The paleobiogeographical distribution of Oxfordian ammonites and coral reefs in northern and Central Europe, the Mediterranean area, North and East Africa, and the Middle East and Central Asia is compared with the distribution in time and space of the most important lithofacies. Interest in the Oxfordian is focused on changes in facies and in biogeographical patterns that can be interpreted as the results of climatic events. Paleotemperature trends inferred from oxygen isotopes and paleoclimatic simulations are tested against fossil and facies data. A Late Callovian-Early Oxfordian crisis in carbonate production is indicated by the widespread absence of Lower Oxfordian reefal formations. There is a gap (hiatus) in deposition on epicontinental platforms, with Middle Oxfordian deposits resting paraconformably on Upper Callovian, while shales accumulated in adjacent intracratonic basins. Simultaneously, in Mediterranean Tethys, radiolarites accumulated in deep troughs while Rosso Ammonitico facies formed on pelagic swells. However, deposition on swells was also discontinuous with numerous gaps (hiatuses) and sequences that are much reduced in thickness. Middle Callovian deposits are generally overlain by Middle Oxfordian limestones. The dearth of carbonates is consistent with a cooling event lasting about 1 My. By the middle Oxfordian a warming, leading to bgreenhouseQ type conditions, is suggested on the basis of both biogeographical (mostly coralreef distribution) and geochemical data. Carbonates spread onto an extensive European platform while radiolarites reached a maximum development in the Mediterranean Tethys. Two distinct latitudinal belts, with seemingly different accumulation regimes, are therefore inferred. Similar latitudinal belts were also present in the late Oxfordian, when carbonates were widespread. The distribution of reefal facies in the late Oxfordian-early Kimmeridgian fits relatively well with GCMs simulations that imply low rainfall in the Tethyan Mediterranean area and slightly higher precipitation in central and northern Europe. Local salinity variations, reflecting more arid or humid conditions, may bias the paleotemperature signal inferred from

Middle to Upper Oxfordian reefs of a shallow marine carbonate platform located in northeastern France show important facies changes in conjunction with terrigeneous contents. The Pagny-sur-Meuse section shows coral-microbialite reefs that developed both in pure carbonate limestones and in mixed carbonate-siliciclastic deposits. Phototrophic coral associations dominated in pure carbonate environments, whereas a mixed phototrophic/heterotrophic coral fauna occurred in more siliciclastic settings. Microbialites occur in pure carbonate facies but are more abundant in mixed carbonate-siliciclastic settings. Reefs seem to have lived through periods favourable for intense coral growth that was contemporaneous with a first microbialitic layer and periods more favourable for large microbialitic development (second microbialitic layer). The first microbialitic crust probably developed within the reef body and thus appears to be controlled by autogenic factors. The second generation of microbialites tended to develop over the entire reef surface and was probably mainly controlled by allogenic factors. Variations in terrigeneous input and nutrient content, rather related to climatic conditions than to water depth and accumulation rate, were major factors controlling development of reefs and their taxonomic composition.

A morphometrical method to quantify and characterize coral corallites using Richardson Plots and Kaye's notion of fractal dimensions is presented. A Jurassic coral species (Aplosmilia spinosa) and Wve Recent coral species were compared using the Box-Counting Method. This method enables the characterization of their morphologies at calicular and septal levels by their fractal dimensions (structural and textural). Moreover, it is possible to determine diVerences between species of Montastraea and to tackle the high phenotypic plasticity of Montastraea annularis. The use of fractal dimensions versus conventional methods (e.g., measurements of linear dimensions with a calliper, landmarks, Fourier analyses) to explore a rugged boundary object is discussed. It appears that fractal methods have the potential to considerably simplify the morphometrical and statistical approaches, and be a valuable addition to methods based on Euclidian geometry.

No abstract