Past approaches and future challenges to the management of fire and invasive alien plants in the new Garden Route National Park

How to cite this article: Kraaij T, Cowling RM, Van Wilgen BW. Past approaches and future challenges to the management of fire and invasive alien plants in the new Garden Route National Park. S Afr J Sci. 2011;107(9/10), Art. #633, 11 pages. doi:10.4102/sajs. v107i9/10.633 The recently established Garden Route National Park (GRNP) along the Cape south coast of South Africa occurs in a landscape where indigenous forests, fire-prone fynbos shrublands and fire-sensitive plantations of alien invasive trees are interspersed. We used the area as a case study in the challenges facing conservation managers in the achievement of biodiversity goals in a fire-prone environment. We explored the context within which fire management was practised during the past century by interviewing former catchment managers and reviewing forestry and catchment management policies. Mountain fynbos adjacent to plantations was subjected to burning regimes aimed at the protection of commercial timber resources rather than the preservation of fynbos biodiversity. Prescribed burning of fynbos adjacent to the plantations was typically done in multiple belt systems at rotations of about 4–8 years during spring, summer and autumn, to avoid the winter berg wind season. Such short-rotation and low-intensity fires favour resprouting graminoids over slow-maturing reseeders, and likely account for the compositional impoverishment observed in fynbos near plantations. Current and future challenges faced by the GRNP include (1) balancing conflicting fire management requirements for plantation safety against fynbos conservation; (2) the continual invasion of fynbos by fire-propagated alien pines sourced from plantations; (3) inadequate resources to redress the ‘invasion debt’ caused by the socio-economic legacy and past management neglect; and (4) fragmentation of land use between conservation and forestry threatening the sustainability of the region at large. We provide recommendations for management actions and research priorities to address these challenges.

The recently established Garden Route National Park (GRNP) along the Cape south coast of South Africa occurs in a landscape where indigenous forests, fire-prone fynbos shrublands and fire-sensitive plantations of alien invasive trees are interspersed.We used the area as a case study in the challenges facing conservation managers in the achievement of biodiversity goals in a fire-prone environment.We explored the context within which fire management was practised during the past century by interviewing former catchment managers and reviewing forestry and catchment management policies.Mountain fynbos adjacent to plantations was subjected to burning regimes aimed at the protection of commercial timber resources rather than the preservation of fynbos biodiversity.Prescribed burning of fynbos adjacent to the plantations was typically done in multiple belt systems at rotations of about 4-8 years during spring, summer and autumn, to avoid the winter berg wind season.Such short-rotation and low-intensity fires favour resprouting graminoids over slow-maturing reseeders, and likely account for the compositional impoverishment observed in fynbos near plantations.Current and future challenges faced by the GRNP include (1) balancing conflicting fire management requirements for plantation safety against fynbos conservation; (2) the continual invasion of fynbos by fire-propagated alien pines sourced from plantations; (3) inadequate resources to redress the 'invasion debt' caused by the socio-economic legacy and past management neglect; and (4) fragmentation of land use between conservation and forestry threatening the sustainability of the region at large.We provide recommendations for management actions and research priorities to address these challenges.

