5.

Examples of projects that were successfully completed or with which considerable progress was made (Annexures I and II)

5.1

The isolation and characterisation of the gene complex regulating the response of soybean (Glycine max (l.) merr.) to water deprivation, salinity stress and temperature extremes

J van Staden and WA Cress

The increasing importance of technology transfer has already been mentioned. As follow-up to the PRF's existing guidelines for drafting management summaries, researchers are expected to present their summaries in the format of a scientific article. The Board decided that these management summaries should be referred to the Marketing Committee for recommendations to the Board and to determine how the report should be further distributed. This will ensure that all research projects reach the target market for which they are intended.

Low-temperature stress subtraction libraries were constructed in a Bluescript vector with the two-step PCR amplified cDNAs using subtractive hybridisation. One insert cs18 was obtained and the sequence analysis of insert cs18 revealed that the insert cDNA had a 76% homology with the sequences of the corresponding portion of glucose dehydrogenise from Thermoplasma acidophilum and 62,0% homology with a genomic DNA of Arabidopsis. Four clones, csl8-13, csl8-14, cs18-15 and cs18-16 from low-temperature stress soybean root conventional cDNA library have been confirmed to have inserts that could hybridise to the csl8 insert. One cDNA with an Xba 1 and Xho 1 fragment of approximately 3,500 bp in length corresponded to the insert cs18, which probably encodes for glucose dehydrogenase, was obtained. Northern blot analysis indicated that cs18 mRNA was highly expressed in soybean root but moderately expressed in leaves under low temperature.

Changes in the nuclei of meristematic root cells in response to severe salinity were studied. Roots are in direct contact with the surrounding solution. Thus, they are the first to encounter the saline medium and are potentially the first site of damage or line of defence under salt stress. Nuclear deformation or degradation in the soybean root meristern with 150 m1\4 or higher NaCl led to sequential cell degradation, cell death and cessation of plant growth. However, this study indicates that an increase in CaC1₂ concentration up to 5mm could partially prevent salt injury to the cells.

Tissue culture is an excellent tool for elucidating the correlation between plant organisational levels and salt tolerance because of the possibility it offers for studying the physiology of intact plantlets together with that of organs and single cells using homogenous plant material under uniform environmental conditions. One NaCl-tolerant cell line (R100) was isolated during this study. The R100 callus cell fine was significantly more tolerant to salt than the salt-sensitive fine (S100) during exposure to salt stress. Salt tolerance in this culture was characterised by an altered growth behaviour, reduced cell volume and relative water content, and accumulation of Na⁺, C1-, K⁺, proline and sugars when grown under salt stress and with its subsequent relief. The selection of this salt tolerant cell fine has potential for contributing new genetic variability to plant breeders.

Sugars are not only important energy sources and structural components in plants, they are also central regulatory molecules controlling physiology, metabolism, cell cycle, development and gene expression in plants. The concentrations of glucose and fructose increased during salt stress and after relieving salt stress, at a rate closely corresponding to the increase in relative water content. Their accumulation was the earliest response detected during the removing of salt stress indicating that glucose and fructose may play important roles during salt stress.

5.2

The effect of enhanced proline synthesis in soybean with special reference to drought stress, comparing antisense and sense technology

JA de Ronde

Control Ibis plants were compared with antisense and sense transgenic plants containing the P5CR gene, through evaluation of different physio­logical and phenotypic techniques. An experiment on mild heat stress demonstrated that although the free proline analysis were not able to detect significant differences between the transgenics, some changes in the photosynthetic functions as measured through chlorophyll fluorescence were already detectable.

Chlorophyll fluorescence transients (JIP test) were thus used to quantify the response of the different transgenics towards their reaction to heat stress. It was proved that the antisense plants were more sensitive than the control plants and the sense plants more tolerant.

The antisense plants responded to a period of heat and drought stress by reducing their photosynthetic efficiency, increasing proline slightly and reducing NADP levels as their defence mechanisms become activated. The defence mechanisms were, however, not able to prevent the plants from incurring irreversible damage as proline degradation and photo inhibition occurred even after the stress was removed. The NADP levels drop lower than before the stress. On the other hand, the sense plants seem to increase the proline levels, maintaining a high photosynthetic activity and increase the NADP levels as their defence system become activated. With dehydration of the sense plants, the proline returned to pre-stress levels, but the NADP levels increased. The sense plants performed as tolerant to the combination stress in all the physiological tests and the antisense plants as sensitive, when compared to the control plants.

The sense and antisense transgenic lines subjected to the combination stress in a pot trial resulted in a 100% recovery of the sense plants, 65% in the control plants and 16% in the antisense plants. This indicated that the sense plants performed more tolerantly than the control plants to a combination of drought and heat treatment. An early drought screening in a wooden box displayed a 2 times higher recovery rate of the sense plants than the control plants. All these results demonstrate that plants with a higher copy number of the proline gene are more drought tolerant than plants with the normal copy of this gene.

5.3

Validation of genetic markers linked to resistance for Meloidogyne javanica and selection characteristics of economical significance in soybean

CMS Mienie

Several soybean growth characteristics require special attention for soybeans to be adapted to the South African environment. In order to be competitive on the world market, locally grown soybeans need to be adapted to typical South African environmental factors such as low humidity or high shattering risk, low night temperatures, diverse growing conditions impacting on plant structure and seed chemical and physical quality, local pests, local disease pressures, etc. Soybean cultivars introduced from the USA or elsewhere cannot compete with cultivars selected locally. Some characteristics however require special logistics in order to breed for. Using laboratory procedures to screen for genetic markers which correspond to the desired characteristic will cut down on complicated and labour intensive plant screening methods and provide a more reliable and cost effective way to identify progeny with the desired trait.

A set of molecular markers was developed in a previous study, which is now being tested for consistency over generations, and is also being tested against traditionally screened material in the existing germ plasm bank. F3 progeny tests revealed stable inheritance of the markers. A correlation of 80% between greenhouse tests and molecular marker screening was obtained with 35 soybean cultivars. Only six genotypes were misclassified as resistant and one as susceptible. This can be due to the difference in genetic background between the cultivars and the mapping population, and means that the resistance probably originated from a different source with a different gene/s. It is recommended that an extended selection of genotypes further be screened with the molecular markers to identify potential breeding material with resistance to M javanica.

5.4

Generation advancement of soybean breeding

CMS Mienie

Traditional plant breeding advances soybean lines through six to seven generations to obtain homozygosity. Plant breeders are in need of a low cost technique to shorten the generation time in order to produce new cultivars more rapidly and more cost effectively. Due to the photoperiod sensitivity of soybean, the vegetative cycle can be shortened with manipulation of day length. The reproductive cycle represents more than 50% of the total fife cycle and can be shortened if the seed-filling and ripening period can be bypassed. This study was conducted to develop a cost effective procedure to shorten the generation time, with the reproductive phase as the main objective. A two-pronged approach was followed to explore the possibility of using in vitro culture of alternatively harvesting the immature seed. Pods were harvested for tissue culture at different time intervals after flowering date and the embryos isolated and cultured on an artificial medium. These results were compared to seed of the same age, which was dried in intact pods and planted in the greenhouse. Preliminary results indicate that direct planting of immature seed would be much more efficient than tissue culture techniques. Seed sampled as early as 18 days after flowering exhibited 80% germination after slow drying. Studies are under way to shorten the time of desiccation.

