Vegetation has been identified to play a significant role in river environments by providing a wide range of ecosystem services. For this reason, the use of plants has become relevant in river restoration projects. However, the presence of plants in channel beds increases the flow resistance and, thus, the water levels during flood conditions. Additionally, river vegetation, whether instream or riparian, influences the morphological evolution of rivers. Observations show that instream vegetation has a strong impact on bedload transport. Yet, there is a scarcity of sediment transport predictors that directly account for the effects of plants, and existing methods, based on re-calculation of roughness coefficients, may present some inconsistencies. Therefore, an approach that extends Einstein’s (1950) parameters to include the effects of vegetation geometry and spatial density on sediment transport is herein proposed. The new formulations of the dimensionless transport parameter Φ and the flow intensity parameter Ψ were derived for their implementation in existing bedload predictors of the form Φ = (Ψ). The applicability of this new approach considers the presence of submerged and emergent vegetation, but reduces to the original Einstein’s model if vegetation is absent. The research methodology was carried out in four phases. First, a comprehensive literature review for the identification of, mainly, the different effects of vegetation on river morphodynamics, the state-of-the-art knowledge on the flow-sediment-vegetation interactions, and the current approaches to bedload estimation in channels with vegetated beds. Second, the derivation of the extended Einstein’s parameters, starting from a momentum balance for a control volume of a generic channel with instream submerged vegetation (as proposed by Petryk and Bosmajian, 1975). Third, an extensive experimental program carried out on a tilting flume with a mobile bed and with plants being represented by series of aluminum cylinders. Different scenarios of vegetation spatial density were tested while measurements of bedload rate, water level, bed level and flow velocity were periodically performed in order to assess conditions of stationarity and morphodynamic equilibrium. Last, a deep analysis of experimental results allowed for the calibration of the new approach, whereas external datasets from the literature were used to assess its performance in a wide variety of conditions. A study based on four statistical measures showed that the extended Einstein’s parameters are significantly more suitable for bedload rate estimation when compared to the original ones, since predicted and measured values have, on average, the same order of magnitude. Additionally, the new approach outperformed the widely-adopted method of Baptist (2005), which consists of the re-calculation of bed roughness in vegetated settings. Finally, the experimental observations suggest that the submergence ratio and the stem spatial density are the most important traits of river plants to display influence on bedload transport, channel bed stability, and bed form dimensions and patterns. A better understanding of these traits might lead to better prediction capabilities of river evolution. / La vegetazione svolge un ruolo fondamentale negli ambienti fluviali, poiché fornisce un ampio spettro di servizi ecosistemici; per questo essa è una componente rilevante dei progetti di riqualificazione fluviale. Tuttavia, la presenza di piante in alveo aumenta la resistenza al moto e di conseguenza anche il tirante idrico durante gli eventi di piena. Inoltre, la copertura vegetale in alveo e nelle zone riparie influenza l'evoluzione morfologica dei corsi d'acqua. Nonostante le evidenze sperimentali mostrino che la vegetazione in alveo ha un forte impatto sul trasporto dei sedimenti, sono poche le formule di trasporto che tengono conto in modo esplicito dell'effetto della vegetazione e i metodi esistenti, basati sulla determinazione di un coefficiente di scabrezza, possono dare luogo a incongruenze. Per questa ragione, in questa tesi si propone un approccio che estende la formulazione di Einstein (1950) e include l'effetto della geometria e della densità spaziale della vegetazione sul trasporto solido. Sono state derivate nuove espressioni per il parametro di trasporto adimensionale Φ e il parametro di intensità del trasporto Ψ, che possono essere introdotte in modelli di trasporto esistenti del tipo Φ = f(Ψ). Questo nuovo approccio consente di considerare l'effetto della presenza di vegetazione sommersa ed emergente e si riduce al modello originale di Einstein in assenza di vegetazione. L'attività di ricerca si è svolta in quattro fasi. Nella prima fase si è svolta un'analisi approfondita della letteratura mirata soprattutto a identificare gli effetti della vegetazione sulla morfodinamica fluviale, definire lo stato dell'arte relativo alle interazioni fra flusso liquido, sedimenti e vegetazione, ed analizzare gli approcci esistenti per la stima del trasporto di fondo in alvei vegetati. Nella seconda fase si sono derivati i parametri della formulazione di Einstein estesa a partire dal bilancio di quantità di moto per un volume di controllo di un canale generico con vegetazione sommersa (come proposto da Petryk e Bosmajian, 1975). Nella terza fase è stato condotto un esteso set di esperimenti, utilizzando