This study was undertaken to evaluate genotype x environment interaction (gxe) and yield stability of eight bambara groundnut genotypes in three locations at two planting dates; to compare the stability parameters used in determining stable yields, to correlate yield and related characters of the crop, to determine the effect of planting date on yield and yield components; and to assess the effect of location and genotype on protein content. Field trials were planted with two different planting dates in three localities. A randomized complete block design with three replications was used. Data collected was days to 50% flower, number of pods per plant, grain yield, haulm yield, 100 seed weight, and maturity days. Data was subjected to simple ANOVAâs for all measured characteristics. Significant differences were found for most of these characteristics. Combined analyses of variance were computed across locations to determine the performance of yield and related characteristic. Significant differences were found for genotype, environment and Gxe interactions. The best genotype was SB4-4 across locations. Four stability parameters namely Eberhart and Russell regression model, Lin and Binnsâ Cultivar Superiority Measure (Pi), Wricke ecovalence (Wi) and Additive Main Effects and Multiplicative Interaction (AMMI) were performed to determine yield stability. SB16-5A was regarded the most stable by Eberhart and Russell, SB19-3 by Wricke ecovalence, SB7-1 by ASV and SB4-4 by cultivar superiority and yield rankings The results showed that SB19-3 was the third most stable genotype according to Eberhart and Russell and Lin and Binns yield ranking; and the first and second most stable genotypes by the Wricke and AMMI model. Therefore SB19-3 proved to be the most stable genotype. ASV and Wricke ecovalence value ranked SB4-4 and SB1-1 to be the first and second most stable genotypes for 100 seed weight, while ranking SB19-3 first and second respectively for number of pods per plant. The AMMI model summarizes patterns and relationship between gxe interactions and helps to obtain a good yield estimates. The results for stability of yields are not conclusive since the data is for one season, therefore the trials must be repeated to validate the results. Correlation analyses were first computed for separate trials and then for combined trials across locations. Hundred seed weight, haulm yield, number of pods per plant and root weight were positively correlated with grain yield with haulm yield and 100 seed weight having strong positive correlations to grain yield. Planting date did not significantly affect yield and protein content was also not significantly affected by gxe interactions.

Wheat (Triticum aestivum L.) is widely cultivated over large areas and is an important food crop worldwide. Wheat is extensively used during the production of many different types of foods. Wheat rusts (leaf, stem and stripe rust) are important foliar diseases of wheat worldwide, causing large losses and damage to the wheat industry. The ability of these rust pathogens to change and be dispersed over long distances pose a continual global threat. Annually millions around the world are spent on fungicides in an attempt to control wheat rusts. Resistant cultivars have proven to be the most effective, economical and environmentally friendly form of rust control. Although many resistant cultivars have been developed historically, a need for more durable resistance exists. The application of molecular markers and marker-assisted selection (MAS) strategies in breeding programmes can support plant breeders in accomplishing pyramiding of several rust resistant genes into new cultivars. The aim of this study was to pyramid seven rust resistant genes/QTL (leaf, stripe and stem rust) into a single genotype using five SSR, three STS and two AFLP markers. The study focused on wheat genes applicable to the local wheat industry and markers used and/or developed in South African breeding programmes. In this study four bread wheat cultivars or lines (AvocetYrsp, Blade, CSLr19-149-299 and Kariega) were used as parental sources of five resistance genes (Sr2, Sr26, YrSp, Lr19 and Lr34) and two QTL (QYr.sgi-7D and QYr.sgi-2B). Selection after each cross was done using a MAS approach with SSR and STS markers linked to the different resistant genes/QTL. The study was conducted over a two year period, involving the development of two different sets of F1 offspring and one double cross generation from a series of directional crosses. Before crosses were made, the presence of the expected rust resistance genes was confirmed in the parental lines using specific SSR, STS and AFLP markers. The SSR and STS markers amplified the expected allele sizes in the parental lines, except for the unexpected detection of the Lr34 gene in AvocetYrSp. Results indicated that the Sr2 marker was not consistently present in Blade, suggesting the Blade cultivar was heterogeneous for Sr2. The AFLP markers linked to the YrSp resistance gene did not detect differences between the parental lines and were excluded from further experiments. The F1 generations were