017. 22

RSPO No deforestation consultancy: High Forest Cover Countries

is drawn based on a combination of forest type and patch size, but can be generalised as Young Regenerating Forest with core patch size of >10ha33.

Figure 3. HCSA threshold and "definition" of forest

It is important to note that the HCSA forest methodology essentially provides for a top-down, global definition of forest. The approach does allow parameters/characteristics of the vegetation classes to be refined for different countries (with different forest types) – but the threshold is always YRF. This is a distinct difference from the HCV approach which allows for more national interpretation.

4.1.4 Why forest definitions matter The choice of forest definition impacts on assessment of forest cover and forest cover change. For example, in many cases, forest assessments (such as for the UN FAO or using Hansen et al. data) do not distinguish between land covered by natural and planted forests28. Thus, if natural forests are cleared and replaced with plantations, no net loss of forest cover occurs28. Furthermore, tree harvesting from managed plantations is not distinguished from clearance of natural forest.28

Using widely adopted structural forest definitions based solely on tree height, minimum area, and crown cover without considering other parameters or forest use, countries can show zero net deforestation or even a gain in forest extent, even while having converted considerable areas of natural forest within the same time interval.28 In mapping global tree cover, Hansen et al. (2014) included plantations of oil palm, rubber, and tree monocultures in their definition of forest cover.

Another major policy consequence of using forest definitions based solely on indices of forest structure is the inability to differentiate forests disturbed by logging operations from ‘secondary’ forest regrowth on former agricultural land.28 The ramifications of the choice of forest definition are discussed in Box 1.

33 This is a generalisation and ignores specific permutations of the HCSA Decision Tree. Actual areas defined as HCS forest depend also on a risk assessment and biodiversity value 23

RSPO No deforestation consultancy: High Forest Cover Countries

Box 1: Comparing forest cover statistics for Liberia Liberia provides an interesting example of the ramifications of choosing different forest definitions. We compared five different forest cover statistics for Liberia, as presented below: Forest cover statistics (% of land area)

FAO USGS REDD+ Forest Reference Level35 (>30% REDD+ Forest Reference Level IFL EROS34 canopy density = national definition) (>10% canopy density)

43% 68% 68% 83% 3%

This comparison indicates the huge differences that can be seen when using different definitions of forest cover, and we highlight the following key points of particular relevance to this paper:

Accuracy of GIS methods. Two recent studies (USGS and REDD+) provide very similar estimates, lending support to current methods based on classification of , Unreliability of FAO statistics (self-reported) and estimates based on satellite imagery classification, Is YRF forest? Liberia has 15% of land of scrub/forest with 10-30% canopy density. This class aligns quite well with HCSA’s YRF forest class (see maps below), but is not included in Liberia’s national forest definition (>30%), indicating that HCSA’s definition of forest is more strict than the national definition – the same is true for other countries. This is supported by HCSA field trials in Liberia which identified minimal areas of non-HCS. The following maps indicate overlap between HCSA’s definition of YRF and above (non-white areas in top map) and the Liberian 10% forest class (light green in bottom map).

34 Tappan, G.G., Cushing, W.M., Cotillon, S.E., Mathis, M.L., Hutchinson, J.A., and Dalsted, K.J. 2016. West Africa Land Use Land Cover Time Series: U.S. Geological Survey data release, http://dx.doi.org/10.5066/F73N21JF

35 Metria & GeoVille 2016. Final Report LIBERIA LAND COVER AND FOREST MAPPING FOR THE READINESS PREPARATION ACTIVITIES OF THE FORESTRY DEVELOPMENT AUTHORITY; Goslee et al 2016. Development of Liberia’s REDD+ Reference Level Draft Final Report for Republic of Liberia Forest Development Authority. Winrock Int. 24

RSPO No deforestation consultancy: High Forest Cover Countries

4.2 Carbon and GHGs Carbon storage is one of the many ecosystem services provided by forests, and concerns about the contribution of deforestation to GHG emissions have resulted in various policy measures to reduce deforestation and to protect forests for their existing carbon stores and roles as carbon sinks. These include:

- Governmental measures related to reducing emissions from Land Use, Land Use Change and Forestry (LULUCF), such as REDD+ under the UNFCCC, - Measures to limit private sector emissions such as: o Corporate commitments to conduct Life Cycle Analyses (LCA) of GHG emissions and reduce GHG emissions (including Scope 3 “indirect” emissions such as those caused by deforestation or land use change), and o Requirements of voluntary certification schemes. These may include requirements within sector specific schemes such as the RSPO, or dedicated carbon accounting or emission reductions schemes, such as the Voluntary Carbon Standard (VCS).

The broad goal of these initiatives is to reduce GHG emissions, although specific objectives or targets vary from “reducing deforestation” to achieving “net zero GHG emissions”. The specific objective in turn affects what commercial activities or mitigation measures are required, and we again return to the “value” of forest, because a focus purely on carbon emissions may allow a company to deforest if this is offset through efficiencies elsewhere in the product lifecycle.

Considering these approaches can help the RSPO TF to reflect on 1) the importance to RSPO of the “carbon” value of forest, and 2) the value of carbon as a quantifiable metric or proxy to define and measure forest. 4.2.1 RSPO and carbon stock RSPO encourages development on low carbon stock areas. Low carbon stock areas according to RSPO are areas meeting zero emission standards over one crop rotation.

RSPO does not require reductions or caps in GHG emissions, merely encouraging efforts to reduce emissions. RSPO also does not require specific methodologies for carbon monitoring, although it does provide the PalmGHG tool for use by companies.

4.2.2 Carbon stock to define forest: history In the 1980s, concerns about climate change led to the establishment of the Intergovernmental Panel on Climate Change (1988) and the creation of the UN Framework Convention on Climate Change (1992), initiating a new forest management objective: forests as carbon stocks. The Kyoto Protocol contains the terms reforestation and which subsequently had to be defined and

25

RSPO No deforestation consultancy: High Forest Cover Countries

operationalized in this context. The adoption of the Bali Action Plan in 2007 gave rise to the Forest Carbon Partnership Facility and the UN-REDD Programme. Biomass and carbon density became the metrics of forest monitoring and assessment.28,36 Attempts to quantify and monetize and other ecosystem services were expanded to incentivize and reforestation through payments for ecosystem services.37

4.2.3 Pros and cons of carbon thresholds Pros The major advantage of using carbon as a proxy for forest value is that it can easily be quantified and measured, simplifying assessment and monitoring over time. Cons Carbon greatly oversimplifies “forest value”, for example by undervaluing local people’s use of low- and medium-carbon forests, which are typically those areas closest to villages where carbon stocks are lower but where most farming, collecting and hunting activities take place. This means that if a very strict carbon threshold is used it may technically mean that local people can’t clear young , rendering even typical smallholder palm plantations ‘unsustainable’. Furthermore, if companies seek out low-carbon stock areas for development this is likely to increase the risk of conflict with communities given that such areas are more likely to be under community use. Furthermore, many ecologically unique or important forests or ecosystems are characterized by low carbon stock,38 e.g. Kerangas (heath) forest or savannah .

These risks are amplified in highly forested countries and where globally defined thresholds (such as that of the HCSA) are applied. For example, Gabon has 88% forest cover, and the average carbon stock in its forests is 180 tC/ha.39,40 Application of a threshold like 35 tC/ha (as previously, but no longer solely, used by HCSA) leaves only some marginal areas of savannah in the country or tiny patches of young regrowth in forest areas. As such it has been argued that such strict ‘High Carbon Stock’ definitions make sense for protecting the little remaining forest in highly fragmented landscapes, but are incompatible with

36 Saatchi et al. 2011. Benchmark map of forest carbon stocks in tropical regions across three continents. Proceedings of the National Academy of Sciences 108: 9899.

37 Wunder, S. 2007. The efficiency of payments for environmental services in tropical conservation. Conservation Biology 21: 48–58.

38 Putz & Redford. 2009. Dangers of carbon-based conservation, editorial. Global Environmental Change 19 (2009) 400–401.

39 https://www.linkedin.com/pulse/high-carbon-stock-forests-carbon-neutral-palm-roadmap- stewart

40 Burton et al. 2016. Reducing Carbon emissions from forest conversion for oil palm agriculture in Gabon. Conservation Letters 26

RSPO No deforestation consultancy: High Forest Cover Countries

development agendas of highly forested nations like Gabon.39 Studies in Gabon have encouraged the application of national thresholds to greatly restrict forest conversion through national-level zoning/identification of productive, low carbon agricultural land, coupled with requirements for finer-scale carbon, biodiversity and community use assessment at the concession-level.40

Other risks identified of using strict carbon thresholds include a blind focus on increasing carbon stocks of forests to the detriment of forest structure and composition. As such, many have argued for the importance of forest definitions that consider multiple aspects including biodiversity and social use.38

According to some scientists, forest definitions focusing on attributes of living trees, combined with regenerating processes such as recruitment and succession, whilst useful, are ecologically incomplete. Definition of forest should incorporate both attributes of the living trees and turnover in the dead- component to more effectively characterize an ecosystem that is dynamic. This would allow to infer whether a tree-covered land unit is likely to be in a static, degrading, or unstable state, and potentially vulnerable to tipping into a ‘non-forest’.41

4.3 High forest cover The concept of high forest cover countries and landscapes has been around for some time, originating from discussions within the United Nations Framework Convention on Climate Change (UNFCCC) around the policy of Reducing Emissions from deforestation and forest degradation (REDD+). More recently the concept has been further discussed within the HCSA. These discussions and lessons are synthesised here. 4.3.1 REDD+ High-Forest, Low-Deforestation countries The concept of REDD+ was agreed in negotiations of the UNFCCC, based on recognition of the significant contribution of forest degradation and deforestation to global climate change. It is a policy designed to reduce these emissions through improved forest conservation, sustainable forest management and enhancement of forest carbon stocks in developing countries.42 A core component of REDD+ was for developed countries to transfer results-based payments to forested developing countries to support their implementation of REDD+ programmes and activities.

