Eucalyptus grandis é uma das espécies mais plantadas no Brasil apresentando excelente crescimento e forma do fuste, contudo é susceptível ao fungo Puccinia psidii Winter causador da \"ferrugem do eucalipto\" e apresenta defeitos nas toras, como a presença de nós, que dificulta a utilização da espécie para a produção de madeira serrada. O estudo teve como objetivo definir uma estratégia de melhoramento para disponibilizar material genético que apresente simultaneamente alto potencial de crescimento inicial, menor grau de infecção à ferrugem e melhor qualidade da madeira nessa primeira etapa do programa de melhoramento. Foram instalados dois experimentos nas Estações Experimentais de Ciências Florestais da ESALQ/USP nos municípios de Itatinga-SP e Anhembi-SP, com 176 progênies de polinização aberta de E. grandis. Os testes foram em blocos ao acaso, com 29 repetições em Anhembi-SP e 30 repetições em Itatinga-SP, com uma planta por parcela e uma linha de bordadura externa. Além das progênies de E. grandis, foi utilizado três matérias genéticos comerciais como testemunhas para o ranking das melhores famílias. Foram realizados três inventários. A avaliação do crescimento foi feita em todos os inventários através da sobrevivência e mensuração do DAP e altura. As avaliações para grau de infecção à ferrugem foram feitas no primeiro e terceiro inventário através de escala de notas. As avaliações indiretas de qualidade da madeira foram feitas no segundo inventário com a contagem do número de galhos e mensuração do diâmetro dos galhos no teste de progênies instalado em Anhembi-SP, e no terceiro inventário com a avaliação da bifurcação e tortuosidade para os dois locais. Os parâmetros genéticos foram estimados utilizando o software SELEGEN. Com base nos resultados foi definida uma estratégia de melhoramento visando otimizar os ganhos genéticos para crescimento inicial, menor grau de infecção a ferrugem e indiretamente para qualidade da madeira utilizando métodos de seleção simultâneos como seleção indireta, níveis independentes e índices combinados com pesos iguais e diferentes para cada variável. Na estratégia foi definida que as duas localidades representam uma zona de melhoramento, sendo escolhido o teste de progênies instalado em Anhembi-SP para o cálculo dos ganhos genéticos. O melhoramento genético foi baseado na formação de um Pomar de Sementes por Mudas (PSM) e de um Pomar Clonal de Sementes (PCS). Ambos os pomares atenderão a duas regiões para a produção de sementes melhoradas. Como as correlações genéticas entre o DAP, variável do crescimento, grau de infecção à ferrugem, e número de galhos, variável indireta para qualidade da madeira, foram fracas, foi necessário considerar as três variáveis para a construção dos índices combinados de seleção e níveis independentes. Para as duas estratégias de seleção, os níveis independentes de seleção são o que melhor otimizam a seleção simultânea para crescimento, resistência à ferrugem e qualidade da madeira. / Eucalyptus grandis is one of the most planted species in Brazil presenting excellent growth and straight, however it is susceptible to the fungus Puccinia psidii Winter that causes \"eucalyptus rust\", and presents defects in the logs, such as the presence of knots, which difficult the use of the species for the production of sawn wood. The aim of the study was to define a breeding strategy to provide genetic material that exhibits high initial growth, less degree infection to rust and better wood quality simultaneously, in the first stage of the program. Two experimental tests were planted at Experimental Stations of Forest Sciences of ESALQ/USP in Itatinga-SP and Anhembi-SP, with 176 open pollinated progenies of E. grandis. A randomized blocks design, with 29 blocks in Anhembi-SP and 30 blocks in Itatinga-SP, with one plant per plot and one external border line. In addition to the progenies of E. grandis, three genetic material was used as witnesses to the ranking of the best families. Three inventories were carried out. Growth assessment was done on all inventories through the survival and measurement of DBH and height. The evaluations for degree of infection to the rust were made in the first and third evaluations through scale of notes. The indirect wood quality evaluations were done in the second inventory with the counting of the number of branches and measurement of the diameter of the branches in the progenies test installed in Anhembi-SP, and in the third inventory with the evaluation of the bifurcation and straightness for the two sites. The genetic parameters were estimated using the SELEGEN software. Based on the results, a breeding strategy was defined aiming at optimizing the genetic gain for growth, lower degree infection to rust and indirectly to wood quality using simultaneous selection methods such as indirect selection, independent levels and selection index with equal and different weights for each variable. It was chosen the progenies test installed in Anhembi-SP to calculate the genetic gains, because the two sites represent only one breeding zone. The improvement was based on the formation of a Seed Orchard by Seedlings and a Clonal Orchard of Seeds. The orchards will serve for production of improved seed for both sites. As genetic correlations between DBH, variable for growth, degree of infection to rust, and number of branches, indirect variable for wood quality, were weak, it was necessary to consider the three variables for the selection with selection indexes and independent levels. For both selection strategies, independent levels are the method that best optimize the simultaneous selection for growth, rust resistance and wood quality.