5.5

Response of soybean to temperature stress and its interaction with light intensity evaluated by photosystem 11 function, photosynthetic gas exchange and xanthophyll cycle activity

GHJ Krüger and PDR van Heerden

For the first time, results were presented which demonstrated the existence of pronounced and consistent genotypic differences in the physio­logical and biochemical response of two chilling tolerant soybean cultivars, Maple Arrow and Fiskeby V, to dark chilling. In most of the processes studied, Maple Arrow was characterised by a less severe or more delayed response than Fiskeby V.

The superior dark chilling tolerance of Maple Arrow compared to Fiskeby V was even further emphasised during exposure to simultaneously induced dark chilling and drought stress. It is hypothesised that the mechanism conveying superior dark chilling tolerance in Maple Arrow is also, at least in part responsible for conveying superior tolerance to simultaneously induced dark chilling and drought stress.

In conclusion, the large volume of evidence acquired during this project strongly implies a more dark chilling tolerant physiological and biochemical make-up in Maple Arrow compared to Fiskeby V. In this investigation two cultivars, which are both regarded as chilling tolerant, were used, emphasising the sensitivity of our experimental approach and methodology. We are of the opinion that the sensitivity of our experimental approach in revealing intra-specific differences should be of great value in future screening and genetic transformation programmes aimed at increasing stress tolerance in soybean. The prospects for further research are exciting, since we will extend our investigations to a highly chilling sensitive cultivar (Java) of tropical origin, which should reveal genotypic differences even more clearly. In addition, experiments with Glycine soya, the wild ancestor of cultivated soybean could be very important in revealing crucial stress tolerance traits that were lost through breeding for high-yielding cultivars. The re-introduction of lost stress tolerance traits is a possibility that could increase the stress tolerance of soybean in the future.

5.6

National soybean cultivar trials

MA Smit and GP de Beer

The soybean (Glycine max (l) Merr.) is one of the most widely grown crops in the world today. However, individual cultivars are generally very restricted in adaptability due to sensitivity to photoperiod. Genotypic interaction with environmental factors such as temperature, water use, soil criteria and cultivation practices further necessitate on-site cultivar evaluation.

The cropping of any given soybean cultivar, without knowledge of its tested performance in that particular environment, may well be an absolute disaster; even though such a cultivar may do well elsewhere. Cultivar evaluation programmes are therefore common practice in all soybean producing countries. A standardised national soybean cultivar evaluation programme was initiated by the ARC Grain Crops Institute in South Africa in 1980 under government directive in order to stimulate and support the local soybean industry. As reported, the National Soybean Cultivar Trials (project O10/01) were planted at 54 localities this past growing season as a uniform test. The objectives of the project are:

• To test all new entries and advise the registrar of varieties on adaptability and features of agronomic importance;
• To characterise all entries in terms, of important agronomic criteria;
• To demonstrate the important cultivar photothermal interaction so typical of soybeans and characterise each entry in terms of area of adaptation;
• To record long term risk under standardised cultivation methodology;
• To introduce the crop to potential cropping areas.

A total of 36 cultivars were evaluated on a national basis and an additional 24 as a phase I test at nine locations. Trials were requested by PDA research stations, Seed Companies and Cooperatives and were also placed in areas where information with regard to relative cultivar performance is lacking. A comprehensive dataset will entail phenological. data such as dates on emergence, flowering, physiological and harvest maturity, maximum plant height and minimum pod clearance, lodging, shattering and green stem; relative disease or pest susceptibility; and seed characteristics such as yield, physical and chemical quality as well as soil and weather data. At chosen sites the effect of planting date is also recorded. All acquired data are captured in a computerised data basis. Data from individual trials were subjected to analyses of variance and combined yield data to the additive main effects and multiplicative interaction (AMMI) statistical method. At the cultivar evaluation committee meeting held in September 2000, 6 entries were cleared for registration on the variety list. The 99/00 report on the National Soybean Cultivar Trials was also cleared for publication. The mean seed yield for all localities tested was 2 462 kg/ha (2 448 kg/ha for 1998 and 2 200 kg/ha for 1999). A pamphlet detailing performance and cultivar recommendations was released for publication and commercial use. The objective test results are used widely by the soybean industry as a source of information on cultivar adaptation, yield performance, biological yield potential and seed chemical quality. Over the past twelve years the National Soybean Cultivar Trials clearly demonstrated the genotype by latitude interaction for the South African context. The increase in local production from some 20 000 t in the early 80's to 200 000 t in 1998/99 would not have been possible had it not been for the National Soybean Cultivar Evaluation Programme.

5.7

Soybean row width and plant population trials in different production areas

MA Smit and GP de Beer

The aim of this project is to demonstrate the agronomic acceptability of recommended soybean cultivars for different production areas and to evaluate the effect of row width and plant population on crop growth and production risk. Trials were grown successfully at five sites and the results of three sites proved statistically reliable. First year results have indicated that an interaction exists between cultivar, row width and environment which makes it difficult to categorise cultivars into typically narrow or wide row type cultivars. These trials will be expanded to more sites in future and will also be used to evaluate seed inoculums using a split plot experimental design.

5.8

The development of soybean cultivars with improved adaptability and seed quality

AJ de Lange

This breeding programme uses the single seed descent (SSD) selection method, which is the most common selection method for soybean breeding. The yield testing system utilised in the programme entails, check row trials, elite trials and phase trials. The single plant selections (2 400) of the previous season yielded 346 single rows that were selected to be included in the check row trial. Eighty-three breeding lines have been included in the elite trials that will be planted at six localities in the coming season. Four breeding lines with superior quality were offered for registration in the past season and one in the new season. The yield trials indicated that the breeding lines compete favourably with the best cultivars on the market. The best breeding lines yielded between 4 666 kg/ha and 5 987 kg/ha, which were on average 12,7 percent better than the best control cultivars. The agronomic traits of these breeding lines are also acceptable.

5.9

Super soya competition for KwaZulu-Natal

C Havenga

Northern KwaZulu-Natal

The Super Soya Competition in Northern KwaZulu-Natal (N-KZN) is now in its eleventh year and continues to attract interest. The number of participants this season had to be limited to a total of thirty-two. The extension wards that participated this season were: Vryheid, Newcastle, Dundee, Bergville and Estcourt.

As a result of the dry conditions during October and the greater part of November, especially in the northern part of KZN, planting dates were relatively late in these areas.

The average yield of participants this season was 2,93 ton/ha, which is slightly better than the 2,69 ton/ha of the previous season. Yields varied between 1,55 ton/ha and 4,6 ton/ha.

Protein quality this year is the best to date, with an average analysis of 42,72% on a dry matter basis ranging from 37,8% to 45,9% dry matter.

From the viewpoint of Super Soya Competition participants, it appears that protein content of soybeans is no longer a serious problem.

Southern KwaZulu-Natal

The Super Soya Competition was initiated in Northern KwaZulu-Natal eleven years ago. This season saw the competition expand for the first time to Southern KwaZulu-Natal (S-KZN).

This first season the competition in Southern KwaZulu-Natal had 14 entries, mostly from the Midlands. We hope to expand the competition to 40 farmers over the next few years.

The average yield of participants in S-KZN was 3,57 ton/ha, ranging from 2,83 ton/ha to 4,56 ton/ha.

