HAL ENVT (Ecole Nationale Vétérinaire de Toulouse)
Not a member yet
16577 research outputs found
Sort by
L'holobionte, une nouvelle dimension pour la sélection porcine ?
No full textInternational audienceThe role of the microbiota, in particular the gut microbiota, on host performance is increasingly documented in farm animals. These advances have been made possible by the use of new high-throughput sequencing techniques and the development of data-analysis methods to characterize the microorganisms present (diversity, relative abundance, ecosystem assembly). They have provided new insights into the relations between host genetic diversity and host-microbiota interactions, as well as into the estimation of phenotypes or genetic values that consider microbiotic variability. These results fuel a new paradigm based on the definition of the holobiont as a biological system that assembles a host and its microbiota, thus providing a new vision of what an individual represents. We propose to assess the swine holobiont through modelling and experimental results to understand, predict and steer phenotypes to improve, for example, the health but also the productive and environmental efficiencies of pigs. These advances are also discussed as paving the way for developing selection strategies that simultaneously take consider information about the host and its microbiota, while reflecting on the sustainable improvement of populations at the scale of a holobiont.Le rôle du microbiote et en particulier du microbiote intestinal sur les performances de l’hôte est de plus en plus documenté chez les animaux d’élevage. Ces avancées ont été permises par l’utilisation des nouvelles techniques de séquençage à haut débit et le développement de méthodes d’analyse des données pour caractériser les microorganismes présents (diversité, abondances relatives, assemblage en écosystèmes). Elles ont apporté de nouvelles connaissances sur les liens entre la diversité génétique de l’hôte et les interactions hôte-microbiote, mais aussi sur l’estimation des phénotypes ou des valeurs génétiques en prenant en compte la variabilité du microbiote. Ces résultats alimentent un nouveau paradigme fondé sur la définition de l’holobionte en tant que système biologique qui assemble un hôte et son microbiote, offrant une nouvelle vision de ce que représente un individu. Nous proposons de faire le point sur l’holobionte porc, en termes de modélisation et de résultats expérimentaux pour comprendre, prédire et piloter les phénotypes dans le but d’améliorer par exemple la santé mais aussi l’efficience productive et environnementale des animaux. Sont également discutées ces avancées comme ouvrant la voie pour élaborer des stratégies de sélection prenant simultanément en compte les informations sur l’hôte et son microbiote en raisonnant l’amélioration durable des populations à l’échelle de l’holobionte
Optimiser génétique et nutrition pour répondre à la diversification des ressources alimentaires dans les systèmes de production porcins
No full textThe diversification of breeding systems and feeding strategies in response to the volatility of feed ingredient prices and the competition for resources foreshadows a wide set of resource profiles in the future. Feed production is a major contributor to the environmental impact of pig farms. The aim of this thesis was to individually characterize the pigs that achieve best economic and environmental performances in breeding systems with diversified feeding strategies, to propose selection strategies that support the sustainability of the pig industry. A multi-criteria assessment approach based on an individual-based model of the fattening unit was used to assess economic performances and environmental impacts of each pig using Life Cycle Assessment. To ensure that each pig fully expresses its impacts, an innovative fair evaluation procedure based on feed formulations according to individual nutritional requirements was implemented, with objectives of least cost, or joint minimization of feed costs and impacts. The performance of 732 Large White pigs representing the genetic variability of a commercial population was evaluated in a conventional reference system, an alternative conventional and soy-free system with low-impact co-products and innovative resources, a conventional local system promoting low-processed ingredients, and an organic system including forages. Analysis of the economic and environmental responses obtained in all systems identified three clusters of pigs, studied at phenotypic and genetic level. Efficient, lean pigs with moderate digestible amino acid requirements at the start of fattening had the best phenotypic and genetic potential for low environmental impacts, irrespective of the system. The least technically efficient pigs always had high environmental impacts. Efficient pigs with high requirements had contrasting climate change (CC), freshwater and marine eutrophication (EUfreshw, EUmar), fossil resource use (RU) and land use (LU) impacts across systems. Technical performances were the main determinants of the impacts when the variability of feed impacts was reduced (local and organic systems). In that case, impacts of all efficient pigs were low. The genetic variability of environmental impacts predicted at system level was estimated for the first time in pigs. Environmental impacts were globally heritable, with low heritabilities for some impact x system combinations (0.06 ± 0.11), and up to 0.54, indicating that a selection potential would exist for most situations. All environmental impacts were positively correlated with each other, and moderately and favourably correlated with technical traits