Response of hemp (Cannabis sativa L.) varieties to
conditions in Northwest Province, South Africa

M.C. Dippenaar, C.L.N. du Toit and M.S. Botha-Greeff
Tobacco and Cotton Research Institute, Private Bag X82075, Rustenburg 0300, South Africa

        Dippenaar, M.C., C.L.N. du Toit and M.S. Botha-Greeff 1996.  Response of hemp (Cannabis sativa L) varieties to conditions in Northwest Province, South Africa. Journal of the International Hemp Association 3(2): 63-66.
        Cannabis sativa L. is illegal in South Africa due to its cannabinoid content and its potential psychoactivity.   However, an increased interest in this bast fibre plant has developed in South Africa because hemp can provide future job opportunities.  Five improved hemp varieties from Europe with low cannabinoid content, but which are high in bast fibre, were planted at Rustenburg, 25°43'S, 27°18'E, at 1157 m elevation in the beginning and middle of November and again in mid-December 1995.  Each variety was sown at densities which would produce plant populations of 150 and 250 plants m 2 respectively.  Extreme warm and dry conditions prevailing directly after planting in early November dried the topsoil and a proper plant stand was not achieved. Rhizoctonia seedling disease destroyed the seedlings of the mid-November planting.  Even a fungicide sprayed directly after replanting in December did not improve the plant stand sufficiently to measure the effects of plant population and nitrogen levels that were included as treatments.  Flower initiation was induced in all the varieties within 34 days after sowing.  At 60 days after planting, some seeds were mature, the plants were cut and the biomass, leaf areas and fibre content of the bark and woody stalks were determined.   A fresh biomass of 9 to 13 t ha-1 and a stalk yield of 3.5 to 9.5 t ha-1 were achieved.  Plant height ranged from 96 to 123 cm.  The bast fibre content varied between 15 to 21%.  Chemical analysis of the plant material revealed that absorption of nitrogen and potassium was very high.  At the end of the season, the plants were attacked by American boll-worms, aphids and stink-bugs.  Hemp roots were infested by nematodes, even at the low natural population present.  This first trial on hemp in South Africa indicated that the cultivation of hemp is not as easy as popular articles claim.  Systematic research on the most limiting production aspects is essential to provide future South African farmers with locally tested cultivation practices.

      Hemp, as an industrial crop, is unknown to South African farmers.  However, Cannabis sativa L. has been grown in Southern Africa for medicinal purposes for centuries and during the past half-century it has also been cultivated as an illegal drug crop.   This plant species was declared illegal in South Africa in 1928 due to its high cannabinoid content and potential psychoactivity.  New improved varieties, characterized by very low levels of THC, but high in bast fibre content, have been developed in countries such as France and Hungary.  In South Africa, there is renewed interest in the special properties of this bast fibre crop as it could provide an income for the small farmer.  The optimum cultivation practices for hemp as a fibre crop, and the bio-diversity of the different local landraces present in South Africa have not yet been scientifically established.
        The purpose of this paper is to indicate problems in the cultivation of hemp as a fibre crop and the importance of research and test plot plantings before commercial production can commence.

      This project was initiated and sponsored by the Southern Africa Hemp Company and special permission was granted by the State Department of Health.  These trials were planted at the Tobacco and Cotton Research Institute at Rustenburg (25° 43'S, 27° 18'E, 1,157 m elevation).
        Five industrial varieties ('Fedora-19', 'Futura-77', 'Felina-34', 'Kompolti', and 'Secuini') were supplied by the International Hemp Association and were planted at three planting dates, namely on 1st November, 15th November and 20th December 1995.  An attempt to establish the effect of two plant population densities in combination with three nitrogen fertilizer levels formed part of the trial, but will not be discussed in this paper.  The seed was evenly sown by hand in shallow furrows spaced 10 cm apart, made by a manually-pulled furrow-maker and covered with dry soil.  To control seedling diseases on the trial area planted during December, the fungicide "Rhizolex" was sprayed at a dosage of 37 g in 15 l of water as a complete cover spray and washed in with sprinkle irrigation.
        The trial was planted on a red soil containing 46% clay (Hutton 3100 soil form).  This soil is characterized by an orthic topsoil overlying a red apedal B horizon (Soil Classification Working Group, 1991).  Before planting, 500 kg superphosphate and 100 kg K per ha (as potassium sulfate) were incorporated with a rotavator.  Two weeks post-germination, limestone, ammonium nitrate and potassium sulfate were applied as a top-dressing at 100 kg/ha each and were washed in by irrigation.
        Drip irrigation was applied at three-day intervals and soil moisture was kept close to field capacity.  Rainfall, and minimum and maximum air temperatures were measured at a weather station located 700 m from the trial site.
        A randomized block design was used with a plot size of 2 x 3 m.  These factorially arranged treatment combinations were replicated four times.
        When the first seeds started to mature, plants were cut at ground level from a 2 m 2 area in the middle of each plot and the fresh biomass was determined.  Five plants were picked at random and divided into stalks, leaves and branches, and the fresh and dry mass determined.  After the removal of the slender 10 cm tip of each of these five plants (containing undifferentiated fiber elements), subsamples, 10 cm in length, were taken from the base and end of the stalks.   The bark was separated from the woody part of these samples by hand.  The wood was oven-dried, weighed and ground.  Fibre content of both the bark and wood was determined by dissolving all soft tissue, boiling it in a 2% sodium hydroxide solution.   The fibre so obtained was washed under running water, dried and its mass determined.  Leaf area was determined with a leaf-area meter (Model Li-3000 from Li-Cor).  The dried plant parts from two replications of all the varieties were chemically analyzed for N, P, K, Ca, Mg, B, Cu, Fe, Mn and Zn.  Various insect species were observed at harvest and collected for identification.  Soil and root samples were collected and the population density of the dominant nematode genera determined.

