Les symbioses ont un role majeur dans le fonctionnement et l'equilibre des ecosystemes. Dans les oceans, qui couvrent pres de 70 % de la surface de la planete, vivent une multitude d'organismes incapables de lutter contre les courants et la plupart sont microscopiques, il s'agit du plancton. Les organismes du plancton, comme ceux d'autres ecosystemes, entretiennent des symbioses, mais la nature et l'ampleur de ces interactions sont encore mal connues dans le plancton du fait la petite taille de ces organismes et de la difficulte d'echantillonnage des ecosystemes planctoniques, surtout dans les zones les plus eloignees des cotes. Les principaux objectifs de cette these sont de donner un apercu global de la place occupee par ces symbioses dans le plancton et de proposer des methodes originales permettant leur detection. Les travaux presentes dans ce manuscrit s'appuient sur l'analyse des donnees generees lors de l'expedition Tara Oceans (2009-2013) pendant laquelle 210 stations oceaniques ont ete echantillonnees a travers le monde. Ils concernent plus precisement le jeu de donnees environnemental obtenu grace au sequencage a haut debit (Illumina) de la region hypervariable V9 (130 nucleotides) de la sousunite 18S de l'ADN ribosomique des organismes eucaryotes (metabarcoding). Dans un premier temps, un etat des lieux de la diversite et de la structure des communautes du pico-nano-micro-mesoplancton (0,8-2000 μm) eucaryote de la zone photique des oceans temperes a tropicaux est realise. Il met en evidence la place importante occupee par les symbiotes au sein de ces communautes. Ensuite, l'etude de deux cas de symbiose (Blastodinium- Copepodes et Symbiodinium-Tiarina) montre les difficultes inherentes a la detection de couples symbiotiques a partir d’un jeu de donnees issue d'etudes par metabarcoding du plancton (flexibilite de la specificite des symbioses dans le plancton), mais aussi la possibilite de distinguer les differentes phases de vie des symbiotes (libres et symbiotiques) lorsque les echantillons etudies ont ete fractionnes. Enfin, un ensemble de methodes est propose afin d'ameliorer l'efficacite de la detection de symbioses dans le cadre d'etudes par reseau de cooccurrences des communautes planctoniques. L'analyse de la distribution des metabarcodes le long des fractions de taille (piconano- (0.8-5 μm), nano- (5-20 μm), micro- (20-180 μm), et meso-plancton (180-2000 μm)) permet de differencier ceux provenant d'organismes symbiotiques de ceux d'organismes libres, sans a priori. De plus la comparaison de l'abondance de groupes genetiques definis a differents niveaux de resolution permet de detecter des associations symbiotiques peu specifiques et d'apprecier leur niveau de specificite. / The oceans, which cover nearly 70 % of the earth's surface, is host to a myriad of mostly microscopic organisms that drift with the currents and are collectively called plankton. As in other ecosystems, symbioses play a major role in the functioning and equilibrium of the plankton. But the exact nature and strength of those symbiotic interactions are still poorly known, not only due to the small size of most planktonic organisms, but also because of the inherent difficulty of sampling planktonic ecosystems, especially in the high-seas. The main goals of this thesis are to give a global view of the importance of planktonic symbioses and to propose novel methods for their detection. The work presented in this manuscript is based on analyses of data generated during the Tara Oceans expedition (2009-2013), during which sea water was collected and size fractionated by filtration at 210 sampling locations distributed across the world's oceans. The data analyses presented herein mostly focus on an environmental metabarcoding dataset obtained from next-generation sequencing (Illumina) of the V9 hypervariable region (~130 nucleotides long) of the 18S small ribosomal subunit of eukaryotic organisms. We begin by assessing the diversity and structure of pico-, nano-, micro and meso-planktonic eukaryotic communities (0.8-2000 μm) in the photic zone of tropical to temperate sea regions. Then, we present two cases of symbioses (Blastodinium-Copepods and Symbiodinium-Tiarina) to illustrate both the difficulties encountered when trying to detect symbiotic relationships using metabarcoding data due to varying specificities of symbiotic relationships, but also the potential solutions offered by size-fractionated sampling to distinguish between the different stages of the life cycle of symbiotic organisms (free living and symbiotic stages). Finally, we propose a set of methods to improve the detection of symbioses by studying the co-occurrence of organisms in planktonic communities: we use the distribution of metabarcodes along size fractions ((piconano- (0.8-5 μm), nano- (5-20 μm), micro- (20-180 μm), and meso-plancton (180-2000 μm)) to distinguish likely free living organisms from those that have a symbiotic life style, and we compare the abundance of genetic groups constructed by clustering metabarcodes at different resolution levels, which allows us to detect interactions occurring above the species level and to evaluate their level of specificity.

