Gondwana Research 25 (2014) 1660–1679 Contents lists available at ScienceDirect Gondwana Research journal homepage: www.elsevier.com/locate/gr Seamounts off the West Antarctic margin: A case for non-hotspot driven intraplate volcanism A. Kipf a,⁎, F. Hauff a, R. Werner a, K. Gohl b, P. van den Bogaard a, K. Hoernle a, D. Maicher a, A. Klügel c a b c GEOMAR Helmholtz Centre for Ocean Research Kiel, Wischhofstr. 1-3, D-24148 Kiel, Germany Alfred-Wegener-Institute for Polar and Marine Research, Postfach 120161, D-27515 Bremerhaven, Germany University of Bremen, Postfach 330440, D-28334 Bremen, Germany a r t i c l e i n f o Article history: Received 21 December 2012 Received in revised form 28 May 2013 Accepted 11 June 2013 Available online 16 July 2013 Handling Editor: I. Safonova Keywords: Antarctica Marie Byrd Seamounts Intraplate volcanism 40 Ar/39Ar age dates Major and trace element and Sr–Nd–Pb–Hf Isotope geochemistry a b s t r a c t New radiometric age and geochemical data of volcanic rocks from the guyot-type Marie Byrd Seamounts (MBS) and the De Gerlache Seamounts and Peter I Island (Amundsen Sea) are presented. 40Ar/39Ar ages of the shield phase of three MBS are Early Cenozoic (65 to 56 Ma) and indicate formation well after creation of the Paciﬁc–Antarctic Ridge. A Pliocene age (3.0 Ma) documents a younger phase of volcanism at one MBS and a Pleistocene age (1.8 Ma) for the submarine base of Peter I Island. Together with published data, the new age data imply that Cenozoic intraplate magmatism occurred at distinct time intervals in spatially conﬁned areas of the Amundsen Sea, excluding an origin through a ﬁxed mantle plume. Peter I Island appears strongly inﬂuenced by an EMII type mantle component that may reﬂect shallow mantle recycling of a continental raft during the ﬁnal breakup of Gondwana. By contrast the Sr–Nd–Pb–Hf isotopic compositions of the MBS display a strong afﬁnity to a HIMU-type mantle source. On a regional scale the isotopic signatures overlap with those from volcanics related to the West Antarctic Rift System, and Cretaceous intraplate volcanics in and off New Zealand. We propose reactivation of the HIMU material, initially accreted to the base of continental lithosphere during the pre-rifting stage of Marie Byrd Land/Zealandia to explain intraplate volcanism in the Amundsen Sea in the absence of a long-lived hotspot. We propose continental insulation ﬂow as the most plausible mechanism to transfer the sub-continental accreted plume material into the shallow oceanic mantle. Crustal extension at the southern boundary of the Bellingshausen Plate from about 74 to 62 Ma may have triggered adiabatic rise of the HIMU material from the base of Marie Byrd Land to form the MBS. The De Gerlache Seamounts are most likely related to a preserved zone of lithospheric weakness underneath the De Gerlache Gravity Anomaly. © 2013 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved. 1. Introduction Seamounts are common bathymetric features on the seaﬂoor and most are of volcanic origin. Although only a fraction of them have been mapped by ship-based echo-sounding, satellite altimetry has identiﬁed more than 13,000 seamounts taller than 1.5 km and predicts more than 100,000 seamounts higher than 1 km (e.g., Smith and Sandwell, 1997; Wessel et al., 2010). Seamounts are important probes of the composition and dynamics of the oceanic mantle and, if they form parts of hotspot tracks, they can also be important recorders of past plate motions (Hofmann, 2003; Tarduno et al., 2003; Koppers et al., 2012). They also form oases for marine life and biodiversity (e.g., Shank, 2010 for a recent review) and are signiﬁcant components of hydrogeological systems focusing the exchange of heat and ﬂuids between the oceanic lithosphere and the oceans (e.g., Fisher et al., 2003; Harris et al., 2004; Hutnak et al., 2008; Klügel et al., 2011). The latter processes can lead to the formation of economically important ⁎ Corresponding author. Tel.: +49 431 600 2645. E-mail address: firstname.lastname@example.org (A. Kipf). mineral deposits (e.g., Hein et al., 2010), which are, for example, commercially mined in some accreted seamount complexes (e.g., Safonova, 2009). Seamounts are also sites of geological hazards such as tsunamis through sector collapse during their growth stage (e.g., McMurtry et al., 2004). Upon subduction of the ocean ﬂoor, seamounts can also serve as prominent asperities generating earthquakes (e.g., Watts et al., 2010 for a recent review). As the subduction process can lead to crustal accretion of seamounts, they can be preserved in the accessible geological record, providing important insights from the evolution of hotspot tracks and continental margins to biological exchange between continents (e.g., Hoernle et al., 2002; Geldmacher et al., 2008; Portnyagin et al., 2008; Buchs et al., 2011; Safonova and Santosh, 2012). Despite the manifold contributions of seamounts to the dynamics of diverse earth systems, their process of formation is still debated. Most commonly the occurrence of isolated volcanoes distant from plate boundaries is attributed to the upwelling of mantle plumes (e.g., Wilson, 1963; Morgan, 1971; Courtillot et al., 2003). The absence of linear volcanic chains and lack of spatially age progressive magmatism in many areas has stimulated a vigorous debate on the origin of intraplate volcanism (e.g., Anderson, 2000; Foulger and Natland, 2003; see also “Great 1342-937X/$ – see front matter © 2013 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gr.2013.06.013 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Plume debate”, www.mantleplumes.org). Other important mechanisms of seamount formation include off-axis volcanism in the vicinity of spreading ridges by lateral expansion of the ridge melting regime (e.g., Batiza et al., 1990; Brandl et al., 2012 and references therein), recycling of delaminated continental lithosphere (Hoernle et al., 2011) and plate fracturing (e.g., Winterer and Sandwell, 1987; Natland and Winterer, 2005). In this paper, we report for the ﬁrst time an integrated bathymetric, geochronological and geochemical data set from three seamount provinces off West Antarctica and show that these intraplate volcanoes are not directly linked to the activity of a mantle plume but rather reﬂect remobilization and transfer of fertile mantle from beneath West Antarctica. The Marie Byrd Seamounts (MBS), located in the western Amundsen Sea north of the continental shelf of Marie Byrd Land, West Antarctica (Fig. 1), are a good example of enigmatic intraplate volcanism. They are located on oceanic crust possibly older than 72 Ma (Heinemann et al., 1999; Eagles et al., 2004a,b) and form an elongated cluster of volcanic ediﬁces, that extends for more than 800 km between ~114° and ~131°W, and ~68° and ~71°S. Based on rock fragments found in corers and dredges carried out at a single MBS (Hubert Miller Seamount), Udintsev et al. (2007) assumed that this structure represents a relict fragment of continental crust which was destructed and altered by a mantle plume. The authors, however, admit that the material recovered cannot unambiguously be interpreted as in situ rocks. Although the MBS form a vast seamount province covering over 200,000 km2, their remote location made sampling difﬁcult, inhibiting elucidation of their age, magma sources and volcanic evolution. Moreover, the relationship of the MBS to the magmatism associated with the ﬁnal break-up of Gondwana and/or to the widespread but low volume intraplate volcanism in the SW Paciﬁc region (e.g., Weaver et al., 1994; Storey et al., 1999; Rocchi et al., 2002a; Finn et al., 2005; Hoernle et al., 2006, 2010; Timm et al., 2010) was poorly constrained. In 2006, the R/V Polarstern cruise ANT-XXIII/4 conducted a bathymetric mapping and dredge sampling survey of ﬁve MBS and associated structures. Samples from two other volcanic complexes in the Amundsen Sea, namely the previously studied ocean island volcano Peter I Island (e.g., Prestvik et al., 1990; Prestvik and Duncan, 1991; 1661 Hart et al., 1995) and the Belgica Seamount (De Gerlache Seamounts, Hagedorn et al., 2007) (Fig. 1), are included in our study to more fully characterize the sources and spatial evolution of intraplate magmatism in this region. Both Peter I Island and the De Gerlache Seamounts have been related to hotspot activity by most previous authors. Here, we present results of the bathymetric surveys together with 40 Ar/39Ar ages and geochemical data (major and trace element and radiogenic Sr–Nd–Pb–Hf isotope ratios) of the recovered rocks. We show that magmatism in the Amundsen Sea occurred at distinct time intervals in spatially conﬁned areas ruling out an origin through a single stationary hotspot. Notably this volcanism appears predominantly inﬂuenced by HIMU (high time-integrated 238U/204Pb) type mantle, requiring emplacement and upwelling of such material in the depleted upper oceanic mantle well after the breakup of Zealandia from Antarctica. After brieﬂy summarizing the tectonic and magmatic evolution affecting this part of the SW Paciﬁc over the past 100 Ma, we discuss our results and evaluate processes, which may cause non-hotspot related HIMU-type intraplate volcanism in the Amundsen Sea. 2. Tectonic and magmatic evolution of the SW-Paciﬁc over the past 100 Ma Plate-kinematic reconstructions (Fig. 2) demonstrate that Marie Byrd Land was attached to the southeastern margin of Zealandia prior to the ﬁnal breakup of Gondwana (Fig. 2a; e.g., Eagles et al., 2004a). After the collision of the Hikurangi Plateau with the Gondwana margin (e.g., Davy et al., 2008; Hoernle et al., 2010) and cessation of subduction along the northern margin of Zealandia at c. 100 Ma (e.g., Weaver et al., 1994), extensional processes set in, causing Zealandia to rift from Marie Byrd Land (e.g., Larter et al., 2002; Eagles et al., 2004a; Boger, 2011 for a recent review). The continental breakup initiated with the Chatham Rise separating from the Amundsen Sea Embayment sector during the Cretaceous Normal Polarity Superchron (CNS) at about 90 Ma (Fig. 2b). Thereafter the southwestward rift propagation jumped farther south and separated the Campbell Plateau from Marie Byrd Land just before chron C33 (83–79 Ma), leaving a rifted West Antarctic continental Fig. 1. Overview map of West Antarctica and the Amundsen Sea. The three seamount/ocean island volcanic provinces of the Amundsen Sea are marked by yellow circles. Dashed red lines indicate major tectonic lineaments (WARS — West Antarctic Rift System from Müller et al. (2007); DGGA — De Gerlache Gravity Anomaly). The map is based on the GEBCO_08 Grid (version 20091120, http://www.gebco.net). 