Steve Pope Safaris

Semi-participatory Camping, Bushwalking, & Canoe Safaris


 Report by Professor Kirkman after four day visit to Zambezi Valley 

There are two levels of decisions in the decision making process.

  1. A decision on whether fire is a part of the ecosystem or not is an extremely important decision and should be based on all available evidence with particular reference to historical records as well as relating the woody vegetation to other similar areas known to be fire driven or non-fire driven systems.
  2. If fire is considered to be part of the system, then the next level in the decision making process is the level of intervention:
    1. No intervention will eventually result in system dominated by woody species that will largely exclude fire because of low herbaceous fuel loads. Thresholds may be crossed in which the dominant vegetation type in certain areas may change from one which can carry a fire (e.g. grasslands) to one which would not (except in exceptional circumstances) carry a fire (e.g. closed woodland with no or little fuel at ground level). This is likely to result in substantial changes to the vegetation structure and composition in affected areas.
    2. Deciding to intervene necessitates additional levels of decision making that will not be considered here at this stage. 

It must be emphasised that non-intervention is the result of a decision, and not merely the default option. 

Risk analysis

The two levels of decision considered here both carry inherent risks:

  1. The level one decision on whether fire is or was a part of the system or not can be either correct or incorrect. This decision can be made as a collective decision after considering all available evidence and consulting all available expertise. The risk of making the incorrect decision at this level is relatively small, although an incorrect decision may have serious consequences.
  2. The decision at level two on whether to intervene or not carries the same consequences as the level one decision. However, the decision on whether to intervene, and on the level of intervention is usually made in a different context with unique constraints. The risk of making an incorrect decision is relatively large, the consequences similar to level one and the additional risks when implementing the decision are large.
    1. It must be noted that the ecological consequences of non-intervention may be as great as the consequences of inappropriate intervention.
    2. The ecological consequences of inappropriate intervention need to be understood. 

Intervention decision (reasons for burning)

The reasons for burning need to be considered in some detail because both the decision to burn or not, and the implementation of a decision to burn both carry substantial risks. The reasons for burning include the following:

  1. Acceptance that fire is part of the system
  2. Realisation that fire is necessary for specific reasons
    1. Bush encroachment (bush dominating in areas previously grassland)
    2. Bush thickening (increase in bush density to the extent that grass and other vegetation layers decrease/disappear
    3. Decrease in grassland area
    4. Decrease in grass productivity
    5. Increase in forb/shrub layer (usually precursor to bush encroachment/thickening)
    6. Changes in numbers of herbivores in particular classes (e.g. grazers, browsers) or ratios between classes
    7. Changes in predator numbers as a result of changes in herbivore numbers 

Should a decision to apply a fire management program be made, the following points should be considered:

Fire intensity

Fire intensity can vary tremendously depending on conditions, and can be manipulated. The impact of fire intensity on vegetation can be significant.

  1. Low intensity
    1. Low fuel load (grass), high humidity, low temperature
    2. Low impact on trees
    3. May impact shrubs
    4. Patchy burn pattern
  2. High intensity
    1. High fuel load (grass), low humidity, high temperature
    2. May kill small trees (up to 2 m)
    3. Will impact shrubs
    4. More uniform burn pattern
    5. Greater risk of spreading beyond intended areas 

Fire frequency

The frequency of fire can be controlled to a large degree. Decisions on fire frequency should be made after considering all the relevant factors, including the following:

  1. Accept that some vegetation types are not fire driven, including:
    1. Sodic areas
    2. Riverine areas
    3. Thicket
    4. Mopane woodlands (maybe)
    5. Jesse bush
  2. Accept that fire plays a role in some vegetation types, including:
    1. Grasslands
    2. Open savanna
    3. Grassland to savanna transitions
    4. Shrub dominated grasslands
  3. Accept that fire frequency and intensity should vary spatially. 

Most of the areas visited where fire could be considered were dominated by shrubs/forbs had a low grass fuel load. Fire will be low intensity and patchy. In many of the shrubby areas the first fire may not kill the shrubs, but should give the grass a competitive advantage. Subsequent fires (two or three years later) may kill a significant proportion of the shrubs. 

Fire frequency should always be flexible and based on a specific decision with defensible reasons as opposed to a regular program. Decisions can often be based on fuel load. 

