Statement of work: In essence the research work was to focus on the development of an ionospheric model suitable for real time HF frequency prediction and direction finding applications. The modelling of the ionosphere had to be generic in nature, sufficient to ensure that the CSIR could simultaneously secure commercial competitiveness in each of the three niche market areas aforementioned, while requiring only minimal changes to software architecture in the case of each application. A little research quickly showed that the development of an ionospheric model capable of driving a HFDFSSL system in "real time" would result in one having to make only slight re-structuring of the software to facilitate application of the same model in the areas of real time frequency prediction and spectrum management. The decision made at the outset of the project to slant the research toward the development of a model best suited for HF direction finding applications is reflected in the avenues followed during the course of the modelling process

No description available.

No description available.

MIMO, ou Multiple-Input Multiple-output, est une technique de communication radio qui repose sur l'utilisation conjointe de réseaux , d'antennes à l'émission et à la réception, Elle permet d'améliorer le débit ou la robustesse d'un lien radio sans augmenter la puissance d'émission et la bande de fréquence allouée. Cependant, les performances de cette technique sont largement dépendantes des propriétés du canal de propagation, Elle n'apporte une amélioration substantielle par rapport aux techniques mono-antennes classiques que si le canal de propagation est suffisamment riche en trajets multiples, Ainsi, la caractérisation et la modélisation du canal de propagation MIMO sont deux étapes indispensables pour un déploiement optimal des systèmes MIMO. Les résultats publiés ces dernières années montrent que les modèles de propagation existants dédiés aux communications MIMO comme par exemple celui basé sur une approche géométrique stochastique sont très complexes à mettre en œuvre, L'objectif de ce travail est de proposer une nouvelle méthode simple et originale de modélisation basée sur la décomposition en matrice diagonale par bloc de la matrice de transfert du canal. L'approche proposée nécessite des estimations précises des caractéristiques du canal. C'est la raison pour laquelle des algorithmes de pré traitement des données ont également été développés pour l'estimation du nombre de sources et des puissances des composantes cohérentes et diffuses contenues dans le signal reçu. Les résultats de la caractérisation des canaux MIMO en Gare, Halle, tunnel et environnement de bureaux permettront d'illustrer les différents points soulevés dans l'étude théorique. / Multiple- Input Multiple-output (MIMO) is a radio communication technique that is based on the use of antenna arrays at both the receiver and transmitter. This tedmique yields a higher throughput and robustness without increasing the ernitting power and allocated bandwidth, Nonetheless, the performance of this technique is strongly dependent to the channel propagation properties. Hence, improvement is only obtained with respect to classical mono-antennas for the case where a large number of paths are embedded within the propagation channel. Therefore, the characterization and modelling of the MIMO channel are the two critical steps necessary to alleviate the deployment ofMIMO systems. CUITent MIMO channel models such as the geometrical-based stochastic model have been shawn ta be difficult to exploit. The work aims at proposing a novel yet simple modelling scheme based on the decomposition of the channel transfer matrix into block-diagonal matrices. Because an accurate estimation of the channel characteristics is required, dedicated post-processing algorithms were developed ta estimate the number of sources, power of the coherent and diffuse components of the measured propagating paths. MIMO channel measurements were performed in train station, atrium, tunnels and office environments to validate the proposed scheme.

