Search

Assessments of aquatic paleoproduction and pigment preservation require accurate identification and quantification of sedimentary chlorophylls. Using chromatographic techniques to analyze long records at high resolution is impractical because they are expensive and labor intensive. We have developed a new rapid and low-cost approach to infer the concentrations of chlorophyll a, chlorophyll b and related chlorophyll derivatives (pheopigments a) from the mathematical decomposition of UV–VIS measured bulk spectrophotometer absorption spectra of standard solutions and sediment extracts. We validated our method against high-performance liquid chromatography (HPLC) measurements on standard solutions and on varved, anoxic sediment from eutrophic Lake Lugano (Ponte Tresa sub-basin, southern Switzerland), where the history of productivity is relatively well known for the twentieth century. Our mathematical approach quantifies the concentration of chlorophyll b ( $$ {\text{R}}_{{{\text{ad}}_{\text{J}}}}^{2} $$ = 0.99, RMSEP ~ 5.9%), chlorophyll a ( $$ {\text{R}}_{{{\text{ad}}_{\text{J}}}}^{2} $$ = 0.98, RMSEP ~ 5.0%), and pyropheophorbide a ( $$ {\text{R}}_{{{\text{ad}}_{\text{J}}}}^{2} $$ = 0.99, RMSEP ~ 7.8%) in standard solutions. We obtain comparable results for total chloropigment a (chlorophyll a + pheopigments a), chlorophyll a and diagenetic products (pheopigments a) in the sediment samples of our case study (Ponte Tresa). Here, HPLC concentrations of chlorophyll b are very low. The method has, however, the potential to achieve values for chlorophyll b concentrations in sediments with chlorophylls a/chlorophylls b ratios lower than 3.4. The pigment stratigraphy of the Ponte Tresa sediments correspond very well with the paleoproduction and eutrophication history of the twentieth century. The ratio between chlorophyll a and pheopigments a used as a qualitative indicator of sedimentary chlorophyll preservation (chlorophyll a/{chlorophyll a + pheopigments a}) is only weakly correlated with aquatic paleoproduction (radj = 0.35, p-value = 0.045) and remained remarkably constant in the recent century despite strong anthropogenic eutrophication. The new method is useful for obtaining, in a cost- and time-efficient way, information about major sedimentary pigment groups that are relevant to inferring paleoproduction, potentially green algae biomass, pigment preservation and early diagenetic effects.

The last 6000 years are of particular interest to the understanding of the Earth System because the boundary conditions of the climate system did not change dramatically (in comparison to larger glacial–interglacial changes), and because abundant, detailed regional palaeoclimatic proxy records cover this period. We use selected proxy-based reconstructions of different climate variables, together with state-of-the-art time series of natural forcings (orbital variations, solar activity variations, large tropical volcanic eruptions, land cover and greenhouse gases), underpinned by results from General Circulation Models (GCMs) and Earth System Models of Intermediate Complexity (EMICs), to establish a comprehensive explanatory framework for climate changes from the Mid-Holocene (MH) to pre-industrial time. The redistribution of solar energy, due to orbital forcing on a millennial timescale, was the cause of a progressive southward shift of the Northern Hemisphere (NH) summer position of the Intertropical Convergence Zone (ITCZ). This was accompanied by a pronounced weakening of the monsoon systems in Africa and Asia and increasing dryness and desertification on both continents. The associated summertime cooling of the NH, combined with changing temperature gradients in the world oceans, likely led to an increasing amplitude of the El Niño Southern Oscillation (ENSO) and, possibly, increasingly negative North Atlantic Oscillation (NAO) indices up to the beginning of the last millennium. On decadal to multi-century timescales, a worldwide coincidence between solar irradiance minima, tropical volcanic eruptions and decadal to multi-century scale cooling events was not found. However, reconstructions show that widespread decadal to multi-century scale cooling events, accompanied by advances of mountain glaciers, occurred in the NH (e.g., in Scandinavia and the European Alps). This occurred namely during the Little Ice Age (LIA) between AD 1350 and 1850, when the lower summer insolation in the NH, due to orbital forcing, coincided with solar activity minima and several strong tropical volcanic eruptions. The role of orbital forcing in the NH cooling, the southward ITCZ shift and the desertification of the Sahara are supported by numerous model simulations. Other simulations have suggested that the fingerprint of solar activity variations should be strongest in the tropics, but there is also evidence that changes in the ocean heat transport took place during the LIA at high northern latitudes, with possible additional implications for climates of the Southern Hemisphere (SH).