Size of area to be burned

Area burned should be related to number of animals expected to congregate on burned area after the fire. High numbers of animals attracted to a small burned site may impact the vegetation through grazing and trampling. 

Ultimate objective of burning program

Create a mosaic of grassland – shrubland – open savanna – closed savanna – woodland (forest).

Frequent fire                                                                                     No fire


  Grassland    Shrubland    Open savanna    Closed savanna    Woodland


Spatial diversity

  • Mosaic

    • Spatial diversity
      • Species (within and between structural layers)
      • Structure (grass layer, shrub layer, tree layer)
    • Vegetation diversity
    • Bird diversity
    • Mammal diversity
    • Invertebrate diversity 


    Impressions of the state of the vegetation in the Mana Pools (specifically Chitake Springs area) after a short, first visit should be considered within the following constraints:

  1. No previous knowledge of the state of the vegetation or its management can be both advantageous and disadvantageous
    1. Advantages – fresh opinion based largely on comparisons with other areas and management in other areas
    2. Disadvantages – no knowledge of previous history of land use and management and no experience in the specific area.
  2. Coverage of the area was limited due to time constraints.
  3. It was not possible to visit other neighbouring areas that may be managed differently for comparison purposes. 

Chitake Springs

The vegetation comprised a mosaic of relatively thick, closed canopy savanna woodland, through open savanna and shrubby areas to grassland. The grassland areas appeared to be quite heavily utilised by grazing animals. Dominant grasses included Digitaria species and a range of annual grasses. In terms of the classical succession models of grassland function, there was no evidence of under utilisation with succession through to Increaser I type species that would be considered a precursor to bush encroachment or thickening. In contrast, the impression is one of heavy utilisation with a tendency to be dominated by annuals. Many of these annuals are probably largely driven by rainfall. In wet years they will probably increase in abundance and in dry years they will probably decline. Fire would probably not be considered necessary in such a system if the shrub and woody layer were ignored. 

However, there are a number of shrub species that dominate the grass layer in many areas. The shrubs appear able to compete effectively with the grasses for nutrients and light, dominate and act as a precursor to woody species. This is probably exacerbated by the relatively heavy utilisation of the grass layer. There was not much evidence of the shrubs being utilised by any herbivores. In areas utilised by buffalo, particularly corridors, shrub abundance declines dramatically due to trampling. Controlled burning (correct intensity and frequency and suitable area in relation to animal numbers) may play a role in allowing the grass layer to compete more effectively with the shrubs and maintain or increase the area of grassland relative to shrub and woody dominated landscapes. Incorrect use of fire may result in increased grazing pressure on the grass layer in burnt areas and may exacerbate the situation. 

Mana Pools (floodplain)

The presence of the exclosure plot allows some assessment of the impact of animals on the vegetation. Apparently the plot was fenced in 1986. A survey was carried out inside and outside the plot in 1994. The results indicated that there were differences between the plot and the surrounding areas. Ten years later, there is a substantial visual difference. It would be very useful to have the plots resurveyed. 

There is evidence of heavy utilisation of both the browse and grass layer. The appearance and dominance of the Indigophera sp in recent years is striking. The role of the Indigophera on an ageing floodplain is an interesting topic for debate. It may be that it is an integral part of the succession process initiated by the northward river migration, or that it is an opportunist plant which has appeared in response to a stressed system. Either way, it is there, and it is part of the system. It would be useful to examine similar areas up- and downstream with differing levels of herbivory in an attempt to better understand the origin and role of the Indigophera

Whatever the role of fire may have been or should be on the floodplain area, there is currently probably not enough of a fuel load to carry a fire, particularly in the areas dominated by Indigophera. With the obvious herbivore pressure in the area, this is not likely to change in the short term. 


The decision making process surrounding the role of fire in the Zambezi Valley is critically important and will determine whether controlled burning should and will take place or not. If a decision is made to implement a burning management program, then further decisions have to be made regarding area, intensity and frequency of fire. The final step is the implementation of the program with due consideration to safety.