un modello fisico costituito da una canaletta di laboratorio a pendenza variabile e fondo mobile, in cui le piante sono state simulate tramite cilindri in alluminio. Sono stati riprodotti diversi scenari di densità spaziale della vegetazione e sono stati misurati periodicamente la portata solida, la quota della superficie libera e del fondo e la velocità della corrente per valutare le condizioni di stazionarietà ed equilibrio morfodinamico. Infine, il nuovo approccio è stato calibrato sulla base di un'analisi approfondita dei risultati sperimentali e quindi applicato a set di dati di letteratura per valutarne l'accuratezza in un ampio intervallo di condizioni. Un'analisi statistica basata su quattro indicatori ha mostrato che i parametri della formulazione di Einstein estesa producono stime di trasporto solido sensibilmente più accurate rispetto ai parametri originali, in quanto i valori calcolati sono, in generale, dello stesso ordine di grandezza dei valori misurati. Inoltre, il nuovo approccio dà risultati migliori rispetto al metodo di Baptist (2005), ampiamente adottato, che consiste nel ricalcolo della scabrezza per gli alvei vegetati. Infine, le osservazioni sperimentali suggeriscono che il rapporto di sommergenza e la densità spaziale delle piante sono i parametri che influenzano in modo più significativo il trasporto solido, la stabilità del fondo dell'alveo, la scala delle forme di fondo e la loro organizzazione spaziale. Una conoscenza più approfondita di questi aspetti può contribuire a una maggiore capacità di prevedere l'evoluzione dei corsi d'acqua. / Se ha identificado a la vegetación como un actor importante en ambientes fluviales al proporcionar una amplia gama de servicios ecosistémicos. Por esta razón, el uso de plantas se ha vuelto cada vez más relevante en proyectos de restauración de ríos. Sin embargo, su presencia en lechos fluviales impacta la resistencia al flujo, aumentando los niveles del agua en condiciones de inundación. Además, este tipo de vegetación, ya sea que esté en el lecho o en las márgenes, influye en la evolución morfológica de los ríos. Diversas observaciones han mostrado que la vegetación fluvial tiene un fuerte impacto en las tasas de transporte sólido de fondo. A pesar de ello, hay una escasez de métodos confiables para la estimación de este tipo de sedimentos que tome en consideración el efecto de las plantas y, aquéllos que existen, los cuales se basan en la corrección del coeficiente de rugosidad del canal, suelen presentar resultados inconsistentes. Por tanto, se propone aquí un método que extiende las definiciones fundamentales de Einstein (1950) en modo que se incluyan los efectos de la geometría y la densidad espacial de las plantas sobre el transporte sólido. Las nuevas ecuaciones del parámtero de transporte, Φ, y el parámetro de movilidad, Ψ, fueron obtenidas para su implementación en métodos predictores de transporte de fondo de la forma Φ = (Ψ). La aplicabilidad de este nuevo enfoque considera la posibilidad de vegtación fluvial tanto emergente como sumergida, y se reduce a las ecuaciones originales de Einstein si ésta fuera inexistente. La metodología de investigación se llevó a cabo en cuatro fases. Primero, una revisión exhaustiva de la literatura para la identificación, principalmente, de los diferentes efectos de la vegetación en la morfodinámica de ríos, los avances más recientes en el conocimiento sobre las interacciones flujo-sedimento-vegetación, y los métodos actualmente existentes para la estimación del transporte sólido de fondo en canales naturales vegetados. En segundo lugar, la obtención de los parámetros de Einstein extendidos a partir de un balance de momentum para el volumen de control de un canal genérico con vegetación sumergida (según lo propuesto por Petryk y Bosmajian, 1975). En tercer lugar, un extenso programa experimental realizado en un canal de fondo móvil y pendiente variable, con las plantas siendo representadas por series de cilindros metálicos. Se probaron diferentes escenarios de densidad espacial de vegetación, mientras que periódicamente se realizaron mediciones transporte sólido, niveles del agua, topografía del fondo y velocidad del flujo con el objeto de evaluar las condiciones de flujo uniforme y equilibrio morfodinámico. Por último, un análisis profundo de los resultados experimentales permitió la calibración del nuevo método, mientras que se utilizaron datos externos disponibles en la literatura para evaluar su desempeño bajo diversas condiciones. Un estudio basado en cuatro medidas estadísticas mostró que los parámetros extendidos de Einstein son mucho más adecuados para la estimación del transporte de fondo en comparación con los originales, ya que los valores estimados y los medidos muestran, en promedio, el mismo orden de magnitud. Además, el nuevo método superó al propuesto por Baptist (2005), ampliamente utilizado, el cual consiste en la corrección de la rugosidad del canal en presencia de vegetación. Finalmente, las observaciones experimentales sugieren que la sumergencia de las plantas y la densidad espacial de los tallos son las variables más influyentes en el transporte sedimentos de fondo, la estabilidad del lecho, y las dimensiones y patrones de la forma de fondo. Una mejor comprensión de estas variables puede significar una mejor capacidad para predecir la evolución de un río.