Lachenalia is a genus endemic to South Africa and Namibia. It is propagated vegetatively by daughter bulbs, bulbi Is, tissue culture and leaf cuttings. In Europe, the demand for Lachenalia bulbs is estimated at 20 million per annum and thus the need for a rapid and cheap way of propagation. Of all the vegetative methods for propagating Lachenalia, the easiest and cheapest method is by leaf cuttings. Cook first reported this method of propagation in 1931. Nothing was done at that time, as the natural way of propagation sufficed for the demand. No efforts were, however, made to optimize the technique of leaf cuttings. In this study various aspects aimed at optimizing the production of bulblets by leaf cuttings of Lachenalia were studied. These included: stage of the donor (mother) plant, leaf section position, medium, disinfectants, and starch deposition in leaves and bulbs during the growing season. Other aspects studied included the ontogenetic origin of bulblets and roots on leaf cutting. The optimal physiological stage to take leaf cuttings was found to be when the inflorescence was visible between the sheath of the leaves and the worst stage was at full flowering. As the donor plant matures, there was a reduction in the number, size and mass of bulblets produced by leaf cuttings. The proximal sections performed better than their distal counterparts in all variables evaluated. Considering cost and time, applying disinfectants when planting the leaf cuttings seem not to be necessary. Decomposed bark was the best medium for Lachenalia leaf cuttings. Bulblets were mostly formed on the adaxial leaf surface on Lachenalia leaf cuttings and developed as a result of the division of the epidermal cells. Both the leaf cutting and the bulblet formed roots. On the leaf cuttings, roots originated from the parenchyma cells associated with the vascular system while on the bulblets they developed from the base of the meristematic mass of cells, which formed the bulblets, and were attached to the bulblet. In a TEM study of the leaf, no starch was observed in leaf sections. Starch was, however, observed in all sections from bulb parts. More starch grains were observed on the inflorescence stalk than in other bulb parts. / Dissertation (M Inst Agrar ( Horticulture Science))--University of Pretoria, 2006. / Plant Production and Soil Science / unrestricted

The many medicinal, nutritional, industrial, and agricultural uses of moringa are well documented, although studies remain focused on moringa as a wild, uncultivated plant and little horticultural research has been devoted to its culture. Greenhouse experiments were conducted at the State University of New York at Cobleskill, New York, USA during the summer of 2006. Soilless culture has been investigated as an alternative production system for medicinal and aromatic plants, and moringa seedlings were grown using hydroponic and aeroponic culture at a pH of 5.5-6.0 and electrical conductivity levels of 2.0-2.3 dS˙m-1 in order to maximize leaf biomass production and gain access to root biomass. Moringa leaves are valued for their human nutritional benefits and as a livestock feed, however the monopodial growth habit of moringa requires removal of the terminal growing point in order to encourage branching and maximize leaf production. Seedlings are often subjected to extreme levels of vegetative harvest when grown in high-density agroforestry production systems, often resulting in high levels of mortality. Seedling decapitation induced outgrowth of lateral buds, and no lateral buds were released by intact plants that were not decapitated. The release from apical dominance was repressible by the application of exogenous auxin (0.5% IAA) to the stumps of decapitated seedlings. Seedlings decapitated to node 6 (15 cm, counted acropetally) released buds more quickly and produced three times the dry weight per plant than seedlings decapitated to node 1 (5 cm). Seedlings decapitated to 30 cm (approximately 12 nodes) released 2 more buds per plant and twice the dry weight compared to those decapitated to 40 cm (approximately 16 nodes). Pinching seedlings to remove 5 or 10 cm of apical stem resulted in little regrowth with few buds released and low dry weights. The percentage of stem removed was more important than the actual height to which a seedling was decapitated in order to maximize regrowth. Twice-decapitated plants released more buds (6.7/plant) than once-decapitated plants (2.7/plant), demonstrating enhanced lateral bud response of previously decapitated seedlings. The effect of four different pruning strategies on leaf biomass production of young, pollarded moringa trees was also examined.