Introduction
Fire has been a key process and evolutionary force shaping plant traits and vegetation communities across the globe for much of its history. 1,2It has been the most ubiquitous terrestrial disturbance, surpassed more recently only by human transformation of the landscape. 3Anthropogenic changes in land use have in turn resulted in modifications in the way fire occurs in space and time, 1 along with changes in our perceptions of fire and demands placed on land management agencies to protect lives and property. 4,5,6 the exceptionally diverse 7 and threatened 8 Mediterranean-climate (summer drought, winter rain) biotas of the world, 9 fire is the most important ecological disturbance factor and predates humans in these ecosystems. 2,10Mediterranean floras have evolved specialised post-fire persistence traits, which are sensitive to the specifics of fire regimes, such as seed banking in the soil or canopy, resprouting, and fire-stimulated flowering and germination. 9,11Fire is instrumental in maintaining diversity in the fynbos of the Cape Floral Kingdom (CFK) of South Africa, 12 and may be considered the most important fynbos management practice, being both a key ecological factor and a practical tool for resource manipulation.
The fire ecology of fynbos has been well researched since the early 1970s and by the 1990s fairly detailed fire management prescriptions were available. 13,14,15However, much emphasis was on the western, strictly winter-rainfall part of the CFK 12,16 and the inland arid mountains, 17,18,19 whereas the eastern coastal part of the CFK has been neglected.The climate of the eastern coastal part is less seasonal (rainfall is bimodal 20,21 ) and species' phenology, 22 and possibly plant growth and maturation rates, differ accordingly, which has implications for the management of fire regimes.
Fuel-reduction burning in fynbos (as in south-eastern Australia 23 ) largely developed from the early 1900s in response to the need to protect commercial timber resources.Early fire legislation has hence been embedded in forestry acts. 24,25Fire management practices aimed at hazard reduction are often in conflict with ecological objectives. 6,23Simple management compromises intended to reconcile conflicting objectives may ultimately not achieve either hazard reduction or biodiversity conservation. 23Therefore, management agencies have to set very clear and realistic objectives to determine the most appropriate management practices for each particular area. 24,26In the fynbos, as elsewhere, invasion by fire-adapted plants can complicate fire management.Invasion of fynbos ecosystems by invasive trees and shrubs, and notably by pines, is one of the largest threats to conservation. 15These invasive pines originate from commercial plantations that have been established throughout the CFK, and they often exacerbate the potential negative effects of both altered fire regimes and invasive species. 1,10,15proved understanding of fire ecology is becoming increasingly important as climates change, protected area networks expand, pressure from alien invasive biota increases, the wildland-urban interface enlarges, and demands to manage fuel loads of natural habitat for asset protection grow. 4,15,27,28Where new protected areas are established, systems are likely to be poorly researched and management prescriptions may have to be made in the absence of a clear understanding of ecosystem processes and responses.Recent additions to protected areas are often not in pristine condition and are affected by historic management that has implications for management into the future.
The recent establishment of the Garden Route National Park (GRNP; ~130 000 ha 29 ) along the Cape south coast of South Africa involved the amalgamation of certain parcels of land.These land parcels have in the past been variously managed for water conservation, biodiversity conservation, plantation forestry using alien invasive species, and the harvesting of natural resources, mainly timber from indigenous forests. 30e used the fynbos areas of the GRNP as a case study for exploring the challenges facing conservation managers in the achievement of biodiversity goals in a fire-prone environment.Our analysis was underpinned by a number of issues.Firstly, the GRNP is located within the southeastern coastal part of the CFK where fynbos fire ecology is inadequately understood.Secondly, it is a new park, and a reconstruction and critical review of past fire management approaches is needed.Thirdly, the institutional history and landscape context of the park pose particular difficulties that need to be addressed by the fire management policy, including high levels of invasion by alien trees, and significant pressure from the adjacent plantation industry to reduce wildfire hazard.Being a new park, there is still opportunity to influence the management policy and practice in the interest of biodiversity conservation.
In this paper we present the environmental and institutional context within which historical catchment management practices in the region of the GRNP evolved, and we consider how the adoption of a new mandate, with conservation as its central goal, will require changes to research priorities, management actions and land-use practices.

Biophysical environment
The study area is broadly defined as the southern slopes of the Outeniqua Mountains east of the Touw River, and the southern slopes of the Tsitsikamma Mountains (22.59°E -24.26°E; hereafter collectively referred to as the Garden Route coastal mountains, GRCMs), with emphasis on those areas recently proclaimed as part of the GRNP 29 (Figure 1).The GRCMs form part of the Cape Fold Belt Mountains 31 and run in an east-west direction, parallel to the coast. 32The highest peaks in the eastern Outeniqua and Tsitsikamma Mountains are 1469 m and 1675 m, respectively.Deeply incised remnants of an early Cenozoic peneplain form the coastal foreland south of these mountains. 32The Table Mountain Group rocks of the Cape Supergroup are the main mountain-forming substrata. 31Acidic (pH 3.3 -pH 5.5) lithosol soils, which are moderately deep, dark-coloured loamy sands, generally poor in bases, phosphorus and nitrogen, predominate. 33,34ing to maritime influence, the climate of the GRCMs is relatively temperate.In contrast to the strictly Mediterranean climate in the western part of the CFK, rain falls throughout the year. 21,22Rainfall peaks in spring and autumn, winter months are the driest, and the proportion of summer rain increases eastwards. 35Annual rainfall increases eastwards, the mean for the Outeniqua and Tsitsikamma Mountains being 820 mm and 1078 mm, respectively. 34,36Rain is mostly cyclonic and orographic with occasional thunderstorms. 34ightning occurs throughout the year and at a density of < 2 flashes/km 2 /year. 32Temperatures are mild with mean minima and maxima ranging from 7 °C and 19 °C in June to 15 °C and 26 °C in January. 34,36South-easterly winds prevail in summer and north to north-westerly winds (hot dry berg winds) prevail in autumn and winter, whilst south-westerly winds occur throughout the year. 34,36Berg wind conditions in particular increase the fire potential 19 and are associated with a higher incidence of fire, as well as an increase in the severity and size of fires. 21e coastal slopes of the eastern Outeniquas contribute run-off to various rivers, some of which support estuarine systems of national and international importance 37 (Figure 1).The coastal slopes of the Tsitsikammas are drained by short rivers flowing through deep gorges incised through the coastal plain. 36In 1986, the water draining from the statemanaged GRCMs was estimated to be 1046 million cubic metres per annum, then valued at R178 billion. 30Sustained yields of water from these catchments are vital to agriculture and coastal towns in the area, 38 a dependency that has been highlighted by recent severe droughts. 39e vegetation of the area comprises fire-prone and firedependent fynbos shrublands, interspersed with lesser areas of fire-free and fire-resistant Afrotemperate forest.The fynbos mostly occurs between the mountain crests and the belt of indigenous forest on the coastal platform at the foot slopes of the mountains 32 (Figure 1).South Outeniqua Sandstone Fynbos and Tsitsikamma Sandstone Fynbos occur to the west and east of the Keurbooms River, respectively. 33oth these vegetation types are tall, medium-dense proteoid shrublands, with medium-tall, dense ericoid-leaved shrub understoreys and a prominent restioid component.Ericaceous fynbos dominates at high altitudes, grassy fynbos at lower altitudes, and forest in fire refugia. 32,33ndian Ocean The extant fynbos of the GRCMs thus typically abuts the commercial plantations to the south.Depending on the topography and altitudinal reach of the plantations, the fynbos forms a belt of 2 km -6 km wide, and is generally narrower along the Outeniquas than along the Tsitsikammas (Figure 1).Considerable tracts of unprofitable plantations in the Outeniqua Mountains are in the process of being phased out, scheduled for rehabilitation to fynbos and incorporation into the GRNP during the next decade. 40,41The landscape setting of the GRNP is thus a mosaic of fynbos and forest amongst remaining and decommissioned plantations, most of which were formerly owned and managed by the state (the extent of these land groupings is indicated in Table 1 and Figure 1).The mix of fynbos and fire-prone pine plantations in the landscape has led to two major management problems: (1) the need to protect the plantations from fire, which kills the adult trees; and (2) invasions by fire-adapted pines, where fire drives the rapid spread and proliferation of these species 42 (Figure 2).

