The DUPAL anomaly (Hart 1984) in the South Atlantic has been attributed variably to deep sources related to mantle upwellings or shallow sources from continental material either from recent Gondwana breakup or ongoing erosions of the cratonic keels. Spatial distribution of the DUPAL anomaly provides an important constraint to distinguish between these possibilities. However, to this point sampling of South Atlantic mantle sources has been limited to the mid-Atlantic Ridge, young ocean islands, the Walvis Ridge and a single sample from Discovery Seamount, leaving uncertainties in the extent of the DUPAL anomaly, particularly its southern limit is poorly defined. Dredge samples from the effectively un-sampled Shona Ridge Meteor Rise Agulhas Ridge Cape Rise and Discovery seamounts were collected by the ANT XXIII/5 cruise of the FS Polarstern. Isotopic compositions of the new Discovery seamount samples form endmembers to the so-called Discovery and LOMU geochemical anomalies on the southern Mid-Atlantic Ridge, suggesting that the latter are formed from the same plume source through plume-ridge interaction (Class et al. 2009). Sr-Nd-Hf-Pb isotope data as well as preliminary Ar-Ar ages (J. OConnor work in progress) on the Shona aseismic ridges indicate the longevity of the Shona plume forming a zig-zag plume track (Hartnady & le Roex 1985). The new data integrated with literature data demonstrate an isotopically strongly heterogeneous source region for South Atlantic intraplate volcanism, including DUPAL, extreme EMI and HIMU. To this point the Shona plume signature on the mid-Atlantic Ridge was taken to be outside of the DUPAL region as its geochemical signature has HIMU affinity. All the Shona aseismic ridges sample this HIMU-like signature, however, each ridge has samples with DUPAL signature as well, suggesting that all the Shona-related bathymetric anomalies tap the DUPAL source. All components contributing to the Shona ridges are found in the 3000 km long Tristan plume trail as well where the single location with HIMU affinity might simply reflect limited sampling. The extent of the DUPAL anomaly is discussed in the context of constraints on the mantle flow field as well as the composition of subcontinental lithospheric mantle and lower crust. Dynamic models require a shallow continental DUPAL anomaly to originate from the African continent, which is not supported by available data. Ultrapotassic rocks have been argued to represent the composition of the South American subcontinental lithospheric mantle showing the DUPAL signature. We argue this an insufficient argument for a shallow origin of the anomaly. Instead, the extent of the DUPAL anomaly along the mid Atlantic ridge as well as the bathymetric anomalies supports a deep origin of the DUPAL signature. Hart, S.R., 1984. Nature 309, 753-757.Class, C. and le Roex, A.P., 2009. Geochim Cosmochim Acta 73 (13) A229.Hartnady, C.J.H. and le Roex, A.P.,, 1985. Earth Planet. Sci. Lett. 75, 245-257.