Novel ecosystems emerge from alterations to historic abiotic regimes and contain new species combinations. Everglades canals offer an opportunity to understand the function of novel habitat for native and non-native fishes and how novel conditions in turn influence distribution, abundance and assembly patterns. I examined native and non-native fish assemblages collected across a gradient in novelty, defined by the loss of wetland connectivity and habitat complexity. As novelty increased, native species richness and abundance strongly declined, and the contribution of non-natives increased. Community structure vastly differed among canals and was strongly influenced by spatial factors and secondarily by hydrological factors. Natives and non-natives had opposing responses to key hydrologic and habitat parameters. This study represents the first comprehensive assessment of Everglades canal fishes, providing insight into the factors influencing native and non-native abundance and assembly patterns and contributing to our understanding of this novel but permanent habitat.

The main objective of this study was the characterization of surface associated microbial communities in breweries. In addition, the beer-spoiling potential of isolated strains and biofilm samples was investigated. Some studies reported the identity of cultivatable organisms from industrial plants. However, there were no data available about the composition of biofilm communities from these habitats for cultivation-independent techniques. Consequently, the fatty acid methyl esters (FAMEs) analysis, the fluorescence in situ hybridization (FISH) and the construction and investigation of 16S rRNA gene clone libraries were applied to reveal the structure of these communities. All of these methods have different advantages and therefore, they complement each other to get a more reliable picture of the biofilm communities. The cultivation method was included in this study because it enables a verification of results from other studies. Furthermore, the obtained strains are genuine brewery isolates and can be used for physiological tests. Isolates were obtained from seven different sample sites (Chapter 1 and 5). They were identified and affiliated to 25 different genera. Some of these strains were inoculated in beer but none of them was able to grow in it (Chapter 1 and 5). However, these strains can still be harmful for the industry, e.g. if they are able to form biofilms. This aspect was investigated by analyzing the potential of the isolates to produce acyl-homoserine lactones (AHLs) (Chapter 6). These quorum sensing mediating molecules are involved in the maturation process of biofilms. Indeed, some strains were found to secrete these autoinducer molecules, they mainly belonged to the genus Pseudomonas. An abundant proportion among the isolates was constituted by members of the Enterobacteriaceae (Chapter 7). In the beginning of this study, there was a minor suspicion concerning their beer-spoiling potential. Indeed, all isolated Enterobacteriaceae were found to be able to multiply in non-alcoholic beer under access of oxygen but they represented no risk for filled beer. The beer-spoiling potential of biofilm communities was investigated by inoculating them in beer (Chapter 3). These enrichments allowed the detection of minor proportions of beer-spoiling organisms. About 25% of the biofilms contained microorganisms which were able to multiply in beer with 4.8% of ethanol (v/v). The absence of anaerobic beer-spoiling bacteria in most of the biofilms was confirmed by using specific FISH probes for Pectinatus and Megasphaera cells (Chapter 9). However, Pectinatus cells constituted one of the most abundant groups in two biofilm communities. These samples clearly demonstrated that brewery biofilms can become hazardous for the quality of the product. The acetic acid bacteria were supposed to be abundant brewery biofilm organisms. This was not confirmed by any method used (Chapter 8). Instead, FISH signals were found for many other taxa in considerable proportions, e.g. communities from the conveyors consisted of members of the Eukarya, Archaea, Alpha-, Beta-, Gammaproteobacteria, Cytophaga-Flavobacteria, Planctomycetales, Actinobacteria and Firmicutes (Chapter 1). Such diverse communities were also evidenced for three other biofilms analyzed by FISH (Chapter 2 and 9). Whereas the FISH technique allows the specific detection of single cells, the FAME analysis targets all organisms present, except the Archaea. The fatty acid profiles of 78 biofilms indicated significant differences between the communities, even between those which were exposed to similar conditions. In addition, repeated sampling of identical sites revealed a temporal variability of the microbial communities (Chapter 3). Characteristical fatty acids of beer-spoiling bacteria were almost absent. Typical fatty acids of Eukarya dominated nearly half of all biofilms. The high proportions of Eukarya in some biofilms was not confirmed, as these samples were also investigated by FISH. This divergence was found to be due to the higher biomass of eukaryotic cells compared to bacterial cells (Chapter 3). As some wild yeast strains were isolated and characterized, they are a potential source of these fatty acids. In contrast to the revealed bacterial diversity, most of the isolated yeasts were assigned to Saccharomyces or Candida spp. (Chapter 4). The Saccharomyces spp. showed a high beer-spoiling potential and many Candida species were able to form biofilms. The construction of 16S rRNA gene clone libraries and the analysis of the clones with amplified ribosomal DNA restriction analysis (ARDRA) was performed with two biofilm communities (Chapter 2). Clones with identical ARDRA patterns were grouped and some representatives were identified by sequencing. These clone sequences were affiliated to 30 different genera, most of which were members of the Alpha- and Gammaproteobacteria and the Bacteroidetes. In addition, some clone sequences were assigned to uncultured organisms. Despite of the presence of 53 and 59 different ARDRA patterns in the two clone libraries, respectively, they had only four patterns in common. This result underlined the differences in the microbial composition of these communities. In conclusion, breweries represent a habitat with high cleaning and disinfecting pressure, which might have selected for a limited number of more resistant or adopted species. Instead, the community structures of biofilms in industrial environments were found to be diverse and variable in their compositions.