A critical issue emerging from the REDD+ discussions surrounded the topic of “High-Forest, Low-Deforestation” (HFLD) countries. These are countries that remain mostly forested and where historic deforestation rates were low, and in many cases were relatively undeveloped often with high poverty levels. These HFLD countries’ forests were critically important to protect, but they also have

41 Buettel et al. 2017. Missing the wood for the tree? New ideas on defining forests and forest degradation. Rethinking Ecology

42 http://redd.unfccc.int/fact-sheets/unfccc-negotiations.html 27

RSPO No deforestation consultancy: High Forest Cover Countries

development needs that would typically use logging or agriculture as means of growing their economies and improving livelihoods.

This became particularly important when defining national forest reference levels or baselines of GHG emissions for REDD+, which are the basis for determining results-based payments. Countries would receive payments only if they could demonstrate that they were reducing deforestation from the historic or reference level. But for HFLD countries (including the Coalition for Nations) that have kept most of their forest standing, was it fair to expect them to keep all of their forests standing and potentially forego development without any compensation? This meant ensuring that:

1. Reference levels took this into account, and 2. Safeguards were in place to ensure local livelihoods could be met and improved in HFLD countries.

Ultimately under REDD+, it was agreed that countries could define their own reference levels based on some flexible guiding rules. There has been extensive research into different methodologies for defining reference levels based on a combination of forest cover, deforestation rate and other variables.

4.3.1.1 Defining HFLD Da Fonseco et al. (2007)43 classified developing countries based on remaining forest cover and deforestation rate (see Figure 4): 1. Low forest, high deforestation 2. Low forest, low deforestation 3. High forest, high deforestation 4. High forest, low deforestation

Values for forest cover, forest area and deforestation rate taken from the UN Food and Agriculture Organization (2005).44 To define the four categories, the authors use cut-offs of 50% remaining forest and 0.22% forest loss per year. Remaining forest of 50% was selected as the “simplest arbitrary cut-off.” A cut-off of 0.22% per year was selected because it represented the global average rate of deforestation for the reference period of 1990-2000.45 Therefore, HFLD countries have forest cover >50% and deforestation rates <0.22% (Figure 4).

43 Da Fonseca, et al. (2007) No forest left behind. PLoS Biol 5(8): e216. doi:10.1371/journal.pbio.0050216

44 Food and Agriculture Organization of the United Nations (2005) Global forest resources assessment 2005: Progress towards sustainable forest management.

45 Griscom et al. 2009. Sensitivity of amounts and distribution of tropical forest carbon credits depending on baseline rules. Environmental Science and Policy 28

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 4. Summary of countries categorised by the da Fonseca methodology

Another idea has been to adopt reference levels indexed to the global deforestation rate for countries with little or no historic deforestation, allowing HFLD countries to receive “preventive credits” that would be lost if the deforestation rate were to increase.45

Griscom et al. have suggested that the da Fonseca method may be overly simplified, and so conducted a quantitative analysis of ‘‘natural’’ groupings of countries according to the same variables. Using a multivariate cluster analysis they calculated the two variables as (1) proportion of original forest cover remaining based on data from the FAO 2005 National Forest Resource Assessment (FRA), and (2) rate of forest change combining FAO 2005 data on forest loss and WRI data on original forest cover (Bryant et al., 1997). The analysis came up with five natural categories.

29

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 5. Clusters of countries identified by Griscom et al. See caption included in graphic for more details. 4.3.2 models Another methodology proposed is based on the forest transition model, as widely discussed in development and environmental literature. Griscom et al. considered the da Fonseca categories as part of a time sequence:

- HFLD countries (Fonseca Type 4) typically shift to… - Increased rates of deforestation as they develop (Fonseca Type 3), but… - Then reach a change point where forest loss starts to decline (Fonseca Type 2) and.. - Finally reverses due to forest regeneration, restoration or reforestation (Fig. 1).

Figure 6. Generalised forest transition model, taken from Griscom et al.

30

RSPO No deforestation consultancy: High Forest Cover Countries

4.3.3 Deforestation baseline approaches Two other approaches proposed for defining REDD baselines are retrospective and prospective baselines.46

Retrospective baseline methods are based on the trend of historic deforestation rates being extrapolated into future commitment periods. This ideally uses multiple-year periods for estimates to avoid biases from high interannual variation. A weakness of retrospective methods is that they assume linear rates of deforestation, so not considering changes linked to forest developmental transitions.

Examples of a retrospective baseline approach (or Joint Research Centre approach) is Mollicone et al.47, who calculate a baseline from satellite imagery from the period 1990 to 2005, including a method to separate intact forest, non- intact forest and non-forest land. The method considers a global baseline rate requiring emissions reductions for those above the global average and allowing those below it to benefit from ‘avoided deforestation’ payments.

Huettner et al. also discuss other retrospective methods they refer to as a Simple Historical Approach (SiHA) and Spatial Historical Approach (SpHA), which use global satellite imagery combined with either global biomass and carbon conversion factors (for SiHA) or local factors (for SpHA).

Prospective approaches use predictive and dynamic spatial land-use modelling to estimate future land use changes. They attempt to incorporate the dynamics of deforestation drivers and various different models to most effectively estimate deforestation and future land use scenarios. They can readily be tailored to different national contexts or drivers. Examples include CLUE-S. Interestingly, these approaches have been less popular amongst policy makers, seemingly due to their complex methods (hence perceived opaqueness) and impractical requirements for detailed data – such models have been seen as more viable at a project level, e.g. within voluntary carbon schemes. Other baseline models are shown in Figure 7.

46 Huettner et al. 2009. A comparison of baseline methodologies for 'Reducing Emissions from Deforestation and Degradation'. Carbon Balance and Management.

47 Mollicone et al. 2007. An incentive mechanism for reducing emissions from conversion of intact and non-intact forests. Climatic Change, 83:477-493 31

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 7. Comparison of different baseline methodologies taken from Griscom et al.

4.3.4 HCS High Forest Cover Landscape (HFCL) The High Carbon Stock Approach (HCSA) methodology considers High Forest Cover Landscapes (HFCL) as landscape with more than 80% of forest cover (HCS Approach Steering Group 2017). Note that unlike discussions in the REDD+ context, where HFLD were defined at national (or sometimes sub-national jurisdictional scales), HCSA defines HFCLs at a much smaller landscape scale.

The HCSA steering group set up this figure based on research on landscape-level impacts of deforestation in the Amazon48. The study reveals that forest structure is weaken when approximately 20% of the forest cover has been removed. A second study about effects of on birds and mammals49 highlights the fact that habitat loss increases once total habitat drops below 30% because of increasing habitat fragmentation consequences. These figures served as a basis for setting up HCSA’s forest cover thresholds. Whilst the use of ecological evidence is clearly important, it is worth noting that these definitions and studies are based on patterns of clearance and fragmentation in the Amazon, which may not necessarily be reflective of patterns elsewhere (e.g. in palm growing regions). In addition, these variables arguably have greater bearing on how the size and connectivity of conservation areas should be designed than they do on how to define highly forested areas. What is more, these studies do not consider the socio-economic dimension or trajectory in HFC landscapes or countries that are considered critical in discussions about equity of ‘no deforestation’ policies.

48 Francisco José Barbosa Oliveira de Filho, Jean Paul Metzger. 2006. Thresholds in landscape structure for three common deforestation patterns in the Brazilian Amazon, Landscape ecology

49 Henrik Andrén. 1994. Effects on habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: a review, Oikos 71: 355-36 32

RSPO No deforestation consultancy: High Forest Cover Countries

Forest cover landscape categories in the HCS Approach toolkit

Low <30%

Medium 30-80%

High >80%

Table 2: Forest cover categories in the HCS Approach toolkit

Box 2. Lessons from the HCSA High Forest Cover Landscape Working Group

Diverse recent case studies presented during HCSA working groups raised the importance of developing an adapted methodology to be applied to HFCL. It was argued that in HFCLs, the HCSA toolkit often results in economically non-viable projects for industrial exploitation. Moreover, it creates “stranded assets” from the perspective of the companies involved. Some of the implications highlighted were as follows:

• Producer abandons HCS or NDPE commitment, • Producer or owner divests (sells plantation or production company), • Companies bought up by companies with no NDPE commitment, • Producer allows lease to lapse, and return to Government for reallocation, • Trader /supply chain might drop suppliers with no NDPE commitments

All of these scenarios suggest a high probability of deforestation unless NDPE companies aspire to actively protect these concessions under forest cover for perpetuity – but which of course comes with huge cost and management implications.

