6.   Projects finalised successfuly or that showed progress (Annexures I and II)

6.1   Evaluation of PRF soybean elite lines under South African conditions

GP de Beer and WF van Wyk

The PRF soybean elite trials (2016/17) were planted at the following six (6) localities:

• Stoffberg – Representing the northern Highveld (cool area)
• University of Pretoria (Hatfield) – Representing the southern Highveld (moderate to warm area)
• Brits – Representing the northern irrigation area (warm area)
• Potchefstroom – Representing the western production area (moderate to cool area)
• Bethlehem – Representing the eastern and northern Free State (cool area)
• Ukulinga (Pietermaritzburg) – Representing KwaZulu-Natal (warm area)

The following four (4) local cultivars were used as standards for the trials:

Seed institutions of South America (Argentina, Uruguay and Brazil) entered 58 elite soybean lines that were evaluated with the four (4) standards at the six (6) localities according to grain yield and general adaptation to South African conditions. The maturity groups (M.G) varied between M.G 4.0 to M.G. 7.4. The trial at Ukulinga was controlled preventively, using registered fungicides, against soybean rust.

A number of the 58 lines produced relatively high grain yields of 5 716kg/ha, 5 702kg/ha and 5 084kg/ha. The best overall yield was produced in Pretoria (UP) by the control DM 5953 RSF at 6 335kg/ha.

The project creates the opportunity for participating institutions to test their materials and to consider local registration of cultivars. As such it expands the choice of soybean cultivars in South Africa to the benefit of soybean producers and the soybean industry in general.

6.2   Etiology and population structure of Macrophomina phaseolina (charcoal rot) in sunflower and soybeans in South Africa

E Jordaan and JE van der Waals

This project investigates interactions between the environment, host (sunflower and soybeans) and Macrophomina phaseolina causing charcoal rot in South Africa. The main objective of this work is to understand the drivers of this disease in order to develop a decision support system for charcoal rot management. In vitro trials, pathogen identification, characterization and greenhouse pot trials will be used to create a holistic picture of charcoal rot on in SA. The results from the in vitro trials have been discussed in previous reports. Unfortunately, the planned grower survey had to be removed as we were unable to obtain enough feedback to make statistically sound conclusions regarding growers' perceptions, the occurrence of this disease and the subsequent control practices that are in place in South Africa. Isolate identification is underway, after several unforeseen setbacks.

Greenhouse pot trials were conducted to investigate the effect of drought on charcoal rot incidence and severity in soybean and sunflower crops and subsequent yield reduction. Although the results have not been statistically analysed yet, results from the water stress trial showed there was no disease incidence for either sunflower or soybeans at seedling stage. Latent colonisation within the stems was found up to 1cm above the root zone even in plants that were not water stressed, suggesting that drought conditions are not required for infection. During flowering, colonisation was measured up to 2cm above root zone for soybean and 3cm for sunflower. No disease was observed in the sunflowers, which could be due to tolerance within the cultivar planted. In soybeans, disease was observed at flowering in both the water stressed and non-water stressed treatments – the latter with lower severity. At harvest disease incidence and severity were high in the inoculated and water stressed soybean and sunflower plants, and stem colonisation was observed 3cm up the stems in all inoculated plants irrespective of water stress. Yields from the water stressed non-inoculated plants as well as the water stressed inoculated plants were half that of the yields from the control plants (non-inoculated and non-water stressed). Plants that were inoculated but not water stressed showed similar yields to that of the control. Another pot trial evaluated the effect of urea and limestone ammonium nitrate (LAN) applied pre-plant at the recommended rate (15kg/ha for soybeans; 50kg/ha for sunflowers) and half the recommended rate on disease development. For soybean, long and medium growth class cultivars were selected and for sunflower medium and medium-late seasonal cultivars were selected. Results have not been statistically analysed yet, however growth, stem width, total nitrogen and number of seeds/pods seemed to be unaffected by the pathogen under the different N applications in different growth habit cultivars. The pathogen colonised soybean and sunflower stems up to 3cm from the root zone throughout all the treatments. No disease was observed and higher yields were recorded in the long growth class soybean cultivar that did not receive nitrogen pre-plant. Higher disease incidence and severity were associated with urea applications in soybeans. In the sunflower trial disease incidence and severity were higher in the medium cultivar than in the medium-late cultivar irrespective of the nitrogen applications. No disease was observed in medium-late cultivar sunflowers treated with LAN at half the recommended rate or urea at the recommended rate.