Heavily-selected livestock production traits rarely come without compromise; altered physiology arising from intensive selection often gives rise to concern of a welfare trade-off. A particularly clear example of welfare challenge caused by genetic selection in chickens is the ‘broiler-breeder paradox’, wherein breeding populations of broiler-type birds selected for fast growth are feed-restricted in order to reduce growth and maintain reproductive viability at sexual maturity. In order to better-inform management and breeding strategies for alleviating reproductive problems resulting from genetic selection for growth, it is essential to develop a better understanding of the physiological processes underpinning growth. Whereas the molecular mechanisms governing energy balance in mammals have been relatively welldescribed, analogous avian systems have not received as much research attention and remain somewhat poorly understood. The broad aim of this doctoral project was to contribute to understanding of avian energy balance, particularly in the context of selection for high growth. Using an advanced broiler-layer intercross chicken line (AIL), high- and low-growth haplotypes at the locus encoding the cholecystokinin A receptor (CCKAR), underlying the most significant QTL for growth in chickens, were characterised. Of over 300 variations detected, a select panel spaced across the CCKAR locus were tested for prediction of bodyweight in a diverse cohort of chicken populations. One intronic SNP was found to be significant (p < 0.05) and proximal to transcription factor binding sites. The effect of this locus on gross bodyweight remained significant into the 20th AIL generation (~20% at 10wk, p < 0.05). In this otherwise effectively genetically homogeneous population, several specific physiological traits were predicted by CCKAR haplotype alone, yielding some clues as to the significance of perturbed cholecystokinin (CCK) signalling in broiler strains. While birds with high-growth CCKAR haplotype (HG) did not appear to consume more, feed conversion efficiency (FCE) was improved, at least for males, compared to low-growth (LG) (p < 0.05). Visceral organ anatomies were morphologically disparate, with HG individuals exhibiting ~1/3 less gallbladder mass (p < 0.01), and ~10% shorter GI tract (p < 0.01) and metatarsal bone (p < 0.05). Further gaps in knowledge of the expression of peripheral satiety hormones in chicken are addressed in this thesis. Tissue distributions for expression of CCK, gastrin, pancreatic polypeptide (PPY) and peptide YY (PYY), were mapped and their respective dynamic responses to nutritive state examined. CCK was found to be most highly expressed in the brain, whereas PYY, PPY and gastrin were far more abundant in distinct regions of the periphery. Interestingly, peripheral CCK was not responsive to short-term (< 10h) satiety in experimental populations where PYY and gastrin were. PYY expression was found to be greatest in the pancreas and consistently upregulated within hours after feeding (p < 0.01), whereas gastrin expression was confined to the gastric antrum and paradoxically highest in fasting birds (p < 0.01). PPY expression is strictly limited to the pancreas and appears dependent on longerterm energy state. These results highlight similarities and differences to mammalian systems; notably, the avian pancreas seems to fulfil an exceptional role as a site of signal integration, perhaps unsurprising considering its disproportionate size compared to mammals. Indeed, pancreatic PYY appears to act as a primary peripheral short-term satiety hormone in birds. This body of work contributes to the understanding of avian energy balance and growth. An invaluable foundation for future research is formed by the identification of the major locations of production, and basic nutrient-responsive trends, for several peripheral avian hormones. Information on the growth role of CCKAR is consolidated and expanded upon, demonstrating a clear genetic contribution to maintenance organ morphology and overall growth. Such knowledge can be used to reliably assess and advise on selection and management of chickens to stem welfare concerns without compromising production. Comparisons between avian and other vertebrate endocrine systems make for interesting insight into the adaptive role of energy homeostatic mechanisms in divergent evolution of mammals and non-mammalian vertebrates. In some aspects, birds might better represent the ancestral phenotype from which each vertebrate clade arose.