The issue of ongoing “leakage” or scope for the oil palm industry to continue with deforestation beyond the control of the current HCS approach was also highlighted, with some working group members fearing that demand from markets that accept “unsustainable” products, such as China and India, would continue to drive BAU or worse deforestation in HFCLs in the absence of a market demand from NDPE companies to reduce deforestation. It was mentioned that many palm oil growers are still not committed to deforestation and that some companies are exploring development of vertically integrated supply chains to avoid having to trade through NDPE companies.

Another critical point of discussion was around the importance of development in many poor HFCLs where palm production could help to create jobs, build social infrastructure such as schools, health clinics and roads (which are often sorely lacking). It was argued that palm does not always provide these benefits locally and that alternative development or economic activities should be prioritised that don’t require deforestation. Growers in turn argued that no other actors or companies are currently providing realistic alternatives to palm or other land-based development options.

4.3.5 IFLs in HCV 2 and FSC The methods discussed above focus more on definitions related to % coverage of jurisdictions or landscape of forest. An alternative approach is simply to map the most important landscape-level forests directly, irrespective of jurisdictional boundaries. This is the approach taken for IFLs.

33

RSPO No deforestation consultancy: High Forest Cover Countries

The HCV common guidance refers to a definition of large undisturbed landscape- level forests, known as Intact Forest Landscapes (IFL). The concept was coined by Greenpeace, working with other organisations50 to map Frontier Forests and Intact Forest Landscapes at a global and regional level since the 1990’s. The definition of an IFL is:

“a territory within today’s global extent of forest cover which contains forest and non-forest ecosystems minimally influenced by human economic activity, with an area of at least 500 km2 (50,000 ha) and a minimal width of 10km (measured as the diameter of a circle that is entirely inscribed within the boundaries of the territory)”

IFL are considered as HCV 2 that are landscape-level ecosystems and ecosystem mosaics51.

HCVNIs (see 4.1.2) are interesting regarding HCV 2 because the way it is understood and applied for forests will be different in Canada (where the country retains large tracts of undisturbed forest) from the way it will be treated in Ghana (where there are only a few remaining forest blocks, none of which are undisturbed). This means that different countries use different thresholds for how large a forest block should be to be considered HCV 2, ranging from a global standard of 50,000 ha down to 10,000 ha in countries with lower forest cover or 1 million ha in high forest countries like PNG.

Figure 8. Countries that have HCV National Interpretations. Note several HFC countries, such as Papua New Guinea, Canada, Liberia and others.

50 The IFL mapping team is an alliance of research and non-governmental ecological organizations and is constantly improving and updating the Intact Forest Landscapes dataset. The following organizations contributed to the year 2000 IFL mapping and 2000-2013 map update: Greenpeace; Global Forest Watch; Transparent World; The Global Land Analysis and Discovery (GLAD) laboratory, Department of Geographical Sciences at the University of Maryland, WWF Russia, Luonto Liitto (Finnish Nature League), Forest Watch Indonesia, and other regional NGO.

51 Large landscape-level ecosystems and ecosystem mosaics that are significant at global, regional or national levels, and that contain viable populations of the great majority of the naturally occurring species in natural patterns of distribution and abundance; https://www.hcvnetwork.org/about-hcvf 34

RSPO No deforestation consultancy: High Forest Cover Countries

4.3.6 Analysis of High Forest Cover definitions Each of the definitions and methodologies discussed above was developed for a different purpose (Table 3), ranging from mapping important contiguous blocks of forest to identifying countries at an early stage of the forest transition to inform policy development. It is important that RSPO considers its objectives when defining a HFC threshold, to ensure appropriate variables are considered. Indeed, RSPO has said that a key motive is to recognise the development status of countries whilst contributing to forest conservation. As such, other variables beyond the biophysical may need consideration. These are discussed further under sections 7 and 8.

One method emerging from REDD+ discussions of potential interest to RSPO’s site-level focus are the prospective baseline models. These have been increasingly used at a site/project-level for carbon projects to extrapolate potential development scenarios. Given concerns about palm developments resulting in a slippery slope of deforestation, perhaps RSPO could consider requiring growers to use these prospective models to better understand and mitigate future deforestation (including land requirements of communities).

Table 3. Comparison of relevant high forests cover thresholds or definitions Source Forest Deforestation Rationale cover rate

HCSA 80% at Not included Amazon study (Oliveira de Filho and Metzger 2006). landscape Threshold after which patch size & isolation reduces rapidly

REDD+ 50% 0.22% p.a. Griscom et al. 2009. “Arbitrary” forest cover. Mean national deforestation rate (2005)

FCPF Not Not specified “achieve net emission reductions … and to pilot Carbon specified REDD+ across a diverse set of countries, including Fund those … with high forest cover and low deforestation”

IFLs NA NA Different approach focused on mapping important, intact forest of at least 50,000 ha, 10 km wide

4.4 Spatial scale The RSPO P&C is applied at a site-level, however, the spatial scale being discussed in this paper is about how you define HFC contexts which necessarily require looking at a broader scale.

35

RSPO No deforestation consultancy: High Forest Cover Countries

RSPO has decided a priori to focus on HFC Countries, and so we will not discuss scale at length, only to provide a brief analysis of some of the pros and cons of a national Vs a sub-national spatial scale. Landscape It is now widely acknowledged that issues of deforestation and development happen beyond the scale of individual concessions, properties or management units. They occur in interconnected landscapes with interacting drivers, indirect impacts and ecological dynamics. Furthermore, the mix of these drivers and dynamics are typically unique to each landscape. This is a major motivation for why the HCV and HCS approaches require particular attention to be given to the wider landscape around individual developments.

In recognition of this, the HCSA defines HFC at a landscape-level but the HCSA methodology allows only very minimal flexibility for ‘more conversion’ in high versus medium or low forest cover landscapes.

The advantage of looking at a landscape scale is that it considers the real dynamics mentioned above and can allow for a greater scale of implementation beyond simply site-level without requiring the huge level of coordination needed for national or jurisdictional scale.

A major risk of RSPO defining landscapes at a landscape (or jurisdictional) scale and then allowing exceptions for deforestation is that it would allow many areas on the frontier of deforestation in otherwise heavily deforested countries to justify further clearance. For example, the Heart of Borneo or Leuser ecosystems at a local landscape level are still highly forested, but most would argue that they should be largely protected and that any (if at all) further palm development should focus on nearby areas that are already degraded or under agricultural use.

Of course, whether having a strict stance for “no deforestation” for palm in these sub-national landscapes actually provides a strong enough incentive to stop deforestation is another question.

Biogeographic region A less widely considered approach would be to consider biogeography in defining scale. Forested landscapes stretch over national boundaries which don’t usually follow natural features and so a biogeographic scale would better capture this.

For examples, the WWF Ecoregion 200 approach uses biogeography as its basis in analysing global patterns of biodiversity, resulting in 238 ecoregions comprising 142 terrestrial, 53 freshwater, and 43 marine priority ecoregions (Figure 9). As an example, the island of New Guinea contains 2 priority forest ecoregions: New Guinea Montane Forests and New Guinea Mangroves. These ecoregions are spread across PNG, and West Papua and Papua in Indonesia.

36

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 9. WWF Ecoregion 200 map

In Southeast Asia, different biogeographic zones have been defined based on the distribution of fauna, divided into 3 distinct regions: • Sundaland: Peninsular Malaysia, Borneo, Sumatera, Java, Palawan

• Wallacea: Central Indonesian islands, (Philippines), (Eastern Indonesian Islands)

• New Guinea, (Eastern Indonesian Islands).

Figure 10. Biogeographic boundaries in Southeast Asia as defined by different biologists.

37

RSPO No deforestation consultancy: High Forest Cover Countries

A biogeographic approach could potentially justify including the Indonesian part of New Guinea (i.e. West Papua and Papua provinces) as an exception to country- specific HFCL rule:

• The island of New Guinea is a distinct biogeographic region on its own – the division of the island into two different countries is completely arbitrary

• The ecoregions within New Guinea are contiguous areas extending across the Indonesia-PNG border

• The ecoregions in the Indonesian side of New Guinea are distinct from those in other parts of Indonesia

• The island as a whole has a high forest cover including the Indonesian portion which may have an even a higher forest cover than PNG

A similar approach could also justify including the Amazon as a HFC landscape.

