Eco Study 2020

Management report: 17 July 2020
Prof Edward. C. Webb, PhD (Pr.Sci.Nat.)

An initial extensive game ranch management plan was complied in 2006 by the Centre of Wildlife Management, Department of Animal and Wildlife Sciences at the University of Pretoria. The report provides an update and assessment of the current status of the game, grazing browse and other aspects of game ranch management.

Date of assessment
This follow-up assessment was done on the 11th and 12th July 2020.

Description of assessment
All watering and feeding points were visited and an assessment was done of the most important grazing and browsing areas on the game farm. Game counts were verified with reported counts from owners and visitors during the previous year. In the subsequent assessment and calculations, the following criteria were employed to make recommendations about the current state and strategic decisions to maintain the game farm in a sustainable and ecologically sound facility.

  • Total farm area and hectarage
  • Veld, browse and ecological assessment by evaluating different ecological
    regions and transects of the game farm.
  • The 5 management units previously identified in the Eco-study were
  • Game counts (drive counts, on days of assessment)
  • Known group counts (on days of assessment)
  • Frequent drive counts of game by owners / manager
  • Veld quality and browse condition assessment
  • Ecological impact of the game on watering and feeding areas
  • Aerial survey and assessments
  • Previous Ecological study by the Centre for Wildlife Management at the
    University of Pretoria.

Follow-up assessments are required in September this year when the lowest veld condition is anticipated, as well as a veld and browse quality assessment in Summer (January) to assess the species composition and state of decreaser and increaser veld species.

Game farm region, size and description
The “Highlands Wilderness Game ranch” is located about 126 km north of Pretoria and about 20 km northwest of Bela-Bela. It is situated in the foothills of the Waterberg mountain range, consisting predominantly of a few undulating koppies with steep slopes, interwoven by indwelling plains and valleys.

There are approximately 49 houses on the game farm, as well as a homestead with workshops and sheds for a manager. The built areas comprise approximately 1.13 ha x (49 + 1) = 56.5 ha. An additional non-usable area includes stand no 100 comprising about 70.76 ha which is not included in that game ranch plan.

All calculations were done based on a farm size of 1172.74 ha. Roads and fire breaks were not subtracted from the total size of the farm, but it should be borne in mind that the total size of usable grazing and browse of the farm probably slightly exceeds the estimates 1172.74 ha due to the presence of about 4 koppies which generally increases the surface area marginally, although it makes it difficult to estimate its exact surface area.

Figure 1. Highlands Wilderness game ranch management map.

Highlands Wilderness game ranch is reached through a gravel road a few kilometres after the Bela-Bela boat club to the property. The farm has several gravel roads to the different stands as well as to watering and feeding points. Gravel roads also service a large part of the perimeter fence, although not all those roads are in a good condition. The farm also has several 4 x 4 roads accessible only by 4-track vehicles.

The perimeter fence of 2.4 m is in a reasonably good condition given its age – the perimeter fence of the farm is outlined in Figure 2 below. A well-maintained firebreak of about 6 m was observed for most of the perimeter of the farm, which is commendable in this mixed bushveld area, where veld fires often occur in the dry season (autumn and winter).

Figure 2. Highlands Wilderness game ranch (areal map from Google Earth,

The Highlands Wilderness game farm is situated in the mixed sour bushveld (Acocks, 1988; Mucina & Rutherford, 2006), between longitude 28º 33’S and latitude 24º 90’E. The altitude of the farm varies between 1397 m and 1522 m (Figure 3). The geology the farm consist of granite (lower altitude southern parts of the farm), rhyolite (higher altitude northern part of the  farm) and a small portion of arenite formations. Steep slopes may hold some problems in terms of soil erosion if overgrazed or disturbed. For this reason, stocking rates should be managed carefully.

Figure 3. Topographical map of the Highlands Wilderness game ranch (from arcGIS).

The rainfall of Highlands Wilderness game ranch falls within the 600-700 mm bracket as indicated in Figure 4. The area is known for typical and consistent spring and summer rainfall, with infrequent droughts. The veld type is classified as Waterberg-Magaliesberg summit sour veld in the savanna biome (Scott, 2016). The rainfall and soil types are important since these influence the type and quality of grazing and browse in the area. The veld types (predominant grass species) and trees on the farm have been described in detail in the Ecological study of

Figure 4. Mean annual rainfall of the Limpopo province. (Highlands Wilderness game ranch falls in the 600-700mm rainfall per annum category).

