The main goals of this study were to ascertain farmers' preferred traits in rice landraces and their perception of Rice yellow mottle virus, to collect rice landraces across Burkina Faso, investigate their genetic diversity, and to exploit this diversity in a search for varieties resistant and tolerant to RYMV, for their utilisation in rice breeding. Farmers' preferred traits, approaches to crop management, and disease perceptions were assessed using a Participatory Research Appraisal (PRA) approach. In the main rice growing regions of Burkina Faso, 330 rice landraces were collected. The agro-morphological diversity of the germplasms was evaluated in the field with 20 quantitative and 30 qualitative agro-morphological parameters. Thereafter, 22 Simple Sequence Repeat molecular markers were used to assess the genetic diversity and the population structure of the collection. Finally, the rice landraces were screened against four RYMV isolates to assess the susceptibility, tolerance and resistance of the landraces in the collection using visual assessment and Enzyme Linked Immunosorbent Assay. The PRA identified sweet taste, grain expansion when cooking, easy cooking and yield as paramount selection criteria in rural rice farming communities in Burkina Faso. Drought and disease resistance are characters that farmers wish to have in their varieties. The PRA also highlighted that farmers are conscious of RYMV disease in their fields. However, they are unaware about the epidemiology of the disease. An agro-morphological study of the phenotypic diversity of the collection confirmed the presence of the two cultivated rice species: O. glaberrima and O. sativa. There were more O. sativa accessions than O. glaberrima landraces. There were 48 O. glaberrima and 282 O. sativa accessions in the collection. Both species were divided into four clusters, reflecting the richness of the collection. The underlying genetic diversity of the collection was confirmed by the use of 22 Simple Sequence Repeat molecular markers. The neutral markers confirmed the existence of two substructures, namely O. glaberrima and O. sativa, and the presence of admixture varieties. However, a core collection of 52 individuals was developed. This included 13 O. glaberrima and 39 O. sativa accessions. It reflects the genetic diversity of the sub-clusters present in each species. This core collection contains 89% of the allelic richness of the collection. Its small size will facilitate the maintenance and active use of diversity of germplasm in the core collection. The entire collection was utilised to search for varieties resistant and tolerant to RYMV disease. The screening of the collection with different RYMV isolates exposed the susceptibility of most of the accessions in the collection. Most of the O. sativa indica accessions were highly susceptible. However, ten O. glaberrima accessions displayed a delay of symptom expression, and moderate resistance. However, their resistance was overcome later by a particularly virulent RYMV isolate BF1. Remarkably, a single moderately resistant cultivar, BM24, showed that partial resistance and tolerance to RYMV can be found in an O. sativa variety. Serological evaluation of this local variety in comparison with the partially resistant variety, Azucena, showed that BM24 and Azucena expressed similar resistance patterns. A genetic profile of both varieties showed that both had an identical allele status at RM101, which is a marker bracketed in the same zone as the QTL12. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2011.

Rice is a staple food in many West African countries, including Niger. However, both regional and national rice production have failed to meet demand due to several constraints, among which is the Rice yellow mottle virus (RYMV). Moreover, attempted intensification of rice cultivation and the introduction of modern cultivars are encouraging farmers towards abandoning local landraces for high yielding, but often susceptible varieties. The study was primarily oriented towards rice pre-breeding, and identifying priorities for rice breeding in Niger in relation to farmers' preferences and their environment. A secondary aim was the development and evaluation (for release at the regional level) of new breeding lines with resistance to RYMV. This study aimed to: 1) Establish farmers' perception of rice varieties as well as the main constraints on rice production in Niger and particularly those posed by RYMV; 2) Create a collection of rice species from Niger for ex- situ conservation, and to determine the phenotypic variability within this collection; 3) Determine the genetic diversity and population structure of the collection; 4) Screen the collection for resistance to RYMV, so that new sources of resistance could be detected; 5) Improve five elite varieties from West Africa for resistance to RYMV using marker-assisted selection (MAS). The germplasm collection and PRA of this study were conducted in 2008 and 2009 in Niger, while the field and the laboratory researches were conducted in 2008 and 2009 at the Africa Rice Center (AfricaRice) in Benin. For the PRA, data was obtained from a semi-structured group discussion carried out in 14 villages, individual questioning of 153 farmers and visits to farmers' field and storage facilities. The local farmers' union was the only formal seed dissemination system. Seed exchanges between farmers and the use of seeds from previous harvests were important. The RYMV and the bacterial leaf blight (BLB) were cited as the prevalent biotic stresses in the irrigated agrosystem, where the varieties IR1529-680-3 and Waihidjo were found to be the most popular. Flood, birds and hippopotamus were the most damaging agents in the lowland cropping system, and the landrace Degaulle/ D5237 was the preferred variety. Apart from the yield, farmers preferred varieties with good grain quality (milling quality and good taste), high market value, stress tolerance (drought, flood, disease, birds, rodents), and those recommended by the local farmers' association. These findings should be included in