screened with one SSR or STS marker each to identify successful crosses. Genotyping of the F1 generations indicated than on average, 85.5% of the tested F1âs were true hybrids. Phenotypic screening was done on the parental lines and F1 individuals for the three rust types and confirmed the presence of the expected genes in the parental lines as well as selected F1 individuals. A total of 900 individuals of the double cross generation were screened with five SSR and three STS markers associated with resistance genes and QTL to identify whether gene pyramiding within a single genotype was successful. The number of individual plants of the double cross population containing markers linked to the desired resistance gene(s)/QTL ranged from two individuals containing none of the markers to three individuals containing all eight markers. The three individuals containing eight markers confirmed the presence of markers associated to the presence of the four single genes (Sr2, Sr26, Lr19 and Lr34) and two QTL (QYr.sgi-7D and QYr.sgi-2B.1). Due to the inefficiency of the AFLP markers, the presence of the seventh gene (YrSp) could not be confirmed on genotypic level. The future uses of the developed rust resistance gene pyramided lines of this study are countless. The use of these lines in combating the continual threat of wheat rusts in some manner should be helpful in future.

Stripe rust, caused by Puccinia striiformis West. f. sp. tritici is one of the most damaging diseases of wheat (Triticum aestivum L.) globally. The South African wheat cultivar Kariega expresses APR and has retained yield levels acceptable for commercial production, which is of great importance to plant breeders. A Kariega x Avocet S partial linkage map has made a significant contribution to understanding the genetics underlying APR to stripe rust (Yr) in Kariega. Two major YrQTL with indications of different resistance mechanisms were identified on chromosomes 2B and 7D. In this study we investigated the effectiveness of identifying AFLP markers closely linked to the YrQTL using a targeted bulk segregant analysis (tBSA) approach in F1 doubled haploid (DH) individuals. Individual Kariega x Avocet S DH lines were characterised and bulked based on stripe rust phenotypes and DNA marker allele profiles. In agreement with standard BSA, an extreme resistant bulk (both QTL present: +7D+2B) and extreme susceptible bulk (both QTL absent: -7D-2B) were constructed based on phenotypic data and verified with marker allele data. Three additional bulks (+7D-2B; -7D+2B and narrow down +7D±2B with marker recombinations in 7D QTL interval) were constructed based on a combination of phenotypic and marker data, with a strong emphasis on the presence or absence of marker alleles representing a specific QTL interval as required by a specific bulk. A total of 184 AFLP primer combinations (SseI and MseI) were tested on the two parental lines and five bulks. Thirty-one of these primer combinations detected 32 putative markers that could discriminate between the extreme resistant and susceptible bulks and that were putatively linked to either the 7D or 2B QTL regions. After validation of these markers on all individuals used in the extreme resistant and extreme susceptible bulks, nine markers were identified that were present in the extreme resistant and the specific -7D+2B bulk, but absent in the extreme susceptible bulk. Another two markers were identified that were present in the extreme resistant, +7D-2B and narrow down +7D±2B bulks, but absent in the extreme susceptible bulk. These markers were mapped onto the existing Kariega x Avocet S partial linkage map using Map Manager QTXb20. Six AFLP markers mapped within or close to the QYr.sgi.2B and one close to the QYr.sgi.7D QTL regions. The tBSA approach was efficient since 10 of the 11 markers (91%) putatively identified after screening of the individuals constituting the bulk samples mapped to either chromosome 2B or 7D. AFLP analysis in combination with tBSA was shown to be reproducible, faster and a more cost effective approach compared to a traditional BSA since tBSA lead to a reduction of 28.2% of primers that need to be tested. Following the tBSA approach, marker s23m53d mapped 3 cM from marker gwm148 previously shown to be significantly associated with mean host reaction type for final field data as well as leaf area infected of the QYr.sgi-2B QTL region. This resulted in an increase in LOD score from 20.1 to 23.9 using interval mapping. Even though two markers were added to the 7D chromosome, both mapped outside the QYr.sgi-7D QTL region. Marker s20m38b mapped 9 cM from the SSR marker gwm295 and 20 cM from the Ltn gene previously shown to be associated with the trait of interest on chromosome 7D. In summary, the combination of AFLP analysis and a tBSA approach has proved to be useful in the identification of QTL, the placement of closely linked markers to known QTLs and targeting chromosome areas with low marker numbers. Indications are that a large number of AFLP primer combinations need to be screened to successfully target a specific QTL interval for increased map resolution.