Country Defining HFC Countries would follow logic applied in REDD+ discussions. It recognises that in most countries the major macro-level economic, social and environmental policies and legislation are made at a national level, and that any large-scale zoning to identify protected areas or define important forests also happen at this scale. Importantly, efforts to alleviate poverty and improve rural livelihoods are typically stimulated by national policy measures or initiatives, e.g. to subsidise certain crops or invest in social infrastructure. Therefore, recognising this allows HFCCs more flexibility and sovereignty to develop such national strategies – whilst allowing responsible RSPO growers to operate in and support these national initiatives.

The main risk of a national definition relates to the fact that many countries, although in theory governed centrally, are in fact highly heterogenous geographically, socially, economically and environmentally. For large countries and those with inequal forest distribution this is particularly true, for example, Indonesia, Brazil and the other Amazonian countries.

Within Indonesia there are huge differences between Sumatra and Kalimantan where huge deforestation which occurred over the past 50 years has allowed major development of palm and other commodities and also supporting economic growth and, in places, rural development. Conversely, Papua and West Papua provinces remain mostly forested, have lower life expectancy than other provinces and less access to infrastructure (physical and social).52 Furthermore, national definitions often overlook cultural and historical factors, for example, that former Irian Jaya (Papua and West Papua) is culturally and ethnically distinct from western Indonesia.

52 E.g. http://www.ifa-fiv.org/wp-content/2014/09/Indonesian_Ageing_Monograph-print-version1.pdf 38

RSPO No deforestation consultancy: High Forest Cover Countries

If countries such as Brazil, Indonesia, Ecuador, Peru and Colombia are all defined as low forest cover countries and not allowed to deforest because national forest cover is approximately 50%, risks may be as follows:

- Is it equitable for jurisdictions that are still highly forested and perhaps poorer for a top-down imposition of strict local requirements based on global values? - If these jurisdictions already have development strategies and have allocated land to companies, what is the BAU trajectory for these areas in terms of deforestation? - Will RSPO have greater impact and influence on changing this trajectory if it a) does not allow growers in these areas to be certified, or b) if it allows growers to do some limited conversion that is less than may happen under BAU? 4.4.1 Analysis of different spatial scales A table comparing the pros and cons of different scales is provided below.

Scale Advantages Disadvantages

Country • Quicker to agree amongst RSPO • Unequal development in country stakeholders • RSPO risks losing influence over known • Link to national initiatives and policy new frontiers e.g. Papua. Most certified making areas are long established – how to stay • Avoids perceived risk of “opening relevant for new development? up” Papua and West Papua • Highly likely leakage to other commodities and non-RSPO palm in high forest areas within low/medium forest cover countries

Jurisdiction • Aligned with implementable • Criticism over Papua, Amazon & jurisdictional approaches ‘greenwash’ • RSPO able to influence trajectory in • Risk of actors justifying Leuser, HoB as key areas like Papua “HFCLs” • Option to make case for reduced deforestation Vs BAU

Biogeographic • RSPO able to influence trajectory in • Criticism over Papua, Amazon & key areas like Papua ‘greenwash’ • Ecological perspective on forests to protect

4.5 Development status and poverty Social equity is a key factor in the HFC debate. To crudely summarise, many highly forested countries and landscapes are also those with higher levels of poverty. Various approaches have been explored to try and quantify development status or poverty as a way of identifying areas or countries where some deforestation

39

RSPO No deforestation consultancy: High Forest Cover Countries

may be acceptable on the grounds of poverty alleviation and improving livelihoods.

One approach is simply to consider forest cover and/or deforestation rate itself as a proxy for development status or status in the “forest transition”, e.g. if a country has a high remaining forest cover it could be assumed to be relatively undeveloped. In affect this is the approach that was taken in REDD+, where reference levels are typically defined based on deforestation rate and forest cover.

The Sustainable Palm Oil Manifesto (SPOM) group explored using specific social indicators as a way of quantifying development status. These findings were captured in the resulting HCS+ Study. Not all of these learnings were captured in the “converged” HCS Approach that is now used, and so are captured here.

Ultimately HCS+ did not include an explicit social threshold. In short, they decided that it was not possible or advisable to quantify a social threshold because:

• Virtually all social indices, such as the Human Development Index and Multi-Dimensional Poverty Index, are calculated at a national level. This means they do not capture often significant differences in development status or poverty levels between rural and urban areas. In fact, they may be skewed by urban areas because that is where most people live. An example is Gabon, which scores fairly highly on HDI and other indices because the majority of the population is urban and relatively well-off, but the minority of the population living in rural areas remains significantly poorer than their urban compatriots,

• The use of a single hard cut-off or threshold risks creating severe ‘boundary effects’ whereby two countries with only a 1 or 2% difference in the value face wildly different requirements,

• In practice, virtually all tropical countries and especially all potential HFCCs have enough rural poverty to “justify” some development. Life expectancy, health and income levels are invariably relatively low in rural tropical areas compared to either national or global expectations.

The main conclusion from the HCS+ Study was the importance of ensuring two fundamental processes or outputs:

1. Minimise social harm during new developments. Primarily this can be achieved through an effective FPIC process, for example, referring to the HCSA’s social requirements, and 2. Maximise social benefits of developments. Once a development is established it is vital that benefits are adequately shared amongst customary rights holders, local community and workers. This requires effective benefit-sharing models as well as indicators to effectively monitor social benefits. As a monitoring indicator the HCS+ study developed the Palm Oil Welfare Index, which has since been adapted into the Welfare Impact Assessment (WIA).

40

RSPO No deforestation consultancy: High Forest Cover Countries

4.6 Statistics on potential high forest cover countries Based on the variables discussed above, statistics for a number of key variables have been collated here for known current or major potential future palm producer countries. Clearly low forest cover countries such as Ghana, Ivory Coast, Nigeria and Sierra Leone are omitted for brevity. Source of Deforestation Oil palm Deforestation deforestation % % forest Source for Deforestation Deforestation rate (%/yr; (% total rate trend rate (if different Country forest % IFL (national/ national/regional rate (%/yr; rate trend national/ land area) (national/ from national/ (FAO) regional) forest cover*** FAO) (FAO) regional (FAO) regional) regional forest source) cover) Gabon 89.3 26.3 88.3 OFAC 0.0 -0.9 Decreasing 0.1 Decreasing Solomon 78.1 10.6 76.6 MinFR 2010 0.6 0.3 Increasing 0.3 Constant Islands PNG 72.5 23.8 71.0 PNGFA 2012 0.4 0.0 Constant 0.1 Constant CCDA 2017 DRC 67.3 22.0 68.8 OFAC 2010 0.1 0.2 Constant 0.2 Constant Metria & GeoVille; Liberia 43.4 2.7 68.0 0.2 0.7 Constant 0.9 Constant Winrock 2016 R of Congo 65.4 27.0 67.1 OFAC 0.0 0.1 Constant 0.5 Increasing Myanmar 42.9 4.0 63.0 Bhagwat et al 2017 0.2** 1.8 Increasing 0.3 Constant Malaysia 67.6 3.9 55.2 NRE 2014 15.2 -0.1 Decreasing 0.4 Decreasing REDD Desk Indonesia 53.0 15.0 54.2 MinEF 2015 5.4 0.7 Increasing 0.9 Constant 2013. Brazilian Brazil 58.0 21.8 54.0 0.0 0.2 Decreasing 0.5 Decreasing Forest Service (SFB) 41

RSPO No deforestation consultancy: High Forest Cover Countries

Peru 57.8 33.5 53.0 MINAM 2015 0.0 0.2 Constant 0.1 Constant Costa Rica 53.9 4.9 52.4 REDD Desk 1.4 1.1 Increasing -0.9 Decreasing Sistema de Monitoreo de Colombia 51.0 24.2 51.7 0.3 0.0 Decreasing 0.3 Increasing Bosques y Carbono (2016) Ecuador 50.5 16.3 51.6 MAE 2014 1.1 0.6 Constant 0.7 Decreasing 2013. OFAC (de Cameroon 39.8 6.8 51.4 0.3 1.1 Increasing 0.9 Constant REDD Desk Wasseige et al) Honduras 40.8 3.5 48.0 UN REDD 1.4 2.4 Increasing 0.9 Constant Guatemala 32.5 3.6 34.2 INAB 2012 1.3 1.0 Decreasing 1.5 Constant REDD Desk Thailand 32.0 2.7 32.0 RFDT 2016 1.3 -0.2 Decreasing 0.1 Constant CAR 35.6 0.7 15.1 OFAC 0.0 0.1 Constant 1.8 Decreasing **Myanmar oil palm stats not available from FAO, taken from Baskett 2015

***Full list of references provided here: Baskett, J.P.C. 2015. Myanmar oil palm plantations: A productivity and sustainability review Report no. 28 of Myanmar palm the Tanintharyi Conservation Programme, a joint initiative of Fauna & Flora International and the Myanmar area Forest Department. Forest cover Min of Env, Brazilian Forest Service. 2013. BRAZILIAN FORESTS at a glance. Available at: Brazil http://www.florestal.gov.br/publicacoes/571-brazilian-forests-at-a-glance-2013