Varved lake sediments are excellent natural archives providing quantitative insights into climatic and environmental changes at very high resolution and chronological accuracy. However, due to the multitude of responses within lake ecosystems it is often difficult to understand how climate variability interacts with other environmental pressures such as eutrophication, and to attribute observed changes to specific causes. This is particularly challenging during the past 100 years when multiple strong trends are superposed. Here we present a high-resolution multi-proxy record of sedimentary pigments and other biogeochemical data from the varved sediments of Lake Żabinskie (Masurian Lake District, north-eastern Poland, 54°N–22°E, 120 m a.s.l.) spanning AD 1907 to 2008. Lake Żabinskie exhibits biogeochemical varves with highly organic late summer and winter layers separated by white layers of endogenous calcite precipitated in early summer. The aim of our study is to investigate whether climate-driven changes and anthropogenic changes can be separated in a multi-proxy sediment data set, and to explore which sediment proxies are potentially suitable for long quantitative climate reconstructions. We also test if convoluted analytical techniques (e.g. HPLC) can be substituted by rapid scanning techniques (visible reflectance spectroscopy VIS-RS; 380–730 nm). We used principal component analysis and cluster analysis to show that the recent eutrophication of Lake Żabinskie can be discriminated from climate-driven changes for the period AD 1907–2008. The eutrophication signal (PC1 = 46.4%; TOC, TN, TS, Phe-b, high TC/CD ratios total carotenoids/chlorophyll-a derivatives) is mainly expressed as increasing aquatic primary production, increasing hypolimnetic anoxia and a change in the algal community from green algae to blue-green algae. The proxies diagnostic for eutrophication show a smooth positive trend between 1907 and ca 1980 followed by a very rapid increase from ca. 1980 ± 2 onwards. We demonstrate that PC2 (24.4%, Chl-a-related pigments) is not affected by the eutrophication signal, but instead is sensitive to spring (MAM) temperature (r = 0.63, pcorr < 0.05, RMSEP = 0.56 °C; 5-yr filtered). Limnological monitoring data (2011–2013) support this finding. We also demonstrate that scanning visible reflectance spectroscopy (VIS-RS) data can be calibrated to HPLC-measured chloropigment data and be used to infer concentrations of sedimentary Chl-a derivatives {pheophytin a + pyropheophytin a}. This offers the possibility for very high-resolution (multi)millennial-long paleoenvironmental reconstructions.

Afin de favoriser l’utilisation des résultats de la recherche dans le processus politique, les scientifiques doivent communiquer avec différents types de publics ayant des besoins et des caractéristiques propres. Alors que la phase de mise à l’agenda politique d’un problème tend à retenir l’attention en raison de sa forte visibilité, l’action publique ne s’arrête pas aux portes de la prise de décision politique. Pour assurer la mise en œuvre d’une réponse publique à des problématiques telles que le changement climatique, il s’agit d’une part de susciter la prise de conscience des élu.e.s et du grand public (communication scientifique axée sur le problème), mais aussi d’autre part de mettre à disposition des administrations une expertise portant sur le contenu des politiques publiques (communication scientifique axée sur les politiques publiques). Les savoirs et savoir faire accumulés en évaluation des politiques publiques constituent à cet égard une ressource essentielle.