Institutional context
Since the early 1900s, water and soil conservation have been central to the management (and acquisition in places) by the South African state of fynbos mountain catchments, including the approach to fire management in these  areas. 15,24,30However, plantation-based timber production has from early on been recognised as a major land use and catchment management consideration in the GRCMs, 30 perhaps more so than in other fynbos catchments.The first state plantations in the area were established in 1883 near the town of Knysna. 24,43The state has since then, until the early 1990s, actively encouraged and subsidised the industry as a strategic move to provide for the country's demand for timber and to alleviate pressure on the limited and overexploited indigenous forests. 43,44Considerable expansion occurred in response to timber shortages experienced during the First and Second World Wars.Afforestation was further promoted as a means to relieve poverty amongst unemployed White people (1918 -1938) and Italian prisoners of the Second World War. 24,43,44'Off-site planting' resulted from the application of poverty relief labour, whereby some of the areas planted were unsuitable for commercial plantation forestry. 44The former Cape Province (including the present Western and Eastern Cape Provinces) was the only province where state plantations covered more area (three times more) than private plantations. 45The state was thus the main player in the forestry industry in the Cape Province and largely focused on growing pines. 46e generally held notion at the turn of the 19th century was that forests increase rainfall. 24Afforestation was accordingly seen as a beneficial land use in mountain catchments which otherwise had limited economic potential. 24However, droughts, public complaints about reductions in stream flow, and alleged desertification of South Africa during the first half of the 20th century prompted the Department of Forestry to undertake hydrological research in mountain catchments. 24,47his research exposed the high water consumption of plantation trees, 48 which led to the introduction in 1972 of an afforestation permit system to regulate new afforestation according to impacts on catchment run-off. 43,49Although it was by then recognised that afforestation competes with the conservation of water supplies and floral diversity, plantation forestry in the GRCMs was justified because 'surplus water [is] presently still flowing into the sea from the humid coastal mountain ranges' 30 .
For most of the 20th century, the state-owned mountain catchments of the Outeniquas and Tsitsikammas were managed by the national Department of Forestry (DoF) in its various forms 38,45 -either an independent department, or a branch of the national departments of Environment Affairs or Water Affairs 25,50 (for simplicity, we henceforth refer to DoF).The state managed a combination of fynbos, indigenous forests and exotic timber plantations on state land.During the mid-1970s, DoF adopted a formal policy of 'multiple use', which included timber production (from both indigenous forests and plantations), fire management and alien invasive plant control in the catchments, soil and water conservation and recreation. 24,25,30,51The same labour force that planted and tended pine trees also cleared alien invasive plants and conducted prescribed fires.The managers of these forestry estates were compelled to consider all aspects of this 'multiple use'.However, governmental restructuring of DoF during the mid-1980s, separated plantation forestry (a commercial undertaking, earmarked for privatisation), catchment fynbos and indigenous forest management (conservation undertakings), and research. 25,49,52,53The restructuring divided the land into (1) plantations, which were transferred to the South African Forestry Company Limited (SAFCOL); (2) fynbos areas that had not been afforested and were to be devolved to provincial nature conservation agencies; (3) indigenous forests, which remained under the jurisdiction of DoF; and (4) research areas, which went to the Council for Scientific and Industrial Research. 49,54llowing restructuring, large tracts of mountain fynbos within the former Cape Province were transferred to Cape Nature Conservation (which later split into CapeNature in the Western Cape Province and Eastern Cape Parks and Tourism Agency in the Eastern Cape Province), accompanied by reductions in state funding of conservation management functions.However, DoF retained responsibility for large  areas of the state-owned catchments in the GRCMs to the north of plantations for the purposes of fire protection. 53,55Prior to the restructuring, DoF largely achieved the fundamental principles of catchment management, that is, nature conservation 51 and sustainable water delivery, by applying appropriate fire and alien plant control regimes.With the commercialisation of state-owned timber plantations and the establishment of SAFCOL (owned by the Department of Public Enterprises) in 1993, the focus shifted to explicitly growing trees for profit. 44,54Nationally, privatisation of SAFCOL was largely finalised by 2002, although a buyer could not be found for the Western and Southern Cape plantations (including the study area), packaged as Mountain to Ocean (MTO) Forestry. 56MTO Forestry was only sold to Cape Timber Resources in 2004, based on a 70-year land-lease agreement with DoF. 54,56However, SAFCOL retained a 25% share on behalf of the government in all privatised plantations. 41The costly burden of catchment management (and in particular fire and alien plant management), which yields no financial return on investment, was not the primary mandate of the plantation companies.Likewise DoF, the landowner, has neglected to assume or delegate management responsibility for those areas originally retained for the purposes of fire protection of the plantations. 55In the process, large tracts of unafforested fynbos in the GRMCs were left without a custodian and became known as the 'no-man's land'. 53anagement of the GRMCs languished for almost 20 years. 44nly recently has the management of this 'no-man's land' been assigned to South African National Parks (SANParks) with the incorporation of the formerly DoF-managed forests and fynbos into the new Garden Route National Park. 29,41 2000, the South African government decided to phase out plantation forestry in much of the Western Cape Province, including the Outeniqua region, as these plantations were economically and environmentally unsustainable. 40,41,54pproximately 45 000 hectares of plantations (most of which were in the Outeniqua region) were to be felled over a 20-year period and converted to other land uses. 41,54,56,57However, this decision was partly reversed in 2006, 77 with approximately half the area recommissioned for plantation forestry (Table 1; Stehle T 2010, personal communication, November 25) on account of changing markets and the national demand for timber exceeding the growth of plantations. 41The decision was based on the presumption that productivity of the plantations in the Outeniqua region could be substantially increased by appropriate silviculture, site-specific soil preparation and continuous fertilisation. 58