Season of Burning 

Research in southern Africa has clearly indicated that that the least damage is caused to the grass sward if prescribed burning is applied when the grass is dormant. Therefore it is recommended that when burning to remove moribund and/or unacceptable grass material burning should preferably be applied after the first rains of >13mm at the commencement of the growing season i.e. when the grass is still dormant and the fire hazard is low. Conversely when burning to control encroaching plants burning should be applied before the first rains initiating the commencement of the growing season i.e. when  the grass is very dry and dormant to ensure a high intensity fire. The actual time of the year when prescribed burning will be applied in Africa will depend upon the latitude and rainfall pattern of the region. For example in the central highlands of Kenya which receives a bi-modal rainfall the main burning windows are the dry period between May and September  and a shorter period  during January and February (Trollope & Trollope, 1999). Conversely in the summer rainfall areas of southern Africa the recommended season of burning is approximately before and immediately after the first spring rains in September/ October (Trollope, 1999). 

Excerpts from Oliver West’s booklet “Fire, Man and Wildlife as the interacting factors limiting the development of climax vegetation in Rhodesia”     Published in 1971

 1) Grassland is the ultimate product of fire because it is composed of the plants, grasses and forbs most tolerant of fire. 

2) Fire, which usually originates in grassland, destroys forest, produces savanna and ultimately grassland. 

3)    Descriptions written by Selous (1893) and other travelers --- stress the openness of the countryside and indicate that during the Iron Age occupation, a small human population, with fire and shifting cultivation, had established a trend which favored the development of open savanna and grassland while it discouraged dense bush and forest. 

 Letter to the Editor of the Rhodesia Science News     Vol 3 No 11 November 1969.

Dear Sir


During the past nine years, a marked and progressive degradation of the
annual grass cover (mainly Urochioa sp), has been observed over some twenty
square miles of bushland, which is dominated by a small tree species,
(Combretum elaeagnoides), surrounding and extending Westwards from
Rekometjie Field Research Station in the Zambezi Valley. The affected area
lies within the boundaries of a controlled hunting area and a game reserve,
and domestic stock in the form of cattle, sheep and goats are involved only
in the immediate vicinity of the Research Station. No systematic assessments
of changes in the flora have been made, but the deterioration is visibly
obvious. Earlier in each successive year the cover of annual grasses has
been removed by harvester termite, to be replaced by an increasing
population of herbaceous weeds which effectively prevent hot clean grass

It has been the practice in this area to exclude fire completely. As a
sequel and possibly a consequence, the harvester termites have been provided
with a copious source of food each year in the form of annual grasses which
they have utilised until the ground has become denuded of these with only
weeds remaining, the time at which this time has been reached becoming
progressively earlier in each succeeding year. The increase in harvester
termite activity has led, it is believed, to the progressive degradation,
and the situation has now been reached where attempts to introduce fire are
quite ineffective because of the absence of burnable grass.

The exclusion of fire, followed by increased harvester termite activity has
deprived grazing animals of an important element of their dry season forage.
Unless controlled early fire can be encouraged when burnable grass is still
available, the situation may deteriorate still further.

              Yours faithfully,

                                    R. D. Pilson

Branch of Tsetse and Trypanosomiasis Control,
           Department of Veterinary Services. 


Excerpts from documents compiled by Professors Winston Trollope and Kevin Kirkman (Grassland Science, School of Applied Environmental Sciences, University of Natal) 

. It is concluded that human beings have used fire for over a million years and in Africa fire has extended the grasslands and savannas at the expense of evergreen forests. This reinforces the fundamental conclusion that fire is a general and influential ecological phenomenon throughout the world (Bond & van Wilgen, 1996) and cannot be ignored when considering the management of rangeland ecosystems for both domestic livestock and wildlife purposes. It is a factor of the environment that has been occurring since time immemorial in the savanna and grassland areas of Africa

Africa is where fire and humanity first interacted and the factor that makes fire on this continent distinctive from other regions is the antiquity of anthropogenic fire (As a result of the aforementioned factors fire is regarded as a natural ecological factor of the environment that has been occurring since time immemorial Pyne, 1995) 

 Sweet (1982) in Botswana and Boultwood & Rodel (1981) in Zimbabwe found that annual burning resulted in a significantly greater reduction in the density of bush than less frequent burning. 

On the contrary the withdrawal of fire for extended periods of time appears to have a more predictable effect. For example on the Accra Plains in southeastern Ghana protection of moist savanna from fire for 29 years has resulted in the development of forest type vegetation with a fairly closed canopy. 

This led to the conclusion that in all the burnt savannas of Lamto the pressure of forest elements on savanna vegetation is very high and the exclusion of fire initiates the development of forest (Menaut, 1977). Similar trends have been found in the more arid savannas (500-700 mm p.a.) in southern Africa where in the Kruger National Park the exclusion of fire caused both an increase in the density and size of tree and shrub species (van Wyk, 1971).

