Last update: August 18, 2011 03:58:02 PM E-mail Print




CH Donaldson

Grootfontein College of Agriculture



The choice of plant species and ecotypes or varieties is one of the most important considerations in growing and managing dryland pastures or reseeding poor and denuded veld. The success of such practices will not only depend on the use of climatically and edaphically adapted species, but also on factors such as the amount and distribution of the rainfall, temperature, seed viability, and the use of correct sowing techniques.

Despite the harsh and dry conditions of most Karoo areas, there are indications that the plant production of these areas can be improved by the growing of grasses and shrubs as dryland pastures or by reseeding denuded areas with these plants.

The biological potential (during the establishment phase) of a number of grasses and shrubs was assessed under dryland conditions at the Grootfontein College of Agriculture. This paper is concerned with the results of the growth rate, winter survival, seed and dry matter (DM) production of 15 grasses and 10 shrubs during the establishment period and under conditions of a well-prepared seedbed and above average rainfall.



The experiment was conducted on a sandy loam soil (shorrocks (-) series) with good levels of phosphorus (36 ppm), potassium {370 ppm) and a soil pH (H2O 1:I) of 8,4.

The experimental area was ploughed and rendered weed-free before sowing. The seeds of 15 grasses and 10 shrubs were sown in single rows, each 3m long and 1,45m between rows, in a randomized blocks design with three replications. A 1m patch was provided at either end of the rows. The grasses and shrubs were grown in two separate blocks.

The wheel tracks of a 30 Kw farm tractor were used to mark the rows on a fine weed-free seedbed. The seed was thereafter sown by hand in the tractor wheel-tracks on dry soil on 26 January 1987. The seed was pressed into the soil by going over the rows for a second time with the tractor wheels.

The quantity and viability of the seed of the grasses Anthephora pubescens. Erharta calycina, Cenchrus ciliaris (Jansenville type) and Themeda triandra were less favourable than the other plant species.

The species planted and the source of the seed used were as follows:



The grass and shrub seedlings were thinned out on 13 May 1987 to spacings of 4 to 8 cm and 10 to 20 cm between plants, respectively. The density of plants in each row was subjectively estimated according to the spacing of the plants and denseness of foliage on 9 April 1987 .

The intensity of frost damage to grasses and shrubs was rated on 29 August 1987 and 16 August 1988 according to the following scale:

0 = Leaves green with very little frost damage.

1 = Less than 20 % of growth dead.

2 = 20 % - 50 % of growth dead.

3 = 50 % - 80 % of growth dead.

4 = More than 80 % of growth dead.

5 = Total root and top growth kill.


Seed heads were harvested on 12 January 1988. Dry matter yields of two plants per row per replication were determined from plants cut at ground level on 13 May 1987, 16 November 1987 and 18 February 1988.

DM yields of grasses and shrubs were also determined for each 3 m row on 18 February 1988 and 16 August 1988 - all grasses were cut at a height of 5 cm and shrubs were cut at a height of 10 cm. DM yields and seed data were subjected to analysis of variance.



1.   Rainfall and Temperatures

A summary of the climatological data over the study period January 1987 to August 1988 is given in Table 1.


It is evident from the climatological data (Table 1) that the climate of the area is characterized by a low rainfall (363.1 mm per annum), moderate to high day temperatures (average maximum for January = 30,3°C), low minimum temperatures in winter (average for July = -O,4°C) and frosts over the period April to October. During the study period the total rainfall for 1987 and 1988 was average and above average, respectively. Higher than average amounts of rain were received during June, September and November of 1987 and February, March and April of 1988. Hot and dry conditions were experienced during January and December 1987 and January 1988. The mean maximum and minimum temperatures were above average for most months (Table 1).


2.   Establishment, frost damage and winter survival

Visual estimates of the stand density of young plants on 9 April 1987 and frost damage ratings on 29 August 1987 and 16 August 1988 are given in Table 2.