Results and discussion
        Only the December planting produced a population which could be utilized to assess biomass yields.  The early November trial ended in total failure.  Directly after planting, the weather changed and became extremely hot and dry (average of 33.3° C).  This lasted for 16 days with only 9.5 mm of rain.  Although additional sprinkle irrigation was applied twice a week, the germinating seeds were damaged.  However, various weed species, established themselves on the trial site during this time and overgrew the few hemp plants that managed to emerge.  Good germination was obtained with the replanting of the trial in the middle of November, but damping-off seedling disease, Rhizoctonia, reduced the stand to only a few plants per plot.  Even spraying the complete trial area with a fungicide did not control the disease and all the varieties seemed to be equally susceptible.  Table 1 summarizes the number of seedlings which emerged 13 days after planting and indicates that only an 8 to 40% stand was obtained.  The standard deviation of the mean values reflects the huge variation experienced.  The number of plants per m 2 decreased further with time and, at the time of harvesting, no evaluation of the effect of plant population density on biomass production could be made.

Table 1. Seedling population and number of plants harvested at the two seeding rates. (Standard deviations in parentheses.)



Plants harvested/m2






60 (5.43)
13 (4.36)
39 (16.22)
27 (21.82)
13 (1.42)
35 (19.75)
37 (24.97)
50 (28.30)
34 (36.98)
38 (8.33)



        The reproductive growth stage was induced 34 days after planting, although day length was at the maximum of 13.6 hours at the 25°'43' S latitude.  This early flowering retarded stalk growth.  Table 2 shows that the average plant height at harvest, 62 days after planting, was only 114 cm, with a standard deviation of ±30 cm.  This high standard deviation is evidence of the variation caused by the poor plant stand.  All five varieties seemed to be very sensitive to photoperiodic induction of flowering.
        The hemp varieties produced between 10 to 13 tons/ha dry biomass (see Table 2) within the 62 days of plant growth, with a dry stalk weight of 3.5 to 9.5 t/ha.  The average fibre content of the bark and woody part of the stalks was 17.6 and 53.0% respectively.  These figures corresponded to the lower limits of similar determinations done by Höppner and Menge-Hartmann (1995).  Popular newspaper articles on hemp gave the misleading impression that hemp is easy to grow.   Business Day (1995) stated that a major advantage of this plant species ("dagga") was that it was hardy, grew in relatively poor soils and on steep slopes, and did not need horticultural or agronomic expertise.  Similarly, Danylchuk (1995) referred to an old farmer at Cold Lake, north of Edmonton in Canada who said, "You don't have to spray it, you don't have to fertilize it.  It builds up the land.  You just have to plant it and it grows fast.  It chokes the weeds out".  These comments seem to contradict our experience during this initial trial, but may be due to either the introduction of varieties unsuited for the regional daylength or the weak growth of these particular varieties.
        Vogl and Hess (1995) reported that farmers in upper Austria sowed 250 ha with the French variety 'Felina-34' under the impression that it just grew everywhere.  These farmers experienced poor germination due to soil compaction.  Where nutrients were not easily accessible, the hemp stopped developing and weeds flourished.  Low soil pH also caused poor growth and the sporadic presence of deer, birds, beetles and slugs caused local damage.
        The leaf area per plant is summarized in Table 2.  'Fedora-19' tended to have a smaller leaf area than the other varieties at the lower seeding rate.  On the other hand, the plant population density of this variety tended to be the best.