For the decision support system, historical weather data, literature mining and various calculations were utilised. The optimal growing temperatures of South African M. phaseolina isolates were found to be between 25 and 30⁰C. Coupled with the reduced rain and constraints on irrigation we would expect higher incidence of this disease than in rainy seasons. There is very little information available on the incidence of charcoal rot on sunflower or soybeans in terms of location and year of outbreak. However, during the 2011/2012 growing season where losses in maize due to charcoal rot of up to 60% were recorded. For this reason, the decision support system model will be based on maize data, using the 2011/2012 incident to validate the model. From this the model can be adjusted for sunflower and soybean crops.

Future research from this project could be focused on investigating epidemiological aspects of the disease to refine the decision support system for use at farm level. A lay article on this disease and research was published in The Conversation in August 2016 (theconversation.com). Upon completion of the project, results will be published in peer reviewed articles in scientific journals, articles in local media such as Farmers Weekly and Oilseeds Focus, and presented at farmer days.

6.3   Studies on Lecanicillium muscarium as a mycoparasite of the soybean rust fungus, Phakopsora pachyrhizi and its use as a biocontrol agent against soybean rust

KS Yobo

In this study, a Isolate N-08, a mycoparasitic fungus, was isolated from Assagay coffee farm, Cato Ridge, KwaZulu-Natal, South Africa, where it was observed parasitizing Hemileia vastatrix, the causal agent of coffee rust. Based on morphological and molecular studies the Isolate N-08 was identified as Lecanicillium muscarium and it was deposited into the National collection of fungi (Accession number PPRI 13715).

Co-inoculation studies of L. muscarium and P. pachyrhizi were done in UKZN Plant Pathology disease garden. The Environmental Scanning Electron Microscope (ESEM) observation of the interactions showed a mycophilic attraction of L. muscarium to P. pachyrhizi urediniospores. Long L. muscarium phialides were observed penetrating and wrapping tightly around P. pachyrhizi urediniospores.

In vitro studies to test the effect of the L. muscarium strain N-08 on P. pachyrhizi, the soybean rust fungus, were done. L. muscarium strain N-08 was observed colonizing P. pachyrhizi under light microscope and ESEM. Laboratory experiments were conducted to assess the effects of different growing conditions (temperatures, artificial growing media, natural substrates and UV radiation) on colony growth and conidia production.

Optimization of growing conditions is one of the essential aspects which must be taken into consideration to produce an effective biocontrol agent. L. muscarium strain N-08 grows best at temperatures between 21 to 25ºC. The highest radial growth was observed at 24ºC (46.54mm). V8 juice agar was the best media for colony growth with the mean value of 42.75mm followed by SDA with 37.86mm. When the isolates were exposed to UV light, the results did not show a significant difference between different media on mycelia growth. The highest conidia production occurred on millet cereal (4.2 x 109 conidia/ml) followed by wheat bran (3.2 x 109 conidia/ml) and pearled barley (2.9 x 109 conidia/ml).

The optimal dose level for disease control was assessed in the greenhouse and in field trials. It was found that 106 and 108 conidia/ml were more effective and 106 conidia/ml was chosen as the optimum dose for the field application.

Effect of L. muscarium against soybean rust was evaluated in the field. Two field experimental trials (2014/2015 and 2015/2016) were run at Ukulinga Research Farm. Compared to the pathogen inoculated control, all the three L. muscarium doses (10⁴, 10⁶, 10⁸ conidia/ml) and the fungicide control (Score) decreased disease severity by 73.3%, 88.2%, 89.1%, and 90% respectively. The Area Under the Disease Progress Curve (AUDPC) for the treatments were as follows: 1st trial, Score (172.2 units), 10⁸ conidia/ml (186.2 units), 10⁶ conidia/ml (202.16 units), 10⁴ (457.8 units) and pathogen inoculated control (1 716.8 units). 2nd trial, score (259.7 units), 10⁸ (284.9 units), 10⁶ (319.9 units), 10⁴ (462.7 units) and the pathogen inoculated control (1 053.5 units). Treated plots showed higher yield increase compared to the pathogen inoculated pathogen. However, dry seed weight did not significantly differ between the L. muscarium strain N-08 treated and score fungicide treated plots.

6.4   Chemical manipulation of vegetative growth, reproductive development and grain yield in canola

GA Agenbag

Canola (Brassica napus L.) is one of the most important sources of plant oil in the world and is rapidly becoming an important crop in the South Africa as well. Although yield per hectare increases during recent years due to the introduction of hybrid cultivars and improved production techniques, yield per hectare is still low compared to leading world producers such as Canada.

