This master-thesis presents an approach to track and count the number of fruit incommercial mango orchards. The algorithm is intended to enable precision agri-culture and to facilitate labour and post-harvest storage planning. The primary objective is to develop an multi-view algorithm and investigate how it can beused to mitigate the effects of visual occlusion, to improve upon estimates frommethods that use a single central or two opposite viewpoints. Fruit are detectedin images by using two classification methods: dense pixel-wise cnn and regionbased r-cnn detection. Pair-wise fruit correspondences are established between images by using geometry provided by navigation data, and lidar data is used to generate image masks for each separate tree, to isolate fruit counts to individual trees. The tracked fruit are triangulated to locate them in 3D space, and spatial statistics are calculated over whole orchard blocks. The estimated tree counts are compared to single view estimates and validated against ground truth data of 16 mango trees from a Bundaberg mango orchard in Queensland, Australia. The results show a high R2-value of 0.99335 for four hand labelled trees and a highest R2-value of 0.9165 for the machine labelled images using the r-cnn classifier forthe 16 target trees.

The importance of European honey bees in improving fruit set, yield and fruit weight of kiwifruit on the central east coast of Australia was investigated. Field investigations were carried out using different bee saturations and different types of male pollen parents. These investigations confirmed the importance of honey bees in kiwifruit fruit set, yield and fruit weight. However, the results suggested that increasing bee activity alone may not increase pollination of kiwifruit by honey bees. Many factors need to be understood before introducing bees into the orchard. Bees were more effective during the early part of the flowering period, and bee activity varied according to the sex of the vine, planting design and the time of day. The type of male pollen parents also influenced fruit size and quality. Flowers pollinated by different pollen parents were assessed for pollen tube growth and histochemical changes. The resulting fruit were also examined for weight and seed numbers. Honey bees play the major role in the size and yield of kiwifruit, but the design of male vines, their age and type of male pollen may also contribute. The kiwifruit pistil also possesses important features that can be considered as adaptations to insect pollination. / Doctor of Philosophy (PhD)

The marula tree has already become a very important alternative crop in South Africa with a great potential for further improvement. There are, however, no accurate records that have been reported on the fruit yield and quality of marula tree in Southern Africa. To utilise the fruit fully, to make proper selections from the wild trees and to develop new selections/cultivars, there is a need to have information on yield potential and fruit quality aspects. The overall objective was to determine the yield potential and to evaluate the fruit quality of different marula selections putting more emphasis on fruit quality parameters such as the fruit mass, size, total soluble solids (TSS), acidity and juice content. Four marula selections were identified during their peak production season(2004/2005/2006/2007) at Schoeman Boerdery (Orchard) between Marble-Hall and Groblersdal in Mpumalanga Province (South Africa). From each of the selections, namely; Pharulani, Toularula and Swarula, five trees were selected randomly and marked, while from the Wild marula only four trees were selected and marked. For the purpose of experimental design, the four selections were regarded as four treatments, and the marked trees per selection as the replicates. All selected trees were more or less the same in size and were in full production. The trees were growing in an orchard with planting distance of 9 m (interspacing) and 4.5 m (intraspacing). To determine yield potential, the number of fruit per square meter per tree was counted on pre-labelled branches on five positions around the tree canopy at three stages of fruit development. Collected data were converted to grams of fruit per square metre. Tree canopy size (CS), expressed as m2, was calculated by using the radius (r) of each canopy in the following formula: CS= (╥.r2) 4. For calculating the yield per tree, the average canopy size was multiplied by the number of fruit counted per m2. Twenty fruits from each labelled marula tree per selection were randomly selected; collected and taken to the Ecophysiology Laboratory at University of Pretoria for fruit quality traits, weight (mass), size, stone mass, peel mass, juice content, Total Soluble Solids (TSS) and Titratable acidity (TA). The yield efficiency/potential for both tree unit and hectare (ha) basis in 2006/2007 season, showed that the Swarula selection had the highest total yield as compared to other selections. Results showed that there was a positive relationship between fruit weight and size, that is, the bigger the weight, the bigger the size and the opposite. Generally, three selections, Swarula, Toularula and Pharulani had a higher fruit mass and size than “wild” marula. Pharulani had the highest seed/stone mass whereas “wild” marula had the lowest seed mass during the 2004/2005 and 2005/2006 fruiting seasons. Pharulani selection had the highest juice mass during both seasons whereas Swarula selection had the lowest juice mass. Results also showed that ‘wild’ marula had the highest TSS/TA ratio during both seasons whereas Pharulani had the lowest TSS/TA ratio. Optimal traits were therefore found in different selections and not in the same selections as envisaged. Copyright / Dissertation (MInst Agrar)--University of Pretoria, 2009. / Plant Production and Soil Science / unrestricted