Last update: August 17, 2011 03:50:36 PM E-mail Print


The short-term response of Merxmuellera disticha and Aristida diffusa dominated veld to fire and follow-up clipping treatments


C. H. Donaldson



Fire is generally accepted as a factor which has long shaped the nature of South African grasslands. It is also regarded as a useful rangeland tool for the control of undesirable woody plants, macchia and Karoo  encroaching species (Trollope, 1978). There are also strong indications that a single veld fire is able to reduce drastically the population of key pioneer Karoo shrubs like Chrysocoma tenuifolia (Donaldson & Mostert, 1958) and Eriocephalus eriocoides (personal observations). Under present-day conditions veld fires in the Karoo areas are rare occurrences, except in the grassy traditional Karoo mountain areas. The grassveld areas which may be burned to advantage occur mainly in the Karroid Merxmuellera mountain veld (Acocks, 1975). These areas are the Suurberg in the Noupoort, Colesberg and Steynsburg districts, the Sneeuberg in the Middelburg, Richmond and Graaff-Reinet districts and the sour-grass mountain veld in the Molteno district. The subject of veld burning in the Karoo mountain regions has been reviewed by Roux & Smart (1979). Except for the preliminary work by Roux (1969), virtually no planned veld burning trials have been conducted in the Karoo Region.

Guidelines for veld burning of Karoo mountain veld are available (Du Preez, 1983). These stipulate the proviso that land users must obtain the necessary permission from their local extension officers to burn veld. Usually it is laid down that burning should be done when the soil is damp (after winter snow or rains) and preferably after spring rains of approximately 25 mm but not before the middle of August or after the end of October.

The object of this paper is to report on the short-term results of a veld burning trial designed to obtain some quantitative data on the effects of three burning treatments, conducted during the months of July, September and October respectively, on the subsequent growth, as measured by three different post-fire clipping treatments.



The short term effects of three veld-burning treatments on grass production were investigated in two separate experiments conducted on the Grootfontein College of Agriculture (near Middelburg) and on the farm Wintersrus, situated approximately 30 kin west of Molteno. The three fire treatments comprised burning during August, September and after substantial rains in October, respectively. The October burns were applied 3 days after a total rainfall of 44,4 mm for the Grootfontein site and 4 days after 60 mm for the Wintersrus site. The experimental sites are situated in False upper Karoo and Karroid Merxmuellera mountain veld (Acocks, 1975), respectively. Aristida diffusa was the dominant grass species of the Grootfontein site (Table 1), whilst Merxmuellera disticha dominated the Wintersrus site (Table 2).

The three burning treatments were randomly allocated to plots measuring 25 m x 25 m at the Grootfontein site and 20 m x 18 m at the Wintersrus site. There were two replications of each treatment at both sites. An estimate of the species composition of each plot was obtained by a 200 point survey with points spaced approximately 1 m x 2 m apart. At each point (allocated from a marker on a measuring chain) the plant species of the nearest plant to the point was recorded. Ten 1 m2 quadrats randomly allocated to each plot were harvested immediately before the burns were applied. This material was used to obtain the moisture content of the herbage and the dry matter fuel loads of the plots before burning. Soil moisture determinations of the top 15 cm soil layer were determined gravimetrically for each treatment before burning. Six samples were taken from each plot.

Three follow-up clipping treatments were applied to two 10 m2 plots randomly allocated to each replicate of the treatment plots. The three clipping treatments were also applied to replicated plots from adjacent unburnt veld. Treatment 1 comprised four cuts taken on 02/02/87, 11/08/87, 23/02/88 and 16/02/89 for the Grootfontein experiment and on 29/01/87, 13/08/87, 24/02/88 and 10/03/89 for the Wintersrus experiment, respectively. The clipping programme for treatment 2 was similar to treatment 1 with the exception that the cuts of 02/02/87 (for the Grootfontein site) and 29/01/87 (for the Wintersrus site) were omitted. Only the last two cuts of treatment 1 were applied as treatment 3.

The DM-herbage production obtained from the abovementioned clipping treatments was used to assess the differences in the quantity of vegetation under the three fire treatments. Representative herbage samples of the grasses Aristida diffusa (Grootfontein site) and Merxmuellera disticha (Wintersrus site) were taken from the February 1988 cuts for crude protein and acid detergent fibre (AD F) analyses. The herbage samples were analysed by the Chemistry Section of Grootfontein using standard analytical procedures.



Pre-treatment conditions pertaining to experimental Sites

The percentage species composition at the onset of the fire treatments of the Grootfontein and Wintersrus plots are given in Tables 1 and 2 respectively. Other useful parameters are also presented in these tables.


Features of the botanical data are the dominance of the strongly perennial and wiry grass species, A. diffusa (Table 1) and M. disticha (Table 2), at the Grootfontein and Wintersrus sites, respectively, and secondly, the relatively small presence of Karoo shrubs at both sites. The fuel loads (Tables 1 and 2) were also high enough to ensure a clean burn of all plots.


Dry matter production of burnt and unburnt plots

The effects of the different burning treatments on the DM-herbage productions of the two veld types as measured by the three follow-up clipping procedures are presented in Tables 3 and 4. Also presented are the rainfalls for the clipping periods as well as the DM yields of unburnt plots.