History of catchment and fire management
The general approach to fire management in fynbos mountain catchments during the 20th century is presented elsewhere. 15,30Here we specifically focus on the approach to catchment and fire management in the GRCMs during the past century, as informed by our review of historical policy and management documentation and interviews with past and current land managers (Table 2).
Prescribed burning of mountain fynbos became fully accepted as a management practice during the 1970s to promote water, soil and biodiversity conservation. 30The general approach to fire management was based on, and continually influenced by, sound ecological principles emerging from a productive fynbos research programme at the time. 14,15However, and in contrast to policies elsewhere in the fynbos, nature conservation was the primary objective only in parts of state catchments zoned as nature reserves.Fire protection of plantations was the primary objective in vast areas of fynbos in the GRCMs to the north of the plantations (Table 2), where high-hazard, berg wind-driven fires characteristically originated.
For the conservation of fynbos, moderate to high-intensity fires at 10-20 year intervals are optimal. 15Conversely, for plantation protection purposes, it is desirable to burn adjacent fynbos vegetation under cool, safe conditions as soon as there is sufficient fuel available to sustain a fire, usually at vegetation ages of 3-8 years, depending on site characteristics.Various systems have been proposed and/or pursued in the GRCMs in an attempt to reconcile this fundamental conflict of interest. 23The most common approach has been the socalled double-(or triple-) belt system (Table 2).Accordingly, fynbos to the north of the plantations was divided into two to three parallel belts, each burnt at a fixed rotation, for example every eight years but four years apart.The vegetation age in one of the belts would consequently always be four years or less, thereby reducing the likelihood of fires spreading from catchment fynbos to the plantations.In some cases, plantation managers aimed to burn the fynbos adjacent to the plantations as soon as it could burn.
It is clear that fire management in the GRCMs has, since the establishment of the plantations until the early 1990s, been primarily aimed at the protection of commercial timber resources (Table 2).Although it is unlikely that catchment managers managed to execute prescribed burning every time and everywhere as planned, it seems reasonable to assume that large parts of the GRCMs' fynbos have at times been burnt at shorter fire return intervals, and lower intensities, than those deemed ecologically desirable for fynbos conservation.Both frequent and low intensity fires in fynbos and other Mediterranean-climate shrublands favour resprouters over slow-maturing reseeders. 59,60This preference may account for the dominance of graminoid sprouters and the paucity of slow-maturing reseeding shrubs of the Proteaceae family in parts of the GRCMs, particularly in areas near plantations (Kraaij T, personal observation).With regard to fire season, prescribed burning has mostly been carried out during spring, summer and early autumn (October to April) in order to avoid the winter berg wind season (May to September) associated with conditions of high fire danger and an increased risk of uncontrollable fires (Table 2).
The conflict of interest in terms of fire management (fire return periods in particular) in the GRCMs has been brought into sharp focus by the separation of plantation