42

RSPO No deforestation consultancy: High Forest Cover Countries

CAR, Gabon, Les forêts du bassin du Congo – État des Forêts 2013. Éds: de Wasseige C., Flynn J., Louppe D., Hiol Hiol F., RoC, Cameroon Mayaux Ph. – 2014. Weyrich. Belgique. 328 p. Dépôt légal : D/2014/8631/30 ISBN : 978-2-87489-298-1

Costa Rica https://theredddesk.org/countries/costa-rica/statistics Ecuador Ministerio del Ambiente. 2014. Ecuador’s Forest Reference Emission Level for Deforestation

Guatemala https://theredddesk.org/countries/guatemala/statistics

Honduras http://www.unredd.net/index.php?option=com_docman&task=doc_download&gid=13374&Itemid=53 MinEF, Republic of Indonesia 2015. NATIONAL FOREST REFERENCE EMISSIONS LEVEL FOR REDD+ In the Indonesia Context of Decision 1/CP.16 Paragraph 70 UNFCCC Metria & GeoVille 2016. Final Report LIBERIA LAND COVER AND FOREST MAPPING FOR THE READINESS PREPARATION ACTIVITIES OF THE FORESTRY DEVELOPMENT AUTHORITY; Goslee et al 2016. Development of Liberia Liberia’s REDD+ Reference Level Draft Final Report for Republic of Liberia Forest Development Authority. Winrock Int. http://www.nre.gov.my/en-my/Forestry/Pages/Statistics-Forest.aspx ; Malaysia https://theredddesk.org/countries/malaysia/statistics Bhagwat T, Hess A, Horning N, Khaing T, Thein ZM, Aung KM, et al. (2017) Losing a jewel—Rapid declines in Myanmar Myanmar’s intact forests from 2002-2014. PLoS ONE 12(5): e0176364. https://doi.org/10.1371/journal.pone.0176364 Peru MINAM 2015 http://sinia.minam.gob.pe/documentos/cobertura-perdida-bosques-nivel-nacional PNGFA 2012 Forest Base Map; Climate Change and Development Authority 2017. Papua New Guinea’s PNG National REDD+ Forest Reference Level Submission for UNFCCC Technical Assessment in 2017. Solomon Islands Ministry of Forestry and Research 2010. http://mofr.gov.sb/foris/forestArea.do#marker Thailand Royal Forest Department of Thailand 2016. http://forestinfo.forest.go.th/Content.aspx?id=72

43

RSPO No deforestation consultancy: High Forest Cover Countries

5 Approaches and interventions The second key part of any HFC approach that RSPO develops, following on from defining where the approach applies, is what the approach itself looks like and requires. For example, what is the assessment process for determining what areas should be developed or protected and what are the desired outcomes?

To inform the development of a robust approach we review a selection of approaches and strategies used elsewhere and try to draw relevant lessons from them. 5.1 Lessons from Forest Stewardship Council No conversion The Forest Stewardship Council (FSC) standard includes a strict requirement for no conversion of natural forest since 1994. In recent years this requirement has come under scrutiny, as stakeholders question its impact in actually preventing deforestation and on the viability of forestry operations.53 It has been strongly argued that having this rule has prevented FSC from having any impact or control over limiting deforestation in developing countries and in the tropics - where huge deforestation has happened since 1994 and there is very little FSC certified production. It is felt by many in the global South that FSC is essentially irrelevant or discriminatory to these contexts and countries, and that such a rule prohibits developing countries from creating sustainable livelihoods from their natural resources.

As a result of these discussions FSC is now considering changing requirements to have a greater focus on net positive conservation outcomes that potentially allow limited conversion in exchange for protection or restoration of equivalently-sized areas.

Is there a risk that RSPO follows a similar path to FSC if it takes a very strict stance on no conversion? IFLs The FSC protects HCV forests, and consequently IFLs, under FSC principle 954. However, in recent years there have been renewed debates about how to protect IFLs within the FSC system with environmental chamber representatives, most notably Greenpeace, demanding the FSC provide additional protections to IFLs beyond those provided to HCVs. FSC decided to reinforce IFLs conservation through motion 65 in 2014.

The motion called for national standard development groups and certification bodies (CBs) to put in place measures to protect the vast majority of IFLs.55 At the same general assembly in 2014, motion 07 was also passed.

53 https://ga2017.fsc.org/changing-the-1994-conversion-rule/

54 FSC Principle 9: The Organization shall maintain and/or enhance the HCVs in the forest management unit (FMU) through applying the precautionary approach (FSC 2018)

55 https://ga2017.fsc.org/history-of-a-motion-intact-forest-landscapes-ifls/ 44

RSPO No deforestation consultancy: High Forest Cover Countries

To work out how to do this FSC established a multistakeholder advisory group including representatives from key IFL areas (Canada, Russia, Amazon, Congo Basin, Indonesia). Initially, the advisory group proposed that the development of IFL indicators work in parallel with the development of national standards (FSC General Assembly 2017 2018).

FSC developed default indicators requiring the full protection of a core area of each IFL within the Forest Management Unit (FMU), where core areas comprising at least 80 per cent of the IFL fall within the FMU. This was later revised to state that forest management operations within IFLs, including harvesting and road building, could proceed as long as they did not impact more than 20 per cent of IFLs within the FMU and did not reduce any IFLs below the 50,000 ha threshold in the landscape. These measures are in place until new national standards, with nationally agreed requirements are in place.

This debate is another example of where national and local rights have been in conflict with global values and rights. Many local and national stakeholders, including social rights groups, have been strongly opposed to environmental restrictions in IFLs that they see as unfairly penalising poor communities and countries and potentially pushing sustainable forestry operators out to be replaced by more unscrupulous actors. 5.2 Defining which areas to protect Different mechanisms and standards have tried different approaches for determining which important forests or ecosystems should be protected. Two approaches that have been used are, firstly, those that require a proportion of the land area or management unit to protected (henceforth ‘proportional protection’) and secondly, value-based approaches that require specific areas or values to be protected. Proportional protection requirements The Sustainable Agriculture Network (SAN) standard requires producers to set- aside 10% of their management units under natural vegetation, and FSC requires producers to maintain a representative sample of natural vegetation.

Similar approaches have also been adopted in some countries’ legislation, for example, in Brazil’s Forest Code. The Forest Code requires land owners to protect a certain percentage of their property depending on which biome they are in (Figure 11), as well as to protect key areas for ecosystem services such as riparian buffers and steep slopes.56

If not coupled with other safeguards, these approaches are applied across the board meaning that all users must comply even if in practice their area consists or all primary forest or all degraded scrubland.

56 Proforest. 2017. Assessing compliance with the Forest Code: a practical guide 45

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 11. Requirements for the percentage of the property that must be set-aside as a legally protected reserve in different Brazilian

Value-based protection Value-based approaches such as the HCV and HCS approaches take a different approach, and are context specific. They place more emphasis on an assessment at the start to identify what is there and then require those values to be either avoided or protected. If you don’t have any HCVs or HCS forest on your land you don’t have to set-aside any area, but if 100% of the area is made up of HCV and HCS areas then you will struggle to have a viable production operation.

By comparison a ratio-based approach can allow companies to have a viable operation whilst contributing to conservation, but without the risk of being overburdened with sometimes costly conservation requirements.

Combining approaches It is worth noting that some standards and policies combine both proportional and value-based approaches to ensure adequate safeguards, and approaches to share the costs to conservation (e.g. net approaches or offsetting). For example, FSC has many elements:

- No conversion of natural forest, - Maintain or enhance HCVs, and - Maintain a representative sample of natural ecosystems.

The HCS+ Study also used a combined approach by having different carbon thresholds designed to protect the more valuable forest (>75tC/ha) and allowing some forests with less carbon to be developed only if the entire operation was carbon neutral in terms of land-use emissions.

This had the effect of requiring companies to be responsible for managing and protecting set-asides – a powerful idea that has been lacking from many policy approaches that have led to leakage because of inadequate efforts to go beyond avoidance of deforestation to protection of avoided areas.

Produce-Protect-Include approaches The main challenge of the HCS approach in HFCLs has been that it simply does not allow companies or producers sufficiently large or viable areas to develop. In large part this is the result of the approach setting a strict, top-down bar for what is considered forest and so allows no flexibility to plan a viable operation that balances conservation, production and community needs. As a result, some companies and organisations working in HFCLs have been exploring alternative approaches that are more bottom-up and participatory, for

46

RSPO No deforestation consultancy: High Forest Cover Countries

example, IDH, Sime Darby and Golden Veroleum in Liberia. The trials have focused on ‘produce-protect-include’ (PPI) approaches that integrate these three pieces through participatory land-use planning processes.

