Portail "HAL-Francophonie Afrique et Océan Indien"
Not a member yet
43052 research outputs found
Sort by
Investigation of PFAS/nPd/μZVI interactions: mechanism, performance, and nPd/μZVI reuse
No full textInternational audienceDue to their persistence, toxicity, and environmental spread, PFAS have been extensively studied for removal since the early 2000s. This study investigates the interactions between nano-sized palladium-coated zero-valent iron microparticles (nPd/μZVI) and perfluorooctanoic acid (PFOA) or perfluorooctanesulfonic acid (PFOS) to examine the mechanism and evaluate the performance of these two PFAS abatement.Batch experiments were performed under anoxic conditions (1–72 h, 90 °C) using 1 g of ultrasonicated (45 Hz) slurries of nPd/μZVI and 9.29 to 485 mg L−1 of PFAS across a pH range of 1–14. After the reaction and subsequent settling, samples of the aqueous phase were collected for pH measurement, fluoride analysis using a fluoride ion-selective electrode (F-ISE), and quantification of residual PFOA or PFOS and potential byproducts using ultra-high-performance liquid chromatography coupled with mass spectrometry (UHPLC-MS). Methanol used to wash nPd/μZVI under oxic conditions was also analyzed to differentiate between PFAS adsorption and their defluorination. The washing step enabled reuse of the nPd/μZVI, which were characterized by XRD and SEM-EDS before interaction, after interaction, and following methanol washing.PFAS sorption, governed by surface complexation and hydrophobic interactions, is the predominant PFAS abatement mechanism. It is driven by the anoxic galvanic corrosion of nPd/μZVI (AGC-nPd/μZVI), which modifies the physical and chemical conditions of the liquid phase and the surface chemistry of nPd/μZVI as a function of pH. Under the investigated conditions, a linear sorption model accurately described PFAS abatement at pH 4, where nPd/μZVI were shown to be reusable for at least 4–5 cycles without significant loss of sorption efficienc
Evidence of vertical stratification in marine environments: insights from passive acoustic monitoring in French Polynesia
No full textInternational audienceThe stratification of species distribution in marine environments is well-documented, reflecting the influence of depth-related factors such as light availability, temperature, pressure, and habitat structure on the organization of marine communities. This study aimed to determine whether this stratification can also be highlighted using Passive Acoustic Monitoring (PAM), a technique proven effective in the study of various marine habitats. Sounds were recorded at 300 m in July 2022 off Moorea Island (French Polynesia) using a SNAP acoustic recorder. These data were compared to both those collected at shallower depths at Moorea Island (20, 60, and 120 m) and to data from other Polynesian islands obtained during a previous, non-concurrent study. Fish sounds in the recordings were manually identified. Richness and abundance of sounds from 20 m to 300 m depth were compared to assess whether fish communities in the altiphotic reef, upper mesophotic reef, lower mesophotic reef, and rariphotic zone could be distinguished based on the sounds they produce. Comparisons between zones over a 24-h period revealed for the first time the existence of acoustic stratification, with a fourfold decrease in species richness, from the altiphotic reef to the rariphotic zone, along with a general decline in sound abundance. However, this general pattern was not uniform in terms of abundance throughout the day, as sound abundance increased with depth at dusk from 950 sounds per hour at 20 m to 3600 at 300 m. It further confirms that diel activity patterns were also evident in light-deprived zones. This first exploration of vertical acoustic stratification based on fish sounds highlights the undeniable contribution of PAM to studies on the ecology of deep-sea communities
Perceptions of tuberculosis and home-based infection screening among families and providers in Madagascar: a qualitative study
No full textInternational audienceBackground Madagascar is a country with a high tuberculosis (TB) burden, and the incidence of the disease has remained unchanged since 2013. TB detection largely relies on passive strategies, with little attention given to TB infections (TBIs), which are silent reservoirs that fuel future active TB cases.Method This qualitative study investigated how Malagasy communities perceive TB and an innovative approach for TBI screening and care. This approach was developed through the APRECIT operational research project in partnership with the National Tuberculosis Control Program (NTP). Semi-structured in-depth interviews were conducted between March and May 2023 with a sample of stakeholders and health workers (26 HWs) at the national, regional and community levels and focus groups with household contacts (46 HHCs) of TB patients.Results. The fight against TB remains challenging in Madagascar. Community knowledge of TB is limited, and the disease is perceived as shameful. Individuals with TB and their families face stigmatization within their own family and the community. At the provider level, HWs reported challenges in promptly detecting and treating TB people with TB due to structural constraints within the health system and barriers at the household (HH) level. The introduction of home-based screening for TBI was perceived as a way to interrupt disease progression and help reduce stigma. For HWs, this strategy helps to limit TB transmission. However, HWs expressed concerns about giving treatment to asymptomatic people and the replicability of the strategy at the national level. ConclusionImproving communication about TB and implementing a home-based screening strategy for HHCs to identify individuals at high risk of TB progression have the potential to improve TB control in Madagascar.</div
Les inégalités scolaires à Mayotte et à La Réunion : vers des pratiques éducatives plus équitables
