HAL - La Rochelle Université, Archives Ouvertes
Not a member yet
19516 research outputs found
Sort by
Breaking family bonds: pair disruption alters female adolescent spatial neophobia but not other personality traits or corticosterone stress response in zebra finches (Taeniopygia guttata)
International audienceAdolescence is a sensitive period because it is associated with the ontogeny of key neurological, physiological and behavioural systems. These systems can be permanently altered by social disruption during adolescence and therefore impair an individual's ability to cope with its environment later in life. We tested whether pair disruption of captive zebra finches (Taeniopygia guttata) during emancipation affects the family social structure, with potential consequences for the nutritional status, personality and corticosterone stress response of juveniles. We experimentally manipulated the social environment of 22 families during emancipation by replacing fathers with unfamiliar males (experimental families) or leaving them unchanged (control families) and monitored the prevalence of affiliative, agonistic and sexual interactions between family members. We assessed offspring growth, timing of nutritional independence, body condition as well as five personality traits and the corticosterone stress response to isolation. While we observed more agonistic and sexual behaviours in experimental families, we also observed more affiliative behaviours between experimental siblings and more maternal provisioning of the experimental juveniles. Among all the traits we tested, we only found a sex-dependent effect of the experimental treatment on spatial neophobia, suggesting that pair disruption may have long-term consequences for females' ability to cope with new environments. However, our findings suggest overall that the emancipation phase is less sensitive to social environment in comparison to the prenatal and early postnatal periods and that nutritional and social buffers may mitigate the lasting impacts of pair disruption on adolescent behavioural and stress response profiles in altricial species such as the zebra finch
Urban mixtures of non-essential trace metal elements affect the behaviour and stress levels of a passerine bird in a laboratory experiment
International audienceTrace Metal Elements (TME) are known for their toxicity, which is higher for non-essential TME (nTME) than for essential TME (eTME). However, little is known about the toxicity of eTME and nTME mixtures at concentrations relevant to the urban environment. Since birds are commonly used for biomonitoring, we investigated the behavioural and physiological responses of a passerine model -the zebra finch (Taeniopygia castanotis formerly known as Taeniopygia. guttata) -to TME mixtures in a laboratory setting. Eighty males were divided into 3 exposure groups and 1 control group. Each group was then divided into ad libitum and intermittent feeding groups to better represent challenging natural conditions. The birds were exposed to TME at urban relevant levels via drinking water for 2 months. Three mixtures were tested: one containing only nTME (cadmium, arsenic, lead), one containing the same nTME plus eTME (zinc, copper, nickel), and one containing the same nTME and double the amount of eTME. Our results showed that the more contaminated the birds were with nTME, the less time they spent at the feeders and the more stressed they were. The birds showed an aversion to the mixtures containing TME, particularly eTME. This suggests that birds are able to detect TME in water. Birds exposed to eTME accumulated more copper and less cadmium than the birds exposed only to nTME, highlighting the antagonism between the two types of TME. This study demonstrates that nTME can directly impact the behaviour and stress levels of birds at urban relevant doses
Critical assessment of an aging phenomenological model for building materials
International audienceThe description of aging of construction materials represents a significant issue. According to previous studies, aging is addressed by using experimental approaches which consist in submitting materials to accelerated aging designs in the laboratory. Then, the effect is highlighted by performing measurements at different stages of material aging. These discrete configurations are considered to evaluate the impact of aging at the building scale. Within this context, this research proposes a phenomenological model that describes the aging of a material over time. It assesses the variation in porosity depending on drying and imbibition loads, each expressed by means of an indicator. Also, a cumulative and a moving approach are used to express each load. Afterwards, the model unknown parameters are estimated by solving an inverse problem. The results of the estimation are obtained based on experimental observations of an aging experimental design on hemp concrete. For each load approach, it is shown that the model fits good most of porosity measurements. In particular, the moving approach is better adapted to study the aging of materials. Further, the model of porosity is integrated into equations of heat and mass transfer within porous media. The objective consists in evaluating the impact of aging, with a continuous approach, on the energy efficiency of the material at the wall scale. Based on a case study, the impact is evaluated by assessing the energy efficiency for a first configuration without aging and a second with. Computations of sensible and latent heat fluxes reveal that thermal losses through the wall are underestimated by about 30% when aging effects are neglected. This finding highlights the significance of this phenomenon and encourages its consideration in future building energy simulations