Mangrove forests have one of the highest capacities of any ecosystem to sequester carbon. Mangroves in the Red Sea exist in a uniquely saline, high temperature, nutrient limited environment and the effects on carbon storage and greenhouse gas (GHG) emissions from arid mangrove soils is understudied. The flux of carbon dioxide (CO2) and methane (CH4) has the potential to enhance or reduce the carbon storage capacity of mangroves, which is an important nature-based solution for carbon drawdown to limit global warming. To determine the magnitude of CO2 and CH4 flux from mangrove soil in the Red Sea, soil cores were incubated on a monthly basis for over a year in light and dark conditions. Soil properties such as salinity, organic carbon, water content, bulk density, and stable isotopes, along with environmental variables such as inundation frequency and temperature were measured to resolve the drivers and variation of GHG flux over time. Additionally, 16S and 18S rRNA metabarcoding was conducted to determine the relative influence of prokaryotes and eukaryotes in the microbial mat within this mangrove ecosystem, and the microbial contribution to GHG flux. Oxygen microsensors were used for fine-scale resolution of the microbial mat, to determine photosynthetic rates and oxygen profiles. Fluxes were found to be highly variable, with the highest correlation between GHG flux and soil water content (p<0.05). Both prokaryotic and eukaryotic components of the microbial mat had a significant relationship with GHG flux, with mixed impacts depending on the taxa. These findings show that Red Sea mangroves, despite their lower carbon storage capacity, are a negligible source of GHG to the atmosphere unlike other regions where GHG emissions offset a greater proportion of carbon storage potential. Additionally, the importance of the microbial mat in this ecosystem is demonstrated, and an important consideration for future studies on mangroves and their potential as a nature-based solution against global warming.

Underwater visual censuses (UVCs) are one of the most widely used methods of studying species-rich coral reef fish assemblages. However, a considerable portion of reef fish diversity is missed or underrepresented by these traditional survey techniques. Environmental DNA (eDNA) sampling is an emerging technology that can detect traces of animal DNA from environmental samples, such as water and sediment, potentially including taxa that are missed by UVCs. Here, we assess the complementarity of eDNA to UVCs in surveying coral reef fish communities, particularly for cryptic and cryptobenthic taxa. We further investigate the effect of environmental sample source (water and sediment) and depth (10m and 30m). We conducted UVCs and eDNA sampling in three islands of the Farasan Banks, southern Saudi Arabia. A metabarcoding protocol was applied to environmental samples using a broad-spectrum fish assay targeting 16S mitochondrial DNA. Our eDNA surveys revealed 94 fish species, across 86 genera, 38 families, and 14 orders. Of the species detected by eDNA, 48.9% were also recorded on transects and 60.6% on roving diver surveys. eDNA also detected 6 cryptic, 10 cryptobenthic, and 13 pelagic species. Of these, only one (Eviota guttata) was recorded by UVCs. eDNA species composition was found to be significantly influenced by collection site (islands), and sample source (more species detected from water samples than sediment samples), but not by collection depth (10 versus 30 m depth). Our study provides further evidence that eDNA is an effective tool for the biomonitoring of tropical coral reef fish communities. However, we also stress that improvements are needed in methodology and reference sequence coverage for eDNA to realize its full potential of capturing cryptic and cryptobenthic diversity.