Fire return periods Fire season and intensity 1920s
Deemed desirable that catchments be acquired and protected by the State.
Contemplated a policy of fire exclusion from Cape mountains (for water preservation), except in the vicinity of afforested areas where burning was to be carried out to reduce fire risk to plantations.
Mountain vegetation adjacent to plantations (often on private land) to be burnt as widely (100 m -200 m or more) and frequently as possible and grazing permits to be issued freely.
Vegetation in exterior fire breaks (50 m -100 m wide) to be slashed such that it could be burnt at a very young age.
Gum (Eucalyptus spp.) belts (producing little litter) to be planted on ridges as fire breaks.
Communal land adjacent to plantations to be ploughed and planted with kikuyu (Pennisetum clandestinum) grass to make it fireproof.
Mountain fynbos adjacent to plantations: as frequently as possible.
Forestry approach: summer is safest for prescribed burning.
Agriculture approach: winter is best to promote grazing.

1930s to mid-1940s
Exclusion of all fires from catchment fynbos except for prescribed burning of mountain belts to protect plantations.

Safety of plantations took precedence over fynbos conservation.
A series of fire breaks along northern boundary of plantations and in places along the mountain crest.
Mountain belts (each 300 m -500 m wide, often on private land) to be burnt (not ploughed) in double-belt or triple-belt systems where and whenever possible.
Limited burning of fynbos deeper into catchments, but reported that belt burning often spread upslope to the first crest.
Grazing permits issued freely.
Fire breaks and mountain belts: whenever possible.
Mountain belts: spring and early summer.
Burn during warmest, driest time of day to obtain clean burn.
Fire danger (berg wind) season to be avoided: winter (May-August).

Mid-1940s to late 1950s
Fire no longer had to be excluded from fynbos.
Adopted the principle of prescribed burning of more than just fire belts on a rotational basis to create a mosaic of veld ages within catchment.
State-managed catchments north of plantations throughout GRCMs to be burnt as soon as they could burn.
Double-belt or triple-belt fire breaks along the northern boundary of plantations and along mountain crest, and the area in-between divided into one or two (east-west) parallel belts (each about 1 km -2 km wide).North-south cut-offs subdivided these mountain belts into blocks, with the aim to create a mosaic of veld ages.
In some areas short rotation burning of belts along the northern slopes was additionally prescribed (e.g. at Witelsbos).
Private land adjacent to plantations to be burnt as often as possible.
Fynbos not immediately adjacent to plantations: 10 years.
Spring and early summer (September-December).

to late 1970s
Prescribed burning of mountain belts and, where necessary, blocks.
Double mountain belt along southern slopes up to the crest (belts each 300 m -500 m wide).
Fynbos close to plantations on private and state land to be 'tamed' by burning at 7-10 year rotation.
Fire breaks on northern boundaries of plantations were to be ploughed or hoed rather than burnt.
Reported that ~500 km of fire breaks or fire belts were burnt annually in the catchments of the Southern Cape.
All lightning fires and unplanned anthropogenic fires were to be extinguished, whether or not these threatened plantations.
Insurance of private plantations against fire damage became the norm. 41untain belts: 5-6 years (3 years apart).
Fire danger season: April-September.
Agriculture recommended winter burning to promote grazing value of vegetation.

Early 1980s
Fire protection of plantations considered the primary objective in GRCMs.
Fynbos conservation the primary objective only in designated nature reserves (i.e.Millwood Nature Reserve in the Outeniquas).
Block burning was common practice (mountains regarded as 'tamed').
Well-maintained double-belt or triple-belt systems (belts each 100 m -1000 m wide).
In places, an additional single-or double-belt system north of crest on private land.
Significant areas burnt in block burns.
State plantations were not insured against fire damage.
Increasing motivation for block burn rotations of ≥ 12 years.
12 year rotation triple-belt system (4 years apart) proposed to reconcile fire protection and fynbos conservation objectives.
Block burning: summer, and not from June or July to September or November as this was regarded as detrimental to reseeding plants.
Fire break burning: September-November after good rains.
Prescribed burning to be done under warm dry conditions but not with dropping air pressure.

Mid-1980s to mid-1990s
Plantation and conservation functions split within DoF, the former subsequently operated on a Trading Account.
Plantation management was instructed to cut down on fire protection expenses (i.e.external fire breaks or fire belts), as this was deemed the responsibility of another division within DoF.
Nationally, catchments were transferred to provincial conservation authorities, but GRCMs largely remained with DoF (except for Formosa and Millwood Nature Reserves).
Plantation management wanted prescribed burning of mountain belts (double-belt system with belts each 200 m wide) to continue as before, but neglect commenced in what became the 'no-man's land' and implementation fell behind.
A single belt or fire break of 200 m wide was maintained along the northern boundary of some plantations.Block burning ceased during the early 1990s.
All lightning fires were to be extinguished.
State plantations were not insured against fire damage.
Belt burning: October or November to March.
Block burning: January-April.Fire danger season: April-October.
'Mountain belts' (normally > 100 m wide) refer to the fynbos north of the plantations and south of the main mountain crests where fire management was generally geared towards protection of the plantations, and where fuel was regularly reduced by prescribed burning in order to reduce the risk and severity of fires.'Fire breaks' refer to narrower strips (normally < 100 m wide) that were burnt, hoed or brush cut (i.e.virtually devoid of fuel) along the boundaries of the plantations to prevent the spread of fire into the plantations, or to serve as access points and safety zones during fire fighting.The fire protection measures proposed or applied sometimes varied in different areas, accounting for ostensibly conflicting measures listed here.Burning frequency, season and intensity indicated are those deemed appropriate for prescribed burning operations.
Table 2 continues on the next page → and catchment management functions within DoF, and later the privatisation of plantations and handover of catchment management to conservation agencies.Catchment management has further been complicated, ecologically and economically, by the extensive infestations of invasive alien plant species (particularly Pinus spp.) almost exclusively sourced from adjacent plantations, with hardly any invasive alien plant control carried out in the 'no-man's land' in almost two decades 61 (Figure 2).

