Last update: November 17, 2010 12:39:10 PM E-mail Print




PCV du Toit

Grootfontein Agricultural Development Institute

Private Bag X 529, MIDDELBURG EC 5900



Stock farmers and pasture scientists have grappled with the problem of estimating realistic current grazing capacities for the different areas of South Africa for many years.  This problem is not unique to South Africa, in the United States of America, scientists have given this selfsame problem a lot of attention (Stoddart 1952).


In the Nama-Karoo, however, the problem of estimating grazing capacities is exacerbated.  Here, two very different plant life growth forms constitute the forage base available to grazing animals.  These two life forms are the herbaceous plants (hemicryptophytes), with grasses the most common representative plant family.  These plants have their resting buds at, or very close to the soil surface and, the woody Karoo sub-shrubs (chamaephytes), with their resting buds within 20 to 30 cm from the soil surface (Bell & Coombe 1965).  To this latter group a relatively large number of plant families contribute readily available forage.  Not only do these sub-shrubs differ with respect to their contribution to the available forage base, but also as regards their growth cycles, which make it extremely difficult to balance the forage ration throughout the year.  Whilst Karoo sub-shrubs, for the moment not differentiating between different groups on account of their acceptability differences, more-or-less continuously supply high quality forage throughout the year, grasses do so mainly from spring to autumn, being dormant and usually of a low quality in winter.  Although the quality of Karoo sub-shrub forage is high, the quantity supplied at any given moment varies tremendously throughout the year and this is further influenced by the total amount of rainfall received, as well as the temperatures prevailing at the time.


The grazing management of these different plant life growth forms not only makes the rationing out of forage extremely difficult, in order not to over-graze the constituent parts, but the growth buds, borne at different heights above the soil level, complicates grazing management.  In the case of the chamaephytes, a lenient grazing pressure is indicated by the different plants= morphology.  This fact has not received sufficient attention in the past with the planning and execution of the different paddocking systems and the length of the planned grazing period.  This ipso facto resulted in the more palatable species being overgrazed as a result of long periods of occupation usually associated with these early pauci-paddocking grazing management systems.  This state of affairs inevitably led to the demise of the palatable species component over vast tracts of the Nama-Karoo, whilst the less palatable and unpalatable, especially the spinescent species becoming more prominent in the veld.


Added to the problem mentioned above, is the problem of relatively rare plant species.  These species, whilst significantly contributing to the overall plant species diversity and, probably to the quality base of the vegetation, contribute rather little to the available forage base in quantity.  To the grazing capacity estimator, these species therefore cause a significant amount of Anoise@ in the data set used to calculate a veld condition index, from which the current grazing capacity is to be estimated.  It is also further suspected that these relatively rare species, cause serious computational problems, where inappropriate methods of botanical surveying are used to describe the veld condition with regard to past grazing management or incidents of fire and severe droughts.  Under these circumstances, vegetation scientists describing the different trends in the development of the vegetation which came about as a result of different grazing and veld burning management systems and different applied stocking rates, or on account of varying climatic influences, experience serious statistical problems in their calculations (Reilly & Panagos 2002a), usually leading to outrageous demands with regard to the time the botanical surveyor spends on surveys.


It is proposed that the vegetation scientist, whether botanist, pasture scientist or nature conservator, should clearly define the ultimate objective of the study, before selecting a method of botanical survey.  At this stage, the statistical tests to be used to describe significant differences between two sets of data, should be clearly defined in order to select an appropriate survey method.  When the current grazing capacity is to be estimated for use by the grazier, in order to control domestic livestock, the attention needs to be focused on the most common forage producing plant species.  The method of surveying, collection of data and analysis of the data set to describe the veld condition from a grazing point of view, differs significantly from that to be used by botanists to measure species richness and diversity, for the same area.


Discussion of different methods of grazing capacity estimation


Score card systems


The earliest attempts to describe veld condition in any detail, were probably the score card method.  Since the mid-1960's to the mid-1970's, a great deal of research effort was primarily aimed at the description of veld condition as to species composition, soil condition, poisonous plants, insect activity and the degree of veld utilization.  The veld was rated in terms of these various properties and was subjectively scored, using the score card rating system.  From these scores, a current grazing capacity was then estimated.  These researchers were stationed in most of the agro-ecological regions of South Africa.  Some of the more well known score cards were developed in the Highveld Region (serving the highveld of the western-Transvaal and the north-eastern Free State) by Edwards and Coetzee (1971), who developed a score card to measure veld vigour.  In the Free State Region (serving the northern Cape and the western and southern Free State), Fourie and Roberts (1977) developed the score card system to assess the natural pastures for beef production.  Similarly, during that period, Ivy (1969) developed a score card to assess veld condition and productivity in the grasslands of Rhodesia for beef production (Zimbabwe).  Presently, there is renewed interest in the score card systems of veld evaluation.