Lower than expected yields may be the result of several factors such as low and uneven plant populations, insect pests, poor plant nutrition management and weed control as well as harvesting losses. The highest yields in South Africa are achieved with early plantings on high fertility sites, but this practice often produce bulky crops, which when combined with high plant populations may result in lodging during pod development.

Research done in Australia, showed that shorter plants are much more resistant to lodging than taller plants. By shortening the stem and changing the canopy structure with the use of plant growth regulators (PGR's), an even, compact pod canopy can be produced. As a result, competition for assimilates and light can be reduced, ripening will be more uniform, pod shattering will be reduced and harvesting will be more efficient.

No PGR's are at present registered for use in canola, in the RSA, but preliminary research done recently with the PGR's, Primo Maxx® and Moddus® as well as liquid seaweed extract (Kelpak®) showed promising results in both pot and field trails. For this reason field trials were conducted during 2015 and 2016 at 3 localities in the Swartland (2) and Southern Cape (1) canola producing areas. Three spraying treatments (control, Kelpak®and Moddus®) in combination with different application rates and two different plant densities were tested.

In contrast to 2015 when generally poor growing conditions due to low rainfalls were experienced at all localities but large responses with treatments were noted, applications of Kelpak® and Moddus®) at budding stage did not increase grain grain at any locality in either low or high plant population plots during 2016 when generally favorable growing conditions prevail at all localities. However, when applied at both budding stage and at the onset of flowering, increases in grain yield were reported at Altona and Langgewens with applications of 4.0 L of Kelpak® per hectare, while Moddus®) did increase grain yield by up to 380kg per hectare at Langgewens when applied at a rate of 0.4 or 0.8 L per hectare. Similar but smaller yield increases with the application of Moddus®) were also recorded at Roodebloem. These results indicated that both Kelpak® and Moddus®) showed better results when growing conditions were not optimal.

6.5   Nitrogen top dressings in canola: time of application and rates

GA Agenbag

Previous research projects in the Western Cape showed that application rates of 80-120kg of N ha^-1 and 15-30kg S ha^-1 are needed to produce canola grain yields of more than 2.0 ton ha^-1 in soil with low organic C contents. These high fertiliser requirements increases production costs and often make nitrogen and sulphur fertilisation the most costly production factor in canola.

The efficiency of applications is affected by soil properties and climatic conditions and very importantly by time of application. Research done in Canada showed that although the nitrogen uptake by canola is the highest from the 5-leaf to 50% flower (which in the Western Cape is reached at 80-90 days after planting), uptake remain high till 50% podded stage (120-130 days after planting). These results indicated that nitrogen topdressing the flowering stage of canola may be important in high yielding canola crops.

In order to determine optimum nitrogen application strategies for different soil and climatic conditions, field trials were conducted during 2015 and 2016 at 3 localities in the Swartland (2) and Southern Cape (1) canola producing areas. Four nitrogen rates namely 60, 90, 120 and 150kg N ha^-1 were tested, with 20kg N ha^-1 applied at planting and the remaining nitrogen applied as a single top dressing at 30 days after planting (dap); divided between 30 and 60 dap or divided between 30, 60 and 90 dap (full flowering stage). Control plots did not receive any nitrogen fertiliser.

In contrast to 2015, most canola producing areas in the Western Cape experienced rainfall during the 2016 growing season which were close to the long term means. These conditions together with generally low maximum temperatures during September resulted in high canola grain yields. Mean yields in trails varied between 1681 kg ha^-1 at Langgewens (low rainfall locality) to 3 300kg ha^-1 at the high rainfall locality of Altona. In the Southern Cape (Roodebloem locality) a mean yield of 1 877kg ha^-1 was recorded.

At the high rainfall locality of Altona, the highest yield of 3 420kg ha^-1 was obtained with a nitrogen application of 120kg ha^-1 of which 20kg N ha^-1 was applied at planting and the rest (100kg N ha^-1) divided between 30 and 60 days after planting.

At Roodebloem, the highest yield of 2 228kg ha^-1 was also obtained with 120kg N ha^-1, but at this locality it was achieved with only one topdressing at 30 days after planting. At Langgewens (low rainfall locality), the highest grain yield of 1 779kg ha^-1 was obtained with only 90kg N ha^-1, but with topdressings at 30, 60 and 90 days after planting.

