Life on Earth could have existed already 3.8 Ga ago, and yet, more complex, multicellular life did not evolve until over three billion years later, about 700 Ma ago. Many have searched for the reason behind this apparent delay in evolution, and the dominating theories put the blame on the hostile Precambrian environment with low oxygen levels and sulphide-rich oceans. There are, however, doubts whether this would be the full explanation, and this thesis therefore focuses on a new hypothesis; the levels of the redox sensitive element arsenic increased in the oceans as a consequence of the change in weathering patterns that followed the upcome of free oxygen in the atmosphere at about 2.4 billion years ago. Given its toxicity, this could have had negative effects upon the life of the time. To test the hypothesis, 66 samples from drill cores coming from South Africa and Gabon with ages between 2.7 and 2.05 Ga were analysed for their elemental composition, and their arsenic content were compared with carbon isotope data from the same samples. These confirmed that a rise in arsenic concentration following the upcome of free oxygen in the atmosphere and the onset of oxidative weathering of continental sulphides. Arsenic, which is commonly found in sulphide minerals, was weathered together with the sulphide and delivered into the oceans, where it in the Palaeoproterozoic increased to over 600% compared to the older Archaean levels, at least locally. Iron had the strongest control over the arsenic levels in the anoxic (ferruginous and sulphidic) oceans, probably due to its ability to remove arsenic through adsorption. During oxygenated conditions, sulphur instead had the strongest influence upon arsenic, likely because of the lack of dissolved iron. The highest arsenic levels were found in samples recognised as coming from oxygenated conditions, although this might be due to the oxygenation state of arsenic affecting its solubility. Arsenic is toxic already at low doses, especially if the necessary arsenic detoxification systems had not yet evolved. However, the lack of correlation between arsenic and changes in δ13C indicated that the increase of arsenic did not affect the primary production between 2.7 and 2.05 Ga. Thus, whether arsenic could have affected the evolution of life during the Mesoproterozoic remains to be shown.

Les conditions oxydantes de la surface terrestre actuelle sont dues à la teneur élevée en dioxygène de l’atmosphère. Au début de l’histoire de la Terre il y a 4.54 milliards d'années (Ga), l’oxygène n’était pas stable dans l’atmosphère. Il a fallu deux épisodes d’augmentation brutale de ce gaz atmosphérique pour qu’il atteigne son niveau actuel : l’un vers 2.4 Ga, nommé le Grand Evènement Oxydant (GOE) qui fait l’objet de ce projet, l’autre 2 milliards d’années plus tard, nommé l’Evènement Oxydant Néo-protérozoïque (NOE). Le GOE est vraisemblablement le résultat de l’émersion généralisée de larges continents dont l’érosion libère le phosphate dans l’océan, un nutriment nécessaire à la production biologique, qui a donc permis l’explosion de la photosynthèse oxygénée. Ces deux hausses d’oxygène atmosphérique coïncident avec deux évolutions majeures dans l’histoire de la vie : (i) peu après le GOE, les eucaryotes sont apparus, alors que (ii) le NOE correspond à l’apparition des métazoaires et à l’explosion cambrienne. L’étude de ces phénomènes atmosphériques primitifs peut avoir d’importantes répercussions sur notre compréhension de l’origine et de l’évolution de la vie, qu’on estime principalement marine à cet âge. Les seules archives de ces temps primitifs sur Terre sont les roches sédimentaires. Pour savoir comment l’oxygénation de l’atmosphère a pu être reliée à cette vie marine, il faut tout d’abord comprendre comment l’océan a interagi avec l’atmosphère lors de cet évènement d’oxygénation. Cette question est au coeur de ce projet : comment le GOE a-t-il affecté les cycles biogéochimiques océaniques dont la vie est dépendante ? Nous nous sommes intéressés aux formations ferrifères litées ou BIFs (Banded Iron Formations). La chimie de ces roches marines fait écho à celle de l’océan contemporain à leur formation. Déterminer quantitativement la composition de l’océan à partir de celles des sédiments, même chimiques, est un défi quasiment impossible à relever y compris dans l’océan moderne. C’est pourquoi nous avons proposé de déterminer le temps de résidence d’éléments sensibles aux conditions redox de la surface, le soufre, le fer et le cuivre dans l’océan pré-GOE. Nous avons obtenu, par des séries temporelles, le spectre des fluctuations isotopiques de ces éléments enregistrées dans des carottes de formations ferrifères litées. La limite inférieure du spectre donne le temps de résidence de ces éléments dans l’eau de mer et fournit donc une indication solide sur la teneur de ces éléments dans l’océan à cette période. Nous avons analysé des échantillons protérozoïques proches de la limite Archéen-Protérozoïque du Transvaal (Afrique du Sud) et d’Hamersley (Australie). Des échantillons eoarchéens de Nuvvuagittuq (Canada) ont été récoltés mais n'ont pas pu être analysés faute de temps. / The present-day oxidizing conditions at Earth's surface are due to the high oxygen content of the atmosphere. However, oxygen was not always stable in the terrestrial atmosphere. Two distinct periods during which oxygen increased in a step-like manner were required to reach the current atmospheric oxygen level. The first, at about 2.4 Ga, is known as the Great Oxidation Event (GOE) and is at the core of this Ph.D. thesis. The other, occurring almost two billion years later, is called the Neo-Proterozoic Oxidation Event (NOE). The GOE likely is the result of the beginning widespread emergence of large continental expanses whose subsequent erosion gradually released phosphate into the ocean. Phosphate, a nutrient essential to organic production, in turn allowed the explosion of oxygenated photosynthesis. The GOE and NOE coincide with two major changes in the history of life. Shortly after the GOE, eukaryotes appeared, while the NOE corresponds to the appearance of metazoans and the Cambrian explosion. A better grasp of the GOE hence may have important implications for the understanding of the origin and evolution of life, which is thought to have been mainly marine at this stage in Earth history. The only records of the oxygen level during these ancient times are found in terrestrial sedimentary rocks. To understand how oxygenation of the atmosphere relates to marine life, we must first understand how the ocean was connected to the atmosphere during the GOE and how the GOE affected life-dependent ocean biogeochemical cycles. To this end we focused on banded iron formations (BIF). The chemistry of these sedimentary marine rocks directly reflects the chemistry of the contemporary ocean. Deriving quantitatively the composition of the ocean from a hydrogenous sediment is a challenge almost impossible to meet, even for the modern ocean. This is why we instead determined the residence time of redox-sensitive elements (in this case sulfur, iron, and copper) in the pre-GOE ocean. We specifically targeted the periods of isotopic fluctuations in these elements as recorded in BIF cores. The lower limit of the spectrum provides the residence time of these elements in seawater, hence giving a robust indication of their contents in the pre-GOE ocean. We sampled early Proterozoic BIF near the Archean-Proterozoic boundary in Transvaal (South Africa) and Hamersley (Australia), as well as Archean BIF from Nuvvuagittuq (Canada), though the latter were not analyzed during this thesis due to shortage of time.