Maps indicating the infrastructure, management units, geology and veld types previously included in the Ecological study are presented in Figures below to clarify some of the points of discussion further on in this report.

Figure 5a. Map indicating the infrastructure, roads and watering points at Highlands Wilderness game ranch (from Ecological study of 2006).

Figure 5b. Map indicating the soil potential of Highlands Wilderness game ranch (from Ecological study of 2006).

Figure 5c. Map indicating the veld types at Highlands Wilderness game ranch (from Ecological study of 2006).

Figure 5d. Map indicating the management units at Highlands Wilderness game ranch as defined in the Ecological study of 2006.

The trees previously described in the Ecological study namely bush willows (Compretum spp.) silver cluster leaf (Terminalia sericea), weeping wattles (Pelotophorum africanum), wild seringa (Burkea africana), and velvet raisin (Grewia flava) and some Ocna pulcra are generally in a good state, with some dense areas that may have to be managed in future. Ocna pulcra is an indicators species of poison leaf, but the risk of plant poisoning is generally very low for the game compared to livestock.

The grass species are dominated by thatch grass species (Hyparrhenia hirta, Hyparrhenia dissolute), Brooms love grass (Eragrostis pallens), purple spike cat’s tail (Perotis patens), but with indications of some increaser species including Aristida spp.

The lower altitude areas in the southwestern part of the farm (described as Management unit 1 in the previous Eco-study) consist of a larger open savanna of about 350ha (30% of farm area), which is characterised by increasing open and barren patches, void of any decreasers or increaser I & II grass species. Much of this occurs close to water and feeding troughs (please refer to Figure 6), which indicates overuse and overgrazing.

These areas should be managed more conservatively, e.g. by moving water and feeding points to other locations on the farm. This is an important area to assess since bulk grazers (zebra) and selective grazers (blesbok, waterbuck and blue wildebeest) are more dependant on grazing in this area, while many mixed feeders (impala, kudu and eland) will also utilise this area, which increases the pressure on the area and lowers the overall carrying capacity of the farm.

Figure 6. Southwestern part (described as Management unit 1 in the previous Eco-study) of the Highlands Wilderness game farm, illustrating increasing open and bare patches in the
savanna area of the farm.

Veld condition assessments of the central and eastern parts of the farm (Management units 3, 4.1 and 4.2) are comparable to previous assessments and demonstrate reasonable veld condition scores (Figure 7). The quality of grazing (decreasers vs increasers should be assessed again in Summer of 2021 to update the grazing capacity of the farm).

Figure 7. Central and eastern parts of Highlands Wilderness game ranch (Management units 3 and 4.1 &4.2), demonstrate moderate veld condition scores.

Farm carrying capacity and stocking rate
The carrying capacity of the farm is not synonymous with stocking rate, e.g. carrying capacity is the potential of the game farm to sustain different species of animals, while stocking rate refers to the manager’s or owners’ decision to stock the farm at a determined rate.

Game farms in the mixed bushveld are renowned for their variation in terms of topography and vegetation, which makes it possible to stock such farms with a larger variety of game species. Although the amount of grazing and browse are generally high, the veld quality is often marginal and characterised by a significant decrease in nutrient composition (protein, energy, mineral and vitamin content) as well as decreasing digestibility due to an increase in crude fibre content of grazing. Also, the availability of browse decrease during winter months.

The carrying capacity of farms are often calculated using large stock units (LSU) approach, which was originally developed to calculate the number of hectares required to carry a large stock unit. This method has been adapted to convert different species of different sexes and ages to LSU’s with reasonable accuracy (Meissner et al., 1983). In the game industry, the use of the grazing animal units (GAU) and browsing animals units (BAU) provides a slightly more detailed method and makes provision for more accurate estimation of browse. The description
of the available biomass from browse is done by use of Biomass Estimates from Canopy

Volume (BECVOL) method (Smit, 1996). Both LU’s and GAU & BAU’s were employed in the calculations of carrying capacity of the farm.

The current carrying capacity of the farm in terms of LSU’s, GAU’s and BAU’s is presented in Table 1. The grazing capacity of the farm is estimated at 147 LSUs (e.g. 147 x 450 kg bovine cows weighing 450 kg and consuming 10kg feed/d). Based on GAU’s and BAU’s, the aforementioned GAUs is estimated at 170 which under current circumstances seem somewhat overestimated based on current grass cover and species composition, as well as
moderate overuse of the southwestern part of the farm (management units 1 and 4). The main limitation of this, is the fact these management unit are important for grazers, but are also overexploited by mixed feeders and some browsers like kudu.