Formal Concept Analysis beyond binary data: past and future in the GALACTIC platform
International audienceThe GALACTIC platform (GAlois LAttices, Concept Theory, Implicational Systems, and Closures) is a conceptual knowledge environment designed to support Formal Concept Analysis (FCA) at various levels of complexity. In this talk, we present the current state of development, highlighting both the capabilities of the stable version 0.4.0.0 and the trajectory introduced by the ongoing 0.5.0.0 redesign.Version 0.4.0.0 supports a mature FCA pipeline capable of handling complex and heterogeneous data. It extends the classical binary context framework to allow analysis over non-binary structures such as numerical intervals, multi-valued attributes, and symbolic sequences. This is made possible through a flexible plugin system, which enables the definition of custom data interpretations and the extraction of formal concepts using generalized predicate logic. Users can explore concept lattices enriched by sophisticated measures such as entropy or support.In contrast, the emerging version 0.5.0.0—currently under active development—focuses initially on a reimplementation of classical FCA mechanisms based strictly on binary contexts. While this version introduces significant architectural improvements (modularity, maintainability, and clearer separation of concerns), it does not yet integrate the advanced heterogeneous data analysis features of version 0.4.0.0. Our goal is to progressively reintroduce these capabilities within the new framework, while preserving mathematical rigor and performance.Through this presentation, we aim to showcase how GALACTIC bridges traditional FCA and modern data needs, and to share insights from our ongoing migration efforts between conceptually rich but older implementations and more sustainable contemporary architectures
Les archives de l’administration centrale du ministère des Affaires étrangères et du réseau diplomatique et consulaire
International audienc
Inventory of landslides affecting the coastal cliffs of Charente-Maritime (France): Typology, frequency, surface area and observation methods
International audienceCharente-Maritime is a French department featuring 90 km of rocky coastline, of which 60 km remain unprotected by artificial longitudinal coastal defenses (Figure 1.A). The dynamics of these rocky coasts have been relatively understudied compared to other French regions, such as Normandy (Lissak et al., 2013; Letortu et al., 2014) or the Basque Country (Prémaillon, 2018; Martins et al., 2022; Guillen, 2024). As a result, the processes driving cliff retreat and coastal evolution in Charente-Maritime remain poorly understood, despite the fact that these cliffs represent a substantial portion of the department's coastline. Both human interests (e.g., coastal pathways, residential areas) and environmental priorities (e.g., sensitive natural habitats) are directly affected by the retreat of these rocky coasts. Without adequate adaptation to climate change, coastal erosion could endanger approximately 46.000 homes and 5.000 businesses along all types of coastlines in Charente-Maritime by 2100, with potential financial losses estimated at 15 billion euros (Hédou et al., 2024). In this context, it is crucial to identify rocky coastlines exposed to a high erosion hazard in order to implement appropriate coastal management and adaptation strategies.This study investigates the retreat mechanisms of the rocky coasts in Charente-Maritime through an inventory of landslides. It aims to address both scientific and methodological questions: (i) How are these rocky coasts retreating? (ii) What is the relationship between the frequency and the volumes of landslides? (iii) Which acquisition method — pedestrian surveys or drone-based inventories — proves to be the most comprehensive?Seven study sites were selected in Charente-Maritime based on their diverse geographical and geomorphological characteristics (Figure 1.A). Quarterly data acquisition was conducted through field surveys, complemented by drone surveys performed every six months. These two methods were selected for their ability to capture data over short time intervals and to detect landslides involving small volumes (≤ 10 m³). Since November 2023, five field surveys and two UAV (unmanned aerial vehicle) acquisitions have been conducted