1662 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Fig. 2. Plate-tectonic reconstruction from 100 Ma to 22 Ma, using rotation parameters by Eagles et al. (2004a). Illustrated are the collision of Hikurangi Plateau with Zealandia at around 100 Ma, the breakup between Zealandia and West Antarctica at 90–80 Ma, the development of the Bellingshausen Plate and the subsequent volcanism along the West Antarctic margin. Double lines mark spreading ridge plate boundaries, single solid lines mark other plate boundary types, and dashed lines in West Antarctica illustrate lineaments of the West Antarctic Rift System (Eagles et al., 2009; Gohl et al., 2013). Abbreviations are: SNS South Island New Zealand, HIK Hikurangi Plateau, CP Campbell Plateau, CR Chatham Rise, GSB Great South Basin, BS Bollons Seamount, BT Bounty Trough, WA West Antarctica, MBL Marie Byrd Land, AP Antarctic Peninsula, ASE Amundsen Sea Embayment, WARS West Antarctic Rift System, PAC Paciﬁc Plate, PHO Phoenix Plate, BP Bellingshausen Plate, MBS Marie Byrd Seamounts (red area marks volcanic activity of the shield phase), DGS De Gerlache Seamounts, PI Peter I Island, DGGA De Gerlache Gravity Anomaly (suture of former PHO-BP ridge jump). margin bordering the Amundsen Sea (Fig. 2c, d; Larter et al., 2002; Eagles et al., 2004a). During the late Cretaceous/Early Tertiary the southern Paciﬁc region was sectioned into a minimum of three major tectonic plates (Bradshaw, 1989; Larter et al., 2002; Eagles et al., 2004a; Wobbe et al., 2012), the Paciﬁc Plate, the Bellingshausen Plate, and the Phoenix or Aluk Plate adjacent to the Antarctic Plate (Fig. 2e). While the Phoenix Plate subducted beneath the eastern portion of the Antarctic Plate, the other plate boundaries were divergent or transform margins. During C27 (61 Ma) the Bellingshausen Plate ceased from being a separate A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 plate and became incorporated into the Antarctic Plate (Fig. 2f; Eagles et al., 2004a,b; Wobbe et al., 2012). Heinemann et al. (1999) and Stock (1997) suggest that the MBS province formed in the vicinity of the Antarctic–Paciﬁc–Bellingshausen triple junction. Between C27 and C25/ C24 (57–54 Ma), a substantial drop in spreading rate occurred at the Paciﬁc–Antarctic Ridge, and together with a gradual rotation of the spreading direction (Müller et al., 2000), an increase in fracture zone density is notable (Eagles et al., 2004a). At the same time, the West Antarctic Rift System (WARS) continued its crustal extension in Marie Byrd Land and possibly into the Amundsen Sea Embayment just south of the MBS (Gohl et al., 2013). The De Gerlache Seamounts and Peter I Island are aligned along the so-called De Gerlache Gravity Anomaly (DGGA) (Gohl et al., 1997a; McAdoo and Laxon, 1997; Hagedorn et al., 2007) (Fig. 2g + h) which was initially interpreted as a fracture zone of the earlier Phoenix–Antarctic Ridge (Hart et al., 1995). However, magnetic seaﬂoor spreading data imply that this is a tectonic scar caused by a westward jump of the Paciﬁc–Phoenix ridge at chron C27 (Larter et al., 2002; Eagles et al., 2004a). Müller et al. (2007) suggested that this zone of possible lithospheric weakness was reactivated by a northward extension of a later WARS branch (Figs. 1 and 2h). The Late Cretaceous tectonic events were accompanied by intense volcanism in East Gondwana and Marie Byrd Land at c. 95–110 Ma (e.g., Hart et al., 1997; Storey et al., 1999). This magmatism has been related to large-scale mantle upwelling in conjunction with extensioninduced rifting (Finn et al., 2005). Others assume an active mantle plume in the area of the Bellingshausen–Amundsen Sea or beneath East Gondwana (Hole and LeMasurier, 1994; Weaver et al., 1994; Rocholl et al., 1995; Hart et al., 1995, 1997; Panter et al., 2000; Hoernle et al., 2010; Sutherland et al., 2010), which may have caused the ﬁnal break-up of Zealandia from Antarctica (e.g., Weaver et al., 1994; Storey et al., 1999; Hoernle et al., 2010). As the region underwent further plate reorganization, a second phase of volcanism occurred (Rocchi et al., 2002a,b; Nardini et al., 2009 and references therein, LeMasurier et al., 1990). This younger magmatism (30–25 Ma until recent) is mainly of alkaline nature and has been related to rifting and crustal extension associated with the WARS. Based on a HIMU (high time-integrated U/Pb) component found in many WARS volcanics, many authors suggest reactivation of old plume material embedded at the base of the continental lithosphere (e.g., Weaver et al., 1994) others favor a metasomatic origin (e.g., Nardini et al., 2009). 3. Bathymetry and morphology of Marie Byrd Seamounts and Peter I Island During cruise ANT-XXIII/4, the onboard Atlas Hydrosweep DS-2 multi-beam echo-sounding system of the R/V Polarstern was used to generate maps of ﬁve MBS (summarized in Table 1; Fig. 3a) and of the submarine base of Peter I Island (Gohl, 2007). Combined with bathymetric 1663 data of previous cruises (RV Nathaniel B. Palmer in 1996, RV Polarstern ANT-XI/3 in 1994, and ANT-XVIII/5a in 2001; e.g., Miller and Grobe, 1996; Feldberg, 1997), these data reveal that the MBS are characterized by steep sides with relatively ﬂat tops and additional small cones on the upper ﬂanks and/or on the platforms. The guyot-like morphology of the main ediﬁces is attributed to seamount growth above sea level to form ocean island volcanoes, which subsequently eroded to sea level and then subsided to their present position. The small cones must have formed after subsidence of the erosional platforms below wave base and therefore represent a late stage or post-erosional phase of volcanism. The westernmost studied seamount, Seamount 6 (informal name), has an elongate WNW–ESE striking base (Fig. 3b). The steep-sided ediﬁce is topped by a ﬂat plateau, on which several well-preserved small volcanic cones are scattered, rising up to 200 m above the plateau. Seamount 9 (informal name) is located about 45 km east of Seamount 6. One track was surveyed across Seamount 9 (not shown in Fig. 3), which revealed an oval shaped guyot and a c. 10 km long WNW–ESE-trending ridge emanating from its western base. This ridge is composed of several aligned small volcanic cones and interpreted as volcanic rift zone. Haxby Seamount (named by the ANT-XIII/4 cruise participants) (Fig. 3c), which has been mapped previously on RV Nathaniel B. Palmer Cruise in 1996 (Feldberg, 1997), has a slightly curvilinear volcanic rift system with numerous cones on its top emanating from the eastern ﬂank of the guyot and extending N 30 km to the east. Two less pronounced, c. 12–15 km long chains of cones and ridges emanating from the western ﬂank may be the western continuation of the volcanic rift. Hubert Miller Seamount (Fig. 3d) is located ~75 km ESE of Haxby Seamount. This seamount is the largest MBS with frequent small cones and ridges scattered along its ﬂanks but infrequent on the plateau. Several up to 8 km long volcanic rift zones extend from the base of Hubert Miller Seamount. The easternmost mapped seamount, Seamount C (informal name; Fig. 3e), is the smallest of the studied volcanoes. Its guyot-shaped ediﬁce has a crudely circular base and a plateau of ~7 km diameter. Volcanic rifts extend from the base in northern and southern directions and NNE–SSW trending, curvilinear graben and ridge structures are adjacent to its eastern ﬂank. The existence of further, most likely sediment covered, volcanic cones and ridge-like basement structures between the main MBS cluster and Marie Byrd Land are predicted from satellite gravity data (Smith and Sandwell, 1997) and observed in seismic data (Gohl et al., 1997b; Uenzelmann-Neben and Gohl, 2012). The original volume of MBS magmatism, however, remains unclear because of incomplete data and the largely unknown initial volume of the eroded islands. Based on the available bathymetric data (multi-beam and satellite gravity), the total volume of all present MBS can roughly be estimated to more than 20,000 km3. The aerial extent of the former Marie Byrd Islands were similar in size to Canary Islands, such as La Palma (compared to Hubert Miller Seamount) or El Hierro (compared to Table 1 Morphological features of the Marie Byrd Seamounts. Seamount 6 Seamount 9 Haxby Seamount Hubert Miller Seamount Seamount C Coordinates (center) Shape Secondary features 69°47′S, 126°17′W Oval shaped guyot Small cones on ﬂanks and plateau 69°40′S, 124°45′W Oval shaped guyot Small cones on ﬂanks and plateau; WNW–ESE trending rift zone 69°17′S, 121°20′W Oval shaped guyot Small cones on ﬂanks and plateau; several rift zones, most of them ~SW–NE Base level (mbsl) Diameter at base (km) Water depth of plateau (mbsl) Ediﬁce height (m) Volume estimate (km3) Dredge samplesa 3000–2800 80 × 20 1600–1350 ~1650 ~2000 – 3600–3400 Long axis 25 1600–1400 ~2200 – – 69°07′S, 123°35′W Oval shaped guyot Small cones on ﬂanks and plateau; WNW–ESE trending major rift zone, minor W–E and WSW–ENE rifts 4000 30 1800–1600 ~2400 ~1600 PS69/317-1 69°12′S, 117°30′W Crudely circular guyot Small cones on ﬂanks and plateau; rift zones mainly trending from SSW–NNE to SSE–NNW 3500 17 2400–2200 ~1300 ~200 PS69/327-1 a A detailed description of dredge operations and recovered material is provided in Gohl (2007). 