In a second experiment, the NUE components of seven parental irrigation cultivars and their F2-offspring were studied in 2004 at a cooler (Bethlehem) and a warmer (Vaalharts) irrigation environment. The genotypes were tested at two N treatments; a control (LN) and 180 kg N/ha (HN) applied. Agronomic characteristics, N uptake and NUE components were measured. From the analysis of variance, the agronomic characteristics (BM, grain yield, HM, TKM and GP) showed significant genotypic variation. Harvest index, BM, HM and TKM showed no significant genotypic variation at the N treatment. The N uptake and NUE components showed significant genotypic variation with the exception of NutEYld. The HN treatment resulted in significant variation in Ngrain, Ntotal NHI, NUEYld and NutEYld. The HN treatment increased BM, grain yield and GP values, but reduced the HM, TKM, HI and NHI values of the genotypes. The NUE components NupE%, NUEYld and NutEYld were also reduced by the HN treatment. Cultivars responded differently at the two N treatments in measured and calculated agronomic and NUE components. Olifants, Kariega and Marico showed good responses for the agronomic components at the HN treatment. Inia, SST 806, SST 822, Olifants, Kariega and Marico showed high values for NUE components. The crosses Olifants x Steenbras, SST 806 x SST 822, Marico x Steenbras, Olifants x SST 822 and Marico x SST 806 showed positive responses in the measured agronomical and NUE components. The correlation coefficients were calculated to study the phenotypic resemblance between characteristics. The calculated correlation matrixes for the environments indicated positive correlation coefficients between several of the agronomic and calculated components. Biomass and grain yield was highly correlated with the NUE components, NupE%, NUEYld, Ngrain and Ntotal at both environments. Grain protein was negatively correlated to NutEYld, while NHI was positively correlated to this component. This points to the first order importance of grain yield production before grain protein under limiting N availability conditions. The correlation coefficients of components BM, grain yield, NPE, NRE and NUEYld increased at the HN treatment. The effect of N treatments on the calculated GCA and SCA effects was studied from a diallel analysis. Significant GCA effects for the parental cultivars were found for the majority of the agronomic and NUE components at both N treatments. Significant SCA effects for the tested hybrids were also found at BM, grain yield, Ngrain, Ntotal, NupE%, NUEYld, NAEYld, PHE and NRE. The application of N fertilizer (LN versus HN treatments) affected the significance of the calculated combining abilities. For the agronomic characteristics that were measured at Bethlehem, GCA effects for BM, HM, TKM, GP and HI were significant at the HN treatment. Significant GCA effects were found at the LN treatment for TKM and GP. The N uptake components Ngrain, Ntotal and NHI showed significant GCA effects at the HN treatment. For the NUE efficiency components significant GCA effects were calculated for NupE%, NUEYld and NutEYld at the HN treatment. No significant cultivar differences were found for the N uptake and NUE components at the LN treatment. For the agronomic components SST 822, Inia and Kariega generally had the highest GCA values. The cultivars Marico, Olifants, SST 806 and Inia showed the highest GCA values at the HN treatment for the agronomic, N uptake and NUE components. At Vaalharts significant variability for GCA effects for the agronomic components BM, grain yield, HM, TKM and HI were found at both N treatments. For the N uptake components, significant GCA effects were calculated for Ngrain and Ntotal at both N treatments and for NHI at the HN treatment. Significant GCA effects were found for the NUE components NupE%, NUEYld at both N treatments and for NutEYld at the HN treatment. The cultivars Kariega, Olifants, Inia, Steenbras and SST 806 had the highest GCA values for the agronomic, N uptake and NUE components at the LN treatment. At the HN treatment Olifants, SST 806 and Inia had the highest GCA values for these components, as well as for the N uptake and NUE components. The calculated SCA effects at Bethlehem were significant for BM, grain yield, Ngrain, Ntotal, NUEYld at both N treatments, and for NAEYld, NPE, NRE, HM and NupE% at the HN treatment. At Vaalharts significant SCA effects were calculated for BM, grain yield, Ngrain, Ntotal, NupE%, NUEYld at both N treatments and for NHI at the HN treatment. The calculated GCA:SCA