Fire management
Sound ecological management of fires in the study area is constrained by various factors.Repeated institutional disruptions, that is, transfers of DoF between different state departments and changing land management agencies, resulted in poor record preservation and data continuity.The history of fires in the study area has therefore been inadequately documented.Fynbos fire ecology in the southeastern CFK is poorly understood in terms of appropriate fire season and return interval. 15,22Moreover, fundamental conflict exists between fire intervals required for the reduction of fire hazard to commercial timber plantations, and those deemed necessary for the conservation of fynbos diversity. 6,23gative feedback mechanisms between fire, alien invasive trees and water resources threaten the achievement of the goals of sustained yields of water 48 from the GRCMs, which are important catchment areas. 53Fire behaviour and ease of access for the purposes of prescribed burns or fire fighting are negatively affected by the invasion of fynbos by alien shrubs and trees.The invaders increase fuel loads and consequently fire intensity, particularly under extreme conditions. 62Extreme fire intensities increase the risk to infrastructure and assets (such as plantations) and fire fighters, and detrimentally affect post-fire recovery of fynbos, soil and water conservation. 63,64,65Alien pines are spread and proliferated by fire, 42 which in turn exacerbates their detrimental impacts.Areas in close proximity to plantations, where fire safety measures in the GRCMs ought to be focused, are often the worst invaded, thereby rendering prescribed burning and fire fighting operations impossible.Pines had invaded 54% of the catchment of the Keurbooms River in 1999 to some degree, causing an estimated 22% reduction in river flow. 66It was further estimated that in the absence of

Mid-1990s to early 2000s
Complete separation of catchment and plantation management with the establishment of SAFCOL.
Block burning in catchments largely ceased as funding declined.
Conservation agencies adopted natural fire zone management in some fynbos areas, whilst neglect prevailed in the 'no-man's land'.
Alien invasive plant clearing regarded to be a 'Working for Water' function.
Mountain belt system replaced with ~50 m wide fire breaks (tracer belt and brushcut) along northern boundary of plantations.
Plantation management (what became SAFCOL) wanted the fynbos north of the plantations to be burnt at 8-10 year rotations as before, but with institutional separation, catchment management no longer had an interest in this expensive undertaking without direct benefits.
To cut down on expenses, SAFCOL largely reduced or stopped block and belt burning, as well as post-harvesting burning within plantations to suppress weedy growth of coral fern (Gleichenia polypodioides).
SAFCOL plantations were insured against fire damage.
Prescribed burning: October-November and February-March.

Early 2000s to present
SAFCOL plantations sold to Mountain to Ocean Forestry (MTO; DoF remained the landowner), and management of a large portion of GRCMs fynbos transferred to SANParks.
Conflict of interest between requirements of plantation industry for short rotation burning of fynbos adjacent to plantations, and ecological objectives (longer fire rotations and allowing natural fires) of protected areas.
Limited resources available to SANParks to rectify fire and invader plant management after 20 years' neglect.
MTO intended to reinstate fire break and block burning, with 10 m -30 m wide fire breaks (slashed, hoed or burnt) along northern boundaries of plantations, whilst larger mountain belts would revert back to fynbos conservation management.
MTO and SANParks agreed on jointly implementing prescribed burning in high fire danger zones (divided into double-belt) within Tsitsikammas.
History of catchment neglect (particularly the extent of invader plants) makes implementation of prescribed burning under conditions of acceptable fire danger largely unachievable in these zones.
The approach to fire management in remote parts of SANParks-managed catchment is that of adaptive interference, 13 with lightning regarded as the main source of ignition and interventions limited to where or when necessary.
MTO plantations are not insured against fire damage, only public liability coverage.
MTO sued the land owner (DoF) and neighbouring catchment management agencies (SANParks and Eastern Cape Parks Board) for plantation losses suffered during the large 2005 fire in the Tsitsikamma Mountains.
High fire danger zones in Tsitsikammas: 8 years (4 years apart) ; not yet established for Outeniquas.
Prescribed burning by plantation managers: October-November and February-March (i.e.no berg winds and less erratic weather conditions).
Prescribed burning by protected area managers: November-April, but may be broadened 22 pending local research.
'Mountain belts' (normally > 100 m wide) refer to the fynbos north of the plantations and south of the main mountain crests where fire management was generally geared towards protection of the plantations, and where fuel was regularly reduced by prescribed burning in order to reduce the risk and severity of fires.'Fire breaks' refer to narrower strips (normally < 100 m wide) that were burnt, hoed or brush cut (i.e.virtually devoid of fuel) along the boundaries of the plantations to prevent the spread of fire into the plantations, or to serve as access points and safety zones during fire fighting.The fire protection measures proposed or applied sometimes varied in different areas, accounting for ostensibly conflicting measures listed here.Burning frequency, season and intensity indicated are those deemed appropriate for prescribed burning operations.