breeding goals, seed production and dissemination systems. During collection, a total of 270 rice accessions were assembled, comprising the two cultivated rice species Oryza sativa L. and O. glaberrima Steud. and its two wild relatives Oryza barthii A. Chev. and O. longistaminata Chev. et Roehr. The region of the Niger River and its tributary (the Dallol Maouri) provided the majority (80.7%) of the accessions. Apart from a few wild O. barthii accessions, the accessions found around Lake Chad and the Komadougou river (South-East) were also collected in the Niger River area. Farmers' naming and ecological classification of rice varieties was generally consistent. Three major phenotypic groups were found during the field trials, and the overall phenotypic variability of the collection (as measured by the Shannon-Weaver Diversity Index) was relatively high. There was no significant difference in diversity between the main eco-geographical zones of collection, as well as between the identified phenotypic groups, suggesting a high level of germplasm exchange between the regions in Niger. From the collection, 264 accessions were genotyped from the collection using 18 well distributed SSR markers and two main genetic compartments were detected, comprising O. sativa subsp. indica varieties and O. glaberrima and its wild relative O. barthii and O. longistaminata. The O. sativa group in Niger was divided into irrigated and floating rice, bound by lowland rice. The wild progenitor O. barthii was widespread but without any clear genetic differentiation from O. glaberrima, probably due to the presence of admixtures within the collected samples of O. barthii. Allelic diversity was relatively high, despite the geographical distance from the centre of domestication of African rice, and the points of entry of Asian rice to Africa. The findings reflect the underuse of Niger's rice landraces genetic potential for rice breeding, given that all the "improved" varieties released during the last 25 years in Niger were clustered together on the dendrogram. The response of a set of the rice collected from Niger and some accessions from Mali to inoculation by RYMV was evaluated using five different virus isolates from Niger (3), Benin (1) and Burkina Faso (1). All rice varieties were susceptible to the disease. However, depending on the virus strain, a few O. glaberrima accessions displayed partial resistance, similar to the highly resistant TOG5681. Allelic research based on primers derived from the RYMV1 gene revealed one accession with allele rymv1-3, and two accessions with allele rymv1-4, and one accession with a different resistance gene. The implications of the finding were discussed and a strategy proposed for breeding varieties with a comprehensive resistance to RYMV. After three generations of backcrossing, the major resistance gene of the variety Gigante was successfully introgressed into five elite rice varieties of West Africa by Marker-Assisted Backcross (MABC). The newly developed BC3F3 progenies were screened for resistance to RYMV in farmers' fields in Guinea and Mali and also under controlled conditions in a screenhouse in Benin. As shown by low virus content and level of disease incidence, low tiller number and plant height reduction, the transferred gene was fully functional in the new genetic background. Moreover, some lines also displayed a high level of resistance to rice blast (Pyricularia oryzae) and stem borer infestation in Guinea. Four of those lines are in the second year of multi-location trial in seven West African countries. Therefore, effective deployment of the newly developed varieties, coupled with good cultural practices, should reduce the damaging effects of RYMV in lowland and irrigated rice cropping systems and thereby increase the income of small scale farmers from rice cultivation. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2012.

This study investigated the establishment of vegetation cover planted in plug and seedling form in the closure phases of the Umlazi Landfill. It also investigated the various facets of the closure process of the Umlazi Landfill and the effect these have on the establishment and choice of vegetative cover, and the grass technology used to make the establishment of vegetation a success. The setting up of trials and the gathering of basic data were undertaken to assess the alternative vegetation options available to researchers. The cover provided by the grasses was assessed in the investigation. The capping of landfill sites is a relatively new approach and it is soon to become a mandatory requirement by the Department of Water Affairs and Forestry (Minimum Requirements for Waste Disposal) (DWAF, 1998). This systematic investigation used in the closure of the Umlazi Landfill, will provide a model for the capping of landfills in South Africa. Seeing that this was the first hazardous (H:h) landfill site in the country to be closed according to the Minimum Requirements for Waste Disposal (DWAF, 1998), every attempt was made to ensure that all aspects in the closure of the site met with the Minimum Requirements. The Minimum Requirements document mentions only briefly that the landfill must be vegetated with some grass type. Prior to 1994, capped landfill sites were usually planted with traditional grass seed mixes and these were not widely successful, as seen on many older landfills that have been partially or completely capped, and where vegetation cover is sparse. There is much literature in the developed countries on the closure of landfills (e.g., Erickson, During the site inspections in June 2001 and February 2002, it was noted that many species of alien plants had established themselves in the poor soil conditions. This made it even more important to find indigenous vegetation to vigorously establish itself that would prevent the establishment of alien invaders. Samples of grass species established on some part of the site were also taken for identification. The dominant grass was identified as Cynodon dactylon. In view of establishing a balanced vegetative cover on top of the