Last update: November 24, 2010 08:36:02 AM E-mail Print


Animal production under different grazing regimes at Grootfontein

P. C. V. du Toit

National Department of Agriculture

Grootfontein Agricultural Development Institute

Private Bag X529, MIDDELBURG, 5900



The relative merits of rotational grazing over continuous stocking, as well as the advantages of multi-paddock rotational grazing systems over some of the pauci-paddock systems received much attention in extensive, published reviews (Gammon 1978; O’Reagain & Turner 1992). An opportunity to study these aspects in the False Upper Karoo (Acocks 1988), presented itself some years after a trial was initiated where cattle and sheep grazed together in a multi-paddock system (Du Toit 1995). This experiment was compared to two facets, each of two separate experiments, started earlier; a stocking rate trial and the Camp Number 6 grazing experiment (Tidmarsh 1951; Du Toit 1998). In order to study the effects of grazing only, the important principle of rotational resting was not incorporated into the multi-paddock rotational grazing system. The incorporation of yearlong rotational resting leads to improved plant production ascribed to an improvement in veld condition and hence to improved animal production. This facet was therefore not incorporated in the present set of experiments and the results, regarding grazing, are directly comparable.


Three systems of animal production on the fragile Karoo veld were studied. Paddocks in which these grazing systems were applied are situated on both pediments and plains. Plains are generally described as broad watercourses, with deep fertile soils supporting a dense, rich species composition (Du Toit 1998). Pediments are moderately steep, occur at the foot of hills and are similar to the veld on hillocks, are stony, with shallow, sandy soils and supports a poor, sparse, species composition(Du Toit 1998). Both terrain types are found in all three grazing systems.

Experiment number one 

The grazing management system in experiment number one follows a multi-paddock rotation with the period of occupation 14 days per paddock. A stocking rate around the norm was chosen for this large, unreplicated experiment, in order to study the effect of cattle and sheep grazing the veld together and to serve as a demonstration to farmers in the area. This experiment was laid out on the Boesmanskop grazing area of Grootfontein. It comprises an area of 206 ha, is divided into 11 paddocks, which varies in size from 13 ha to 36 ha. The Boesmanskop experiment is unreplicated, on account of the large tract of land necessary to accommodate sufficient animals to obtain meaningful production responses. In addition to the plains and pediments, a fairly steep sided, flat topped hill which covers about 90 ha of the 206 ha is included in the Boesmanskop area. The rest of the area consists of approximately one third plains and two thirds pediment. A large part of this hill, on the top and on the north facing slope, supports sparse vegetation of poor species composition, growing in shallow, rocky soils. Due to problems with the provision of water, the hill area is subdivided into only three large paddocks, hence leading to a grazing period of 28 days per paddock, which operates as follows: 14 days grazing, 28 days recovery and a second round of 14 days grazing before the grazing recommences on the rest of the experiment. A herd of six head of Afrikaner cross cattle and 36 Merino wethers rotationally grazed the area as a unit in a 1:6 ratio i.e. a 1:1 ratio on a large stock unit basis, which is the ratio found to be the most beneficial to both the vegetation and animal performance in the sour grassveld of Natal (Hardy 1995; Hardy & Tainton 1995). This ratio is also recommended by the Department of Agriculture in farm planning exercises. The experiment commenced in 1991 and still continues on veld previously grazed in a two- to three-paddock system, with occupation periods varying from 3 to 5 months. Cattle and sheep graze together at a mean seasonal stocking rate of 15.3 ha LSU-1. This is the approximate grazing capacity norm of 16 ha LSU-1 which is applicable to veld of this part of the False Upper Karoo (Acocks 1988). Since stocking rate is such an important variable in veld research work (Edwards 1980), a stocking rate around the norm was chosen for this large, unreplicated experiment. The applied stocking rate is the same as the medium stocking rate applied in experiment number two.