京都大学 / 新制・課程博士 / 博士(農学) / 甲第24111号 / 農博第2516号 / 新制||農||1093(附属図書館) / 学位論文||R4||N5402(農学部図書室) / 京都大学大学院農学研究科森林科学専攻 / (主査)教授 北島 薫, 教授 井鷺 裕司, 准教授 東樹 宏和 / 学位規則第4条第1項該当 / Doctor of Agricultural Science / Kyoto University / DGAM

This thesis is divided into two distinct chapters. In the first chapter, we apply network representation learning to the field of materials science in order to predict aluminum grain boundaries' properties and locate the most influential atoms and subgraphs within each grain boundary. We create fixed-length representations of the aluminum grain boundaries that successfully capture grain boundary structure and allow us to accurately predict grain boundary energy. We do this through two distinct methods. The first method we use is a graph convolutional neural network, a semi-supervised deep learning algorithm, and the second method is graph2vec, an unsupervised representation learning algorithm. The second chapter presents our dynamic global value chain network, the combination of the dynamic global supply chain network and the dynamic global strategic alliance network. Our global value chain network provides a level of scope and accessibility not found in any other global value chain network, commercial or academic. Through applications of network theory, we discover business applications that would increase the robustness and resilience of the global value chain. We accomplish this through an analysis of the static, dynamic, and community structure of our global value chain network.

Phylogenetic Community Structure (PCS) metrics are becoming more common in community ecology. PCS metrics estimate the phylogenetic relatedness among members of an ecological community or assemblage. If ecological traits are conserved, then phylogenetic clustering (i.e., taxa are more closely related than expected by chance) indicates habitat filtering as the key process in community assembly. On the other hand, a pattern of phylogenetic overdispersion (i.e., taxa are more distantly related than expected by chance) suggests competition is dominant. Most studies to date have used PCS of unmanipulated ecosystems, but the value of PCS metrics will be best revealed in experiments. This project used PCS for aquatic beetle (Coleoptera) assemblages in experimentally manipulated seasonal wetlands on a cattle ranch in south-central Florida, and compared PCS metrics to standard ecological metrics. Wetlands were experimentally treated with all combinations of pasture management, fencing to exclude cattle, and controlled burning during 2006-2009. Beetle assemblages in fenced wetlands were significantly more overdispersed compared to non-fenced wetlands, suggesting that this treatment decreases habitat filtering, causing competition to become the dominant process in community formation. There was also a significant pasture x fence x burn interaction effect, with assemblages in wetlands differing in PCS depending on what combination of the three treatments were applied. Phylogenetic Diversity (PD – a measure of branch length of a community or assemblage on a phylogenetic tree) was highly correlated with genera richness (number of genera), and these metrics along with the expected number of genera (D – an ecological diversity index) found significant differences among burn treatments and a pasture x iii burn interaction. The results of this study indicate that PCS metrics complement classical ecological methods and should be widely applied.

