Search

The creation and deployment of digital technologies throughout the agro-food system contribute to achieving Sustainable Development Goal nine. However, various digitalization studies offer limited insight into the policy issues and solutions around emerging technologies and big data systems in agriculture. There is a need for an additional understanding of how agricultural policymaking should respond to the usage of digital technologies in the agri-food sector. Furthermore, evidence is limited on how existing agricultural government programs such as land reform can be linked with digitalization policy. This systematic review of literature sought to determine the transformation that is needed in the political and economic environment for the digital revolution to take place within South African land reform agriculture. A thematic analysis of data sampled from ProQuest Central, Scopus, Dimensions, and Google Scholar reveals five areas of intervention from agricultural policymaking. The digital revolution in agriculture can be brought about by transforming digital infrastructure, data interoperability and governance, digital markets, the compatibility of government incentives with the private sector, and the digital cultural landscape. This essay contributes to agricultural policy and decision-making dialogues that pay attention to digital technologies and land reform programs in South Africa.

This study sought to develop an index for agricultural digitalization by applying composite confirmatory analysis (CCA). Another aim was to determine the factors that affect the development of digitalization in PLAS farms. Data on the indicators of the three dimensions of digitalization were collected from 300 Proactive Land Acquisition Strategy (PLAS) farms in South Africa using semi-structured questionnaires. Confirmatory composite analysis (CCA) was employed to reduce the items into three digitalization dimensions and ultimately to a digitalization index. Standardized digitalization index scores were extracted and fitted to a linear regression model to determine the factors affecting digitalization. The results revealed that the model shows practical validity and can be used to measure digitalization as measures of fit (geodesic distance, standardized root mean square residual, and squared Euclidean distance) were all below their respective 95%quantiles of bootstrap discrepancies (HI95 values). Therefore, digitalization is an emergent variable that can be measured using CCA. The average level of digitalization in PLAS farms was 0.02 and varied significantly across provinces. Although farmers have attempted to digitalise their farms, there are still minimal levels of digitalization in PLAS farms. The results further reveal different digitalization patterns. As judged by the estimated weights of various dimensions of digitalization, the use of digital technologies to collect, store, analyse, and disseminate (CSAD) farm-related data contributed more towards the digitalization index. The second most important component of digitalization was automation digitalization. In contrast, value chain digitalization was the least significant contributor. The factors that significantly influence digitalization were age, gender, farm type, network type, and cellular data type. Since PLAS farmers have not embraced much digitalization, it is important to focus on awareness and capacity building. A balanced approach to digitalization would benefit PLAS farms by ensuring that strategies to integrate digital solutions within the value chain are developed. To foster and support the digitalization in PLAS farms, policymakers and stakeholders should tailor their strategies to fit specific socioeconomic factors.

Ontlametse T Bareng,1,2 Wonderful T Choga,1,2 Segomotso T Maphorisa,3 Sekgabo Seselamarumo,1 Kaelo K Seatla,1 Patrick T Mokgethi,1,4 Dorcas Maruapula,1,4 Mompati L Mogwele,1 Doreen Ditshwanelo,1,5 Natasha O Moraka,1 Irene Gobe,2 Modisa S Motswaledi,2 Joseph M Makhema,1,6 Rosemary Musonda,1 Roger Shapiro,1,6 Max Essex,1,6 Vlad Novitsky,1 Sikhulile Moyo,1,2,6 Simani Gaseitsiwe1,6 1Botswana Harvard AIDS Institute Partnership, Gaborone, Botswana; 2School of Allied Health Professions, Faculty of Health Sciences, University of Botswana, Gaborone, Botswana; 3Ministry of Health and Wellness, Republic of Botswana, Gaborone, Botswana; 4Department of Biological Sciences, Faculty of Science, University of Botswana, Gaborone, Botswana; 5Department of Biological Science and Biotechnology, Botswana International University of Science and Technology, Palapye, Botswana; 6Department of Immunology and Infectious Diseases, Harvard T.H. Chan School of Public Health, Boston, MA, USACorrespondence: Simani Gaseitsiwe, Botswana Harvard AIDS Institute Partnership, Private Bag BO320, Bontleng, Gaborone, Botswana, Tel +267 390 2671, Fax +267 390 1284, Email sgaseitsiwe@bhp.org.bwPurpose: Monitoring HIV-1 drug resistance mutations (DRM) in treated patients on combination antiretroviral therapy (cART) with a detectable HIV-1 viral load (VL) is important for the selection of appropriate cART. Currently, there is limited data on HIV DRM at low-level viremia (LLV) (VL 401– 999 copies/mL) due to the use of a threshold of VL ≥ 1000 copies/mL for HIV DRM testing. We here assess the performance of an in-house HIV drug resistance genotyping assay using plasma for the detection of DRM at LLV.Methods: We used a total of 96 HIV plasma samples from the population-based Botswana Combination Prevention Project (BCPP). The samples were stratified by VL groups: 50 samples had LLV, defined as 401– 999 copies/mL, and 46 had ≥ 1000 copies/mL. HIV pol (PR and RT) region was amplified and sequenced using an in-house genotyping assay with BigDye sequencing chemistry. Known HIV DRMs were identified using the Stanford HIV Drug Resistance Database. Genotyping success rate between the two groups was estimated and compared using the comparison of proportions test.Results: The overall genotyping success rate was 79% (76/96). For VL groups, the genotyping success was 72% (36/50) at LLV and 87% (40/46) at VL ≥ 1000 copies/mL. Among generated sequences, the overall prevalence of individuals with at least 1 major or intermediate-associated DRM was 24% (18/76). The proportions of NNRTI-, NRTI- and PI-associated resistance mutations were 28%, 24%, and 0%, respectively. The most predominant mutations detected were K103N (18%) and M184V (12%) in NNRTI- and NRTI-associated mutations, respectively. The prevalence of DRM was 17% (6/36) at LLV and 30% (12/40) at VL ≥ 1000 copies/mL.Conclusion: The in-house HIV genotyping assay successfully genotyped 72% of LLV samples and was able to detect 17% of DRM amongst them. Our results highlight the possibility and clinical significance of genotyping HIV among individuals with LLV.Keywords: in-house genotyping, low-level viremia, samples, HIV-1C drug resistance testing