South Africa: 1998 A Year of Destructive Wildfires in South Africa (IFFN No. 20 – March 1999)
1998:
A Year of Destructive Wildfires in South Africa
(IFFN No. 20 – March 1999,p. 73-78)
1. Introduction
This year more wildfires occurred in South Africa than in decades, in all corners of the country. At least fourteen people lost their lives directly as a result of the fires,while more than twenty people lost their lives in road accidents caused by the smoke from grassland fires. Hundreds of thousands of hectare of agricultural land, nature reserves and timber plantations were lost in the process, while numerous houses, farms and even factories were destroyed by the runaway fires experienced.
Unfortunately no suitable records were kept of all the wildfires that occurred during 1998, but a summary of the major wildfires, their causes, damage created and impact, will be provided in this report. As a result of the wide contrasts in climate, altitude, vegetation cover and other fuel and landuse characteristics between the different regions in South Africa, the information about the wildfires experienced will be discussed by biome.
2. The Fynbos Biome
2.1 Biome characteristics
This biome comprises evergreen sclerophyllous heathlands and shrublands. Trees and evergreen succulent shrubs are rare, while grasses form an insignificant part of the biomass (Huntley, 1984). Fynbos normally takes about 4 – 6 years before sufficient biomass has developed to carry a fire, and optimum burning rotations range from 12 to 18 years.
2.2 Locality
Fynbos covers approx. 5.3% of South Africa, and is almost exclusively occupying the coastal areas of the Western Cape and Eastern Cape Provinces, as far as (just) east of Port Elizabeth. Most of it occurs along the southern mountain ranges, but coastal communities are also present, in an area where rainfall ranges from 200 to 3000 mm per year.
Bordering the natural fynbos areas we also find some industrial pine plantations (particularly on the S foothills and plateau of the Outeniqua and Tsitsikamma mountain ranges). Although their total area is restricted to approx. 110 000 ha, they contribute significantly to the local industrial output of the Cape regions.
2.3 Wildfires experienced and causes
Dry Bergwind conditions contributed mostly to the extremely hazardous conditions experienced, and during two such periods – not lasting longer than a few days – most of the destructive fires occurred. During both periods people had to be evacuated from certain communities, and the lives of six people was lost during one of the Tsitsikamme fires that raged right through the Tsitsikamma mountain range, through about 4000 ha of plantations, and in the process destroying a whole rural settlement.
The extremely dry conditions and absence of rainfall during the period preceeding the wildfires contributed significantly to the fires, which were mainly caused by human beings, by allowing prescribed burns to become uncontrolled, and through arson. However, during the last spell of fires experienced in the Tsitsikamma during August, dry thunderstorms caused most of the fires in the mountains.
A steady decrease in prescribed burning application in the mountains also had created areas with abnormal fuel loading in some of the mountain fynbos areas, which contributed significantly to the high fire intensities (and even fire storms) experienced. In the plantations a lack of weed control and proper fuel management indirectly left the stands affected with no chance of survival.
Fig.1. and 2. Prescribed burning operations in Fynbos shrubland (upper) and pine plantation (lower) in the Western Cape Province. Photos: J.G.Goldammer
2.4 Recommendations
Burned areas in the mountains should be mapped, together with the mapping of remaining areas, by average vegetation age. Special care should be taken for the urban interface of these areas, and to identify “blowup situations” well in time. There should be concerted effords by the provincial authorities for a more intensive fuel management programme by means of more regular prescribed burning. The “natural fire occurrence policy” can still be applied in areas suitable for this purpose provided they are recognized as “least disturbed” by man, but elsewhere the creation of old fynbos land, with abnormal fuel loadings, should be avoided by means of a proper fuel management programme.
Timber growers should now change their fire protection plans to a more flexible system, starting by mapping burned areas, and rating fire hazard by plantation unit on a compartment level. Thereafter proper prescribed burning programmes should be put in place as part of an effective fuel management and weed control regime on a regional scale, particularly in the Tsitsikamma (de Ronde, 1990).
3. Grassland Biome: Higher Rainfall Area
3.1 Biome characteristics
Although 24.1% of South Africa is covered by this biome (Huntley, op cit.), less than 10% of this area falls in the higher than 800 mm per year rainfall margin. These grasslands are dynamic, with most of the biomass being “cured” during the dry winter season. Most of these grasslands can burn after one winter (curing) season, providing dangerous fuels if fire is not allowed in controlled form within 2 – 3 years if left unburned. However, the rate of biomass production is also significantly influenced by seasonal parameters, particularly rainfall occurrence.
