abstract: Here I present a phylogeographic study of at least six reproductively isolated lineages of harvester ants within the Pogonomyrmex barbatus and P. rugosus species group. The genetic and geographic relationships within this clade are complex: four of the identified lineages are divided into two pairs, and each pair has evolved under a mutualistic system that necessitates sympatry. These paired lineages are dependent upon one another because interlineage matings within each pair are the sole source of hybrid F1 workers; these workers build and sustain the colonies, facilitating the production of the reproductive caste, which results solely from intralineage fertilizations. This system of genetic caste determination (GCD) maintains genetic isolation among these closely related lineages, while simultaneously requiring co-expansion and emigration as their distributions have changed over time. Previous studies have also demonstrated that three of the four lineages displaying this unique genetic caste determination phenotype are of hybrid origin. Thus, reconstructing the phylogenetic and geographic history of this group allows us to evaluate past insights and plan future inquiries in a more complete historical biogeographic context. Using mitochondrial DNA sequences sampled across most of the morphospecies' ranges in the U.S. and Mexico, I employed several methods of phylogenetic and DNA sequence analysis, along with comparisons to geological, biogeographic, and phylogeographic studies throughout the sampled regions. These analyses on Pogonomyrmex harvester ants reveal a complex pattern of vicariance and dispersal that is largely concordant with models of late Miocene, Pliocene, and Pleistocene range shifts among various arid-adapted taxa in North America. / Dissertation/Thesis / M.S. Biology 2012

abstract: Gene-centric theories of evolution by natural selection have been popularized and remain generally accepted in both scientific and public paradigms. While gene-centrism is certainly parsimonious, its explanations fall short of describing two patterns of evolutionary and social phenomena: the evolution of sex and the evolution of social altruism. I review and analyze current theories on the evolution of sex. I then introduce the conflict presented to gene-centric evolution by social phenomena such as altruism and caste sterility in eusocial insects. I review gene-centric models of inclusive fitness and kin selection proposed by Hamilton and Maynard Smith. Based their assumptions, that relatedness should be equal between sterile workers and reproductives, I present several empirical examples that conflict with their models. Following that, I introduce a unique system of genetic caste determination (GCD) observed in hybrid populations of two sister-species of seed harvester ants, Pogonomyrmex rugosus and Pogonomyrmex barbatus. I review the evidence for GCD in those species, followed by a critique of the current gene-centric models used to explain it. In chapter two I present my own theoretical model that is both simple and extricable in nature to explain the origin, evolution, and maintenance of GCD in Pogonomyrmex. Furthermore, I use that model to fill in the gaps left behind by the contributing authors of the other GCD models. As both populations in my study system formed from inter-specific hybridization, I review modern discussions of heterosis (also called hybrid vigor) and use those to help explain the ecological competitiveness of GCD. I empirically address the inbreeding depression the lineages of GCD must overcome in order to remain ecologically stable, demonstrating that as a result of their unique system of caste determination, GCD lineages have elevated recombination frequencies. I summarize and conclude with an argument for why GCD evolved under selective mechanisms which cannot be considered gene-centric, providing evidence that natural selection can effectively operate on non-heritable genotypes appearing in groups and other social contexts. / Dissertation/Thesis / M.S. Biology 2012

Les Hyménoptères sociaux (abeilles, fourmis et guêpes) doivent leur succès écologique à une division dutravail marquée par l’existence de castes. Les reines sont spécialisées dans la reproduction, tandis que les ouvrières sont stériles et prennent en charge la construction du nid et sa défense, la récolte de nourriture et l’élevage des jeunes. Le développement d’un œuf femelle en une reine ou une ouvrière est généralement régi par des facteurs environnementaux, tels que la qualité ou la quantité de la nourriture.En réalisant une étude de génétique des populations chez la fourmi Cataglyphis hispanica, nous avons mis en évidence un mode de reproduction inhabituel appelé hybridogenèse sociale. Cette espèce est caractérisée par la co-existence de deux lignées génétiques au sein des populations. Les reines de chaque lignée