Freshwater fungi are vital to the aquatic food web, essential for nutrient cycling, energy flow, and ecosystem regulation. Their distribution is particularly contingent upon agricultural runoff, which can carry agrochemicals capable of influencing the freshwater mycobiota and potentially impacting the ecosystem services which they provide. While such impacts are well documented for freshwater bacterial communities, fungal communities are critically understudied. Here, we address this research gap by assessing the impact of anthropological and environmental perturbations on the freshwater mycobiota in the agriculturally dominated South Nation River basin in Eastern Ontario, Canada. We undertook biweekly water sampling from 2016-2021, complemented by rich ancillary data including water properties, hydrology, weather conditions, and fungal ITS2 metabarcoding. Our study yielded 6,571 Operational Taxonomic Units from 503 water samples, spanning 15 fungal phyla, dominated by Ascomycota, Basidiomycota, and Chytridiomycota. Agricultural land use decreased the mycobiota alpha diversity and distinct fungal communities were observed at agricultural ditch and mixed-use sites compared to the forested site. Notably, river discharge emerged as a predominant influencer of both alpha and beta diversity, likely transporting fungi via precipitation, especially from plant-rich catchment basins. Intriguingly, environmental data only explained a fraction of fungal community variation, underscoring the significance of unmeasured factors such as fungicide application, alongside stochastic community assembly processes. This work highlights the complex interplay of factors influencing the freshwater fungal community in agriculturally impacted watersheds and shows the need for further investigation for a more comprehensive understanding of the freshwater fungal ecology.

Yes / Marine plastic pollution has a range of negative impacts for biota and the colonization of plastics in the marine environment by microorganisms may have significant ecological impacts. However, data on epiplastic organisms, particularly fungi, is still lacking for many ocean regions. To evaluate plastic associated fungi and their geographic distribution, we characterised plastics sampled from surface waters of the western South Atlantic (WSA) and Antarctic Peninsula (AP), using DNA metabarcoding of three molecular markers (ITS2, 18S rRNA V4 and V9 regions). Numerous taxa from eight fungal phyla and a total of 64 orders were detected, including groups that had not yet been described associated with plastics. There was a varied phylogenetic assemblage of predominantly known saprotrophic taxa within the Ascomycota and Basidiomycota. We found a range of marine cosmopolitan genera present on plastics in both locations, i.e., Aspergillus, Cladosporium, Wallemia and a number of taxa unique to each region, as well as a high variation of taxa such as Chytridiomycota and Aphelidomycota between locations. Within these basal fungal groups we identified a number of phylogenetically novel taxa. This is the first description of fungi from the Plastisphere within the Southern Hemisphere, and highlights the need to further investigate the potential impacts of plastic associated fungi on other organisms and marine ecosystems. / Conselho Nacional de Desenvolvimento Científico e Tecnológico. Grant Numbers: 407889/2013‐2, PQ 310597/2018‐8, PQ 312470/2018‐5, SWE 206250/2017‐7. Programa de Excelênca Acadêmica – PROEX. University of Salford Career Development Fellowship. Chevron Brasil Upstrem Frade Ltda. Brazilian Interministerial Commission for the Resources of the Sea (CIRM).

Cette thèse s'inscrit dans le cadre du traitement des données issues de séquençage haut débit, et en particulier des données produites en metabarcoding ADN. Le metabarcoding ADN consiste à identifier des taxons ou des groupes taxinomiques à partir de l'ADN présent dans des échantillons environnementaux (eau, sol, fèces...). Après extraction de l'ADN, de courtes séquences utilisées comme marqueurs taxinomiques sont amplifiées par PCR puis séquencées en utilisant les nouvelles techniques de séquençage haut débit. De très importants volumes de données sont ainsi générés, le plus souvent, de plusieurs milliers à plusieurs centaines de milliers de séquences par échantillon. L'objectif principal de cette thèse était le développement de méthodes d'analyse de ces séquences. Les méthodes de classification permettent de traiter de nombreuses problématiques en metabarcoding ADN. La classification supervisée est utilisée pour assigner les séquences à des taxons en les comparant aux séquences de bases de données de référence. Les méthodes de classification non supervisée permettent de créer des groupes taxinomiques (MOTU) à partir des séquences, afin de faire des estimations de biodiversité. Ces méthodes sont aussi employées pour identifier les séquences erronées produites par la PCR et le séquençage notamment, où les séquences erronées dérivent souvent des vraies séquences et leur sont très similaires. Les méthodes de classification demandent une méthode de comparaison des séquences qui soit idéalement à la fois très rapide et exacte. Une telle méthode a été développée, en utilisant un algorithme d'alignement global de type Needleman-Wunsch calculant la longueur de la plus longue sous-séquence commune entre les séquences à aligner, associé à un filtre sans perte permettant d'éviter l'alignement de certaines paires de séquences n'ayant aucune chance de présenter une similarité supérieure à un seuil