included in the current selection objectives. Simulation of selection for a production index thus showed a favourable effect on the reduction of environmental impacts, in particular CC, acidification, terrestrial eutrophication, and RU, and could punctually degrade EUfreshw, EUmar and LU impacts. Correlations between genetic values between systems show strong rerankings for highly feed-dependent impacts, indicating genotype x system interactions. The transfer efficiency of genetic gain between systems is also affected by differences in the genetic variance of impacts between systems. These initial estimates need to be strengthened on a larger dataset, and supplemented by simulations of breeding schemes integrating environmental impacts to support changes in breeding objectives adapted to the diversity of systems.La diversification des systèmes d’élevage et des stratégies d’alimentation des porcs en réponse à la volatilité des prix des matières premières et à la compétition pour les ressources laisse anticiper des profils de ressources variés. La production d’aliments est un poste majeur des impacts environnementaux des élevages porcins. L’objectif de cette thèse était de caractériser individuellement les porcs qui réalisent les meilleures performances économiques et environnementales dans des systèmes d’élevage à stratégies alimentaires diversifiées, pour proposer des stratégies de sélection qui soutiennent la durabilité de la filière porcine. Une démarche d’évaluation multicritère s’appuyant sur un modèle individu-centré de l’atelier d’engraissement a été utilisée pour évaluer les performances économiques et les impacts environnementaux par Analyse de Cycle de Vie de chaque porc. Pour que chaque porc exprime complètement ses impacts, une procédure innovante d’évaluation équitable basée sur des formulations d’aliments en fonction des besoins nutritionnels individuels a été mise en œuvre, avec des objectifs de moindre coût, ou de minimisation conjointe des coûts et impacts des aliments. Les performances de 732 porcs Large White, représentatifs de la variabilité génétique d’une population commerciale ont été évaluées dans un système conventionnel de référence, un système conventionnel alternatif sans soja avec des coproduits et ressources innovantes à faibles impacts, un système local conventionnel privilégiant des matières premières peu transformées, et un système biologique incluant des fourrages. L’analyse des réponses économiques et environnementales obtenues dans l’ensemble des systèmes a discriminé trois groupes de porcs, étudiés au niveau phénotypique et génétique. Les porcs efficaces, maigres, avec des besoins en acides aminés digestibles modérés en début d’engraissement, avaient un meilleur potentiel phénotypique et génétique à avoir des impacts environnementaux faibles dans tous les systèmes. Les porcs les moins bons techniquement avaient toujours des impacts élevés. Les porcs efficaces mais à besoins élevés avaient des impacts changement climatique (CC), eutrophisations de l’eau douce et marine (EUfreshw, EUmar), utilisation de ressources fossiles (RU) et utilisation des terres (LU) contrastés selon les systèmes. Les performances techniques étaient le déterminant majeur des impacts quand la variabilité des impacts des aliments était réduite (systèmes local et biologique). Les impacts de tous les porcs efficaces étaient alors faibles. La variabilité génétique des impacts environnementaux prédite à l’échelle du système a été estimée pour la première fois chez le porc. Les impacts étaient globalement héritables, avec des valeurs faibles pour certaines combinaisons impacts-systèmes (0,06 ± 0.11), et jusqu’à 0,54, indiquant qu’un potentiel de sélection existerait pour la plupart des situations. Tous les impacts environnementaux étaient corrélés positivement entre eux, et modérément et favorablement aux caractères techniques inclus dans les objectifs de sélection actuels. La simulation de sélection pour un index de production montrait ainsi un effet favorable sur la réduction des impacts, en particulier CC, acidification, eutrophisation terrestre, et RU, et pouvait ponctuellement dégrader les impacts EUfreshw, EUmar et LU. Les corrélations entre les valeurs génétiques entre systèmes montrent des reclassements forts pour les impacts très dépendants des aliments, pointant des interactions génotype x système. L’efficacité de transfert du gain génétique entre systèmes est aussi affectée par des différences de variance génétique des impacts entre systèmes. Ces premières estimations devront être consolidées sur un plus large jeu de données, et complétées par des simulations de schémas de sélection intégrant les impacts environnementaux pour soutenir des changements d’objectifs de sélection adaptés à la diversité des systèmes
glmtools -an R Package to compute closed-form estimator for Generalized Linear Models with categorical explanatory variables
No full textThe glmtools package computes the closed-form estimator (CFE) introduced in Brouste et al. ( 2022) for generalized linear models (GLM) with categorical explanatory variables. Performances of our methodology are presented in several simulation studies for models with one, two or more categorical explanatory variables. The CFE implemented in the glmtools package are far less computationally intensive than MLE which uses classical iteratively re-weighted least square algorithm. Models with two categorical explanatory variables are discussed more deeply in the GLM with and without interaction for which the closed-form estimators are the MLE or not
Investigating the role of urban vegetation alongside other environmental variables in shaping Aedes albopictus presence and abundance in Montpellier, France