The very nature of the score card system leads to the subjective evaluation of the veld, a fact which Einstein (undated) would have criticized, had he been a pasture scientist.


Point based methods


A relatively large number of researchers at the different universities and agricultural research stations of the Department of Agriculture, spent a great deal of time developing simple, yet effective methods to estimate objective grazing capacities.  These actions took place from the mid-1970's to the mid-1980's.  Centres of these actions were dispersed over a number of the former agro-ecological regions into which the erstwhile South African Department of Agriculture was organized.  This, ipso facto, led to different approaches to the problem, with every region approaching the problem from a slightly different perspective, on account of the vast differences encountered in their veld=s composition.


In the Natal Region, Booysen, Tainton, Edwards, Mentis and Foran developed a method (Foran et al. 1978; Tainton et al. 1978; Tainton et al. 1980).  Similarly in the Karoo Region, Vorster was working on a model (Vorster 1982), while Fourie, Visagie and Fouche worked on models in the Free State Region (Fourie & Visagie 1985; Fourie & Fouché 1985).  One common factor these methods shared, was, without exception, that they were almost entirely based on proportional plant species composition from which current grazing capacities were estimated by appropriate computation.  On account of this fact and the fact that no reference was made to actual dry matter production, these methods have lately been severely criticized (Kirkman 2002; Letty et al. 2002; Reilly & Panagos 2002b).  Objections to the recording of species composition and the computation of a current grazing capacity without any reference to available dry matter production, have been raised as early as 1982 to 1984 (Du Toit, unpublished departmental memo=s to directors of soil conservation, Natal Region), and is an inherent weakness of all the systems developed during the 1970's and the 1980's.


At the root of the problem with the earlier proposed method of grazing capacity estimation, lies the fact that the prescribed survey method is rather imprecise.  The plant nearest to the actual point, is recorded.  On account of the fact that it is extremely difficult to define the nearest plant, i.e. should the nearest perennial plant be recorded or should the nearest mature plant be recorded.  In practice it happens that the timing of the survey greatly influences the outcome thereof.  In spring with favourable rains more annuals are recorded.  During the following year, at the same time, on account of dry climatic conditions, these annuals are absent.  This fact exerts a tremendous influence on the veld condition index and the resultant estimated current grazing capacity.  This state of affairs have serious complications for the statistician.  Because of this rather serious shortcoming in the nearest plant method of survey, the scientist should refer to Roux (1993) with respect to the different possibilities in the recording of strikes which will be to advantage.  Serious statistical problems are encountered when  the data of the survey results of the nearest plant method are analyzed.  Almost by definition, a definition which is not precisely circumscribed, this method is inherently, unrepeatable (Reilly & Panagos 2002a & Reilly & Panagos 2002b).  Repeatability is the single most important aim of vegetation scientists in order to determine progress or retrogression in the vegetation due to management and climatic variables.


The step point method (Mentis 1981) which is also a nearest plant method is too crude to consider for use in dense vegetation, in rather sparse vegetation, it may have a role to play.  The Apsion@ method (Bosch et al. 1989), based on the nearest plant survey method, is so imprecise and lacking in repeatability, despite protests as to its significant prowess, that it hardly warrants discussion.  Despite pleas to the effect that surveys consisting of at least 500 points be conducted (Du Toit 1995), the Apsion@ method goes even further, in that it decides on statistical probability when sufficient points are recorded and that no further plant variability is to be expected.  This point usually falls far short of the commonly recorded 100 points (Bosch et al. 1989; Booysen 1989).


The nearest plant method, with most of the surveys, but excluding the Apsion@ method, obviously leads to the recording of a hundred percent cover, to which the Apsion@ method contributes its own significant degree of error.  This cover value, usually not mentioned, must on no account be confused with the recording of basal cover as described by Tidmarsh and Havenga (1955), with the result that their statistical tables cannot be invoked to prove or disprove any statistical significance in cases where differences due to management are suspected, whatsoever.  It should also not be confused with the recording of the canopy spread cover propounded at the moment, the measurement of which is defined very precisely, is largely repeatable within the limits laid down and bears up to statistical scrutiny very well (Roux 1993; Du Toit 1995).