choisi. L'utilisation d'instructions Single Instruction, Multiple Data, de même que le multithreading optionnel des calculs, permettent d'associer rapidité et exactitude. Cette méthode de comparaison est implantée dans SUMATRA, un programme calculant toutes les similarités deux à deux d'un jeu de données ou entre deux jeux de données, avec possibilité de fixer un seuil de similarité en dessous duquel les similarités ne sont pas rapportées. Elle est aussi utilisée dans SUMACLUST. SUMACLUST est un programme regroupant les séquences en utilisant un algorithme de clustering en étoile, où chaque groupe possède une séquence représentative. Il peut être utilisé pour créer des MOTU, ou pour détecter les séquences erronées dérivant de vraies séquences. Plus spécialisé, le programme SUMACLEAN a été développé pour détecter les séquences contenant des erreurs ponctuelles de PCR. Pour cela, des graphes orientés acycliques sont générés, dont la topologie correspond parfaitement aux cascades d'erreurs générées par les erreurs ponctuelles de PCR. Par ailleurs, une réflexion a été menée pour le développement d'une nouvelle approche de classification supervisée pour l'assignation taxinomique des séquences. Aujourd'hui, la plupart des approches d'assignation utilisent des méthodes mal adaptées au polymorphisme important des marqueurs, et ne considèrent pas suffisamment l'incomplétude et les erreurs inhérentes aux bases de données de référence. Une nouvelle approche a été testée, basée sur l'idée d'un départ depuis la racine de l'arbre taxinomique, suivi d'une descente jusqu'à un arrêt possible lorsque descendre à un niveau taxinomique plus précis semble irraisonnable. Cela permettrait en théorie de mieux gérer les problèmes inhérents aux bases de données de référence, mais pose le problème de la représentation des séquences aux différents niveaux de l'arbre, et du modèle de choix du chemin à prendre, pour lesquels aucune solution complètement satisfaisante n'a été trouvée à ce jour. / This thesis positions itself in the context of the processing of high-throughput sequencing data, and specifically DNA metabarcoding data. DNA metabarcoding consists of the identification of taxa or taxonomic groups from DNA extracted from environmental samples (water, soil, animal feces). After extraction of the DNA, short sequences used as taxonomic markers are amplified by PCR, then sequenced using high-throughput sequencing technologies. Important volumes of data are produced that way, usually from several thousands to several hundreds of thousands sequences per sample. This thesis aimed for the development of methods for the analysis of these sequences. Classification methods allow the treatment of numerous problems in DNA metabarcoding. Supervised classification is used for the taxonomic assignment of sequences to taxa, by comparing them to the sequences of a reference database. Unsupervised classification methods are used to create taxonomic groups (MOTUs) from the sequences, in order to estimate biodiversity. They are also used to identify the erroneous sequences generated during the PCR and sequencing steps in particular, where erroneous sequences often derive from true sequences and remain very close to them. Classification approaches used in the context of DNA metabarcoding necessitate a sequence comparison method that should be both fast and exact. Such a method was developed, using a Needleman-Wunsch type global alignment algorithm computing the length of the longest common subsequence between the two sequences being aligned, associated with a lossless filter allowing to avoid the alignment of some pairs of sequences that have no chance to present a similarity superior to a chosen threshold. The use of Single Instruction, Multiple Data instructions, as well as the availability of multithreading speed up the calculations. This comparison method is implanted in SUMATRA, a program computing all the pairwise similarities of a dataset or between two datasets, with the possibility to set a threshold under which similarities are ignored. It is also used in SUMACLUST, a program grouping sequences using a star clustering algorithm, where each cluster possesses a representative sequence. This algorithm can be used to generate MOTUs, or to identify erroneous sequences deriving from true sequences, by using the fact that true sequences tend to end up as the representative sequences of their cluster. More specialized, the SUMACLEAN program was developed to identify sequences containing ponctual PCR errors. To that end, directed acyclic graphs are created, whose topology matches perfectly the successions of errors generated by ponctual errors during PCR. A new approach for the taxonomic assignment of sequences with a supervised classification method was also studied. Nowadays, most taxononomic assignment approaches use methods that are badly suited for the important polymorphism of markers, and don't take in account enough the incompleteness and errors inherent to reference databases. A new approach was tested, based on the idea of a start from the root of the taxonomic tree, and a descent in it with a possible stop before reaching a leaf, if descending to a more precise taxonomic level seems unreasonable. This approach would theoretically allow for a better handling of the problems inherent to reference databases, but poses a few issues, such as the representation of sequences at intermediate tree levels, and the model used to make choices regarding the path to take in the tree, for which no satisfying solutions have been found yet.