ratios for the respective components at Bethlehem showed that the values for most of the components increased at the HN treatment, with only grain yield, Ngrain and NUEYld at values less than one. The calculated NUE components NAEYld, NPE and NRE also showed non-additive gene action. At Vaalharts all the components showed additive gene action at both the LN and HN treatments, except for GP, NHI and NutEYld at the LN treatment. Also, in contrast to the response calculated at Bethlehem, the calculated NAEYld, NPE and NRE showed additive gene action. These differences in responses between the two environments can be attributed to the different growing conditions and N availability scenarios inducing the changes in response of agronomic and NUE components. High broad-sense heritabilities were calculated for BM, grain yield, HM, TKM and HI at the LN and HN treatments at both environments. At Vaalharts the heritability value for GP was high at the HN treatment, and at Bethlehem at both the LN and HN treatments. Broad-sense heritability estimates improved at the HN treatment except for TKM compared to the LN treatment. The heritability values for the N uptake components, Ngrain, Ntotal and NHI were high, and were increased at the HN treatment. With the exception of the NutEYld component at the LN treatment at Vaalharts, the broad-sense heritability values of all the NUE components were high and increased by the HN treatment. At Bethlehem the narrow-sense heritability value of TKM was high at both N treatments, and values of GP, NHI and NutEYld at the HN treatment. At Vaalharts high narrowsense heritability values were found for HM, TKM, NAEYld and NPE at the LN treatment. All the agronomic characteristics and calculated components showed increased narrowsense heritability values at the HN treatment. These responses indicate that the estimation of heritability of the studied components were improved when N were sufficiently available.

The ability of wheat flour to be processed into bread and other products is largely determined by the gluten proteins, which confer unique visco-elastic properties to dough. These proteins are influenced by genetic make up and the environment in which the plants are grown. The objective of this study was to gain a better understanding of the influence of extreme temperatures on the gluten proteins, quality characteristics, starch and starch components. Two bread wheat cultivars, a soft biscuit wheat and a durum wheat were grown two plants to a pot, with 12 pots for each replication, with three replications in a greenhouse for two consecutive years. Cold and heat treatments were applied during grain filling. Size exclusion and reverse phase HPLC were used to measure the different protein fractions. Starch, amylose, amylopectin and various quality characteristics were measured on all entries after harvest in both years. The results from this study showed that in both years the softest cultivar, Snack showed the largest reaction to low temperature stress specifically, where the monomeric proteins were significantly increased, and the polymeric proteins were significantly decreased. In the second year SST86 showed the same trend as Snack under cold stress conditions. Oranje, the durum wheat cultivar, reacted to both heat and cold stress in the second year, where the small polymeric proteins (SDS soluble) were significantly reduced, and the small monomeric proteins were increased with stress, and the SDS insoluble large monomeric proteins were increased only under heat stress. The protein fractions of the tetraploid Oranje were very different from that of the other three hexaploid cultivars. The soft wheat Snack had significantly lower large polymeric protein, and higher large monomeric protein fractions (SDS insoluble) than the bread wheat cultivars for both years. Summed over cultivars, per treatment for the two years, all the fractions, except for the small polymeric and small monomeric proteins (SDS insoluble) were affected by the cold treatment. The monomeric proteins were consistently increased, and the polymeric proteins decreased. Only the SDS soluble large polymeric proteins were significantly affected by the heat treatment, where it was significantly decreased. Both kernel weight and diameter were significantly decreased at both stress treatments for the two bread wheat cultivars, indicating that their kernel characteristics were sensitive to extreme temperatures. Across the two seasons, the starch content in Kariega was significantly reduced by both heat and cold temperatures, as