The Southern Ocean (SO) has long been recognised as a key player in regulating atmospheric CO2 variations and hence global climate. Here, the biological utilisation of nutrients regulates the preformed nutrient inventory for most of the deep ocean and, therefore, the global average efficiency of the biological pump. Marine sediment records from the Subantarctic Atlantic and Pacific document that higher mineral dust flux, increased bioproductivity, and lower atmospheric CO2 co-varied on glacial-interglacial time scales, which has been associated with iron fertilisation. It has been suggested that up to 40-50 ppmv of past atmospheric CO2 changes are related to iron fertilisation in the Subantarctic Ocean. The main objective of DustIron is an improved characterisation of the modern and past dust cycle and its link to SO iron fertilisation and atmospheric CO2 through a closely coupled novel datamodel approach. Within this project we want to extend the available geographic coverage of modern dust deposition, provenance and marine productivity records as well as the spatial and temporal variability during past glacial-interglacial cycles across all SO sectors. For the modern ocean we will explore dust fluxes, grain-size, and geochemical source area fingerprints (iron mineralogy, isotopy). Iron fertilisation and productivity will be assessed with a variety of both traditional (e.g., fluxes of biogenic barium or opal) and novel proxies for nutrient utilisation (δ15N in foraminifera). Our paleostudies will focus on the last glacial-interglacial climate transitions from the western Atlantic proximal to the Patagonian sources and from the central Indian SO (Kerguelen) to the eastern Indian Ocean sector, in order to obtain a circum-Antarctic view of dust-productivity changes. This will be complemented by a modelling study, simulating glacial-interglacial changes of atmospheric dust concentrations, deposition fluxes and linked SO bioproductivity.