Experiment number two 

In the second experiment, part of which is a pauci-paddock rotational grazing management system, Merino wethers, Dorper wethers and Angora goat kapaters graze the veld at three stocking rates, low (24 ha LSU-1), medium (16 ha LSU-1) and high (8 ha LSU-1), in order to test the validity of the grazing capacity norm of 16 ha LSU-1 laid down for this area. This experiment is known as the Afrikaner/Hereford stocking rate trial. This experiment commenced in 1988 and still continues, on veld previously also grazed on a two- to three-paddock system, with periods of occupation varying from 3 to 5 months. The Afrikaner paddocks are situated on hillocky plains, while the Hereford paddocks are situated on hillocky pediments. Alternated grazing between the pediment and plains was applied in such a fashion as to ensure that no paddock was grazed during the same time of the year and in the same season (Du Toit 1996a & b), for two successive years. The medium stocking rate of the Merino sheep only, was duplicated and it approximates the grazing capacity norm of 16 ha LSU-1. Only the two medium stocking rate facets of the Merino sheep were included in the present study. Both treatments consisted of groups of ten sheep. The entire area was divided into two blocks, known as Afrikaner block (Block A) and Hereford block (Block B) respectively. The 20 ha allocated to each group was further subdivided into two blocks of 10 ha each (Blocks A1 and B1 and Blocks A2 and B2).

The group of 10 Merino’s grazed Block A1 uninterruptedly for 120 days before moving to Block B1 for the following 120 days following the slow rotation method. This group was stocked at a seasonal mean stocking rate of 15.5 ha LSU-1. The experimental treatment approximates the previous two-paddock grazing system, but with fixed periods of grazing.

The 20 ha allocated to the second group was divided into eight paddocks of 2.5 ha each. Four paddocks in the Afrikaner block (Block A2) were grazed in a quick rotation, the grazing period being 14 days. After 120 days of grazing these four paddocks, the group moved to the next set of four paddocks, the Hereford block (Block B2). These four paddocks were again rotationally grazed for 120 days, with the grazing period being 14 days. The mean seasonal stocking rate of this group was 15.3 ha LSU-1. This quick rotation grazing treatment is directly comparable to the first experiment, except for the exclusion of the large stock factor.

Experiment number three 

The main aims of the third experiment is to demonstrate various management options on the Karoo veld. 1. Fixed season grazing of two-paddock and three-paddock rotational grazing systems on pediment. 2. Continuous grazing (Booysen 1967); continuous stocking of pediment, plains and a combination of pediment and plains. 3. Yearlong continuous stocking alternated by yearlong rest of a paddock on pediment. This last treatment refers to the rotational rest treatment advocated by the Department of Agriculture.

The continuous rest treatment of a paddock was included to study the effects of resting on the species composition of karoo pediment (Du Toit 1998). Only two of the eight facets of the Camp Number 6 grazing experiment were selected for this study. The two facets being the continuous high and the continuous low stocking rate treatments. These two treatments were each situated on two thirds pediment and one third plains (Du Toit 1998)(refer also to the Boesmanskop treatment). Most of the treatments of this experiment continued unchanged since 1934 (Du Toit 1998). In the two continuously stocked treatments, two groups of nine and seven Merino sheep each grazed 13 ha, at mean seasonal stocking rates of 9.9 and 12.1 ha LSU-1 respectively. These stocking rates are higher than the norm but correspond to the stocking rates applied by farmers practicing continuous stocking in the region. Relative to the ungrazed control plot, the veld condition of these two treatments deteriorated significantly since the beginning of the experiment (Du Toit 1998). Plant species composition has stabilized and now represents the early developmental phase of vegetation development, being Aristida congesta and A. adscensionis.

Experimental procedures 

Animals were weighed fortnightly without prior fasting (Hardy & Tainton 1995).

The study period ran from the end of October 1996 to the end of July 1997. Total rainfall for the study period was 582 mm for the 1st experimental area, 552 mm for the 2nd experimental area and 544 mm for the 3rd experimental area, which is almost 200 mm above the norm for the area (Table 1). Trends in animal production in the three production regimes were studied and for this reason it was reasoned that the differences in starting weights of the different groups could be ignored, because young, growing Merino wethers were used in all three experiments. Animal weights were computed to an index of weight change to facilitate interpretation. Animal handling such as shearing and dosing was similar in all three systems during the experimental period. In previous years, similar grazing periods were very short, often interrupted by drought and comparable results as to body weights and grazing days difficult to obtain on account of the different times of shearing.