Table 1. Size and carrying capacity of game species on the Highlands Wilderness game reserve

The total browsing animal units have increased a bit due to the recovery of the veld and browse in management unit 5 which was previously affected by a veld fire. A rough estimate suggests that the BAU of this management unit 5 has increased to about 14, but this needs to be confirmed again in summer. Nevertheless, the quality of treas has not decreased and the number of invader species such as sickle bush (Dicrostachys cinerea) is not common, which is commendable since the latter is an indicator of bush encroachment.

The total BAUs could safely be set at about 104.24, but as mentioned before, this needs to be assessed again in summer to estimate the canopy sizes and available biomass, especially in Management unit 5.

As mentioned previously, Ocna pulcra which serves as an indicator species of poison leaf, was observed in the sandy areas, which indicates the presence of poison leaf on the farm. However, poison leaf does not pose a serious risk to the game as compared to livestock species.

Game numbers, sex ratios and reproduction rates
Game numbers generally agree well with recent game counts by observations from vehicles by owners. A drive count was performed and known groups of game were also counted. The numbers and presence of game spoor at the water and feeding troughs were also assessed.

These counts confirmed the numbers of blesbok, blue wildebeest and giraffe, while it also provided a good indication of impala and kudu. Recent photos confirm the presence of Njala, rooihartbeest, oryx and eland. (One rooihartbeest mortality was confirmed by photograph).

Among the tracks at water points, a clear leopard track was identified, confirming the presence of a large adult leopard on Highland Wilderness game rach. Tracks of smaller carnivores like caracal and jackal were also observed. Based on these observations, the data presented in Table 2 provide a good estimate of the current game numbers present on Highlands
Wilderness game ranch.

In assessing the numbers of game in confined areas, it is important to determine the total numbers of each species classified into their typical feeding preference, e.g. bulk grazers, selective grazers, mixed feeders and browsers.

It is also useful to assess the locations where game frequently occurs on the farm, as well as the numbers of male and female animals, to predict normal sex ratios and risk of inbreeding and loss of fitness. The numbers of progeny is also useful in the typical lambing season to indicate normal reproduction patterns.

Table 2. Estimated numbers, sex ratios and reproduction rates of game on Highlands Wilderness game ranch as on July 2020.

Although mortalities, culling and predation should be considered, the current numbers of zebra indicate that this species has adapted well and are quite prolific at Highlands Wilderness game ranch. This is beneficial from a game production point of view, but the numbers may increase beyond the normal carrying capacity (expressed either as LSU’s or Grazing animal units) of the farm, and sex ratios may become problematic, as is demonstrated in the current zebra herd. Inbreeding depression and poor reproductive fitness could become more problematic if
new genetic material (e.g. new unrelated males from other game farms) are not imported to maintain genetic fitness.

The ratio of male to female zebra of 37% far exceeds the norm of 15%, so it is recommended that at least about 10 males are sold, and at least 2 or 3 new stallions bought to prevent possible inbreeding depression in the herd.

The same situation is clear for giraffe, which has adapted well and is quite prolific. However, in the case of the giraffe, it is recommended that the bull be replaced with a new bull (new genetic material), but the numbers are not currently a major concern because they browse at a browsing height far exceeding that of other species on the farm and do not pose a serious risk at this stage. Considering the current good physical condition and reproduction rate of giraffe, there is a large enough browse resource to sustain the current number of giraffe on
the farm of about 10 animals.

Some chewing and mild symptoms of pike were observed in some giraffe (and also observed in kudu), which indicates possible mineral deficiencies (probably phosphorus), which will require a more aggressive NaCl-phosphate lick supplementation strategy for these animals.

Blesbok and Impala
Both blesbok (grazers) and impala (mixed feeders) are generally in a good condition (especially at this time of year) and the numbers of animals seem to have increased since the Ecological study. However, the reproduction rates of both blesbok and impala are below that expected for both these species, and it is known that there has been culling and sales of the game. Also, it is suspected that carnivores such as leopard are causing more losses in these two groups of game at Highlands Wilderness game ranch. So although the number of grazers
are approaching the estimated grazing capacity of the farm, culling is not recommended at this stage since predation probably maintains an acceptable balance between births and mortalities.