Eucalyptus grandis é uma das espécies mais plantadas no Brasil apresentando excelente crescimento e forma do fuste, contudo é susceptível ao fungo Puccinia psidii Winter causador da \"ferrugem do eucalipto\" e apresenta defeitos nas toras, como a presença de nós, que dificulta a utilização da espécie para a produção de madeira serrada. O estudo teve como objetivo definir uma estratégia de melhoramento para disponibilizar material genético que apresente simultaneamente alto potencial de crescimento inicial, menor grau de infecção à ferrugem e melhor qualidade da madeira nessa primeira etapa do programa de melhoramento. Foram instalados dois experimentos nas Estações Experimentais de Ciências Florestais da ESALQ/USP nos municípios de Itatinga-SP e Anhembi-SP, com 176 progênies de polinização aberta de E. grandis. Os testes foram em blocos ao acaso, com 29 repetições em Anhembi-SP e 30 repetições em Itatinga-SP, com uma planta por parcela e uma linha de bordadura externa. Além das progênies de E. grandis, foi utilizado três matérias genéticos comerciais como testemunhas para o ranking das melhores famílias. Foram realizados três inventários. A avaliação do crescimento foi feita em todos os inventários através da sobrevivência e mensuração do DAP e altura. As avaliações para grau de infecção à ferrugem foram feitas no primeiro e terceiro inventário através de escala de notas. As avaliações indiretas de qualidade da madeira foram feitas no segundo inventário com a contagem do número de galhos e mensuração do diâmetro dos galhos no teste de progênies instalado em Anhembi-SP, e no terceiro inventário com a avaliação da bifurcação e tortuosidade para os dois locais. Os parâmetros genéticos foram estimados utilizando o software SELEGEN. Com base nos resultados foi definida uma estratégia de melhoramento visando otimizar os ganhos genéticos para crescimento inicial, menor grau de infecção a ferrugem e indiretamente para qualidade da madeira utilizando métodos de seleção simultâneos como seleção indireta, níveis independentes e índices combinados com pesos iguais e diferentes para cada variável. Na estratégia foi definida que as duas localidades representam uma zona de melhoramento, sendo escolhido o teste de progênies instalado em Anhembi-SP para o cálculo dos ganhos genéticos. O melhoramento genético foi baseado na formação de um Pomar de Sementes por Mudas (PSM) e de um Pomar Clonal de Sementes (PCS). Ambos os pomares atenderão a duas regiões para a produção de sementes melhoradas. Como as correlações genéticas entre o DAP, variável do crescimento, grau de infecção à ferrugem, e número de galhos, variável indireta para qualidade da madeira, foram fracas, foi necessário considerar as três variáveis para a construção dos índices combinados de seleção e níveis independentes. Para as duas estratégias de seleção, os níveis independentes de seleção são o que melhor otimizam a seleção simultânea para crescimento, resistência à ferrugem e qualidade da madeira. / Eucalyptus grandis is one of the most planted species in Brazil presenting excellent growth and straight, however it is susceptible to the fungus Puccinia psidii Winter that causes \"eucalyptus rust\", and presents defects in the logs, such as the presence of knots, which difficult the use of the species for the production of sawn wood. The aim of the study was to define a breeding strategy to provide genetic material that exhibits high initial growth, less degree infection to rust and better wood quality simultaneously, in the first stage of the program. Two experimental tests were planted at Experimental Stations of Forest Sciences of ESALQ/USP in Itatinga-SP and Anhembi-SP, with 176 open pollinated progenies of E. grandis. A randomized blocks design, with 29 blocks in Anhembi-SP and 30 blocks in Itatinga-SP, with one plant per plot and one external border line. In addition to the progenies of E. grandis, three genetic material was used as witnesses to the ranking of the best families. Three inventories were carried out. Growth assessment was done on all inventories through the survival and measurement of DBH and height. The evaluations for degree of infection to the rust were made in the first and third evaluations through scale of notes. The indirect wood quality evaluations were done in the second inventory with the counting of the number of branches and measurement of the diameter of the branches in the progenies test installed in Anhembi-SP, and in the third inventory with the evaluation of the bifurcation and straightness for the two sites. The genetic parameters were estimated using the SELEGEN software. Based on the results, a breeding strategy was defined aiming at optimizing the genetic gain for growth, lower degree infection to rust and indirectly to wood quality using simultaneous selection methods such as indirect selection, independent levels and selection index with equal and different weights for each variable. It was chosen the progenies test installed in Anhembi-SP to calculate the genetic gains, because the two sites represent only one breeding zone. The improvement was based on the formation of a Seed Orchard by Seedlings and a Clonal Orchard of Seeds. The orchards will serve for production of improved seed for both sites. As genetic correlations between DBH, variable for growth, degree of infection to rust, and number of branches, indirect variable for wood quality, were weak, it was necessary to consider the three variables for the selection with selection indexes and independent levels. For both selection strategies, independent levels are the method that best optimize the simultaneous selection for growth, rust resistance and wood quality.