Soybean seed protein quality has been excellent this season. Protein content calculated on a dry matter basis varied between 37,5% and 44,3% with an average of 42,3%.

The opportunity that this project creates for farmers to exchange knowledge and experience contributes towards the noticeable improvement in the cultivation of soybeans in KwaZulu-Natal.

5.10

Super soya competition for Mpumalanga

J Middel

The competition for the Mpumalanga farmers was launched during 1996 and participation over the four-year period increased from 11 to 31 participants.

The abnormal dry weather conditions that prevailed during February, March and April caused a substantial variation in yields, resulting in sub-optimal average yields.

The mean yield recorded for the Mpumalanga entries was 2,17 ton/ha.

Mr J. van Niekerk of Amersfoort achieved the top yield of 3,94 tons/ha as well as the highest gross margin of R3 857,96/ha.

5.11

Biological nitrogen fixation of soybean: inoculation problems

JF Bloem¹ and A Booysen²

Bradyrhizobium japonicum strains, with the ability to nodulate soybean, are not indigenous to South African soils. The establishment of effective nodulation and nitrogen fixation is essential for profitable soybean production. The cost of inoculation is minimal compared to the application of mineral nitrogen. The protein content of locally produced soybean has raised questions in the past. This project was initiated. in collaboration with the ARC-GCI to determine the reasons for variable and disappointing utilisation of the symbiosis in commercial plantings.

Surveys at eighteen plantings in four production areas revealed that some inoculants used by farmers were below acceptable standards. The number of rhizobia per seed and per ml of liquid inoculant was nevertheless sufficient to realise proper nodulation.

Recommended inoculation procedures were followed, but inoculants are still kept in close proximity to herbicides, insecticides, etc. Poor nodulation was mainly attributed to high residual N and low P levels. Residual herbicides such as alachlor, atrazine, metribuzin and chloritnuron were found in commercial soybean plantings, and the effect on yield, protein and oil content is currently being investigated. Poor yields could be attributed to ineffective weed control and drought during the flowering stage.

Soybean production has increased significantly and farmers must be informed continuously on the correct handling and application of soybean inoculants. The established bradyrhizobia population in soybean fields varied between 0 and more than 70 000 kg soil. Soybeans must be inoculated every year, as regular planting of soybeans will not guarantee sufficient rhizobial numbers in the soil (especially under low pH conditions).

5.12

Super soya production competition

MA Smit

With the focus on crop specialist support for soybean production in targeted areas, the super soya project is very much a producer-orientated technology transfer project. For the year 2000 a total of 8 short courses on soybean production technology were hosted, 7 field days addressed, 14 days spent visiting production units, an exhibition and information centre staged at the annual NAMPO agricultural show and a high profile super soya event held at Ermelo on 21 August 2000. Guest speakers at the Ermelo super soya day covered the pertinent topics of long-term weather forecasting, soil acidity and profitability prospects for soybean production on the Highveld. The event was attended by some 160 people and received good news coverage. Two TV talk shows concerning soybean production were aired on the GSA channel in February and May respectively and a lecture on soybean production was given to third year BSc Agric students at the University of Pretoria.

5.13

Evaluation and standardisation of analytical methods for determining the degree of soya processing and the estimation of the available lysine content in soya

D Palic

With regard to full fat soya (FFS) processing, the major concern is over or under processing, as both reduced amino acid availability.

The aim of this project is:

• To evaluate and standardise a number of existing analytical procedures for determining the effects of Full-fat soybean processing.
• To derive regression equations for predicting the response of animals fed on processed soy beans, using results obtained by standardised laboratory procedures and values obtained in in vivo trials.

In ongoing Phase 1 of the project, full fat soya was processed by dry-extrusion in the temperature range 115-177ºC.

Evaluation of currently available laboratory methods has shown that the Urease activity index (UDP), Protein solubility in KOR (Araba and. Dale, 1990 as modified by Palic and Potgieter, 2 000) and Nitrogen solubility index (NSI) showed a clear correlation with. the temperature of processing.

Trials with chickens fed seven dry-extruded FFS showed that only the sample processed at 115ºC was under processed, while in the range from 125-177ºC there were no significant differences between samples as measured by body weight gain, cumulative feed intake, feed conversion ratio and amino acid digestibility.

The in situ nylon bag and mobile bag techniques were used to estimate ruminal protein degradability and rumen undegradable protein (UDP) digestibility of the same dry-extruded FFS samples. The results showed that the heat treatment decreased ruminal protein degradability (increased UDP) from 56% (at 115ºC) to 21% (at 177ºC) without negatively affecting UDP digestibility.

The growth and amino acid digestibility trials with chickens will be repeated to confirm the above-mentioned findings.

5.14

Cultivar evaluation trials of oil and protein seeds in the winter rainfall region (lupin)

DJ Hanekom

The Chief Directorate Agriculture: Western Cape was again, as with the previous two seasons ('98 and '99), responsible for the lupin cultivar evaluation project during the 2000 season. Twenty-two cultivars were tested at 12 sites – 6 in the Southern Cape and 6 in the Western Cape region. An additional planting date was planned at each of the two experimental farms in the two regions, i.e. Tygerhoek and Langgewens. Only the trial at Tygerhoek could eventually be carried out as the only significant rainfall in the Swartland came late in May, which made an early planting date impossible. Two new yellow lupin cultivars, as well as a narrow leaf cultivar obtained from Poland, were included in the programme. Data pertaining to the flowering date and yield were collected and analysed for statistical differences.

The trial at Swellendam, as well as the second planting date at Tygerhoek in the Southern Cape region, had to be abandoned due to bad germination, low soil moisture conditions and a problem with weeds. The average yield of lupin cultivars at the different sites in this region ranged from 946,1 kg/ha (Tygerhoek, first planting date) to 1717,9 kg/ha (George). Mean yields of cultivars for all the sites ranged from 594,8 kg/ha (Markiz) to 1 789,1 kg/ha (Gail). The best average yield for a narrow leaf cultivar was registered for Eureka (1545,1 kg/ha).

The highest yield by a lupin cultivar was recorded by Esta (2 405,4 kg/ha) at the Napier site.

Severe moisture stress conditions prevailed for a significant period during the season at most of the sites in the Swartland region. The above average rainfall that was measured during July at all the sites did, however, lead to good recovery of lupins at most of the sites. The late planting date, together with a lengthy period of moisture stress after planting, resulted in poor establishment at the Malmesbury site and eventually in the trial being abandoned. Bad spots, probably caused by poor drainage or high salinity in the trials at Porterville and Eendekuil caused visible reduction in vigour, resulting in the data of these trials not being included in the final analysis.

The average yields of lupin cultivars at the different sites in this region ranged from 1630,3 kg/ha (Langgewens) to 1944,9 kg/ha (Philadelphia). Mean yields of cultivars for all the sites ranged from 710,8 kg/ha (Markiz) to 2 733,6kg/ha (EI6). The best average yield for a narrow leaf cultivar was registered for Tanjil (1 815,6 kg/ha). The highest yield by a 1upin cultivar was recorded by E16 (3 045,7 kg/ha) at the Philadelphia site.

Yield assurance data for the 2000 season showed that the cultivars Astra, Kiev, Vladimir and Cedara 6150 gave good yield assurance at a wide spectrum of potentials ranging from 0,5 to 3,0 tons/ha. Gail and Wonga gave good yield assurance at the lower potentials, while Kali and KS-3 gave good assurance at the higher yield potentials.