s’accouplent systématiquement avec un mâle de l’autre lignée génétique. Ainsi, les reines de la lignée 1 s’accouplent toujours avec un mâle de la lignée 2, et les reines de la lignée 2 s’accouplent avec un mâle de la lignée 1. Les ouvrières sont issues du croisement entre les deux lignées :il s’agit donc d’individus hybrides. A l’inverse, les individus reproducteurs (nouvelles reines et mâles) sont produits exclusivement par reproduction asexuée. La caste femelle (reine ou ouvrière) est donc déterminée génétiquement; les ouvrières portent un génome hybride, alors que les reproducteurs possèdent un génome non-hybride hérité exclusivement de la mère.Les travaux réalisés dans le cadre de cette thèse de doctorat visent à comprendre l’origine et l’évolution de l’hybridogenèse sociale chez les fourmis Cataglyphis par le biais d’approches variées (génétique des populations, phylogéographie et manipulation de colonies en laboratoire). Ils sont articulés selon plusieurs axes complémentaires :la description du phénomène, l’étude de sa distribution géographique et phylogénétique au sein du genre Cataglyphis, et l'étude de ces mécanismes génétiques sous-jacents. / In eusocial Hymenoptera, such as bees, wasp and ants, it is commonly accepted that the diploid female eggs are bipotent and develop either into a queen or a worker depending on environmental factors. While conducting a population genetic study of the ant Cataglyphis hispanica, we discovered an unusual reproductive system called social hybridogenesis. Under this system, queens and workers develop from eggs with different genetic make-up. Two divergent genetic lineages coexist within a population. Queens of each lineage mate with males originating from the other lineage. Workers are produced by sexual reproduction; hence, they are first generation hybrids of the two lineages. By contrast, new queens and males are produced by thelytokous and arrhenotokous parthenogenesis, respectively. By using alternative modes of reproduction, queens appear to maximize their reproductive success by increasing the transmission rate of their genes to the reproductive offspring while benefiting from genetic diversity and heterosis effects in their worker force. Males, however, do not contribute genes to the next generation as they only sire sterile workers. This system is expected to be evolutionary short-lived as lineages may be selected to stop the production of males with null direct fitness. This PhD project aims at understanding the origin and evolution of social hybridogenesis in Cataglyphis. Using several different approaches including population genetics methods based on microsatellite makers, phylogeography and experimental manipulations on lab colonies, I explored tthe characteristics of social hybridogenesis, the genetic mechanisms underlying the system and its distribution within the genus Cataglyphis. My results, together with other recent findings, question the paradigm of environmental caste determination in eusocial Hymenoptera and suggest that genetic influences on caste determination may be more common than previously thought. / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Com o avanço da meliponicultura, a utilização das abelhas sem ferrão, assim como de seus subprodutos, tem abrangido novas áreas, como a polinização de culturas agrícolas. Assim, a demanda pelo aumento do número de colônias tem sido constante, porém devido ao pouco conhecimento sobre a biologia reprodutiva dessas abelhas, há dificuldade de produção de colônias em larga escala, acarretando uma séria limitação quanto à utilização comercial desses polinizadores. Buscamos com este trabalho, oferecer ferramentas que possibilitem a multiplicação de colônias de abelhas sem ferrão em grande quantidade em um curto período de tempo. Nas abelhas indígenas sem ferrão (exceto no gênero Melipona e nos casos onde ocorrem rainhas-miniatura) a quantidade de alimento ingerido pelas larvas fêmeas é o fator responsável pela diferenciação das castas, pois as larvas que se tornam rainhas ingerem mais alimento que as larvas de futuras operárias, não havendo diferença qualitativa entre o alimento fornecido às larvas que originarão ambas as castas. De acordo com este modelo de determinação de castas, buscamos estabelecer a produção in vitro de rainhas em Tetragonisca angustula oferecendo maior quantidade de alimento às larvas de operárias, induzindo seu desenvolvimento em rainhas, que após a emergência foram introduzidas em mini-colônias órfãs para verificação da sua viabilidade (fecundação natural e postura de ovos). Em condições naturais as larvas de T. angustula que se tornam rainhas e operárias recebem respectivamente 55 µL e 8 µL de alimento em média. Assim, para a produção in vitro de rainhas oferecemos 55 µL de alimento a larvas de operárias coletadas em estágio pré-alimentação de ninhos naturais e conseguimos uma taxa de