No full textInternational audienceUrban greening helps address urbanization challenges and provides health benefits, but it may also favor mosquito species, vectors of pathogens causing human diseases. This study examines the relationship between urban vegetation and the presence and abundance of Aedes albopictus in Montpellier, the second greenest city of France, while accounting for meteorology, microclimate, air quality, human socio-demography, and landscape. From May to October 2023, adult mosquitoes were collected monthly in urban parks, residential areas, and the impervious city center using BG-Pro traps with odor and CO? attractants. Microclimate data were recorded on-site. Vegetation, land cover, meteorological, air quality, and human demographic data were gathered from open-access databases. Ae. albopictus presence and abundance were analysed according to environmental variables taken at different time lags and spatial distances using a two-stage modeling approach: bivariate analyses using generalized linear mixed models were conducted to select variables for inclusion in a multivariate random forest model, aiming to identify the factors that best explain Ae. albopictus presence and abundance. While urban vegetation had a limited effect on Ae. albopictus presence, number of patches and the percentage of area covered by low vegetation were among the most important predictors of abundance. The main predictors for presence were minimum hourly temperature (24h-48h before sampling), minimum atmospheric pressure during sampling, and the weekly cumulated rainfall recorded six weeks before sampling. The most important predictors of abundance were maximum hourly temperatures during sampling, cumulated growing degree days (GDD) during two weeks preceding sampling and the length and patch size of roads. To our knowledge, this is the first study examining urban vegetation’s influence on Ae. albopictus in France, offering insights for urban planning and suggesting further research on vegetation's role in mosquito-borne disease transmission, particularly in the context of increasing dengue incidence in Europe
Développement et industrialisation d'un outil de génotypage adapté à la filière porcine et fondé sur l’utilisation des données issues duséquençage du génome
No full textProjet de Recherche et DéveloppementContexte -Le marché du génotypage chez les porcs est aujourd'hui dépendant de la technologie des puces à SNP. Cette technologie qui a permis une amélioration significative de la précision et de l'efficacité de la sélection porcine, est cependant fondée sur une approche figée et incomplète, dépendant d'un faible nombre de fournisseurs étrangers. Il est donc essentiel de sortir de cette dépendance de marché en développant un nouvel outil répondant aux attentes des acteurs de la filière porcine française. De plus, à terme, cette technologie pourrait être adaptée à d'autres filières animales, avec en premier lieu la filière bovine.</div
High-resolution mass spectrometry-based untargeted approach: A cutting-edge analytical technique for the detection and identification of organophosphorus synthesis sources
No full textInternational audienc
Influences de multiples composantes des aptitudes maternelles de la truie sur la survie et la croissance précoces de la portée
No full textInternational audienceTo define which components of maternal ability contribute most to piglet production during early lactation, we maximised the observable variability of traits by applying a crossbreeding scheme to 21 Large White sows and 22 Meishan sows reared in individual pens in the farrowing unit. We measured 100 maternal traits and grouped them into 11 blocks according to the biological component they described. Sows' reactivity and peripartum activity were determined using on-farm notations and image analysis, respectively. The functional and behavioural explanatory variables for the proportion of dead piglets relative to live born and litter weight gain, which were calculated for three periods after birth (D0-D1, D1-D3 and D3-D7), were adjusted for the effects of breed and litter size. Partial least-square multi-block analyses were applied separately to mortality traits and growth traits. Several behavioural traits (standing activity, reactivity to several stimuli, postural activity) and functional traits (body reserves, teat characteristics) on the day of farrowing (D0) contributed substantially to within-litter D0-D7 growth and survival. Sow aggressiveness towards piglets on D0 and the time spent standing were negatively correlated with D1-D3 growth. During D3-D7, growth was positively correlated with feed intake. Aggressiveness towards piglets on D0 decreased survival to D7. During D1-D7, time spent lying laterally with the udder exposed and feed intake on D0 increased survival.Influence de multiples composantes des aptitudes maternelles sur la survie et la croissance précoces de la portéePour définir quelles composantes des aptitudes maternelles contribuent le plus à la production de porcelets en début de lactation, nous avons maximisé la variabilité observable sur les différents caractères par un dispositif de croisements appliqué à 21 truies Large White et 22 truies Meishan, libres en case de maternité. Nous avons mesuré 100 caractères maternels et les avons regroupés en 11 blocs selon la composante biologique qu'ils décrivent. La réactivité des truies et leur activité peripartum ont été obtenus avec des notations en élevage et de l'analyse d'image, respectivement. Les variables fonctionnelles et comportementales explicatives de la proportion de morts par rapport aux nés vivants et du gain