4000–3600 75 × 50 1600–1200 ~2800 ~8000 PS69/321-1 PS69/324-1 PS69/325-1 1664 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Fig. 3. (a) Overview of the Marie Byrd Seamount Province. Red arrows mark the MBS surveyed during R/V Polarstern cruise ANT-XXIII/4 in 2006, letters in bold signify those which have been successfully dredged. Predicted bathymetry is after Smith and Sandwell (1997). (b) Multi-beam bathymetry of the eastern part of Seamount 6. This is the westernmost studied during ANT-XXIII/4 and has not been mapped before. It appears to be one of the largest MBS. (c) Haxby Seamount (named by the ANT-XIII/4 cruise participants) has completely been mapped on the R/V N.B. Palmer cruise in 1996 and morphologically studied in detail by Feldberg (1997). (d) Combined ANT-XVIII/5a (2001) and ANT-XXIII/4 multi-beam bathymetry of Hubert Miller Seamount. This Seamount is located ~40 nm ESE of Haxby and appears to be the largest of the MBS. (e) Combined ANT-XVI/3 (Miller and Grobe, 1996) and ANT-XXIII/4 multi-beam bathymetry of Seamount C. This seamount does not appear in the bathymetric maps derived from satellite gravity data. Note that Seamount C differs in size, high, and morphology from the other surveyed MBS guyots. The red dots with numbers mark dredge station of cruise ANT-XXIII/4 which yielded in situ volcanic rocks. Seamount 6), which are believed to be the product of a mantle plume (e.g., Montelli et al., 2006). The submarine base of Peter I Island was only partially surveyed prior to ANT-XXIII/4 and, except for dredge hauls directly off the coast of the island (Broch, 1927), un-sampled (Fig. 4). The island is elongated in N–S direction and represents the top of a large volcano, which measures ~ 65 km in diameter at its base and rises from the abyssal plain at ~3500–4000 m to an elevation of 1640 m above sea A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 1665 Fig. 4. Multi-beam bathymetry of the base of Peter I Island. The map reveals several small cone- and ridge-like structures on its ﬂanks and a steep canyon at its eastern side which most likely has been formed by slope failure or sector collapse. The red dots indicate locations sampled during cruise during R/V Polarstern cruise ANT-XXIII/4 in 2006 (244 — dredge station at the eastern base of Peter I Island; PI 1–4 — subaerial samples taken on Michajlovodden Peninsula). level. Volcanic rifts emanate from the submarine ﬂank of the island mainly in northern and southern directions. A striking feature of Peter I Island is a c. 9 km wide depression in the eastern ﬂank of its base, which most likely has been formed by a major slope failure or sector collapse. 4. Sample background Volcanic rocks were recovered at ﬁve dredge stations of the main MBS ediﬁces and associated small cones (Fig. 3b–e). In addition a single dredge haul along the northeastern submarine ﬂank of Peter I Island has been carried out (Fig. 4). At all dredge sites discussed here, the angular shape of the rocks, freshly broken surfaces and homogeneity of rock types within a single dredge were taken as evidence for an in-situ origin (and non-ice rafted) of the rocks. Our samples represent the ﬁrst in-situ volcanic rocks recovered from the MBS. A detailed description of dredge operations and recovered material is provided in chapter 7 of Gohl (2007). At Haxby Seamount, dredge haul PS69-317-1 from the upper southern slope beneath the plateau edge contained freshly broken carbonate cemented breccias, which consist of aphyric basaltic clasts up to 8 cm in size (Fig. 5a). At Hubert Miller Seamount, three dredges yielded mainly lava fragments; dredge PS69-321-1 along a steep slope below the SE plateau edge gave olivine (ol)–clinopyroxene (cpx)–phyric lava Fig. 5b, dredge PS69-324-1 at the lower SE slope beneath a cone like structure provided dense feldspar (fsp)–cpx–phyric basalt lava and (carbonate) cemented Mn-encrusted volcanic breccia, and dredge PS69-325-1 obtained vesicular fsp–phyric lava from the upper southern ﬂank. At Seamount C, vesicular ol–fsp–phyric and dense fsp–phyric pillow fragments (Fig. 5c) were dredged from a cone on the lower western ﬂank. At Peter I Island, a 150 m high ridge located in ~ 1800 m water depth on the NE slope of the volcano was dredged (PS69-244-1). The rocks are predominantly vesicular pillow and sheet ﬂow lava fragments (Fig. 5d). Both are feldspar (fsp)–phyric and have up to 1 cm thick, fresh, glassy rims. Vesicles are generally unﬁlled and only a few glassy surfaces show early stages of palagonitization. The subaerial samples from Peter I Island were taken at the Michajlovodden Peninsula (Fig. 4). They comprise vesicular lava (up to 15% vesicles; sample PI-1), aphyric agglutinates of a N 1.5 m thick, partially red oxidized layer outcropping in the northern part of the peninsula (sample PI-3), and part of a reddish volcanic bomb with 10–20% vesicles (sample PI-4). Belgica Seamount is the easternmost ediﬁce of the De Gerlache Seamount group. It is guyot-shaped and has a N–S elongated base diameter of c. 60 × 90 km with a ﬂat-topped summit at c. 400–500 m below sea level. Belgica was dredge-sampled during Polarstern cruise ANTXII-4 in 800 to 600 water depths (Hagedorn et al., 2007). Hagedorn et al. (2007) initially determined K–Ar ages and major and trace element geochemistry on the recovered samples. Here we complement the existing data with Sr–Nd–Pb isotope data on a subset of newly prepared sample material. 5. Petrography and rock classiﬁcation The petrography of the MBS volcanics is quite uniform being slightly phyric with a few large phenocrysts of altered olivine and zoned plagioclase in a ﬁne-grained groundmass of olivine, plagioclase and clinopyroxene. Occasionally, ilmenite and magnetite occur as accessory phases. Olivine is commonly altered to iddingsite and the latter is sometimes replaced by calcite. The groundmass is variably altered by low temperature processes ranging from hydrated glass at Haxby Seamount to replacement by secondary minerals such as zeolite and dolomite at Hubert Miller Seamount and Seamount C. The altered state of the MBS volcanic rocks is also manifested in elevated H2O contents of up to 2 wt.% in most samples, except that samples from Dredge 324 at Hubert Miller Seamount have b1 wt.% H2O and those from Seamount C have 3 wt.% H2O (Table 2). CO2 contents are generally low (b0.3 wt.%) and only two samples show slightly elevated CO2 N 0.5 wt.%, due to secondary carbonate. Unusually high phosphorous contents were detected in 5 samples (marked with a in Table 2) and are interpreted to reﬂect the presence of secondary phosphate that is, however, not detected in thin section. Only samples 1666 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Fig. 5. Basaltic rocks dredged at the MBS and the submarine base of Peter I Island. (a) Typical monomict breccia from Haxby Seamount composed of dense, aphyric irregular-shaped and angular coarse lapilli set in a carbonaceous matrix (PS69/317-1). (b) Dense ol–cpx–phyric basaltic lava fragment from Hubert Miller Seamount, note angularity and freshly broken surfaces of the sample (PS69/321-1). (c) Dense fsp–phyric pillow fragment of Seamount C (PS69/327-1). (d) Sheet lava ﬂow fragment with fresh, 1 cm thick glassy rim from the submarine base of Peter I Island (PS69/244-1). with P2O5 ≤ 1 wt.% are considered meaningful when treating the major element chemistry further below. The submarine samples of Peter I Island are slightly porphyric with zoned plagioclase laths and small, homogeneously distributed clinopyroxene phenocrysts, set in a microcrystalline groundmass of clinopyroxene and plagioclase. Magnetite occurs as an accessory mineral and fresh glass is common. The subaerial volcanics of Peter I Island are more aphyric than those from the submarine ﬂanks. The ﬁne-grained crystalline groundmass of these samples contains pyroxene, plagioclase and possibly glass, and accessory minerals of magnetite, ilmenite and hematite. All samples from Peter I Island are generally very fresh as manifested by low H2O (0.3–0.9 wt.%) and CO2 (b 0.06 wt.%) contents (Table 2). The silica content of the entire sample suite ranges from 53.1 to 45.7 wt.% SiO2. On a total alkali vs. silica diagram (TAS; Fig. 6), the majority of the samples plots above the alkalic–sub-alkalic division line and are classiﬁed as basalts, trachybasalts and basaltic trachyandesites. All but one of MBS samples lie along an alkali basaltic differentiation trend. The samples from the submarine ﬂank of Peter I Island are tholeiitic basalts (SiO2 ~ 49 wt.%; Na2O + K2O 4.1–4.3 wt.%), whereas the subaerial samples are slightly more alkaline transitional tholeiites (SiO2 ~ 47 wt.%; Na2O + K2O 4.3–4.5 wt.%). 6. Analytical results 6.1. 40Ar/39Ar age dating The 40Ar/39Ar age dating results are summarized in Table 3. Age and alteration index spectra (based on the measured 36Ar/37Ar ratios after Baksi, 2007) are shown in Fig. 7. A detailed description of the methods and the full analytical data are provided in Appendix 1. Glasses from two hyaloclastite breccia samples at Haxby Seamount yield plateau ages of 64.2 ± 0.9 Ma (317-1-1gls) and of 62.3 ± 0.4 (317-1-2gls) and 61.2 ± 0.5 Ma (317-1-2gl2), slightly outside of the two sigma analytical errors. Alteration indices are relatively high even in the plateau sections (0.001 to 0.01), reﬂecting partial hydration of the basalt glass and uptake of atmospheric 36Ar. Three samples of porphyric lava from Hubert Miller Seamount yield plagioclase step-heating plateau ages of 56.7 ± 1.9 Ma (321-1-2), 56.5 ± 0.6 Ma (325-1-2B) and 57.0 ± 0.9 Ma (321-1-5). Alteration indices are high in the low-temperature heating steps indicating partial alteration of the feldspars, but systematically low in the plateau steps (b0.0002) indicating degassing from little or un-altered sites. Matrix step-heating analyses from the same rock samples yield plateau age results within error of the feldspar step-heating results (321-1-2: 58.9 ± 0.6 Ma; 321-1-5: 55.7 ± 0.5 Ma), but are considered inferior with respect to scatter and alteration effects. The matrix step-heating analysis of aphyric basalt lava sample 324-1-3, in contrast, yields a plateau age of 3.0 ± 0.5 Ma. Alteration indices are high in the plateau section (0.003 to 0.01), possibly indicating a partial loss of radiogenic 40Ar. Nevertheless, the analysis shows that Hubert Miller Seamount comprises both Paleocene and Pliocene lavas. The least-altered aphyric lava sample from Seamount “C” (327-1-2) yields a low-probability plateau age of 58.7 ± 0.8 Ma, with intermediate plateau-step alteration indices (0.002 to 0.008). Fresh basaltic glass from Peter I Seamount yields plateau steps alteration indices of b0.0009 (244-1-1) and b 0.0001 (244-1-3), and plateau ages of 1.9 ± 0.3 Ma, 1.7 ± 0.3 Ma respectively. 