is also reflected in the reduction of kernel weight and diameter. Amylose content and the amylose:amylopectin ratio increased under heat and cold treatment. There were highly significant positive and negative correlations between certain RP-HPLC peaks and quality characteristics for all three the temperature treatments in both years. The number of significant correlations was reduced under the two stress conditions, but there is a possibility to use certain peaks for quality selection. The profiles must, however, be standardised so that peaks can be scored and compared on any gel according to their elution times. Stepwise regression also emphasized the importance of some of the RP-HPLC peaks in explaining variation in flour protein content and SDS sedimentation for all three temperature treatments.

The aims of this study were to evaluate the functionality and the practicality of hydroponics in the process of wheat breeding compared to a conventional glasshouse cultivation method. Furthermore for assessing the screening capability of drought tolerance in wheat cultivars using a hydroponic system was used and morphological yield components were measured and protein concentration fluctuations by means of SE-HPLC were analyzed. x In the first study, five South African wheat cultivars, SST 88, Baviaans, Steenbras, SST 876 and Kariega, were planted into two identical hydroponic systems. A second set of the five cultivars where planted two months after the first planting to separate the two stress periods, the first being drought stress before seed fill and in the second planting, stress was induced at seedling stage. This was done simultaneously. x When maturity was reached, several yield component measurements were taken. x The influences of the drought treatment on the different yield components were established. x Protein extracts of wheat flour were analyzed by SE-HPLC. x SDS was used for the first protein extraction step, and the rest of the proteins were extracted by sonication. 89 x The SE-HPLC method fractionated the storage proteins (both SDS-soluble and SDS-insoluble) into four distinct peaks of decreasing molecular size range, representing mainly larger polymeric proteins (mainly HMWglutenins), smaller polymeric proteins (mainly LMW-glutenins), larger monomeric proteins (mainly gliadins), and smaller monomeric proteins (mainly albumins and globulins). x Relationships between the amount and size-distribution of polymeric and monomeric proteins and flour quality properties were established. x The influence of the drought treatment on protein fractions and quality characteristics was also determined by SE-HPLC. x The results showed that both the genotype and drought conditions had a significant influence on the yield components and protein characteristics. x The use of a hydroponic cultivation method for the induction of drought conditions at any stage of development of wheat can be done at a highly significant level, not only for inducing stresses but for re-establishing and maintaining optimum conditions as well. x SE-HPLC provided a simple and an objective test for measuring the relative size-distributions of wheat storage proteins. x In the second study two South African wheat cultivars, Baviaans and Steenbras, were planted into two identical hydroponic systems (differing only in the makeup of the nutrient solution) and in pots. x Emergence percentages and dates were recorded for a one month period after the emergence of the first seedling in the experiment. x When maturity was reached several yield component measurements were taken. 90 x The two hydroponic systems provided a better emergence environment compared to the potting solution, indicating that there is already a higher potential number of crosses that can be made in the breeding program using a hydroponic cultivation method. x The effect of treatment was highly significant for most of the yield characteristics, indicating that the effectiveness of the treatments played a significant role in the yield potential of each cultivar. x Findings indicated that by using a hydroponic cultivation method instead of the potting method for growing wheat, parent lines can increase the amount of crosses that can potentially be made, thus increasing the amount, potential success and quality of the crosses made, together with the potential of producing seeds that will have the potential of a higher emergence rate. x The hydroponic cultivation method using the chemicult solution proved to be the most effective method of cultivation during this study.

