Throughout the world, increasing pressure is being exerted on producers to limit the use of agrochemicals and apply practices that will impact as little as possible on the environment.

According to Stimmann and Ferguson (1990) it seems likely that in California, the future use of various pre-emergence herbicides will be prohibited by legislation. As recently as the end of September 2001, the French government gave orders for triazines (inter alia simazine and terbutylazine) to be withdrawn from the French market, due to the occurrence of breakdown products from these substances in the soil and surface water. Although natural enemies have been considered for some time as a means of controlling certain weed species (Daniël et al., 1973; Woodhead, 1981; Cullen, Kable & Katt, 1973 and Phatak et al., 1983), it is important to realise that the gap left by such a weed will inevitably be filled by another (Putnam, 1990). Therefore cover crop cultivation, being non-specific, is the preferred practice for biological pre-emergence control of weeds.

There is increasing interest in systems where the source of nutrients is organic matter, rather than inorganic substances (Soane,1990). In the case of a cover crop, nitrogen is made available to the vines mainly in the form of complex organic compounds. In this form the nitrogen is not leached readily out of the topsoil, but made available slowly to the crop’s roots in the course of the season (Moulds, 1986). Leguminous cover crops therefore have the added advantage of fixing nitrogen and releasing it in a slow release form.

Approximately 60% of the vineyards in the Olifants River valley are situated on so-called “buitegronde”, of which Karoo soil and aeolic sands constitute the biggest percentage. A shift in emphasis with regard to soil use in this region is currently taking place, with vineyards being planted on the “buitegronde” and the silty soil being used increasingly for the cultivation of vegetables. Where soil preparation has been done on the aeolic sands, wind may cause large scale erosion. The establishment of a cover crop is therefore essential to stabilise the soil surface.

Due to a lack of sufficient scientific knowledge about the cultivation of other species as cover crops in the Olifants River valley and surrounding regions, only rye was recommended as a cover crop. Selected alternative species were therefore evaluated as cover crops in the Olifants River valley to present the producer with a variety of species for cover crop cultivation in the region. This study provides guidelines for optimum dry matter production (DMP) using these species, within the framework of viticulture.


The effect of seeding date on the DMP of the various cover crops

The trial was conducted on a sandy soil with a clay content of 0,03% and an organic matter content of 0,22% on the Nietvoorbij experimental farm at Lutzville (Table 1). The potential cover crop species were established at three seeding dates (Fig.1). Seed bed preparation with a disc harrow was done approximately six weeks before sowing. Just before sowing the seeds, the crust on the soil surface was broken with a cultivator and after sowing, a cultivator was used to cover the seeds with a thin soil layer. During seed bed preparation, superphosphate was applied at 100 kg/ha (19.5 kg P) to provide in the requirements of the broad leaf species in particular (Moulds, 1986; Wooldridge & Harris, 1987), without exceeding the fertilisation requirements of the vines. Paraggio medic and pink Seradella were established at a sowing density of 25 kg/ha, grazing vetch at 50 kg/ha and the grain species at 100 kg/ha. The grain species received 100 kg/ha LAN (28 kg N) at the two to four leaf stage as recommended by Van Huyssteen & Van Zyl (1984). Irrigation of 18 mm was applied by overhead sprinkler irrigation on a weekly basis up to 10 weeks after sowing, followed by a fortnightly irrigation of 18 mm. If rainfall occurred during this period, the irrigation quantities were adjusted accordingly. The DMP of the cover crops was determined at the end of August by measuring the above-ground growth in a 0,5 m2 site, randomly selected in each experimental plot. The samples were dried for 48 hours at 1050 C, whereafter the dry mass was determined.

[View Fig. 1]

The effect of various cover crop management practices on cover crop performance

In a follow up experiment, selected cover crop species (Fig. 2) were established in a Sauvignon Blanc/Ramsey vineyard on the same soil type at the above-mentioned seeding densities the second week of April. Irrigation of 18 mm was applied on a weekly basis up to four weeks after the seeds were sown, whereafter the same amount of irrigation was applied on a three-weekly basis. If rainfall occurred during this period, the irrigation quantities were adjusted accordingly. The DMP of the selected cover crop species was determined. The effect of various fertilisation levels (Fig. 3 & 4) on the performance of Saia oats and grazing vetch was also determined.