6.2. Major and trace elements A total of 19 samples from the MBS and Peter I Island were analyzed for major and trace elements compositions and the results are Table 2 Results of major and trace element analyses. PS69/317-1-1 Haxby Smt a 46.13 3.39 16.33 14.81 0.21 3.46 7.72 3.82 1.54 0.91 1.91 0.06 100.29 28.1 330 4.79 1.52 70.3 4.21 49.6 106 2.75 13.1 55.6 760 12.1 8.06 404 3.79 10.8 1.45 7.98 1.43 42.3 3.59 0.457 3.13 0.450 PS69/317-1-2 Haxby Smt PS69/321-1-2 Hubert Miller Smt PS69/321-1-4 Hubert Miller Smt PS69/321-1-5 Hubert Miller Smt PS69/321-1-12a Hubert Miller Smt PS69/324-1-3 Hubert Miller Smt PS69/324-1-4 Hubert Miller Smt PS69/324-1-6 Hubert Miller Smt 28.1 330 4.72 1.52 70.6 4.18 48.9 106 2.35 13.1 55.2 756 12.1 8.02 402 3.75 10.7 1.47 7.89 1.46 42.0 3.72 0.479 3.14 0.447 42.03 3.57 18.2 15.91 0.21 3.41 9.02 3.51 0.74 1.44a 2.38 0.17 100.59 6.66 264 4.60 0.427 68.1 4.27 64.4 114 2.41 14.4 57.6 1219 13.3 8.35 413 4.14 11.4 1.56 8.36 1.54 51.6 3.84 0.475 3.84 0.554 50.22 1.89 17.15 11.49 0.22 2.58 6.39 4.57 2.61 0.73 2.23 0.30 100.38 53.4 725 6.89 1.75 123 6.60 76.8 154 3.81 17.0 63.2 825 12.6 8.83 475 3.92 10.2 1.40 7.50 1.33 43.6 3.47 0.455 3.54 0.496 46.23 3.59 17.59 14.19 0.19 2.85 8.84 3.54 2.09 0.92 1.80 0.08 101.91 37.3 462 4.57 0.816 78.5 4.59 52.4 106 2.44 12.2 46.0 969 9.86 6.36 323 3.09 8.00 1.05 5.40 0.97 32.5 2.49 0.310 2.57 0.359 45.95 3.44 16.62 13.95 0.22 3.96 8.44 3.53 2.05 0.79 1.86 0.07 100.88 50.4 493 4.43 0.695 77.2 4.45 51.4 103 2.65 11.6 44.3 925 9.49 5.94 317 2.91 7.45 0.979 5.04 0.92 31.5 2.31 0.291 2.45 0.332 50.19 2.09 18.46 11.09 0.17 2.07 6.76 4.53 2.15 0.87 2.10 0.07 100.55 23.9 784 7.31 0.296 119 6.97 65.9 146 3.37 16.1 61.6 891 12.5 9.36 446 4.12 9.66 1.29 6.58 1.14 35.5 2.79 0.345 2.76 0.366 45.80 2.60 15.02 13.05 0.19 8.13 10.65 3.29 1.04 0.48 0.74 0.13 101.12 24.1 311 2.99 0.743 46.0 2.78 32.1 66.2 2.47 7.74 30.1 626 6.69 4.30 205 2.19 5.89 0.814 4.38 0.787 25.2 2.07 0.261 2.08 0.278 52.85 0.85 17.18 8.63 0.18 5.52 8.68 3.99 1.45 0.21 0.98 0.07 100.59 35.8 455 4.72 0.819 78.5 4.60 53.1 107 2.30 12.3 46.5 967 9.97 6.35 321 3.04 8.10 1.02 5.49 0.971 31.6 2.50 0.325 2.49 0.350 46.74 3.26 14.86 14.73 0.20 4.87 10.31 3.65 1.24 0.54 0.82 0.08 101.30 28.0 320 3.91 0.982 56.6 3.39 38.3 78.4 1.89 8.98 35.5 568 8.08 5.55 261 2.49 7.17 1.02 5.74 1.02 31.8 2.66 0.350 2.65 0.364 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 SiO2 TiO2 Al2O3 Fe2O3 MnO MgO CaO Na2O K2O P2O5 H2O CO2 Total Rb Ba Th U Nb Ta La Ce Pb Pr Nd Sr Sm Hf Zr Eu Gd Tb Dy Ho Y Er Tm Yb Lu PS69/317-1-1 Replicate ICP-MS Results with unusual high values not shown in Fig. 6. 1667 1668 PS69/325-1-2b Hubert Miller Smt PS69/327-1-1 Seamount C PS69/327-1-2 Seamount C PS69/244-1-1 Peter I submarine PS69/244-1-3 Peter I submarine PS69/244-1-5 Peter I submarine PS69/PI-1 Peter I subaerial PS69/PI-3 Peter I subaerial PS69/PI-4 Peter I subaerial 47.72 2.52 16.78 11.58 0.14 2.54 7.93 3.88 2.74 1.69a 2.12 0.59 100.23 90.1 469 6.93 1.34 95.6 5.38 73.5 141 3.38 16.3 62.4 674 12.9 7.99 412 3.59 10.9 1.48 7.39 1.35 48.7 3.50 0.458 3.63 0.515 47.72 2.49 17.24 12.02 0.14 1.34 7.40 4.19 3.24 2.12a 2.13 0.18 100.21 57.3 480 7.02 1.33 96.8 5.44 73.8 144 3.21 16.3 62.3 717 13.3 8.02 417 3.56 10.9 1.41 7.51 1.36 46.1 3.52 0.449 3.54 0.506 41.98 2.26 18.15 8.92 0.08 0.97 12.84 3.55 1.74 5.02a 2.92 0.56 98.99 27.2 404 5.21 1.78 55.6 3.48 66.2 80.4 3.07 10.4 39.0 811 8.09 4.99 250 2.45 7.78 1.07 6.32 1.33 66.5 3.71 0.509 4.21 0.674 46.00 2.70 20.31 11.30 0.15 1.24 8.73 3.44 1.38 1.62a 3.03 0.24 100.14 21.5 427 5.06 1.43 58.7 3.58 44.6 84.5 2.24 9.36 35.6 843 7.89 5.37 258 2.49 6.87 0.957 5.30 0.979 32.0 2.70 0.349 2.72 0.377 49.48 2.79 13.68 12.58 0.14 8.18 9.00 3.17 1.16 0.50 0.86 0.05 101.59 20.3 240 2.85 0.887 30.2 1.88 27.2 58.5 2.30 7.49 32.7 624 7.57 5.22 233 2.43 6.78 0.874 4.69 0.792 22.5 1.83 0.226 1.53 0.192 49.09 2.74 13.26 12.55 0.15 9.2 8.68 3.03 1.08 0.51 0.87 0.05 101.21 19.8 228 2.68 0.725 29.4 1.83 26.7 57.1 2.26 7.20 30.8 622 7.54 5.19 230 2.48 6.55 0.882 4.54 0.737 21.6 1.72 0.207 1.50 0.193 49.26 2.75 13.61 12.85 0.14 8.55 8.91 3.00 1.15 0.48 0.94 0.06 101.70 19.2 226 2.57 0.704 28.9 1.76 26.1 55.6 2.22 6.95 29.2 632 7.48 5.03 227 2.39 6.38 0.854 4.40 0.706 21.6 1.73 0.204 1.49 0.186 47.79 3.53 12.79 13.48 0.15 9.64 9.03 3.07 1.37 0.64 0.28 0.04 101.81 23.9 308 3.53 0.920 48.5 2.82 38.0 80.9 2.26 10.2 43.1 797 9.80 7.03 315 3.14 8.50 1.14 5.70 0.924 24.9 2.14 0.255 1.58 0.206 47.26 3.48 12.52 13.36 0.15 9.86 8.83 3.03 1.42 0.69 0.45 0.04 101.09 26.5 315 3.74 0.984 48.7 2.89 40.3 85.2 2.46 10.4 43.6 804 10.2 7.42 326 3.21 8.63 1.13 5.52 0.894 24.9 2.05 0.246 1.54 0.209 48.15 3.46 12.73 13.03 0.14 9.42 8.67 3.17 1.47 0.78 0.38 0.04 101.44 26.3 332 3.78 1.02 50.6 2.90 43.1 92.4 2.31 11.4 48.5 871 11.0 7.59 344 3.52 9.61 1.19 5.85 0.963 25.7 2.14 0.256 1.56 0.203 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 SiO2 TiO2 Al2O3 Fe2O3 MnO MgO CaO Na2O K2O P2O5 H2O CO2 Total Rb Ba Th U Nb Ta La Ce Pb Pr Nd Sr Sm Hf Zr Eu Gd Tb Dy Ho Y Er Tm Yb Lu PS69/325-1-2a Hubert Miller Smt A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 1669 The new trace element data from Peter I Island also display trace element patterns similar to OIB (Fig. 8b) that compares well with the data of Prestvik et al. (1990) and Hart et al. (1995). Overall the subaerial samples are slightly more enriched in incompatible elements than the submarine sample but show similar HREE abundances (Fig. 8d), which is consistent with lower degrees of melting for the subaerial lavas. In contrast to the MBS, Peter I Island samples are slightly less enriched in LREE and the most incompatible elements (Rb through Ta) and show lower La/Sm ratios (Fig. 9a), indicating higher degrees of partial melting than observed for the MBS. Notably the LREE are more strongly enriched relative to the HREE through a more pronounced HREE depletion. The higher (Sm/Yb)n ratios of the Peter I Island melts suggest that their source had a higher garnet content (Fig. 9b). 6.3. Sr–Nd–Pb–Hf isotopes Fig. 6. Total alkali versus SiO2 diagram illustrating the alkali basaltic to basaltic trachyandesitic composition of most samples from MBS, Peter I Island, and Belgica Seamount. Subdivision between alkalic and sub-alkaline rock suites is after Irvine and Baragar (1971). All data are normalized to a 100% volatile free basis. Samples displaying unusual high phosphor contents are not shown in this diagram (cf. Table 2). Tb — Trachybasalt. shown in Table 2. Descriptions of methods and uncertainties are given in Appendix 1. A full table with sample locations, radiometric ages and geochemical data is provided in Table A4 of the Appendix. The majority of MBS samples are fairly evolved (8 to 2 wt.% MgO), whereas samples from Peter I Island are more primitive and cluster between 10 and 8 wt.% MgO. Al2O3 shows a good negative correlation with decreasing MgO, suggesting fractionation of pyroxene and olivine. In the most evolved MBS lavas (b3 wt.% MgO), FeOt and TiO2 signiﬁcantly decrease which may reﬂect fractionation of ilmenite in late stage melts. Subaerial and submarine samples of Peter I Island exhibit small compositional differences. The submarine samples have higher SiO2 and Al2O3 and slightly lower MgO, FeOt and TiO2 contents than the subaerial samples. Trace elements patterns of the MBS are typical for ocean islands basalts (OIB; Fig. 8a) with characteristic troughs for Pb and K and strong enrichments for Nb and Ta relative to primitive mantle. The Nb and Ta enrichments are most pronounced in samples from Hubert Miller Seamount while Haxby Seamount and Seamount C are less enriched in the most incompatible elements (Rb, Ba and Th). All MBS samples show strong enrichment of the light REE (LREE) relative to the heavy REE (HREE) (see Fig. 8c), suggesting small degrees of partial melting while differentiation of the HREE indicates melting within the garnet stability ﬁeld (N70–80 km). Sr–Nd–Pb–Hf isotopic ratios of representative samples from the MBS, Belgica Seamount, and Peter I Island are shown in Table 4. Descriptions of analytical methods and accuracy along with initial isotopic ratios are given in Appendix 1 and Table A4. Figs. 10 and 11 compare the new MBS, Peter I Island and Belgica Seamount isotope data with data of West Antarctic volcanic rocks, related to the WARS (for data sources see ﬁgure captions) and the Hikurangi Seamounts (Hoernle et al., 2010). Excluding two samples with anomalously high 87 Sr/86Sr isotope ratios that may have been affected by seawater alteration, the MBS samples form a crude negative array on the Sr–Nd isotope diagram (Fig. 10). The samples from Seamount C have the most radiogenic Nd and least radiogenic Sr isotope ratios and fall between Paciﬁc MORB and the high 238U/204Pb (HIMU) mantle endmember. Samples from Hubert Miller Seamount have the least radiogenic Nd isotope ratios and trend vaguely towards an enriched mantle (EM) type component (Fig. 10). The Belgica samples plot within the Paciﬁc MORB ﬁeld and the Peter I Island samples lie within the published ﬁeld for this island (Prestvik et al., 1990; Hart et al., 1995) and are displaced to slightly more radiogenic Sr and less radiogenic Nd isotope ratios i.e. to faintly more EM ﬂavored compositions than the majority of Hubert Miller Seamount samples. In Pb–Pb isotope space (Fig. 11a), the MBS volcanic rocks do not form a simple two component mixing array as the majority of samples extends from a HIMU-type component with radiogenic Pb towards enriched mantle one (EMI) while two samples having signiﬁcantly lower 207Pb/204Pb which displaces them towards the extension of the Paciﬁc MORB ﬁeld. Sample 324-1-4 from Hubert Miller Seamount has the least radiogenic Pb composition of all MBS and plots above the Paciﬁc MORB ﬁeld away from the main MBS array while samples from Haxby Seamount possess the most radiogenic Pb composition. The Belgica Seamount samples plot near the unradiogenic end of the main MBS ﬁeld in Pb–Pb isotope space (Fig. 11a) but possess more radiogenic 143Nd/144Nd compositions than the MBS (Fig. 11b). The majority of MBS samples and all Belgica Table 3 40 Ar/39Ar step heating analyses results. Seamount Sample ID PS69- Analysis ID Dated material Plateau ± 2 sigma age (Ma) 39 Haxby 317-1-1 317-1-2 317-1-2 321-1-2 321-1-2 321-1-5 321-1-5 325-1-2B 324-1-3 327-1-2 244-1-1 244-1-3 gls gls gl2 fss mx2 fss mx2 fss mxs mx2 gls gl2 Glass Glass Glass Plag Matrix Plag Matrix Plag Matrix Matrix Glass Glass 64.7 62.3 61.2 56.7 58.9 57.0 55.7 56.5 3.0 58.7 1.9 1.7 63.1 56.8 72.9 73.9 95.6 58.4 63.5 61.1 84.6 55.7 83.8 96.4 Hubert Miller “C” Peter Ia a Dated samples from Peter I Island are dredge samples from its submarine base. ± ± ± ± ± ± ± ± ± ± ± ± 0.8 0.4 0.5 1.9 0.6 0.9 0.5 0.6 0.5 0.8 0.3 0.3 Ar fraction MSWD Probability 0.95 0.99 1.19 0.61 1.10 1.30 0.78 1.20 1.30 1.90 1.03 0.58 0.46 0.44 0.27 0.72 0.34 0.23 0.45 0.30 0.21 0.04 0.41 0.87 1670 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Table 4 Sr-Nd-Pb-Hf-Isotope analyses. 