Selon l'hypothèse de l'effet de dilution, les écosystèmes les plus riches en espèces seraient également les moins propices à la circulation des agents infectieux du fait de la présence d'hôtes non compétents constituant des impasses épidémiologiques. Dans cette thèse, nous explorons l'existence de ce phénomène sur le modèle des leishmanioses cutanées zoonotiques en Guyane. Des outils moléculaires basés sur l'utilisation des technologies de séquençage haut-débit ont été développés pour étudier le système épidémiologique. Ces outils ont ensuite été employés pour caractériser le cycle de transmission des leishmanioses dans des sites forestiers sujets à différents niveaux de perturbation d'origine humaine. Nos résultats semblent globalement congruents avec l'hypothèse de l'effet de dilution, et indiquent un risque infectieux plus élevé sur le site le plus perturbé. Toutefois, les différences observées entre les sites ne sont pas significatives et d'avantage de données seraient nécessaires pour tirer des conclusions générales. / The dilution effect hypothesis states that more diverse ecological communities are less prone to pathogen transmission because of the presence of non-competent hosts acting as epidemiological dead-ends. In this work, we investigate the existence of this phenomenon in the case of zoonotic cutaneous leishmaniases in French Guiana. Molecular tools based on high-throughput sequencing technologies have been developed to study the epidemiological system. These tools were employed to explore leishmaniases transmission cycles in forest sites undergoing different levels of human-induced perturbations. Our results seem generally congruent with the dilution effect hypothesis, indicating higher disease risk in the most perturbed site. However, differences observed between sites were not significant, and more data is needed to draw general conclusions.

Les diatomées (Stramenopiles, Bacillariophyceae) jouent un rôle important sur le plan écologique et sont l'un des groupes phytoplanctoniques les plus divers, avec environ 1800 espèces planctoniques estimées. Bien que largement étudiées, leurs modèles de diversité et de distribution biogéographique ne sont pas bien connus. L'avènement du séquençage de l'ADN à haut débit a révolutionné les études de biodiversité moléculaire facilitant la compréhension de la biogéographie, de la structure des communautés et des processus écologiques. Les deux principaux objectifs de cette thèse sont (1) d'enquêter sur les modèles de la biodiversité mondiale et la structure des communautés de diatomées planctoniques à travers les océans du monde, et (2) de comprendre les mécanismes et processus déterminants la structure de la communauté. Cette thèse présente également une première tentative de discerner la répartition des espèces rares dans les communautés de protistes. L'étude a été réalisée en utilisant les données de metabarcoding générées à partir des échantillons biologiques et des données environnementales associées recueillies au cours de la campagne Tara Oceans (2009-2013), une circumnavigation globale couvrant toutes les principales provinces océaniques. Le matériel d’étude pour cette thèse est constitué d’un total de 12 millions de séquences de la sous unité V9 du 18S ribosomal (barcode), récoltées à partir de 46 stations soit 293 échantillons. Basée sur 63371 metabarcodes de diatomées uniques, cette étude présente une évaluation approfondie de la distribution mondiale des diatomées et de leur diversité. Les analyses révèlent des faits marquants liées à la biogéographie des diatomées, par exemple une nouvelle estimation du nombre total d'espèces de diatomées planctoniques, une diversité considérable inconnue, une diversité exceptionnellement élevée en haute mer, et des patrons de diversité complexes entre les provinces océaniques. La thèse examine ensuite les facteurs qui déterminent les modèles de bêta-diversité. Les résultats suggèrent que les diatomées sont des communautés structurées et réglementées par l'hétérogénéité de l'environnement et des processus spatiaux. Néanmoins, la majorité de la variation totale dans la composition de la communauté ne peut être expliquée ni par les facteurs environnementaux, ni par les distances spatiales, ce qui justifie les analyses futures se concentrant sur les interactions biologiques, les événements historiques, et d'autres facteurs qui ne sont pas considérés. La thèse décrit en outre une approche pour caractériser les clusters significativement associés de ribotypes concomitants. Enfin, une étude préliminaire de communautés de protistes fractionnées par taille révèle que la queue (de leurs distributions rang abondance) semble suivre un comportement en loi de puissance dans presque toutes les communautés de protistes. Cette observation peut indiquer un mécanisme universel potentiel qui peut expliquer l'organisation de communautés planctoniques marines. De façon générale, ce travail présent une perspective globale et complète de la distribution et de la diversité des diatomées dans les océans du monde. La thèse propose un cadre global pour l'évaluation de la diversité