5.15

Development of production technology for sweet white lupins in the summer rainfall areas

JAM van der Mey

Since the Protein Research Trust has appointed lupin promotion officers in the Free State, KwaZulu-Natal, Mpumalanga and the Western Cape regions, this report does not cover those areas. The ARC-GCI team has stimulated the planting of lupins, in the North West and Northern Provinces, which are new areas where lupins can be grown with success under irrigation during winter. This year approximately 75 ha were planted to lupins in the Viljoenskroon, Vryburg, Ventersdorp, Brits and Swartruggens areas. It is expected that this area will increase to about 150 ha during the 2000/2001 season. If what the producers consider fair prices is obtained, lupin cultivation could increase dramatically in this area. The sweet white lupin cultivar Hantie yielded between 3 and 6 t/ha in semi-commercial blocks, while the recently introduced Australian narrow-leaved cultivars, Wonga and Merrit, with better resistance to anthracnose, yielded about 2,5 t/ha. Protein and oil content of the narrow-leaved lupins is generally lower than that of sweet white lupins. The ARC-GCI team also produced seed for the National Lupin Cultivar Trials, which were planted in the Western and Southern Cape, as well as in the summer rainfall areas during the winter. The trial results were also analysed by the ARC-GCI. This season it was possible to produce a combined statistical analysis of the 1998 and 1999 seasons. A total of 24 trials were acceptable, with either the coefficient of variation (CV) and/or intra-class correlation (tn), the entry repeatability (t), and the ratio of t over the standard error of t, within acceptable ranges. The best cultivars over all localities, exceeding 1,8 t/ha, were: CED 6150 (1,7 t/ha), Vladimir (1,84 t/ha), Ronell (1,83 t/ha) and KS-3 (1,91 t/ha). The same cultivars performed best in the Western Cape, exceeding 2 t/ha: CED 6150 (2,2 t/ha), KS-3 (2,02 t/ha) and Vladimir and Ronell (2,01 t/ha). In the Southern Cape the three top producing cultivars were: Eureka (2,0 t/ha), Helderberg (1,91 t/ha) and Hantie (1,90 t/ha). It should be noted here that both Eureka and Helderberg are narrow-leaved sweet lupins.

The winter plantings in the summer rainfall region did not include any high potential environments this season. Ronell produced the best overall yield with 1,75 t/ha. The Lindley locality was extremely drought-stressed with a mean yield of 0,71 t/ha. Here again, the two sweet narrow leaved lupins Helderberg (1, 18 t/ha) and Eureka (1, 17 t/ha) performed best. Yield reliability calculations show the cultivar Vladimir to have the greatest reliability over all yield targets. The yield reliability values generally lie about 10% below mean yield. Ronell and Astra are best adapted to environments with low to medium high yield potential. The cultivar CED 6150 is best adapted to environments varying from medium low to high potentials.

5.16

Selection and breeding for (Colletotrichum gloeosporioides) in germ plasm of Lupinus albus and the development of commercial anthracnose resistant cultivars

JAM van der Mey

The 4th generation of sub-populations has been selected within a medium (SAL 60) and a long season (SAL 160) population for anthracnose (Colletotrichum gloeosporiodes) resistance. These sub-populations were subjected to further selection under both field and glasshouse conditions. These environments were Caledon (W. Cape), Bethlehem (Free State), Ermelo (Mpumalanga Highveld), Potchefstroom (NW Province), as well as in the Potchefstroom glasshouse. At Caledon and in the Potchefstroom glasshouse, where anthracnose occurred, about 800 single plant selections were made. At the other localities, where no anthracnose occurred, about 40 row selections were made, using the degree of resistance encountered at Caledon and in the Potchefstroom glasshouse in the same lines and agronomic criteria as criteria. These selections will be used in a check plot trial to be planted at Caledon (under both dry land and irrigated conditions), Bethlehem and Potchefstroom during the 2001 season. The single plant selections will be tested further under artificial inoculation in the glasshouse to obtain greater homogeneity for the anthracnose resistance character, which appears to be highly polygenic.

5.17

Pathological characterisation and selection of lupin anthracnose causing strain(s) to be used in resistance screening

SH Koch

Isolations and re-isolations were made from all 1upin and weed samples collected since the outbreak of the disease in 1993. Isolates were also obtained from five new samples collected during 2000 from different localities.

An international set of 16 1upin accessions, with various genetic backgrounds, was constantly multiplied to produce seed. These seeds are being used to distinguish pathogenic variability in the fungal isolates.

A more pathogenic reaction was observed in the international set when inoculated with a mixture of 1999 cultures, compared to results that were obtained in different parts of the world. Lupinus angustifolius cvs, in particular, tested more susceptible than before. Due to limited amounts of seed available from the international set, a new set of cultivars was compiled representative of the accession being used in the local cultivar trials. The previously tested isolates and some of the new isolates (seven in total) are currently being compared separately on a set of 16 cultivars. Based on the statistical analysis of two sets of data no obvious differences in pathogenicity between isolates were recorded.

Over a period of six months Lupinus albus cv Esta (susceptible) and L. angustifolius cv Merrit (less susceptible) were regularly inoculated with different concentrations of the fungus, different wetting agents and under different incubation conditions. Inoculated material was dissected, stained and destained with various stains and destains. The material was then mounted and observations were made with the aid of a fight microscope. Initially difficulties were encountered in obtaining even germination of spores. This problem was overcome, but the fungus still did not develop past the appresoria formation stage. The spores showed a tendency to accumulate in and around the trichomes, especially on the stems. As information on the infection processes is of utmost importance in understanding the nature of resistance, this work is being continued in collaboration with the Departments of Botany and Plant Pathology at the University of the Free State.

In addition to the 142 lupin accessions received from KwaZulu-Natal Department of Agriculture during 1999, 68 breeding lines were received from ARC-GCI and 6 cultivars, from PRF. The seed collection currently contains 250 1upin accessions. During the report period another 48 accessions from the KwaZulu-Natal collection were screened for resistance to the aforementioned mixture. All the accessions still tested moderately susceptible to very susceptible.

5.18

Genetic engineering of lupins (Lupinus sp.) for resistance to anthracnose

TG Watson

This project is a collaborative project between CSIR Bio/Chemtek and ARC Roodeplaat. The aim of the project is to genetically engineer lupins for resistance to anthracnose caused by a distinct Colletotrichum sp. Work in previous years focused on optimisation of a regeneration and transformation protocol for lupin explants, and the identification and efficacy of the antifungal protein, apple PGIP, in inhibiting Colletotrichum sp growth.

The apple PGIP gene was isolated from apple genomic DNA, and was cloned into an Agrobacterium tumefaciens binary vector. Transformation of tobacco with this construct and a subsequent sugar reducing assay indicated that apple PGIP exhibited up to 73% inhibition of Colletotrichum sp polygalacturonases. These results demonstrate that the cloned apple PGIP is expressed and synthesised in an active form in plants, and confirm that the apple PGIP gene is a suitable candidate for genetically engineering 1upins for resistance to anthracnose. However, a combination of the apple PGIP gene and other antifungal genes with a different mode of action will be adopted to obtain long-term durable resistance to Colletotrichum sp.