3.2 Locality
Mostly situated in the Drakensberg mountain area, from the NE Cape (southern slopes of the Drakensberg, S of Lesotho) to the Drakensberg escarpment in the N, in the Mpumalanga Province. Most of the area is used for agricultural purposes (particularly grazing) but industrial timber plantations are also found there in abundance, covering more than 1 million hectare.
3.3 Wildfires experienced and causes
Although the area covered by this biome experienced more fires than usual, the occurrence did not exceed normal patterns as much as in the fynbos and grassland falling within the lower rainfall margins (see next paragraphs). However, the summer rainfall stopped very early (during April already, compared to during June/July during normal seasons) and this caused early grassland curing, and a subsequent early start of the fire season.
In the absence of a proper fuel moisture and grass curing monitoring system coupled to Fire Danger Index, foresters, farmers and nature conservators – used to burning each year during June and July – were not aware of the dangerous situation that had developed, and as a result most wildfires in this area originated from controlled prescribed burning.
The most serious case of mortality occurred during an arson-caused grassland fire in the Mpumalanga Highveld area, near Badplaas, when four people were killed when attempting to drive through a grassland fire front in a van without a protective canopy . Some serious plantation fires were also experienced in the NE Cape, in Mpumalanga, and in Swaziland, and in the worst cases more than 1000 ha of plantation was lost at a time.
3.4 Recommendations
Fortunately selective use of prescribed burning is mostly used by farmers and foresters, but the sequence of burning needs improvement to arrive at a regional, integrated burning plan with pre-determined priorities, concentrating on regional buffer zoning (de Ronde, 1997). The burning season should also be more flexible, but (most important) staring and ending dates should be determined scientifically after a monitoring programme, and determined in such a way that it can give foresters and farmers early earning when unusual seasonal changes are recorded that can affect the safety of fire application.
Upgrading of the existing fire danger rating system – to include regional characteristics, land use, grass curing patterns, fuel build up and moisture – is recommended to privide safer prescribed burning conditions, and subsequent less problems with situations where controlled prescribed fires turn uncontrolled (de Ronde, 1998).
4. Grassland Biome: Marginal Rainfall Area
4.1 Biome characteristics
The most important characteristic of this part of the grassland biome, is that it falls in the marginal (600 – 800 mm per year) rainfall area. The grassland is as dynamic as in the higher rainfall area, but produces on average less biomass per year, and thus the optimum fire need will be approximately 5 – 6 years. However, this can range within the region from 2 – 8 years depending on soil type, locality in the landscape and microclimate. Most important is that seasonal rainfall can fluctuate more than in the higher rainfall regions, and thus contribute significantly variation in yearly biomass production. This might be as low as 500 kg/ha/year during dry seasons, and as much as 2500 kg/ha/year during seasons with above-average rainfall. This aspect in particular determines the extend of fire hazard during a specific fire season. Most of this grassland is utilized for agricultural purposes, and where grassland is maintained, for it is mostly used for grazing purposes. A relatively small area within this biome is utilized for nature conservation purposes, particularly in the Eastern Free State.
4.2 Locality
Situated mostly in the SW of the Mpumalanga Highveld, the E Free State and Gauteng Province, this biome forms a very important part of the agricultural sector, with vast areas covered by the grassland, although covering less than 3% of South Africa. Prescribed burning is here less commonly used by farmers compared to the frequent use of fire in the higher rainfall area, but the use of fire is not uncommon in farming communities.
4.3 Wildfires experienced and causes
The biome has experienced some very serious wildfires this season, mostly caused by arson. In the Vrede area (SE Mpumalanga) and near Warden (E Free State) 30 000 – 40 000 ha of farmland were wiped out by single fires, also burning down farm homesteads and equipment worth millions of Dollars. In Gauteng Province, one Nature Reserve was burned to the extend that animals had to be transported to other reserves to save them, while the most serious road accidents, caused this year by smoke from grassland fires, were recorded in Gauteng, killing at least 20 people indirectly. Most of the latter were caused by smoke from prescribed as well as from uncontrolled burns.
One of the main indirect causes of the wildfires experienced in this region was that farmers were unaware that the last 3 – 4 years of above-average rainfall had caused an very high grass biomass production pattern to develop, and no steps were taken before the 1998 winter to counteract this problem by means of increased prescribed burning programms or the creation of proper fire belt systems and buffer zones. In Gauteng – as elsewhere – the traffic authorities were also not taking the necessary action to safeguard roads exposed to fire and smoke from grassland fires. Where sections of public roads are exposed to serious smoke from fires, these must be closed for traffic immediately, and should remain closed until the smoke has lifted.