sobrevivência de até 51% e de 19% na obtenção de rainhas fisiogástricas, configurando um avanço em relação à taxa natural de emergência de rainhas, que foi de 0,21%. Das mini-colônias onde foram introduzidas rainhas virgens, 41% tiveram sucesso e se tornaram colônias perenes. A utilização de larvas de operárias na produção in vitro de rainhas é possível devido ao fato de as larvas serem totipotentes, assim como a utilização de alimento coletado de células de cria de operárias/machos, pois os resultados das análises comparativas do conteúdo protéico e de aminoácidos totais e livres dos alimentos contidos em células de cria de operárias/machos e células reais mostraram não haver diferenças significativas. A utilização de alimento larval de Scaptotrigona aff. depilis na criação in vitro de rainhas de T. angustula mostrou a possibilidade de produção de rainhas viáveis com esta nova técnica embora os perfis protéicos dos alimentos larvais de ambas as espécies sejam diferentes. Os experimentos de produção in vitro de rainhas com diferentes quantidades de alimento oferecido às larvas mostraram a existência de uma quantidade limite de alimento entre 35 µL e 45 µL acima da qual todos os indivíduos se tornam rainhas e abaixo da qual todos se tornam operárias, não havendo a ocorrência de indivíduos intermediários (intercastas). Introduzindo as rainhas produzidas in vitro em mini-colônias órfãs, conseguimos o estabelecimento de 16 colônias perenes a partir de seis colônias doadoras de material no período de seis meses, o que configura um avanço de 33% em relação às técnicas tradicionais de multiplicação de colônias que conseguiriam formar, no máximo, 12 colônias no mesmo período. A verificação da freqüência de produção de sexuados nas colônias naturais ao longo do ano mostrou que os machos são produzidos sazonalmente com alta taxa no período de fevereiro a abril, e embora as rainhas possuam uma produção baixa e homogênea ao longo do ano, concluímos que aquelas produzidas no período onde ocorre maior disponibilidade de machos possuem maiores chances de serem fecundadas. Assim, foi extremamente importante sincronizar a produção in vitro de rainhas com o período de maior disponibilidade de machos, uma vez que a fecundação destas ocorreu naturalmente. A produção in vitro de rainhas e a multiplicação de colônias nas abelhas sem ferrão se tornam ferramentas importantes para as técnicas de manejo que visam obter colônias em larga escala, com emprego na polinização, meliponicultura e conservação. / With the improvement of meliponiculture, the use of stingless bees, as well as their byproducts, has comprised new areas, such as the pollination of crops. Thus, the requirement by increasing the number of colonies has been constant, but due to poor knowledge about reproductive biology of these bees, have been difficult produce colonies on a large scale, what causes serious limitations on the commercial use of these pollinators. Our goals were to offer tools that allowed the multiplication of stingless bees colonies in large quantities in a short time. In stingless bees (except in the genus Melipona and in cases where there are miniature-queens) the amount of food ingested by female larvae is the responsible feature for caste differentiation, because the larvae that will become queens must to ingest more food than the larvae of future workers, and there is no qualitative difference between the food provided to larvae of both castes. According to this model of caste determination, we aimed at to establish the in vitro production of queens in Tetragonisca angustula, offering a large quantity of larval food to the workers larvae, thus inducing their development into queens that after rearing were introduced in orphaned mini-colonies to verify their feasibility (natural mating and egg laying). Under natural conditions the larvae of T. angustula that become queens and workers receive respectively, 55 µL and 8 µL of food on average. Thus, for the in vitro production of queens we offered 55 µL of food to workers larvae collected in pre-feeding stage from natural nests and we were able to get a survival rate up to 51 % and 19% in obtaining viable queens, which is an improvement compared to 0,21% of rearing queens observed in natural hives. About 41% of mini-colonies, where viable queens were introduced were successful and have become perennial colonies. The use of workers larvae on in vitro production of queens is possible because the larvae are totipotent, as well as the use of food collected from workers/males brood cells, because the results of comparative analysis of protein content and total and free amino acids of food stored in worker/males brood cells and royal cells showed no significant differences. The use of larval food from Scaptotrigona aff. depilis on in vitro rearing of T. angustula queens showed the possibility of producing viable queens, with