The effect of frequency of burning on the quality of forage is that generally frequent fires improve and maintain the nutritional quality of grassland particularly in high rainfall areas making it highly attractive to grazing animals. 

 West (1965) stated that the fresh green shoots of new growth on burnt grassland are very high in protein and quotes Plowes (1957) who found that the average crude protein content of 20 grasses after burning at the Matopos Research Station in Zimbabwe was 19%. This is approximately twice the protein content of mature grasses that have not been burnt at the end of the dry season.

. At the turn of the century the Serengeti-Mara area was described as open grassland with lightly wooded patches, much as it is today. Following the great rinderpest epidemic of 1890, human and animal populations are thought to have been reduced to negligible numbers in the Serengeti-Mara region. Fires were infrequent due to the low human populations and elephant numbers were low having suffered from heavy ivory poaching during the previous decades. Over the next 30 to 50 years these prevailing conditions of low fire frequencies and low elephant numbers saw the establishment of dense woodlands and thickets. This woody vegetation provided ideal habitat for infestations of the tsetse fly that further prevented any significant human settlement within the Serengeti-Mara ecosystem. Consequently by the early 1940’s the area had become densely wooded and the Serengeti National Park and the Masai Mara National Reserve were characterised by dense woody vegetation and remained in this condition for over 20 years. These woodlands began to decline significantly during the late 1950’s and early 1960’s in response to two factors. Firstly there was a marked increase in the frequency of burning as a result of the dramatic increase in the human population that was recovering from the secondary effects of the rinderpest epidemic. In addition the low ungulate populations were unable to reduce the standing crop of grass particularly with the unusually high rainfall experienced in the late 1950’s and early 1960’s. This resulted in increased burning by Masai pastoralists to create better grazing for their livestock, tribesman used it to facilitate hunting, some fires were inadvertently ignited by honey hunters and European hunters used fire to attract game. The resultant high intensity fires proved devastating and helped to clear the area of bush and attract ungulates to the highly nutritious grazing created by the fires. The other effect of the increased human population was to compress elephant populations into the protected conservation areas. The Serengeti-Mara woodlands declined accordingly in response to increased utilization by elephants similar to what had occurred in other parts of Africa when elephant numbers increased to very high densities. The decline in the area of woodland in the Mara was greatest during the period 1961 to 1967 but continued into the 1980’s. Subsequently the situation changed in the Serengeti where the elephant population declined by 81 % (2460 vs 467) between 1970 and 1986 as a result of poaching and resulted in the recovery of the woodland vegetation. Conversely the elephant numbers continued to increase in the Mara due to immigration and natural population growth. These changes in elephant densities have also been accompanied by a steady increase in the wildebeest population that has risen from 250000 in the 1960’s to its current level of 1.5 million. The situation at present is that woodland vegetation is increasing in the Serengeti in response to low elephant densities and reduced frequencies and intensities of fires. Conversely in the Mara the woodlands continue to decline and the Themeda dominant grasslands are being maintained by the large number of resident elephants despite a decline in the frequency and intensity of fires.  

Finally it was shown that frequent fires favour the abundance of the highly productive and palatable grass species, Themeda triandra in southern African grasslands (Scott, 1970; Dillon, 1980; Forbes & Trollope, 1990). ) This raises the possibility of using fire to improve range condition by increasing the abundance of valuable forage species like Themeda triandra.  


“Veld and Pasture Management in South Africa” N.M. Tainton 1981 

On page 21 - 1.4.4 it says that there is a general belief that fire is harmful to vegetation and upsets the natural ecological balance and should be prevented.  Burning is often condemned but many ecologists believe that it is natural and that no fire causes profound negative changes to the vegetation. It is an accepted fact that fire maintains grass cover and prevents undesirable thickets. Foresters and amateur ecologists and preservationists are the people who automatically condemn burning without scientific backing, playing on the emotions of the public. Bush fires have been part of the African ecology for thousands if not millions of years. According to Sim (1907) the Portuguese called the interior “Terra de Fume” because of the smoke from fires. Roux (1966) wrote that burning in winter was common long before European settlement.