The veld condition score, in the case of the nearest plant survey, is calculated from the percentage contribution of the different species to the final tally of proportionally recorded plant species.  These species may, or may not be the species which exert the strongest influence on the soil and environment, nor contribute the most to the vegetation base in the form of available forage.  The difficulties with regard to the recording of the nearest plant were previously discussed (Vorster 1982; Du Toit 1995). 


The author is currently using two methods at Grootfontein, to great advantage, the line-point survey of 500 point observations per survey and the ten metre square quadrat where the aboveground plant production is harvested on a species basis.


As guiding principles with regard to these various surveys, the author was constantly reminded by the words of Einstein (undated) who said: AWhat cannot be measured, is not science, but speculation@ and Lord Snowdon (early 1900's) who said: Aif you cannot express what you are talking about in numbers, you know nothing about it@.  These quotations most of all, should be constant reminders to every botanical surveyor, to be consistent in the application of the chosen method, meticulous and strict with regard to the observation and recording of a strike and, accurate with plant identifications during surveys.


The current method of line-point surveys, as carried out in the Nama-Karoo, record strikes on the canopy spread cover of the grasses and the Karoo sub-shrubs.  This measurement, in addition to providing an indication of the size of the different plants, thereby affecting the probability of a strike, is also an indication of the amount of readily available dry matter that can be grazed by domestic stock (Du Toit 2001).  The percentage cover recorded during the survey is a valuable tool indicating past management to a certain degree.  The model estimating the index values attached to the individual plant species, used inter alia, size, available dry matter and various computations of the chemical composition, to describe the ranking order of the different species and, the grazing value of a specific species to the grazing animal (Du Toit 1997a).  The described model therefore allows for different Apalatability/nutritional scenarios@, for the different species.  Whilst an individual species may be highly palatable and have a favourable chemical composition but a rather low dry matter production, a less palatable species, with a slightly less favourable chemical composition but a much higher dry matter production, will in all probability have a higher index value than the palatable species and therefore contribute significantly more to the veld condition index and the estimated current grazing capacity.  


Both methods, line-point and quadrat, estimate reasonably objective current grazing capacities.  However, it should always still be necessary to regard the estimated grazing capacities with a certain degree of reserve, because, with regard to veld grazing and, in the interests of the sustainable use of the natural resources, it is necessary to err on the conservative side and therefore to apply the estimated grazing capacities realistically.  Estimated current grazing capacities should always be compared to the natural ability of the veld to produce available forage in response to long term climatic conditions, as well as to the grazing capacity norm set for the region.


The line-point method of botanical survey

The simplified line-point method of survey currently in use, evolved out of the wheelpoint method (Tidmarsh & Havenga 1955) and the later modification of the wheelpoint method which incorporated the descending point (Roux 1963).  Currently a line-point survey of 500 point observations per survey (Du Toit 1997b; Du Toit 1997c) is conducted in the specific area for which a grazing capacity has to be estimated, by using a non-stretch rope of 50 to 100 m, suitably marked off at one metre-intervals, and the use of a hand-held rod, lowered onto the soil surface at the marked metre intervals.


The theory and principles governing the current methodology of point surveying remain the same as previously described (Tidmarsh & Havenga 1955; Roux 1963) and this method is comparable with both the wheelpoint and descending-point methods of vegetation survey (Roux 1993).  As far as Karoo vegetation is concerned, this method is sufficiently robust and accurate to record different veld composition scores by measuring the canopy-spread cover before and after grazing events (Du Toit 1997b).  During the survey all the species Astruck@ by the point of the descending rod are identified, recorded and counted.  To each of these species, a specific, unique grazing index value has been allocated (Du Toit 2000; Du Toit 2002a).  The number of individuals of a species is then multiplied by its corresponding grazing index value.


When doing the calculations, it is important to remember that the number recorded, divided by five, in this case with the survey of 500 points, should be multiplied by the grazing index value.  On no account should the recorded number of strikes on a specific species be converted to a percentage value for that species of all the strikes obtained during the survey, here, Vorster (1999) is in error in his description of the grazing index method of grazing capacity estimation.