Umlazi Landfill, Acacia karoo trees (in seedling form) were also planted. Three bunch grass species, Melinis nerviglumis, Melinis minutiflora and Hyparrhenia hirta, were tested to see if thatching grass could be grown on the site to generate a cash crop for local residents of Umlazi township. Preparation and planting of the capped areas took place in the latter part of 2003 and were completed in early 2004. Measurements and field data were recorded and statistically analysed. The trials revealed three key findings: Firstly, both creeping grasses studied, namely Cynodon dactylon var. “Sea Green” and Panicum natalense var. ”Natal Buffalo Grass” grew well on the site. Initially P. natalense grew faster but after a month, C. dactylon overtook it. At the end of the trial (six months, P. natalense provided a higher level of soil cover. However, C. dactylon grew more consistently over this period. Hence both species provided good growth and cover on this site. Secondly the three bunch grasses, Melinis nerviglumis, Melinis minutiflora and Hyparrhenia hirta, all grew well and had similar survival rates. Hence the potential for growing these grasses as a cash crop has potential. Thirdly, all the Acacia karoo trees survived, i.e., they achieved 100% survival. The average height increase and stem width was similar in all trials and growth was consistent over the six month growing period. Hence the tree species would be a good choice for planting on landfills in its ecologically suitable zones. It is therefore feasible to envisage the planting of a mixture of grasses under the cover of A. karoo trees, to provide a balanced mixture of indigenous grasses to cover a freshly capped landfill. Such a system should provide for stable growth of vegetation for many years. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2008

The Sugarcane Industry contributes approximately 400 000 jobs and ZAR 8 billion annually to South Africa’s economy. Due to climate change and the subsequent threat posed by disease, these figures have been on the decline. Brown rust, a contributor to this decline is caused by the basidiomycete Puccinia melanocephala Syd. and P. Syd., which previously resulted in 50% yield losses in susceptible varieties. This highlighted the need for improved screening and breeding techniques which will result in the replacement of susceptible varieties. The objectives of this study were to: a) Adopt and optimise a glasshouse whorl inoculation screening technique applicable for mass screening of large populations. b) Develop a rapid and cost effective rust resistance screening technique using detached leaves. c)Utilise two flanking marker sets (R12H16 and 9O20-F4-PCR primers) for the rust resistance Bru1 gene in a diagnostic polymerase chain reaction (PCR) to identify rust resistant genotypes lacking Bru1 and possessing either quantitative resistance or an alternative major qualitative resistance gene. d) Correlate rust phenotypic data to AFLP marker data for the Linkage Disequilibrium (LD2) mapping population. e) Utilise suppression subtractive hybridization (SSH) profiling on rust challenged genotypes to discover differentially expressed genes between susceptible and resistant (susceptible Bru1 negatives taken away from resistant Bru1 negatives); and resistant genotypes (resistant Bru1 positives taken away from resistant Bru1 negatives). 4 Results from the glasshouse whorl inoculation trials showed the technique could be reliably used to screen large populations, as two independently conducted pot trials showed highly correlated rust ratings. A visually assessed detached leaf assay (DLA) was developed using selected genotypes. Chlorophyll fluorescence and SPAD readings were used in the DLA to determine the leaf photochemical efficiency (PIABS) with relation to chlorophyll content, resulting in reduced assessment time of at least two days. PCR diagnostics revealed 31% of LD2 did not possess either flanking marker, 8% had one or the other marker, and 61% had both markers. The overall rust phenotypic ratings (rating scale of 0-10) and Bru1 status of the genotypes was used to group the population, with the Bru1 negative genotypes containing all three rating categories (resistant 0-3.5; intermediate 3.51-6.5; susceptible 6.51-10); while the Bru1 positive genotypes were all resistant. The phenotypic data was correlated to AFLP data using the Pearson product-moment correlation coefficient and stepwise multiple linear regression employed to build marker based models to use for predicting non-Bru1 mediated resistance. SSH analysis was then subsequently conducted on genotypes selected on the basis of Bru1 status and AFLP correlation data. Two subtraction cDNA libraries were constructed and the cDNA inserted into electro-competent Escherichia coli cells. PCR on transformed cells revealed cDNA inserts ranging from 200- 1300bp. BLAST analysis of the cDNA sequences indicated the presence of high proportions of disease and drought stress related sequences in the libraries. Analysis of the sequences in both libraries showed that the resistant Bru1 negative genotypes contained oxidative stress related sequences which were however absent in the Bru1 positive resistant genotypes. The library comparing the Bru1 negative resistant genotypes against the Bru1 negative intermediate and susceptible genotypes showed a higher proportion of differentially expressed sequences coding for putative disease resistance proteins, highlighting their presence in the resistant genotypes. Both subtraction libraries also contained high proportions of a leucine rich repeat protein coding cDNA which contained a conserved domain homologous to that of a disease resistance protein conferring resistance to Pseudomonas syringae in Arabidopsis thaliana. The outcomes of this study will subsequently enable an improved understanding of sugarcane-rust resistance mechanisms and improved breeding and screening techniques for sugarcane by identifying SSH and AFLP markers linked to rust resistance QTLs or alternative R genes. / Thesis (M.Sc.Agric)-University of KwaZulu-Natal, Pietermaritzburg, 2012.