Este trabalho teve como objetivo analisar o transporte de sedimentos na região central da Baixada Santista (SP) através do uso do modelo numérico ROMS (Regional Ocean Modeling System) acoplado ao modelo numérico SWAN (Simulating WAves Nearshore) em simulações representativas dos meses de agosto de 2009 e janeiro de 2010. Os resultados indicaram haver picos no transporte de sedimentos, principalmente na plataforma continental e na Baía de Santos, durante a passagem de sistemas frontais, chegando a atingir 4,2 kg m2 s1 na região costeira. As menores taxas de concentração de sedimentos em suspensão na coluna d\'água ocorreram durante os períodos de quadratura, sem a presença de ventos fortes e incidência de ondas com grande energia. O transporte por carga de fundo encontrado na área costeira de Praia Grande foi predominantemente para sudoeste, em ambos os meses simulados, onde, as maiores taxas de transporte foram observadas no mês de agosto com valores em torno de 1 x 10 kg m2 para o mês. A taxa de sedimentação na região do alto estuário foi da ordem de 1 cm ano1, concordando com os dados disponíveis na literatura. O modelo numérico representou de forma satisfatória os padrões de transporte de sedimentos encontrados na região. / This study aimed to analyze the sediment transport in the central region of Baixada Santista (SP) using the numerical model ROMS (Regional Ocean Modeling System) coupled to the numerical model SWAN (Simulating Waves Nearshore) in simulations representing the months of August 2009 and January 2010. The results suggested peaks in sediment transport, especially on the continental shelf and the Bay of Santos, during the passage of frontal systems, achieving values of 4,2 kg m2 s1 nearshore. The lowest rates of concentration of suspended sediment in the water column occurred during periods of neap tides, without the presence of strong winds and incidence of waves with high energy. In both simulated months, the bedload transport found in the coastal area of Praia Grande was predominantly to the southwest, with the highest rates of transport observed in the month of August, with values of 1 x 10 kg m2. The rate of sedimentation in the upper estuary has been on the order of 1 cm year1, in agreement with the data available in the literature. The numerical model represented the transport patterns of sediments found in the region in a satisfactory way.

Este trabalho teve como objetivo analisar o transporte de sedimentos na região central da Baixada Santista (SP) através do uso do modelo numérico ROMS (Regional Ocean Modeling System) acoplado ao modelo numérico SWAN (Simulating WAves Nearshore) em simulações representativas dos meses de agosto de 2009 e janeiro de 2010. Os resultados indicaram haver picos no transporte de sedimentos, principalmente na plataforma continental e na Baía de Santos, durante a passagem de sistemas frontais, chegando a atingir 4,2 kg m2 s1 na região costeira. As menores taxas de concentração de sedimentos em suspensão na coluna d\'água ocorreram durante os períodos de quadratura, sem a presença de ventos fortes e incidência de ondas com grande energia. O transporte por carga de fundo encontrado na área costeira de Praia Grande foi predominantemente para sudoeste, em ambos os meses simulados, onde, as maiores taxas de transporte foram observadas no mês de agosto com valores em torno de 1 x 10 kg m2 para o mês. A taxa de sedimentação na região do alto estuário foi da ordem de 1 cm ano1, concordando com os dados disponíveis na literatura. O modelo numérico representou de forma satisfatória os padrões de transporte de sedimentos encontrados na região. / This study aimed to analyze the sediment transport in the central region of Baixada Santista (SP) using the numerical model ROMS (Regional Ocean Modeling System) coupled to the numerical model SWAN (Simulating Waves Nearshore) in simulations representing the months of August 2009 and January 2010. The results suggested peaks in sediment transport, especially on the continental shelf and the Bay of Santos, during the passage of frontal systems, achieving values of 4,2 kg m2 s1 nearshore. The lowest rates of concentration of suspended sediment in the water column occurred during periods of neap tides, without the presence of strong winds and incidence of waves with high energy. In both simulated months, the bedload transport found in the coastal area of Praia Grande was predominantly to the southwest, with the highest rates of transport observed in the month of August, with values of 1 x 10 kg m2. The rate of sedimentation in the upper estuary has been on the order of 1 cm year1, in agreement with the data available in the literature. The numerical model represented the transport patterns of sediments found in the region in a satisfactory way.