It is important to remember to divide the number of strikes on a particular species, by the multiple of the total number of point observations observed during the survey.  Expressing the number of strikes on a particular species, as the percentage of the sum of all the strikes obtained, will cause a skewed picture of the specific condition of the veld being evaluated and this will consequently lead to an incorrect grazing capacity being estimated.  For instance, say line-point surveys of 100 points per survey were conducted with the total number of strikes being 50 in the first survey, with the count for Themeda in the survey totaling ten and, the canopy spread cover percentage being 50% therefore, i.e. the number of all the strikes divided by the multiple of 100.  The percentage contribution of Themeda to the veld condition index in this case would then be 20%. That is, 10 divided by 50 multiplied by 100, which is quite erroneous.  When the 20 count on Themeda is now multiplied by its grazing index value, it renders a veld condition index value of 136.20 for Themeda in this survey.  In a separate survey, also of 100 points, conducted in rather degraded veld, a canopy spread cover of only 25% was recorded, with a total count of only five for Themeda.  The percentage contribution of Themeda to the veld condition index of this veld would again be 20%, i.e. 5 divided by 25 multiplied by 100.  Again, in this veld, Themeda would also have a veld condition index value of 136.20.  Multiplying these percentage values by the grazing index value of Themeda, would render both sections of veld with a similar current grazing capacity.  This condition is quite unacceptable and is obviously erroneous, especially when the two surveys are compared to their respective recorded canopy spread cover values, their obvious differences in physiognomical condition and to their past grazing management.  It is patently obvious that Vorster (1999) is in error in this instance.


The products of the number of individuals of a species recorded, multiplied by that specific species= grazing index value, represents the index value of that particular species in the section of veld being surveyed.  All the index values of the individual species are then summed, with this sum representing the veld condition index of the particular section of veld.  This veld condition index is then compared to the veld condition index of the benchmark, i.e. the known standard (Du Toit 2002b).  The veld condition index value of the benchmark, is then divided by the veld condition index value of the veld sample.  The answer obtained here, represents the grazing capacity of the veld sample in ha/SSU (hectare per small stock unit), specifically Merino sheep.  Since a defined small stock unit doesn=t exist, this value has to be multiplied by a factor which represents the average number of small stock units which represents the defined large stock unit (LSU).  This number is 7.14, and is a constant, currently used to describe the average number of producing Merino ewes which are equivalent to the defined large stock unit, while it also simultaneously convert the veld condition index of the benchmark, originally described and defined in morgen, to a value in hectares.  The answer obtained after these computations, represents the grazing capacity of the sample of veld in ha/LSU.  This value can then be compared to the grazing capacity norm applicable to the area.


Botanical veld survey method


The method of botanical surveying currently in use, is based on the observations of a dimension-less plot and the recording of strikes, satisfies all the statistical requirements (Du Toit 1997b; Du Toit 1997c).  The method consists of a line being stretched out in a predetermined direction, while the observer walking along the line lightly spears the point.  It should be decided before the survey on which side of the knot the recording shall be made.  In this fashion, practically complete objectivity is attained in locating the systematically arranged point samples.  There is no reason to fear that samples were being consciously selected.  Rigorously following the principles of the method, makes the method statistically robust and reliable.  It is therefore of the utmost importance that the observer be strict in the observations of whether a strike should be recorded or not, the observer should practice a great deal of restraint and discipline and rigorously observe a strike in the same manner throughout the entire survey.


With experience, a survey of 500 points should take no longer than 20 to 30 minutes.  A homogeneous section of the veld to be surveyed is selected.  In instances where this section of veld does not represent the veld of the whole paddock for which a grazing capacity is to be estimated, a separate survey should be conducted in that portion of the paddock which differs.  The whole idea of surveying to estimate the current grazing capacity, is to repeat the surveys at a later stage, in order to compare the results obtained during each of the follow-up surveys, relative to the management system applied.  It is therefore important to conduct repeat surveys at the same site, at the same time of the year.  The best time in the karroid areas was found to be at the end of the growing season, when most of the Karoo sub-shrub species have flowered and it is a fairly simple matter to identify the different species.  Under these conditions, the few ephemeral annual plant species present, will not materially influence the estimated current grazing capacity.


The canopy spread cover is observed throughout and when the surveying rod passes through the specific species= canopies= circumference, i.e. the perimeter of the cover projected vertically down onto the soil surface, a strike is recorded.  It is important to note that it is not necessary to only record strikes where the rod is in physical contact with the grass or sub-shrub (refer in this regard to Roux 1993).  It is recommended that 500 point observations should be observed during each survey (Du Toit 1997c), unlike the recommendations by Reilly and Panagos (2002) the mean percentage error that will be committed when recording this low number of point observations, is three percent.  This percentage error is quite acceptable in grazing capacity estimations, taking into account the vast number of factors impacting on grassland management.  These recommendations do not hold for surveys where the objective is the recording of species diversity and species reaction to applied conservation measures.  A completely different method of survey should be conducted to satisfy statistical requirements.