Good conditions for seed germination were experienced during the establishment period. The first rains of 1,0 mm fell on 8 th February 1987. This was followed up by 10,0 mm, 2,9 mm, 7,1 mm and 22,2 mm on the 9 th, 10th, 11 th and 12 th of February respectively. Good follow-up rains during March 1987 (Table 1) ensured good stands for most plant species by the beginning of April (Table 2). The moderate stands obtained for the grasses Themeda triandra and Erharta calycina (Table 2) could largely be ascribed to the poor quality of seed used in this trial. Despite good seed and good germination for all shrubs, poor stands were obtained for Pollicha campestris, Eriocephalus africanus, and Herpicium alienatum and, to a lesser extent, Atriplex halimus and Monochlamys albicans. It is therefore likely that these shrubs may either require cooler conditions during the germination phase or other sowing procedures for good establishment.

From the data in Table 2 it is clear that only one grass, Cenchrus ciliaris (Jansenville), and two shrubs, Eriocephalus africanus and Hirpicium alienatum failed to survive the first winter, whilst all Monochlamys plants and about forty per cent of the Osteospermums died during the second winter. It was also noted that approximately fifty per cent of the seedlings of Cenchrus ciliaris (Molopo) succumbed to the frosts of the first winter. Approximately thirty per cent of the young plants of Osteospermum sinuatum and a few seedlings of Digitaria eriantha (Smutsii) and Heteropogon contortus also died during the first winter period. Observations made on O. sinuatum plants that were sown, in adjacent plots, during February 1986 and 1988, revealed that more than ninety per cent of the young plants died during the colder winter of 1986 and 1988. With the exception of the three introduced grass species, Anthephora pubescens, Cenchrus ciliaris (Jansenville and Molopo) and Digitaria eriantha (Smutsii), and the shrubs, Hirpicium alienatum, Atriplex nummularia, A. canescens and E. africanus, all the species listed in Table 2 are indigenous to the veld of the Middelburg district. Three of the shrub species (0. sinuatum, P. campestris and M. albicans) that were susceptible to frost damage (Table 2) normally grow in isolated numbers in protected areas such as between stones, on the northern side of hills or underneath taller shrubs.

A further noteworthy feature of the data in Table 2 is the very high tolerance of the grasses, Themeda triandra and Tetrachne dregei, to frost. The high tolerance to frost damage of the temperate grass species, Erharta calycina, is however to be expected.


3. Seed production and flowering habits

Seed yields and notes on the flowering habits of the grasses and shrubs are given in Table 3. S. microphylla produced significantly (p ≤ 0,01) more seed than the other species (Table 3). Other prolific seed producers were E. obtusa, O. sinuatum and F. africana, these plants yielded significantly (p ≤ 0,01) more seed than the other species, with the exception of S. microphylla. Fairly good seed yields were also obtained from L. pauciflora, 7: dregei, E. curvula var. conferta, S. fimbriatus and A. pubescens. C. ciliaris (Molopo), H. contortus and 7: triandra yielded significantly lower amounts of seed than all other species (Table 3).


From the phenological observations and seed yield data of the grasses in Table 3 it would appear that c. ciliaris (Molopo), H. contortus and T triandra are either slow or poor seeders. It is likely, however, that flower producing tillers (on 16 August 1988) could have been removed by the clipping treatment of 18 February 1988. The shrubs that did not seed probably flower at a more advanced age or stage of growth. The results of this trial indicate that most of the plants that set seed produced large amounts of seed in their first year, suggesting that large quantities of seed can be produced off relatively small areas.


4. Dry matter yields of single plants

The mean DM yields of the primary growth of single plants of each plant species on 13 May 1987, 16 November 1987 and 18 February 1988 as well as rainfall and temperature data are given in Table 4.

For period 1, the seedlings of C. ciliaris (Molopo), c: ciliaris (Jansenville) and A. pubescens produced significantly (p ≤ 0,01) more DM than the yields of all the other grasses (Table 4). Further important features of the data of this period were the extremely slow growth rate and low yield of E. calycina and the high yields obtained from the shrub seedlings. S. microphylla was the most productive shrub, yielding significantly (p ≤ 0,01) more DM than the yields of all other shrub seedlings.