Les mosaïques de paysages dessinées par les cours d’eau sont le fruit de fonctionnements complexes. Ces fonctionnements sont appréhendés à divers échelles spatio-temporelles (e.g, bassin versant, secteurs fonctionnels, macroformes alluviales ou particules sédimentaires) décrites au sein du concept de système fluvial. Le fonctionnement des systèmes fluviaux s’établissent autour d’équilibres sédimentaires mettant en balance les flux liquides et solides. C’est pourquoi, l’état géomorphologique des cours d’eau à charge sédimentaire graveleuse peut être décrit au travers d’un prisme sédimentaire par l’observation de tendances générales à large échelles, de phénomènes plus ou moins localisés tels que l’incision, l’engravement, la migration des formes sédimentaires ou des modifications des propriétés géométriques des particules constituant la charge sédimentaire. Ainsi, la meilleure connaissance du rôle des flux sédimentaires actuels et passés dans l’état géomorphologique pour la prédiction des trajectoires hydrosédimentaires des cours, répond à des objectifs à la fois écologiques (la perméabilité de la nappe d’alluvions et de la zone hyporhéique exercent des fonctions auto-épuratrices (Datry et al., 2008), la faune benthique est la base de la chaîne alimentaire des écosystèmes fluviaux), politiques (liés à la gestion des risques d’inondation et d’érosion des terres ou d’ouvrages d’ingénierie) ou encore philosophiques (quel fonctionnement de référence pour quel système fluvial ? Quelle implication et quel rôle pour l’Homme dans le fonctionnement de ces systèmes anthropisés à divers degrés ?). Aussi, le développement d’outils et de méthodologie(s) pour la caractérisation des flux sédimentaires ou des processus observés en rivières constitue un axe de recherche important en constante émulsion. Les recherches que nous avons engagés sont articulées autour de trois axes interdépendants : (1) les propriétés physiques des particules sédimentaires, (2) leur mobilité et (3) leur transformation par abrasion. Dans le cadre de ce typique, nous avons cherché à lever le doute subsistant concernant le rôle de la fracturation de traceurs naturels (calcaires) équipés de transpondeurs passifs basses fréquences (PIT-tag) et à tester l’intérêt de l’utilisation de galets synthétiques pour la conception de nouveaux traceurs équipés de transpondeurs actifs ultra hautes fréquences (a-UHF). Nous avons ensuite testé le potentiel de ces transpondeurs (a-UHF), dont les caractéristiques d’émission sont mieux adaptées que celles des PIT-tag à l’étude du fonctionnement des macroformes fluviale à l’échelle de tronçons fonctionnels, puis avons développé une méthodologie de recherche en rivière spécifique à ces nouveaux traceurs. Par ailleurs, Nous avons testé la sensibilité à divers paramètres d’entrée d’un programme de calcul d’émoussé de particules sédimentaires grossières, en avons validé son intérêt pour l’analyse (1) des relations entre évolution des indices d’émoussé et abrasion de la masse sédimentaire et (2) du fonctionnement hydrosédimentaire d’un bassin versant et avons formulé les recommandations nécessaires de respecter pour sa bonne utilisation. Enfin, nous avons cherché à comparer les influences respectives de la densité et de la forme des particules sédimentaires pendant leur transport par charriage. Ces résultats soulignent l’influence au moins aussi grande de la forme comparée à la mobilité, sur le déplacement des particules sédimentaires. / The influence of fluvial systems dynamics on riverine landscape patterns is the consequence of complex ecosystem dynamics. These driving factors can be analysed at different spatiotemporal scales of the fluvial system (for example: river basin, functional units, alluvial megaforms or sedimentary particles). The dynamic of fluvial systems depends on the equilibrium between solid and liquid fluids. For this reason, the geomorphological status of watercourses can be described from a sedimentary perspective taking into account general trends at large scale, local characteristics such as incision, aggradation, sedimentary migration or the alteration of particles’ geometrical properties that constitute bed load. Thus, a better understanding of the influence of past and present sedimentary flows on the geomorphological status of watercourses has at the same time an ecological impact (permeability of both the alluvial layer and the hyporheic zone and its self-filtration function (Datry et al., 2008)), a political impact (related to flood management, erosion risks and issues or hydraulic facilities) or even a philosophical issue (what is the appropriate reference for what fluvial system? What is the role of Humanity on these anthropized environments and at what extent?).Thus, the development of methodological tools to characterize sedimentary flows or the observed dynamics on rivers, is considered a major and rising research issue. The research studies described in this thesis are structured in three distinct topics: (1) physical properties of sedimentary particles; (2) their mobility; (3) their transformation as a consequence of abrasion.In this context, we have analysed the relationship between breakage and low recovery rates for natural tracers (limestone pebbles) equipped with low frequency passive transponders (PIT-tag). In addition, we have tested the use of synthetic pebbles to create new tracers equipped with high frequency active transponders (a-UHF), more performant than PIT-tags. Secondly, we have developed a new search and location field methodology specially, adapted to these new transponders. Thirdly, we have tested the sensibility of the toolbox developed by Roussillon et al. (2009), design to measure the roundness of coarse sedimentary particles, to different entry parameters. We have confirmed its interest to (1) analyse the relationship between particles normalized abrasion and its roundness trends; (2) analyse the hydrosedimentary dynamics at a river basin scale. We have included several recommendations to apply this toolbox in the appropriate conditions. Finally, we have performed a comparative analysis between particles’ density and shape and its influence on bedrock transport. These results highlight the significant influence of shape on travelled distance in comparison with density.