Evidence that vacuum infiltration of legume seeds with A. tumefaciens led to increased transformation efficiencies motivated us to change our research focus and optimise vacuum infiltration mediated A. tumefaciens transformation of L. angustifolius cv Helderberg. Initial experiments indicated that this approach yielded high transient transformation frequencies. We have therefore adopted this method to introduce the apple PGIP gene in L. angustifolius. Once transformed plants have produced seed, progeny will be screened for the presence of the transgenes.

5.19

Vacuum infiltration of lupin seed for genetic engineering for resistance to anthracnose

JA de Ronde

Lupin seed were transformed with a pCambia construct using the Agrobacterium-vacuum infiltration method optimised on soybean. Various criteria that could influence the transformation efficiency were investigated. It was found that a dilution of 50% of the Agrobacterium culture in a phosphate buffer resulted in maximum transformation, least contamination of the seed and least lysed cells. Two types of Agrobacterium strains were used to transform the 1upin: LBA4404 and AGL 1.

While growing the LBA4404 culture it was noted that flocculation took place, which resulted in a lower transformation efficiency than the AGL I, which grew more evenly. No efficiency difference was observed with or without Vir G. The construct transformed into the lupin contained 2 marker genes: a bar gene, coding for herbicide resistance and can be tested with Ignite and the GUS-intron gene, which gives a blue colouring when tested with X-GLUC. The presence of the GUS-intron gene was demonstrated in leaf tissue (TO) as well as seed (TI), indicating that transformation took place and that the offspring expressed the GUS gene.

The putative transformants were tested for the presence of the bar gene through 0,1% Ignite leaf painting and after 3 days the difference between transformed and non-transformed plants was clearly visible. The transgenics displayed significantly less damage than the control plants. Results indicated that an 11% transformation rate was accomplished with lupin plants using the marker genes. The pgip gene will now be transferred into the lupin seed.

5.20

Inoculation of lupins in South Africa

IJ Law

Lupins are protein-rich crop legumes that can obtain most of their nutritional nitrogen through the activity of symbiotic N2-fixing root-nodule bacteria called rhizobia. Lupins are not indigenous to South Africa and symbiotic rhizobia must be inoculated onto seed in newly planted soils. Lupins have, however, been cultivated for many decades in the Swartland region of the Western Cape and naturalised populations of rhizobia have become established in these soils through repeated inoculation. Because the natural populations in the Western Cape are effective, many farmers in the region do not inoculate and instead rely on soil rhizobia for nodulation of their crop.

It is well known that a response to inoculation is easily achieved when legumes are grown in soils devoid of rhizobia, whereas inoculation is invariably futile where large resident populations of rhizobia exist. A recent survey of eight Western Cape farm soils provided evidence that both situations exist in this region as soil populations of 1upin rhizobia ranged from undetectable to about 3000/g. Field trials were subsequently carried out to determine the effect of inoculation at sites that had a history of lupin cultivation.

The experimental sites were located in districts of the Swartland (Eendekuil, Hopefield, Moorreesburg) and Southern Cape (Riviersonderend) regions and all contained large rhizobial populations (10 000 to 100 000/g). Glasshouse tests showed that each population was comparable in nitrogen-fixing effectiveness to the inoculant strain VKl0. Although about 1 million cells of VK10 were inoculated onto individual seeds at planting it is estimated that this strain was outnumbered 100:1 by competing soil populations.

Subsequent analysis of nodulated plants confirmed that relatively few nodules were formed by VKI0 at each site. At Hopefield, for example, only 20% of the nodules of inoculated plants contained VK10 even though this strain formed part of the existing soil population, being detected in 12% of uninoculated plant nodules.

This explains why such inoculation did not significantly increase seed yields (kg/ha) at any of the localities following harvest. Application of ferti­liser nitrogen (45 kg nitrogen/ha) also did not significantly increase yields over that of uninoculated treatments. This showed that the soil rhizobia were capable of providing the plant with adequate nitrogen. Lupin yield did not benefit when molybdenum was included in inoculated, uninoculated or fertiliser nitrogen treatments, although this trace element is recommended for application to Lupin.

Although the field trials failed to demonstrate significant benefits from inoculation, nodulation of uninoculated plants at one site (Eendekuil) was observed to be less uniform than that of inoculated plants. This indicated that the soil rhizobial population was not evenly distributed. A sig­ni­ficant (5% probability) increase in seed yield per plant was achieved by inoculation at this site, possibly because plants of this treatment were uniformly nodulated. This suggests that the farmer may obtain benefit from inoculation even in fields containing effective rhizobia.

5.21

Quantifying the energy values and composition of potential alternative protein sources

TS Brand

Samples of sweet yellow (Lupinus luteus; n=4), broad leaf (Lupinus albus; n=12), narrow leaf (Lupinus angustifolius; n=8), faba beans (Vicia faba; n=2), peas (Pisum sativum; n=4) and narbon beans (Vicia narbonesis; n=2) were collected over a two-year period. The physical characteristics (thousand seed and hectolitre mass), chemical composition (dry matter, ash, crude protein, ether extract, acid detergent fibre, neutral detergent fibre and mineral content), energy values (TMEn for roosters) as well as the lysine and methionine availabilities (with roosters) of the samples were determined. L. albus had the highest nitrogen corrected true metobolisable energy content (TMEn) (12,49 MJ/kg) value, followed by that of field peas (11,35 MJ/kg) and narbon beans (11,25 MJ/kg), faba beans (10,90 TvU/kg), L. angustifolius (10,46 TvU/kg) and L. luteus (10,20 MJ/kg). L. luteus had the highest crude protein (CP) content (39,36%), followed by that of L. albus (38,19%), faba beans (26,00%), field peas (24,74%) and narbon beans (23,76%). L. luteus had the highest lysine content (2.22%), followed by the value of L. albus (1,96%), field peas (1,93%), L. angustifolius (1,86%), narbon beans (1,75%) and faba beans (1,70%).

Important correlation coefficients with the TMEn for lupins were ether extract (EE) (0,82) and thousand seed mass (TSM) (0,80). TMEn for roosters could be predicted by ash and EE contents (R2=0,74) or by TSM (R2=0,62).

It was concluded that grain legumes may be a valuable potential protein source for monogastric animals. The study presents a range of values for different cultivars of these type of grain legumes cultivated in South Africa. The values are important to determine the nutritive and financial value of these grain legume types.

5.22

Cultivar evaluation of oil and protein seeds in the winter rainfall region (canola)

DJ Hanekom

As in the previous two seasons ('98 and '99), the Chief Directorate Agriculture was responsible for cultivar evaluation of canola during the 2000 season. Twenty-one cultivars were tested at 12 sites, 6 in the Southern Cape and 6 in the Western Cape region. An additional planting date was planned in each of the regions at Tygerhoek and Langgewens experimental farms. Only the trial at Tygerhoek could eventually be carried out, as the late start of the season in the Swartland made early planting at Langgewens impossible. Several new canola varieties as well as interesting breeding lines were obtained and included in the programme. Data pertaining to flowering date and yield were collected and analysed for statistical differences.