this new technique, although the protein profiles of larval food of both species were different. The experiments of in vitro production of queens with different amounts of food offered to the larvae showed the existence of a threshold quantity of food between 35 mL and 45 mL above which all individuals become queens and below which all individuals become workers, without the occurrence of intermediate individuals (intercaste). Inserting the queens produced in vitro in orphaned mini-colonies, we were able to establish 16 perennial colonies as from six colonies donors of material in six months, constituting an increase of 33% compared to traditional techniques of multiplication of colonies which would form up only 12 colonies in the same period. Verification of frequency of sexuals reared in natural colonies over the year showed that males are produced seasonally with high rate in the period from February to April, and although the queens have a low and homogeneous occurrence over the year, we concluded that the queens produced during the period where there is greater availability of males, have higher chances of being fertilized. Thus, was extremely important synchronize the in vitro production of queens with the period where there is greater availability of males, since fertilization of these queens occurred naturally. The in vitro production of queens and multiplication of colonies in stingless bees become important tools for handling techniques that aim to obtain colonies on a large scale, with employment in pollination, meliponiculture and conservation.

Com o avanço da meliponicultura, a utilização das abelhas sem ferrão, assim como de seus subprodutos, tem abrangido novas áreas, como a polinização de culturas agrícolas. Assim, a demanda pelo aumento do número de colônias tem sido constante, porém devido ao pouco conhecimento sobre a biologia reprodutiva dessas abelhas, há dificuldade de produção de colônias em larga escala, acarretando uma séria limitação quanto à utilização comercial desses polinizadores. Buscamos com este trabalho, oferecer ferramentas que possibilitem a multiplicação de colônias de abelhas sem ferrão em grande quantidade em um curto período de tempo. Nas abelhas indígenas sem ferrão (exceto no gênero Melipona e nos casos onde ocorrem rainhas-miniatura) a quantidade de alimento ingerido pelas larvas fêmeas é o fator responsável pela diferenciação das castas, pois as larvas que se tornam rainhas ingerem mais alimento que as larvas de futuras operárias, não havendo diferença qualitativa entre o alimento fornecido às larvas que originarão ambas as castas. De acordo com este modelo de determinação de castas, buscamos estabelecer a produção in vitro de rainhas em Tetragonisca angustula oferecendo maior quantidade de alimento às larvas de operárias, induzindo seu desenvolvimento em rainhas, que após a emergência foram introduzidas em mini-colônias órfãs para verificação da sua viabilidade (fecundação natural e postura de ovos). Em condições naturais as larvas de T. angustula que se tornam rainhas e operárias recebem respectivamente 55 µL e 8 µL de alimento em média. Assim, para a produção in vitro de rainhas oferecemos 55 µL de alimento a larvas de operárias coletadas em estágio pré-alimentação de ninhos naturais e conseguimos uma taxa de sobrevivência de até 51% e de 19% na obtenção de rainhas fisiogástricas, configurando um avanço em relação à taxa natural de emergência de rainhas, que foi de 0,21%. Das mini-colônias onde foram introduzidas rainhas virgens, 41% tiveram sucesso e se tornaram colônias perenes. A utilização de larvas de operárias na produção in vitro de rainhas é possível devido ao fato de as larvas serem totipotentes, assim como a utilização de alimento coletado de células de cria de operárias/machos, pois os resultados das análises comparativas do conteúdo protéico e de aminoácidos totais e livres dos alimentos contidos em células de cria de operárias/machos e células reais mostraram não haver diferenças significativas. A utilização de alimento larval de Scaptotrigona aff. depilis na criação in vitro de rainhas de T. angustula mostrou a possibilidade de produção de rainhas viáveis com esta nova técnica embora os perfis protéicos dos alimentos larvais de ambas as espécies sejam diferentes. Os experimentos de produção in vitro de rainhas com diferentes quantidades de alimento oferecido às larvas mostraram a existência de uma quantidade limite de alimento entre 35 µL e 45 µL acima da qual todos os indivíduos se tornam rainhas e abaixo da qual todos se tornam operárias, não havendo a ocorrência de indivíduos intermediários (intercastas). Introduzindo as rainhas produzidas in vitro em mini-colônias órfãs, conseguimos o estabelecimento de 16 colônias perenes a partir de seis colônias doadoras de material no período de seis meses, o que configura um avanço de 33% em relação às técnicas