The plants and animals that inhabit river channels may act as zoogeomorphic agents affecting the nature and rates of sediment recruitment, transport and deposition. The impact of benthic-feeding fish, which disturb bed material sediments during their search for food, has received little attention, even though benthic feeding species are widespread in rivers and may collectively expend significant amounts of energy foraging across the bed. A series of experiments were conducted to investigate the impacts of benthic feeding fish on the structure and composition of gravel-bed river sediments, and the implications for bed material transport. An ex-situ experiment was conducted to investigate the impact of a benthic feeding fish (European Barbel Barbus barbus) on particle displacements, bed sediment structures, gravel entrainment and transport fluxes. In a laboratory flume, changes in bed surface topography were measured and grain displacements examined when an imbricated, water-worked bed of 5.6-16 mm gravels was exposed to feeding juvenile Barbel. For substrates that had been exposed to feeding fish and control substrates which had not, grain entrainment rates and bedload fluxes were measured under a moderate transport regime. On average, approximately 37% of the substrate, by area, was modified by foraging fish during a four-hour treatment period, resulting in increased microtopographic roughness and reduced particle imbrication. Structural changes caused by fish increased bed load flux by 60% under entrainment flows, whilst on average the total number of grains transported during the entrainment phase was 82% higher from substrates that had been disturbed by Barbel. An ex-situ experiment utilising Barbel and Chub Leuciscus cephalus extended this initial study by considering the role of fish size and species as controls of sediment disturbance by foraging. Increasing the size of Barbel had a significant effect on measured disturbance and bedload transport. Specifically, the area of disturbed substrate, foraging depth, microtopographic roughness and sediment structure all increased as functions of fish size, as did bedload flux and total transported mass. In a comparison of the foraging effects of like-sized Barbel and Chub 8-10 in length, Barbel foraged a larger area of the riverbed and had a greater impact on microtopographic roughness and sediment structure. Foraging by both species was associated with increased sediment transport, but the bed load flux after foraging by Barbel was 150% higher than that following foraging by Chub and the total transported mass of sediment was 98% greater. An in-situ experiment quantified the effects of foraging fish, primarily Cyprinids (specifically Barbel and Chub), on gravel-river bed sediment structures, surface grain-size distributions, sediment transport fluxes and grain entrainment in the River Idle, Nottinghamshire, UK. This was achieved by installing large experimental sediment trays seeded with food at typical densities. The experiments yielded data about 1) topographic and structural differences between pre- and post-feeding substrates using DEMs interpolated from laser scans, 2) modifications to surface and sub-surface grain-size distributions as a function of fish foraging and 3) differences in sediment entrainment from water-worked substrates exposed to feeding fish and control substrates, without fish. Small sections of the substrate trays were recovered in tact from the field and for substrates that had been exposed to feeding fish and control substrates which had not, grain entrainment rates and bedload fluxes were measured under a moderate transport regime in the laboratory. On average, approximately 74% of the substrate, by area, was modified by foraging fish during a twelve-hour period, resulting in increased microtopographic roughness and substrate coarsening which had significant implications for bed material transport during the steady entrainment flow. Together, results from these experiments indicate that by increasing surface microtopography, modifying the composition of fluvial substrates and undoing the naturally stable structures produced by water working, foraging can influence sediment transport dynamics, predominately by increasing the mobility of river bed materials. The implication of this result is that by influencing the quantity of available, transportable sediment and entrainment thresholds, benthic feeding may affect sediment transport fluxes in gravel-bed rivers. In addition, three discrete studies were performed alongside the core experiments described above. A quantitative examination of habitat conditions favoured by feeding Barbel was conducted in the River Idle (Nottinghamshire, UK) which served to supplement existing literature pertaining to Barbel ecology, and inform experimental design during the core experiments. Two further studies considered the potential importance of foraging as a zoogeomorphic activity in terms of spatial extent, at a variety of scales, thereby extending core experiments to larger spatial scales in-situ.

