Abstract Surface currents constitute an efficient transport agent for (larvae of) marine faunas, while the circulation of water masses in the ocean interior drives nutrient redistribution, ventilates the ocean and contributes to shaping surface biological productivity and the benthic redox landscape. Therefore, a robust understanding of ocean circulation, both shallow and deep, and of its response to climate change, is required to interpret palaeobiogeographic signals, biological productivity patterns and biodiversity trends. This is especially critical during periods of dynamic biological change, such as the Ordovician. Yet, oceanic circulation patterns leave no direct evidence in the geological record and can therefore be reconstructed solely based on indirect indicators, such as the distribution of faunas and geochemical proxies. General circulation models offer independent, physically robust insights onto the coupling between climate change and ocean circulation. Integrated approaches based on the assimilation of geological data in numerical models thus constitute a promising way forward. We here provide a literature review and updated synthesis of the current understanding of the Ordovician ocean circulation, based on data and models.

Abstract U–Th–Ra isotope analyses of whole rocks and mineral separates were conducted in order to perform isochron dating of three morphologically young lavas from Tatun volcano, northern Taiwan (from Mt Cising, the Shamao dome and the Huangzuei volcano). The data do not yield tight U–Th isochrons, indicating open-system magmatic processes. However, crystallization ages of two samples can be constrained: namely, less than about 1370 years for the Shamao dome, based on 226 Ra– 230 Th disequilibrium in magnetite, and less than approximately 70 ka (but potentially Holocene) for a Huangzuei flow, based on 238 U– 230 Th disequilibrium in plagioclase. Discordant Ar–Ar, 238 U– 230 Th and 226 Ra– 230 Th ages are best explained by young lavas having inherited some crystals from older lithologies (crystal mushes or rocks), and indicate that the above ages represent maxima. Our study provides the first evidence of effusive volcanism at the Tatun Volcano Group in Late Holocene times. All separates from the Shamao dome and Huangzuei volcano are in 234 U– 238 U equilibrium. Minerals in the Mt Cising sample are in 234 U– 238 U disequilibrium, despite the 234 U– 238 U equilibrium of the whole rock. We interpret this as uptake of a hydrothermally altered, old crystal cargo into fresh melt prior to eruption. A different dating approach will thus be required to constrain the eruption age of Mt Cising. Supplementary material: Ar–Ar plateaus from Mt Cising and the Shamao dome, reproduced from Lee (1996) , are available at www.geolsoc.org.uk/SUP18817