The main motivation for these trials was the need for something more flexible in these contexts, given that the HCSA was found to be too limiting. These organisations recognise the need to protect important forest and other values in these landscapes but felt that a more bottom-up approach would be more effective by getting community buy-in and that necessarily allowed some production to ensure income and livelihoods were met and improving. There is growing interest in PPI approaches in HFC contexts as a means of:

1. Allowing companies or investors to stay engaged by supporting some development, and 2. Ensuring careful planning that prevents creeping deforestation and ensures biodiversity is protected.

5.3 Assessment requirements A HFC approach will require some kind of due diligence or assessment process prior to any new developments to identify which areas to protect. The RSPO system already requires various assessments as part of its New Planting Procedure, namely HCV, Land Use Change, GHG (including carbon stock), soil suitability and ESIA assessments.

The HCSA now represents a similar methodology for conducting a pre- development baseline assessment, specifically aimed at identifying forest areas that should be set-aside from development. Realistically, the RSPO has two options that could be used in a new HFC approach:

1. Use a modified version of the existing NPP assessments to better integrate and strengthen the land cover assessment requirements with HCV requirements. This could enable requirements that protect HCVs and certain forests, or 2. Use the integrated HCV-HCSA assessment methodology.

These would be reliable options that would avoid creating new methodologies, although RSPO would still need to agree on what thresholds would be applied for areas or forests to be protected. 5.4 Analysis of options discussed by RSPO P&C TF The RSPO P&C TF has come up with various ideas during TF meetings for what a HFC approach could look like. Some of the more prominent ideas are listed below with a short analysis of their pros and cons.

Proposed idea Pros Cons

47

RSPO No deforestation consultancy: High Forest Cover Countries

Shift HCS boundary up to LDF • No need for new Limited developable area in methodology some contexts, e.g. Gabon • Aligns with initial HCSA legacy ideas

Use a higher carbon threshold Option to define with a new HCSA moved away from fixed RSPO specific approach (e.g. thresholds and not using only with more flexibility than HCSA) carbon because carbon is a poor approximation of forest’s full value

Define relative, national level Fits with RSPO model of Impractical, costly to do full vegetation thresholds through inclusivity & NI approach mapping at national level. national level mapping (e.g. Alternative could be to YRF in Gabon = LDF in validate/adjust thresholds Indonesia) through pilots

Offsetting mechanism Allows positive messaging & • Baggage of HCS+ aligns with net commitments, • ZND not supported by e.g. CGF campaign NGOs

Allow NIs to set their own Fits with RSPO model of Risk of ‘sanctioning’ threshold inclusivity & NI approach, deforestation if global RSPO allowing local ownership guidance does not ensure some safeguards or max limits

6 Scenarios/options Based on the literature review, interviews and discussions with the RSPO NODIG, this section puts forward proposals to the RSPO NODIG for how to develop a HFC approach. Recommendations are made for guiding principles, potential HFC thresholds/definitions and for requirements of the HFC approach itself.

Supporting justification is provided in Annex 1 for the final text agreed by the NODIG and RSPO taskforce on HFC countries for public consultation. 6.1 Guiding principles Based on lessons learnt from other discussions, standards and stakeholder concerns the following guiding principles for a robust HFC approach were identified: • Applies only in HFC areas (as defined by RSPO)

• Time-limited • Net positive carbon impact

• Net positive or neutral biodiversity e.g. through HCVs secured, reduced hunting, compensation and/or a minimum ratio of conserve : develop

48

RSPO No deforestation consultancy: High Forest Cover Countries

• Forest conservation e.g. assured protection of HCV and HCS areas in concessions & wider landscape • Socially positive both before, during and after the development process:

• Pre-development: FPIC & social requirements

• Once established: new models for benefit sharing, e.g. smallholder model

• Requirement to support57 wider landscape initiatives to reduce leakage risk Several additional ideas raised or identified during consultations included:

- Avoid absolute thresholds. Lessons from FSC and the HCS+ Study suggest that using single fixed thresholds can be problematic because of major boundary effects. It was recommended using either bounded categories or continuous variables with a bottom cut-off. For example, RSPO could determine eligibility for a HFC approach by considering remaining forest where the HCSA class available for development gradually gets stricter as forest cover declines: e.g. 90-100% forest cover allows up to MDF to be cleared, 70-90% allows up to LDF, 50-70% allows up to YRF and <50% requires full HCSA implementation - Require “no conversion” of all natural ecosystems, but with a slightly less strict bar e.g. protect 50% of all natural ecosystems in the management unit - Mandatory compensation required for all conversion of HCS or other ecosystems - How to include smallholders? Have the same rules for all but give everyone a 10ha discount for conversion. This would act to support smallholders by allowing them enough clearance to have a viable livelihood. The threshold could of course be adjusted to reflect the farm size needed for a viable livelihood in different contexts.

6.2 Defining HFC countries The following options were proposed for how to define HFC countries: Option 1 Combine % forest cover (FAO) with % IFLs – measure of forest cover and development status. Countries must be above XX % for both variables to qualify: e.g. 70% forest, 20% IFL. • Pro: Uses both national and global datasets and spreads ‘risk’ • Con: Reliability of FAO data, controversy of IFLs

Option 2

57 “Support” to be defined by RSPO 49

RSPO No deforestation consultancy: High Forest Cover Countries

Combine % forest cover (FAO) with % IFLs, % oil palm area AND deforestation rate, to give a more comprehensive measure of forest cover and development status.

One scenario for this would be to use the following thresholds, resulting in only Gabon and PNG qualifying: • >70% forest cover AND; • IFL>20% AND;

• oil palm <0.5% AND; • Deforestation rate>0.1%

Annual % forest IFL (% land Oil palm (% deforestation Country (FAO) area) land area) (%; FAO)

Gabon 89.3 26.3 0.02 1.1

PNG 72.5 23.8 0.37 0.1

Option 3 Use a continuous or banded approach based on forest cover, e.g. >70% use one HCS threshold, 50-70% use another, <50% use HCSA. If the principle is supported it may be possible to explore including other variables as considered above.

Option 4 No threshold used. Countries specified a priori or based on qualitative factors, e.g. known palm frontiers, deforestation threat, higher forest cover.

Option 5 A combined semi-quantitative approach is used, combining quantitative variables of forest cover and % oil palm cover, with qualitative variables to assess whether the country is a palm frontier. Countries would have to meet all of these criteria to qualify. This approach was supported by the NODIG as follows:

Parameter Threshold

50

RSPO No deforestation consultancy: High Forest Cover Countries

Forest cover (REDD+ national (recent, >60% trusted))

Oil palm cover <1%

Deforestation trajectory Historically low, but increasing or constant

‘Threat’ from palm expansion Known frontier area or major areas allocated for development

HCSA trials indicate HFCL Trials of the HCSA toolkit indicate non-viable for development

6.3 HFC approaches 6.3.1 Site-level approach Various options for an RSPO HFC approach were presented to the NODIG and discussed within the group. A refined list of options presented prior to the TF5 meetings is presented here. Note that these options each include slightly different combinations of elements and principles. In practice RSPO could pick different combinations from each of these options:

Option Description

1 HCSA applies across the board: • Toolkit for low/medium FCLs • Legacy case procedure for HFCLs

• No new development in HFCLs • What about smallholders?

2 HCSA with threshold shifted, with or without cut-off date? • Ensures social benefit through HCSA social requirements, WIA • Carbon neutral, HCVs secured etc.

• Where is HCS threshold? LDF or MDF?

3 HCV only until 2020, with safeguards – HCSA applies everywhere from 2020? • Conserve: Develop = 1:1

• Apply HCSA social requirements and WIA to demonstrate social benefits • Carbon neutral

51

RSPO No deforestation consultancy: High Forest Cover Countries

• Compensation payments US$2,500 x ha developed

4 Smallholder only model – applies now to 2025? • Conserve: Develop: Food Security = 1:1:1

• Carbon neutral

• HCVs secured. Hunting pressure reduced. Min. 1km wide corridors connect intact areas

5 Zero net deforestation – applies now to 2025/30?

• Restore/conserve 1 ha for 1 ha developed. In same biome (ideally country)

• Carbon neutral, HCV approach, smallholder only model, HCSA social requirements & WIA

6.3.2 Wider landscape or jurisdictional requirements One concern about site-level approaches is that they may not adequately address deforestation drivers in the wider landscape and that conservation efforts in isolation are unlikely to be effective. Therefore, it was proposed to include some requirements as part of a HFC approach that require companies to act at these wider scales. Examples of requirements proposed were: • At a landscape level:

• Maintain HCV and HCS areas in the wider landscape (at least 5km from the concession boundary)

• Stabilise deforestation in wider landscape in 3 years

• Implement community livelihood and conservation programmes to demonstrate reduced deforestation in landscape

• Collaborate with other actors to ensure effective connectivity (at a minimum through physical corridors) • At a jurisdictional level:

• Engage actively with governments to agree policies for capping deforestation at 5% (TBC) from current level, and ensure no more concessions are allocated

• Actively participate in jurisdictional approaches to implement No Deforestation across all sectors by latest 2030 • Jurisdictional programme funded & implemented • Financial contribution from grower

52

RSPO No deforestation consultancy: High Forest Cover Countries

• Deforestation reduced in 3 years, halted by 2030… 6.3.3 Outstanding questions for the RSPO NODIG To finalise proposed RSPO P&C text for public consultation the above approaches and options will need to be reviewed by the NODIG and it will be necessary for the group to agree on the following elements:

• No HCSA smallholder approach yet. RSPO needs to agree what approach can be used to “limit deforestation” by smallholders in HFC contexts Minimum outcomes or Qs for sub-group • For an approach applied to smallholders or community land. What definitions of smallholder and community land?