Table 1 Rainfall received in the experimental blocks during the 1996/97 season, recorded in mm

Experimental block




July 1996




August 1996




September 1996




October 1996




November 1996




December 1996




January 1997




February 1997




March 1997




April 1997




May 1997




June 1997









Discussion of results 

In Figure 1 the starting weights for the different groups of animals were recalculated as an index and all the weights start at one. The recomputation of animal weights on an index basis was necessary because not enough animals of an even starting weight were available to fulfil the requirements of all three systems. Despite the different starting weights of the animals in the different groups this did not prove to be a problem, all the animals were young, growing animals, with no restrictions placed on their ability to produce on the veld. Research indicates that on this type of veld, the lighter animals perform better than the heavier animals (Table 2). Trends in the production ability of the animals in the different systems were the main concern and this is the aspect that was studied. Rainfall did not limit either plant production nor animal production on the veld. The above average rainfall led to satisfactory plant growth, enabling good conclusions to be drawn from the animal production data. The animals in experiments two and three were expected to perform better than the animals of experiment one when the values in Table 2 are compared. Gains per ha and average daily gains (A.D.G.’s) for the animals in all the experimental treatments are given in Table 3.

Table 2 Sheep weights for the Boesmanskop area, Experiment 1 for three seasons 1996/97 to 1998/99


Starting mass


End mass


Gain (kg)

Rain (mm)





















Table 3 Production per ha, average daily gain (kg or g/animal/day) and the seasonally applied stocking rates for the three different experiments



Seasonally applied

stocking rate

(ha LSU-1)

Production per ha


Production per animal per day


Boesmanskop (Experiment 1)




(cattle 8.66;

sheep 6.2)

Cattle 0.49 kg/a/d

Sheep 58 g/a/d


slow rotation (Experiment 2, Blocks A1 and B1)



32 g/a/d


quick rotation (Experiment 2, Blocks A2 and B2)



45 g/a/d

Camp 6 continuous low (Experiment 3)



21 g/a/d

Camp 6 continuous high (Experiment 3)



5 g/a/d

The fact that all the animals gained weight during the summer growth period, indicates that sufficient dry matter was available on the veld for grazing during the summer period, following the favourable summer rains. Soon after autumn, the continuously stocked treatments started losing weight, with the high stocking rate treatment losing weight more rapidly than the lower stocking rate treatment. This can be ascribed to the fact that during the previous years of continuous stocking, the species composition in these paddocks deteriorated. The current vegetation, consists mainly of annual grasses, such as Aristida spp and less palatable and unpalatable karoo subshrubs, that were unable to support the animals during the winter (Du Toit 1998). Insufficient dry matter remained available after the initial summer grazing period, to ensure weight gains during the winter period in these two treatments. This statement is in direct opposition to earlier work and reports, where continuous stocking was found to be superior in maintaining plant and animal production (O’Reagain & Turner 1992; Kreuter et al 1984).

Irrespective of the treatment differences and different starting weights, all the animals initially gained weight during the summer growth period (refer to Fig. 1 and Table 1). However, soon after autumn, animals in the continuously stocked treatments started losing weight, with animals in the high stocking rate treatment losing weight more rapidly than animals in the lower stocking rate treatment (Table 3 and Fig.1), despite the fact that rainfall did not limit plant production. Regression analysis of all the data points of production per ha on the seasonally applied stocking rates indicates that in the continuously stocked trials, stocked at the grazing capacity norm of 16 ha LSU-1, a production of 5.5 kg ha-1 could theoretically have been realised. However, a regression of only the two continuously stocked data points, revealed that, even after extrapolation of the regression to the grazing capacity norm, a production of only 3.4 kg ha-1 would have been realised. This figure is appreciably lower than the production realised by any of the three rotational grazing systems. From the foregoing, it seems that the continuously stocked treatments on this degraded type of veld, with its species composition, mainly consisting of annual Aristida spp, have reached their peak production.

Fig. 1 Indices of weight change for the three experiments. BSMN refers to the Boesmanskop experiment, K6H refers to the Camp number 6 contiuous high stocking rate, while K6L refers to the Camp number 6 continuous low stocking rate, AH2/4 refers to the slow 4 month rotation of the Afrikaner/Hereford experiment, while AH8/14 refers to the 14 day rotation of the Afrikaner/Hereford experiment

The animals of the Afrikaner/Hereford medium stocking rate treatments did not lose weight after the initial growth period, but compared to the Boesmanskop experiment, the quick rotation gained at a slower rate, while the weights of the slow rotation remained fairly constant over the autumn/winter period. The quick rotation treatment gained longer into the winter than did the slow rotation (Fig. 1). The vegetation of these treatments consisted of a good species mix of annual and perennial grasses and various palatable and less palatable Karoo subshrubs. The grazing period of the slow rotation of the Afrikaner/Hereford treatment was too long and the animals grazed out the palatable species component leaving only less palatable and unpalatable species for grazing later in the season. These species are presumably inferior in quality and grazing them led to an initial drop and a lower rate of weight gain during the latter part of the season.

