Eduardo Mondlane University

The South-West Indian Ocean (SWIO) displays marked spatial variations in physical forcings, driving in turn distinct seasonal biogeochemical regimes. Mesoscale eddies in the Mozambique Channel drive large-scale redistribution of nutrients and plankton, while eastern Madagascar waters are highly oligotrophic. However, the spatial organization and diversity of biogeochemical seasonal cycles across the SWIO remain insufficiently characterized on the basin scale. This study examines the seasonal and interannual variability of surface chlorophyll-a (Chl-a), used as a proxy for primary production. A phenoregion approach is employed, involving the grouping of areas exhibiting similar Chl-a phenology, that is, bloom timing (austral winter and/or summer) and the number of blooms per year (0, 1, or 2). 22-year of weekly satellite-derived Chl-a from the OC-CCI data set was used to generate normalized climatological seasonal cycles, which were subsequently clustered using a k-means algorithm. This approach identifies six distinct Chl-a phenoregions across the SWIO. Winter blooms dominate the basin and are likely associated with mixed-layer deepening and enhanced vertical nutrient supply, leading to temporally stable phenoregions. In contrast, summer blooms are restricted to coastal regions and south-eastern Madagascar, coinciding with enhanced terrestrial nutrient inputs during the wet season. In the Mozambique Channel, Chl-a exhibits limited seasonal variability, likely reflecting strong eddy-driven exchanges between coastal and offshore waters. Interannual analysis reveals stable phenoregion cores with consistent seasonality, while their boundaries exhibit variability, highlighting contrasts between persistent and highly intermittent biogeochemical regimes. The results obtained provide a basin-scale framework for linking physical dynamics to ecosystem variability in the SWIO.

Bioactive peptides (BP) have been investigated as potential functional feed additives due to their reported immunomodulatory and metabolic properties. This study evaluated the effects of dietary BP supplementation on growth performance, non-carcass components, serum biochemical profile and intestinal morphology in Texel lambs. Forty-five weaned lambs (31 castrated and 14 intact) were housed in pairs and fed for 58 days following a 20-day adaptation period. At the beginning of the experimental period, lambs averaged approximately 30 kg body weight, reaching about 45 kg at the end of the trial. Animals were assigned to three treatments: a control diet without BP and diets supplemented with 1–2 g BP per animal per day. Lambs were fed a total mixed ration based on corn silage and concentrate. Performance parameters, non-carcass traits, serum biochemical variables and distal jejunum morphometry were assessed. Data were analyzed using linear mixed models in software R. Dietary BP supplementation did not affect final body weight, average daily gain, feed efficiency, non-carcass components or intestinal morphometric parameters (P > 0.05). A significant increase in serum albumin concentration was observed in BP-supplemented groups; however, no consistent changes were detected in other biochemical indicators. Under the conditions of this study, BP supplementation at practical inclusion levels did not improve growth performance or intestinal morphology in Texel lambs. Future studies should explore different dosages, feeding durations and stress conditions to further evaluate the potential benefits of BP supplementation in ruminants.

Rapid urbanization has significantly increased traffic congestion, adversely impacting travel efficiency, fuel consumption, and environmental sustainability. To address this challenge, this paper proposes RATM-FL, a multi-layer RSU-assisted traffic management framework based on federated learning. In the proposed system, roadside units (RSUs) locally collect and process real-time traffic data to predict congestion within their coverage areas, while a cloud server aggregate selected RSU model updates to construct a global traffic prediction model. Since RSU deployment is constrained by cost, the cloud server manages regions not covered by RSUs. Unlike conventional federated learning, RATM-FL employs a traffic-aware RSU selection strategy, formulated as an integer linear programming problem, which dynamically selects influential RSUs based on vehicle density and energy consumption. This approach reduces communication overhead and energy usage without degrading system performance. Extensive simulations conducted using the SUMO traffic simulator demonstrate that RATM-FL outperforms existing traffic management schemes by significantly reducing travel time, fuel consumption, and CO2 emissions.

This article analyses how governance deficiencies shape the development of informal settlements in Mozambique, with a particular focus on informal self-construction of housing and basic infrastructure in contexts where the state’s presence is simultaneously limited and complicit. The study aims to analyse how residents effectively take on governance functions by producing housing and collectively organising basic services that would ordinarily fall under public authority mandates. The analysis focuses on the metropolitan area of Maputo, Mozambique, drawing on two neighbourhoods—Chamissava and Malhangalene—as case studies that represent, respectively, expanding peripheral areas and consolidated inner-city contexts. Through these cases, the study identifies systemic deficiencies in urban planning, construction oversight, and policy implementation, thereby revealing the limited effectiveness of existing housing strategies in responding to rapid urban change. Based on fieldwork conducted under challenging socio-political conditions and developed in collaboration with municipal officials, technical staff, and residents, the study highlights the everyday interactions between informal housing practices and weak governance arrangements. The findings show how communities strategically adapt to institutional shortcomings—and, at times, leverage political moments—to address pressing housing needs. In doing so, the article underscores the need for improved institutional coordination and more context-sensitive regulatory approaches to address the persistence of informal urban development in Maputo and similar urban contexts.

Introduction. The rapid integration of artificial intelligence (AI) into biomedical education has created an urgent need for a framework that can guide both learners and academic staff. In Spain, where institutions are adopting AI at speed but without homogeneous regulatory or pedagogical guidance, structured approaches are required to prioritize competencies and address ethical, legal and educational challenges. Methods. A Delphi study was coordinated by the Universidad Europea de Madrid, using the REDCap platform to manage iterative rounds. The project received financial support from the Spanish Society for Medical Education (SEDEM). Experts in medical education, including some with specific AI expertise, participated in a sequence of online questionnaires. Through anonymized feedback, participants evaluated, refined and prioritized a core of areas where the impact of AI seems essential. Results. Experts agreed on the need for skills extending beyond technical literacy, emphasizing critical thinking, interpretation of AI-generated outputs, bias awareness and professional responsibility. Ethical and legal considerations, particularly concerning privacy, transparency and decision-making, were strongly prioritized. Participants also highlighted the transversal nature of AI, suggesting that competencies should be embedded across curricula rather than treated as isolated content. Discussion. Despite institutional heterogeneity, consensus converged on areas that balance innovation with ethical safeguards. The results support the development of a competency-based framework capable of guiding curriculum design, informing faculty development and promoting responsible, evidence-informed use of AI while safeguarding professional autonomy.