An example of a botanical survey conducted on a site in the Eastern Mixed Karoo


Calculation of the veld condition index of a sample site

Species                                Number            OGIV                 Species VCI-value

Aristida congesta        4         1.04                4.16

Asparagus suaveolens     3         0.90                2.70

Chrysocoma ciliata       2         1.12                2.24

Cynodon incompletus      5         1.88                9.40

Eberlanzia ferox         4         1.54                6.16

Eragrostis bicolor       4         3.35               13.40

Eriocephalus spinescens  2         2.12                4.24

Eriocephalus ericoides   4         2.43                9.72

Lycium cinerium          2         1.63                3.26

Pentzia globosa          5         2.30               11.50

Pentzia incana           3         2.88                8.64

Rosenia humilis          2         1.77                3.54

Salsola calluna          4         3.55               14.20

Walafrida geniculata     6         3.52               21.12


Total                   50                          114.28


OGIV = objective grazing index value

VCI = veld condition index


Estimation of the current grazing capacity from the veld condition index, for the surveyed veld sample


The veld condition index of the benchmark site, situated on the experimental site known as Boesmanskop, which is part of the farm Grootfontein, has been estimated at 500 (Du Toit 2002b).  It has previously been estimated that two morgen of this veld could sustain one producing Merino ewe for the whole year, without deterioration taking place in the veld condition.


Therefore: 500 ) 114.28 x 7.14 = 31.24 ha/LSU.


The grazing capacity of the sample site is therefore 31.24 ha/LSU.  From the list of species enumerated during the survey it can be seen that the pioneer grasses and less palatable and unpalatable Karoo sub-shrubs dominate the vegetation composition.  The low veld condition index value together with the low canopy spread cover percentage, explains the large deviation from the grazing capacity norm of 16 ha/LSU applicable to this area.


Quadrat based methods


The quadrat harvest method of botanical survey


The easiest method of evaluating the veld by means of a quadrat, is to use a ten square metre quadrat.  Various surveyors use quadrats of varying sizes.  The author found that by using the 10 metre square quadrat, calculations later on are simplified, in addition it has less edge effect than 10 one-metre square quadrats, and especially in sparse vegetation such as in the Arid Karoo and Desert Grassveld (Acocks 1988), a more realistic estimate of aboveground plant production is obtained by the larger quadrat.  A portion of the veld, representative of the paddock for which a grazing capacity is to be estimated, is selected and the quadrat laid out.  All the species included in the quadrat are clipped at stubble height for the grasses and approximately 50 mm above soil level for the Karoo sub-shrubs.  Clip each species separately and put the clippings in separate bags.  The material is usually dried at 60EC, overnight.  The material is then weighed, species by species and multiplied by either their particular rule-of-thumb usage factors.


The usage factor for highly desirable species is 0.5.  Desirable species have a factor of 0.3, less desirable species have a factor of 0.2, while undesirable species have a factor of 0.1.


The plant species specific unique objective grazing index values can also be used during these computations, but for this exercise they need to be divided by ten in order to render them of the same order of magnitutde as the rule-of-thumb values.


Now the product of every species= weight is multiplied by its usage factor and summed.  This represents the usable material harvested in every quadrat.  The value obtained is expressed in grammes dry matter per ten square metres.  Since there are 1000 quadrats measuring ten square metres in a hectare, the product of the species weight multiplied by its usage factor obtained in grammes, can immediately be expressed as that amount of dry matter in kilogrammes per hectare.  Now according to the given formula; that the defined large stock unit consumes 10 kilogrammes of acceptable dry matter per day, to grow at a rate of 500 grammes per day, one large stock unit therefore needs 3650 kilogrammes of suitable forage per year (Meissner et al. 1983).


Simply divide the total annual needs by the production per hectare to calculate the number of hectares necessary to keep one large stock unit in fodder for a year.  Care must be taken when applying this method of grazing capacity estimation because this method is inclined to over-estimate the current grazing capacity of the veld, on account of a number of inherent difficulties during the harvest process.  For instance, only about 50% of the harvested sub-shrub material contributes to the forage base (Du Toit 2001).  Likewise, only 50% of the grass material produced contributes to the forage base, should more than this amount be taken off at any given stage, the specific grass tuft is in fact over-grazed at that stage.