Nevertheless, the ratios of males to females are much higher for both blesbok and impala compared to the norms. This together with possible inbreeding depression, necessitates removal or culling of about 50% of the males of each of these species and replacement with about 5 blesbok rams and 5-8 impala rams. This is a good time to buy game due to the relatively low prices for these types of game species.

Sable antelope and Oryx
Based on recent photo’s of sable antelope and oryx on the farm, it appears that both species are in an excellent condition and should be maintained as such. No drastic management measures are recommended at this stage for sable antelope or oryx, apart from strategic supplementary feeding which will be discussed a bit later in this report in the section on feed supplementation.

The leopard may likely have caused some mortalities in these herds, which is why attention should be given to the placement of water points and supplementary feeding of pellets in closer proximity to where Sable antelope and oryx generally graze (closer to the koppies). Sable for example often hide the newborns and juvenile animals in bushes before going to watering and feeding points.

The kudu observed at Highlands Wilderness game ranch is in an excellent condition for this time of year. Beautiful mature bulls and well-conditioned cows with several calves and juvenile animals were observed in herds of about 10 to 12 animals per herd. The exact number of kudu could not be determined accurately, but the observed numbers of about 69 animals seems plausible, especially considering the large number of kudu tracks (including large bulls) that were observed at every water trough during my 2-day visit and drive counts.

Some shewing and symptoms of pike were observed in kudu, as indicated through excessive chewing of bark and occasionally objects from the ground. This indicates possible mineral or or nutrient deficiencies) and most likely phosphorus deficiency which is quite common this time of year continuing to mid-summer. The protein content of browse (Combretum spp.) tend to be relatively high compared to grazing so a protein deficiency is less likely than a phosphorus (P) deficiency. Besides, it is known that the Ca:P ratio in browse is much larger compared to grazing, which may aggravate this situation in browsers.

It is recommended that a phosphorus (P12)-salt lick should be supplemented to kudu (and giraffe) to remedy this problem. Lucern or good quality hay could be supplemented from late winter onwards, but a P12-salt lick is even more crucial at this stage. P-deficiencies are well known to adversely affect growth and reproduction in animals.

Game numbers versus carrying capacity of the farm
The numbers of game expressed as game animal units (GAU’s) and browse animal units (BAU’s), and relative to the farms grazing and browse capacity are summarised in Table 3. 

Table 3. Numbers of game expressed in graze units, browse units and % of graze and browse capacity.

An assessment of the veld and browse quality indicate that Highlands Wilderness game farm has been managed reasonably well over the past 14 years. Some improvements and some retrograde veld succession have been observed compared to the Ecological study of 2006 (see Table 3).

The browse quality and biomass production of trees in Management unit no 5 has improved since the veld fire. This improved the estimated total browsing capacity of the farm to about 104.24. However, it is estimated that the current browse capacity of the farm is exceeded by about 31.98 BAU’s. In this regard, it should be remembered that the impact of giraffe seems high in terms of their 25% contribution to the browsing impact of the farm, but giraffe utilises the biomass of trees at a much higher browsing height, which is of reasonable quality and quite assessable to the giraffe at highland Wilderness game ranch. This probably explains why giraffe is in good physical condition and maintain high reproduction rates.

At this stage, the management of giraffe would include mostly replacing the large bulls with unrelated bulls from elsewhere to prevent inbreeding.

Kudu has numerically the highest browsing impact namely 59% of the browsing capacity of the farm. This species should be monitored carefully in terms of physical condition and possible deficiencies. Inbreeding may be a problem, although it is known that large bulls easily clear high game fences and do migrate among farms. A reduction of the numbers of kudu or culling is not recommended before the numbers have been confirmed more accurately. Also, leopard probably cause some predation losses of young kudu calves and maintain numbers
to the current observed 8% reproduction rate, which is well below expected rates reported for kudu of about 20-25%.

Eland currently represent about 21% of the browsing capacity of the farm, but they do browse at a different height compared to impala and kudu. The Eland seem to struggle more than other species at highlands Wilderness game ranch, and may be attributed to previous inbreeding and loss of fitness. However, it appears that a dominant bull jumped to a neighbouring farm and similarly a bull from a neighbouring farm jumped back to Highlands Wilderness game ranch. Based on this occurrence, no immediate action is recommended since the breeding of the unrelated eland bulls with cows on the ranch should improve the fitness of Eland.