The sheep in the Boesmanskop treatment gained weight at a fairly constant rate throughout the summer, maintaining their growth rate well into the autumn and winter period. This came about not only as a result of the good species composition available to them for grazing, but also grazing a higher bulk of younger, more palatable regrowth, in most of the pediment and plains paddocks. This constant rate of gain is ascribed to the fact that the cattle possibly opened up the vegetation by grazing the long, rough vegetative material, enabling the sheep to selectively graze the younger more nutritious material which grew out following earlier defoliations and by so doing, to constantly gain in weight (Fig. 1). This is in line with present research findings. The quick rotation, 14 day period of occupation per paddock, led to sufficient dry matter being produced during the fairly extended periods of absence.

The grass-rich False Upper Karoo (Acocks 1988), is locally referred to as the Eastern Mixed Karoo (Du Toit 1996c) on account of the high proportion of late developmental stage grasses which occur together with the more palatable Karoo subshrubs throughout this area. It is generally recommended that cattle and sheep should graze this veld together, in order to utilise the available dry matter to advantage and by so doing, raise the production per ha (Botha et al 1983). However, contrary to recommendations (Botha et al 1983), the stocking rate in this instance was not raised simultaneously. From the results obtained in this experiment, it seems that the advice given to graze cattle and sheep together was warranted, because a higher production per ha was realised. The decision not to raise the stocking rate was equally vindicated in that no overgrazed patches occurred in the veld. The cattle produced 8.66 kg ha-1, while their A.D.G. was 0.49 kg/animal/day and the sheep 6.2 kg ha-1, with an A.D.G. of 58 g/animal/d giving a mean production of 7.4 kg ha-1 for the cattle and sheep combined. The Afrikaner/Hereford sheep of the quick rotation produced 6.85 kg ha-1, at an A.D.G. of 45 g/d, as opposed to the 6.2 kg ha-1 of the Boesmanskop sheep only (Table 3), while the sheep of the slow rotation produced 4.9 kg ha-1 at an A.D.G. of 32 g/animal/d (Table 3). Over the experimental period on the 206 ha of the Boesmanskop experiment it means that 370.8 kg more live weight can be produced than would have been produced on the same area by sheep alone, following the comparable quick rotation grazing system. The extra 1.8 kg ha-1 live weight that was produced by the cattle and sheep in the Boesmanskop experiment during the experimental period, over the sheep only of the Afrikaner/Hereford quick rotation, represents a production of 370.8 kg more live weight gain than could be realised by sheep of the comparable quick rotation system alone, on similar veld. This represents quite a substantial increase in production on these relatively poor Karoo soils.

The continuously stocked treatments did not perform as well as the rotationally grazed treatments, A.D.G.’s of 5 and 21 g/animal/day, as opposed to 32, 45 and 58 g/animal/day (Table 3). Production in the continuously stocked treatments approaches the production achieved by the slow rotation of the two-paddock grazing system, but only after regression of all the data points.

The pauci-paddock rotational grazing system produced better than the continuous stocked treatments, but the average daily gain is only about half that realised by the multi-paddock system. This is in direct opposition to the statement that there is no experimental evidence to indicate that one form of rotational grazing is better than the other (Gammon 1978). The multi-paddock system was defended in the past quoting more flexibility, higher animal production and the development of a better species composition on the veld (Rennie 1975).

The above discussion of results stands in stark contrast to the findings that of all the southern African grazing experiments reviewed (O’Reagain & Turner 1992), most favoured continuous stocking. The reasoning was flawed with respect to the Karoo experiments, in that continuously stocked treatments on a combination of pediment and plains, were compared to fixed season two- and three-paddock grazing systems on pediment only (Du Toit 1998; Donaldson 1986). It has already been pointed out previously (Du Toit 1998), that the two- and three-paddock systems did not perform as well as some of the continuously stocked treatments in this experiment, on account of the fixed seasonal nature of the grazing and as a result of the different classes of soil and veld potential on which the paddocks were laid out. The conclusions that data, collected over the short term, indicated the superior performance of continuous stocking over rotational grazing in A.D.G. as well as in live weight production ha-1 (Kreuter et al. 1984) does not agree with the present findings. Although only one year’s data of the three experiments are presented here and used in the comparison, the systems have stabilized under present management, where these differed from past management. Where continuous stocking has been compared to rotational grazing in the past, this was done on veld which has not stabilized under the applied management system and therefore the conclusions drawn were in the main, invalid (Rennie 1975). In the present study, continuous stocking over several decades has led to a reasonably stable, early developmental stage veld (Du Toit 1998) and it is this veld that was used in the comparison of the animal production potentials. Taking into account the mid-seral stage of the vegetation condition overall, it is gratifying to note that improved veld management systems, i.e. multi-paddock veld management systems, lead to improved animal production.