Research into the micropropagation of black wattle in South Africa is important for two reasons. Firstly micropropagation technology allows breeders to select and propagate mature tissue, which in turn allows them to better capture selected traits. Secondly, tissue culture may control the highly invasive nature of black wattle. If triploid black wattle can be developed, foresters will then have to rely on clonal propagation to supply material for their growing operations. This research was part of the Institute for Commercial Forestry’s Acacia mearnsii vegetative propagation programme. The main focus of this research was to overcome various problems associated with direct organogenesis of ex vitro material. The shoot apex region was used as the explant in all studies because this region is thought to harbour relatively few internal microbial contaminants and is of sufficient size to withstand stresses associated with micropropagation. The initial research was focussed on the screening of sterilants, searching for a viable alternative to mercuric chloride. Surface sterilisation is integral to any micropropagation technique. This process should do the least amount of plant damage, whilst reducing microbial contamination to an acceptable level. Explants were cultured on Murashige and Skoog (MS) medium supplemented with 2.0 mg L-1 BA and monitored for signs of contamination and shooting. Household bleach proved an excellent alternative to mercuric chloride because it did significantly less damage to the explants than mercuric chloride and is handled easily. There was no significant effect of sterilant exposure time on explant decontamination levels, whilst the shortest exposure time resulted in significantly higher levels of shoot development than the other two times tested. The results of this initial research was developed into a protocol and utilised in subsequent investigations. Due to a considerable variation in the success of the developed surface sterilisation protocol according to different times of the year, a further investigation into the effects of season and mother plant material on shoot apex culture of Acacia mearnsii was undertaken. The success of any tissue culture technique depends on a large array of ex vitro and in vitro variables. The objective of this research was to determine the ii effect of two ex vitro variables, season and mother plant, on shoot apex culture of Acacia mearnsii. Explants from individual mother plants were cultured on MS medium supplemented with 2.0 mg L-1 BA during four separate seasons and monitored for signs of contamination and shooting. Spring was found to be the best harvesting season because spring explants showed significantly higher decontaminated explant levels and shooting levels than explants harvested in the other three seasons. The effect of mother plant selection on the performance of Acacia mearnsii explants during shoot apex culture was also found to be significant, especially with regard to shooting levels. Finally factors influencing shoot elongation of A. mearnsii during shoot apex culture were investigated. In the past, induction of shoot elongation during micropropagation of A. mearnsii was attained through the addition of plant growth regulators and other supplements to the basal culture medium. However, some micropropagation methods in other species have utilised red light as a means of promoting shoot elongation. The objective of this study was to test the effects of an alternative basal medium, red light and differing concentrations of chemical additions to the culture medium on shoot elongation of Acacia mearnsii during shoot apex culture. Four independent experiments were undertaken comparing: shoot elongation on Woody Plant Medium (WPM) to the MS basal medium control; shoot elongation under a red cellophane box compared to control culture light conditions; shoot elongation on media supplemented with various concentrations of GA3 to the un-supplemented control and shoot elongation on media supplemented with combinations of BA and IBA compared to a control. Although no significant effects were observed, many trends were noted. The results indicated that there was no advantage to using WPM instead of MS medium when attempting to elongate shoots, rejuvenated through shoot apex culture of A. mearnsii, whilst the effect of GA3 showed a negative trend. The effects of red light and some BA and IBA combinations showed positive trends on the elongation of initiated shoots. This research successfully addressed some of the problems associated with micropropagation of A. mearnsii. Shoot apex culture shows promise and further research into this technique should be considered. A viable surface sterilant alternative to mercuric chloride was successfully identified. This alternative is not only iii safer to use but shows a large reduction in phytotoxic effects. The effects of season and mother plant on shoot apex culture was successfully investigated, resulting in a better understanding of mother plant influences on tissue culture as well as the identification of an optimum season for explant selection. Finally two possible shoot elongation promoters were identified for further research and a more affordable alternative to red light sources and screens was identified. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2007.

