L'Institut Agro Archives Ouverte
Not a member yet
4294 research outputs found
Sort by
Le productivisme sur la sellette. Ce que la colère des agriculteurs nous dit des impasses du système agri-alimentaire actuel
No full textInternational audienceNote de contenu : • Discours sur la crise et contre-feu critiques• Des réponses captives d’un modèle économique à bout de souffle• Écologie, alimentation et santé comme variables d’ajustement• Syndrome de Stockholm : des agriculteurs otages et collaborateurs d’un marché mondialisé• Soutenir les modèles agri-alimentaires émergents à l’échelle des territoire
Enactive Design-Based Research in Vocational and Continuing Education and Training
No full textInternational audienceThe purpose of this article is to introduce a design-based research (DBR) approach developed in the field of vocational and continuing education, which is grounded in a pragmatic and phenomenologically inspired enactivist approach to activity. As a design-based methodology, our activity-centered and enactive DBR approach aims to generate knowledge related to design and to identify relevant design principles. After detailing the particularities of an activity-centered and enactive DBR approach, we focus on the results pertaining to design knowledge by identifying two broad design principles for vocational education and training, and five enactivist inspired principles for training design. A significant practical implication for researchers and practitioners in vocational and continuing education and training is that these enactivist inspired design principles provide promising pathways to enhance the connectivity between (i) work experiences, (ii) work and training practices, and (iii) learning contexts
School Food Politics, Identity, And Indigeneity in the Peruvian Amazon
No full textThis content is PDF only.International audienceIn Peru, a National School Food Program called Qaliwarma provides food in public schools across the country since 2012. This chapter draws up the transformation of school food in two Amazonian indigenous peoples: Maijuna and Napuruna. Mothers appear to play a significant role in its concretization since they take turns cooking the meals every school day. They constantly negotiate with the Peruvian government as they grapple with the constraints of the program and deploy strategies to adapt it to their own agendas. In doing so they follow local logics and discretely or profoundly deviate from how the program has been designed by state policies. School food is understood by both groups as a key feature of becoming mestizo – non-indigenous – while they also perceive some mestizo food items as of low value or as suspicious. With each meal they negotiate the frontier between appropriation and resistance on their own terms
Adapting FLORSYS for climate change: implementing plant-plant competition for water in a 3D mechanistic model for predicting future crop/weed interactions and their consequences in arable cropping systems
No full textInternational audienceIntroductionWeeds compete with crops for light, minerals and water, and weed-related yield losses will probably increase with climate change (drought, heat waves), under the influence of competitive weeds, in particular for increasingly scarcer water resources (Storkey et al., 2021). Thus, cropping systems must be redesigned to make weed management low-input and climate-resilient. FLORSYS is a 3D mechanistic model (Colbach et al., 2021) that can be used to investigate such issues. It simulates the multi-annual dynamics of weeds and their harmfulness (e.g. yield losses) and benefits (e.g. trophic supply for crop auxiliaries) from cropping system (rotation, cultivars, cropping techniques) and pedoclimate. However, FLORSYS does not include all mechanisms relevant to climate change, in particular plant-plant competition for water. To better forecast future weed dynamics in arable cropping systems, this work aimed at developing a water competition submodel for FLORSYS.Material and methodsThe submodel was customised for FLORSYS and connected to other submodels (e.g. growth and phenology, light and nitrogen competition). When possible, it reused existing formalisms. The submodel was designed to be generic across crop and weed species, with few parameters.ResultsAs in Aschehoug et al. (2016), the submodel modelled: (1) water availability, demand, competition and uptake at the voxel (3D pixel) scale, as for light and nitrogen competitions (Munier-Jolain et al., 2013; Moreau et al., 2021) and (2) the consequences of water stress on photosynthesis and morphology at the plant scale (Figure 1).The potential water uptake of a plant in each voxel occupied by its roots is the minimum of (1) its water demand in the voxel, downscaled from total plant demand according to soil water distribution (formalisms from ‘Virtual Grassland’, Louarn and Faverjon (2018)), (2) the amount of water available to plants in the voxel (linked with the STICS soil submodel, Brisson et al. (2008)) and (3) the maximum amount of water roots can take up (experiment of Cournault et al. (2024)). Competition among plants with roots in the voxel only occurs if the available water is insufficient to meet their potential uptakes. After completing the loop across soil voxels, it is possible that some plants did not take up enough water to fulfil their demands in some voxels, while water remains in other occupied voxels. Thus, a second uptake loop is run across voxels to compensate for the initial insufficient uptake.The plant's total water uptake is the sum of water uptakes over all occupied voxels. For each plant, a daily water stress index is computed as “1 - the ratio of water uptake to water demand”. To account for past stresses, the daily indexes are combined into a relative linear combination over the plant life (since emergence), with a greater weight for recent stresses. Together with shading and nitrogen stress indexes, the water stress index can affect photosynthesis (according to DSSAT/APSIM formalisms, Ritchie (1998)), and plant morphology (new formalisms from Cournault et al. (2024), first presentation in this congress).DiscussionFLORSYS becomes the first crop-weed model to account for light, water and nitrogen competitions, with new formalisms accounting for maximum root water uptake and water-stress effects on plant morphology. The new submodel includes only 7 new parameters, in line with FLORSYS’ spirit. It disregards daily lateral soil water flows, but this is consistent with FLORSYS' focus on multiannual cropping system evaluation (Colbach et al., 2021).ConclusionThe new FLORSYS version is expected to improve the credibility of flora projections in the context of climate change. Once the model has been evaluated with field observations, it will be used in simulation studies and participatory workshops with farmers and crop advisors, to design sustainable low-input and climate-resilient cropping systems