The fact that many plants have evolved to withstand fire indicates that fire has  influenced the grassland, woodland and savanna for millennia. Many flowers only appear after fire and disappear if there is no fire. Many species of tree resist fire when mature even though they are susceptible as seedlings whereas species in the Evergreen Forests succumb to fire even when they are mature. Most grasses have evolved so that if they are not burnt they become moribund and die.  Many birds and other animals are therefore regulated by fire according to the environmental changes therefore fire is a natural ecological factor and as a management tool would be natural and without it would lead to undesirable changes in the vegetation and animals. 

On page 374 17.2 it says that fire is used to prevent bush encroachment which reduces the amount of grass (du Toit 1972; Trollope 1977). 


Global International Geosphere-Biosphere Programme (IGBP) Change Report No 31 – 2 – 5th June 1993, Victoria Falls Zimbabwe.


1)    Soot particles reduce global warming (p 8)

2)     Fire is a source of tropospheric ozone (p 8)

NB: there is only 10% of the original North American Prairies left due to fire prevention during the last 200 years (National Geographic, vol.184,no4 October 1993) – that means there is 90% less grass fire smoke in the atmosphere – therefore there is 90% less soot particles to contribute to global cooling and chemicals contributing to ozone – has this contributed to global warming and do anti fire policies compound the problem? 

3)    Tool for bush control (p 8)

4)     Humans have burnt savannas for thousands of years (p 8)

5)    African savannas have highest biomass and biodiversity of large mammal herbivores on earth (p 26)

6)    Exclusion of fire leads to woody component of savanna (p27) 

Early burning in Brachystegia – Fay Robertson 

1)    Restoration of woodland by early burning (p68)

2)    Termite population low because frequent fires destroy food supply (p 69)

3)    Woody stems vulnerable to fire in late dry season when food reserves are being transferred to new leaves (69)

4)    Hot fire can prevent the recovery of shrub land into woodland (p 69)

5)    After 10 to 15 years of protection from fire new saplings are sufficient to suppress grass. (p 70)

6)    Low elephant population and early burning encourage woodland over 40 years – (was the Kariba cull in 1984

      to establish woodland?- S. Pope)

7)    Introduction – 1) In 15 years on burnt plot the trees declined by only 28 % but on unburnt plots trees increased

      by 87 %  (p 8)

                            2) Burning increases nutritional value of grass (p 14) 


Eleven Years in Central South Africa – Thomas Morgan Thomas, Published 1873 

Page 85 – In the spring and summer, vegetation grows so luxuriantly and rapidly that the thousands of cattle cannot keep one tenth of even the sweet part of it under, and they have soon to be closely followed and kept in sight, that they may not be lost. In the autumn, these pasturages resemble immense fields of ripe corn; and in winter, when seedless and dry, they are swept away, one after the other, by conflagrations that are always raging in some part or other of the country. During the day, the smoke is seen at a great distance, ascending cloudlike towards the skies; and in the night, the flare is reflected brilliantly upon the ascending clouds of smoke. With this yearly clearance, by the cleansing element, comparatively few of the trees are destroyed; but, while the larger timber escape with little injury the shrubs and underwood are destroyed, and all dry plants are reduced, in some localities, to complete ashes. 


Wankie the Story of a Great Game Reserve by Ted Davison – (1967) 

Edward Hartley Davison was given the task of establishing Hwange Game Reserve in 1928. He held the post of Warden for 34 years and “did more than anyone else for the preservation of our wild game” -  Ian Douglas Smith. 

“The game itself did not suffer to any great extent. More often than not animals would walk calmly ahead of the flames about a mile away. When the flames died down at night (as they usually did) the game would find gaps in the chain of fire and get back onto the burnt – out areas. 

Game of all types show a marked preference for burned off country, when the new grass grows. Localities which have not been burned for two or three years are not nearly so popular as feeding grounds as those which have been burnt and cleared of old grass. Quite apart from the grazing aspect, an area which had not been burned or heavily grazed for a number of years takes on an unhealthy appearance. Dead grass, leaves, and other types of litter accumulate, and in many cases this layer of dead material suffocates the growing grasses. 

To exclude fire altogether might have a very detrimental effect on many animals, particularly those such as gemsbuck and tssessebi who prefer the sparse but succulent early grasses after a fire; as well as buffalo, who show a marked preference for grazing veld which was burnt the year before. 