87 2 sigma 143 0.703093 0.704186 0.703384 0.703277 0.703335 0.703230 0.703094 0.704027 0.704043 0.703108 0.000003 0.000003 0.000003 0.000002 0.000002 0.000003 0.000003 0.000003 0.000003 0.000003 0.703502 0.703417 0.702888 0.702805 0.702831 De Gerlache Smts. PS-2693-1_(1)e PS-2693-1_(2)e PS-2693-1_(3)e PS-2693-1_(4)e PS-2693-1_(5)e PS-2693-1_(6)e Peter I Island PS69/244-1-1 PS69/244-1-3 PS69/PI-1f PS69/PI-3 PS69/PI-4 Marie Byrd Smts. PS69/317-1-1 PS69/317-1-2 PS69/321-1-2 PS69/321-1-4 PS69/321-1-5 PS69/321-1-12a PS69/324-1-3 PS69/324-1-4 PS69/324-1-4d PS69/324-1-6 PS69/324-1-6d PS69/325-1-2a PS69/325-1-2b PS69/327-1-1 PS69/327-1-2 PS69/327-1-2d d Sr/86Sr Nd/144Nd 2 sigma 206 0.512885 0.512881 0.512798 0.512806 0.512816 0.512811 0.512906 0.512881 0.512870 0.512899 0.000003 0.000003 0.000003 0.000002 0.000003 0.000003 0.000003 0.000003 0.000003 0.000003 0.000003 0.000002 0.000003 0.000003 0.000003 0.512809 0.512817 0.512913 0.512918 0.512927 0.000003 0.000003 0.000006 0.000009 0.000004 0.703015 0.703023 0.703029 0.703029 0.702998 0.702994 0.000005 0.000005 0.000005 0.000004 0.000005 0.000005 0.512967 0.512966 0.512966 0.512957 0.512983 0.512975 0.703748 0.703752 0.703759 0.703837 0.703871 0.000003 0.000003 0.000005 0.000002 0.000002 0.512759 0.512750 0.512805 0.512782 0.512775 Pb/204Pb 2 sigma 207 20.7725 20.4116 20.2467 20.1075 20.1005 19.9595 19.6713 18.7063 18.7110 19.8645 19.8661 20.1218 20.0871 19.8152 19.8530 0.0015 0.0020 0.0016 0.0006 0.0008 0.0008 0.0011 0.0006 0.0006 0.0009 0.0013 0.0008 0.0008 0.0010 0.0008 0.000002 0.000003 0.000003 0.000003 0.000004 0.000003 19.8515 19.8376 19.8278 19.9238 19.7441 19.7447 0.000002 0.000003 0.000002 0.000002 0.000003 19.2360 19.2517 19.3013 19.3015 19.3244 Pb/204Pb 2 sigma 208 15.7739 15.7561 15.7229 15.7153 15.7160 15.7082 15.6140 15.6189 15.6184 15.6339 15.6331 15.7193 15.7180 15.6835 15.6903 0.0016 0.0022 0.0017 0.0006 0.0008 0.0009 0.0013 0.0007 0.0007 0.0009 0.0014 0.0006 0.0006 0.0010 0.0008 0.0021 0.0025 0.0032 0.0037 0.0025 0.0016 15.6878 15.6795 15.6804 15.6834 15.6641 15.6601 0.0015 0.0006 0.0012 0.0007 0.0007 15.7437 15.7520 15.7216 15.7409 15.7456 Pb/204Pb 2 sigma 176 Hf/177Hf 40.1472 39.9679 40.0393 39.9521 39.9815 39.9274 39.5216 38.4900 38.4973 39.7163 39.7173 39.9100 39.8863 39.2970 39.3447 0.0056 0.0075 0.0057 0.0018 0.0028 0.0028 0.0044 0.0020 0.0019 0.0027 0.0048 0.0016 0.0017 0.0031 0.0024 0.282871 0.282875 0.282879 0.282880 0.000005 0.000004 0.000003 0.000004 0.282875 0.282787 0.282877 0.000004 0.000007 0.000004 0.283003 0.000004 0.282862 0.000004 0.282927 0.000004 0.0018 0.0021 0.0027 0.0030 0.0024 0.0012 39.5057 39.4705 39.4777 39.4845 39.3550 39.3438 0.0049 0.0056 0.0063 0.0081 0.0069 0.0032 0.0017 0.0006 0.0010 0.0006 0.0006 39.3779 39.4162 39.3230 39.3865 39.4265 0.0056 0.0019 0.0026 0.0015 0.0014 0.282798 0.000003 2 sigma Replicate analyses. Sr/86Sr determined on MAT262 TIMS. Pb isotope ratios without Pb DS. e 87 f Seamount samples largely overlap with the ﬁelds of the West Antarctic volcanics and the Hikurangi Seamounts (Fig. 11). The Peter I Island samples overlap the published data from this island and have Pb isotope compositions near the enriched mantle two (EMII) component. The above mixing relations are also seen in co-variations of 206 Pb/204Pb versus 143Nd/144Nd (Fig. 11b) and εNd versus εHf (Fig. 12). On the Pb vs Nd isotope diagram, it is clear that at least three distinct components are required in the source of the MBS seamounts. Haxby Seamount has radiogenic Pb and intermediate Nd isotope ratios, similar to the HIMU mantle endmember. Seamount C and two Hubert Miller Seamount samples have less radiogenic Pb and intermediate Nd, trending toward Paciﬁc MORB (or depleted mantle = DM). The remaining Hubert Miller seamount samples except sample 324-1-4 have radiogenic Pb but the least radiogenic Nd, so that they are somewhat displaced toward EM like compositions. On the Nd–Hf isotope diagram, the MBS seamounts show a relatively restricted range in Nd but a large range in Hf isotope ratios that fall between Paciﬁc MORB (DM) and the HIMU and EM mantle endmembers. The Belgica Seamount samples have the most MORB-like compositions in Nd, but their 206Pb/204Pb isotopic compositions are more radiogenic than commonly found in MORB. The Peter I Island samples have a clear EMII-type isotope signal with respect to Pb while Sr, Nd and Hf isotopes are just EM indicative. 7. Discussion 7.1. Spatial distribution of Cenozoic volcanism in the Amundsen Sea and Bellingshausen Sea 40 Ar/39Ar dating of six samples from the MBS yielded Early Cenozoic ages ranging from 64 to 57 Ma. A clear spatial age progression between the three dated MBS is not observed. The oldest ages are from Haxby Seamount in the west (64–61 Ma) and clearly younger ages are from Hubert Miller Seamount to the east (57 Ma, three feldspar ages). Seamount C, the easternmost seamount, yielded an intermediate age (59 Ma), but this matrix age with a very low probability should be treated with caution. The Pliocene age of 3.0 ± 0.5 Ma determined for sample 324-1-3 was collected right beneath a small volcanic cone along the upper slope of Hubert Miller Seamount (Fig. 3d) and most likely represents the age of this cone. Similar cones are scattered on the plateau and slopes of all mapped MBS (cf. Fig. 3), indicating widespread and possibly long-lasting low volume post-erosional volcanism, as has been observed at other seamount provinces worldwide (e.g., Hoernle et al., 2004; Geldmacher et al., 2005; Hoernle et al., 2010). Assuming that the 40Ar/39Ar ages obtained at the three MBS are close (within a few million years) to the time when these islands were eroded and submerged below sea-level, a minimum subsidence rate can be calculated for each seamount taking the age and present water depth of the plateau of the seamount into account. Seamount C, the smallest and deepest ediﬁce, displays the highest subsidence rate of ~ 41 m/Ma if it is actually a guyot. In contrast, the larger Haxby and Hubert Miller Seamounts both yield minimum subsidence rates of ~ 28 m/Ma despite their apparent age difference of ~ 5 Ma. We note that the plateau edges of the westernmost Seamounts 6 and 9 lie at roughly similar water depth (1600–1350 m, Table 1) as observed for Haxby and Hubert Miller Seamounts (1800–1200 m), which in turn may indicate a comparable subsidence history provided similarities in lithospheric age and structure west of Haxby Seamount as well as analogous formation ages of 60 ± 5 Ma. The new ages (1.9 ± 0.3 Ma to 1.7 ± 0.3 Ma) for samples from the eastern submarine ﬂank of Peter I Island are signiﬁcantly older A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 1671 Fig. 7. Age spectra and alteration indices (A.I.) from 40Ar/39Ar laser step-heating experiments. Plateau steps and corresponding range of alteration index values are accentuated by gray shading. Stated errors are ±2σ. 1672 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Fig. 8. Multi-element diagram normalized to primitive mantle after Hofmann (1988) for (a) MBS and (b) Peter I Island samples. The trace elements patterns of all the samples are similar to those of ocean islands basalts (OIB). OIB and E-MORB patterns are after Sun and McDonough (1989). REE diagrams normalized to C1 are after McDonough and Sun (1995) for (c) MBS and (d) Peter I Island samples. Fig. 9. (Sm/Yb)n (n = normalized to primitive mantle after Hofmann, 1988) versus (a) La/Sm and (b) Zr/Hf ratios. Lower La/Sm ratios indicate slightly higher degrees of partial melting for Peter I Island and Belgica Seamount than for MBS. Residual garnet and pyroxene and/or eclogite in the magma source is indicated by high (Sm/Yb)n ratios of 2 to 8 and relatively high Zr/Hf ratios (N40), respectively. Zr/Hf ratios for depleted MORB mantle (32–40) are after Salters and Stracke (2004) and Workman and Hart (2005). Fig. 10. 87Sr/86Sr versus 143Nd/144Nd isotope correlation diagram for MBS, Peter I Island, and Belgica Seamount samples. Symbols are as in Fig. 9. The ﬁeld for West Antarctic volcanics is deﬁned by data of the WARS (Rocholl et al., 1995; Rocchi et al., 2002a), the Jones Mountains in Ellsworth Land (Hart et al., 1995), and the Marie Byrd Land Volcanic Province (Hart et al., 1997; Panter et al., 1997, 2000), which extends along the Paciﬁc margin of Marie Byrd Land. Most authors consider the volcanism at the Marie Byrd Land Volcanic Province and Jones Mountains as related to the WARS (e.g., Hart et al., 1995, 1997; Panter et al., 2000). The ﬁeld for the Hikurangi Seamounts is based on data by Hoernle et al. (2010), published data for Peter I Island comprise analyses of subaerial basaltic lavas from Prestvik et al. (1990) and Hart et al. (1995). The ﬁeld for Cretaceous volcanics of New Zealand is based on Tappenden (2003); Panter et al. (2006) and McCoy-West et al. (2010). HIMU, EM I, and EM II are after Zindler and Hart (1986) and Hart et al. (1992). Fields for Paciﬁc MORB are from PetDB (http:// www.earthchem.org/petdb) based on analyses of fresh glass. A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 1673 since ~ 2 Ma at Peter I Island. The three seamount/ocean island groups are spatially arranged in a highly elongated triangle with the MBS lying at its western tip and the De Gerlache Seamounts and Peter I Island forming the eastern limit (Fig. 1). The age distribution neither shows a correlation with spatial distribution nor a correlation with the age of the underlying ocean crust (e.g., Eagles et al., 2004a). A relationship between ages and plate motion cannot be observed, because neither the relative motion between the Bellingshausen Plate and the Antarctic Plate nor the absolute plate motion of the Antarctic Plate was signiﬁcant for this time period (Eagles et al., 2004a,b; Wobbe et al., 2012; Doubrovine et al., 2012). Therefore, the irregular spatial distribution of seamount ages in the Amundsen Sea and Bellingshausen Sea indicates that this magmatism occurred at distinct time intervals in spatially conﬁned areas. This observation excludes an origin through a single stationary hotspot sensu Morgan (1971). Instead this regional age pattern of intraplate volcanism favors the presence of three melting anomalies independent in space and time. Before we explore possibilities of non-plume related intraplate volcanism, we will ﬁrst brieﬂy reiterate geochemical constraints on the origin of the magma sources. 