The first serious outbreak of brown rust of sugarcane caused by Puccinia melanocephala Syd. & P. Syd. was reported in India in 1907. It was first reported in South Africa (SA) in 1941 on the variety Co301 and is now present in almost all the sugarcane growing areas of the world. In SA, it is now described as an important disease of sugarcane, causing yield losses of up to 26% in susceptible varieties. Within the SA sugar industry, rust is controlled through the use of resistant varieties as it is the most economical method of control. However, most of the newer varieties that are being released have an intermediate resistance rating for rust. An integrated management approach for the control of rust is therefore being investigated. Aspects investigated in this study included environmental conditions required for development of the disease i.e. epidemiology, the use of silicon (Si) as a cultural control method against brown rust and identification of gene sequences expressed in response to brown rust infection. For the epidemiology study, inoculated plants were incubated in a dew chamber at different temperatures and leaf wetness periods. The choice of leaf wetness duration and temperature was based on urediniospore germination studies. The optimum temperature for urediniospore germination and disease development at > 98% relative humidity was found to be between 20 and 25°C with nine hours of leaf wetness. Silicon has been shown to reduce the incidence of diseases and pests in a number of crops. The ability of sugarcane to accumulate Si and the location of Si deposition was established using two uptake and deposition trials. Different concentrations of Si were applied to the plant and accumulation in the roots, stalks, old leaves and young leaves was determined using inductively coupled plasma optical emission spectrometry, with accumulation found to be roots > old leaves > stalks > young leaves. Silicon deposition in the leaves was determined using energy dispersive X-ray mapping on freeze dried specimens and significant differences were found between the upper epidermis, lower epidermis and mesophyll with the most Si being deposited in the lower epidermis. For disease severity, plants were naturally infected with rust and rated weekly. A significant decrease in disease severity and area under disease progress curve was noted when the Si concentration increased, indicating that Si has potential in reducing rust incidence. Currently, the most reliable and economical method of managing brown rust is with the use of resistant varieties. Identification of resistance within breeding lines is therefore important. For this part of the study, suppression subtractive hybridization was used as a tool to identify differentially expressed genes between a susceptible and resistant variety and a susceptible and intermediate variety, in response to brown rust infection. Two efficient subtracted cDNA libraries were generated and differentially expressed sequences were identified within each library. The results of this study show potential for the development of molecular markers which could be used for the early identification of brown rust resistance during the breeding process. This study forms a firm basis on which an integrated management strategy, for the management of brown rust in the SA sugar industry, could be designed. The cDNA sequences identified could be further investigated and used to develop molecular markers to select for rust resistant varieties, the epidemiology results together with further field data could be used to develop a disease prediction model and Si has potential in the field to reduce brown rust severity. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2009.

Pepper (Capsicum annuum L.) is an important vegetable crop grown and consumed worldwide. Potato virus Y (PVY) is a globally economically important pathogen which significantly reduces the yield and quality of cultivated pepper. The virus is considered as a major limiting factor to the economic production of pepper in the province of KwaZulu-Natal (KZN) in the Republic of South Africa (RSA). Many applied practices to control the spread of PVY are ineffective to mitigate the losses incurred by many farming communities across the KZN province. Therefore, the objectives of this study was to determine the full genome sequence of a PVY isolate from KZN, to identify resistance alleles in commercially available pepper varieties in KZN and to develop a pepper hybrid variety with resistance to PVY using a molecular breeding strategy The first part of the study was conducted to determine the first full genome sequence of a PVY isolate (JVW-186) infecting pepper from KZN. The complete genome sequence of JVW-186 was assembled from overlapping RT-PCR clones using MEGA 5 software. Individual ORFs were identified using the nucleotide data base NCBI and aligned using CLUSTALW. RDP4 software was used to identify recombination junctions in the sequence alignment of JVW-186. CLC Main Workbench 6 software was used to determine the nucleotide sequence similarity of recombinant and non-recombinant fragments of JVW-186 in conjunction with ten PVY parental isolates. Based on sequence data, virus morphology and the coat protein size as determined by SDS-PAGE analysis, the identity of the isolate JVW-186 was confirmed as PVY. Phylogenetic trees were constructed from all recombinant and non-recombinant segments of the sequence by the maximum likelihood method using MEGA 5 software. The full length sequence of JVW-186 consisted of 9700bp. Two ORF’s were identified at position 186 and 2915 of the sequence alignment encoding the viral polyprotein and the frameshift translated protein P3N-PIPO, respectively. RDP4 software confirmed two recombination breakpoints at position 343 and 9308 of the sequence resulting in four segments of the genome. At each recombination event, a 1021-bp fragment at the 5’ end