Ce travail est fondé sur l’analyse de 28 000 restes osseux, provenant de deux sites inédits, Archontiko Giannitson et Angelochori Imathias (Macédoine, Grèce) et les résultats obtenus intégrés à une synthèse de la documentation publiée à l’échelle de l’ensemble de la Grèce septentrionale sur toute la période de l’âge du Bronze. Les proportions entre les espèces exploitées indiquent que l’économie vivrière était largement centrée sur le petit bétail avec toutefois un taux d’espèces sauvages relativement important. Les activités pastorales étaient dominées par l’exploitation des caprinés et du porc et mettent en évidence la recherche d’une rentabilité maximale dans la gestion des cheptels ainsi qu’une généralisation de l’utilisation des produits secondaires : lait, laine, force de travail etc. Les pratiques de chasse étaient très variées comme en témoigne la diversité des taxons déterminés. La capture d’oiseaux, de poissons, de mollusques, de grands et de petits mammifères démontre l’intérêt des populations pour la faune sauvage. En diachronie, une nette évolution est observable à travers une diversification des spectres fauniques et des productions. Au Bronze Récent, l’économie animale se caractérise par une nouvelle augmentation de la diversité et par une concentration des activités pastorales sur le petit bétail. Enfin, ces adaptations aux nouvelles demandes des populations ont également conduit à l’apparition d’innovations techniques comme le charriot et la charrue qui coïncident avec l’arrivée des premiers équidés domestiques dans le monde égéen. / This work is based on the analysis of 28 000 bones from two sites : Archontiko Giannitson and Angelochori Imathias (Makedonia, Greece). The results have been included to a synthesis of the documentation of the Bronze Age in Northern Greece. Species proportions indicate that the supplying was largely focused on small animals with a relatively high rate of wild fauna. Pastoral activities were dominated by the use of sheep/goat and pig and highlights the search for maximum efficiency in the management of livestock as well as an increased use of « secondary products » : milk, wool, labor etc. Hunting practices were varied as shown by the diversity of taxa identified. The capture of birds, fishes, shellfishes, large and small mammals demonstrate the interest of societies for these faunas. A clear trend is seen in diachrony through a diversification of wild fauna’s spectra and productions. In the Late Bronze Age, the animal economy is characterized by a further increase in the diversification and a concentration of pastoral activities on small cattle. Finally, the adaptation to new demands have also led to the development of technical innovations such as the plow and the cart, coinciding with the appearance of domestic horses in the Aegean world.