The analysis of variance for the DM-yield data of both sites revealed no significant differences in the total DM production obtained from the three fire treatments, irrespective of which follow-up clipping treatment was applied to evaluate these treatments. The lack of differences between the total DM yield obtained from either four cuts, three cuts or two cuts for each respective burn treatment in the case of either site (Tables 1 and 2), suggests that burning during late winter (August) does not appear to be more detrimental to veld growth than burning in spring, before (September burns) or after (October burns) rains.

It is further suggested that the lack of difference in the yields of these three burns was brought about by one or more of the following factors:

Firstly, the burning treatments were applied at a time of the year when soil temperatures are generally not high enough for active grass growth; secondly the little growth that did take place during the pre-burn and early post-burn stages of grass growth was adequately provided for by the presence of some moisture in the soils (Tables 1 and 2); thirdly that very favourable rainfalls were recorded during the post-burn growing period (Tables 3 and 4).


It is interesting to note that the total DM yield of 5299 kg (= 226 + 536 + 1249 + 601 + 1364 + 1 323) for the unburnt plots of A. diffusa veld (Table 3) was 29,6 % higher than the corresponding total mean yields of the burnt plots of 3728 kg (= 185 + 349 + 800 + 489 + 958 + 947) (Table 3), whilst the differences in the total yield of the unburnt plots (5 725 kg) and burnt plots (5118 kg) of M.disticha veld (Table 4) were only 10,6 %. The results suggest that either the A. diffusa veld reacted differently to burning and clipping than M. disticha veld, and/or that the differences in climatic and soil conditions between the two sites were mainly responsible for these findings.


Crude protein (CP) and acid detergent fibre (ADF) content of herbage

Data showing the mean effect of burnt and unburnt veld on CP and ADF of herbage cut at three different clipping intervals for the dominant grass species of the Grootfontein and Wintersrus sites are presented in Table 5.


A striking feature of the data (Table 5) is the very low CP and ADF contents of herbage (at the end of the summer period) for all treatments. Veld dominated by either A. diffusa or M. disticha should therefore be grazed selectively at this time of the year and later on. The application of a controlled selective type of grazing practice may therefore ensure that the animals' nutritional requirements are mainly met by selectively eating the nutritious parts of these grasses as well as other more palatable plants.

The data in Table 5 also indicate differences in CP and AD F of the herbage of burnt and unburnt plots. These differences also increased with an increase in the cutting interval. The chemical analysis of the burnt plots, for both grasses, were higher for CP and lower for ADF than those of the unburnt plots. It is to be expected that the quality of the grass herbage after a burn will be much higher than the standing D M herbage such as that of the unburnt plots of the two cut follow-up clipping treatment (Table 5). In this case it would appear that the mean CP of A. diffusa for burnt plots of 3,5 % was, for example, much higher than the 2,7 % of the unburnt plots. Similar results were also obtained for M. disticha (Table 5). Of interest, however, is that the quality (CP and ADF values) of the herbage of both grasses from the burnt plots of the four cuts and three cut follow-up clipping treatments was generally higher than that of the same age material obtained from the unburnt plots.

Apparently very little data are available on the effect of veld fires on the herbage quality of burnt veld under South African conditions. Tainton et. al. (1977) and Preller (1950) recorded a higher quality for burnt than for unburnt veld herbage.

It can be concluded that the data presented in this paper are in accord with the general veld burning guidelines advocated for Karoo mountain veld. The unreliable and cyclic nature of the rainfall of these areas emphasizes the importance of burning veld only when the soil is very moist or after very effective rains in spring, to ensure rapid sprouting and active growth of grasses. For obvious reasons the plant cover must be dense enough to obtain a good clean burn throughout a camp, thus avoiding patchy burns. The idea of grazing burnt M. disticha veld intensively over a short period of three weeks when the regrowth after the burn is 8 to 10 cm tall and then resting the veld until autumn or winter, with the aim of giving the sour grass a set-back and hopefully also favouring the palatable grasses, may be a sound policy. There is, however, a need for detailed studies on the effects of this and other practices on the vegetation and on animal performance.



Messrs A. Barnhoom, F. Roux and G. Jordaan are thanked for their assistance in the field work.



ACOCKS, J.P.H., 1975. Veld types of South Africa. 2nd Ed. Mem. Bot. Surv. S. Afr. No 40. Dept Agricultural Technical Services, Pretoria.

DONALDSON, C.H. & MOSTERT, J.W.C., 1958. Alarming encroachment of bitterbos in the Orange Free State. Fmg. SA 34:53-56.

DU PREEZ, G., 1983. Riglyne vir die brand van veld in die Karoo-berggebied. Karoo Agric 3(2),8-11.

PRELLER, J .H., 1950. Pasture and Pasture Crops. Sci. Bull. No 280, Dept of Agric. Union of South Africa.

ROUX, P.W., 1969. The autecology of Tetrachne dregei Nees. Ph.D. Thesis. Univ. Natal, Pietermaritzburg.

ROUX, P.W. & SMART, C.W., 1979. Veld burning in the Karoo mountains, Unpubl. Roneo. Dept Agricultural Technical Services. Director Karoo Region, Middelburg CP.

TAINTON, N.M., GROVES, RH. & NASH, RC., 1977. Time of mowing and burning of veld: short term effects on production and tiller development. Proc. Grassld. Soc. sth. Afr. 12:59-64.



Karoo Agric, Vol. 4, No 3, 1991 (28-32)