The objective of this study was to obtain and interpret field data related to the establishment of cranberry (Vaccinium macrocarpon Ait.) plantings, as well as examine the impact of conventional weed management practices on yield components and weed control. Integrated weed management, recently recognized by weed scientists as a desired goal for research and extension, is an important part of current cranberry production. This research was designed to permit the incorporation of the collected data into practical grower recommendations, as well as to expand our general knowledge about invasion ecology and plant species composition in new and established commercial plantings. Data from four years of repeat annual applications of 0, 1.8, and 4.5 kg ai/ha dichlobenil in low-weed and high-weed density areas indicated minimal negative impact on cranberry vines. Herbicide application did not adversely affect upright productivity, biomass, fruit set, or other yield parameters; in addition, no improvements for these parameters were noted. No consistent treatment effect on cranberry root length was seen. The presence of weeds, rather than herbicide application, was the important determinant of yield. Vines in low-weed density areas produced more marketable fruit and had higher percentage fruit set than vines growing in high-weed density areas. Results suggest that repeat annual applications of dichlobenil to commercial cranberry beds may be considered as part of a viable integrated weed management program with no adverse effect on crop growth or yield. One specific goal of this research was to identify the most beneficial combination of nitrogen rate, vine planting density, and weed management option that would promote quick and economical vine establishment of the cultivar, Stevens, while providing adequate and cost-effective weed control. After two years, several treatment combinations seemed promising for commercial implementation. However, the most cost-effective production scheme for establishing a new planting was to plant vines at a low density, use moderate rates of nitrogen, and apply a yearly application of napropamide for weed control. This combination produced substantial vine coverage at very low cost, reduced weed biomass by 85% compared to the untreated plots, and gave the best control per dollar spent.

In the first part of the thesis, experiments are described where self-compatible tetraploids of L. peruvianum were isolated from tissue culture and the expression and inheritance of their S-related proteins was explored. The S-related protein profiles of styles of self-compatible tetraploids were indistinguishable from the diploid self-incompatible explant source based on SDS-PAGE. Progenies obtained from self-fertilization of two tetraploids were all found to be self-compatible. Cloned cDNA sequences of the S-related proteins were used to determine the inheritance at the locus in these progenies through Southern hybridization. The allelic ratio consistent with the predicted ratio obtained if only the pollen bearing two different alleles was successful in achieving fertilization. This work demonstrates that the observed self-compatibility in the tetraploids was due to failure of recognition of heterogenic pollen by the style, while the expression and activity of the S-related stylar proteins remained unaffected. In the second part of the thesis, periclinal chimeras between the SI L. peruvianum (P) and the SC L. esculentum (E) were utilized to analyze the relative influence of apical cell layers as they relate to SI and Unilateral incompatibility (UI). Irrespective of the expression of S-proteins in the chimeral styles, the chimeras were compatible with P component indicating that there was failure of recognition of the "self" pollen (the chimeras and the P component are genetically similar at the S-locus). The breakdown of SI was not related to the RNase activity of the S-proteins which was intact in all chimeras. Therefore S-proteins are not sufficient to maintain the SI response. Since the response was lost if either of the two layers (L1 and L2) was composed of E component, we conclude that both L1 and L2 are required for the SI phenotype. On the other hand, UI response was maintained in all the chimeras in which the L1 or L2 layer was P. This seems to suggest that SI and UI responses are developmentally unrelated in these chimeras.

6-benzylaminopurine (BA), which was applied at the 10 mm stage of fruit development at 50 and 100 mg.liter$\sp{-1}$ to 'McIntosh' apples (Malus domestica Borkh), effectively thinned fruit, and increased fruit size and caused seed abortion. Regardless of BA treatment, there were two distinguishable peaks of fruit abscission during 'June drop' on 'Morespur McIntosh' apples. There was a large peak about 14 days after treatment and a second smaller one 24 days after treatment. The first abscission peak was increased by BA, and approximately half of then existing fruit abscised from treated trees between 10 and 14 days after treatment, whereas only one fourth abscised from the control trees. The number of seeds per fruit was lower in abscising fruit than in persisting fruit regardless of BA treatment. Seed number of abscising fruit was increased linearly by BA application. BA enhanced the translocation of $\sp{14}$C-sorbitol from leaves to fruit when applied directly to the fruit, but not when applied directly to the leaves. Therefore, it is unlikely that BA thinned apple fruit directly through restricting the translocation of photosynthates moving from leaves to fruit. Net photosynthesis was decreased and dark respiration was increased when temperature following BA application was high (30$\sp\circ$C), whereas there was no effect when temperature was lower (20$\sp\circ$C). Total nonstructural carbohydrates, total soluble sugars, reducing sugars, and starch in the leaves decreased dramatically over the 12 or 13 day observation period, regardless of BA treatment. These carbohydrate levels were further lowered by BA application. Abscising fruit, based on specific reddening of pedicel, had higher reducing sugars, total soluble sugars, starch, and total nonstructural carbohydrates. The levels of these carbohydrates in fruit from BA treated trees were similar to that from untreated control trees. Foliar application of 5% sorbitol solution had no effect on fruit set. Removal of bourse shoot tip increased fruit set regardless of BA treatment. BA thinned fruit regardless of whether shoot tips were removed or not, and there was no interaction between BA application and shoot tipping. Experiments were conducted to evaluate the effects of BA on fruit set and fruit development on small fruiting branches supported by different numbers of leaves per fruit. BA thinned fruit only when one leaf per fruit was on the girdled small fruiting branch, but not when the leaf number was greater than two. Fruit weight and soluble solids content increased quadratically with increasing leaf number per fruit. BA reduced fruit growth rate when less than 16 leaves were present on the girdled branches between 3 and 7 days after treatment, but it did not affect fruit growth rate when 32 leaves were on the girdled branches. Increasing leaf number also increased normal seed number while decreasing the number of aborted seeds, but it had no effect on the number of total seeds per fruit. BA reduced the number of normal seeds per fruit only when the number of leaves per fruit was less than 4. These results were discussed in relation to carbohydrate supply and source-sink relationships.