The influence of channel morphology and hydraulics on sediment transport within steep upland streams is investigated. Step-pools are the most common bedform in such streams. Bedload transport processes operating in a perennial step-pool stream in central England are established using magnetically-tagged particles and bedload pit samplers. Water-surface slope in two step-pool streams is monitored in detail. Water-surface slope measurements show that step-pool bedforms cause large spatial variations in flow depth during floods, and that the temporal variation of watersurface slope during a flood wave differs diametrically from the pattern found in lowgradient rivers that have little roughness. These results demonstrate that it is difficult to estimate the forces acting on the streambed in steep, rough streams and, therefore, conventionally established flow variables cannot be used to predict bedload transport . rates in step-pool channels. Particle tracing experiments are used to determine the control that step-pool morphology and flow hycfraulics have on particle displacements. Step-pool streams exhibit smaller particle travel distance per flood event than low-gradient rivers. This difference reflects the greater flow resistance and bed rugosity of step-pool streams. Bed morphology dictates flood particle displacement at step-pool bedform scale. It is shown that entrainment, pool export rates and transport distances of coarse sediment are dependent on downstream step dimensions, step permeability and pool form. Large variations in bedload transport rates at any given flow strength are associated with variable sediment supply, and particle tracing demonstrates that sediment availability is controlled at unit pool scale. Bedload consists of fine gravel moving over a stable bed surface at low discharges, while, at higher discharges, bedload coarsens, approaching the size of the pool bed material as the bed surface is partially mobilized. Step framework clasts are stable features of the streambed. Results from this study are incorporated into a descriptive model of discharge-related bedload transport phases in step-pool streams.

Signal crayfish are an internationally widespread invasive species that can have important detrimental ecological impacts. This thesis aims to determine whether signal crayfish have the potential to also impact the physical environment in rivers. A series of experiments were undertaken in purpose-built still-water aquaria using a laser scanner to obtain Digital Elevation Models (DEMs) of narrowly-graded gravel surfaces before and after exposure to crayfish. The difference between DEMs was used to quantify volumetric changes in surface topography due to crayfish activity. Two distinct types of topographic change were identified. The first was the construction of pits and mounds which resulted in an increase in surface roughness and grain exposure. The second was the rearrangement of surface material caused by crayfish brushing past grains when walking and foraging, reorientating grains and altering friction angles. A series of 80 flume runs were undertaken to quantify alterations made by crayfish to water-worked, as well as loose, gravel substrates at low velocity flows. Crayfish significantly altered the structure of water-worked substrates, reversing the imbrication of surface grains to a more random arrangement. Surfaces were entrained at a relatively high velocity flow subsequent to crayfish activity in order to directly link topographic and structural alterations to substrate stability. Nearly twice as many grains were mobilised from surfaces which had been disturbed by crayfish in comparison to control surfaces that were not exposed to crayfish. A field investigation aimed to determine the potential significance of the geomorphic impact of crayfish in rivers. Signal crayfish were tracked through a 20 m reach of a small, lowland alluvial river for 150 days using a Passive Integrated Transponder (PIT) system. Crayfish were active throughout the channel, although their activity became limited as water temperature dropped and flow stage increased. Substrate was not an important determinant of crayfish activity at this scale. Instead, crayfish tended to be found along the inner bank of a meander bend where there was a substantial cover of macrophytes. Consequently, signal crayfish were active for extended periods on substrates of a similar size to those that they could disturb in flume experiments. These results suggest that signal crayfish could have important geomorphic effects in rivers, disturbing bed structures and increasing the mobility of coarse material. This may have important implications for both the management of some rivers and benthic organisms that reside on the river bed.

