18105 research outputs found
Sort by
Innovation in ecological restoration techniques: Enhancing Portulacaria afra survivorship in degraded arid thicket
Get PDFThe Albany Thicket Biome has undergone extensive degradation over the past century, particularly from overstocking of livestock in the arid types of thicket. The degradation of the biome, coupled with little to no natural recovery, prompted the South African Government to implement the Subtropical Thicket Restoration Programme (STRP) in the early 2000s to rehabilitate degraded thicket using unrooted Portulacaria afra truncheons. The STRP also sought to create job opportunities, bring about social upliftment, promote biodiversity conservation and incentivise farmers and landowners in the region to promote carbon trading and farming of P. afra using carbon credits. However, the survivorship of P. afra planted using the STRP planting protocol has been less than ⁓30% because they are planted in degraded areas, where the truncheons are subjected to harsh biophysical conditions such as hard, capped soil, high soil temperatures, drought and herbivory. The aim of this study, therefore, was to improve the survivorship of P. afra planted in degraded thicket to at least 30%, using modified planting methods. In this study, four treatments were applied in 210 pondings (or micro-dams), each with P. afra planted in them (the first three treatments used rooted cuttings). The first treatment involved watering regimes where pondings were watered at varying frequencies. The second treatment involved planting companion species alongside P. afra while the third treatment involved planting P. afra underneath a nurse canopy. In the fourth treatment, unrooted P. afra truncheons were planted inside pondings. The mean survivorship of the P. afra cuttings was assessed 18 months after planting. The watering regime treatments, including the control, all yielded a survivorship of over 75%, with the significantly highest levels of survivorship being displayed in the weekly watering treatment (90.9 ± 6.8%) (p < 0.01). The companion plant treatment also produced a high mean survivorship of P. afra (94.9 ± 3.6%), significantly higher than that of the nurse plants (87.5 ± 6.3%) (p < 0.01). Among the unrooted truncheons it was found that untreated truncheons had the highest survivorship (76.2 ± 17.6%), with the lowest mean being found in truncheons that were both pruned and scarified (70.8 ± 20.8%), suggesting that this treatment, out of all of them, is the least successful under the prevailing environmental conditions in the study area. However, there was no significant differences among the unrooted truncheon treatments and their respective survivorship values to further substantiate this assertion (p = 0.26). Findings in this study also suggest that planting P. afra cuttings under a nurse plant was ideal for P. afra survivorship due to the nurse plant’s ability to ameliorate the microclimate under which the cuttings can establish and grow. Despite the soil under the nurse canopy having a significantly lower mean soil water potential (-160.9 ± 200.5 kPa), compared to the open areas (-73.4 ± 55.7 kPa) (p = 0.04), the pondings under the nurse canopy still had a lower mean soil temperature (31.4 ± 5.25°C) than the treatments in the open areas (38.5 ± 2.7°C), during the harsh midday sun. These favourable characteristics are reflected in the P. afra cuttings under a nurse canopy having a higher chlorophyll fluorescence (0.76 ± 0.06). compared to those planted in open areas (0.73 ± 0.13), suggesting that the P. afra cuttings in the former treatment had a higher photosynthetic efficiency compared to those in the latter, despite there being no statistically significant difference in chlorophyll fluorescence between the treatments (p = 0.14). Although these findings suggest that applying these modifications to planting P. afra, having achieved its objective of increasing survivorship to well over 50%, is likely to produce favourable results not only in overall survivorship, but also in fast-tracking arid thicket rehabilitation and restoration, further research on these rehabilitation techniques and their effectiveness is required. Furthermore, the downside to these treatments is that they are costly and time consuming, which puts the feasibility of large-scale programmes using these applications into question. Further investigation is required to determine ways in which the cost-effectiveness of these applications can be enhanced.Thesis (MSc) -- Faculty of Science, Environmental Science, 202
Semantic segmentation of astronomical radio images: a computer vision approach
Get PDFThe new generation of radio telescopes, such as the MeerKAT, ASKAP (Australian Square Kilometre Array Pathfinder) and the future Square Kilometre Array (SKA), are expected to produce vast amounts of data and images in the petabyte region. Therefore, the amount of incoming data at a specific point in time will overwhelm any current traditional data analysis method being deployed. Deep learning architectures have been applied in many fields, such as, in computer vision, machine vision, natural language processing, social network filtering, speech recognition, machine translation, bioinformatics, medical image analysis, and board game programs. They have produced results which are comparable to human expert performance. Hence, it is appealing to apply it to radio astronomy data. Image segmentation is one such area where deep learning techniques are prominent. The images from the new generation of telescopes have a high density of radio sources, making it