Results shown that the optimal application rate and time of application differs between localities and most probably also between years. For this reason it will be necessary to develop criteria to enable producers to decide on the best application strategy in different growing seasons.

6.6   Grain Yield competition for canola 2016

GA Agenbag

The PRF Canola Grain Yield Competition was held during 2016 for the second year and tough competition and high grain yields were experienced amongst competitors.

In the Swartland production area, the 14 competing producers showed on average a grain yield of 2.03 ton per hectare on a total area of 459 hectares. In the Southern Cape a mean grain yield of 2.34 ton per hectare was recorded on a total area of 272 hectare and 8 out of the 11 competing producers recorded yields of more than 2.0 ton per hectare.

Mr Dirk Lesch (Elim Boerdery near Malmesbury) won the Swartland competition, while Mr Pieter Beukes (Yahshua Boerdery near Caledon) won the competition in the Southern Cape.

Mr Lesch recorded a grain yield of 2.66 ton per hectare from a canola field of 18.7 hectare, planted on 26 April 2016 with the DuPont-Pioneer canola-cultivar 45Y88 at a planting density of 2kg seed per hectare. Mr Beukes also used the DuPont-Pioneer canola-cultivar 45Y88. His canola field of 15.7 hectare was planted on 27 April 2017 at a density of 4kg seed per hectare and he recorded a record winning grain yield of 3.09 ton per hectare.

6.7   Canola technology transfer

GA Agenbag

The timely transfer of technology and information on the production of canola are very important to ensure that producers are optimizing their production techniques to obtain maximum yields. For this reason canola-information days are organized annually by the PRF. During 2016 the third annual canola-information day was held for agricultural advisors of the Agro-chemical industry.

The program for the day included a range of production related subjects, but special attention was given to the control of diseases such as Sclerotinia. There can be little doubt that information presented during this day contributed to the highly successful control of this disease during 2016 and more optimal production techniques used by producers which ensured that several producers obtained grain yield of more than 2 ton per hectare.

6.8   An evaluation of continuous cash crop production (including small grains, canola and other alternative broadleaf crops) under conservation agriculture principles on high potential soils of the Riversdale flats

JA Strauss

2016 was the 5th year of continuous cropping research at the Riversdale site. Six cash crop systems are tested including shortened canola rotations and cover crops. A total of 60 plots are planted. The 6 systems tested are replicated 3 times and all crops within each system are represented on the field each year.

Riversdale received excellent summer rainfall of just over 200mm in the pre-season which resulted in enough available moisture to plant at the end of April. Unfortunately there were issues with the rainfall data generated by the ARC weather stations and we could not get an accurate description of the rainfall though out the production season. General indication was that the rain during the 2016 production season was less than average. The Western Cape Department of Agriculture has invested in our own weather stations at all the main crop rotation trial sites.

Hyola 555 was planted at Riversdale at 3.9kg/ha. A total of 45kg N/ha was applied to each plot (24kg N/ha at planting and 21kg N/ha top-dressing). Canola plots following the legume cover crop did not receive the topdressing. Canola yields at Riversdale averaged 2 275kg/ha with all plots showing oil yield above 43%. Canola yields ranged from 1 324kg/ha to 2 833kg/ha. This was on average 855kg/ha more than the 2015 season (1 420kg/ha) and 845kg/ha more than the 2014 season (1 430kg/ha). The canola following the legume cover crop in the system cover crop – canola – wheat has outperformed the canola in other systems every year since the inception of the new rotation systems at the Riversdale site, even when other plots received an extra topdressing.

6.9   Investigating the potential of indigenous nematode to control invasive molluscs in canola

A Pieterse, JL Ross and AP Malan

European molluscs (slugs and snails) have become significant pests in South Africa, especially in terms of targeting canola crops in the winter months. The local production of canola is an important objective for the Protein Research Foundation, as it is the only successful method for the large-scale production of animal feed protein in the Western Cape Province. Therefore, protecting this crop is important. Current methods for controlling molluscs, which rely on synthetic chemical pesticides, are often overused, ineffective and toxic to non-target organisms and the environment. Therefore, identifying a biological remedy is of key importance. The use of mollusc-parasitic nematodes is a feasible environmentally friendly alternative.