All the trials in the Southern Cape region, with the exception of the Caledon site, were completed. Extremely low soil moisture conditions after planting resulted in poor establishment of this trial, which was eventually abandoned. The average yield of canola at the different sites ranged from 734,7 kg/ha (Heidelberg) to 1 683,3 kg/ha (Tygerhoek early planting date). The mean yield of cultivars for all of the sites in this region ranged from 885,9 kg/ha (Hylite 200) to 1 753,4 kg/ha (Hyo1a 42). The highest yield recorded was for the cultivar Rainbow at the Napier site.

Severe moisture stress conditions prevailed for a significant period during the season at most of the sites in the Swartland region. This condition was extremely severe at the Hopefield trial, which led to uneven establishment and eventually to the trial being abandoned. The above-average rainfall that was measured during July at the sites did, however, lead to good recovery of the canola at some of the sites.

The first significant bird damage was also experienced for the first time in this region. This was probably caused by the late start of the season resulting in canola being the only available food source for birds at that time of the season.

Because of the aforementioned only the data obtained from the Langgewens and Eendekuil trials were included in the final analysis. An average yield of 1 757,6 kg/ha was recorded at the Langgewens site where the cultivar Varola 44 gave the best yield of 2 703,8 kg/ha. At the Eendekuil trial the cultivar Varola 44 gave a yield of 1 083,8 kg/ha in spite of an average yield of only 461,7 kg/ha.

Yield assurance data for the 2000 season showed that the cultivars Monty, Hyola 42, Varola 50, Varola 44 and the breeding fine 54801 gave the best yield assurance at potentials that ranged from 0,5 to 3 tons/ha, while Insignia showed good potential at yields above 1.5 tons/ha.

5.23

Epidemiology of blackleg on canola in the winter rainfall region of South Africa: a prerequisite for sustainable disease management

EE Auret

Blackleg (caused by Leptosphaeria maculans) is endemic in most rapeseed producing countries, causing losses of up to 100%. Leptosphaeria maculans attacks cotyledons, leaves, stems, crowns, roots and seedpods. Cotyledon and crown infections are considered the most important phases of this disease, resulting in high plant mortality. To ensure an increase in the production of canola, new cultivars/accessions should be evaluated for resistance / tolerance to blackleg under different South African conditions using local strains of the pathogen.

Blackleg disease nurseries were established at Langgewens (Moorreesburg), Outeniqua (George) and Tygerhoek (Riviersonderend) experimental farms to evaluate blackleg resistance / tolerance of cultivars / accessions under local conditions. Three replicates of 22 cultivars / accessions were planted at each site, and subsequently blackleg-infected stubble was distributed over the experimental plots. Due to insufficient rain at the Langgewens and Outeniqua experimental farms, seedling establishment was poor and low spore release rates from infected stubble occurred.

Although sufficient rain occurred at Tygerhoek to ensure excellent seedling establishment, the situation changed later in the season with low incidences of blackleg on stems and crowns of the more resistant/tolerant cultivars / accessions and negligible infection of seedpods. Blackleg incidences on cotyledons and crowns were significantly higher in cultivar Hyola 420 than all other cultivars / accessions, resulting in low yield and blackleg rating of 1 (0 = susceptible and 10 = resistant / tolerant). Blackleg classification of the other cultivars/accessions was 6 and higher.

Cultivars commercially planted in the winter rainfall region during the 2000 season included Charlton, Dunkeld, Grousse, Hylite 200 TT, Hyola 42, Karoo, Monty, Oscar, Rainbow, Scoop, Varola 50 and Varola 54. Although low incidences of blackleg were recorded for most of these cultivars, high cotyledon infection was recorded for Oscar, and high crown infection for Hylite 200 TT and Hyola 42.

This confirms the importance of blackleg disease and the need for disease nurseries in order to recommend only resistant / tolerant cultivars, thus ensuring an increased and sustainable production of canola in the winter rainfall region.

5.24

The extent and role of canola and lupin soilborne diseases on yield depression: a prerequisite for developing an effective and sustainable control strategy

SC Lamprecht

Soilborne diseases are major constraints to both sustainability and expansion of canola and 1upin production in the winter rainfall region of South Africa. These diseases cause serious economic losses worldwide and annual yield losses of more than 20% have been reported in other countries. However, little or no information about these diseases is available in South Africa. During 1999 the incidence and extent of soilborne diseases of canola and lupin were determined in the major production areas in the winter rainfall region. The farming areas in the Southern Cape (Bredasdorp / Napier, Heidelberg and Swellendam) and the Swartland (Moorreesburg, Piketberg / Eendekuil and Porterville) included in this survey, represented different soil types and rainfall regimes. Plants were sampled at 4-6 weeks (seedling stage), 8-10 weeks (vegetative stage) and 14-16 weeks (flower / pod stage) after planting.

This survey confirms the results obtained in other countries, viz. that blackleg (caused by Leptosphaeria maculans) are the most important disease of canola. Disease incidence of up to 100% was observed in the Swartland, but was less severe in the Southern Cape. Damping-off can be considered the second most important seedling disease of canola. Several fungi that are also highly pathogenic on other crops were associated with damping-off viz. Fusarium avenaceum, Pythium irregulare, P. tracheiphilum and Rhizoctonia solani. Low incidences (0,1 and 0,3% respectively) of black spot and Sclerotinia stem rot were recorded on the vegetative stage of canola. This situation changed with the subsequent flower/pod stage. Black spot with incidences of 1,5% for the Southern Cape and 14% for the Swartland, may therefore be considered the second most important disease of mature canola, followed by Sclerotinia stem rot (0 and 4% incidences for the Southern Cape and Swartland, respectively). According to this survey, root rot is the fourth most important disease of canola in the winter rainfall region.

Damping-off is the most important disease of lupin during the seedling stage, with an incidence of up to 43% recorded. In these cases, large bare patches were observed in the affected fields. Fungi associated with the disease were: E avenaceum, E solani, E oxysporum, Phytophthora megasperma, P. irregulare, P. tracheiphilum and R. solani. Sclerotinia stem rot was not recorded in lupin fields included in this survey. Leaf blotch (caused by Alternaria spp.) was observed at both the vegetative and flower / pod stages of 1upins with incidences of 2 and 16% respectively in the Southern Cape, and 2 and 10% respectively in the Swartland. Crown rot of lupin was observed only in the Swartland and incidences increased from the vegetative (1%) to the flower / pod stage (3%), resulting in bare patches in lupin fields. Rhizoctonia solani was frequently isolated from the diseased crown tissue. According to this survey, root rot is the most important soilborne disease of lupin in the winter rainfall region. High incidences (8%) were recorded in the vegetative stage with a slight increase (10%) towards the flower / pod stage in the Southern Cape. Although the incidence of 1upin root rot was lower in the Swartland, the same trend of higher infection of the flower / pod stage occurred. Similar to damping-off, several fungi pathogenic on other crops were isolated from root rot lesions including E avenaceum, E solani, F oxysporum, P megasperma and P. irregulare.

During this survey the incidence and extent of soilborne diseases of canola and 1upin in the winter rainfall region, as well as organisms associated with these diseases, were determined. This information is indispensable for developing sustainable disease management strategies.

5.25

Optimal planting methods for oil and protein seed crops in the Western and Southern Cape

GA Agenbag

Production of canola did not reach expectations set for the Western and Southern Cape because yields of 2,0 ton/ha^-1 and more were not obtained generally. Uneven germination, which results in poor establishment and low plant populations, is one of the main causes of lower than expected yields. This problem was studied by comparing planting methods where seed was broadcasted and where a presswheel planter was used on two localities during the period 1998-2000. In spite of the unfavourable planting conditions (hot and dry) that prevailed during all years tested, higher plant densities were obtained where the presswheel planter was used.