in the region of the P1/HC-Pro protein and a 392-bp fragment in the region of the coat protein shared a high sequence similarity of 91.8 % and 98.89 % to the potato borne PVYC parental isolate PRI-509 and the PVYO parental isolate SASA-110 respectively. The non-recombinant fragment 1 clustered within the C clade of PVY isolates; however the large 7942-bp fragment 3 did not cluster within any of the clades although it shared > 80% nucleotide sequence similarity to other PVY isolates used in this study. Our results suggest that isolate JVW-186 is a novel recombinant strain of PVY that could have evolved due to the dynamics of selection. The second part of the study aimed to evaluate different pepper lines for resistance to PVY. Two recessive alleles (pvr21 and pvr22) located on the pvr2-elF4E locus are known to confer resistance to the virus. To this end, six pepper lines were challenged with PVY infected Nicotiana tabacum cv. Xanthi leaf material using mechanical inoculation under greenhouse conditions. Each line was assessed for resistance to PVY by visual screening for disease severity and quantitative enzyme linked immunosorbent assay (ELISA) for virus load. Pepper lines were further characterized using tetra-primer ARMS-PCR (amplification refractory mutation system polymerase chain reaction) to identify and differentiate the presence of homozygous/heterozygous resistance alleles that confer PVY resistance. Evaluations revealed two resistant pepper lines (Double Up and Cecelia) and varying levels of susceptibility in the other four pepper lines challenged with PVY. The most susceptible pepper line was Benno, although high levels of susceptibility were observed in three other lines (IP, Mantenga and Excellence). The pvr2+ allele was positively identified in all the susceptible pepper lines using the T200A tetra-primer which confirms that the presence of this allele is dominant for PVY susceptibility. Double Up and Cecelia were genotyped homozygous pvr21/pvr21 and pvr22/pvr22 respectively, and remained asymptomatic throughout the trial which indicates that these alleles confer resistance to the isolate of PVY used in this study. The information generated in this study can be incorporated into breeding programs intended to control PVY on pepper in KZN. The final part of the study focused on the development of resistant varieties as the best alternative to manage PVY diseases on pepper. Homozygous F2 pepper lines were developed from local germplasm carrying PVY resistance genes (pvr21 and pvr22) using marker assisted selection (MAS). The F1 progeny was obtained by crossing a homozygous pvr21 (resistant) ‘Double Up’ cultivar with a heterozygous susceptible (pvr2+/pvr22) ‘Benno’ cultivar. F1 and F2 generations were assessed for the presence of PVY resistance/susceptibility alleles (pvr2+/pvr21/pvr22) at the pvr2-elF4e locus using the tetra primer amplification refractory mutation system – polymerase chain reaction (ARMS-PCR) procedure. Negative selection was carried out using the tetra-primer T200A marker to detect the pvr2+ (susceptible) allele. All F1 progeny displaying the pvr2+ allele were eliminated from further study. All 302 plants belonging to 29 F2 families expressing homozygous recessive traits were tested via mechanical inoculation for their response to PVY infection and resistance to PVY was confirmed in all selected families based on symptomatology in greenhouse house screens using double antibody sandwich enzyme linked immunosorbent assay (DAS-ELISA). These results show that ARMS-PCR can be used to successfully screen pepper genotypes for alleles that confer PVY resistance thereby contributing to the improvement of pepper production using molecular breeding approaches. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2013.

Potato virus Y (PVY) is an economically important virus worldwide. In South Africa, PVY has been shown to be a major limiting factor in the production of important solanaceous crops, including potato (Solanum tuberosum L.), pepper (Capsicum annuum L.), tomato (Lycopersicon esculentum Mill.) and tobacco (Nicotiana spp). The variability that PVY displays, wherever the virus occurs, merits the study of the isolates occurring in KwaZulu-Natal (KZN) in the Republic of South Africa (RSA). This characterization will provide a clear understanding of strains/isolates from local vegetables and how they relate to the other PVY strains already identified, as well as information that can be used to manage the diseases they cause. Hence, the aim of this project was to study the biological and genetic properties of PVY isolates infecting potato, tomato and pepper in KZN. Enzyme-linked immunosorbent assay (ELISA) using monoclonal antibodies and reverse transcription polymerase chain reaction (RT-PCR) using primers specific to all PVY strains were used to detect the virus in plant material showing PVY-like symptoms collected from various locations in KZN. A total of 39 isolates (18 isolates infecting tomato, 12 infecting potato and 9 infecting pepper) were further differentiated into strains by means of ELISA using strain specific antibodies and RT-PCR using primers specific to the different strains of PVY identified around the world. All PVY isolates infecting tomato and pepper tested positive for the ordinary PVYO strain with both ELISA and RT-PCR. PVY isolates infecting potato were more diverse and comprised the PVYN, PVYNTN and PVYNWilga strains, with mixed infections noted in some cases. The biological properties were studied by mechanically inoculating Chenopodium quinoa, Nicotiana tabacum cv Xanthi, N. tabacum cv Samsun, N. glutinosa, and N. rustica with leaf extracts from plants infected with the different PVY strains detected in this study. All inoculated C. quinoa plants did not show symptoms. All tobacco plants showing symptoms were tested for the presence of PVY by means of ELISA using monoclonal antibodies targeting all