“Creatures of Habit” – Peter Apps (PhD University of Pretoria)     Doctor of Zoology

Each year in Africa south of the equator fires sweep across an average of
one point 6 million square kilometres. Nearly all of southern Africa’s
vegetation is subject to periodic or sporadic fires, and considering the
fury of a wind driven fire in grasslands, savannah woodland or fynbos, cases
of mammals being killed by fires are remarkably rare.  Small mammals simply
disappear down their burrows- the heat from even the fiercest of blazes
penetrates no more than a few centimetres into the soil.   A radio-tracked
genet survived a fire in its cape fynbos habitat by sheltering between two
boulders.   The main danger comes not from the fire itself, but from the
destruction of food supplies and cover that offers shelter from predators.
The wind that drives the fire along also carries for kilometres the smell of
smoke that warns of its approach, and larger animal species have the ability
to move out of the fires path, onto rock outcrops or across rivers.  Fires
are likely to catch them only when fences block their escape routes. 


“Sometimes When it Rains” 

Keith Meadows - Writing about Charlie Mackie – In 1984 Mr. Mackie headed the resurrection of Zaire’s Garamba National Park – “It essentially consists of sweeping savanna, laced by a snakes and ladders unravelment of rivers that make vehicle - travel complicated. Dotted here and there over the grassland are big, solitary, shade - giving trees, usually sausage trees.”  “The goal of the Garamba Rehabilitation Project was to, in time, refurbish one of the most beautiful remote game parks in Africa”.  “It is elephants and fire that shape the Garamba grasslands”. 


Factors affecting the ratio of Lions to Spotted Hyaena – Dr. Gianetta Purchase 

On Page 44 - the Liuwa Plains National Park (Zambia) - the grassland burns annually as a result of fires begun by the local people to encourage new growth for their cattle to graze. 


From: "Justine Blackbeard" <>

To: <>

Subject: Bush Encroachment

Date: Thursday, July 20, 2006 5:05 PM 

Gavin Blackbeard - Botswana 

Hi Steve, 

Many thanks for your email and information that you have collected and forwarded to me.               

I agree with your findings but unfortunately the Botswana Government have been swayed by the young ecologists from other countries and has made a law that anybody starting a fire whether by accident or intent will be liable to a jail sentence for no less than 10 years.  As a result there have been no bush fires and the bush encroachment has now become a very serious problem to the cattle ranchers and wild game in our country... 

I do hope that your research in the Zambezi Valley is conclusive and if so maybe we can use your findings to persuade the Botswana Government to allow us to do controlled burning at our cattle posts. 

Many thanks, 

Gavin Blackbeard 


Ian Nyschens referring to his time in the Zambezi Valley in the 1940s and ‘50s 

 It was not just the valley that used to burn but the savannah country as well – you could sometimes see the whole length of the escarpment on both sides of the valley glowing. These fires were lit by the tribesmen who feared that the dense cover gave lions hiding opportunity. These fires were especially noticeable along the footpaths between the villages – they were fanned by the August winds

and the air currents as you know always come up the Zambezi river. In most wilderness areas the vegetation has got to be hammered. We all know that the dam wall must have had an effect on all the flood plains but if you care to go into history you will realise even hundreds of years before our occupation that Africa was regularly put to the torch, in fact the middle Zambezi valley was known to the

Portuguese as Tierra Del Fume – land of smoke. Read “The Months of the Sun” – chapter 12 – page 157 to 160. I hope this information will be of some value to you. 

Ian Nyschens. 

 Claude Meredith who hunted and managed the Dande North Concession stated that because of continued anthropomorphic burning there is still abundant grassland and large herds of Buffalo in that region. Most fires are lit by villagers while collecting honey, clearing fields for cultivation or for protection from dangerous game. A total of 80% of the Dande North concession which is an area of 98 100 hectares is burnt annually. These fires occur in the late dry season and are therefore “hot fires” which benefit grassland by inhibiting bush encroachment. The result is that there is sufficient grassland for numerous herds of up to 600 buffalo with a total for the relatively small area of 3 374 buffalo (aerial strip counts conducted by National Parks and WWF in 2001). This compares with 3179 buffalo in Mana Pools, which is more than double the area at 218700 hectares. Bear in mind that one herd of 5000 was recorded in 1976 on the Mana River Terraces.

 Before the villagers were relocated out of the Zambezi valley, anthropological fires were a regular occurrence, (pers.comm. Ian Nychens who hunted the Zambezi valley in the 1940s and 1950s).