The advice usually given to graze cattle and sheep together in the Eastern Mixed Karoo (Du Toit 1996c) is warranted and the narrow ratio of cattle to sheep produced good results, because an increased production was realised. The increase of 1.8 kg ha-1 in live weight production on these relatively poor Karoo soils is substantial. From the production figures it can be seen that continuous stocking on the fragile, drought prone Karoo soils cannot be recommended, not even as in this case, during an above average rainfall season.

Many authors claim veld degradation following years of continuous stocking with either cattle or sheep while others claim that continuous stocking is not inferior to rotational grazing in terms of its impact on the vegetation (Gammon 1978; O’Reagain & Turner 1992; Rennie 1975; Kreuter et al. 1984). Therefore, in the light of the above findings, if valid comparisons are to be drawn between rotational grazing and continuous stocking, it must be done on veld which have extended histories of both methods of treatment on the veld, as in the cases cited above, where the vegetation species compositional structure has stabilized within each grazing management system.

Literature references 

Acocks JPH 1988. Veld Types of South Africa. Memoirs of the Botanical Survey of South Africa 57. Government Printer, Pretoria.

Booysen PdeV. 1967. Grazing and grazing management terminology in Southern Africa. Proceedings of the Grassland Society of southern Africa 2:45-57.

Botha P, Blom CD, Sykes E & Barnhoorn ASJ 1983. A comparison between the diets of small and large stock on mixed Karoo veld. Proceedings of the Grassland Society of southern Africa 18:101-105.

Donaldson CH 1986. The Camp No. 6 veld grazing trial: an important milestone in the development of pasture research at the Grootfontein College of Agriculture. Karoo Agric 3:1-6.

Du Toit PCV 1995. Die bepaling van die optimum weidingskapasiteit in die Oostelike Gemengde Karoo deur middel van verskillende kleinveerasse. Unpublished progress report, Grootfontein Agricultural Development Institute, Middelburg.

Du Toit P C V 1996a. Seasonal grazing of Camp 6, in the False Upper Karoo. Unpublished Grootfontein Agricultural Development Institute Report, Middelburg.

Du Toit PCV 1996b. Seasonal grazing of a Karoo Mountainpaddock in the False Upper Karoo. Unpublished Grootfontein Agricultural Development Institute Report, Middelburg.

Du Toit PCV 1996c. Development of a model to estimate grazing-index values for Karoo plant species. Ph.D. thesis, University of Pretoria.

Du Toit PCV 1998. Effects of grazing and resting treatments on animal performance and vegetation condition in the False Upper Karoo at the Grootfontein Agricultural Development Institute, Eastern Cape. South African Journal of Science 94:507-512.

Edwards P J 1980. The use of stocking rate/animal performance models in research and extension. Proceedings of the Grassland Society of southern Africa 15:73-77.

Gammon DM 1978. A review of experiments comparing systems of grazing management on natural pastures. Proceedings of the Grassland Society of southern Africa 13:75-82.

Hardy M B 1995. Short-term effects of cattle-to-sheep ratio and stocking rate on patch grazing in sour grassveld. African Journal of Range and Forage Science 12:121-127.

Hardy MB & Tainton NM 1995. The effects of mixed-species grazing on the performance of cattle and sheep in Highland Sourveld. African Journal of Range and Forage Science 12:97-103.

Kreuter UP, Brockett GM, Lyle AD, Tainton NM & Bransby DI 1984. Evaluation of veld potential in East Griqualand using beef cattle under two grazing management systems. Journal of the Grassland Society of southern Africa 1:5-10.

O’Reagain PJ & Turner JR 1992. An evaluation of the empirical basis for grazing management recommendations for rangeland in southern Africa. Journal of the Grassland Society of southern Africa 9:38-49.

Rennie JG 1975. More beef from pasture I Veld management for a sustained increase in meat production. Proceedings of the Grassland society of southern Africa 10:151-154.

Tidmarsh CEM 1951. Veld management studies at the Grootfontein College of Agriculture. Unpublished progress report no 111, 1934 - 1950, Grootfontein Agricultural Development Institute, Middelburg.



Karoo Agric Vol 4 (1)