Moore and Shannon's reliability polynomial can be used as a global statistic to explore the behavior of diffusive processes on a graph dynamical system representing a finite sized interacting system. It depends on both the network topology and the dynamics of the process and gives the probability that the system has a particular desired property. Due to the complexity involved in evaluating the exact network reliability, the problem has been classified as a NP-hard problem. The estimation of the reliability polynomials for large graphs is feasible using Monte Carlo simulations. However, the number of samples required for an accurate estimate increases with system size. Instead, an adaptive method using Bernstein polynomials as kernel density estimators proves useful. Network reliability has a wide range of applications ranging from epidemiology to statistical physics, depending on the description of the functionality. For example, it serves as a measure to study the sensitivity of the outbreak of an infectious disease on a network to the structure of the network. It can also be used to identify important dynamics-induced contagion clusters in international food trade networks. Further, it is analogous to the partition function of the Ising model which provides insights to the interpolation between the low and high temperature limits. / Ph. D. / The research presented here explores the effects of the structural properties of an interacting system on the outcomes of a diffusive process using Moore-Shannon network reliability. The network reliability is a finite degree polynomial which provides the probability of observing a certain configuration for a diffusive process on networks. Examples of such processes analyzed here are outbreak of an epidemic in a population, spread of an invasive species through international trade of commodities and spread of a perturbation in a physical system with discrete magnetic spins. Network reliability is a novel tool which can be used to compare the efficiency of network models with the observed data, to find important components of the system as well as to estimate the functions of thermodynamic state variables.

No description available.

No description available.

El uso de las redes para modelar sistemas complejos es creciente en multitud de ambitos. Son extremadamente utiles para representar interacciones entre genes, relaciones sociales, intercambio de informaci on en Internet o correlaciones entre precios de acciones burs atiles, por nombrar s olo algunos ejemplos. Analizando la estructura de estas redes, comprendiendo c omo interaccionan sus distintos elementos, podremos entender mejor c omo se comporta el sistema en su conjunto. A menudo, los nodos que conforman estas redes tienden a formar grupos altamente conectados. Esta propiedad es conocida como estructura de comunidades y esta tesis doctoral se ha centrado en el problema de c omo mejorar su detecci on y caracterizaci on. Como primer objetivo de este trabajo, se encuentra la generaci on de m etodos e cientes que permitan caracterizar las comunidades de una red y comprender su estructura. Segundo, pretendemos plantear una serie de pruebas donde testar dichos m etodos. Por ultimo, sugeriremos una medida estad stica que pretende ser capaz de evaluar correctamente la calidad de la estructura de comunidades de una red. Para llevar a cabo dichos objetivos, en primer lugar, se generan una serie de algoritmos capaces de transformar una red en un arbol jer arquico y, a partir de ah , determinar las comunidades que aparecen en ella. Por otro lado, se ha dise~nado un nuevo tipo de benchmarks para testar estos y otros algoritmos de detecci on de comunidades de forma e ciente. Por ultimo, y como parte m as importante de este trabajo, se demuestra que la estructura de comunidades de una red puede ser correctamente evaluada utilizando una medida basada en una distribuci on hipergeom etrica. Por tanto, la maximizaci on de este ndice, llamado Surprise, aparece como la estrategia id onea para obtener la partici on en comunidades optima de una red. Surprise ha mostrado un comportamiento excelente en todos los casos analizados, superando cualitativamente a cualquier otro m etodo anterior. De esta manera, aparece como la mejor medida propuesta para este n y los datos sugieren que podr a ser una estrategia optima para determinar la calidad de la estructura de comunidades en redes complejas. / Aldecoa García, R. (2013). Detección de comunidades en redes complejas [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/31638 / Premios Extraordinarios de tesis doctorales