across all study sites. This approach has already documented hundreds of events, largely due to a succession of storms during the winter of 2023–2024 combined with very high tidal coefficients.According to the scientific literature (Sunamura, 1992; Dikau et al., 1996; Hungr et al., 2014), five main types of landslides responsible for the dismantling of the cliffs in Charente-Maritime have been identified: rockfall, rock topple, slide, flow, and debris flow. Initial results from surveys conducted between November 2023 and December 2024 reveal that rockfalls (x̄ = 15 m³) and debris flows (x̄ = 11 m³) are the most frequent modes of cliff retreat at the departmental scale, despite involving relatively small volumes due to the low height of the cliffs. At the scale of individual study sites, the hierarchy of processes varies depending on the specific characteristics of each site, with lithology playing a key role in determining the mechanisms of cliff dismantling.The statistical sample obtained from the inventory confirms that the rockfalls affecting the cliffs of Charente-Maritime are characterised by small volumes (x̄ = 17 m³; σ = 37.4 m³) but a high spatio-temporal frequency (about 20 rockfalls per kilometre per year). An initial assessment of the frequency/intensity relationship can be proposed at both the departmental and site-specific scales
R project code - Agazella.nutrient.release.Ker
Assessing the variability in the multi-nutrient content of scats of Antarctic fur seals in Kerguelen and using this data to estimate their nutrient release.This R project is associated to the article Gilbert et al. (2025), https://doi.org.10.1016/j.scitotenv.2025.179195.This R project contains all the analysis of data conducted for this project from data exploration to modeling and result analysis.Targets set upThe project uses the targets packages (https://books.ropensci.org/targets/), and is therefore organized as follows:the R folder contains scripts containing functions and only functions (no direct command)functions are called in the _targets.R script at the root of this project, in items of a unique list. Each item in a list is defined by function tar_target(y, function(x)), in which y is a target object created by function(x) where function must be defined in the R folder and x must be a known object (usually, a target objet previously defined in the list)Each object from the list can be read using targets::tar_read(object) or loaded in the local environment with targets::tar_load(object).The full analysis can be run using targets::tar_make(), and only tasks associated to objects that are not up-to-date will be run.Note that the folder "work-in-progress" contains R scripts with direct commands which were used throughout the development of the project.Renv set upThe project also uses the Renv package (https://rstudio.github.io/renv/articles/renv.html). It helps in creating reproducible environments into R projects. Basic commands regarding Renv (e.g., update the environment) are contained in /dev-scripts/set-up-project.
Canevas performe!
Consortium Huma-Num CANEVAS https://canevas.hypotheses.org/5788The article discusses methodological advancements in the e-spect@teur project, which focuses on the annotation and analysis of audiovisual recordings related to the performing arts and live performance. It emphasizes the crucial importance of structuring these contents through appropriate knowledge organization systems, due to the large volumes of data generated by collective annotation facilitated by the e-spect@teur tool. The project adopts a thesaurus-based model, offering the necessary flexibility to account for the diversity of artistic practices while ensuring data interoperability. The collaborative and evolving management of these vocabularies is carried out using the OpenTheso tool. This work is supported institutionally by Huma-Num, the CANEVAS consortium, and the Institut des Amériques, all of which contribute to enhancing research, analysis, and valorization capabilities for audiovisual corpora in live performance.Le billet traite des avancées méthodologiques du projet e-spect@teur, consacré à l’annotation et à l’analyse de captations audiovisuelles liées aux arts de la scène et au spectacle vivant. Il souligne l’importance cruciale de structurer ces contenus par le biais de systèmes d’organisation des connaissances adaptés, en raison des volumes importants de données générés par l’annotation collective facilitée par l’outil e-spect@teur. Le choix s’est porté sur un modèle de type thésaurus, offrant la flexibilité nécessaire pour rendre compte de la diversité des pratiques artistiques tout en garantissant l’interopérabilité des données. La gestion collaborative et évolutive de ces vocabulaires est assurée grâce à l’outil OpenTheso. Ce travail s’inscrit dans un cadre institutionnel soutenu, notamment par Huma-Num, le consortium CANEVAS et l’Institut des Amériques, qui accompagnent cette démarche visant à enrichir les capacités de recherche, d’analyse et de valorisation des corpus audiovisuels du spectacle vivant