tradicionais de multiplicação de colônias que conseguiriam formar, no máximo, 12 colônias no mesmo período. A verificação da freqüência de produção de sexuados nas colônias naturais ao longo do ano mostrou que os machos são produzidos sazonalmente com alta taxa no período de fevereiro a abril, e embora as rainhas possuam uma produção baixa e homogênea ao longo do ano, concluímos que aquelas produzidas no período onde ocorre maior disponibilidade de machos possuem maiores chances de serem fecundadas. Assim, foi extremamente importante sincronizar a produção in vitro de rainhas com o período de maior disponibilidade de machos, uma vez que a fecundação destas ocorreu naturalmente. A produção in vitro de rainhas e a multiplicação de colônias nas abelhas sem ferrão se tornam ferramentas importantes para as técnicas de manejo que visam obter colônias em larga escala, com emprego na polinização, meliponicultura e conservação. / With the improvement of meliponiculture, the use of stingless bees, as well as their byproducts, has comprised new areas, such as the pollination of crops. Thus, the requirement by increasing the number of colonies has been constant, but due to poor knowledge about reproductive biology of these bees, have been difficult produce colonies on a large scale, what causes serious limitations on the commercial use of these pollinators. Our goals were to offer tools that allowed the multiplication of stingless bees colonies in large quantities in a short time. In stingless bees (except in the genus Melipona and in cases where there are miniature-queens) the amount of food ingested by female larvae is the responsible feature for caste differentiation, because the larvae that will become queens must to ingest more food than the larvae of future workers, and there is no qualitative difference between the food provided to larvae of both castes. According to this model of caste determination, we aimed at to establish the in vitro production of queens in Tetragonisca angustula, offering a large quantity of larval food to the workers larvae, thus inducing their development into queens that after rearing were introduced in orphaned mini-colonies to verify their feasibility (natural mating and egg laying). Under natural conditions the larvae of T. angustula that become queens and workers receive respectively, 55 µL and 8 µL of food on average. Thus, for the in vitro production of queens we offered 55 µL of food to workers larvae collected in pre-feeding stage from natural nests and we were able to get a survival rate up to 51 % and 19% in obtaining viable queens, which is an improvement compared to 0,21% of rearing queens observed in natural hives. About 41% of mini-colonies, where viable queens were introduced were successful and have become perennial colonies. The use of workers larvae on in vitro production of queens is possible because the larvae are totipotent, as well as the use of food collected from workers/males brood cells, because the results of comparative analysis of protein content and total and free amino acids of food stored in worker/males brood cells and royal cells showed no significant differences. The use of larval food from Scaptotrigona aff. depilis on in vitro rearing of T. angustula queens showed the possibility of producing viable queens, with this new technique, although the protein profiles of larval food of both species were different. The experiments of in vitro production of queens with different amounts of food offered to the larvae showed the existence of a threshold quantity of food between 35 mL and 45 mL above which all individuals become queens and below which all individuals become workers, without the occurrence of intermediate individuals (intercaste). Inserting the queens produced in vitro in orphaned mini-colonies, we were able to establish 16 perennial colonies as from six colonies donors of material in six months, constituting an increase of 33% compared to traditional techniques of multiplication of colonies which would form up only 12 colonies in the same period. Verification of frequency of sexuals reared in natural colonies over the year showed that males are produced seasonally with high rate in the period from February to April, and although the queens have a low and homogeneous occurrence over the year, we concluded that the queens produced during the period where there is greater availability of males, have higher chances of being fertilized. Thus, was extremely important synchronize the in vitro production of queens with the period where there is greater availability of males, since fertilization of these queens occurred naturally. The in vitro production of queens and multiplication of colonies in stingless bees become important tools for handling techniques that aim to obtain colonies on a large scale, with employment in pollination, meliponiculture and conservation.