Résumé: Deux facteurs principaux, une fois combinés, permettent de comprendre l’assemblage des communautés, soient i) l’environnement (abiotique et biotique), qui agit comme un filtre sélectionnant les espèces les mieux adaptées aux conditions données, et ii) les traits fonctionnels, sur lesquels s’effectue ce filtrage environnemental puisqu’ils représentent les adaptations des espèces aux conditions données. Il est donc essentiel d’établir des relations fiables entre les conditions environnementales et la structure fonctionnelle des communautés afin de pouvoir identifier et comprendre les mécanismes régissant l’assemblage des communautés. Cependant, plusieurs facteurs tels que les interactions entre variables environnementales à différentes échelles spatiales (par exemple le climat et la fertilité des sols) peuvent complexifier la situation et c’est pourquoi, malgré une quantité grandissante d’études sur le sujet, les processus d’assemblage des communautés restent difficiles à définir et à généraliser. Cette thèse vise donc à i) mieux définir et quantifier les relations trait-environnement des systèmes d’herbacées au travers de différentes échelles spatiales et ii) déterminer l’influence de ces relations sur l’assemblage des communautés et le fonctionnement des écosystèmes. Pour réaliser ces objectifs, j’ai travaillé avec les données de programmes ayant collecté des données taxonomiques et fonctionnelles sur les communautés d’herbacées à travers la France (DivHerbe et DivGrass) et, à moindre mesure, l’Europe (VISTA). Ces bases de données couvrent donc de larges gradients climatiques régionaux ainsi que des gradients environnementaux plus locaux relatifs à la qualité des sols et aux perturbations. J’ai d’abord testé l’importance de considérer à la fois des variables environnementales locales et régionales ainsi que leurs interactions pour déterminer la structure fonctionnelle et taxonomique des communautés. J’ai ensuite étudié comment l’importance relative des processus menant soit à la convergence ou à la divergence fonctionnelle peut changer le long d’une succession, puis comment ces deux types de processus influencent notre capacité à prédire l’assemblage des communautés à partir des traits fonctionnels. Finalement, j’ai présenté comment les résultats au niveau des communautés peuvent être utiles pour étudier le niveau des écosystèmes. Les résultats de cette thèse démontrent que les variables climatiques régionales interagissent fortement avec les variables environnementales locales pour influencer les processus locaux déterminant l’assemblage des communautés. Évaluer le contexte régional semble donc nécessaire afin d’éviter des interprétations erronées des patrons d’assemblage observés. Travaillant avec ces deux niveaux de variation environnementale, une dissociation important entre la variation taxonomique et fonctionnelle des communautés a été mise à jour, reflétant l’importance de considérer plusieurs facettes de biodiversité pour comprendre la dynamique des communautés. Les résultats ont aussi démontré que les processus d’assemblage menant à la convergence et à la divergence ont un impact très différent et prédictible sur les relations liant les traits et les abondances des espèces, c’est-à-dire notre capacité à prédire l’assemblage des communautés à partir des traits. Finalement, toutes ces notions, relatives aux relations trait-environnement et à l’assemblage des communautés basé sur les traits, ont été utilisées dans un contexte de biogéographie fonctionnelle. Il a été démontré qu’il était possible de construire des cartes de valeurs de traits fonctionnels dans les prairies permanentes à l’échelle de la France, à partir de variables environnementales. Certaines propriétés écosystémiques ont ensuite pu être prédites à partir de ces cartes. Ces travaux ont donc permis d’illustrer les défis à surmonter pour utiliser nos connaissances de l’écologie fonctionnelle en vue d’une conservation et d’une exploitation viables de nos écosystèmes. / Abstract: There are two main factors which, combined together, allow understanding community assembly : i) the environment (both abiotic and biotic), which acts as a filter selecting species according to how well-adapted they are to given conditions, and ii) functional traits, on which this environmental filtering occurs since they represent species adaptations to particular conditions. It is thus essential to establish reliable relationships between environmental conditions and the functional structure of communities in order to identify and understand the mechanisms driving community assembly. However, several factors such as cross-scale interactions between environmental variables (e.g. between climate and soil fertility) complicate the situation. This is why, despite a growing body of studies on the subject, processes of community assembly are still poorly understood and are difficult to generalize. The purpose of this thesis is to i) better define and quantify the trait-environment relationships in herbaceous systems across different spatial scales and ii) determine the influence of those relationships on community assembly and on ecosystem functioning. To realize these objectives, I worked with data from programs which assembled taxonomic and functional data on herbaceous communities across France (DivHerbe and DivGrass) and, to a lesser extent, Europe (VISTA). These databases thus cover large regional climatic gradients, as well as more local environmental gradients related to soil quality and disturbances. I first tested the importance of simultaneously considering local and regional environmental variables as well as their interactions to determine the taxonomic and functional structure of communities. Then, I studied how the relative importance of processes leading to either functional convergence or divergence can change along a successional gradient, and how these two types of processes influence our ability to predict community assembly from functional traits. Finally, I presented how the results at the community-level can be used to study the ecosystem-level. The results of this thesis demonstrate that regional climatic variables strongly interact with local environmental variables in driving the local processes responsible for community assembly. Assessing the regional context is thus necessary in order to avoid erroneous interpretations of observed assembly patterns. Working with those two levels of environmental variation, important discrepancies were found between taxonomic and functional variations across communities, reflecting the importance of considering several aspects of biodiversity in order to understand community dynamics. The results also demonstrated that the assembly processes leading to functional convergence and divergence have a very different and predictable impact on the relationships between traits and species relative abundances, i.e. on our ability to predict community assembly from traits. Finally, these notions related to trait-environment relationships and to trait-based community assembly were used in a functional biogeography framework: It was possible to build maps of functional traits values in permanent grasslands across France using environmental variables. These maps then allowed predicting particular ecosystem properties. Thus, this work allowed illustrating some challenges that we are facing in using our knowledge in functional ecology to build sustainable conservation and exploitation plans for our ecosystems.