Isozyme analysis was used to identify clones, measure levels of genetic variation within groups of clones, and analyze mating systems in two Cactaceae genera--Hatiora and Schlumbergera. Isozymes were extracted from phylloclades and pollen and were separated by polyacrylamide gel electrophoresis. The inheritance of aspartate aminotransferase (AAT), glucose-6-phosphate isomerase (GPI), malate dehydrogenase (MDH), phosphoglucomutase (PGM), and triosephosphate isomerase (TPI) was examined in Hatiora. Six loci (Aat-1, Gpi-1, Mdh-1, Pgm-1, Pgm-2, and Tpi-2) were analyzed, and results were generally as expected for single loci with codominant alleles. For all six isozyme loci segregation distortion was observed in at least one segregating family. Aat-1 was linked with Pgm-1 (26 cM), but the other isozyme loci assorted independently. The inheritance of leucine aminopeptidase (LAP), phosphoglucomutase (PGM), and shikimate dehydrogenase (SKD) was investigated in Schlumbergera. Three loci were analyzed (Lap-1, Pgm-1, and Skd-1), and results were generally as expected for single loci with codominant alleles. Significant segregation distortion was observed in at least one segregating family for all three isozyme loci. Disturbed segregation at Lap-1 was due to tight linkage (7 cM) with the locus controlling gametophytic self-incompatibility (S). All three loci assorted independently of each other. In a third study, a Hatiora germplasm collection composed of 49 clones was assayed for AAT, GPI, LAP, MDH, PGM, SKD, and TPI. Thirteen putative loci and 42 putative alleles were identified, and 9 of the 13 loci (69%) were polymorphic. Twenty-two clones (45%) could be distinguished solely on the basis of their isozyme profiles, but the other 27 clones shared isozyme profiles with one to five other clones. Thirteen modern H. x graeseri cultivars exhibited less genetic diversity than 40 H. gaertneri, H. x graeseri, and H. rosea clones representing older and modern cultivars plus field-collected specimens. The difference in genetic diversity was primarily attributed to a loss of alleles during breeding. In a fourth study, a Schlumbergera germplasm collection composed of 59 clones was assayed for AAT, GPI, LAP, MDH, PGM, SKD, and TPI. Twelve putative loci and 36 putative alleles were identified, and 10 of the 12 loci (83%) were polymorphic. Forty-one clones (69%) could be distinguished solely on the basis of their isozyme profiles, but the other 18 clones shared isozyme profiles with one or two other clones. Forty-two commercial clones of S. truncata, S. x buckleyi, and S. x exotica exhibited less genetic diversity than 14 field-collected clones of S. kautskyi, S. opuntioides, S. orssichiana, S. russelliana, and S. truncata. The difference in genetic diversity was attributed to limited sampling from wild populations and loss of alleles during breeding.