No full textInternational audienceSi les enquêtes internationales confirment depuis des décennies le rôledéterminant du milieu socioéconomique dans la réussite scolaire, lesinégalités territoriales apparaissent comme un facteur prééminant. C’estnotamment le cas dans les territoires ultramarins, souvent marginalisés auniveau des politiques publiques, comme le soulignent de nombreux rapportsinstitutionnels.Cet ouvrage collectif, porté par plusieurs enseignants-chercheurs dulaboratoire ICARE (Institut coopératif austral de recherche en éducation),à Mayotte et à La Réunion, propose une analyse des inégalités scolaires etd’apprentissage, en croisant des dimensions sociales, linguistiques, culturelleset territoriales. Il s’appuie sur des enquêtes de terrain qui mettent au jourles tensions entre prescriptions institutionnelles et réalités locales, tout envalorisant les initiatives pédagogiques et les ressources locales.Les contributions qui composent cet ouvrage abordent, entre autres, les enjeuxde la formation initiale et continue, la professionnalisation des enseignants,les pratiques langagières en contextes plurilingues, et la nécessaire contextualisationdes supports et des pratiques éducatifs. En articulant les apports dela recherche et les voix d’acteurs de terrain, les auteurs offrent des pistesconcrètes pour mieux comprendre les défis d’un système scolaire confrontéà la diversité des contextes. Ils invitent également à interroger les normes, àreconnaître les compétences parfois invisibilisées, et à penser ces situationsen marge comme des lieux et des situations d’innovation et de transformation.En ce sens, cet ouvrage s’adresse autant aux étudiants qu’aux chercheurs,formateurs et professionnels de l’éducation désireux de découvrir ces deuxdépartements et leurs spécificités ainsi que des perspectives croisées enmatière d’égalité éducative, afin d’aspirer à une école plus inclusive
Microcrédit à fidélisation plasticienne : trajectoiresd’inclusion et optimum de plasticité
No full textEM-SEC: Efficient Multi-head Set-valued Evidential Classification
No full textInternational audienceIn machine learning and deep learning, uncertainty quantification helps to accurately assess a model's confidence in its predictions, enabling the rejection of uncertain outcomes in safety-critical applications. However, in scenarios involving AI-assisted decision-making, proposing multiple plausible decisions can be more beneficial than either not making any decisions or risking incorrect ones. Set-valued classification is a relaxation of standard multiclass classification where, in cases of uncertainty, the classifier returns a set of potential labels instead of a single label. Current methods for set-valued classification often suffer from high computational complexity or fail to adequately quantify uncertainty. In this paper, we introduce a novel, computationally efficient approach to set-valued classification leveraging evidential deep learning and subjective logic, explicitly providing a measure of classification uncertainty. Our method employs a dual-head architecture: one head conducts multiclass evidential classification, while the other suggests candidate label sets when uncertainty is high. The proposed approach has linear worstcase computational complexity with respect to the number of classes. Extensive evaluation on several benchmark datasets demonstrates that our method showcases comparable performance to baseline set-valued methods, while being up to 23 times faster at inference on the benchmark datasets
Aboveground biomass and growth of smallholder Eucalyptus robusta under low starter fertilization and weed competition in Madagascar
No full textSource Agritrop Cirad (https://agritrop.cirad.fr/614702/) * Autres projets (id;sigle;titre): ;DIABE;(EU) Développement Intégré, Aménagement et Bois-Energie//International audienceSmallholder Eucalyptus plantations (EP) in Madagascar Central Highlands (MCH) address substantial fuelwood demand and reduce pressure on natural forests. However, their sustainability is challenged by low soil fertility and inadequate management. While fertilization increases tree growth, high mineral fertilizer costs limit its use by smallholder farmers. Both biomass estimation equation and impact of fertilization on smallholder EP in MCH remain poorly documented. This study aimed to evaluate the impact of low starter mineral fertilization (15 kg.ha-1 N, 12 kg.ha-1 P, and 17 kg.ha-1 K) and weed competition on Eucalyptus robusta growth and aboveground biomass (AGB) while providing allometric equations for AGB estimation. Dendrometric data were collected from four stands aged 2 to 6 years, with fertilized and non-fertilized parts. AGB of 16 trees per stand (8 fertilized, 8 non-fertilized) were destructively measured. AGB of trees were estimated by regression based on tree circumference, height and their combinations. The results indicated circumference as the best single variable predictor (R2 > 0.90) for all tree compartments (leaves, branches, trunk) and treatments. Fertilization significantly improved global tree survival by 7% and increased tree height by 3.1 m and circumference by 8.3 cm in 6 yo stands. AGB per hectare doubled with fertilization, reaching 55.3 Mg.ha-1 at 6 yo compared to 29.2 Mg.ha-1 in non-fertilized plots. Aboveground carbon stock reached 4.6 MgC.ha-1yr-1 in fertilized plots. Weed cover and biomass had likely negative linear relationships with tree AGB and survival rate. This study provides robust allometric equations for biomass estimation and highlights that even low fertilizer application combined with effective weed control can significantly enhance AGB production in smallholder EP
Non-Termination of Term Rewrite Systems Using Pattern Unfolding
No full textInternational audienceWe present a revisit, based on a new unfolding technique, of an approach introduced in term rewriting for the automatic detection of infinite non-looping derivations from patterns of rules
