Can savanna burning projects deliver measurable greenhouse emissions reductions and sustainable livelihood opportunities in fire-prone settings?
Jeremy Russell-Smith, Catherine Monagle, Margaret Jacobsohn, Robin L Beatty, Bibiana Bilbao, Hebe Vessuri, Isabelle Sánchez, Adriana Millán
Supplementary Material
Online resource 1: Savanna conditions in Australia, southern Africa and South America
Australia
Savanna burning in Australia
The savannas of northern Australia occupy 1.9M km2 and occur mostly under markedly seasonal monsoonal rainfall conditions, generally receiving an average of >500 mm rainfall p.a (see Fig. 1a). Soils are mostly weathered and infertile, and rainfall occurs almost exclusively in the 4-5 month wet season period, Nov/ Dec to Mar / Apr. Although land use is dominated by extensive rangeland beef cattle production, it is economically viable only in more fertile settings especially in south-eastern, lower rainfall regions. Despite the appearance that relatively unmodified north Australian eucalypt-dominated savanna systems are structurally intact and healthy, it is increasingly recognised that contemporary land uses and changed fire regimes are having significant regional impacts on sustainable land use and biodiversity values (Russell-Smith et al. 2003, 2012; Woinarski et al. 2007, 2011).
The savannas are sparsely settled by international standards, with densities of <0.1 persons km-2 outside of scattered urban centres; only three regional cities have populations slightly in excess of 100,000 people. Most of the rural population is indigenous and this is projected to double in the next two decades. (Taylor 2006). Indigenous groups own substantial land holdings or, more commonly, share non-exclusive title (‘native title’) with pastoral land holders under recent Australian national and state or territory government legislated arrangements. With relatively few exceptions, indigenous people do not own the more productive pastoral lands and, despite being ‘land rich’, remain severely economically and socially disadvantaged (Whitehead et al. 2009).
The sparse population, limited infrastructure, and low economic base, has resulted today in fire regimes in many regional settings being unmanaged and characterised by the frequent (annual-biennial) recurrence of large (>1000 km2) wildfires occurring predominantly late in the dry season. An average of ~20% of Australia’s savannas region is burnt each year (Russell-Smith et al. 2007), with fire frequencies exceeding 50% p.a. in extensive higher rainfall regions (Felderhof and Gillieson 2006; Russell-Smith et al. 2009, Fig. 1b)). Development of these contemporary burning patterns follows breakdown in customary Aboriginal modes of fire management associated with societal collapse from the late 19th Century (Ritchie 2009). Over much of northern Australia burning traditionally was undertaken throughout the year, with a focus on implementing extensive ‘cleaning of country’ management through intensive application of small patchy burns in the early-mid dry season (Russell-Smith et al. 2003). Finding the economic means to reinstate such prescribed strategic management, and associated socio-cultural and economic opportunities for indigenous custodians, has been at the heart of developing savanna burning projects across fire-prone northern Australia (Whitehead et al. 2009; Heckbert et al. 2012).
Fig. 1 Northern Australia (a) mean annual rainfall isohyets, 1970 to 2009; (b) the frequency of fires from 1997 to 2011 (burnt area mapping derived by Landgate, Western Australian Government, from AVHRR imagery; (c) current savanna burning projects across north Australia. Refer text for details.
Western Arnhem Land
The Western Arnhem Land Fire Abatement (WALFA) project operates over 28,000 km2 of indigenous-owned land in rugged, very remote and fire-prone savanna in the ‘Top End’ of the Northern Territory (Fig. 1c). Commencing informally in 1997 as a landscape-scale fire management project at the behest of senior indigenous land owners, early objectives concerned (a) re-engaging younger and older generations with their traditional lands, (b) building capacity of regional indigenous ranger groups to implement a coordinated and strategic landscape-scale fire management program using both customary (detailed indigenous knowledge) and contemporary (satellite fire mapping, Geographic Information System—GIS, and aerial ignition technologies) toolkits, in order to (c) address a severe unmanaged late dry season wildfire problem, with resultant deleterious impacts on internationally significant biodiversity values (Russell-Smith et al. 2009; Whitehead et al. 2009).
From 2000, the WALFA scientific program incorporated development of a savanna burning GHG emissions accounting methodology, and associated recognition of the potential for strategic landscape fire management in the project area to reduce GHG emissions on an industrial scale. In 2005, a 17-year agreement was reached between WALFA landowners, the Northern Territory Government, and a transnational energy company, to annually offset 100,000 t.CO2-e from the LNG plant for a fee of AUS$1.1M p.a. (indexed to 2006). Over the period 2005-2011, effective fire management in the WALFA project area has delivered substantially in excess of its contracted commitment (Russell-Smith et al. in press). Whereas, to date, WALFA has operated essentially as a voluntary arrangement (sensu Bayon et al. 2006; Wunder et al. 2008), under Australia’s nationally regulated CFI scheme WALFA now has the opportunity to become an accredited offset project.