Predicting Body Weight Through Biometric Measurements in Bolivian Llamas
No full textInternational audienceThe objective of this study was to evaluate the relationship between body weight (BW) and different biometric measurements in llamas (Lama glama) from the Bolivian highlands and to generate prediction models of BW. A total of 515 individual records of BW and biometric measurements were used. The measurements were taken on 202 males and 313 females aged between 0.5 and 5 years, and included: neck length (NL), withers height (WH), rump height (RH), heart girth (HG), body length (BL), abdomen circumference (AC), rib depth (RD), hip width (HW), pin bone width (PBW), thoracic width (TW), and back length (BKL). The relationships between BW and biometric measurements were developed using simple linear and multiple regression. For the evaluation, the relationship between the observed and predicted values of BW was determined by linear regression, the mean squared error of prediction (MSEP) and root MSEP (RMSEP); concordance correlation coefficient analysis was also used. The BW ranged from 22 to 122 kg. Regression equations between BW, HG and RD had an r2 of 0.94 and 0.92, respectively (RMSEP= 6.06 and 6.70 kg, respectively). The equations were highly precise (r2 >0.86) and accurate (Cb>0.98), with a reproducibility index > 0.92. The model efficiency (MEF) indicated a higher efficiency of prediction (MEF ≥ 0.86). Using a single predictor, HG and RD accounted for more than 92% of the variation in BW. Overall, HG may be used as a single predictor to predict BW in llamas maintained under the conditions of the Bolivian highlands
Les transitions : mouvements collectifs à accompagner et transformations silencieuses
No full textInternational audienceDans cet article, j’extrais quelques traits saillants émergeant des expériences collectives pour former aux transitions et à l’agroécologie menées dans les établissements d’enseignement technique agricole et qui ont été présentées lors des rencontres dédiées du 2 au 5 octobre 2023 à Toulouse et font partie du présent dossier. Je pars de deux constats. Le premier est celui de la richesse des expériences conduites et de l’expérience acquise par les acteurs concernant la manière de former aux transitions et à l’agroécologie. Ce qui était « un impensé de “enseigner à produire autrement” » (Mayen et al., 2022), semble désormais une évidence partagée, qui a irrigué les échanges : former aux transitions vers l’agroécologie implique des innovations pédagogiques et didactiques. Et les transformations que cela implique « supposent un temps long » (ibid., p. 29), celui de la transition pédagogique et didactique aux échelles des individus, des équipes, des établissements et, plus globalement, de tous les acteurs de l’institution « enseignement agricole ».Le deuxième constat est celui du non-spécialiste de la question des transitions que je suis. Dans les échanges, j’ai trouvé peu de références explicites à ce qui caractériserait une transition pédagogique et didactique pour former aux transitions et à l’agroécologie, à différentes échelles sur le territoire de l’enseignement agricole
Impact de la sècheresse sur la sinistralité des contrats d’assurance climatique sur récolte en France en contexte de réchauffement climatique récent
No full textInternational audienc
Automatic pre‐treatment and multiblock analysis of flavor release and sensory temporal data simultaneously collected in vivo
No full textInternational audienceProton Transfer Reaction-Time-of-Flight- Mass Spectrometry (PTR-ToF-MS) is an analytical chemistry technique that can be used for measuring the concentration of volatile organic compounds directly in the subjects’ noses (nosespace, in vivo analysis) during a tasting and over time. It can be combined with temporal sensory methods such as Temporal Dominance of Sensations (TDS) or Temporal Check All That Apply (TCATA) in order to obtain simultaneous sensory and physico-chemical signals. This paper aims to provide a methodology to analyze in vivo PTR-MS and temporal sensory data and illustrate it on a real dataset. First, relevant pretreatments of PTR-MS data were established, including breathing correction, blank periods removal and standardization. Then, a statistical multiblock analysis was presented: the Regularized Generalized Canonical Correlation Analysis (RGCCA). The versality of the approach was demonstrated, as it can be used to answer most of problematics (exploratory or supervised). Finally, this methodology is illustrated on a dataset of PTR-MS and TDS or TCATA data collected simultaneously. In this study, 16 semi-trained subjects evaluated 3 chocolates in TDS and TCATA on six flavor attributes (Spicy, Cocoa, Woody, Fruity, Roasty and Dry Fruits) with 2 replicates for each sensory method. Results showed that TCATA and TDS gave similar results, but TDS was shown to slightly better preserve the PTR-MS observed product configuration than TCATA. All computing tools developed in this work are freely available
Antimicrobial properties of porphyrins and corroles functionalized with carboxylic acids
No full textInternational audienc
Forget early exaggeration in t-SNE: early hierarchization preserves global structure
No full textInternational audienceAs a local method of dimensionality reduction, t-SNE requires careful initialization in order to preserve the data global structure to the best extent. In regular t-SNE, the low-dimensional embedding is initialized either randomly or with PCA; next, gradient descent refines the embedding coordinates in two phases. In the first one, called early exaggeration, attractive forces between points are artificially strengthened to delay any detrimental effect of repulsive forces while points are still poorly organized. In this paper, a novel initialization of t-SNE is proposed. It works by hierarchizing the data points into a space-partitioning binary tree and successive runs of t-SNE with 4, 8, 16, ..., N points. Between two runs, the prototypical point in each tree branch is split into its two children prototypes, with some little random noise, and the embedding is rescaled to account for the increased population. Experimental results show the effectiveness of the method. The proposed method is compatible with any method of neighbor embedding (t-SNE, UMAP, etc.) provided early exaggeration can be disabled and initial coordinates can be fed into.</div