       At a time when Tsetse Fly Control Department’s operations are compromised by the economic situation the lack of late dry season hot fires are creating the ideal habitat for tsetse fly. In the last few years there has been a noticeable increase in the Tsetse fly population and incidents of Sleeping sickness. Dense bush and forest is the ideal habitat for tsetse fly, therefore, if the current trend of forestation continues we can expect an even greater increase in tsetse fly problems in the Zambezi valley. On November 10th 2005, Zimbabwe Association of Tour and Safari Operators (ZATSO) circulated a warning that sleeping sickness was coming back into the Zambezi valley in an email that included information from Dr. Vale on prevention, diagnosis and cure.

       It is imperative that a grassland management program be adopted in the Zambezi valley with fire and elephant pressure as tools. Many people will be concerned about the danger to wildlife by fire but Professor Kirkman says that “single point ignition” along a river or road, allowing the fire to burn down wind has minimal casualties to wildlife, even to slow moving animals, as the smoke prompts them to seek shelter. This also results in a “patch work “ of grassland, woodland and forest, which is ideal to provide the different habitats for biodiversity. This method of burning would replicate the anthropological fires that occurred for thousands of years before the villagers were moved out of the Zambezi Valley. The benefit of a much higher biomass and biodiversity in a mosaic of grassland, open woodland,  thickets and forest far outweighs the results of gradual forestation due to lack of fire.


Fire as a Grassland management tool

In the 1960’s there were references to a herd of 5000 buffalo in Mana Pools. More recently two ex National Parks officers, Dolph Sussien and Steve Edwards, have both informed me of recordings of a herd of similar size in the 1970’s. Dolph Sussien’s sighting was in 1978 and the herd stretched from the Lodges to Mana Mouth, a distance of about 3 km.         Currently a herd of 200 is considered to be a large herd. 

Animal population sizes are related to their food supply. 

During most of the 1990’s there was a herd of 1000 buffalo on the Matusadona shoreline in the region of Sanyati West Bay and Gordons Bay. This herd built up from small numbers in the early 1980’s due to the persistent low lake levels that prevailed during the drought years of the 1980’s and 1990’s. Because of the shallow gradient of the Matusadona shoreline there were permanent vast areas of Panicum repens (Buffalo Grass). The herd grazed almost permanently on the shoreline because there is very little grass in the Mopane woodland. As a result of this large herd being established the Lion also increased in number to prides of over 20 and sightings of them hunting and feeding on the Buffalo were a major contributing factor to the development of the Houseboat industry and a valuable natural resource for tourism.  The Matusadona shoreline is unique in enabling game viewing from the comfort of luxury houseboats. 

 In 1999 the lake level was allowed to rise and inundate the tree line. This occurred again in 2000 and 2001. Within 2 seasons the large herd of buffalo had all but disappeared. This was followed by a crash in the Lion population. In 2003 there was only 1 male and 2 females in the vicinity of Fothergill and Spurwing. There were even sightings of Lion hunting and feeding on Hippo and Crocodile. In fact, a Hippo killed one of the 2 females in November 2003. 

The significance of this occurrence in Matusadona is that it illustrates what is happening in the Zambezi Valley.  Bush encroachment is at an advanced stage and the end result will be forestation and the loss of ungulates dependent on grassland. There has been no significant fire in the Zambezi Valley since 1974, a period of more than 30 years. It is a combination of fire and elephant pressure that maintains open woodland with grass (Professors Trollope and Kirkman), - 4000 elephant were culled in the Zambezi valley in 1985. Prior to the Zambezi Valley coming under the control of National Parks and Wildlife there were regular fires, which burnt in the valley ignited by villagers. These anthropomorphic fires were lit while collecting honey, for protection from an encounter with dangerous animals while walking and to improve hunting conditions with a new flush of grass to attract game. 

It is a well-known and researched fact that fire inhibits bush encroachment and forestation but maintains grassland. There is now only 10 % of the North American Prairies left in a patchwork of isolated pockets because of anti fire policies. The growing points of grasses are below ground and within weeks of even a hot fire new shoots appear. They may have begun evolving while the earth was still cooling down. Volcanoes would have started the original fires under which grasses evolved millions of years before man appeared. Since the evolution of man most fires have been anthropological where man has lit fires to provide grass for ease of hunting (the American Indian and in Africa), protection from dangerous game (in Africa) and providing grazing for livestock (the Masai).