A further complication when using this method of estimation, is the fact that it has been determined that small stock will only voluntarily graze twigs of Karoo sub-shrubs to a thickness of 1.5 mm (Du Toit 1996).  Separation of harvested sub-shrub dry matter is usually done on the rule-of-thumb basis that small stock voluntarily graze twigs with a diameter of up to 2 mm (Botha 1981; Botha et al. 1990).

An example of the quadrat survey to estimate the current grazing capacity
Species                                    Production          Factor          Usable forage

Aristida congesta            3          0.1            0.3

Ephemerals                  64          0.1            5.4

Eragrostis lehmanniana     264          0.3           76.2

Eragrostis trichophora      55          0.5           27.5

Schmidtia pappophoroides  1526          0.3          457.8

Stipagrostis uniplumis     155          0.2           31.0

Themeda triandra           432          0.5          216.0


Total usable forage                                  904.2


3650 ) 904.2 = 4.04 ha/LSU


Obtaining the square root of the product of the grazing capacities estimated by means of the line-point method and the quadrat method


It is a sound practice to obtain the square root of the product of multiplication of the grazing capacity obtained by means of the line-point method, with that obtained by means of the quadrat estimate.  This method makes use of the answers obtained by both methods and will provide a fairly representative current grazing capacity value for the area in question.  However, as mentioned previously, it is always advisable to use the estimated grazing capacity value with caution, as it is only an estimate of the true grazing capacity.  Calculating the square root of the product of the two methods, makes use of the species= unique index values as well as the species contribution to forage production per hectare.  It will be found that this Aanswer@, lies very close to the Acorrect@ grazing capacity for the area, taking into consideration the various important complicating factors, such as for instance the time of the year and time since the last grazing took place.


Always keep in mind that the estimated current grazing capacity of any section of veld, must be corroborated by comparing it to:

  • The grazing capacity norm applicable to the area,
  • The production of forage in response to the amount of rainfall received in the area annually, with due regard to the long term mean rainfall recorded in the area (Du Toit 2002c), and,
  • The stocking rate applied by conservative, progressive farmers in the area.


 Simultaneous estimation of the current grazing capacity


1. Line-point botanical survey on pediment at Grootfontein on 11th July 2001, in the False Upper Karoo (Acocks 1988)

Species                                                  number            ogiv*             ogiv/vci*


Aristida diffusa                  1        3.12         3.12

Chloris virgata                   1        1.45         1.45

Digitaria eriantha                4        4.70        18.80

Enneapogon scoparius             11        2.80        30.80

Eragrostis lehmanniana           23        3.26        74.98

Eriocephalus ericoides           16        3.08        49.28

Pentzia incana                    8        3.39        27.12

Tragus koelerioides               3        1.69         5.07


                                 67                   210.62


* see previous explanation                                                16.95


The estimated current grazing capacity for the veld on Grootfontein is therefore 16.95 ha/LSU.  Considering its species composition, the calculated grazing capacity is not far different from the grazing capacity norm of 16 ha/LSU applicable to the area.


2.   Quadrat survey on pediment at Grootfontein carried out simultaneously to the line-point survey on the 11th July 2001


2a.  First calculation, using the objective grazing index values, but these, divided by ten, to render them equivalent in order of magnitude to the rule-of-thumb values

Species                                          mass 1      mass 2       mass 3       ogiv       forage


Aptosimum depressum          378     189      141.8   0.074   10.49

Aristida diffusa             574     287      287     0.318   91.27

Digitaria eriantha           149      74.5     74.5   0.637   47.46

Eragrostis lehmanniana       263     131.5    131.5   0.324   42.61

Eriocephalus ericoides      1920     960      720     0.243  174.96

Felicia muricata              11       5.5      4.1   0.343    1.41

Gazania oxyloba               69     34.5     25.9   0.116    3.00

Melolobium candicans          12       6        4.5   0.112    0.50

Pentzia incana               276     138      103.5   0.288   29.81

Tragus koelerioides          125      62.5     62.5   0.084    5.25




Dry matter needed as forage                                                                              3650 kg

Forage available per hectare                                                                                406.76 kg


3650 ) 406.76 = 8.97 ha/LSU

8.97 X 16.95 = 152.0415

s 12.33 ha/LSU


Comments on mass 1, mass 2 and mass 3 are necessary at this point: Mass 1 refers to the total standing crop of Karoo sub-shrubs harvested.


Mass 2 refers to the portion of this harvested material which contributes to the forage base, separated out by the rule-of-thumb of 2 mm (Botha 1981).  According to this separation the available portion is only approximately one half of the total mass harvested (Du Toit 2001) (Fig. 1).