• Scale of implementation/assessment: If only smallholder or community development is allowed, will assessments need to be conducted at a larger site/landscape scale?

• HCVs identified and maintained

• Carbon/forest: requirements on HCS threshold (MDF?) or carbon neutrality? • Ratio of Conserve: Develop: Food Security areas? E.g. 1:1:1

• Social requirements: FPIC assured and participatory land use planning. Use HCSA social requirements – but need to adapt for more landscape level? • Time-limited? Assessment process • HCV-HCS-FPIC assessment for the site/landscape or HCV for smallholders tool (not designed for landscape participatory planning)?

• Quality assurance? Can it use the existing HCVRN ALS system?

53

RSPO No deforestation consultancy: High Forest Cover Countries

Annex 1: RSPO TF5 HFC approach justification Thresholds for defining High Forest Cover Countries in the context of the RSPO P&C’s proposed criterion 7.13

1. Why use a 60% threshold for High Forest Cover?

The percentage forest cover of a country, i.e. the land area under forest cover as a percentage of the total land area of a particular country, is the most widely used approach in describing forest cover. Our analysis of the best available forest cover data for potential HFC countries indicate that a 50% threshold maybe too low as it would include countries with high historic deforestation and a now fragmented forest cover, while a 70% threshold would exclude some countries that have minimal areas of developable land based on field trials of the HCSA approach, such as Liberia.

2. Why use national or regional datasets as a first priority for determining country-level forest cover and FAO data as a second priority?

For the FAO data, it is largely left to the national governments to provide their forest cover data and there is no clearly defined process for ensuring data accuracy. As such, there is much variability in how national governments compile their statistics with some countries potentially under-reporting their forest cover (e.g. Liberia) or over-estimating their forest cover (e.g. Malaysia).

We opted to use national or regional datasets as it allowed use of more up-to- date statistics and to look for more reliable estimates, for example, developed for national REDD+ programmes using highly accurate land cover classification techniques. For example, REDD+ requirements there are protocols for national governments to provide their statistics for forest cover with a tiered approach for methodologies. There is a push for national governments to advance from Tier 1 to Tier 2 and Tier 3 with a corresponding increase in accuracy.

3. Why use a threshold of less than 1% oil palm cover?

In addition to using forest cover data as a criterion for HFC countries, there is also a need to have a criterion that serves as a proxy for the development status of a country, its current dependence on oil palm for economic development and the potential for oil palm expansion. The oil palm cover of a particular country, expressed as a percentage of the total land area of a country, serves this purpose and the data for a wide range of countries is readily available from the FAO Stats web portal. This threshold of 1% oil palm cover would allow for the inclusion of many countries that are widely perceived to be new oil palm expansion fronts and may benefit from adopting RSPO standards in avoiding the clearance of HCV areas and HCS forests. Countries that have a higher percentage of oil palm cover are assumed to be already deriving economic benefits from oil palm development and therefore may be less dependent on further forest conversion to oil palm in order

54

RSPO No deforestation consultancy: High Forest Cover Countries

to drive further economic growth. Other proxies for a country’s development status (e.g. Human Development Index, poverty levels) were also explored but based on lessons learnt from the Sustainable Palm Oil Manifesto (SPOM)’s HCS Study, these proxies were not very useful at a national level due to spatial income disparities and unequal development within a particular country.

4. Why use a combination of 60% threshold for HFC and 1% threshold for oil palm cover? Using forest cover alone as a criterion for HFC countries is not advisable due to issues with data inaccuracy and its inability to provide a good indication of a country’s development status or potential for oil palm development. Combining it with a criterion for oil palm cover allows for a more accurate selection of countries that still have a relatively high forest cover but are likely to see active conversion of forest to oil palm for the purpose of economic development.

A full list of countries considered and statistics used are provided at the end of this document.

5. Why use a threshold of Medium Density Forest (rather than LDF) for legacy cases and community development?

During discussions within the HCSA’s HFC Landscapes WG it was indicated by several representatives from Technical Support Organisations that the application of an LDF threshold would likely be very restrictive on the potential areas available for development – given that many of the case studies in HFCLs had minimal land cover of the YRF class and below.58

This approach is supported by evidence collected during this consultancy project indicating:

• Based on five HFCL case studies from the HCSA and SPOM, on average only 14% of AoIs was in the YRF land cover class (Range: 4 to 24%; case studies in Papua, PNG and Gabon), and

• Negligible areas of forest with carbon stock <35 tC/ha59 (a proxy for the lower threshold of YRF) at a national level for a sample of potential HFC countries:

58 Although it was also noted that there would need to be further analysis to verify this conclusion

59 It is noted that the HCSA methodology does not use 35 tC/ha as a fixed threshold for HCS forest, but in the absence of national level HCS maps this was used as a proxy for the lower threshold of the YRF class. 55

RSPO No deforestation consultancy: High Forest Cover Countries

Figure 12. Aboveground live woody biomass on the island of New Guinea. White areas are those with carbon stock less than 35 tC/ha. Based on data from Baccini et al. 2000

Figure 13. Aboveground live woody biomass in Gabon. White areas are those with carbon stock less than 35 tC/ha. Based on data from Baccini et al. 2000

56

RSPO No deforestation consultancy: High Forest Cover Countries

National statistics on forest cover, deforestation rates and oil palm area Source of Deforestation % % forest Oil palm Deforestation Deforestation Deforestation deforestation % Source for nat/reg rate (%/yr; Country forest (national/ (% total rate (%/yr; rate trend rate trend rate (if different IFL forest cover*** nat/reg (FAO) regional) land area) FAO) (FAO) (nat/reg) from nat/reg source) forest cover)

Gabon 89.3 26.3 88.3 OFAC 0.0 -0.9 Decreasing 0.1 Decreasing

Solomon 78.1 10.6 76.6 MinFR 2010 0.6 0.3 Increasing 0.3 Constant Islands

PNG 72.5 23.8 71.0 PNGFA 2012 0.4 0.0 Constant 0.1 Constant CCDA 2017

DRC 67.3 22.0 68.8 OFAC 2010 0.1 0.2 Constant 0.2 Constant

Metria & GeoVille; Liberia 43.4 2.7 68.0 0.2 0.7 Constant 0.9 Constant Winrock 2016

R of Congo 65.4 27.0 67.1 OFAC 0.0 0.1 Constant 0.5 Increasing

Myanmar 42.9 4.0 63.0 Bhagwat et al 2017 0.2** 1.8 Increasing 0.3 Constant

Malaysia 67.6 3.9 55.2 NRE 2014 15.2 -0.1 Decreasing 0.4 Decreasing REDD Desk

Indonesia 53.0 15.0 54.2 MinEF 2015 5.4 0.7 Increasing 0.9 Constant

57

RSPO No deforestation consultancy: High Forest Cover Countries

2013. Brazilian Brazil 58.0 21.8 54.0 0.0 0.2 Decreasing 0.5 Decreasing Forest Service (SFB)

Peru 57.8 33.5 53.0 MINAM 2015 0.0 0.2 Constant 0.1 Constant

Costa Rica 53.9 4.9 52.4 REDD Desk 1.4 1.1 Increasing -0.9 Decreasing

Sistema de Monitoreo de Colombia 51.0 24.2 51.7 0.3 0.0 Decreasing 0.3 Increasing Bosques y Carbono (2016)

Ecuador 50.5 16.3 51.6 MAE 2014 1.1 0.6 Constant 0.7 Decreasing

2013. OFAC (de Cameroon 39.8 6.8 51.4 0.3 1.1 Increasing 0.9 Constant REDD Desk Wasseige et al)

Honduras 40.8 3.5 48.0 UN REDD 1.4 2.4 Increasing 0.9 Constant

Guatemala 32.5 3.6 34.2 INAB 2012 1.3 1.0 Decreasing 1.5 Constant REDD Desk

Thailand 32.0 2.7 32.0 RFDT 2016 1.3 -0.2 Decreasing 0.1 Constant

CAR 35.6 0.7 15.1 OFAC 0.0 0.1 Constant 1.8 Decreasing

58

RSPO No deforestation consultancy: High Forest Cover Countries

*Countries chosen are known current or major potential future palm producer countries. Clearly low forest cover countries such as Ghana, Ivory Coast, Nigeria and Sierra Leone are omitted for brevity. Yellow highlighted countries are those that were reclassified compared to the provisional list in the TF5 text for public consultation, based on use of more accurate forest cover statistics: Peru no longer qualifies as a HFCC due to its forest cover being less than 60% and Myanmar is added as a HFCC due to its high forest cover and potential as an oil palm frontier. **Myanmar oil palm stats not available from FAO, taken from Baskett 2015

***Full list of references provided here:

Baskett, J.P.C. 2015. Myanmar oil palm plantations: A productivity and sustainability review Report no. 28 of Myanmar palm the Tanintharyi Conservation Programme, a joint initiative of Fauna & Flora International and the Myanmar area Forest Department.