As of current knowledge, eight nematode families associate with molluscs, including Agfidae, Alaninematidae, Alloionematidae, Angiostomatidae, Cosmocercidae, Diplogastridae, Mermithidae, and Rhabditidae. To date, Phasmarhabditis hermaphrodita is the only nematode to have been developed as a biological molluscicide in Europe. The nematode, which was commercially released in 1994 by MicroBio Ltd, Littlehampton, UK (formally Becker Underwood, now BASF) under the trade name Nemaslug®, is now sold in fifteen different European countries. Due to current legislation, Nemaslug® can neither be sold, nor used in South Africa. A survey was, therefore, conducted in the Western Cape Province of South Africa to locate local nematode isolates capable of causing mortality in invasive mollusc pests.

A total of 1 944 slugs were collected from 12 different study sites. On the identification of the slugs, they were dissected alive, and examined for internal nematodes. The nematodes were identified using morphological and molecular techniques (18S rRNA). Seven of the 12 sites had nematodes present, with 8% of the slugs being found to be infected with nematodes. Six nematode species were identified, including Angiostoma margaretae, Angiostoma sp. SA1, Caenorhabditis elegans, mermitid sp. SA1, Phasmarhabditis sp. SA3, and Phasmarhabditis papillosa (previously referred to as Phasmarhabditis sp. SA4). Of the six species mentioned, four were previously undescribed. The isolation of new Phasmarhabditis spp. indicates the importance of conducting further surveys of mollusc-parasitic nematodes in South Africa.

Nematodes isolated in the survey were tested for their ability to reproduce on decaying organic matter, with results demonstrating that one of the nematodes, P. papillosa, could complete its life cycle under such conditions. Phasmarhabditis papillosa was, then, fully described and characterised by the shape and length of the female tail, and by the presence of males. Phylogenetic analysis demonstrated that P. papillosa was placed in a monophyletic clade, along with Phasmarhabditis sp. SA2 (another isolate found in South Africa), P. papillosa (US strain), and the mollusc-parasitic nematode, Angiostoma dentiferum.

Phasmarhabditis papillosa was established in monoxenic cultures. Five bacterial isolates were isolated from the intestine of slug hosts, identified using 16S rRNA gene sequences, and their pathogenicity tested by means of injecting directly into the haemocoel of D. reticulatum, and monitoring the mortality over time. Kluyvera sp., which was found to cause the highest mortality rate among the slugs concerned, was chosen for monoxenic culturing. Cultures containing P. papillosa and Kluyvera sp. were optimised using temperatures ranging from 15°C to 25°C, with results showing that the former was the optimum growth temperature.

The results of this project have been presented at both national and international levels. The local integrated pest management meetings, held at Stellenbosch University, act as a link between research and industry, and have proved to be a successful method for transferring knowledge to end users. In addition, outputs of this project have also been presented at Stellenbosch University's annual research day, with the PRF-funded student involved winning a prize for best presentation. Furthermore, the student also won a prize for best presentation at the 21st Nematological Society of Southern Africa conference, held in Durban in May 2017. The student's presentation of a summary of her MSc project elicited significant interest from both researchers and industry. Furthermore, the self-same student, since graduating in 2016 cum laude, has acquired funding from the CSIR/NRF to continue this project for her PhD.

The outputs of this project have been published in peer review journals. Three papers have been accepted thus far, with another currently being under review. Furthermore, two popular articles have been published, one in the African Fruit Journal, and another in the Scottish Farmer.

Now that a local nematode isolate has been identified and cultured with a highly pathogenic bacterial species, the next phase of the work is to optimise production systems, and to develop a suitable formulation method. Further pathogenicity tests are also required to determine the nematode's host range, and effect on non-target organisms. The final phase of this project is to develop an in-depth business plan for the proposed product, so as to enable applications to be made for future funding from government seed funding, and to attract future venture capital / angel investors to assist with the final commercialisation scale-up phase, following the finalising of the project.

6.10   Development and evaluation of protein and phosphorous feed ingredients from fish processing by-products

N Goosen

The research in this project aimed to develop technology to transform low value fish by-products into higher value and high quality protein and phosphorous animal feed ingredients, and to evaluate the performance of these ingredients in animal feeding studies. Optimal utilisation of limited marine resources is of global importance from economic and sustainability perspectives, and a country such as South Africa which has rich fisheries resources, can benefit from further unlocking the economic potential from its natural fish stocks.

During the course of the project, successfully optimized processing techniques were developed to recover hydrolysed proteins and bone minerals from typical South African fish processing by-products. The hydrolysed proteins and bone minerals were then evaluated as dietary sources of protein and phosphates respectively, using the African catfish (Clarias gariepinus) as model species. All experimental investigations for the entire project have been completed and the final laboratory and statistical analyses are being performed to characterize bioavailability of the novel feed ingredients, and their ability to sustain good production performance and animal health.