Early leaf spot (ELS) caused by Cercospora arachidicola, late leaf spot (LLS) caused by Cercosporidium personatum, web blotch (WB) caused by Phoma arachidicola and rust caused by Puccinia arachidis, are serious diseases of groundnut in South Africa. The aims of this study were to evaluate ARC-GCI germplasm for resistance to the important foliar diseases such as ELS, LLS, WB, and rust and to ascertain if cytoplasmic factors influence the pattern of inheritance of resistance to ELS, LLS, and WB, testa colour and mutations such as the one responsible for cupleaf phenotypes. Twenty-one reciprocal crosses were made to study the cytoplasmic factors but the F1 progenies gave no evidence of cytoplasmic factors directly responsible for any of the above-mentioned traits. During 2003/04, 138 ARC-GCI germplasm entries were evaluated for resistance to ELS, LLS, WB and rust at Potchefstroom (North West Province; Highveld), Burgershall (Mpumalanga Province), Brits (North West Province), Vaalharts (Northern Cape Province) and Cedara (Kwazulu Natal Province). No fungicides were used to control the foliar diseases. In trials at Vaalharts, Cedara, and Burgershall entries 1, 10-13, 18-21, 23, 30, 52 and 56-58 were resistant to two of the three (LLS, WB, and rust) foliar diseases and entries 18, 19 and 20 (best) to LLS, WB and rust. At Brits, Elite (E) and ICRISAT (I) short/medium growth entries 8I and 7E, were resistant to LLS and WB and the long growth entries 1E and 3E were resistant to WB, LLS and rust. In micro-plots at Potchefstroom entries 6, 9, 10 and 13-15 were resistant to LLS and WB and in the brick plots 1-3, 6 and 7 to ELS and LLS. Entries with resistance to ELS, LLS, WB, and rust as well as other favourable traits, such as a high yield and oleic acid percentage, from the 2003/04 trials were included in the 2004/05 trials for further assessment. All the entries were evaluated for resistance to ELS, LLS, WB and rust, were graded and grouped by using the Canonical Variate Analysis (CVA). From the results on the five locality trials (2004/05) medium growth entries 1, 3, 7, 10, 14, 15, 20-22, 26, 27, 30, 32 and 33 were resistant to three or four of the foliar diseases, gave good yields and choice grade kernels. The CVA identified separate groups, containing entries 3, 20, 21 and 32, with resistance to two or more foliar diseases and with high yield. No choice grade entries were reported for unsound blemished and soiled (UBS) kernels with an UBS>13%. In the Elite short/medium trials at Brits (2004/05) entries 3 and 8 were resistant to rust, 1, 2, 4, 7, 11-13, 15, 17, 19-22, 24 and 25 gave choice grade kernels, and 10, 13 and 12 the highest yield. Entries 1-6, and 8-15 (Elite long) were resistant to LLS, 1-15 to WB, and 1, 4 and 15 to rust, 4, 10 and 12 gave choice grade kernels and 4, 1 and 10 the highest yields. ICRISAT (long) entries 5 and 7 gave choice grade kernels, and 1, 6 and 3 the highest yields. The CVA identified the groups containing medium growth entries 10, 12, 13, 22 and 23, Elite long 1, 4, 10 and 15 and ICRISAT long 5, 8 and 9 as groups with resistance to two or more foliar diseases (ELS, LLS, WB, and rust) and good yield. Twenty-three medium and 11 long growth entries in the Elite trials at Vaalharts (an extra trial planted where fungicides were used to control foliar diseases) gave choice grade kernels and good yields. In this study, some ARC-GCI germplasm entries were resistant to ELS, LLS, WB, and rust, but climatic factors influenced the severity of infection. Foliar diseases lowered the quality and grade of the kernels and resulted in lower yields. CVA simplifies the evaluation and grouping of new germplasm entries and will save valuable research resources.