Two closely-related, different-sized species of geese nest sympatrically south of the Queen Maud Gulf (QMG) in Canadas central arctic. Following a period of high population growth rate in both species within the QMG, the population growth rate of larger-bodied lesser snow geese (Chen caerulescens caerulescens; hereafter snow geese) has slowed most recently to roughly half that observed in smaller-bodied Rosss geese (Chen rossii). I focused on factors that influence productivity and recruitment in these two species, to improve our understanding of life history variation associated with interspecific differences in body size, and to test for density-dependent population responses. I used long-term data (1991 to 2008) to compare spring nutrient reserves, breeding strategies, clutch sizes, nest success, and juvenile survival in Rosss and snow geese breeding at Karrak Lake, Nunavut; a large breeding colony located within the QMG.<p> Long-term patterns of spring body condition (i.e., fat and protein reserves) diverged in prospective breeding female Rosss and snow geese implying that differences in food acquisition ability had become more acute. Snow geese displayed larger reductions in protein and fat reserves through time compared to Rosss geese thereby suggesting a differential density-dependent response in the ability to store nutrient reserves, a prerequisite for breeding in both species. Decreased per capita food availability influenced the timing of reproduction in both species. Nesting phenologies of Rosss and snow geese, adjusted for variation in phenology of local spring climate, have become later by 6.5 and 5.0 days, respectively, since 1991. Nutritional strategies (i.e., reliance on reserves versus local food) used for clutch formation differed between species. Rosss geese displayed greater reliance on stored reserves (i.e., capital breeding) than did snow geese, though both used endogenous reserves (> 62% of yolk protein, > 48% of albumen, and > 73% of yolk lipid) for clutch formation. Rosss and snow geese experienced declines of 28% and 23% in body masses from arrival to post-laying and also until hatch demonstrating that endogenous reserves are the main nutrient sources for incubation. Still, constraints of small size forced Rosss geese to use a mixture of local food plants and reserves for incubation metabolism.<p> I then examined differences in clutch size, nest success, and juvenile survival to understand of the role of recruitment in the interspecific divergence of population trajectories. I did not find strong interspecific differences in clutch size and nest success. Overall, snow geese had a larger mean clutch size, which was expected based on benefits of a larger-body size. Clutch sizes decreased with delays in breeding and decreasing protein reserves of arriving females, although Rosss geese displayed larger declines with decreasing protein reserves. Mean apparent nest success for Rosss geese was 4.5% higher compared to snow geese. Nest success showed large declines (11%) in both species with increasing population size at the breeding colony. However, nest success of snow geese decreased twice as fast with delays in breeding compared to Rosss geese. Last, I found no evidence of negative density dependence in juvenile survival over time. Juvenile survival was higher in snow geese (48%) compared to Rosss geese (38%), consistent with a life history prediction based on body size differences. Despite lower juvenile survival, recruitment by Rosss geese is likely greater than that of snow geese because of earlier sexual maturity, higher breeding probability and/or greater nest success.<p> Ultimately, small body size of Rosss geese may produce an ideal life history schedule under resource limitation at this colony i.e., one that maximizes fitness compared to larger snow geese. Life history characteristics of Rosss geese (e.g., absolutely lower energy requirement, have a flexible breeding strategy, higher reproductive effort, an earlier age of sexual maturity, a shorter breeding cycle allowing delayed arrival and nest initiation on arctic breeding areas, and shorter time required by goslings to attain adult size), in addition to their smaller bill morphology may allow exploitation of a wider niche space (i.e., one that includes marginal quality and low quantity vegetation) relative to snow geese. Because there were no large differences in components of recruitment considered here, other components of recruitment (age of sexual maturity, breeding probability) may be affected more strongly by diminished spring nutrition in snow geese and thus have a larger influence on local population dynamics.

Two closely-related, different-sized species of geese nest sympatrically south of the Queen Maud Gulf (QMG) in Canadas central arctic. Following a period of high population growth rate in both species within the QMG, the population growth rate of larger-bodied lesser snow geese (Chen caerulescens caerulescens; hereafter snow geese) has slowed most recently to roughly half that observed in smaller-bodied Rosss geese (Chen rossii). I focused on factors that influence productivity and recruitment in these two species, to improve our understanding of life history variation associated with interspecific differences in body size, and to test for density-dependent population responses. I used long-term data (1991 to 2008) to compare spring nutrient reserves, breeding strategies, clutch sizes, nest success, and juvenile survival in Rosss and snow geese breeding at Karrak Lake, Nunavut; a large breeding colony located within the QMG.