Fig. 1   Total aboveground dry matter production and its separation into non-grazable and grazable fractions


The correction applied to the mass 2 value comes about through the investigation that sheep will only graze twigs with a mean diameter of 1.5 mm voluntarily (Du Toit 1996), measured experimentally where Afrino sheep at four different stocking rates grazed the veld at Carnarvon in the Arid Karoo (Acocks 1988).  Regressions of animal production on applied stocking rate, yielded an optimum stocking rate value for this veld, around 40 ha/LSU.  At this stocking rate, the sheep were voluntarily grazing twigs of the Karoo sub-shrubs with a diameter of only 1.5 mm (Fig. 2).


Fig. 2  Animal production (Y=6.012+128.8X-360.6X5) and grazed stem diameter (Y=1.225+29.98X-80.4X5), stems voluntarily grazed to a diameter of 1.5 mm at the optimum stocking rate of Afrino sheep grazing the Arid Karoo and Desert Grassveld at Carnarvon


2b.       Second calculation using Arule-of-thumb@ forage usage factors


Species                  mass 1    mass 2    factor   forage


Aptosimum depressum       378      189        0.1     18.9

Aristida diffusa          574      287        0.3     86.1

Digitaria eriantha        149       74.5      0.5     37.25

Eragrostis lehmanniana    263      131.5      0.3     39.45

Eriocephalus ericoides   1920      960        0.2    192

Felicia muricata           11        5.5      0.3      1.65

Gazania oxyloba            69       34.5      0.1      3.45

Melolobium candicans       12        6        0.1      0.6

Pentzia incana            276      138        0.2     27.6

Tragus koelerioides       125       62.5      0.2     12.5




Dry matter needed as forage                                                  3650   kg

Forage available per hectare                                                   419.5 kg


3650 ) 419.5 = 8.70 ha/LSU

8.70 X 16.95 = 147.47

s 12.14 ha/LSU



The current grazing capacity estimated by means of the line-point method of survey indicates a grazing capacity of 16.95 ha/LSU.  This is very close to the grazing capacity norm applicable to this area.  However, the rather high estimated current grazing capacity estimated for the area by means of the simultaneous estimate, indicates that a significant amount of dry matter has accumulated over time, due to the low applied stocking rate and infrequent grazing treatments of this area.




It is possible to estimate reliable current grazing capacities for any section of veld in the karroid areas at the moment.  Due regard being given to the method of surveying and the application of the estimated values.  With proper training and minimal knowledge of the Karoo sub-shrubs and grasses, reliable estimates can be made.




Acocks, J.P.H.  1988.  Veld Types of South Africa.  Memoirs of the Botanical Survey of South Africa, no. 57, Government Printer, Pretoria.

Bell, P. & Coombe, D.  1965.  Strasburger=s textbook of Botany.  Longmans, Green and Co. Ltd., Great Britain.

Booysen, J.  1989.  Personal communication.  Potchefstroom University for Christian Higher Education, Potchefstroom.

Botha, P.  1981.  The influence of species-selection by sheep, cattle and goats on the floristic composition of mixed Karooveld.  Unpublished D.Sc. thesis, P.U. for C.H.E., Potchefstroom. (Published in Afrikaans).

Botha, P., Erasmus, C.H. & Theron, S.C. 1990.  Mean phytomass and chemical composition of a number of plant species in the northwestern Karoo.  Technical Communication no. 227.  Government Printer, Pretoria.

Bosch, O.J.H., Kellner, K. & Scheepers, S.H.E.  1989.  Degradation models and their use in detecting the condition of southern african grasslands.  Proceedings of the XVI International Grassland Congress.  Nice, France.

Du Toit, P.C.V.  1995.  The grazing index method of range condition assessment.  African Journal of Range and Forage Science, 12(2):61-67.

Du Toit, P.C.V.  1996.  Karoobush defoliation in the Arid Karoo.  Journal of Range Management, 49:105-111.

Du Toit, P.C.V.  1997a.  A Model to estimate grazing index values for Karoo plants.  South African Journal of Science 93:337-340.

Du Toit, P.C.V. 1997b.  Description of a method for assessing veld condition in the Kazoo.  African Journal of Range and Forage Science, 14(3):90-93.

Du Toit, P.C.V. 1997c.  Grazing-index method procedures of vegetation surveys.  African Journal of Range and Forage Science, 14(3):107-110.

Du Toit, P.C.V. 2000.  Estimating grazing index values for plants from arid regions.  Journal of Range Management, 53:529-536.