The browsing impact of impala is comparable to that of eland (e.g. about 20% of the farms browse capacity) but obviously at a different height so the competition effect is small. However, impala, blue wildebeest and blesbok are highlighted ion yellow because these species impact that grazing capacity of the farm and appears to overuse certain common areas, especially in management units 1, and 4.1 and 4.2, even though only about 50% of the total grazing capacity of the farm is used.

The most critical aspect of impala, blue wildebeest and blesbok is the high ratio of males in the herds and long absence of newly introduced males of all three species. This poses a risk for inbreeding and loss of fitness in all three species. For this reason, the immediate need is to remove some males form each of these species and introduce unrelated males from elsewhere into the herds at highlands Wilderness game ranch. It is  recommended that about 50% of the adult impala, blesbok and wildebeest males are removed and replaced with about 5 unrelated blesbok rams, 5-8 impala rams and 3 blue wildebeest bulls (these numbers make provision for possible morbidity or mortalities).

Physical condition and fitness of game
Observation of the game at Wilderness highlands game ranch indicates that most game is in a good physical condition, apart from the pike (chewing behaviour) observed in kudu and giraffe. It is always difficult to determine the physical condition and growth of game species accurately. It is not always possible to get close enough to observe and assess the animal’s physical condition (size, conformation and hair coat quality), but the results obtained by Hirst (1975) provide some values for growth data of game at different ages, which may provide an
indication of the condition, growth and fitness of animals in terms of size and estimated weight for age (Table 4, and illustrated in Figures 8 to 10). Comparisons to weight for age with the data provided in Table 4 may assist to assess if growth, development and condition are acceptable.

Table 4. Summary statistics for mean body weights of ungulates at different stages of maturity (from Hirst, 1975)

The size and development of game animals at different ages can be determined by comparing animals of different ages and sexes. The difference in size between males and females are quite pronounced in many game species, which is known as sexual dimorphism. This is a normal feature of many game species that should be observable in game animals that are well adapted and in a good physical condition.

Nutrition and supplementary feeding of game
The most responsible approach in planning game nutrition on game farms is to supplement nutrients that may be deficient during certain seasons of the year, rather than substituting the available grazing of the game farm. Several products are available to safely supplement deficiencies and perhaps improve nutrient utilisation from the natural resource base, e.g. better utilisation of available grazing and browse.

Firtsly, it should be reiterated that Highlands Wilderness game rach is located in the Waterberg mixed sour bushveld, which is known to provide and maintain a reasonable amount and quality of grazing during the spring and summer months of the year, but with a deterioration in nutrient quality in autumn and winter, although the quantity of grazing and browse often remain high.

This slows growth and may adversely affect reproduction specifically of grazers and mixed feeders during autumn and winter, due to increasing nutrient deficiencies, typically protein and energy content of grazing, and certain minerals and vitamins such as calcium, phosphorus, and vitamins A and E.

Mineral deficiencies (and sometimes also protein) often cause pike (chewing behaviour) in grazing mammals. The chewing behaviour involves picking up and chewing foreign objects in the environment such as bones or carcass remains which may contain Clostridium botulinum which causes botulism in animals. The overuse of browse trees often cause an increase in the production of tannins which decreases the palatability of trees. This deters mixed feeders and browsers to feed on such trees and may cause nutrient deficiencies and emaciation. Such animals are more prone to develop pike and are therefore more at risk to be infected with botulism and die.

For this reason, the 1st important supplement at Highlands Wilderness game ranch should be a phosphorus-salt (P12-NaCl) lick supplement provided to especially cloven-hoofed (ruminant) game such as kudu and giraffe. There is some debate if P12-salt licks should be provided all year round, but a safe approach is to provide this at species-specific feeding height (e.g. fed on an elevated platform to the giraffe) for as long as it is consumed. Generally P12-salt licks are supplemented in animals on green grazing, but also in winter and spring in browsers. P12-phosphorus licks are also beneficial since it provide other important minerals (see table 4).

Protein and energy supplements are often required for certain game species in autumn and winter due to the significant decrease in protein and energy content of natural grazing and to some extent also in the browse. Lucern has a relatively high protein content and is often used during winter to supplement protein, especially if the quantity of grazing and available browse becomes scarce.

It is unlikely that there will be serious quantitative nutrient deficiencies at Highlands Wilderness game ranch soon, (e.g. too little biomass for grazing and browsing), but the quality of the grazing and browse will be limiting during autumn, winter and early spring. For this reason, lucern supplementation may be more expensive and cause more waste.