4.4 Recommendations
Region-specific grassland fuel buildup and grassland curing patterns should be monitored as part of an improved fire danger rating system, so that guidelines for prescribed burning intensity can be supplied well in time to the farming communities for them to take timely action. A regional approach should also be followed to fire protection and fuel management measures, so that regional buffer zones can be constructed along strategic lines, and to move away from the individual prescribed burning measures at present being applied.
Traffic authorities should also be advised of any wildfires occurring in particularly areas, and well in advance be told about any prescribed burn that might cause problems along public road sections.
5. Grassland Biome: Lower Rainfall Area
5.1 Biome characteristics
Although this grassland is also dynamic, the grass is more sparsely spaced, and only slowly and systematically adds biomass above ground level. Typical yearly biomass addition is 400 – 600 kg/ha/year, and it takes mostly approx. 12 years before such a community can carry a fire, and then only on steep slopes, and/or if some dominant prevailing wind can move the fire front. The low rainfall is the main reason for this (400 – 600 mm per year average). However, good rainfall years can increase biomass dramatically, creating a fire hazard in regions where for years there was none. Most of this grassland is utilized by farmers for low intensity grazing.
5.2 Locality
This grassland biome covers most of the central Free State, as well as the North West Province. It is estimated that approx. 10% of South Africa is covered by this grassland type, but a high percentage of this has been taken up by cultivated land, particularly in the NW Province.
5.3 Wildfires experienced and causes
Wildfires have been relatively unknown in these areas, or are normally restricted to small controllable fires. However, for quite a few years above-average rainfall was experienced, and the higher-than-normal biomass production had created a situation where a wind-driven fire could run completely out of control. Quite a few fires of this nature were recorded in the Free State area, although none of them caused serious damage to men and property. However, one such fire in the NW Province burned down a range of farms, damaging homesteads, while in another one (near Schweizer Reynecke) two policemen were killed. Most of the fires occurred during strong windy conditions, and almost all of them were caused by arson.
5.4 Recommendations
Although no daily fire danger rating forecast is necessary for this area, some form of biomass addition monitoring is recommended to warn farming communities when a fire hazard situation is reached, with a strategic fire protection plan for the region, that can be used prior to such a season. Simple rainfall pattern monitoring and quarterly grassland biomass observations can then provide a fuel build up index that can set fire protection measures in motion in time.
6. Moist Savanna Biome
Covering about 10% of South Africa, this biome is mostly situated on the plateau, 500 – 1000 m above sea level. The vegetation cover is a combination of woodland and grassland, and some industrial timber plantations are also found there (approx. 300 000 ha). A high percentage of the area is utilised as nature reserves or game farming, and also other for agricultural purposes.
Although some wildfires were experienced in the Mpumalanga Lowveld and Zululand, there was no deviation from normal patterns in this area.
7. Conclusions
Although weather conditions in the Fynbos and Grassland regions were drier than normal, with a longer period of rain absence, it was in most cases the lack of proper fuel management and integrated fire protection that contributed significantly to the seriousness of the 1998 wildfires. There is also a clear indication that the existing fire danger rating system should be adjusted to include fuel moisture and fuel buildup indexes, as well as a grassland curing monitoring system, to provide specific regions with early fire hazard warning provision over and above the daily, weather-based, fire danger rating used at present.
A regional early warning system should also be backed-up by a more appropriate fuel management and regional fire protection system, that is created well in advance of the fire season. In other areas, such as in higher rainfall areas of the Highveld, fire protection measures should be better co-ordinated on a regional scale, to move away from the existing ad hoc individual measures presently being applied.
If the above recommendations are applied at a national level, a repeat of the scale of damage to property and loss of life experienced during 1998, can be avoided in the future.
8. References
De Ronde, C., 1997. Forest fire prevention in industrial plantations. Proc. IX Silvotecna Conf. Concepcion, Chile, Nov 1997: 15 pp.
De Ronde, C., 1998. Proposals: Revision of fire danger class criteria for industrial plantations, nature reserves and the agricultural sector in South Africa. Report submitted to the SA Forest Industry: 5pp.
De Ronde, C., Goldammer, J.G., Wade, D.D. and Soares, R.V., 1990. Prescribed fire in industrial plantations: In: Fire in the tropical biota, Ecological Studies 84 – J.G. Goldammer ed.: 216 – 265.
Huntley, B.J., 1984. Characteristics of South African Biomes. In: Ecological effects of fire in South African Ecosystems: Ecological Studies 48 – P de V. Booysen and N.M. Tainton ed.: 1 – 18.
C. Neels De Ronde
SILVA Forest Services du Toit Street 16
P.O.Box 612
7280 Bredasdorp
SOUTH AFRICA Fax/Tel: ++27-2841-52546
e-mail: nderonde@dorea.co.za