de poids de portée, calculés sur trois périodes après la naissance (J0-J1, J1-J3 et J3-J7) ont été ajustées pour les effets de la race et de la taille de portée. Des analyses multiblocs selon la méthode des moindres carrés partiels ont été appliquées séparément aux caractères de mortalité et aux caractères de croissance. Plusieurs caractères comportementaux (activité debout, réactivité à plusieurs stimuli, activité posturale) et fonctionnels (réserves corporelles, caractéristiques morphologiques des tétines) le jour de la mise bas (J0) influençaient de façon marquée la croissance et la survie intra-portée de J0 à J7. A J0 l'agressivité de la truie envers les porcelets et le temps passé debout étaient corrélés défavorablement à la croissance de J1 à J3. La croissance de J3 à J7 était positivement corrélée à l'ingestion d'aliments sur la même période. L'agressivité envers les porcelets à J0 avait un effet défavorable sur la survie jusqu'à J7. Le temps passé coucher latéral avec exposition de la mamelle et l'ingestion d'aliment à J0 favorisaient la survie de J1 à J7. Influence of multiple maternal-ability components on early litter survival and growthTo define which components of maternal ability contribute most to piglet production during early lactation, we maximised the observable variability of traits by applying a crossbreeding scheme to 21 Large White sows and 22 Meishan sows reared in individual pens in the farrowing unit. We measured 100 maternal traits and grouped them into 11 blocks according to the biological component they described. Sows' reactivity and peripartum activity were determined using on-farm notations and image analysis, respectively. The functional and behavioural explanatory variables for the proportion of dead piglets relative to live born and litter weight gain, which were calculated for three periods after birth (D0-D1, D1-D3 and D3-D7), were adjusted for the effects of breed and litter size. Partial least-square multi-block analyses were applied separately to mortality traits and growth traits. Several behavioural traits (standing activity, reactivity to several stimuli, postural activity) and functional traits (body reserves, teat characteristics) on the day of farrowing (D0) contributed substantially to within-litter D0-D7 growth and survival. Sow aggressiveness towards piglets on D0 and the time spent standing were negatively correlated with D1-D3 growth. During D3-D7, growth was positively correlated with feed intake. Aggressiveness towards piglets on D0 decreased survival to D7. During D1-D7, time spent lying laterally with the udder exposed and feed intake on D0 increased survival.</div
RITHMS : An advanced stochastic framework for the simulation of transgenerational hologenomic data
No full textInternational audienceR Implementation of a Transgenerational Hologenomic Model-based Simulator (RITHMS) is a framework for simulating transgenerational hologenomic data that relies on MoBPS features for genomic data, accounts for the particularities of microbiota transmission, uses real genomic and microbiota data to construct a base population, and is flexible enough to cover a variety of scenarios defined by heritability, microbiability, and microbiota heritability
Genetic and environmental determinants of Immunoglobulin G in kid serum and adult colostrum of dairy goats
No full textInternational audienceImmune competence is a key factor in the health and resilience of dairy ruminants. The aim of this study was to study immunoglobulin G concentration in serum shortly after birth and in colostrum of lactating animals as proxy of passive immune transfer and colostrum quality, respectively, and thus provide new health and resilience indicators for selection.Data from 424 female kids and 203 primiparous goats, reared on an INRAE experimental farm, were analyzed. IgG concentration was measured in kid serum and goat colostrum, respectively, using the radial immunodiffusion technique. Linear models were used to identify factors of variation in serum and colostrum IgG concentration, and health events were monitored for 5 weeks after birth to measure the impact of passive immune transfer on kids' health. We found a significant effect of serum IgG level on hindquarters cleanliness score (as a proxy of digestive troubles) and nasal and ocular discharge. As expected, the absorption of IgG in the days following birth is essential to ensure the passive transfer of immunity from the mother to the youngster, and provide immune protection. Genetic parameters were estimated for serum IgG concentration in kids and for colostrum characteristics in primiparous goats. Heritability of serum IgG in kids was not significantly different from zero. This low value, with large SE, may be due to the limited sample size and large environmental variability likely not accounted for. Especially, we could not adjust for the amount of IgG that was ingested by the kid, which can have a significant effect on serum IgG and possibly on heritability estimates. Heritability of colostrum traits were moderate, with estimates of 0.25 (±0.16), 0.35 (±0.16), and 0.25 (±0.16) for IgG concentration, IgG mass and colostrum volume, respectively. Genetic correlations with milk production showed high standard errors due to small numbers. However, estimates suggest a negative genetic association between production and IgG concentration in colostrum (rg = -0.44 ± 0.33), probably linked to a positive association with total colostrum volume. The significant heritabilities that were estimated for the first time demonstrate the existence of genetic control of colostrum quality in goats, and open up opportunities for selection in this species.</p