mondiale basée sur le metabarcoding, qui pourra être utilisé pour étudier la distribution et la diversité des autres lignées taxonomiques. Par conséquent, ce travail fournit un point de référence pour explorer comment les communautés microbiennes feront face à la variation des conditions environnementales. / Diatoms (Stramenopiles, Bacillariophyceae) are an ecologically important and one of the most diverse phytoplanktonic groups, with an estimated ~1,800 marine planktonic species. Although widely studied, their diversity and biogeographic distribution patterns are not well known. The advent of high-throughput DNA sequencing has revolutionized molecular biodiversity studies facilitating the understanding of biogeography, community assembly and ecological processes. The two major goals of this thesis are (1) to investigate global biodiversity patterns and structure of marine planktonic diatom communities across the world’s oceans, and (2) to understand the mechanisms and processes determining their community structure and assembly. This thesis also presents an initial attempt to discern the distribution of rare species in protist communities. The study was conducted using the metabarcoding data generated from the biological samples and associated environmental data collected during the Tara Oceans (2009-2013) global circumnavigation covering all major oceanic provinces. A total of ~12 million diatom V9-18S rDNA tags from 46 sampling stations, constituting 293 size fractionated samples represent the study material for the thesis. Using 63,371 unique diatom metabarcodes, this study presents an in-depth evaluation of global diatom distribution and diversity. The analyses study draw a number of revelations related to diatom biogeography, e.g. a new estimate of the total number of planktonic diatom species, a considerable unknown diversity, exceptionally high diversity in the open ocean, complex diversity patterns across oceanic provinces. The thesis then looks into the factors determining the beta-diversity patterns. The results suggest that diatoms represent biogeographically structured ecological communities regulated by both environmental heterogeneity and spatial processes. Nonetheless, the majority of the total variation in community composition remained unexplained by either the examined measured environmental factors or spatial distances, which warrants future analyses focusing on biological interactions, historical events, and other factors that are not considered. The thesis further outlines an approach to characterize significantly associated clusters of co-occurring ribotypes. Finally, a preliminary study of size-fractionated protistan communities reveals that the tail (of their rank-abundance distributions) appears to follow a power-law behavior in almost all protistan communities. This observation may indicate a potential universal mechanism which can explain the organization of marine planktonic communities. In general, this work has presented a global comprehensive perspective on diatom distribution and diversity in the world’s oceans. The thesis offers an overall framework for metabarcoding-based global diversity assessments which in turn can be employed to study distribution and diversity of other taxonomic lineages. Consequently, this work provides a reference point to explore how microbial communities will respond/change in response to environmental conditions.

Le développement des plantes est conditionné par leurs interactions avec le microbiote racinaire, dont le fonctionnement repose sur des fonctions élémentaires partagées par plusieurs taxons, au sein de groupes fonctionnels. Certains de ces groupes incluent des PGPR (Plant Growth-Promoting Rhizobacteria), qui stimulent la croissance des plantes grâce à différentes fonctions phytobénéfiques. L'inoculation d'une PGPR peut avoir un impact sur la physiologie de la plante et sur les autres partenaires microbiens de la racine. D'ailleurs, une PGPR inoculée peut modifier la structure génétique du microbiote racinaire global, sans que l'on sache comment cet impact se concrétise au niveau de groupes fonctionnels particuliers. L'hypothèse de cette thèse est que les PGPR inoculées peuvent modifier la structure de communautés microbiennes fonctionnelles importantes pour le développement de la plante. Ces travaux ont évalué l'impact de l'inoculation de la PGPR Azospirillum lipoferum CRT1 sur la taille et/ou la diversité des diazotrophes, des producteurs d'1-aminocyclopropane-1-carboxylate (ACC) désaminase et des producteurs de 2,4-diacétylphloroglucinol (ainsi que de la communauté bactérienne totale) dans la rhizosphère du maïs cultivé en champ sous différents intrants azotés. Nos travaux ont montré que diazotrophes et producteurs d'ACC désaminase pouvaient être co-sélectionnés dans la rhizosphère du maïs, et que l'inoculation modifiait la composition taxonomique de ces groupes fonctionnels. En conclusion, l'inoculation de PGPR module la sélection rhizosphérique de taxons telluriques portant des fonctions phytobénéfiques, ce qui pourrait contribuer aux effets phytobénéfiques observés / Plant