Results from the study have been encouraging, to say the least. Apart from the original aims, the research has also managed to produce a high-value phosphate mineral which is used in medical applications, and investigations into recovery of gelatine from the bones have been initiated. Discussions are currently under way with a potential industry partner to design a piloting facility and complete market studies on the hydrolysed protein and phosphate products. Further acknowledgement of the value and quality of the work was realised through acceptance of two conference proceedings, one to be delivered at the World Aquaculture Society Conference in Cape Town in June 2017, and the other at the Aquaculture Europe Conference in Dubrovnik, Croatia in October 2017. At both of these conferences, the nutritional aspects of the newly-developed ingredients will be presented to an international audience, and it is the belief that the conference papers will be reworked into full scientific journal articles. By the end of 2017 a total of 3 postgraduate students would have graduated through the research completed as part of this project.

6.11   Management strategies for soybean soilborne diseases in South Africa

YT Tewoldemedhin and SC Lamprecht

Seed treatment is a very important part of integrated management strategies against soilborne diseases of field crops. Surveys conducted in the major soybean production areas during 2010/11, 2011/12 and 2013/13 showed that many important soilborne pathogens are present in soybeans in South Africa. Many of these pathogens such as species within Fusarium, Pythium and Rhizoctonia affect seedling survival and establishment of soybean crops. In order to protect seedlings against these pathogens, glasshouse trials were conducted during 2014/15 and 2015/16 to evaluate fungicide seed treatments against damping-off and root rot caused by the most important soilborne pathogens. Three of the most effective treatments were selected for evaluation under field conditions. The current study therefore included the evaluation of the seed treatments Evergol (TR1), Celest XL+Apron XL (TR2), Maxim Quatro (TR3) and untreated seed (TR4) on three soybean cultivars viz DM 6.8i.RR, PAN 1454R and SSS 5052 in the cool (Bethlehem), moderate (Potchefstroom) and warm (Groblersdal) production areas. The field trials at Groblersdal and Potchefstroom were irrigated and the trial at Bethlehem was planted under dryland conditions. Soil was also collected from the trials to conduct similar tests under glasshouse conditions to evaluate the seed treatments on the three cultivars. The survival of seedlings at Bethlehem was significantly higher for treatments with Evergol and Celest XL+ Apron XL than for untreated seed. The same trend was recorded at Groblersdal and Potchefstroom although there were no significant differences in survival of seedlings from the different seed treatments at these localities. In non-pasteurised soil from Bethlehem under glasshouse conditions, all three treatments significantly improved survival of seedlings with TR1 and TR2 treatments resulting in the highest survival rates and in Groblersdal soil TR1 and TR2 also significantly improved survival, but in soil collected from Potchefstroom TR2 was significantly more effective than TR1 and TR3 to improve survival. Although the survival of seedlings was highest six weeks after planting at Bethlehem and Groblersdal for DM 6.8iRR and PAN 1454R, the yields were highest at Bethlehem and Potchefstroom. Unfortunately heavy rain and bird damage at Groblersdal resulted in very low yields in the field trial. It therefore appears that survival of seedlings is not always correlated with yield under field conditions and that other factors also affect yield. At Potchefstroom treatment of seed of cultivar DM 6.8iRR with Evergol (TR1) increased yield with 4.5% and treatment with Celest XL + Apron XL (TR2) increased yield with 17.8%. However, for cultivar PAN 1454R treatment TR1 increased grain yield with 34.5% and TR2 with 13.3%.