<p> Long-term patterns of spring body condition (i.e., fat and protein reserves) diverged in prospective breeding female Rosss and snow geese implying that differences in food acquisition ability had become more acute. Snow geese displayed larger reductions in protein and fat reserves through time compared to Rosss geese thereby suggesting a differential density-dependent response in the ability to store nutrient reserves, a prerequisite for breeding in both species. Decreased per capita food availability influenced the timing of reproduction in both species. Nesting phenologies of Rosss and snow geese, adjusted for variation in phenology of local spring climate, have become later by 6.5 and 5.0 days, respectively, since 1991. Nutritional strategies (i.e., reliance on reserves versus local food) used for clutch formation differed between species. Rosss geese displayed greater reliance on stored reserves (i.e., capital breeding) than did snow geese, though both used endogenous reserves (> 62% of yolk protein, > 48% of albumen, and > 73% of yolk lipid) for clutch formation. Rosss and snow geese experienced declines of 28% and 23% in body masses from arrival to post-laying and also until hatch demonstrating that endogenous reserves are the main nutrient sources for incubation. Still, constraints of small size forced Rosss geese to use a mixture of local food plants and reserves for incubation metabolism.<p> I then examined differences in clutch size, nest success, and juvenile survival to understand of the role of recruitment in the interspecific divergence of population trajectories. I did not find strong interspecific differences in clutch size and nest success. Overall, snow geese had a larger mean clutch size, which was expected based on benefits of a larger-body size. Clutch sizes decreased with delays in breeding and decreasing protein reserves of arriving females, although Rosss geese displayed larger declines with decreasing protein reserves. Mean apparent nest success for Rosss geese was 4.5% higher compared to snow geese. Nest success showed large declines (11%) in both species with increasing population size at the breeding colony. However, nest success of snow geese decreased twice as fast with delays in breeding compared to Rosss geese. Last, I found no evidence of negative density dependence in juvenile survival over time. Juvenile survival was higher in snow geese (48%) compared to Rosss geese (38%), consistent with a life history prediction based on body size differences. Despite lower juvenile survival, recruitment by Rosss geese is likely greater than that of snow geese because of earlier sexual maturity, higher breeding probability and/or greater nest success.<p> Ultimately, small body size of Rosss geese may produce an ideal life history schedule under resource limitation at this colony i.e., one that maximizes fitness compared to larger snow geese. Life history characteristics of Rosss geese (e.g., absolutely lower energy requirement, have a flexible breeding strategy, higher reproductive effort, an earlier age of sexual maturity, a shorter breeding cycle allowing delayed arrival and nest initiation on arctic breeding areas, and shorter time required by goslings to attain adult size), in addition to their smaller bill morphology may allow exploitation of a wider niche space (i.e., one that includes marginal quality and low quantity vegetation) relative to snow geese. Because there were no large differences in components of recruitment considered here, other components of recruitment (age of sexual maturity, breeding probability) may be affected more strongly by diminished spring nutrition in snow geese and thus have a larger influence on local population dynamics.

No description available.

Le pin maritime (Pinus pinaster Ait.) est l’une des principales espèces forestières en France, fournissant près d’un quart de la production nationale de bois. Un programme d’amélioration, mis en place dans les années 1960, propose des variétés génétiquement améliorées pour la croissance et la rectitude du tronc.Cette thèse explore la possibilité d’introduire les marqueurs moléculaires dans les stratégies d’amélioration génétique du pin maritime en Aquitaine. Les marqueurs sont utilisés afin de reconstituer a posteriori les pedigrees au sein d’un test de descendance « polycross », pour d’une part vérifier les hypothèses sur lesquelles repose la sélection backward, et d’autre part, pour proposer une stratégie de sélection innovante. Tout d’abord, la reconstitution du pedigree de 984 individus à l’aide de 63 marqueurs SNPs permet de valider les hypothèses de la sélection backward, et montre que les estimations des paramètres génétiques et des valeurs génétiques maternelles, basées sur l’information d’un pedigree partiel ou complet, diffèrent peu. Puis, les meilleurs descendants du test polycross sont présélectionnés et génotypés pour évaluer la faisabilité d’une stratégie de sélection forward. Enfin, des vergers à graines sont simulés selon différentes stratégies de sélection (backward, forward, mixte) afin de comparer les gains génétiques des variétés améliorées ainsi obtenues.Une stratégie de sélection forward chez le pin maritime permettrait d’accélérer les cycles de sélection et d’augmenter la fréquence des sorties variétales. De plus, le jeu de marqueurs SNPs développé dans cette étude est en cours de valorisation dans différentes étapes du programme d’amélioration. / Maritime pine (Pinus pinaster Ait.) is one of the main economical forest species in France, providing about twenty five percent of the national round wood production. A breeding program, implemented since the 60’s, offers genetically improved varieties for growth and stem straightness.This PhD explores the use of molecular markers in breeding strategies for maritime pine in Aquitaine. Molecular markers were used for pedigree recovery in a polycross progeny trial to test assumptions of backward selection on one hand, and to evaluate the feasibility of a new breeding strategy on the other hand. First, the pedigree of 984 progeny was recovered with 63 SNPs allowing to verify the assumptions of backward selection. We also showed that genetic parameters and maternal breeding value estimates were not much modified by inclusion of full pedigree information. Then, the best progenies in the polycross trial were preselected and genotyped to investigate the possibility of carrying out a forward selection strategy. Finally, establishment of clonal seed orchards were simulated from various breeding strategies (backward, forward, mixed) in order to compare genetic gains from the improved varieties obtained thereby.This study opens new perspectives towards an implementation of forward selection in the French maritime pine breeding program, to speed the selection cycles up and to increase the frequency of variety renewal. Moreover, the set of SNP markers developed is now used in different steps of the breeding program.