strains and electron microscopy using the leaf dip technique. Not all the inoculated tobacco tested positive with ELISA. The symptoms observed were therefore divided into PVY-related and PVY non- related. PVY-related symptoms included vein clearing, mosaic chlorosis, stunting, and vein necrosis. PVY non-related symptoms included wrinkles and leaf distortions. Potyvirus-like particles of about 700 nm were observed under the transmission electron microscope (TEM) from plants showing PVY-related symptoms while rod shaped viral particles of sizes varying between 70 and 400 nm were observed from plants showing non-PVY related symptoms. A portion of the virus genome (1067 bp) covering part of the coat protein gene and the 3’ non-translated region (NTR) of three PVYO isolates infecting tomato, one PVYO isolate infecting pepper and one PVYNWilga isolate infecting potato were amplified, cloned and sequenced. The 5’ NTR, P1, HC-Pro and part of P3 regions (2559 bp) of a PVYN isolate infecting potato were also amplified, cloned and sequenced. Sequence data was compared with selected PVY sequences from different geographical locations around the world. These were available on the NCBI website and subsequently used for phylogenic analyses. The sequenced genomic regions of the PVYN isolate were found to be 99% similar to the New Zealand PVYN isolate (GenBank accession number: AM268435), the Swiss PVYN isolate CH605 (X97895) and the American PVYN isolate Mont (AY884983). Moreover, the deduced amino acid sequence comparison of the genomic regions of the PVYN isolate revealed the presence of five distinct amino acids residues. The three amino acid residues (D205, K400, and E419), which determine the vein necrosis phenotype in tobacco, were also identified. The coat protein and 3’ NTR sequences of all KZN PVYO isolates infecting pepper and tomato were closely similar to each other than to KZN PVYNWilga isolate infecting potato. The phylogenic analysis clustered the KZN PVYN isolate with the European sublineage N, PVYNWilga isolate infecting potato with the American PVYO isolate Oz (EF026074) in the O lineage and all PVYO isolates infecting tomato and pepper in a new sublineage within the O lineage. Taken together, these results point to the presence of PVY in solanaceous vegetables cultivated in KZN and they lay the foundation for the formulation of effective control measure against PVY diseases in KZN. / Thesis (M.Sc.) - University of KwaZulu-Natal, Pietermaritzburg, 2009.

Fusarium root rot (FRR) disease, caused by the fungus Fusarium solani f. sp. phaseoli (FSP), is an important soil-borne disease reducing common bean (Phaseolus vulgaris L.) yields, and hence food security, in Uganda and elsewhere in developing countries where the crop is grown without fungicides. The key aim of this study was to elucidate the significance of bean root rot (BRR), appraise methods for screening germplasm for resistance to FRR, determine the genotypic variability of resistance, and the inheritance of resistance to FRR in common bean. This information was deemed useful in devising an appropriate strategy for breeding FRR resistance in beans. A participatory rural appraisal (PRA) was conducted in south-western and eastern Uganda to ascertain farmers’ awareness of BRR and their influence on preferred bean varieties. Bean root rot is considered to be the most devastating and most recognised disease, especially in south-western Uganda. Control measures for BRR were very minimal, and in some cases, non-existent. Use of resistant varieties to control the disease was not evident, because the most popular varieties were susceptible to the disease. The resistant bean varieties currently available have undesirable characteristics such as small seed size, black seed and late maturity. Large-seeded bean varieties, even though cited as being more susceptible to BRR than the small-seeded varieties, are still very popular. The study highlighted the need for breeding FRR resistance in the large-seeded bean varieties that are highly preferred by farmers. Four isolates of FSP (FSP-1, FSP-2, FSP-3 and FSP-4) were tested for pathogenicity under screenhouse and laboratory conditions. In addition, three methods of storing and maintaining the viability of FSP isolates were appraised. The isolate FSP-3, was found to be the most pathogenic, resulting in 100% disease incidence on all bean varieties tested, with high severity scores. The potato dextrose agar (PDA) slants stored at 5oC were found to be the best method of storage for pathogenic isolates. The FSP-3 isolate was subsequently utilised for screening bean lines for resistance to FRR. The influence of soil composition, irrigation frequency, and inoculation technique on the severity of FRR was studied on six bean lines. Interactions of irrigation frequency, soil composition, and bean lines were not significant. The 50% swamp soil:50% forest soil composition and forest soil alone categorized the varieties most distinctly according to their reaction to FRR. Also, the best distinct classification for the varieties was obtained under treatments that were watered daily and once in a week. Based on economic considerations, the standard forest soil and daily irrigation were subsequently adopted for screening bean germplasm for resistance to FRR. It was also found that sorghum seed as a medium for pathogen inoculation was better than the agar slurry medium. One hundred and forty seven common bean varieties were evaluated for resistance to FRR (isolate FSP-3) under screenhouse conditions. In order to confirm this resistance, 46 common bean lines selected from the screenhouse trial were further evaluated using natural inoculum in a BRR-infested field. Forty-four varieties comprising ten large-seeded, four medium-seeded and 30 small-seeded varieties showed moderate resistance to FRR; but none were resistant or immune to the disease. Based on adaptability, eight moderately