With the implementation of Australia’s emissions trading scheme from in mid- 2012 (albeit with a fixed price during the first 3 years), considerable interest exists for expanding WALFA-style savanna burning projects in other fire-prone regions of northern Australia. Much of that interest focuses on lands owned or managed by indigenous Australians. While it is our contention that, as illustrated by WALFA, it is eminently feasible to (a) operationally implement strategic fire management at landscape scales, and (b) apply robust and transparent GHG emissions accounting procedures, far more challenging are the multi-faceted, cross-cultural requirements for establishing effective and inclusive governance arrangements.
Australia’s NGGI Accounting Methodology
While the Kyoto Protocol includes savanna burning as an accountable activity (UNFCCC 1998: Article 3, Annex A), only Australia as a Tier 1 (developed economy) country currently reports on savanna burning in its national accounts (ANGA 2011). Recently, Australia has also made formal provision for recognising emissions reductions associated with enhanced savanna fire management as an accountable offset activity in its developing emissions trading system (DCCEE 2012).
In its most basic form, for any region, savanna burning emissions (E) are calculated as the product of the mass of fuel pyrolised (FP) and the emission factor (EF) of respective accountable GHG (g) species (E = FP * EF(g)), where FP is the product of the area exposed to fire (A) taking into account spatial patchiness, the fuel load (FL) in respective fuel classes, and the burning efficiency (BEF) is defined as the mass of fuel exposed to fire that is pyrolised. EF(g) is defined relative to the fuel elemental content where, for carbon species, it is expressed relative to fuel carbon, and nitrogen species are expressed relative to fuel nitrogen. Fuel carbon mass is determined from fuel mass by the fuel carbon content, while fuel nitrogen is derived from the fuel mass by the product of carbon content and the fuel nitrogen to carbon ratio (DCCEE 2012).
Critical features of this Australian methodology are that (a) fuel loads (grass, litter, coarse woody debris, shrub components) are defined for specific vegetation / fuel types, calculated with respect to fuel accumulation relationships determined from time-since-fire, as measured from satellite imagery; (b) emission factors for CH4 and N2O gases (i.e. their respective concentrations in smoke) differ between different fuel types (e.g. flaming combustion of grasses vs. smouldering combustion of woody fuels), but not between early and late dry seasons (Hurst et al. 1994; Meyer et al. 2012); and (c) fire spatial patchiness and burning efficiency factors vary significantly with fire severity, which in turn is strongly related to fire seasonality—i.e. early dry season fires typically are less severe than late dry season fires (Russell-Smith and Edwards 2006).
Southern Africa
Savanna burning in southern Africa
The savannas of southern Africa occupy c. 10M km2 (43% of the landmass) occurring under markedly seasonal rainfall conditions up to 1750mm p.a. and extending into areas of lower and less seasonal rainfall than in Australia and South America due to higher soil fertility (Lehmann et al. 2011). They are commonly distinguished as broad-leafed infertile savannas occurring on nutrient poor soils in higher rainfall areas, or fine-leafed fertile savannas on nutrient rich soils in the more arid areas, predominantly in Botswana, Namibia and South Africa (Fig. 2).
Fig. 2 African distribution of broad-leafed infertile savannas (dark shade) on nutrient poor soils with higher rainfall and fine-leafed fertile savannas (striped shade) on nutrient rich soils in more arid areas (adapted from White 1983)
Population density varies widely with areas outside of urban settlements ranging from <5 up to 50 persons km-2 with rural livelihoods dependent on traditional and subsistence farming methods of small-scale, labour intensive, rain-fed crops and livestock production. Natural resources are important for basic everyday needs and include grazing for livestock, croplands, building materials, wildlife, firewood, medicinal plants, wild fruits, honey, natural dyes and materials for crafts. The use of fire to manage or access these natural resources is prevalent for slash-and-burn agriculture, pasture management, forest product harvesting, pest control, hunting and honey collection.
Communal land tenure with various types of traditional / customary authority structures comprises the vast majority of the 16 southern Africa nations[1] south of the equator. Together with protected areas, including national parks and gazetted forests, they commonly occupy the infertile areas. Individualized leasehold or freehold land title is typical of fertile areas.
Savanna burning is regulated, with a few exceptions, by prevention- and suppression-oriented fire management legislation and policies, maintained since colonial administrations revoked local burning practices and control (Frost 1998; FAO 2006). Typically, insufficient and inconsistent land and fire management legislation, administered by centralized governments with limited capacity, inadequately address the appropriate use of fire on communal lands. The absence of clearly defined processes, roles and responsibilities for decision-making, combined with weak local-level governance and community capacity, results in uncoordinated savanna burning throughout southern Africa.