Cette thèse porte sur l’analyse morphodynamique et sédimentaire des méandres du Cher (affluent de rive gauche de la Loire moyenne). L’intérêt pour cette question est né d’un double constat. Le premier était celui d’un manque patent de connaissances sur le fonctionnement géomorphologique des affluents de la Loire en général et du Cher en particulier. Le second se rapportait à la coexistence en apparence contradictoire d’une faible mobilité en plan depuis le début du XIXe siècle et d’une érosion latérale active au cours des trois dernières décennies, au minimum. L’objectif était à la fois de combler une partie de ce déficit de connaissances, afin notamment d’améliorer les pratiques de gestion de la rivière, et de caractériser le plus finement possible le comportement a priori atypique de ces systèmes à méandres, afin d’en comprendre la dynamique. Cette étude aura permis de mettre en évidence que depuis 1950, mais aussi probablement depuis 1830, la faible mobilité du tracé de la rivière relève essentiellement des contraintes exercées par les aménagements présents dans le lit mineur. En outre, ces méandres disposent encore aujourd’hui d’une très forte capacité intrinsèque à balayer leur plaine alluviale. La fréquence de mobilisation de la charge de fond est également élevée. Enfin, la morphogénèse est principalement commandée par les évènements hydrologiques de faible magnitude. Ces résultats sont de première importance pour les gestionnaires puisqu’ils indiquent que les dysfonctionnements qui affectent actuellement cet hydrosystème ne sont pas totalement irréversibles. / This thesis is dedicated to the morphodynamics and sedimentary analysis of the Cher River meanders (left bank tributary of the Middle Loire River). Two main points explain the present study. First, there is a general lack of knowledge concerning the geomorphological functioning of the Loire River tributaries, especially on the Cher River. Second, there is an apparent contradiction between the low mobility of the fluvial land forms since the early 19th century and active lateral erosion observed over the past three decades at least. In comparison with the Loire and its tributaries, the specificity of the Cher River lies in a meandering pattern that is both old and spatially extended. This research also directly concerns the current management issues of the Cher and the Loire River. These issues relate mainly to the maintenance or restoration of alluvial landscapes and to the refill of the bed with coarse sediments. The main objective of the thesis was both to fill a part of this knowledge gap and to characterize as fully as possible the atypical behavior of this meandering system in order to understand its dynamics. The various investigations conducted have focused on the middle valley of the Cher River, upstream of the channelized section, and more specifically on three distinct reaches. They were designed to validate two explanatory hypotheses that are not mutually exclusive of each other; the two hypotheses were initially set out to explain the stability of the meander planforms. The first hypothesis postulated that the migration of meanders was inhibited by the presence of bank protections. The second is related to a decrease in the frequency and / or intensity of hydrological morphogenic events since the 19th century. Our analysis is based on two complementary approaches. Firstly, we undertook the reconstruction of the evolutionary trajectory of these meanders at a historical timescale (period 1830-2005) and we investigated the causes of the observed changes. Secondly, we examined more specifically the relationships between morphogenesis and bedload transport at time scalesspanning from the hydrological event to a few decades (last 25-70 years). This study demonstrates that since 1950, but probably also since 1830, the low meander mobility, essentially results from engineering works in the bed. Further, the meanders still have a very high intrinsic ability to erode the alluvial plain. The frequency of bedload mobilization is also high. Finally, morphogenesis in the system is mainly controlled by low magnitude hydrological events. These results are of primary importance for river managers because they highlight the reversibility of the current hydrosystem degradation.

Cette thèse expérimentale traite du transport sédimentaire provoqué par un fluide (généralement l’air ou l’eau) s’écoulant sur un matériau érodable (lit de sable par exemple) et lui transmettant une force suffisante pour soulever les particules solides et les emporter avec lui en aval. Ici, on se focalise uniquement sur le charriage aquatique : dans ce mode de transport, les grains en mouvement roulent les uns sur les autres, glissent, effectuent une succession de bonds et d’arrêts tout en restant confinés près du fond. On réalise différents types d’expériences dans un canal de laboratoire avec, à chaque fois, plusieurs types de grains (3 tailles, 2 densités) afin d’étudier en détails le charriage sous l’eau. Tout d’abord, on s’intéresse aux trajectoires de grains isolés transportés sur le fond plat et lisse du canal. Alors que les écoulements sont pour la plupart hydrauliquement lisses (les grains sont immergés dans la sous-couche visqueuse), on montre que la turbulence proche paroi impacte fortement la dynamique des grains. Les données accumulées permettent d’estimer les moyennes et les dispersions de la vitesse des particules ainsi que de construire les fonctions de densité de probabilité. On compare ces résultats avec ceux obtenus par Vélocimétrie Laser Doppler (VLD) concernant l’écoulement fluide proche du fond. Dans un second temps, on quantifie le transport par la mesure de la quantité de sédiments charriés sur des lits plats pour différentes contraintes de cisaillement. L’idée qui motive ces expériences est la détermination empirique de la loi de transport(la relation entre le flux de grains et le cisaillement) pour notre configuration expérimentale. Nos résultats diffèrent assez largement des lois de transport habituelles du type Meyer-Peter et Müller et s’interprètent plus aisément dans le cadre