resistant varieties were selected for use as parents in the study of inheritance of resistance to FRR. A 12 x 12 diallel mating design was utilised to develop 66 F1 and F2 populations, plus their reciprocal crosses, with the aim of studying the mode of inheritance of resistance to FRR. The F1 and F2 progeny evaluations showed that FRR resistance was mainly governed by additive genes in most populations. However, there were a few crosses which displayed highly significant specific combining ability (SCA) effects, implying that dominant effects were important in some populations. Maternal effects were also highly significant at both the F1 and F2 generations, suggesting that resistance was modified by cytoplasmic genes. The non-maternal effects were also significant in some populations, suggesting that the cytoplasmic genes were interacting with nuclear genes. The number of genes governing resistance to FRR varied from two to nine among the eight sources of resistance. The allelism test of resistant x resistant populations, and the observation of continuous distributions of severity scores, suggested the presence of many loci governing FRR resistance in beans. Broad sense heritability of disease resistance varied from 0.22-0.69, while heritability in the narrow sense was estimated as 0.35-0.49 in the populations. These results suggested that selection and backcrossing to both parents would be the best breeding procedures for improving resistance in the popular large-seeded bean varieties in Uganda. However, there could be complications in breeding for resistance to FRR in beans, because resistance was modified by cytoplasmic gene effects and their interaction with nuclear genes in some of the populations. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2008.

There is a growing recognition globally that many agrochemicals are hazardous to humans, animals and the environment. Therefore, there is a need to substitute these chemical products with biological and physical treatments, and to change agronomic practices in order to control pests and diseases in agriculture. The primary objective of this thesis was to develop and test in laboratory, field and commercial packhouses trials as alternative control measures against green mould of citrus (caused by Penicillium digitatum Pers: Fr. Sacc) and Penicillium molds of litchi (caused by several Penicillium). A South African isolate of P. digitatum, isolated from an infected orange fruit, was found to be resistant to imazalil (the standard postharvest fungicide used in South Africa). Sixty yeast and 92 Bacillus strains were screened for their antagonistic activity against this isolate of P. digitatum. None of the yeasts or Bacillus isolates produced a curative action against P. digitatum on oranges. However, yeast Isolate B13 provided excellent preventative control of P. digitatum, superior to all the Bacillus isolates, when it was applied to citrus fruit prior to artificial inoculation with P. digitatum. Electron microscopy showed that yeast Isolate B13 inhibited conidial germination of P. digitatum. For the control of P. digitatum pre-harvest, trees were sprayed with a yeast, Isolate B13, a few months or a few days before harvest. However, this treatment alone proved to be ineffective in providing preventative control of green mould on Valencia oranges. For the control of P. digitatum preharvest, trees were treated with potassium silicate for a full season. Regular potassium silicate treatments resulted in a significant preventative control of P. digitatum infection on both navel and Valencia oranges. Treatment of Eureka lemons with potassium silicate as a postharvest treatment for the control of P. digitatum resulted in reduced disease lesion diameters when applied preventatively or curatively. Electron microscopy showed that potassium silicate inhibited germination of P. digitatum conidia and growth of its mycelium. Hot-water dip treatment at 50-58°C for 60-180 seconds (in increments of 15 seconds), significantly reduced infection development in inoculated wounds of Valencia oranges compared with control fruit treated with tap water, without causing any rind damage. The integration of the yeast, a hot water dip and potassium silicate pre-and postharvest applications provided control of P. digitatum control in multiple packhouse trials. The control achieved by the yeast Isolate B13 or hot-water, and potassium silicate in the packhouse alone was superior or equivalent to that provided by imazalil. A similar study was also carried out to determine possible control measures for Penicillium sp. on litchis. In this study, a total of 23 yeast and 13 Bacillus isolates were obtained from litchi fruit surfaces. Ten yeast and 10 Bacillus isolates that had shown good efficacy against P. digitatum of citrus were added to these for screening against Penicillium sp. of litchis. None of the yeasts or Bacillus isolates produced a curative action against Penicillium sp. infection on litchis. However, several yeast isolates (YL4, YL10, YLH and B13) resulted in reduced severity of the pathogen, when applied preventatively, compared with an untreated control. The yeast isolates were superior to all the Bacillus isolates, when applied to litchis prior to artificial inoculation by Penicillium infection on litchis. Potassium silicate as a postharvest treatment for the control of the pathogen caused reduced lesion diameters when applied preventatively or curatively to naturally infected litchis. The results presented in this thesis highlight the use of biological, physical and agronomic practices singly or in combination as an alternative control strategy against citrus postharvest green mould. This thesis also provides an insight into expanding these strategies, partly or fully, for the control of other postharvest Penicillium infections using litchi as an example. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2010.