Table 2. Growth and yield parameters of hemp varieties.



'Fedora-19' 'Futura-77' 'Felina-34' 'Kompolti' 'Secuini'
101.5 119.3 104.3 117.3 129.3
Dry biomass
12.8 12.2 10.8 12.1 10.4
Stalk mass
8.23 9.49 7.31 8.61 3.56
Leaf area
(cm 2 /plant)
418 443 511 685 685
Bark fiber
15.7 20.5 14.2 20.5 17.6
Wood fiber
57.4 50.9 57.0 52.6 47.3

        An indication of the nutrient uptake by the hemp crop at Rustenburg is given in Table 3.  Nitrogen and potassium were absorbed equally in high amounts.  This emphasizes the importance of fertile soil and sufficient N and K in a fertilization program.  Berger (1969) summarized literature on the fertilization of hemp and indicated the dominant role of nitrogen and equivalent supplies of potassium in the nutrition of this plant.  All the hemp varieties were able to accumulate 0.35% P in their leaves from the red clay soil.  The microelement content in different plant portions was typical of that found in many plant species.

Table 3.  Total nutrient uptake (kg/ha) by hemp varieties from a Hutton clay soil at Rustenburg, South Africa, during the 1995/96 growing season.



    'Fedora-19' 'Futura-77' 'Felina-34' 'Kompolti' 'Secuini'  

        The trial planted in December was attacked by many insect species and the leaves and some seed were damaged.   Seven individual Lepidoptera larvae specimens were collected and reared on an artificial dry bean diet until fully grown.  All seven pupated and developed into Helicoverpa armigera adults.  A number of different Hemipteran species also attacked the plants.  Most of the plants were infested by parthenogenetic viviparous alate and apterous aphid females as well as by the various immature stages.  These aphids were tentatively identified as Aphis gossypii.  The aphids, in turn, were attacked by internally developing parasitoids and also ladybird beetle predators.   Some pentatonic stink-bugs, a Nezara species, were also collected on the plants.  The stink-bugs and aphid specimens were sent to The National Insect Collection for confirmation of identification.

        Nematodes that were found in the naturally infested field where the different C. sativa cultivars were planted, consisted of lesion nematodes (Pratylenchus spp.), spiral nematodes (Helicotylenchus and Scutellonema spp.) and the reniform nematode (Rotylenchulus spp.).   Small numbers of nematodes (Table 4) were found on the roots of most of the C. sativa cultivars, except for the cultivar 'Kompolti', where higher numbers of Pratylenchus spp. were found.  Only a small number of spiral nematodes occurred in the soil, in the immediate vicinity of the roots.

Table 4. The occurence of nematodes on different cultivars of C. sativa (per 10 g root) planted in naturally infested soil.


Pratylenchus spp.


'Fedora-19' roots
'Futura-77' roots
'Felina-34' roots
'Kompolti' roots
'Secuini' roots
Combined soil sample
(250 ml)



        According to de Meijer (1994), the root-knot nematode (Meloidogyne hapla) reproduces on C. sativa, but differences occurred among hemp accessions with regard to nematode reproduction. Meloidogyne spp. (with the dominant species being M. incognita races 2 and 4 and M. javanica) are the most important plant-parasitic nematodes in South Africa, attacking a wide spectrum of crops.  As none of these were present in the field where the trial was planted, it is necessary to evaluate the resistance of the hemp cultivars to the local Meloidogyne species.

      As for any other crop, well-adapted cultivars, high quality seed and favorable conditions during germination are required to establish a good plant population.
        The five varieties from Europe and Hungary are very sensitive to day length and this characteristic limits plant development and fibre production under the conditions which prevail at Rustenburg.
        Chemical analysis of the hemp plants confirms that the nitrogen and potassium nutrient requirements are high.  Therefore, a fertile soil of suitable pH is preferred.  Growth of hemp is further determined by sufficient water and enough sunshine.  Arable land in the Republic of South Africa lacks some of these growth factors.  Supplementation with required nutrients, amelioration of the soil acidity, and irrigation will be necessary for production of this crop.
        Northern European hemp varieties with low cannabinoid content seem to be subject to aphid and boll-worm damage.  Promising hemp varieties should be evaluated for resistance to local insects and Meloidogyne species.   Before hemp can be grown economically as a fibre crop in the Republic of South Africa, research and evaluation of cultivation and pest management practices are essential to place hemp production on a sound scientific basis.