Die Arbeit beschreibt ein stochastisches Simulationsprogramm zur Modellierung von Honigbienenpopulationen unter Zuchtbedingungen. Dieses Programm wurde neu implementiert um unterschiedliche Selektionsstrategien zu evaluieren und zu optimieren. In einer ersten Studie wurde untersucht, inwiefern die Vorhersagen, die das Programm trifft, vom verwendeten genetischen Modell abhängt. Hierbei wurde festgestellt, dass das Finite-Locus-Modell dem Infinitesimalmodell in Langzeitstudien vorzuziehen ist. Eine zweite Studie beleuchtete die Bedeutung der sicheren Anpaarung auf Belegstellen für die Honigbienenzucht. Hier zeigten die Simulationen, dass die Zucht mit Anpaarungskontrolle derjenigen mit freier Paarung von Königinnen deutlich überlegen ist. Schließlich wurde in einer finalen Studie der Frage nachgegangen, wie erfolgreiche Zuchtprogramme bei der Honigbiene langfristig nachhaltig zu gestalten sind. Hierbei sind kurzfristiger genetischer Zugewinn und langfristige Inzuchtvermeidung gegeneinander abzuwägen. Durch umfangreiche Simulationen konnten für verschiedene Ausgangspopulationen Empfehlungen für eine optimale Zuchtintensität auf mütterlicher und väterlicher Seite gefunden werden. / The present work describes a stochastic simulation program for modelling honeybee populations under breeding conditions. The program was newly implemented to investigate and optimize different selection strategies. A first study evaluated in how far the program's predictions depend on the underlying genetic model. It was found that the finite locus model rather than the infinitesimal model should be used for long-term investigations. A second study shed light into the importance of controlled mating for honeybee breeding. It was found that breeding schemes with controlled mating are far superior to free-mating alternatives. Ultimately, a final study examined how successful breeding strategies can be designed so that they are sustainable in the long term. For this, short-term genetic progress has to be weighed against the avoidance of inbreeding in the long run. By extensive simulations, optimal selection intensities on the maternal and paternal paths could be determined for different sets of population parameters.

Cette étude archéozoologique porte sur les variations morphologiques de différents éléments du squelette porcin (Sus scrofa domesticus) et s’appuie sur deux présupposés majeurs : (1) les changements morphologiques de l’animal sont le reflet des évolutions du contexte socio-économique ; (2) il existe une grande diversité des formes porcines, conditionnée par différents facteurs d’influence (environnement, stratégies agropastorales, habitudes culturelles, etc.). L’analyse menée, sur la base de tels postulats, tend à mesurer puis décrire ces changements et cette diversité pour tenter d’en comprendre l’origine et les mécanismes de progression. Pour cela il a fallu observer les restes dentaires et osseux de plus de 600 sites archéologiques, situés sur les territoires gaulois, français et italien, datés de La Tène moyenne à la période moderne, à l’aide de différentes méthodes de morphométrie géométrique et traditionnelle. De cette manière, il a été possible de constituer un solide référentiel ostéométrique pour le porc et d’atteindre ici une meilleure compréhension des pratiques d’élevage aux périodes gauloise, romaine et médiévale / This archaeozoological study investigates morphological variations in different skeletal parts of the pig (Sus scrofa domesticus) according to two main assumptions: (1) the morphological changes of the animal reflect socio-economic changes; (2) there is a great diversity of pig’s forms, driven by different factors (environment, agropastoral strategies, cultural habits, etc.). Our analysis, based on these assumptions, tends to quantify and describe these changes and this diversification to try to understand their origin and mechanisms. For this we have observed the dental and skeletal remains of more than 600 archaeological sites, located on the Gallic, French and Italian territories, dated from La Tène moyenne to the modern period, using different methods of geometric and traditional morphometrics. In this way, we were able to build a strong osteometric framework for pigs and get new clues to a better understanding of farming practices for Gallic, Roman and medieval periods