My recommendation is to rather use lick blocks (e.g. Superblok 150, Molateck Trophy block, or similar products) to supplement protein, energy, minerals and vitamins in a more controlled way (avoiding excessive consumption). Such lick blocks should not contain urea since it could cause urea poisoning of monogastric hindgut fermenters such as Zebra.

Game pellets may be an alternative option, but only if used strategically for short periods when the quality of grazing is lowest. Several such products are available e.g. Alzu GP100, Cap-Chur game pellets, Mwedow game pellets or Afgri’s Gametech pellets.

Problem animals
Management of problem animals is problematic and there are few good long-term solutions. Nevertheless, this is an ongoing challenge that requires sustained management. Common “problem species” on game farms include predators such as leopard, caracal and baboons.

It is not recommended managing leopard by trapping or culling, since leopard tend to migrate to where a habitat where they are not threatened, so another leopard will simply replace the current leopard (s). The same applies to caracal, which together with leopard form a normal part of this pristine ecosystem which should be maintained as such. 

Management of baboons is much more challenging. Previous attempts to manage baboons include the use of fences, barbed wire on infrastructure and roofs that are damaged, but with varying success. It is important to ensure that game feeding areas are not to close to houses or other built infrastructure that can be easily reached and invaded by baboons. 

A possible solution is to use the technologies employed to mage grazing livestock using 12V electrified wires. Such wires can be used on the edges of roofs and on walls to prevent baboons from climbing on houses and causing damage. These wires have a low fire or environmental risk and should be safe for use in the current scenario.

Summary of recommendations

  1. The browse capacity of Highlands Wilderness game ranch has improved some what due to the recovery of management unit 5 after the veld fire.
  2. It is a concern that the grazing quality of management units 1 and 4.1 & 4.2 show some retrograde succession, even though grazers and mixed feeders are stocked at only 50% of the grazing capacity. This may require more planning about the strategic locations of watering and feeding points.
  3. P12-salt supplementation is recommended for kudu and giraffe. (It is not a problem if other species also consume some of these mineral licks).
  4. It is recommended that about 50% of the adult impala, blesbok and wildebeest males are removed and replaced with about 5 unrelated blesbok rams, 5-8 impala rams and 3 blue wildebeest bulls.
  5. About 10 zebra males should be removed or sold, and replaced with at least 2 or 3 new unrelated stallions to prevent inbreeding depression in the herd.
  6. It is recommended that the current number of giraffe not be exceeded, but the dominant bull be replaced when possible.
  7. It is recommended that the sable antelope and oryx be maintained as such.
  8. Eland require careful monitoring of condition and reproduction over the next year.

Acocks, J.P.H. 1988. Veld types of South Africa. Mem. Bot. Surv. S. Afr. 57.

AGIS.AGRIC. Accessed 29 Aug 2010.

Van Hoven, W., Orban, B., Deutchlander, N., & Mokoena, J.R., 2006. Ecological study of the
plant communities and animal populations of Highlands Wilderness game ranch, with
management recommendations, Centre for Wildlife Management, Department of Animal &
Wildlife Sciences, University of Pretoria.

Hirst, S.M. 1975. Ungulate-habitat relationships in a South African woodland. Savannah
ecosystem, Wildlife Monographs, The Wildlife Society, No. 41.

Low, A.B. & Rebelo, A.G. 1996. Vegetation of South Africa, Lesotho and Swaziland.
Pretoria: Dept Environmental Affairs & Tourism.

Meissner, H.H., Hofmeyer, H.S., Van Rensburg, W.J.J. & Pienaar, J.P., 1983. Classification
of livestock for realistic prediction of replacement values in terms of a biologically defined
large stock unit, Department of Agriculture, RSA.

Mucina, L. & Rutherford, M.C. 2006. The vegetation of South Africa, Lesotho and
Swaziland., Pretoria: Strelitzia 19. South-African National Biodiversity Institute.

Scott, L., 2016. Fluctuations of vegetation and climate over the last 75 000 years in the
Savanna Biome, South Africa: Tswaing Crater and Wonderkrater pollen sequences
reviewed, Quertenery Science Review, 145, 117-133.

Smit, G.N., 1996. BECVOL: Biomass estimates from canopy volume (version 2) – users
guide. Unpublished manual. University of the Orange Free State. Bloemfontein.