difficult to classify the sources in the image. Identifying and segmenting sources from radio images is a pre-processing step that occurs before sources are put into different classes. There is thus a need for automatic segmentation of the sources from the images before they can be classified. This work uses the Unet architecture (originally developed for biomedical image segmentation) to segment radio sources from radio astronomical images with 99.8% accuracy. After segmenting the sources we use OpenCV tools to detect the sources on the mask images, then the detection is translated to the original image where borders are drawn around each detected source. This process automates and simplifies the pre-processing of images for classification tools and any other post-processing tool that requires a specific source as an input.Thesis (MSc) -- Faculty of Science, Physics and Electronics, 202
Social entrepreneurship and the tensions between social, environmental, and business goals: a case of Uganda
Get PDFClimate change has resulted in a global environmental crisis, with deforestation, plastic pollution, soil degradation, and species extinction increasing. Businesses have contributed significantly to these issues through their manufacturing processes. While many companies have implemented sustainability initiatives, systemic environmental problems persist, and businesses continue to play a significant role. One reason for this is that profit often takes priority over environmental considerations, limiting the effectiveness of sustainability programs. In response, environmental social enterprises are gaining prominence as a potent alternative to conventional business models. These enterprises are commended for placing environmental value generation at their core. Nevertheless, the simultaneous pursuit of financial, socio-economic, and environmental goals can instigate conflicts, including those related to performing, organizing, belonging, learning, and those spanning various levels of analysis. These conflicts impede the intended outcomes of these enterprises. This challenge is particularly under-addressed within contexts distinct from North America and Europe, like Uganda, which possesses unique geographic, political, and cultural attributes. Investigating how environmental social enterprises in Uganda navigate goal-related contradictions is pivotal for comprehending their internal mechanisms and strategies for economic sustainability. This study adopted a qualitative research strategy grounded in the interpretivist paradigm. Employing multi-qualitative methods, including interviews with 20 key decision-makers from 4 environmental social enterprises, the study delved into their experiences and practices. The findings revealed that these enterprises grapple with tensions stemming from their endeavors to balance economic, environmental, and social objectives. These tensions manifest both anticipated and unanticipated effects on organizational performance. Strategies for managing these tensions encompass collaborations with other organizations, experimentation with diverse management approaches, and acknowledging tensions as inherent and potentially beneficial. Nevertheless, the efficacy of these strategies hinges on the organizational context and systemic factors such as government regulations. While facing similar tensions, environmental social enterprises also adopt varied tactics—proactive, defensive, or a combination thereof—to address conflicts rooted in their goals. The chosen response method holds implications for ecological outcomes, underscoring the importance of framing tensions appropriately. Nonetheless, persistent tensions that resurface due to their dynamic nature pose a challenge. Consequently, addressing such tensions necessitates an iterative approach, often entailing a fusion of strategies. To address tensions between social, environmental, and economic objectives in environmental social enterprises, policymakers, governments, and social entrepreneurs should collaborate to institute supportive policies, encourage partnerships, invest in education and capacity building, and establish evaluation frameworks. Implementing these recommendations can adeptly manage such conflicts, fostering sustainable development and success.Thesis (MCom) -- Faculty of Commerce, Rhodes Business School, 202
Please Do Not Call It Human Right: A Southern Epistemological Perspective on the Digital Inclusion of People With Disabilities in South Africa
No full textSouth Africa is one of the world’s most unequal countries and has a well-known past of serious human rights abuses. Despite a progressive legal system, a relatively large economy, and one of the highest rates of Internet penetrations on the African continent, social inequalities extend to the digital domain. While digital inclusion for all is inscribed as a desirable goal in policy documents, in practice this is not the case. The rights based on which policy documents rely and from which much of the formal civil society sector draws is rooted in Western understanding of concepts such as “dignity”, “ability”, “inclusion”, and even “disability” itself. Employing a decolonial approach, this chapter problematises narratives about the digital inclusion of people with disabilities. It argues that for South Africans with disabilities, benefitting from government interventions and enforcing their rights depend to a large extent on their socio-economic background. An emerging body of research documents challenges associated with the new forms of exclusion/invisibility of people with disabilities online, such as inaccessible software/hardware, a steep learning curve, or the practice of hiding one’s disability to avoid pity or stigma. The ubiquity of digital technology makes digital inclusion of people with disabilities more of a necessity rather than a choice or a right