Historical legacy
Past neglect of fire and alien plant management in large parts of the GRCMs has left current conservation agencies financially incapable of correcting the situation given normal operational funding, and thus unable to fulfil their primary mandate of conserving biodiversity.Previous joint government ownership and management of plantations and surrounding catchment fynbos fostered the expectation that contemporary fynbos managers should provide protection from fire to adjacent commercial plantations -at a considerable cost to the former.This political and institutional legacy yielded a plantation industry which is commercially and environmentally unsustainable 44,56 in the absence of subsidisation in the form of fire safety management and alien plant control on adjacent land.Even with government subsidisation (prior to the 1990s), and without taking on the costly burden of catchment management (during the period of neglect of the 'no-man's land'), SAFCOL plantations in the Western Cape Province operated at a financial loss. 44This financial situation is why a buyer could initially not be found at the time of privatisation 56 and why the decision was taken to phase out plantation forestry in the area. 40,41,44,54stainable management at landscape scale Sustainable management of the current conservationplantation matrix will not be achievable in the study area if the continual invasion of the surrounding landscape by self-sown timber species is not controlled.53 The current environmental and social certification system (the Forestry Stewardship Council, FSC), adhered to by the plantation industry 67 inadequately appraises environmental accountability (i.e.spread of invasive trees and their impact on water resources) beyond the borders of the plantation management unit.Indeed, in 2004, environmental NGOs requested that a moratorium be placed on further certification of plantations worldwide, which led to a review of standards for FSC certification of plantations.41,68 The development of new standards and the implementation of trial audits have not resolved this issue, which remains contentious amongst FSC stakeholder groups.Whether plantation forestry can be undertaken sustainably in South Africa and whether it should be certified remains an open question.67 Biological control, which may be seen as the only viable option for the control of vast infestations of invasive pines in remote catchments, continues to be opposed by the plantation industry.42,69,70 Legislation pertaining to the control of alien invasive plants is not enforced, whilst a discrepancy between the capacities to enforce legislation pertaining to fire risk management versus invasive plant control 70 intensifies the conflict between the conservation and plantation sectors.This conflict is evident from substantial legal claims instituted against conservation authorities for fire damage to plantations in recent years (Table 2). Fially, the sustainability of both sectors is additionally compromised by government's decision to partially reverse the plantation decommissioning programme.58 The outcome will be a more fragmented landscape where neither plantations nor protected areas can be suitably consolidated or coherently managed in terms of fire, invasive plants or general operations 44 (Stehle T 2010, personal communication, November 25).

Research priorities
The challenges associated with managing the new national park are substantial, and knowledge and solutions are not always available, which indicates the need for further research.Firstly, understanding of the historical fire regime in the area and how it has changed during the past century 71 has to be improved through the creation and analysis of a database of historical fire records. 72Ongoing accurate mapping of future fires is furthermore necessary to serve as a basis for the design of natural experiments and fire management decisions. 28Secondly, the ecological requirements of eastern coastal CFK fynbos in terms of fire season and minimum fire return intervals need to be determined.To this end, post-fire recruitment success and youth periods of slow-maturing reseeding species (e.g. the Proteaceae 16,18 ) should be studied.Similarly, the youth periods of invasive pine species need to be established under local conditions.In combination, these should inform fire management guidelines aimed at facilitating fynbos conservation and invasive pine control.It should furthermore inform thresholds 73

Fire management
Land owners and managers in the area need to realise that fynbos and plantations cannot be fire-proofed. 71It has been shown for fire-prone shrublands and forests across the globe that large fires are not dependent on a build-up of fuel, and therefore frequent burning to reduce fuel loads will not necessarily reduce the risk of runaway wildfires. 27,74,75,76he combination of fire danger restrictions and financial constraints furthermore makes large-scale prescribed burning of catchment vegetation unattainable. 27The most effective strategy for facilitating fire safety where necessary is to focus effort on (limited) strategic locations.A legal review should furthermore clarify the respective responsibilities of the conservation and forestry sectors under current fire legislation.Official agreements between neighbouring land managing agencies should clearly stipulate fire management protocols, and should be formalised within regional Fire Protection Agencies.