development is influenced by interactions established with root microbiota. Microbiota functioning relies on individual functions carried out by functional groups, which often contain multiple microbial taxa. Some of these groups include plant growth-promoting rhizobacteria (PGPR), which stimulate plant growth thanks to several plant-beneficial functions. PGPR inoculation may impact plant physiology and microbial partners of roots, and inoculated PGPR can modify the genetic structure of the root microbiota. However, it is not known how this impact materializes at the level of particular microbial functional groups. The hypothesis of this thesis is that inoculated PGPR will also modify the structure/composition of microbial functional groups important for plant development. This work assessed the impact of inoculation with the PGPR Azospirillum lipoferum CRT1 on the size and/or diversity of diazotrophs, 1-aminocyclopropane-1-carboxylate (ACC) deaminase producers and 2,4-diacetylphloroglucinol producers (as well as the total bacterial community) colonizing the maize rhizosphere in three fields and under different nitrogen fertilization regimes. Our work showed that diazotrophs and ACC deaminase producers could be co-selected in maize rhizosphere and that PGPR inoculation modified the taxonomic composition of functional groups. To conclude, PGPR inoculation modulates rhizospheric selection of microbial members of functional groups important for plant growth, which may mediate some of the plant-beneficial effects of PGPR inoculants

Doctor of Philosophy / Department of Biology / David A. Haukos / The lesser prairie-chicken (Tympanuchus pallidicinctus) is 1 of 3 prairie-grouse species in North America. Prairie-grouse have undergone local or widespread declines due to a loss of habitat through conversion to row crop agriculture, anthropogenic development, and alteration of ecological drivers that maintain quality grasslands. For lesser prairie-chickens, habitat loss and declines were deemed significant for listing as threatened under the Endangered Species Act in 2014. Despite a judge vacating the listing decision in 2015, the lesser prairie-chicken remains a species of concern. Conservation plans are currently being implemented and developed. To maximize the effectiveness of efforts, knowledge of the distribution of lesser prairie-chickens, regional demography, foods used during critical life-stages, and where to prioritize management is needed. To guide future conservation efforts with empirical evidence, I captured, marked with transmitters, and monitored female lesser prairie-chickens in Kansas and Colorado during 2013–2016 (n =307). I used location data to predict the distribution of habitat. Encounter data from individuals were used to estimate vital rates and integrated into a matrix population model to estimate population growth rates (λ). The matrix model was then decomposed to identify life-stages that exert the greatest influence on λ and vital rate contributions to differences in λ among sites. After assessing demography, I examined the diet of adults and chicks during critical brood rearing and winter periods using a fecal DNA metabarcoding approach. Overall, potential habitat appears to compromise ~30% of the presumed lesser prairie-chicken range in Kansas with most habitat in the Mixed-Grass Prairie Ecoregion. Within occupied sites, populations were most sensitive to factors during the first year of life (chick and juvenile survival), however, the persistence of populations through drought may rely on adult survival. Among regional populations, breeding season, nest, and nonbreeding season survival rates contributed most to differences in λ among sites, breeding season survival contributed to differences in λ among more and less fragmented sites. During critical life-stages, diets were comprised of arthropod and plant foods. Among 80 readable fecal samples, 35% of the sequences were likely from Lepidoptera, 26% from Orthoptera, 14% from Araneae, and 13% from Hemiptera. Plant sequences from 137 fecal samples were comprised of genera similar to Ambrosia (27%) Latuca or Taraxacum (10%), Medicago (6%), and Triticum (5%). Among cover types, lesser prairie-chickens using native grasslands consumed a greater diversity of foods. Last, promising conservation options include the conversion of cropland to grassland through the Conservation Reserve Program (CRP) and tree removal in mixed-grass prairie landscapes. Lesser prairie-chickens mostly used CRP during nesting and the nonbreeding season, during drier periods, and in drier portions of their distribution. Strategic CRP sign-up and tree removal could recover >60,000 ha and~100,000 ha of habitat respectively. In summary, conservation that targets management in areas within broad scale habitat constraints predicted will be most beneficial. In areas occupied by lesser prairie-chickens, management that increases brood survival in large grasslands having optimal nesting structure will elicit the strongest influence on population growth and will likely be the most resilient to stochastic drought-related effects.