[View Fig. 2]
[View Fig. 3]
[View Fig. 4]

The ability of the cover crops to re-establish themselves was determined by allowing the cover crops in some of the treatments to die back naturally and produce seed. The subsequent autumn seed bed preparation was done, but no seed sown. The ability of the cover crops to re-establish themselves was determined by comparing the DMP of these treatments with the DMP of the treatments in which the same species was sown annually. In order to determine how much viable seed was present on the soil surface and in the top 0-100 mm soil layer, an area of 0,25 m2 was sampled. The soil was air-dried, whereafter the organic matter and seed were separated from the sand with a 850 micron sieve. Thereafter the cover crop seed was sorted by hand from the other fibre. The viability of the seed was determined according to the between paper method prescribed by ISTA (1999).


Although various cover crop species were evaluated in the field trials, only the six most successful species in this region are discussed for the purpose of this publication.

The effect of seeding date on cover crop performance

It is important for cover crops to be sown at the right time to ensure maximum DMP. If sufficient irrigation water is available in the post-harvest period, both the grains and broad leaf species must preferably be sown at the end of March / beginning of April (Fig. 1).

The effect of irrigation on cover crop performance

Weekly full surface irrigation or rainfall of 18 mm for a period of 10 weeks, followed by fortnightly irrigations of 18 mm, give DMPs of approximately six tons and more, if adhering to the above-mentioned seeding dates (Fig. 1). Weekly irrigation or rainfall of 18 mm for a period of four weeks after sowing, followed by at least four three-weekly irrigations or rainfall of 18 mm each per hectare, gave acceptable DMPs (Fig. 2, 1996 data). The DMPs of the 2000/2001 season were very low (Fig. 2, 2000 data). This is ascribed to the lower rainfall than that measured in the 1996 and 2001 seasons (Table 1), as well as drastic water restrictions which meant that only 8 mm per week could be applied in the first month after sowing. A minimum of 144 mm per hectare (rainfall and irrigation together) is therefore required in winter to establish cover crops successfully on the sandy soils of the Olifants River valley and surrounding regions. In dry years (2000/2001 season) rye appears to be the best choice as cover crop. The optimum management of seeding date and available irrigation water in winter in this low rainfall area will determine, in drier years especially, the extent to which producers will be able to apply cover crop management successfully in this region.

Table 1. Monthly rainfall for the period March to August, as measured at the Nietvoorbij experimental farm at Lutzville
Rainfall (mm)


The effect of fertilisation on cover crop performance

By doubling the normal quantity of fertiliser (30 kg P, 30 kg K) applied during cover crop establishment on the sandy soil, the DMP of grazing vetch was not increased (Fig.3). In the treatment in which grazing vetch did not receive any fertilisation during establishment, the DMP was not lower than that of the grazing vetch that received the recommended amount of fertiliser. It therefore seems that P at 10 mg/kg and K at 78 mg/kg, as determined in the 0-300 mm soil layer of the sandy soil, are sufficient to satisfy the demands of grazing vetch. The DMP of Saia oats did not improve significantly either when the recommended amount of fertiliser (30 kg P, 30 kg K, 42 kg N) was doubled (Fig. 4). However, where no fertiliser was given during establishment, the DMP was in most years significantly lower than that of the cover crop that received the recommended amount of fertiliser. This indicates that N fertilisation is necessary for Saia oats to produce a significant amount of dry matter. The current recommendation for grains on these sandy soils therefore appears to supply in the fertiliser need.

Ability of the cover crops to re-stablish themselves

Grazing vetch and Paraggio medic produced a fair amount of seed in 1995 and showed the ability, after just two years of sowing, to re-establish itself (Table 2). This may be ascribed to the large amount of viable seed on the soil surface and in the top 100 mm soil layer.

Table 2. The ability of grazing vetch, Paraggio medic and rye to re-establish themselves on sandy soil at Lutzville as illustrated by the dry matter production per m2 and the amount of seed available per m2 on the soil surface and in the 0 – 100 mm soil layer

Cover crop species
Dry matter (tons.ha-1) measured at the end of August
Amount of seed per m2



Grazing vetch, sown annually, chemically controlled before budbreak (BB)

Grazing vetch, sown bi-annually and allowed to die back naturally (DN)

Paraggio medic, BB

Paraggio medic, DN

Rye, BB

Rye, DN

Due to the poor performance of grazing vetch in the treatment in the 1997/98 and 1998/99 seasons, the amount of viable seed was reduced drastically. The re-establishment of vetch in the 1999/2000 season was also considerably poorer than in the 1995/96 and 1997/98 seasons (Table 2). The partial re-establishment of this species is, therefore, dependent on sufficient annual growth and accompanying annual seed production, despite a hard seed coat which allows them to survive in the soil for many years. Although the species has the ability to re-establish itself, it will still be necessary to sow, although at a lower seeding density, in order to ensure that sufficient dry matter for cover crop cultivation will be produced annually.