The recent warming of the global climate is well recognized. Many of the observed changes are unprecedented over decades to millennia. This conclusion could be drawn because an increasingnumber of paleoclimate time series with the detail needed, have and are being developed.At the same time, it is also recognized that further advances in this field of research are stilllimited by the scarcity of well calibrated and quantified time series of reconstructed variables fromnatural paleoclimate archives that cover the last Millennium. For instance, the interaction betweentemperature and precipitation, and its influence on the flood properties remain an open question.For this thesis, two lakes were investigated for their potential as paleoclimate archives: Proglacial Lake Oeschinen in the Bernese Alps that contains clastic varves Glacial Lake Żabińskie in northeast Poland that contains biochemical varvesThe following three research questions were addressed: (i) Can we use information fromrapid surface scanning reflectance spectroscopy in the visible light range (VIS-RS) as a quantitativeproxy for the composition of lake sediments? (ii) Can we calibrate sediment proxy data and VIS-RSdata with climate variables in the instrumental period? (iii) Can we produce a 1000-year longquantitative climate reconstruction from sediment proxy data and VIS-RS data?Scanning VIS-RS methodology was explored on the biochemical varves of Lake Żabińskie andthe clastic varves of Lake Oeschinen. In Lake Żabińskie, the absorption trough centred at 660-670nm(RABD660;670) could be calibrated to the total concentration of chlorins (chlorophyll-a degradationproducts = {Pheophytin-a + Pyropheophytin-a}; r = 0.75, p< 0.01) as measured by High PerformanceLiquid Chromatography (HPLC). In Lake Oeschinen, mean 1stderivative VIS-RS spectra were calibratedto the proportions of calcite/siliciclastics (r = 0.96, p< 0.01) as measured by loss on ignition.The study from Lake Oeschinen revealed that varve thickness could be used as a quantitativepredictor for cumulative alpine warm-season precipitation (rMJ = 0.71, rMJJA = 0.60, pc< 0.01; AD 1901-2008). This calibration model was used to produce a record of precipitation and flood frequency backto AD 884. Beyond answering research question (ii) and (iii), this millennial-long record enabledassessing the relationship between temperature, precipitation and floods over the past 1000 years.In the study of Lake Żabińskie, multivariate statistics were performed on a multi-proxy dataset of sedimentary pigments and other biochemical proxies. Results showed the recent lakeeutrophication could be quantitatively discriminated from climate-driven changes for the period AD1907-2008. Sedimentary pigments related to chlorophyll-a (Chl-a and degradation products) werenot affected by the eutrophication-based processes, but responded to spring March-to-Maytemperature (r = 0.63, pc< 0.05). This suggested that sediment photopigments from Lake Żabińskiecan be used for a climate reconstruction., Le réchauffement global de la planète est aujourd’hui reconnu et sans précédent sur uneéchelle de temps décennale à millénaire. Cette conclusion a pu être formulée grâce notamment aurecours à des données climatiques sur le passé de très bonne qualité, qui ont été et sont encoreproduites. Dans le même temps, il est de plus en plus admis que de nouvelles avancées dans cedomaine de la Recherche restent restreintes par la rareté des variables climatiques couvrant les 1000dernières années et produites de manière quantitative à partir d’archives naturelles du paléoclimat.Pour exemple, l’interaction entre les températures et précipitations, ainsi que leur influencemutuelle sur les propriétés et le contrôle des crues restent à examiner.À travers ce projet de doctorat, deux lacs ont été sélectionnés pour l’étude du paléoclimat: Lac d’Oeschinen, un lac d’altitude des Alpes Suisses qui présente des varves de type clastique Lac Żabińskie, au Nord-est de la Pologne qui présente des varves de type biogéochimiqueNous avons adressé les trois questions de recherche suivantes: (i) est-il possible d’utiliser laspectroscopie à lecture optique (VIS-RS; 380-730nm) afin de déterminer de manière quantitative lacomposition des sédiments lacustres? (ii) Peut-on calibrer les données proxy et VIS-RS de cessédiments avec les données climatiques pour la période instrumentale, et cibler le proxy à utiliserpour reconstruire le climat passé? (iii) Peut-on produire un jeu de données climatiques, à partir desinformations proxy et VIS-RS, qui soit quantitatif et couvrant les 1000 dernières années?Les résultats obtenus à partir de la spectroscopie à lecture optique sur les varves ont révéléque cette technique est un outil solide à la caractérisation d’éléments spécifiques (quantité etcomposition). Pour le lac Żabińskie, la bande d’absorption VIS-RS à 660-670 nm a pu être étalonnéeaux concentrations en Chlorins (produits de dégradation de la chlorophylle-a = {Phéophytine-a +Pyrophéophytine-a}; r = 0.75, p< 0.01) mesurées par Chromatographie en phase Liquide à HautePerformance (CLHP). Pour le Lac d’Oeschinen, la dérivée première du spectre VIS-RS a pu êtreétalonnée aux proportions en calcite/siliciclastics (r = 0.96, p< 0.01) mesurées par perte au feu.L’enregistrement varvé du Lac d’Oeschinen a été utilisé pour reconstruire les précipitationsd’été (Mai à Août) et reconstituer l’évolution de la fréquence des crues pour le Nord-ouest des Alpesjusqu’en 884 (rvarve thickness;MJJA = 0.60, pc< 0.01; AD 1901-2008). Au-delà d’avoir répondu aux questionsde recherche (ii) et (iii), cet enregistrement nous a également permis d’étudier les relations qui lienttempérature, précipitation et fréquence de crue à travers ce millénaire.Nous avons réalisé une étude statistique multi-variée à partir de pigments et autres proxybiochimiques des sédiments du lac Żabińskie afin de séparer quantitativement le signal climatique dusignal d’eutrophisation pour la période 1907-2008. Plus particulièrement, les données pigmentairesrelatives à la chlorophylle-a (matière première et produits de dégradation) ont montré unecorrélation significative avec les températures printanières (Mars à Mai, r = 0.63, pc< 0.05). À des finsde perspectives, ces résultats peuvent servir à reconstruire le climat en Pologne