Forest cover

Min of Env, Brazilian Forest Service. 2013. BRAZILIAN FORESTS at a glance. Available at: Brazil http://www.florestal.gov.br/publicacoes/571-brazilian-forests-at-a-glance-2013

CAR, Gabon, Les forêts du bassin du Congo – État des Forêts 2013. Éds: de Wasseige C., Flynn J., Louppe D., Hiol Hiol F., RoC, Cameroon Mayaux Ph. – 2014. Weyrich. Belgique. 328 p. Dépôt légal : D/2014/8631/30 ISBN : 978-2-87489-298-1

Costa Rica https://theredddesk.org/countries/costa-rica/statistics

Ecuador Ministerio del Ambiente. 2014. Ecuador’s Forest Reference Emission Level for Deforestation

Guatemala https://theredddesk.org/countries/guatemala/statistics

Honduras http://www.unredd.net/index.php?option=com_docman&task=doc_download&gid=13374&Itemid=53

59

RSPO No deforestation consultancy: High Forest Cover Countries

MinEF, Republic of Indonesia 2015. NATIONAL FOREST REFERENCE EMISSIONS LEVEL FOR REDD+ In the Indonesia Context of Decision 1/CP.16 Paragraph 70 UNFCCC

Metria & GeoVille 2016. Final Report LIBERIA LAND COVER AND FOREST MAPPING FOR THE READINESS PREPARATION ACTIVITIES OF THE FORESTRY DEVELOPMENT AUTHORITY; Goslee et al 2016. Development of Liberia Liberia’s REDD+ Reference Level Draft Final Report for Republic of Liberia Forest Development Authority. Winrock Int.

http://www.nre.gov.my/en-my/Forestry/Pages/Statistics-Forest.aspx ; Malaysia https://theredddesk.org/countries/malaysia/statistics

Bhagwat T, Hess A, Horning N, Khaing T, Thein ZM, Aung KM, et al. (2017) Losing a jewel—Rapid declines in Myanmar Myanmar’s intact forests from 2002-2014. PLoS ONE 12(5): e0176364. https://doi.org/10.1371/journal.pone.0176364

Peru MINAM 2015 http://sinia.minam.gob.pe/documentos/cobertura-perdida-bosques-nivel-nacional

PNGFA 2012 Forest Base Map; Climate Change and Development Authority 2017. Papua New Guinea’s PNG National REDD+ Forest Reference Level Submission for UNFCCC Technical Assessment in 2017.

Solomon Islands Ministry of Forestry and Research 2010. http://mofr.gov.sb/foris/forestArea.do#marker

Thailand Royal Forest Department of Thailand 2016. http://forestinfo.forest.go.th/Content.aspx?id=72

60

RSPO No deforestation consultancy: High Forest Cover Countries

Annex 2. Profiles of selected HFC countries Gabon % forest cover 88%

Deforestation rate 0.1%

Deforestation Palm, rubber, subsistence agriculture drivers Degradation: logging

Socio-economic For the region a wealthy country but very inequal with status ~30% of the population living in poverty, most of whom rely on subsistence or small-scale agriculture.60

Types of producers Large-scale palm and rubber; small numbers of subsistence smallholders

Policies in place Ambitions of becoming a leading palm oil exporter by increasing production from 13,000 to 280,000 tons year−1 by 2025 (Republique Gabonaise 2011, 2012, in Burton, et al. 2016).

National palm oil strategy in development, to identify nationally defined HCVs and high carbon stock forest to protect

BAU? Allocated There are currently 130,680 ha of oil palm concessions in land/concessions the country of which 58,980 ha is plantable, based on environmental, social, and agronomic suitability

Large parts of the country allocated for forestry concessions

Papua New Guinea % forest cover 73%

Deforestation rate 0.1%

Deforestation Subsistence agriculture & industrial forestry drivers Degradation: industrial forestry

Socio-economic 38% of Papua New Guineans live in poverty, 41.3% in status rural areas.61

60 https://borgenproject.org/poverty-in-gabon/

61 https://devpolicy.anu.edu.au/pdf/2011/Rural_Poverty_Remote_PNG_Report.pdf 61

RSPO No deforestation consultancy: High Forest Cover Countries

Types of producers A handful of large palm companies, with the two main companies RSPO members or certified. 100% of exported CPO is RSPO certified.

Several new, uncertified members have recently begun developing.

Majority of development on community land leased from communities – communities receive land rent and profit shares.

Policies in place Government Development Strategic Plan 2010-2030, Vision 2050, including a goal to increase palm production from 556,000 tonnes in 2007 to 1.5 million tonnes in 2030.

REDD+ country with a national REDD+ programme in development to reduce deforestation

BAU? Allocated 97% of land is customarily owned in PNG so in theory land for communities decide on how to use their land. However, development corruption has exploited this through Special Agricultural Business Leases. Although these are being clamped down on now, there are also problems with unscrupulous actors making agreements with clan leaders to develop palm or forestry operations without true FPIC of wider community members

Increasingly new palm companies are establishing in PNG that are not RSPO members, including logging companies that are developing palm having exploited all timber

62

RSPO No deforestation consultancy: High Forest Cover Countries

Annex 3: High Forest Cover case studies synthesis The following case studies were shared with the consultants for this study. The full results are considered confidential, and so only a synthesised analysis of the studies is provided in the table below:

1. Hargy/Sipef, New Britain, PNG 2. HCS+ Study, Gabon, Mouila 1&2, Olam 3. Golden Veroleum, Liberia 4. NBPOL, PNG 5. Sime Darby, Liberia 6. Goodhope, Papua Province, Indonesia

For non-confidential topics, the numbers above are used to reference sources of the following information. Issue Summary analysis of case studies

SOCIAL Local communities (most of the time they are land owners) pressure companies for oil palm development because it is No deforestation (1,2,3,4): vs. oil palm for community • A source of income income • A source of employment • Development in a wider sense In some cases pressures from local communities for development can be very severe, e.g. there was one case of violent attacks on a plantation by local communities asking for jobs.

Government Governments are also pushing companies for development and expansion (3, 4)

What were the In most case studies, “sustainable” oil palm is not the main or main threats to only threat to forests. The main identified threats are: forest? Was oil • Logging (1) palm always the • biggest one? Village oil palm (1) What else? • Non-certified oil palm producers (1) • Subsistence agriculture is the primary driver of deforestation (4) • Industrial forestry (forest degradation) (4)

Were there other In certain contexts, e.g. PNG, it is often mentioned that if the oil palm current company does not develop land offered by companies in companies, that communities are likely to offer the land to area? other companies to develop. In East New Britain there are several other non-RSPO companies now

63

RSPO No deforestation consultancy: High Forest Cover Countries

What were the Opportunities opportunities or • Development may reduce pressure in communities to potential open up areas to community logging and farming, solutions allowing support for efforts to promote forest proposed conservation in high priority national parks and targeted protection zones • Potential for poverty reduction • Source of employment • Basic infrastructures Potential solutions proposed in case studies • Adapted carbon threshold to national context • Thresholds to permit agriculture in previously degraded forests while avoiding deforestation of older secondary and primary forests • A combined national level prioritization exercise to zone out critical carbon and biodiversity areas and then more degraded forests open for development but only if companies follow strict site-level due diligence, e.g. HCV • Set-asides ratio at the concession ensuring no-net carbon emissions from plantation development • If the context requires forest conversion for development, emissions could be offset through set-asides at the concession level to prevent immediate emissions and sequester carbon under proper management. Example from Gabon: a rule-of-thumb ratio of 2.4-2.6 conserved hectare to each converted hectare of forest should be set- aside if the 118 Mg C ha−1 threshold is applied, to ensure carbon neutrality (2) • Conduct a landscape study consisting of 3 steps to determine possible development (4): ➔ Due-Diligence: desktop research and initial stakeholder discussion ➔ Draft Land Use plan: HCS, HCV, ESIA and other relevant assessments, a report with various development options is proposed ➔ Draft Management Plan and Agreement: full management Plan including monitoring is proposed and agreed by key stakeholders Risks of not • Externalise deforestation to less sustainable actors allowing some • Social conflicts: development in o Communities with more forest may receive less HFCLs income from oil palm than those with less forest

64

RSPO No deforestation consultancy: High Forest Cover Countries

o Companies prioritising low carbon areas may be more likely to then conflict with community use areas

65

RSPO No deforestation consultancy: High Forest Cover Countries

66