Paraggio medic produced more than one ton of dry matter in 1997, although the amount of viable seed on the soil surface and in the top 0-100 mm soil layer was only 6,3 % of the recommended seeding density (Table 2). The seed population increased by 41 % from 1997 to 1999 and the amount of viable seed by 259 %. Despite this, the re-establishment in the 1999/2000 season was similar to that of the 1997/98 season. It may, therefore, only be possible to adjust the seeding density of the species downwards in instances where the cover crop is left to ripen the seed.

Rye showed a limited ability to re-establish itself in 1997, with the amount of viable seed on the soil surface and in the top 0-100 mm soil layer equal to 22 % of the recommended seeding density (Table 3). Due to the poor performance of rye in the experiment during the 1998/99 season, no further seed accumulation occurred and for all practical purposes, by March 1999 the amount of viable seed was nil. Rye therefore re-establishes poorly since viable seed does not build up on the soil surface and in the top 0-100 mm soil layer over time, inter alia due to deterioration and damaging of the seed by insects during the vineyard’s growing season.

The oats species and pink Seradella did not have the ability to re-establish themselves (data not shown).


If chemical control is postponed to the middle of October, the broad leaf species have the ability to produce additional fibre (Table 3). The grains, on the other hand, have already completed their life cycle and the decomposition of fibre has started.

Table 3. The increase/decrease in dry matter production (DMP) in the period from the beginning of September (before budbreak) to the middle of October as measured in the treatments where the cover crops were left to grow after budbreak

Cover crop species

Increase/decrease in DMP

Saia oats
Grazing vetch
Parabinga medic
Paraggio medic
Pink Seradella

Despite the species being established on sandy soil for nine consecutive seasons, the DMP was still at an acceptable level (Fig. 2). It is safer, however, to apply crop rotation bi-annually, in order to prevent the accumulation of soil diseases that may be harmful to the cover crop.


The grain and broad leaf species should be sown at the end of March/beginning of April, provided sufficient irrigation water (at least 80 mm/ha) is available.

Weekly full surface irrigation of 18 mm per ha for at least four weeks, followed by three-weekly irrigation of 18 mm each, give acceptable DMPs of both the grain and broad leaf species.

100 kg LAN per hectare applied at the two to four leaf stage is essential for the grain species on these sandy soils.

The nitrogen fixing broad leaf species produce additional fibre if they are sprayed dead by mid-October, even if sown at the beginning of April.

Since none of the species show the ability to re-establish itself sufficiently, it is important that the seeding density of 100 kg/ha for the grain species, 50 kg/ha for the grazing vetch and 25 kg/ha for Paraggio medic and pienk Seradella has to be maintained, even if the cover crop is left to produce seed.


CULLEN, J.M., KABLE, P.F., & KATT, M., 1973. Epidemic spread of a rust imported for biological control. Nature (London) 244, 462-463.

Daniël, J.T., TEMPLETON, G.E., SMITH, R.J. & FOX, W.T., 1973. Biological control of northern jointvetch in rice with an endemic fungal disease. Weed Sci. 21, 303-307.

INTERNATIONAL SEED TESTING ASSOCIATION, 1999. International rules for seed testing. Seed Sci & Technol., 27, Supplement, 43-182.

MOULDS, G.A., 1986. Cover crops are very beneficial. Austr. Grapegrower & Winemaker 267, 12.

PHATAK, S.C., SUMNER, D.R., WELLS, H.D., BELL, D.K. & GLAZE, N.C., 1983. Biological control of yellow nutsedge with the indigenous rust fungus Piccinia canalicilata. Science 219, 1446-1448.

PUTNAM, A.R., 1990. Vegetable weed control with minimal herbicide inputs. HortScience 25, 155-159.

SOANE, B.D., 1990. The role of organic matter in soil compactability: A review of some practical aspects. Soil & Tillage Res. 16, 179-201.

STIMMANN, M.W. & FERGUSON, M.P., 1990. Potential pesticide use cancellations in California. California Agric. 44, 12-16.

VAN HUYSSTEEN, L. & VAN ZYL, J.L., 1984. Mulching in vineyards. Farming in South Africa E.12.

WOODHEAD, S.H., 1981. Field efficacy of Phytophthora palmuvora for control of milkweed vine. Phytopathology 71, 913-916.

WOOLDRIDGE, J. & HARRIS, R.E., 1987. Response of subterranean clover to liming and phosphate fertilization on a low potassium soil. S. Afr. J. Plant Soil 4, 145-146.

About the Authors:

J.C. Fourie1, P.J.E. Louw1 and G.A. Agenbag2

1. ARC Infruitec-Nietvoorbij, Stellenbosch
2. Department of Agronomy, University of Stellenbosch, Stellenbosch

You may like to read these:

Go Back