Overall increases in yield of the two cultivars combined at Bethlehem showed an 18.4% increase for TR1 and 6.3% increase for TR3 at Bethlehem and a 16.6% increase for TR1 and a 16.0% increase for TR2 at Potchefstroom compared to the untreated seed treatment (TR4) control. Although these increases were not statistically significant, it is biologically significant and shows the huge impact that seed treatments can have on yield, but also that the same seed treatment can have a different effect on different cultivars. Survival of SSS 5052 seedlings was significantly lower than the survival of seedlings of the other two cultivars at all the localities and it was interesting to note that the yield of cultivar SSS 5052 at Potchefstroom was significantly higher for the untreated seed than seed treated with the three fungicides. This confirms the suspicion that this cultivar was unfortunately double treated and that this caused the plots from the treated seed to have such a poor performance. Treatment of seed with Evergol (TR1) significantly reduced growth of seedlings under glasshouse conditions, especially on seedlings younger than two weeks old. However, despite the growth reduction in young seedlings, this seed treatment proved to be very effective in improving survival of seedlings and grain yield and also appears to be more effective for the control of Fusarium species that are pathogens of soybean seedlings than Celest XL + Apron XL. Soil pasteurisation and seed treatments TR1, TR2 and TR3 significantly reduced cotyledon and root rot severity for all three cultivars under glasshouse conditions. Treated seed plated to determine the effect of seed treatments on the incidence of seedborne fungi showed that, of the eight potential pathogens isolated from untreated seed, F. equiseti and P. longicolla could still be isolated from TR1 treated seed, B. maydis, F. equiseti, F. temperatum, F. verticillioides, Phomopsis sp. and P. longicolla from TR2 treated seed, and F. equiseti and P. longicolla from TR3 treated seed. Surface disinfestation eliminated many of the seedborne fungi, however, F. equiseti and P. longicolla could still be isolated from surface disinfested seed. It is also important to note that fungi that were seedborne such as F. verticillioides and P. longicolla were significantly more often isolated from seedlings planted in pasteurised compared to non-pasteurised soil which demonstrates the transmission of these pathogens from seed to seedlings. The glasshouse test conducted to determine the effect of potential pathogenic fungi isolated from seed and seedlings from treated seed in field soil showed that the fungi that significantly reduced survival were F. andiyazi, F. cerealis, F. oxysporum, P. longicolla and all the Pythium spp. Seed treatments TR1 and TR2 significantly increased survival of seedlings in soil inoculated with F. andiyazi and although seed treatment TR3 also significantly increased survival, it was less effective. Seed treatment TR2 was also less effective compared to TR1 and TR3 to increase survival in soil inoculated with F. oxysporum. For P. longicolla and all the Pythium spp. all three seed treatments TR1, TR2 and TR3 effectively controlled damping-off. In the present study none of the Rhizoctonia isolates obtained caused significant damping-off. Fungicide seed treatment is a common practice for managing soilborne, seed, and seedling pathogens. Evaluating seed treatments for control of soilborne diseases under glasshouse conditions allows the evaluation of single pathogens which is important to determine the efficacy against some of the most important soilborne pathogens. However, since these products are ultimately intended for management of soilborne pathogens under field conditions it is essential to evaluate seed treatments under field conditions. It is well-known that there is a complex of soilborne pathogens that affect soybean in field soil and that these complexes differ in different production areas and are affected by different soils and climatic conditions. From this report it is also clear that there were cultivar by seed treatment interactions indicating that certain treatments may be more beneficial to certain cultivars than others and certain seed treatments are better suited to certain production areas than others. The challenge is to identify a seed treatment that will benefit establishment and yield of most cultivars in most production areas under both dryland and irrigation systems.

During the 2017/18 season the three seed treatments will be evaluated again on three soybean cultivars in the cool, moderate and warm production areas, to confirm results that seed treatment has the potential to increase seedling survival and grain yield of soybean in South Africa.

6.12   Performance of a dual disc and tine planter, soil quality, residue management and rate of nitrogen placement with seed for canola production

PA Swanepoel, PJG le Roux and GA Agenbag

Canola is produced in the Western Cape of South Africa under Mediterranean climate conditions. Although farmers have widely adopted conservation agriculture (CA) in this area, the residue of the previous crop retained as cover becomes problematic during establishment. Stubble obstructs tine openers and farmers have started to consider disc seeders. Theoretically, disc seeders can cut through residue. The aim of this study was to compare establishment of canola with a tine and disc opener through residue on soils of high and low quality. Trials were conducted at Langgewens Research Farm (33.2765 S 18.7048 E) where CA have been implemented for the past 20 years. Three factors were assessed in a split-plot design, i.e. soil quality (high and low as whole plots), opener (tine or disc as split-plots) and residue level (low, medium and high, nestled within split-plots), replicated in four blocks. Soil quality was determined using the Soil Management Assessment Framework and Solvita soil health tests. Following planting on 25 May 2016, soil micro relief was measured. Soil disturbance, plant population density, biomass production and yield were assessed. The tine opener disturbed soil more (p<0.05) than the disc. There was a higher (p<0.05) plant population density for the tine planter at all three residue levels, which indicates more effective establishment. Canola established with the disc seeder compensated well to the end of the growing season, thus the difference (p<0.05) in biomass production at 30 days after planting between the tine and disc seeder were eliminated at physiological maturity (p>0.05). No difference (p>0.05) in yield was observed between any of the three factors. The trial will be repeated for three years before solid conclusions could be drawn.