Although results varied between years, the use of this planting method resulted on average in a 10% higher plant establishment at Roodebloem. Average yield of 1 585 kg/ha^-1 obtained with the presswheel planter, however, did not differ significantly from average yields of 1398 kg/ha^-1 obtained where seed was broadcasted. This was most probably due to the indefinite growth habit of canola, which enables the crop to compensate for low plant populations. The use of presswheel planters to establish canola is nevertheless still recommended because of its effect on plant densities.

5.26

Economical short rotation crops for wheat

PJ Pieterse

The project investigating the growth and yield of wheat and the mineralisation of nitrogen in different rotational systems was continued. The study forms part of the large-scale rotational cropping experiment of the Department of Economic Affairs, Agriculture and Tourism's Chief Directorate Agriculture: Western Cape, that is being carried out on the Langgewens Experimental Farm north of Malmesbury. Trends observed in 2000 differed from trends observed in 1999. In 1999, wheat yield was higher (although not statistically significant) in treatments where lupins were planted in the previous year. In 2000, however, significantly higher yields were obtained in the experimental plots where canola and medics were grown in the previous year. This trend was also reflected in the early vegetative growth of the wheat plants and the nitrogen content of the soils in the early part of the season. An incubation test was also performed in which the nitrogen mineralising potential of soil planted to different crops two years previously was compared. It was clear that soil that was planted to legume crops two years before had a better nitrogen mineralising potential than soil planted to non-leguminous crops. The study will be continued for a further two years.

5.27

An investigation into the production dynamics of eight crop rotation systems, including wheat, canola, lupins and pasture species in the Swartland

M Hardy

The aim of this study is to determine the short and long-term effects of eight of the most feasible crop and crop/pasture rotation systems iden­tified for the Swartland on: crop yields; weed control; disease suppression; soil production potential; sheep production; economically sustainable land-use in the Swartland.

It is envisaged that the experiment will lead to: reduced input costs for wheat production; increased protein and oil seed production; increased diversification of the farming system for greater financial stability; reduced costs to livestock farmers; increased production potential of soils due to increased organic matter content, and micro and macro-organism activity.

Detailed economic analyses of all rotation systems over the first four-year cycle of the experiment were completed during the report period. These analyses have revealed a convincing trend of greater financial stability and profitability from crop rotations that include canola, lupins and annual legume pastures when compared to a wheat monoculture. The economic data have also been extrapolated to simulate the effects of converting from a wheat monoculture to various crop rotation options on a "typical" farm in the Swartland.

Variables include, among other factors, farm size and rainfall patterns. This work was done on a full-time basis by an MSc Agric. Economics student who will be completing his thesis in May 2001.

Crop yield measurements show that a greater quantity and higher quality of wheat is produced when canola, lupin or pasture was grown in the previous season. A season of 1upin production appears to have the greatest positive effect on wheat production in the following season. There are even indications that only one season of canola production in a three-year wheat / one-year canola rotation has advantages for subsequent wheat crops.

In an attempt to understand the (biological) reasons why the production characteristics of one rotation system differ from those of another, various other (component) studies were undertaken during the report period, namely:

• Nitrogen mineralisation and the effects of rotation system on soil nitrogen availability for wheat production. A student completed his M.Sc degree on this aspect during January 2001.
• The effects of crop rotation system on weed-seed banks.
• The effect of rotation system and nitrogen application on wheat production.
• The effects of crop rotation system on soil micro-organism activity.
• The effects of crop rotation system on air and soilborne diseases.
• The state of medic and clover seed banks for each treatment as affected by crop rotation system, pasture management and physical environment.

The trial thus provides a practical "laboratory" for researchers, students, advisors and farmers to gain insight into the advantages of crop rotation for sustainable crop production in the Swartland.

During the 2000 report period information derived from the experiment was published in the Landbouweekblad (2 articles) and presented to farmers at the SKOG farmer's day in the Swartland and at a Farmer's Association meeting in the Southern Cape.

5.28

Application of feed enzymes to lupin, full fat canola and oilcake and containing diets for broilers and weaning pigs

TS Brand

In this experiment diets with either soybean oilcake (control), canola. oilcake, full-fat canola or sweet lupins (L angustifolius) were fed to broilers, while enzymes were either added to or withheld from diets. All diets were formulated on an iso-nutrient basis. The three diets with the test protein sources were blended with the control diet to produce four levels (0% test protein, 33% test protein, 67% test protein and 100% test protein) of each protein source. Ninety-six pens with 85 broiler chickens were used (8 160 chickens). The application of feed enzymes had no effect on the production of broiler chickens. A small negative effect on both growth rate and the efficiency of feed utilisation was found with increasing levels of all test protein sources compared to the soya control diet. The largest suppressing effect on production was observed with increasing levels of canola, oilcake meal.

5.29

Should animals be fed to their genetic potential if this potential is reduced by stress

RM Gous

A sensible approach to determining the nutrient requirements of genotypes of growing pigs is to base these on the potential growth rate of the genotype concerned. The important information required about the genotype for such determinations is the mature protein weight, the potential protein growth rate and the lipid-to-protein content in the gain (or at maturity). Techniques have been devised and tested at the University of Natal for measuring these characteristics in growing pigs, the essential component in such measurements being a non limiting environment. However, under commercial conditions, it is unlikely that the environment will be nonlimiting throughout the growing period, with high temperatures, high stocking densities, disease challenges, lack of feeder space, social stresses and other constraints leading to a reduction in the potential growth rate. The question addressed in the research reported here is whether the nutrient requirements of growing animals should be reduced commensurate with the reduction in the potential growth rate of the animals caused by stress.

An experiment was conducted at the University of Natal to measure the effects of stress (reduced floor space) on nutrient requirements of pigs grown from 25 to 60 kg five weight. Two hundred and sixty four female Large White x Landrace pigs were assigned at 25 kg to one of four dietary lysine treatments (13,3, 11,4, 9,5 and 7,6 g/kg) and one of three floor space treatments (2,0, 1,0 and 0,5 M. 2/pig). Animals were given ad libitum access to dietary treatments from mean pen starting weight of 26 kg to a mean pen finishing weight of 63 kg live weight. Protein and lipid gains were determined by analysing the carcasses of pigs on the various treatments at the end of the experiment.

The results suggest that feeding according to the requirements for maximum protein retention will still produce the best carcass and growth performance irrespective of the level of stress. The improvement in lysine retention associated with higher dietary nutrient levels did not completely offset the adverse physiological effects of stress on reducing maximum protein retention, but may partly counteract the reduced nutrient intake associated with a stress-related suppression of appetite. However, there were indications that feeding crowded pigs a lower dietary lysine concentration may not further reduce the already diminished protein (lysine) growth rate. An additional experiment was performed to test whether the number of feeder bins may have constrained food intake and therefore growth in group-penned animals. The results of this experiment proved that there were no significant differences in FI, ADG or FCR in group-penned pigs provided with 1, 2, 3 or 4 bins per pen, and therefore a single feeder bin was not considered a constraining factor in pigs housed with limited floor space.