A company, Elementol (Pty) Ltd, requested the evaluation of their novel product, Pheroids. Pheroids can apparently facilitate the transport of phytological beneficial substances over membranes. Information regarding the chemical attributes was withheld as patent registration is still pending. Pheroids is apparently a microemulsion containing free fatty acids (FFAâs) and or fatty acid derivatives. It apparently encapsulates a substance and facilitates its transport over the membrane. The exact mechanism involving encapsulation, transport and release of the substances inside the cells is still vague due to little information available on it. The aim of this study was to evaluate the ability of Pheroids to facilitate the transport of additional nitrogen, urea in this case, in tomatoes grown under nitrogen limiting conditions. Tomatoes (Lycopersicon esculentum cv. Rodade Star) were cultivated in a greenhouse using a circulating ebb and flow hydroponic setup, which supplied the plants with either a control- or nitrogen limiting nutrient solutions. The plants cultivated in the nitrogen limiting conditions showed a remarkable reduction in vegetative development and yield. To alleviate the effect of nitrogen limiting conditions on yield, the plants were foliarly sprayed with 0.5% and 1% urea solutions, singly or mixed with Pheroids, once every two weeks. The purpose of these foliar treatments was to determine whether Pheroids can further enhance the absorption and transport of urea across membranes of the leaves to alleviate the effect of limiting nitrogen supply. Plants grown under nitrogen adequate conditions (control) were also foliarly treated with a 0.5% urea solution, singly and mixed with Pheroids, to determine to which extent control plants react to the additional nitrogen supplied. The reduction in yield, as a result of limited nitrogen supply, was partially alleviated by spraying nitrogen deficient plants with the 0.5% and 1% urea solutions. However, mixing the 0.5% and 1% urea solutions with Pheroids, not only improved vegetative growth and generative development, but also improved yield, suggesting that Pheroids indeed has the ability to improve the uptake of urea. The 0.5% urea / Pheroids solution specifically proved to have the best ability in alleviating the effect of nitrogen limiting conditions on yield without compromising fruit quality. Although the reducing effect was not completely alleviated, the yield and loss in income as a result of nitrogen limiting conditions was prevented to a large extent. Spraying control plants with 0.5% urea, singly or mixed with Pheroids, also improved yield, without compromising fruit quality. In addition, Pheroids itself, without mixing it with any substance, also resulted in increased yields in both control- and plants grown under nitrogen limiting conditions. In summary, it appeared that Pheroids has the ability to facilitate the transport of phytological beneficial substances, in this case urea, over plant membranes and enhances cellular nitrogen content, but this needs further detailed analyses. This phenomenon was more evident in plants grown under nitrogen limiting conditions than in plants grown under control conditions. Taking into consideration that most crops frequently may suffer from nitrogen limiting conditions in standard agricultural practices, Pheroids may have numerous potential applications in the agricultural industry.