des lois pour le charriage de faible intensité. Ensuite, une méthode de Profilométrie par Transformée de Fourier (PTF) est implémentée pour mesurer avec précision les surfaces 3D de petites dunes aquatiques (les barkhanes) et leur évolution au cours du temps. On s’intéresse, entre autres, à la formation des barkhanes, à leur morphologie, à leur vitesse de migration ainsi qu’au phénomène d’érosion. On apporte la preuve que les barkhanes aquatiques se déplacent sans se déformer et dès lors, on montre qu’il devient facile d’évaluer le flux de grains sur le dos des dunes (dans le plan de symétrie). Pour finir, on caractérise l’écoulement au-dessus d’une "maquette" de barkhane par VLD et on en déduit le cisaillement sur le dos de la dune (du pied à la crête). Ces mesures de cisaillement couplées aux lois de transport pour notre canal et aux résultats concernant les barkhanes obtenus par PTF nous permettent de discuter du phénomène de saturation du flux de grains. On estime la valeur de la longueur de saturation pour le charriage tout en restant prudent quant à la fiabilité de notre conclusion à ce sujet. / This experimental thesis deals with sediment transport induced by a fluid (air or water generally)flowing over an erodible bottom (sand bed for example) and providing a sufficiently high strengthto whip up solid particles and carry them downstream. Here, one focuses only on aquatic bedload: in this transport regime, moving grains roll, slip, jump or stop alternatively while staying confinedclose to the bed. We achieve different kinds of experiments in a laboratory flume involving everytime several grain types (3 sizes, 2 densities) in order to study underwater bedload in details. First,one considers trajectories of isolated beads transported over the flat and smooth bottom of theflume. Although the studied flows are hydraulically smooth (grains are fully embedded in theviscous sublayer), one shows that near wall turbulence highly impact grains dynamics. From ourdata, we extract mean particle velocities, standard deviations and also probability densityfunctions. One compares these results with those obtained by Laser Doppler Velocimetry (LDV)concerning the near wall fluid flow. Second, one quantifies the bedload by measuring the volumeof grains transported at the surface of flat granular beds for different shear stresses. The goal ofthese experiments is the empirical determination of the transport law (the relationship between thegranular flux and the fluid shear) for the present experimental configuration. Our results differsignificantly from usual transport laws like the one of Meyer-Peter and Müller and are bestunderstood within the framework of weak sediment transport. Then, one implements a FourierTransport Profilometry (FTP) method in order to measure with a good resolution the 3D surfacesof small aquatic dunes (called barchans) and their time evolution. One takes an interest in, amongothers, barchans formation, their morphology, their displacement velocity and the erosionphenomenon. One shows that subaquatic barchans move downstream without deforming and thusit becomes easier to estimate the granular flux on the back of the dunes (in the symmetry plane).Eventually, one characterizes the flow above a barchan "model" by LDV and deduces the shearstress along the dune’s back (from foot to crest). These shear stress measurements coupled withthe empirical transport laws and results concerning barchans obtained by FTP permit us todiscuss saturation effects of the granular flux. One estimates the saturation length for the bedloadand highlights that our conclusion on this point must be taken with care.

Bedload transport studies are essential in the understanding of river forms, functions and processes. These studies have been done using various methods over the past century. In recent years Radio Frequency Identification Technology (RFID) has become popular with researchers to track bedload particles. However, no standard operating procedures are used in the implementation of this technology. Methods used for tagging, seeding and tracking RFID tracers (RFID transponders inserted into a bedload particle) can introduce variability in their detection. In this study, RFID tracers were used to study four sites in Laurel Creek in Waterloo, Ontario. Two hundred RFID tracers were seeded in each of the four sites. Following three major storm events, the tracers were tracked with an antenna and their locations surveyed. The tracers were able to be detected to a precision of 1 m as a transponder used can be detected at a maximum of this distance. Practical tracking in the field highlighted the need for the understanding of how precisely the tag location can be identified. Laboratory experiments were designed and carried out to determine the effects of factors (tracer orientation, antenna orientation, tracer size, clustering of multiple tracers, burial depth, saturation and submergence of the soil matrix) that possibly confounded detection. Of these factors, tracer orientation, clustering and burial depths were determined to be the ones that affected detection distances the most. A transponder in a vertical orientation was found to have as much as 40% larger range of detection than a transponder in a horizontal orientation (i.e., they could be detected from further away). Additionally, “skip zones” were identified during laboratory and field experiments. These are zones of gaps in the electromagnetic field of the transponder that occur directly over the transponder. These zones were experimentally determined to extend to approximately 10 cm on each side of the transponder. Therefore, by identifying the skip zones, the tracers can be located to a precision of 10 cm; this is an order of magnitude smaller than the published detection limit of the transponder. The precision of detection can also be improved by the reduction of the effects of confounding factors. However, the improvement in the precision of detection is a tradeoff with the ease of detection. A tagging, seeding and tracking protocol is recommended to counter the effects of confounding factors.