No description available.

The factors that determine the efficiency of energy transfer in aquatic food webs have been investigated for many decades. The plant-animal interface is the most variable and least predictable of all levels in the food web. In order to study determinants of food quality in a large lake and to test the recently proposed central importance of the long-chained eicosapentaenoic acid (EPA) at the pelagic producer-grazer interface, we tested the importance of polyunsaturated fatty acids (PUFAs) at the pelagic producerconsumer interface by correlating sestonic food parameters with somatic growth rates of a clone of Daphnia galeata. Daphnia growth rates were obtained from standardized laboratory experiments spanning one season with Daphnia feeding on natural seston from Lake Constance, a large pre-alpine lake. Somatic growth rates were fitted to sestonic parameters by using a saturation function. A moderate amount of variation was explained when the model included the elemental parameters carbon (r2 = 0.6) and nitrogen (r2 = 0.71). A tighter fit was obtained when sestonic phosphorus was incorporated (r2 = 0.86). The nonlinear regression with EPA was relatively weak (r2 = 0.77), whereas the highest degree of variance was explained by three C18-PUFAs. The best (r2 = 0.95), and only significant, correlation of Daphnia's growth was found with the C18-PUFA α-linolenic acid (α-LA; C18:3n-3). This correlation was weakest in late August when C:P values increased to 300, suggesting that mineral and PUFA-limitation of Daphnia's growth changed seasonally. Sestonic phosphorus and some PUFAs showed not only tight correlations with growth, but also with sestonic α-LA content. We computed Monte Carlo simulations to test whether the observed effects of α-LA on growth could be accounted for by EPA, phosphorus, or one of the two C18-PUFAs, stearidonic acid (C18:4n-3) and linoleic acid (C18:2n-6). With >99 % probability, the correlation of growth with α-LA could not be explained by any of these parameters. In order to test for EPA limitation of Daphnia's growth, in parallel with experiments on pure seston, growth was determined on seston supplemented with chemostat-grown, P-limited Stephanodiscus hantzschii, which is rich in EPA. Although supplementation increased the EPA content 80-800x, no significant changes in the nonlinear regression of the growth rates with α-LA were found, indicating that growth of Daphnia on pure seston was not EPA limited. This indicates that the two fatty acids, EPA and α-LA, were not mutually substitutable biochemical resources and points to different physiological functions of these two PUFAs. These results support the PUFA-limitation hypothesis for sestonic C:P < 300 but are contrary to the hypothesis of a general importance of EPA, since no evidence for EPA limitation was found. It is suggested that the resource ratios of EPA and α-LA rather than the absolute concentrations determine which of the two resources is limiting growth.

L'évolution topographique des chaînes de montagne résulte d'interactions complexes entre la tectonique et le climat via l'action des processus de surface. La quantification de l'évolution du relief passe par le développement d'outils méthodologiques permettant d'identifier les processus mis en jeu pour des échelles de temps (103-106 années) et d'espace (1-100 km) caractéristiques de contextes orogéniques. Dans cette étude, l'évolution Néogène et Quaternaire du relief des Alpes Européennes est abordée à partir d'une approche couplant la thermochronométrie basse-température in-situ (essentiellement (U-Th-Sm)/He et 4He/3He sur apatite) et la modélisation numérique. Le développement d'une méthode numérique associant la modélisation thermo-cinématique (Pecube) à un algorithme d'inversion (Neighbourhood Algorithm) permet l'interprétation quantitative de données thermochronologiques en termes d'histoires d'exhumation et d'évolution du relief. La quantification de l'évolution du relief reste néanmoins problématique et fortement dépendante du contexte géomorphologique étudié. Cette étude implique la nécessité de repenser a priori la stratégie d'échantillonnage et d'interprétation des données suivant le contexte géomorphologique considéré et le signal d'exhumation que l'on cherche à quantifier via la thermochronologie. Enfin, l'application de cette approche à un jeu de données thermochronologiques échantillonnées dans le massif des Ecrins-Pelvoux (Alpes françaises) révèle l'existence d'un épisode d'exhumation rapide cessant autour de ~5-6 Ma, encadré par des taux d'exhumation plus modérés. Cependant, les données ne permettent pas de conclure quant à l'évolution tardi-Néogène du relief dans le massif des Ecrins-Pelvoux. L'application de la thermochronométrie 4He/3He dans la vallée du Rhône (Alpes suisses), couplée à des données thermochronologiques issues de la littérature, confirme un épisode d'exhumation rapide jusqu'à ~5-7 Ma, et révèle une augmentation majeure du relief local (~1-1.5 km) associée au creusement des vallées par d'importants appareils glaciaires. Le début de cette phase de creusement correspond à la transition climatique Mi-Pléistocène (~1 Ma) depuis des cycles glaciaires symétriques de 40 ka vers des cycles asymétriques (100 ka) de plus forte amplitude. Ces données permettent également de reconstruire la topographie pré-glaciaire du bassin versant du Rhône, et ainsi d'évaluer, à une échelle plus globale, l'impact des glaciations Pléistocènes sur l'évolution du relief. Des résultats préliminaires issus de la modélisation numérique des processus glaciaires mettent en évidence le potentiel d'une telle approche afin de tester quantitativement l'influence de la transition climatique Mi-Pléistocène sur le développement du relief alpin, ouvrant de nouvelles perspectives de recherche. Enfin, l'étude de l'évolution topographique post-glaciaire dans le massif des Ecrins-Pelvoux (modélisation numérique et utilisation du 10Be cosmogénique produit in-situ) met en évidence une dynamique d'érosion fluviale pouvant atteindre localement des vitesses de l'ordre du cm an-1, illustrant l'évolution géomorphologique rapide en réponse à la transition climatique entre le tardi-Pléistocène et l'Holocène. / The topographic evolution of mountain belts results from complex couplings between tectonics, climate and surface processes. Quantifying landscape evolution requires methodological tools to constrain forcing processes over temporal (103-106 years) and spatial (1-100 km) scales characteristic of orogenic systems. This thesis investigates the Neogene and Quaternary relief evolution of the European Alps using in situ low-temperature thermochronometry (mostly apatite (U-Th-Sm)/He and 4He/3He) and numerical modeling. A novel numerical approach combining thermal-kinematic modeling (Pecube) with an inversion scheme (Neighbourhood Algorithm) allows extracting quantitative information on exhumation and relief histories from thermochronological datasets. Quantifying relief evolution remains problematic, however, and strongly depends on the geomorphic setting. Our results show that both thermochronology data sampling and modeling strategies have to be considered a priori, in function of the geomorphic setting and the spatial/temporal scale of the exhumation signal to be constrained. This approach has been applied on a thermochronological dataset collected in the Ecrins-Pelvoux massif (French Alps). The results show a pulse of rapid exhumation until ~5-6 Ma, preceded and followed by more moderate rates of exhumation. However, the data cannot resolve the late-Neogene relief evolution in the Ecrins-Pelvoux massif. New 4He/3He thermochronometry data from the Rhône valley (Swiss Alps), combined with thermochronological data from the literature, also point out an episode of rapid exhumation until ~5-7 Ma, and reveal a major increase in local topographic relief (~1-1.5 km) linked to valley carving by large mountain glaciers. The onset of this phase of relief carving corresponds to the Mid-Pleistocene transition from symmetric 40-ka to asymmetric and high amplitude 100-ka glacial/interglacial oscillations. The new data also permit to reconstruct the pre-glacial topography of the Rhône basin, and to evaluate the net effect of Pleistocene glaciations on relief evolution at the basin scale. Preliminary results from numerical modeling of glacial dynamics highlight the potential opportunity of using such an approach to quantitatively assess the impact of the Mid-Pleistocene climate transition on Alpine relief development, leading to new research avenues. Finally, the post-glacial topographic evolution of the Ecrins-Pelvoux massif has been studied using numerical modeling and in situ cosmogenic 10Be analyses. The results suggest efficient fluvial incision at rates of cm yr-1, illustrating the efficient landscape response to late-Pleistocene/Holocene climate change.