7.2. Geochemical constraints on the origin of seamount magmatism Fig. 11. (a) 206Pb/204Pb versus 207Pb/204Pb, and (b) 143Nd/144Nd isotope correlation diagrams for MBS, Peter I Island, and Belgica Seamount samples. Symbols and data sources are as in Fig. 10. than earlier published K–Ar ages (327 ± 88 ka to 111 ± 36 ka [1 sigma errors], Prestvik and Duncan, 1991) obtained on subaerial samples, which suggests that the fresh pillow glasses belong to an earlier submarine phase of this volcano. Together with published Upper Miocene K–Ar ages for the Belgica Seamount (20–23 Ma, Hagedorn et al., 2007), the three seamount/ ocean island volcanic provinces of the Amundsen and Bellingshausen Sea appear to have formed at distinct age intervals of 64–57 Ma for the MBS, at ~ 22 Ma for the De Gerlache Seamounts and at least Fig. 12. εNd versus εHf isotope correlation diagram for MBS and Peter I Island samples. Figure modiﬁed after Geldmacher et al. (2003), symbols and data sources for Hikurangi Seamounts and Paciﬁc MORB are as in Fig. 10. The New Zealand Cretaceous ﬁeld includes Nd values by Tappenden (2003) and Hf values analyzed by Timm et al. (2010) for the Mandamus Complex, as well as data from McCoy-West et al. (2010) for Lookout Volcanics. Lavas of all three seamount provinces (MBS, De Gerlache and Peter I Island) display a strong enrichment of the LREE relative to the HREE (Fig. 8c + d), clearly indicating partial melting in the presence of garnet. Likewise (Sm/Yb)N, (Gd/Yb)N and (Dy/Yb)N are all N 1 which, is consistent with residual garnet in the source (cf. Fig. 9b). Furthermore, the slight enrichment of Zr relative to Hf on the mantle-normalized plot (Fig. 8a + b) is also consistent with residual garnet (Hauri et al., 1994). Consequently, melt segregation in all three areas must have occurred in the garnet stability ﬁeld N60–80 km or 40–50 km if garnet pyroxenite was in the source (Hirschmann and Stolper, 1996). High Zr/Hf (43–54, Fig. 9b), Nb/Ta (16–19) and low Zr/Sm (31–38) provide additional support for partial melting of eclogite/garnet pyroxenite (i.e. recycled ocean crust), rather than garnet peridotite, consistent with a HIMU component in the mantle source. The isotopic signatures of MBS volcanic rocks are consistent with the presence of a HIMU-type mantle component in the source of these rocks (Figs. 10–12). The extremely radiogenic 206Pb/204Pb of the HIMUendmember requires a high 238U/204Pb in the source; a component unlikely to develop in signiﬁcant amounts within the convecting upper oceanic mantle without crustal recycling (see Stracke, 2012 for a recent review). HIMU is classically thought to reﬂect deep mantle recycling of oceanic crust by mantle plumes, ascending from deep in the mantle (Hofmann and White, 1982; White, 2010), however, from the lack of clear indications for the long-term existence of a classical mantle plume in the Amundsen Sea it is clear that alternative mechanisms are required to explain the occurrence of HIMU-type intraplate volcanism in this area, as has also been proposed for HIMU-type volcanic rocks in New Zealand (Hoernle et al., 2006). Subaerial and submarine samples of Peter I Island exhibit small compositional differences with the submarine samples having higher SiO2 and Al2O3 and slightly lower MgO, FeOt and TiO2 contents than the subaerial samples. The slight differences in MgO, FeOt, Al2O3 and TiO2 between submarine and subaerial lavas could be related through fractionation of olivine, pyroxene and possibly ilmenite from a subaerial melt composition, but this scenario cannot explain the higher incompatible element abundances in the subaerial lavas. Along with the slightly more alkaline character of the subaerial lavas in our sample set, the data indicates that the subaerial lavas could reﬂect slightly lower degrees of mantle melting, which would also explain their higher incompatible element abundances. Variations in the extent of partial melting are common during the life cycle of ocean island volcanoes with more alkaline compositions of lavas during the subaerial stage compared to less alkaline (tholeiitic) compositions during the 1674 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 submarine shield stage (e.g., Frey et al., 1990). Even during the submarine stage, short-term variations in the degree of partial melting have been observed at Loihi Seamount in the Hawaiian Islands (Garcia et al., 1993). The Pb isotopic composition of lavas from Peter I Island carries a clear EM II source signal (Fig. 11a) that is commonly thought to reﬂect contributions from pelagic sediments or upper continental crust (e.g., Zindler and Hart, 1986; Willbold and Stracke, 2010). The maﬁc composition of Peter I Island lavas, negative Pb anomalies and high Ce/Pb (~ 25 in submarine samples, 34–40 in subaerial samples) argue against shallow AFC processes such as sediment assimilation or preferred leaching of sedimentary Pb. This conclusion is similar to that of Hart et al. (1995), who explain the high 207Pb/204Pb signature of Peter I Island melts as evidence for the involvement of a mantle plume with EM II characteristics. We also note that the majority of global pelagic sediments have lower 206Pb/204Pb and 207Pb/204Pb ratios than observed in the Peter I Island lavas and thus involvement of modern pelagic sediment seems less likely. This is consistent with the Hf–Nd isotope ratios, which show that marine sediments did not inﬂuence the submarine sample of Peter I Island (Fig. 12). Ce/Pb lying within (or slightly above) the canonical array of 25 ± 5 for global OIB and MORB (after Hofmann et al., 1986) provides additional evidence for derivation from oceanic mantle rather than involvement of continental crust, which has Ce/Pb of 3–5. The solitary location of Peter I Island suggests that magmatism is related to a localized upwelling of EMII-like mantle but it is unclear whether this is connected to a blob rising from a thermal boundary such as the SW Paciﬁc superswell or melting of a continental raft that drifted into the oceanic upper mantle during the ﬁnal Gondwana breakup. In summary, Cenozoic intraplate volcanism in the Amundsen Sea and Bellingshausen Sea requires involvement of depleted MORB mantle in the source with signiﬁcant contributions of enriched components of HIMU and EM afﬁnity. Due to the lack of clear evidence for the existence of a mantle plume in this region, a model is needed to explain the evidence for enriched (plume like) components in the source of Amundsen Sea intraplate volcanism and a non-plume related process to accomplish adiabatic mantle melting in an intraplate environment. 7.3. Origin of the HIMU component in non-hotspot related Southwest-Paciﬁc and Antarctic volcanic provinces Alkalic volcanism with HIMU-like incompatible-element and isotopic signatures, similar to the samples from MBS, is reported from numerous locations throughout the SW Paciﬁc and West Antarctica. These include the Chatham Rise, Hikurangi Seamounts, intraplate volcanic ﬁelds in New Zealand, sub-Antarctic islands and West Antarctica Fig. 13. Schematic sketch placing the origin of the MBS in a regional geodynamic context. (a) During the ﬁnal stage of subduction of the Paciﬁc Plate beneath the Zealandia/ West Antarctic Gondwana margin, the Hikurangi Plateau approaches the subduction zone. (b) Forces acting upon the plate margin as, for example, the collision of the Hikurangi Plateau with Zealandia (e.g., Bradshaw, 1989; Davy et al., 2008) cause cessation of subduction and slab detachment. The impact of a plume head at that time was accompanied by large scale underplating of HIMU material beneath East Gondwana (e.g., Weaver et al., 1994; Hart et al., 1997) and the Hikurangi Plateau (possibly by deﬂection of the plume material by the subducting plate; Hoernle et al., 2010), triggering volcanism on West Antarctica and Zealandia and the formation of the Hikurangi Seamounts. (c) After subduction ended, extensional processes set in, causing the break-up of Zealandia from Marie Byrd Land at ~90 Ma and subsequent rifting, forming the oceanic crust of the Amundsen Sea (Eagles et al., 2004a). (d) Lateral temperature differences between warm mantle beneath the continental lithosphere and normal upper mantle drove continental-insulation ﬂow (model modiﬁed after King and Anderson, 1995), allowing sub-continental mantle material to rise into the upper mantle beneath the adjacent oceanic lithosphere. At the Cretaceous/Tertiary boundary lithospheric extension at the southern margin of the Bellingshausen Plate (e.g., Wobbe et al., 2012) formed deep reaching faults that allowed rise of plume type melts and formation of the MBS from a magma source similar to that of the Hikurangi Seamounts and the West-Antarctic/Zealandia volcanoes. For further details and references see text. (e.g., Weaver and Pankhurst, 1991; Baker et al., 1994; Weaver et al., 1994; Rocholl et al., 1995; Hart et al., 1997; Panter et al., 2000; Tappenden, 2003; Panter et al., 2006; Nardini et al., 2009; Hoernle et al., 2010). In all these localities, volcanic centers are diffusely distributed and do not show any age progression relative to plate motion. Most commonly, models suggest localized extension/upwelling of asthenosphere that induces melting of metasomatized lithosphere in thin spots to produce the diffuse alkaline magmatism. Finn et al. (2005) postulate a “diffuse alkaline magmatic province (DAMP)”, which formed without any rifting or plume upwelling. They temporally extend the DAMP into the Cenozoic and explain this magmatism by detachment of subducted slabs from the base of Gondwana lithosphere in the late Cretaceous. The sinking of material into the mantle is thought to have introduced Rayleigh Taylor instabilities A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 along the Gondwana margin and activated lateral and vertical ﬂow of warm Paciﬁc mantle. After Finn et al. (2005) the interaction of the warm mantle with metasomatized lithosphere generated the HIMU geochemical characteristics of the DAMP. A shortcoming of this model is, however, that Finn et al. (2005) had to focus their study on old, continental fragments of East Gondwana, and could not include oceanic occurrences like the MBS, the De Gerlache Seamounts or the Hikurangi Seamounts which are situated on top of Hikurangi Plateau off New Zealand. An important difference between the seamounts and the continental alkaline provinces is that the majority of seamount provinces formed on relatively young oceanic crust. While HIMU signatures can be found in old continental terranes, HIMU-type volcanism in the oceans either requires rise of HIMU material from depth or some sort of refertilization of the upper mantle, especially when required shortly after ocean crust