5.30

The response of growing pigs to dietary amino acids

RM Gous

Due to the recent and rapid improvements in potential lean tissue growth in the latest pig genotypes, and the move towards more extensive use of alternative dietary protein sources (e.g. legume seeds), the effect of dietary amino acid balance and environmental temperature has become an area of concern. Of particular concern is the effect of imbalances in amino acids and temperature on voluntary food intake. There is strong evidence to suggest that the latest lean strains of pigs are more sensitive to dietary amino acid imbalances than older genotypes. It is also recognised that the increased use of more fibrous and poorly digestible protein sources has reduced the availability of some essential amino acids, one of which is tryptophan, as well as increased the amount of heat the animal generates in digesting the diet. All the factors will have an impact on protein utilisation and food intake. The effects of temperature on the efficiency of utilisation of both lysine and threonine have been measured in growing pigs at the University of Natal in the past few years.

Unlike lysine and threonine, which are not involved in specific metabolic processes, tryptophan plays an important role in the regulation of appetite through its function as a precursor of brain serotonin. Serotonin is a neuro-transmitter involved in satiation and protein intake regulation. It is postulated, therefore, that decreased dietary concentrations of tryptophan will cause appetite to decline with a consequential decrease in food intake, rather than follow the trend of increasing feed intakes with decreasing dietary crude protein, threonine and lysine, as has been found in the previous experiments. In addition it is not known whether there is an interaction between dietary tryptophan and ambient temperature. Therefore, an experiment was conducted to measure the response of young pigs, between 13 and 26 kg five weight, to dietary tryptophan concentrations and environmental temperatures. The results showed that the response in food intake to increasing dietary tryptophan concentrations was dependent on the environmental temperature. Daily food intake increased significantly with decreasing tryptophan content, reached a maximum and then declined. The concentration at which maximum intake was achieved varied according to the amount of heat the animal could lose. The amount of heat lost will in turn be dependent on the environmental temperature. Maximum food intakes were attained on higher dietary tryptophan levels at higher temperatures. A similar response was reported for protein content of the empty body weight as well as the daily rate of protein deposition. The efficiency of tryptophan utilisation increased from 0,64 at 20ºC to 0,88 at 30ºC. At comfortable temperatures the efficiency was 0,76.

It can be concluded from this and previous research at the University of Natal that when young pigs are faced with a feed marginally limiting in an amino acid, they will increase their food intake in an attempt to satisfy their requirements for the first limiting amino acid, even when the limiting amino acid is tryptophan. However, the extent of this compensation will depend on how much heat the animal can lose to the environment, and this is dependent on the prevailing environmental temperature. At high environmental temperatures the amino acid content and balance in a food become critical if the animal is to utilise this food efficiently.

5.31

A model to predict the usage of agricultural products of the feeding of livestock

RM Gous and E Briedenhann

In today's world, information is power. Incorrect decisions made in purchasing raw materials in the animal feed industry can result in huge losses to a feed company, so there is immense value in being able to predict accurately both the annual requirement for animal and poultry feed, and the most economical means of supplying that requirement from the raw materials available.

Mr Erhard Briedenhann recently obtained a PhD from the University of Natal for producing a model that is capable of accomplishing both these tasks.

The model predicts the demand for animal products, as well as raw material requirements, under a wide range of scenarios, such as changing population dynamics and changes in purchasing decisions among the different population groups in South Africa. It uses linear programming to determine the most economical mix of raw materials that would supply the essential nutritional requirements of the different classes of livestock predicted to be needed to supply the demand for animal products in South Africa. All available sources of these raw materials are considered, as well as all the costs that are incurred in obtaining the materials from the various sources, whether local or international.

The model makes use of the latest computer software platform (Windows 2000), it is available for personal computers, and it is very user-friendly. A wide range of variables may be manipulated, including animal distribution per region in South Africa, animal performance, animal products imported and exported, population growth and per capita consumption of different animal products per population or per income group. The effects of these manipulations on the raw material requirements, and on the optimum economic sources of the ingredients required are made instantly available to the user.

The model will be of benefit to the farming community, giving them a scientific basis on which to decide which crops should be planted by evaluating a realistic selling price of different raw materials relative to predicted prices for competing alternatives. Oilseed crushers would be able to determine market size and acceptable price levels for their products, as the model predicts the volume of oilseed meal that can be absorbed by the market at various prices whilst considering competing alternatives. The milling industry would be able to determine the likely size of markets for their by-products, and trading companies could estimate which commodities would be most likely to be required under different scenarios, and at what price.

This model will be extremely useful, giving those who make use of it the competitive edge in the feed industry in South Africa. The Protein Research Trust, from whom the model may be obtained, met the development costs of the model.

5.32

International supply and demand situation wih regard to protein for animal feed

WI Nieuwoudt

Good progress was made during the report year and a final report will be made available next year in the form of a Masters thesis and man­agement summary. The model on which the study is based has already been completed and articles relating to this work have already been published. Oral presentations were made to interested parties.

In addition to the work on the international model, the model projecting local demand for protein for animal consumption was further refined and upgraded.

The international prices for protein for animal consumption were, for example, integrated into the local model. Adjustments were also made so that constant income elasticities that were applied initially can now vary, while partial price elasticities were converted into total price elasticities with a view to more accurate predictions.

When projections are now made regarding local demand for protein for animal consumption, events in the international arena are thoroughly taken into account. Aspects that are considered include, for example, imports of animal products for human consumption, the effect of protective tariffs, various income growth scenarios and the effect of AIDS on population growth.

It is also clear that this model can provide Mr E Briedenhann's APR model with basic information (demand for animal products and thus the number of animals) with a view to more accurate projections by the latter.

5.33

The extent, importance and potential of the lucern industry in the RSA

HD van Schalkwyk

This project was initiated to gather more information on the lucern industry in order to determine whether the PRF's promotion efforts should also focus on this crop. The lucern industry is already well established in the RSA and although protein per ton is relatively low compared to other crops with which the PRF is involved, protein per hectare lucern is relatively high.

The wider study, which was preceded by a pilot survey, covered all major irrigation areas in the RSA, with the final report being released during the year under review. The study shows clearly that, in terms of size and importance, the lucern industry plays a major role in the South African agricultural sector. Lucern is produced on approximately 208 000-240 000 hectare and it is estimated that 3,08 million tons are produced. Lucern is an important feed crop, especially in the dairy cattle and ostrich industries, but also plays an important role in other livestock sectors. Lucern also makes a positive contribution towards farming in terms of cash flow benefits and as rotational crop.

The majority (65%) of producers indicated that they are not interested in expanding lucern production. According to the study, producers are satisfied with available cultivars. Yield and quality of hay (straw) were identified as problems and are largely influenced by management practices. However, the necessary information is available and must simply be applied.

The vast majority of producers sell lucern directly at farm level with very little value being added. There is good demand for lucern and producers have no difficulty in marketing their crops.

The reasons why producers do not wish to expand lucern production include the following: high input costs, high cost of machinery and implements, relatively low profitability compared to other crops and high water costs.

The conclusion reached by the PRF was that the high priority constraints that are mentioned do not fall within the PRF's research mandate and that the PRF would not become involved in the lucern industry at this stage.

Improved transfer of technology could possibly play a role in upgrading management practices that influence yield and quality. According to feedback received the study was of great value to the major role-players in the lucern industry while the PRF had, in the process, also played a role in promoting the lucern industry.