Invasive plant control
The timber industry should recognise its legal responsibility to control the spread of invasive alien trees from their plantations in terms of the Conservation of Agricultural Resources Act. 42The South African government proposed the introduction of a seed pollution levy on the forestry sector in 2005 under the National Environmental Management Act: Biodiversity Act, which has been vehemently opposed by the industry on account of its potential economic impact and because it is perceived as discriminatory. 77Plantation forestry in the Cape has in the past externalised its environmental costs, and continues to do so by not taking sufficient responsibility for invasive plant clearing beyond plantation borders and by leaving the task to the neighbouring conservation agencies. 44As the intention of government is now clearly to separate the functions of commercial forestry and conservation, the former has an obligation to the latter to offset these previously externalised costs, as government has been cautioned to 'not end up subsidising a single business' 44 .Nonetheless, a 'polluter pays' policy remains to be implemented, possibly through the imposition of a levy on timber products. 42Likewise, the FSC and other certification bodies should consider adopting stricter criteria for the mitigation of the negative effects of invasive trees beyond the borders of forestry estates. 42,67The best approach to management of invasive alien plants would be to integrate various control methods. 35Options include (1) manual clearing; (2) manipulation of disturbance regimes, i.e. fire; (3) future planting of less invasive species and/or sterile varieties; and (4) allowing the release of host-specific, seedattacking biological control organisms for invasive Pinus species. 42Biological control for pines has been considered, 78 but, in a highly conservative move, has been abandoned for fears that it could potentially be detrimental to the forest industry. 69

Landscape rationalisation
The invasive pine problem in the GRCMs is an example of 'invasion debt' 79 where the current pattern of invasion primarily reflects the historical socio-economic and political legacy.Neglect of the 'no-man's land' during the past 20 years has further aggravated the invasive plant problem.Government appropriately stated in its White Paper on sustainable forest development in South Africa 52 that 'the costs and benefits of this [the plantation] industry in terms of water resources and the environment in general need to be properly evaluated' and furthermore that 'Government believes that a responsible attitude in forestry would have plantation forests removed from areas where demonstrable environmental damage has been done'.These statements imply three distinct undertakings by the state: (1) application of sound economic and environmental standards to the commercial forestry industry (DoF is ultimately responsible for overseeing and regulating the industry 25,52 ); (2) appropriate allocation of resources for the rehabilitation of decommissioned plantations and neighbouring areas damaged by invader plants historically sourced from plantations 38 ; and (3) a review of the decision to reverse the plantation decommissioning strategy. 44,58A balanced analysis of the full public costs and benefits of the plantations, and their economic and ecological impacts and future risks, have to be considered at landscape scale. 54,80Land use fragmentation needs rationalisation in the interests of both the plantation and conservation sectors.This was the original intention of DoF with their plantation decommissioning strategy.In this strategy, DoF emphasised the link between plantation withdrawal and the transfer of other conservation land to SANParks, and required that these processes be well coordinated. 57

Conclusion
Owing to successive institutional disruptions in the study area, the collation of fire management related policies, practices and data has been challenging.Our account constitutes a first qualitative regional history of fire management in the CFK published in the primary literature.Such published accounts will become invaluable where there is a reliance on tacit knowledge, and particularly where institutional memory is rapidly fading.Fire management in the fynbos catchments of the new GRNP presents considerable challenges that cannot be overcome without addressing the invasive alien plant problem in the area.In the short and medium term, substantial resources will be required to correct the situation left by decades of management neglect.Longer-term sustainability of the region necessitates rationalisation of the currently highly fragmented land use with their conflicting requirements.

b
Plantations of alien pine trees (Pinus pinaster Aiton and Pinus radiata D.Don) have fragmented and replaced large tracts of fynbos, mostly along the lower and mid-mountain slopes (Figure1).The plantations of Bergplaas, Karatara and Buffelsnek in the Outeniquas, and Kromrivier in the Tsitsikammas, reach deeper into the upper catchments.

FIGURE 1 :
FIGURE 1: Map of the study area showing the distribution of fynbos, indigenous forest and decommissioning plantations (to be clearfelled and rehabilitated) within the boundaries of the Garden Route National Park (GRNP).The shown distribution of remaining plantations, on state land and managed by Mountain to Ocean Forestry, follows the recommendations for the partial reversal of the plantation decommissioning strategy.The map is split into (a) the Outeniqua region and (b) the Tsitsikamma region.

FIGURE 2 :
FIGURE 2: A landscape mosaic characteristic of the Tsitsikamma region, with dairy pastures in the foreground, a pine plantation in the middle ground, and mountain fynbos invaded by pine trees in the background.The lucrative dairy industry and coastal towns in the region are dependent on water emanating from the pine covered catchments.

TABLE 1 :
Extent of Fynbos and Forest Biomes within the Garden Route National Park and the extent of former state-owned plantations, now managed by Mountain to Ocean Forestry.
Source: Photograph taken by Tineke Kraaij

TABLE 2 :
Chronological account of the approach to fire management in the fynbos of the Garden Route coastal mountains (GRCMs), based on interviews with former catchment managers and a review of catchment management policies of the national Department of Forestry (DoF).

TABLE 2 (
Continues...): Chronological account of the approach to fire management in the fynbos of the Garden Route coastal mountains (GRCMs), based on interviews with former catchment managers and a review of catchment management policies of the national Department of Forestry (DoF).