6.13   Projected protein requirements for animal consumption in South Africa

D Strydom¹, W de Jager¹ and E Briedenhann²

The growth in the domestic availability of oilcake is a good measure by which the PRF could ascertain if it was achieving its objectives, by way of supporting the industry with research, new technology and technology transfer. The targets that will need to be met in the future for the PRF to continue to emulate the great progress that has been made thus requires projections of future oilcake demands and what will be required to obtain self-sufficiency, as well as when this goal is likely to be met. The APR model in collaboration with BFAP data is used to calculate these projections.

According to the model feed requirements will increase to 14 624 422 tons in 2025 and 12 767 149 tons in 2020 from 11 736 738 tons. Oilcake requirements will increase from 1 965 291 tons to 2 074 931 in 2020 and 2 561 503 tons in 2025. Soya oilcake requirement will be 1 537 927 tons by the year 2020 and 2 112 764 by 2025.

The poultry sector plays a major role in oilcake and particularly soya oilcake usage. Growth and sustainability in the poultry industry will play a major role in oilcake requirements.

For South Africa to achieve 79% of soya oilcake requirement self-sufficiency by 2020 the country will need to produce 1, 6 million tons of soybeans, to achieve 76% by 2025 the country will need to produce 2,1 million tons of soybeans.

6.14   Income and cost estimates of soybeans, canola and a few competitive crops

SG Ferreira

The 2015/16 research report included a complete description of the composition and functions of income and cost estimates. However, Mr SG Ferreira of Agriconcep indicated that he can no longer do the income and cost estimates due to increased work pressure. The PRF approached Grain SA and BFAP in an attempt to avoid duplication, but also to maintain the same format that all got used to over so many years. The next research report will contain more detailed information about the new approach and will highlight any changes that may become necessary.

The PRF is very grateful toward Mr Ferreira for all the years of handling this very important project on behalf of the PRF.

6.15   PRF website

M du Preez and Y Papadimitropoulos

The PRF was proud to announce its new web site going live on 1 December 2016. The outlay is more visual and colourful, with large banner photos at the top of each page, advertising new publications and interesting news on the site.

Home Page Contents

The Protein Research Foundation's new site consists of a phenomenal total of more than 800 pages, growing monthly as new information is being added. The total includes the dynamic pages such as the research database index ("Research Database"). These form other dynamic pages programmatically when following an index heading.

Another focal point was to expand the section containing the photo gallery and videos published on YouTube.

Home Page Structure

With the development of the PRF's new web site, it was upgraded to HTML5 in combination with the existing HTML4. The objective of this combination was HTML5, because it is not being implemented fully by all web browsers and search engines yet, despite being applied increasingly with each new software upgrade.

The link structure was also adjusted to be accessed faster. Most of the information on the web site contains a lot of in-depth flowing sub-sections. In the past an information page was uploaded, allowing links to be followed as the visitor delved deeper to find information. Now the visitor may delve using the links. The page information is loaded only once the required link is reached.

Search Engine Visibility

In April 2016, Google implemented its algorithm to give preference to web sites that were "mobile ready" or "mobile first". This was one of the driving forces to upgrade the Protein Research Foundation's web site. Each page was tested for mobile adaptability. This included complicated data tables.

Another important requirement of search engines, although not officially implemented yet, was to require any web site that collects information from members of the public using online forms, to install SSL certificates. Previously this was required only from sites that conducted online sales.

The easiest place to see whether a security certificate is in place when using a site, is to compare the web address protocol:

The PRF web site uses online application and subscription forms and as such the security and encryption certificates had been installed on the PRF domain.

The implementation of the Protein Research Foundation's new web site also involved a refined SEO ("Search Engine Optimisation") according to the latest known search engine algorithms. These include, inter alia, better defined file names for photos, diagrams and figures, programmatic photo descriptions (not visible to general web site visitors) and built-in meta photo identification and definitions ("embedded meta data").

Visitor Statistics

One of the measurements to determine user value of the web site is to determine the number of visitors and time spent on the sites, including number of pages visited. The statistics below indicate the number of visitors per reporting year since inception of the website.

* Raw values indicate total interactions measured by the web server, while Google values only measure the visitors that link to the web site using a web browser. Visitors that link to the web site again, using the same browser are not counted again.