Du Toit, P.C.V.  2001.  The relation between canopy spread cover and the aboveground available phytomass of Nama-Karoo sub-shrubs and grasses.  African Journal of Range and Forage science, 18:143-146.

Du Toit, P.C.V.  2002a.  Objektiewe weidingswaardes van Nama-Karoo plantegroei; grasse en bossies van die Karoo.  Grootfontein Agric, 4:8-19.

Du Toit, P.C.V. 2002b. Boesmanskop grazing-capacity benchmark for the Nama-Karoo.  Grootfontein Agric, 5:1-6.

Du Toit, P.C.V.  2002c.  Modelling Nama-Kazoo sub-shrub dry matter production using climatic variables.  South African Journal of Science, 98:541-542.

Foran, B.D., Tainton, N.M. & Booysen, P.deV.  1978.  The development of a method for assessing veld condition in three grassveld types in Natal.  Proceedings of the Grassland Society of southern Africa, 13:27-33.

Fourie, J.H. & Fouché, H.J. 1985.  Die bepaling van weidingskapasiteit vanaf weidingskapasiteit.  Glen Agric, 14(1):12-13.

Fourie, J.H. & Visagie, A.F.J. 1985.  Weidingswaarde en ekologiese status van grasse en karoobossies in die Vrystaatstreek.  Glen Agric, 14(1):14-18.

Ivy, P. 1969.  Veld Condition Assessments.  Reprinted from The Rhodesian Agricultural Journal as Bulletin no. 2051 for the Department of Conservation and Extension, pp. 1-5 plus veld condition score sheet.  First discussion for adoption of the method at the Veld Management Conference, Bulawayo, 27 to 30 May 1969, and published as the proceedings of that conference.

Kirkman, K.P.  2002.  Veld Management.  GSSA/SASAS Joint Congress, Congress Program and Abstracts, 13 to 16 May 2002:37a-37b.

Letty, B., Gumede, S. & Khubone, O.  2002.  The farming systems approach to the management of communal grazing areas.  A case study: The Obonjaneni community grazing project.  GSSA/SASAS Joint Congress, Congress Program and Abstracts, 13 to 16 May 2002:58a-58b.

Meissner, H.H., Hofmeyr, H.S., van Rensburg, W.J.J. & Pienaar, J.P.  1983.  Classification of livestock for realistic prediction of substitution values in terms of a biologically defined Large Stock Unit.  Technical Communication no. 175, pp 1-40.  Government Printer, Pretoria.

Mentis, M.T.  1981.  Evaluation of the wheel-point and step-point method of veld condition assessment.  Proceedings of the Grassland Society of southern Africa, 16:89-94.

Reilly, B.K. & Panagos, M.D.  2002a.  Within stand statistical power of commonly used point methods in grassland monitoring.  GSSA/SASAS Joint Congress, Congress Program and Abstracts, 13 to 16 May 2002:2.

Reilly, B.K. & Panagos, M.D.  2002b.  Statistical power of commonly used point methods in grassland monitoring.  African Journal of Range and Forage Science, 19:117-122.

Roux, P.W. 1963. The descending point method of vegetation survey.  A point sampling method for the measurement of semi-open grasslands and Karoo vegetation in South Africa.  South African Journal of Agricultural Science, 5:273-288.

Roux, P.W. 1993.  The line-point method of survey and measurement of Karoo, grassveld and tall shrub vegetation.  Unpublished manuscript, Grootfontein Agricultural Development Institute.

Stoddart, L.A.  1952.  Problems in estimating grazing capacity of ranges. 6th International Grassland Congress, pp 1367-1373.

Tainton, N.M., Edwards, P.J. & Mentis, M.T.  1980.  A revised method for assessing veld condition.  Proceedings of the Grassland Society of southern Africa, 15:37-42.

Tainton, N.M., Foran, B.D. & Booysen, P. deV.  1978.  The veld condition score: an evaluation in situations of known past management.  Proceedings of the Grassland Society of southern Africa, 13:35-40.

Tidmarsh, C.E.M. & Havenga, C.M.  1955.  The wheel-point method of survey and measurement of semi-open grasslands and Karoo vegetation in South Africa.  Memoirs of the Botanical Survey of South Africa, no. 29, Government Printer, Pretoria.

Vorster, M. 1982.  The development of the ecological index method for assessing veld condition in the Karoo.  Proceedings of the Grassland Society of southern Africa, 17:84-89.

Vorster, M. 1999.  Karoo. (In N.M. Tainton ed.)  Veld Management in South Africa, pp. 207-214.



Karoo Agric Vol 6 (1)