Diversity and Endemism of Southern African Gekkonids Linked with the Escarpment Has Implications for Conservation Priorities
Get PDFSouth Africa is recognised for its high reptile diversity and endemism, specifically among lizards. Phylogenetic diversity, endemism, and richness can have clear implications or raise important questions in a range of fields, and most urgently in conservation. Among squamate reptiles, these indices are very commonly associated with high temperatures and topographic heterogeneity. Indeed, mountainous biogeography has been a critical driver in the radiation of the family Gekkonidae within the subregion. Here, we assess the species richness, diversity, and endemism of Gekkonidae species inhabiting South Africa, Lesotho, and Eswatini, accounting for phylogenetic relationships. We also employ the CANAPE method to identify regions that have neo- and/or paleoendemics. Southern African gekkonids appear to be most diverse and show high levels of endemism in three regions of Southern Africa: the northwestern Richtersveld, the escarpment running west to southeast, and the northeastern escarpment in the Limpopo province. Implications for conservation priorities are discussed
Addressing global challenges in geographical education: perspectives from the CGE Springer book series
No full textA key responsibility of academic scholarship is to further develop the field under investigation and to critically evaluate how it is responding to challenges from both inside and outside of the field of enquiry. This article offers an overview of whether the scholarship published in geography education (GE) from the perspective of the Commission for Geography Education (CGE) Springer book series is sufficiently engaging with contemporary global challenges and education discourses. Guided by the 2022 IGU-CGE Rennes Conference’s three axes: curriculum renewal, pedagogy and teacher training, the article discusses the findings of a meta-analysis of the six books published in the CGE Springer series in the past five years (2017 to 2021).The thematic analysis shows that GE is engaging with global realities of environment and sustainability, climate change, globalisation and the importance of powerful knowledge for understanding and engaging with complex, often controversial issues. But there are also gaps and absences in terms of who is producing knowledge and from where knowledge is being generated. We contend that the book series is well placed to support the community in addressing global concerns and the related transformative pedagogies needed to enact them. However, we also suggest that the publications at this point in time have yet to fully grasp the specifics needed for the challenges that lie ahead. Some preliminary considerations for advancing the field through future publications are offered. These may be useful to GE researchers and scholars in different regions of the world
Learning to make a difference: Small-scale women farmers in social learning spaces for climate action
Get PDFHow do women farmers in Africa learn about climate change? What is quality climate change learning for farmers? How do farmers interface new knowledge with their long-held and trusted traditional knowledge? How do we evaluate learning at farm level and beyond? Using Okoli’s theory mining review, I untangled a tripartite knot of social learning literature to find Social Learning Theory (SLT) suitable for a study to explore my practical and scholarly curiosity as reflected in the above questions. Wenger’s theory of Social Learning emerged as the most appropriate for my research. The second phase of my study explored the climate change learning and practice terrain for small-scale women farmers, analysing the connection between learning, practice, and the resultant value in two case study areas, municipalities in the Amathole District of the Eastern Cape Province of South Africa. In addition to a paper on SLT mining review that unravels and chooses between the strands of social learning, the two case studies resulted in three articles that responded to the study’s objectives and the research questions. The thesis is introduced and synthesised through five 'book-end' chapters, as well as through these four articles. What were my findings? In the first case study, in the drought-stricken Raymond Mhlaba Municipality, I gathered the data through individual semi-structured interviews with farmers, extension officers and representatives of the involved organisations. I also conducted a group interview with farmers and analysed documents to supplement interview data. I analysed the data using concepts of Communities of Practice (CoP) and SLT to map out the learning and practice landscape. I discovered a constellation of CoPs interconnected by the shared drive for adaptive water management. The constellation is made up of tertiary institutions, government departments, non-governmental organisations and farmers of varying experiences and competencies, with women emerging as the more proactive gender, and state-led extension services being willing but overstretched and under-resourced. SLT effectively traced the apparent fragmented learning within and outside the CoPs and the sudden and extensive shifts in the CoP boundaries, especially in the context of COVID-19 pandemic lockdowns and the