Good rainfalls and low average temperatures during period 2 resulted in a good growth and production of E. calcimine (Table 4). The high yield of this grass was greater than the yield of all the other grasses at the 1 % level of significance. E. curvula var. conferta and P. stapfianum also responded well to the climatic conditions encountered during this period by producing significantly (p ≤ 0,01) more DM than all the grasses, excluding E. calycina. High growth rates were obtained from all the shrubs during this period. A. nummularia out yielded (p ≤ 0,01) all other shrubs, whilst L. pauciflora produced the lowest yields (Table 4).


With the exception of E. obtusa, C. plurinodes, T triandra and L. pauciflora thigh growth rates and DM yields were obtained for all species during period 3. The good rains of period 2,followed by favourable precipitations and good temperatures for plant growth during period 3, were probably responsible for the increased levels of DM production during this period (Table 4). P. stapfianum out yielded (p ≤ 0,01) all the grass species. 7: dregei and E. curvula var. conferta produced significantly (p ≤ 0,01) more DM than all the other grasses except P. stapfianum. The herbage yields of C plurinodes, T triandra and E. obtusa were significantly (p ≤ 0,01) lower than all grass species (Table 4). The DM production and growth rate of the shrubs, with the exception of L. pauciflora during period 3, were on the average superior to the yields of most grasses. A. nummularia out yielded (p ≤ 0,01) all shrubs, whilst very good yields were also obtained from A. canescens (Table 4). The data in Table 4 also show that the growth rate of the shrubs was generally much higher than that of the grasses for all three periods.


5.  Dry matter production of established stands

The primary growth yields of grasses and shrubs, representing twelve months growth and production from germination (February 1987) to the first cut (18 February 1988) are given in Table 5. Also presented in Table 5 are the DM regrowth yields for the period 18 February 1988 to 16 August 1988.

The DM yield data, statistical comparisons and rankings in Table 5 show that very high yields were obtained from both cuts for the grasses, T dregei, P. stapfianum, E. curvula var. conferta and A. pubescens. Fairly good yields were also obtained form C. ciliaris (Molopo), D. eriantha, F: africana, S. [fimbriatus, E. calycina and D. eriantha (Smutsii), whilst poor yields were obtained from the climax grasses (H. contortus, C plurinodes and Themeda triandra) and the perennial pioneer, E. obtusa (Table 5). The grass yields of the 18 February 1988 cut were very similar to those of the 16 August 1988 cut for each grass.


A. nummularia yielded more than the other shrubs. As a group the shrubs performed much better than the grasses in the first cut (18 February 1988), whilst the comparatively lower regrowth yields of the shrubs for the second cut (16 August 1988) confirm the generally known fact that shrubs need a longer period to recover than grasses after having been subjected to drastic clipping or grazing treatments.



Of the 15 grasses and 10 shrubs tested, one grass (Cenchrus ciliaris, Jansenville), and three shrubs (Eriocephalus africanus, Hirpicium alienatum and Monochlamys albicans) failed to establish, whilst a poor stand was obtained from Pollichia camprestris. Low temperatures and frosts were responsible for the death of the seedlings of this grass as well as the three shrub species. Osteospermum sinuatum and Cenchrus ciliaris (Molopo) seedlings were also very susceptible to frost damage.

The favourable temperature and moisture conditions that existed during at least the first six days during the germination period, suggest that the growing conditions were favourable to assess the establishment potential of the sown species. It was under these conditions that good stands of seedlings were obtained for all plants except Eriocephalus africanus, Hirpicium alienatum and Pollicha campestris

The ability of all the seedlings to survive high temperatures or long dry periods could not however be put to the test, since these climatic variables, which are so typical of semi-arid areas, did not occur during the experimental period. Finally it can be concluded that on the basis of ease of establishment, growth rate, production and frost tolerance, the following plant species performed the best during the establishment period of this trial: E. curvuIa var. conferta, P. stapfianum, T dregei, A. canescens, A. nummuIaria and S. microphylIa.



My thanks go to Mr A Barnhoorn for his assistance in some of the fieldwork.



Karoo Agric 4 (2) 3-10