The purpose of this study was to investigate the effect of variations in climatic factors such as vernalization periods, minimum- and maximum temperatures and rainfall on the expression of host plant resistance to RWASA1 (original biotype) and RWASA2 (newly identified biotype) in the eastern Free State of South Africa between 2003 and 2006. Climatic records indicate above-average temperatures and reduced rainfall occurred in the Eastern Free State in 2003 to 2005 probably contributing to widespread RWA damage on dryland wheat crops. Greenhouse experiments confirmed that vernalization did not alter RWA plant resistance ranking of wheat entries. However, infestation did produce significant damage by RWASA2 on Elands, Gariep and Limpopo but not on CItr 2401 (containing Dn-4 and another unidentified Dn-gene) and Halt (Dn-4). Resistance rating to RWASA1 of Halt, Komati and Matlabas in greenhouse tests at 18°C/12°C, 22°C/16°C and 26°C/22°C shifted to less resistant- or susceptible categories and changed the ranking of SST 966 as temperature increased. Also, RWASA2 ratings at increased temperature shifted Halt to a more resistant category and changed rankings of Komati and SST 399. Komati, Matlabas, SST 966 or SST 399 did not express any host plant resistance to RWASA2 through leaf area or leaf roll and chlorosis supporting the susceptible ratings that were given to the same varieties in the visual screening test. Seeddressing in AMMI results for 2003 to 2006 were able to increase yields under higheryielding conditions in seasons when moisture- and heat stress occured and the yield performance of cultivars was stabilized irrespective of whether they contained genetic resistance to RWA. Identifying new RWA resistance donors provide an excellent opportunity to exploit novel germplasm with potential new genetic variance for RWA resistance, as well as other beneficial traits such as drought-tolerance. Three germplasm pools were evaluated for this purpose. Twenty-one entries from the Stillwater, Oklahoma/SGI RWA resistance pool expressed resistance against RWASA2 in the seedling test and where generally well adapted to local conditions. Seedling evaluation of the CIMMYT pool revealed twenty entries with acceptable resistance to RWASA2 though field adaptability particularly to biotic stress was occasionally lacking. Twenty entries from the Iranian pool were resistant to RWASA2 though unadapted for resistance to stripe rust; Puccinia striiformis Westend f.sp tritici Eriks., occurring in the Eastern Free State.

The aim of this study was to study various aspects of the defence response of wheat infected by Puccinia triticina. SSH was used to isolate differentially expressed cDNA fragments from P. triticina infected wheat. Once sequenced, two cDNA clones were selected for further analysis. One of the cDNA clones, LRW268, showed similarity to an EST derived from de-etiolated wheat. It showed a 16.5 fold induced expression at 9 hpi in the P. triticina infected wheat. In P. striiformis infected wheat, an induced expression was found in the resistant Avoset-Yr1 plants. Application of chemical elicitors showed an induction of LRW268 expression for plants treated with H2O2, MeJA and menadione. To obtain the full length gene, 5â-RACE was attempted, but it was unsuccessful. However, a contig was assembled using ESTs present in the GenBank database. This yielded a 603 bp contig encoding a 96 amino acid sequence that showed good homology to a RLK interacting protein. The presence of a putative MAPK docking motif and a phosphorylation site indicated that LRW268 could play a role in cell signalling. The presence of LRW268 in different wheat cultivars signified that it could form part of the general defence response of wheat. The second cDNA clone, LRW222, which was similar to an EST from wheat infected with powdery mildew, showed a 4.6 fold induction of expression at 15 hpi in wheat infected with P. triticina, while the susceptible Thatcher cultivar showed an induction at an earlier time interval. P. striiformis infected wheat showed a more constant expression of LRW222. Putative induced expression of LRW222 was observed in H2O2, MeJA and menadione treated wheat. Assembly of a contig using published ESTs yielded a 668 bp contig which encoded an 89 amino acid polypeptide showing homology to various wound-induced protease inhibitors. The presence of a putative MAPK docking motif on LRW222 suggested that it could be a general or pathogen specific protease inhibitor. The effect of light on the wheat defence response was also examined. The photosynthetic capacity of all treatments was measured and fluorescence microscopy performed. Infection caused a decrease in the photosynthetic capacity of the susceptible plants with the resistant plants showing less fluctuation. The infected resistant plants recovered faster and better than the infected resistant plants after the dark incubation. Plants that were additionally dark incubated showed a lower photosynthetic capacity compared to the control treatments. This difference in photosynthetic capacity was not observed on molecular level with photosynthesis related genes showing unaltered expression. The putative expression of certain defence related genes did however show a light dependency. An induced defence was observed in the uninfected plants, confirming a putative volatile signalling event that was detected during an earlier study involving this particular plant/pathogen interaction. The importance of light in wheat resistant towards P. triticina could thus be attributed to the ability of plant to photosynthesise optimally. It must however be emphasized that the expression of all tested genes were not quantitatively determined, since end-point analysis using RT-PCR was used. Future research will include the use of techniques that allows quantitative measurement of gene expression.