Widespread use of slash-and-burn agriculture in the late dry season leads to extensive uncontrolled fires across large expanses of southern Africa’s savannas every year. Most of the area burned is due to the accumulation of many small- to medium-sized fires from numerous ignitions, rather than the occasional extreme fire event; >50% of savanna burning occurs between August and November (Archibald et al. 2010) (Fig. 3). Burned area is greatly reduced outside protected areas that are utilized by humans and their cattle, and further reduced in areas of cultivation and settlement (Archibald et al. 2010).
Uncontrolled fires in southern Africa negatively impact community livelihoods through loss of natural resources, property and life, degrade ecosystems and biodiversity values and significantly contribute to global carbon emissions from savanna burning. Nearly all fires are of anthropogenic origin and in recent decades Community-Based Fire Management (CBFiM) strategies have been introduced in a few areas to address these issues. Success has been varied and is dependent on the existence of sufficient supporting policy and legislation, land tenure or at least legal access to land, institutional and community capacity (FAO 2011).
Fig. 3 The area affected by fire determined from an 8-year satellite burnt area product. Colours indicate the number of times pixels were classified as burned. Grey areas represent pixels that were classified as invalid over the time period; darker grey, less valid data (source: Archibald et al. 2010)
Caprivi Region, Namibia
Namibia is a predominantly arid country with a strong gradient of highly variable rainfall <25mm p.a. in the southwest to markedly seasonal and consistent rainfall of 650mm p.a. in the northeast (Mendelsohn et al. 2002). Savannas cover c. 50% of the country with broad-leafed nutrient poor savannas, supporting a population of 5 - 10 persons km-2, occurring in the northeast regions (Fig. 4a). Customary rights to natural resources are well established in Namibia and community-based natural resource management (CBNRM) policy and legislation are well developed. National programs have established c. 100 legally recognized community-based structures of communal area conservancies and community forests (NACSO 2012; Fig. 4b). In Caprivi Region (c. 15,000km2) these provide employment and supplement livelihoods, through tourism and utilisation of forest products, for communities situated on communal lands (70%) in close proximity to protected areas (28%) (Fig. 5a). They play a key role in sustainable wildlife and high value plant management of the region, particularly in areas of lower population densities (<5 persons km-2) and lower land use pressure / conflicts that enable more effective local-level governance. Experience from an innovative integrated fire management program (Integrated Rural Development and Nature Conservation Caprivi Program) implemented between 2006 and 2010 in over 10,000 km2 provides valuable insights into the potential for a savanna burning offset program in this representative fire-prone setting.
Fig. 4a Biomes of Namibia with broad-leafed infertile savannas in the northeast regions (Mendelsohn et al. 2002)
Fig. 4b Land tenure of Namibia with communal and protected lands comprising the majority of infertile areas
In Caprivi, frequent uncontrolled high intensity late dry season fires affect >50% of the area annually (NRSC 2002; IRDNC 2006) exhibiting the southern African trend of reduced fire frequency and burned area with increased human/cattle density, cultivation and settlement. Frequent uncontrolled fires negatively impact Caprivian community livelihoods through loss of life and property (houses, food stores and fences), and reducing the availability, productivity and long-term sustainability of natural resources leading to increased poverty (Kamminga 2001: FAO 2011). This scenario is important for potential savanna burning offset programs as reduction of area burnt annually and shift in burning seasonality to the early dry season, required for emissions abatement, is needed for providing sustainable livelihood benefits (Fig 5b).
Fig. 5a Land tenure of Caprivi Region with CBNRM structures adjacent to protected areas
Fig. 5b The area of Caprivi affected by fire between 1998 – 2005 where colours indicate the number of times burnt; white indicates no fire (NRSC 2002; IRDNC 2006)
The Caprivi Program demonstrates the use of controlled fire management to provide tangible livelihood benefits to communities in this setting through improved land use productivity and sustainability, reduced hazards, and impact of uncontrolled fires and improved natural resource availability. Controlled burning, typically in the early dry season (April – July), was prescribed to improve grazing, natural product harvesting and agriculture. The timing, intensity and frequency of burning was determined for specific land use objectives in specific locations. As in the Australian context, these controlled burns collectively reduce and fragment fuel load, creating a landscape of interlinked burnt and unburnt patches and corridors that minimize the occurrence and extent of late dry season fires. It is quite feasible to deliver emissions abatement based on this type of communal land use-oriented controlled fire management in these settings. Countries where frequent (annual/biennial) fires affect >50% of the land area, including Angola, Zambia and Mozambique (Archibald et al. 2010), highlight the potential of this scenario in some extensive southern African savanna regions.