increased adoption of digital learning platforms. Despite the richness and diversity brought by the emergent new learning networks that involve participants in the province and further afield, the adoption of digital learning platforms worsened the existing generational digital divide among farmers. iii In the second case study, in the water scarce Buffalo City Metropolitan Municipality, I adopted the Value Creation Framework (VCF) to conduct an ethnographic evaluation that used semi-structured interviews, participant observation and document analysis of the learning experiences of women farmers in a social movement on agroecology. I found that the farmer-centred learning approach of the movement has created value for the farmers involved, evidenced by the adoption of agroecology by over 2700 members (including new urban farmers who are occupying open spaces typically used as dumpsites). The learning approach has facilitated expansive learning, enhanced resource mobilisation, new collaborations, partnerships, and seed sharing networks. Additionally, it necessitated context-appropriate and transformative changes to intersectional justice issues associated with historical inequalities in access to land and water and gender discrimination, leading to improved practices, new access to markets and improved quality yields. These are examples of immediate, potential, applied, realised, orienting, enabling and transformative as well as strategic value, as defined by the VCF. In reflecting on how women farmers learn in these social learning spaces I elucidate the learning impact pathways and local contextual influences in shifting CoP boundaries, domains, and practices during the climate crisis as it intersects with other compounding factors. I generated insights that could be useful for stakeholders in the agricultural (extension) sector to build better pathways for emancipatory and empowering expansive social learning in contexts characterised by resource constraints, but also by strong women-led agency. Such learning could make a difference and cushion small-scale farming from collapse especially in times of unprecedented changes. The agroecology movement and associated communities of practice explored in this study create transformative social learning spaces that are able to respond to climate change, and hence a model that state-led extension might want to adopt in other resource-constrained contexts.Thesis (PhD) -- Faculty of Education, Education, 202
A socio-political and historical perspective of linguistic prescriptivism in relation to the African languages of southern Africa
No full textMajor languages in southern Africa evolved from oral to the written mode within particular socio-cultural and political milieu from the eighteenth century. In the postcolonial period, some African countries established regulatory bodies, while others maintained those established during the colonial period to oversee the development and use of African languages. The quest for uniformity manifested itself in a prescriptive approach to the orthographies and grammars of southern African languages. This chapter looks at how prescriptivism emerged from those socio-political and historical processes to become a feature in the development and use of African languages in southern Africa
Euendolithic Infestation of Mussel Shells Indirectly Improves the Thermal Buffering Offered by Mussel Beds to Associated Molluscs, but One Size Does Not Fit All
Get PDFMussel beds form important intertidal matrices that provide thermal buffering to associated invertebrate communities, especially under stressful environmental conditions. Mussel shells are often colonized by photoautotrophic euendoliths, which have indirect conditional beneficial thermoregulatory effects on both solitary and aggregated mussels by increasing the albedo of the shell. We investigated whether euendolithic infestation of artificial mussel beds (Perna perna) influences the body temperatures of four associated mollusc species during simulated periods of emersion, using shell temperature obtained via non-invasive infrared thermography as a proxy
Darknet Traffic Detection Using Histogram-Based Gradient Boosting
Get PDFThe network security sector has observed a rise in severe attacks emanating from the darknet or encrypted networks in recent years. Network intrusion detection systems (NIDS) capable of detecting darknet or encrypted traffic must be developed to increase system security. Machine learning algorithms can effectively detect darknet activities when trained on encrypted and conventional network data. However, the performance of the system may be influenced, among other things, by the choice of machine learning models, data preparation techniques, and feature selection methodologies. The histogram-based gradient boosting strategy known as categorical boosting (CatBoost) was tested to see how well it could find darknet traffic. The performance of the model was examined using feature selection strategies such as correlation coefficient, variance threshold, SelectKBest, and recursive feature removal (RFE). Following the categorization of traffic as “darknet” or “regular”, a multi-class classification was used to determine the software application associated with the traffic. Further study was carried out on well-known machine learning methods such as random forests (RF), decision trees (DT), linear support vector classifier (SVC Linear), and long-short term memory (LST) (LSTM). The proposed model achieved good results with 98.51% binary classification accuracy and 88% multi-class classification accuracy