formation. No doubt, small-scale heterogeneities exist in the upper mantle away from mantle plumes as is evident from small off-axis seamounts that often have more enriched element and isotopic signatures than associated MORB (e.g., Brandl et al., 2012 and references therein). It seems, however, unlikely that such smallscale heterogeneities are present shortly after formation of the ocean crust to an extent that can explain the c. 1000–8000 km3 of enriched melt required to form individual MBS (see Table 1 for volume estimates). In other words, even if Raleigh Taylor instabilities affected the Gondwana margin it seems unlikely that upwelling of regular Paciﬁc upper mantle that underwent high degrees of melting shortly before can serve as the source of the HIMU-type compositions without refertilization. Alternatively, the superplume beneath the SW-Paciﬁc could have supplied a dense swarm of widely distributed and contemporaneously active secondary plumes causing diffuse alkaline volcanism (Suetsugu et al., 2009). Since it is in principle possible that secondary plumelets or blobs are continuously rising from the SW Paciﬁc superswell (presumed to have stalled at the 660 km transition zone; Courtillot et al., 2003) they may also serve as the cause of volcanism forming the MBS and De Gerlache Seamounts. The age–distance relationship between MBS and the much farther north located Paciﬁc Superswell is, however, unclear. Alternatively, Timm et al. (2010 and references therein) identify a low velocity anomaly extending from Chatham Rise off New Zealand to western Antarctica in at 600–1500 km depth and suggest that this could be the HIMU source polluting the upper mantle in this area since Cretaceous. Still it appears accidental that only the Marie Byrd Land margin was hit by a short-lived swarm of plumelets and no other oceanic region above this low velocity zone. Therefore we explore an alternative scenario for the oceanic seamount provinces off Marie Byrd Land based on reactivation of (HIMU) material, added to the base of continental lithosphere by plume activity during the pre-rifting stage of Marie Byrd Land/Zealandia. On a regional scale, the MBS and Belgica Seamount data overlap with the data ﬁeld of the Hikurangi Seamounts (Hoernle et al., 2010) in most isotope correlation diagrams (Figs. 10–12). A similar HIMU signature of Cretaceous rocks is also found at the Mandamus complex, the Lookout Volcanics in southern New Zealand, and the Chatham Islands (Weaver and Pankhurst, 1991; Tappenden, 2003; Panter et al., 2006; McCoy-West et al., 2010). During the Cretaceous these localities were assembled adjacent to Marie Byrd Land. It has been proposed that a HIMU-type plume or plume head may have caused breakup of the Gondwana margin in this region (e.g., Weaver et al., 1994; Hart et al., 1997; Storey et al., 1999; Hoernle et al., 2010). This plume event may have also inﬂuenced the source characteristics of the Hikurangi Seamounts (Hoernle et al., 2010) and may have been accompanied by large scale underplating of the Zealandia continental lithosphere by HIMU material (e.g. Weaver et al., 1994; Hart et al., 1997; Panter et al., 2000) (Fig. 13). During the mid Cretaceous the plume head expanded and thus forced rifting and the breakup of Gondwana as it impacted at the base of the continental lithosphere (Weaver et al., 1994). We note, however, that in contrast to other continental breakup related mantle plumes such as the Tristan-Gough in the 1675 South-Atlantic (and related Paraná and Entendeka continental ﬂood basalts), a ﬂood basalt event is absent on Zealandia and West Antarctica, possibly reﬂecting the convergent margin setting and associated thick continental lithosphere along the Gondwana margin. Together with the observation that the Cretaceous HIMU volcanism occurred only locally and was of relatively low volume, it seems likely that unmelted HIMU-mantle got attached at the base the Gondwana lithosphere, which underwent extension and rifting during that period. The proposed large-scale underplating of HIMU material beneath East Gondwana is consistent with the HIMU signature of many Cenozoic continental volcanics from West Antarctica (e.g., Hobbs Coast, Marie Byrd Land Volcanic Province, and WARS; cf. Figs. 10 and 11). Accordingly, many authors relate the Cenozoic HIMU similarities in West Antarctica to the reactivation of HIMU material, added to the base of the continental lithosphere during the earliest pre-rifting stage of the Marie Byrd Land through plume activity (e.g., Weaver et al., 1994; Rocholl et al., 1995; Hart et al., 1997; Panter et al., 2000). Alternatively Nardini et al. (2009 and references therein) call upon a late Cretaceous metasomatic event that caused variable elevation of U/Pb ratios in the sub-lithospheric mantle to an extend that explains the high 206Pb/204Pb of b 20 Ma WARS volcanics and generation of their HIMU isotopic source signatures through radiogenic ingrowth over extremely short time scales. The regional context, however, requires the presence of a HIMU component that is already present in the Cretaceous, so that the metasomatic model of Nardini et al. (2009) for the formation of HIMU appears less likely. Notably, the ﬁeld for continental volcanic rocks of West Antarctica overlaps the data of the oceanic seamount provinces (Hikurangi Seamounts, MBS, De Gerlache) (Figs. 10 and 11), which have been formed close to the East Gondwana and West Antarctic margin, respectively. The samples also fall within the range of Cretaceous volcanic rocks of southern New Zealand, suggesting that all the above-mentioned volcanic suites originate from a similar HIMU source. This material may therefore also represent reactivation of fossil Cretaceous plume material that was originally attached to the base of the continental lithosphere during Marie Byrd Land/Zealandia break up. In contrast to the above mentioned onshore occurrences of Cenozoic HIMU volcanism, an additional transport mechanism and mode of reactivation is required to explain the marine equivalents of HIMU volcanism, because this material needs ﬁrst of all be transferred into the oceanic mantle beneath the newly formed ocean basins of the Amundsen/Bellingshausen Sea followed by decompression melting (see Section 7.4 for details). Admittedly, the arguments for an initial upwelling of plume-like material and storage at the base of the Gondwana lithosphere are solely based on geochemistry, which points to a HIMU like mantle. Such a source is unlikely to develop in situ in a mantle region affected by long-term subduction zone volcanism and small scale convection cells operating within the mantle wedge both leading to continuous depletion and replenishment of the arc mantle. On the other hand, upwelling of refertilized sub-continental lithospheric mantle (SCLM), isolated from mantle circulation for several billion years, can lead to the formation of EM type melts (e.g., Rudnick, 1995; Grifﬁn et al., 2009; Hoernle et al., 2011; Soager et al., 2013). Ancient SCLM, however, features low 206Pb/204Pb and 143Nd/144Nd along with high 207Pb/204Pb and 87Sr/86Sr ratios, reﬂecting an ancient source that evolved with low U/Pb, Sm/Nd but high Rb/Sr (see Tang et al., 2013 for a recent review). Mantle regions that underwent such a fractionation and/or metasomatic event early in the earth's history are commonly thought to be involved in the formation of the early continental crust, having resided thereafter in the roots of stable Achaean cratons. In conclusion, SCLM seems to be a very unlikely candidate as source of the Cretaceous HIMU-type intraplate volcanism due the conﬂicting isotopic composition of SCLM (EM-like) and the long-term subduction zone setting of this area. Therefore, our preferred model for the origin of the HIMU component in the MBS and De Gerlache lavas is reactivation of fossil Cretaceous plume material, which was attached and stored at the base of the West Antarctic continental lithosphere during East Gondwana breakup. 1676 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 7.4. Model for the formation of the Marie Byrd Seamounts In the case of the West Antarctic volcanoes, underplated HIMU material may have been reactivated and caused to upwell during the WARS extension (e.g., Hart et al., 1997). For the formation of the c. 99 to 67 Ma Hikurangi Seamounts, Hoernle et al. (2010) propose the rise of HIMU-type material directly beneath the Hikurangi Plateau — a ~118 Ma oceanic LIP (Hoernle et al., 2010) that formed in connection with the Manihiki (Timm et al., 2011) and possibly the Ontong Java Plateaus (Taylor, 2006), through deﬂection of rising plume material beneath Zealandia by the subducting plate towards the Hikurangi Plateau, which was about to collide with the Zealandia margin at that time (Fig. 13a + b). For the MBS and De Gerlache Seamounts, however, a mechanism is required that enables lateral transport of the earlier emplaced HIMU material under Marie Byrd Land beneath the newly formed bordering oceanic lithosphere. When the new oceanic crust of the Amundsen Sea formed, Zealandia (including the Hikurangi Plateau) rifted away from Marie Byrd Land in a northward direction (Fig. 13c), whereas the West Antarctic continental margin remained more or less ﬁxed and developed as a relatively stable passive margin thereafter (e.g., Eagles et al., 2004a; Wobbe et al., 2012). Mutter et al. (1988) proposed a transition zone directly at the edge of thicker to thinner lithosphere where small convective ﬂow is focused. In case of the MBS and De Gerlache Seamounts, the transition zone lies at the edge of the Antarctic continental lithosphere and the beginning of the adjacent oceanic crust. Several mechanisms such as edge-driven convection (EDC), small-scale convection (SSC) or shear-driven upwelling (SDU) have been suggested to explain edge-driven buoyant ﬂow between young, thin and old, thicker lithosphere (e.g., King and Anderson, 1995, 1998; King and Ritsema, 2000; Huang et al., 2003; Dumoulin et al., 2008; Conrad and Behn, 2010 and references therein). For example, (super-) continents may effectively insulate the upper mantle, leading to a buildup of heat (Anderson, 1994; Lowman and Jarvis, 1995, 1996; Gurnis et al., 1998). These lateral temperature differences between the warm mantle beneath the continental lithosphere and normal upper mantle can drive an upper mantle convective ﬂow pattern that leads to upwelling beneath the continent–ocean transition zone (Fig. 13d), the so-called “continental-insulation ﬂow” (e.g., King and Anderson, 1995, 1998). From numerical modeling, King and Anderson (1998) suggest that lateral variations in temperature of at least 30 °C are required for continental insulation ﬂow to signiﬁcantly modify or even shut off the normal, downwelling EDC ﬂow. Higher temperature anomalies (150–200 °C) would drive major upper mantle convection cells. In case of the Cretaceous East Gondwana lithosphere, the impact of the hot plume head may have caused additional heating of the mantle beneath the continental lithosphere and therefore reinforced the lateral variations in mantle temperature and consequently mantle convection. Notably, the pattern of ﬂow resulting from continental insulation is opposite to that of normal EDC ﬂow (King and Anderson, 1995, 1998). At the initial stages of rifting of a continent, upwelling should occur as warm mantle from beneath the continent that occupies the space created by spreading between the continental masses. At the Late Cretaceous Marie Byrd Land margin, this process would transfer mantle material directly from beneath the continent into the upper mantle under the adjacent oceanic lithosphere on which the MBS started to form at that time (Fig. 13d). Therefore we consider continental insulation ﬂow as the most plausible mechanism to bring the HIMU plume-like material previously attached beneath Marie Byrd Land upwards beneath the adjacent oceanic lithosphere of the Bellingshausen/Antarctic Plate. As the HIMU material was transported upwards beneath the newly formed oceanic lithosphere from beneath the thick Antarctica continental crust, the material will melt by decompression. The volcanism forming the De Gerlache Seamounts at ~ 22 Ma and the Pleistocene activity of Peter I Island, on the other hand, was most likely related to the De Gerlache Gravity Anomaly (Figs. 1 and 2), which represents a zone of lithospheric weakness resulting from a presumed WARS activity in this region (Müller et al., 2007), where pre-existing N–S striking faults allowed rise (and decompression melting) of HIMU-type material brought up beneath the oceanic lithosphere by mantle convection. The formation of the MBS may therefore have been triggered by a complex sequence of plate reorganization events that affected the West Antarctic margin and the Bellingshausen Plate in Late Cretaceous and Early Cenozoic (e.g., Eagles et al., 2004a; Wobbe et al., 2012). Shortly before the Bellingshausen Plate became incorporated into the Antarctic Plate at 61 Ma (Eagles et al., 2004a,b; Wobbe et al., 2012), a change in rotation of the Bellingshausen Plate from counterclockwise to clockwise was accompanied by lithospheric extension on its southern margin between 74 and 62 Ma (Wobbe et al., 2012). Contemporaneously the MBS started to form in that area (Fig. 2) (Fig. 13d), suggesting that lithospheric extension lead to upwelling of sub-lithospherically attached HIMU material and deep reaching faults that allowed rise of the HIMU-type melts and formation of large volcanic islands (Fig. 13d). 8. Conclusions Our new morphological, geochronological, and geochemical data for the MBS combined with additional data for the De Gerlache and Peter I Island volcanic complexes (complementing previously published data) permit for the ﬁrst time a comprehensive reconstruction of the origin and evolution of Cenozoic intraplate volcanism in the Amundsen Sea. The most important results are: (1) Intraplate volcanism occurred during the entire Cenozoic at distinct time intervals in spatially conﬁned areas in the Amundsen Sea, excluding an origin of this volcanism by a single stationary hotspot. (2) The MBS and De Gerlache Seamount lavas show OIB signatures and posses a distinct HIMU component in their magma source similar to Late Cretaceous–Cenozoic volcanics of the Hikurangi Seamounts off New Zealand, intraplate volcanic ﬁelds in New Zealand, sub-Antarctic islands and the WARS, suggesting a common mantle source for these volcanic provinces. (3) Peter I Island displays a strong EM afﬁnity probably caused by shallow mantle recycling of a continental fragment. Consequently, the formation of the MBS and De Gerlache Seamounts intraplate volcanism requires an alternative, non-hotspot scenario, which takes distinct melting anomalies independent in space and time and a non-hotspot related HIMU source into account. Placing the morphological, geochronological, and geochemical data in a regional plate tectonic context, we conclude that the most plausible explanation for the HIMU-type intraplate volcanism in the Amundsen Sea is reactivation of HIMU-material, added to the base of the Antarctic lithosphere by a Late Cretaceous plume event. Major tectonic events, namely the separation of Zealandia from Antarctica during the ﬁnal stage of the Gondwana break-up and subsequent formation of ocean crust give way for transport of the sub-lithospheric HIMU material beneath the Amundsen Sea oceanic crust by continental insulation ﬂow. Extension caused by plate tectonic reorganization (MBS) and/or lithospheric weakening underneath the De Gerlache Gravity Anomaly (De Gerlache, Peter I Island) allow rise and adiabatic melting of the HIMU material resulting in the formation of these volcanic ediﬁces. Reactivation of the MBS magmatism resulting in Pliocene low volume volcanism and the Pleistocene formation of Peter I Island documents ongoing magmatism in the Amundsen Sea. The new model for the Amundsen Sea volcanism presented here adds to case examples for non-hotspot intraplate volcanism and provides additional evidence that HIMU-type intraplate volcanism is not necessarily a direct consequence of an actively upwelling, stationary mantle plume or hotspot. A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.gr.2013.06.013. Acknowledgments We are grateful to Captain Pahl, the crew, and shipboard scientiﬁc party for their excellent support during RV Polarstern cruise ANT-XXIII/ 4. R. Gersonde (AWI) kindly provided the dredge samples from Belgica Seamount. D. Rau, S. Hauff, J. Sticklus (GEOMAR) and H. Anders (Uni-Bremen) are thanked for technical assistance during lab work and S. Gauger for help with processing of the bathymetric data. Discussions with Maxim Portnyagin, Jan Grobys, Graeme Eagles, and Christian Timm signiﬁcantly helped to develop this paper. Furthermore, we are grateful for the constructive reviews of John Gamble and Tsuyoshi Komiya that helped to improve an earlier version of the manuscript. We thank Inna Yu Safonova for editorial handling and useful comments that helped to emphasize the importance of seamount formation. The German Research Foundation (DFG; Grants HO1833/15-1 to -3 to KH and FH) funded this research. References Anderson, D.L., 1994. Superplumes or supercontinents? Geology 22, 39–42. http:// dx.doi.org/10.1130/0091-7613(1994)022b0039:SOSN2.3.CO;2. Anderson, D.L., 2000. The thermal state of the upper mantle; no role for mantle plumes. Geophysical Research Letters 27 (22), 3623–3626. http://dx.doi.org/10.1029/ 2000GL011533. Baker, I.A., Gamble, J.A., Graham, I.J., 1994. The age, geology, and geochemistry of the Tapuaenuku Igneous Complex, Marlborough, New Zealand. New Zealand Journal of Geology and Geophysics 37, 249–268. http://dx.doi.org/10.1080/00288306.1994.9514620. Baksi, A.K., 2007. A quantitative tool for detecting alteration in undisturbed rocks and minerals — I: water, chemical weathering, and atmospheric argon. Special Paper of the Geological Society of America 430 (1197), 285–303. http://dx.doi.org/ 10.1130/2007.2430(16). Batiza, R., Niu, Y., Zayac, W.C., 1990. Chemistry of seamounts near the East Paciﬁc Rise: implications for the geometry of subaxial mantle ﬂow. Geology 18, 1122–1125 (doi:1122-112510.1130/0091-7613(1990)018 b 1122:COSNTE N 2.3.CO;2). Boger, S.D., 2011. Antarctica — before and after Gondwana. Gondwana Research 19 (2), 335–371. http://dx.doi.org/10.1016/j.gr.2010.09.003. Bradshaw, J.D., 1989. Cretaceous geotectonic patterns in the New Zealand region. Tectonics 8, 803–820. http://dx.doi.org/10.1029/TC008i004p00803. Brandl, P.A., Beier, C., Regelous, M., Abouchami, W., Haase, K.M., Garbe-Schönberg, D., Galer, S.J.G., 2012. Volcanism on the ﬂanks of the East Paciﬁc Rise: quantitative constraints on mantle heterogeneity and melting processes. Chemical Geology 298–299, 41–56. http://dx.doi.org/10.1016/j.chemgeo.2011.12.015. Broch, O.A., 1927. Gesteine von der Peter I.-lnsel, West Antarktis. Avhandlinger/Det Norske Videnskaps-Akademi, I, Matematisk-Naturvidenskapelig Oslo KL 9, 1–41. Buchs, D.M., Arculus, R.J., Baumgartner, P.O., Ulianov, A., 2011. Oceanic intraplate volcanoes exposed: example from seamounts accreted in Panama. Geology 39, 335–338. http://dx.doi.org/10.1130/G31703.1. Conrad, C.P., Behn, M.D., 2010. Constraints on lithosphere net rotation and asthenospheric viscosity from global mantle ﬂow models and seismic anisotropy. Geochemistry, Geophysics, Geosystems 11, Q05W05. http://dx.doi.org/10.1029/2009GC002970. Courtillot, V., Davaille, A., Besse, J., Stock, J., 2003. Three distinct types of hotspots in the Earth's mantle. Earth and Planetary Science Letters 205, 295–308. http://dx.doi.org/ 10.1016/S0012-821X(02)01048-8. Davy, B.W., Hoernle, K., Werner, R., 2008. Hikurangi Plateau: crustal structure, rifted formation, and Gondwana subduction history. Geochemistry, Geophysics, Geosystems 9, Q07004. http://dx.doi.org/10.1029/2007GC001855. Doubrovine, P.V., Steinberger, B., Torsvik, T.H., 2012. Absolute plate motions in a reference frame deﬁned by moving hot spots in the Paciﬁc, Atlantic, and Indian oceans. Journal of Geophysical Research 117, B09101. http://dx.doi.org/10.1029/2011JB009072. Dumoulin, C., Choblet, G., Doin, M.P., 2008. Convective interactions between oceanic lithosphere and asthenosphere: inﬂuence of a transform fault. Earth and Planetary Science Letters 274, 301–309. http://dx.doi.org/10.1016/j.epsl.2008.07.017. Eagles, G., Gohl, K., Larter, R., 2004a. High-resolution animated tectonic reconstruction of the South Paciﬁc and West Antarctic Margin. Geochemistry, Geophysics, Geosystems 5, Q07002. http://dx.doi.org/10.1029/2003GC000657. Eagles, G., Gohl, K., Larter, R., 2004b. Life of the Bellingshausen Plate. Geophysical Research Letters 31, L07603. http://dx.doi.org/10.1029/2003GL019127. Eagles, G., Larter, R., Gohl, K., Vaughan, A.P.M., 2009. West Antarctic Rift System in the Antarctic Peninsula. Geophysical Research Letters 36, L21305. http://dx.doi.org/ 10.1029/2009GL040721. Feldberg, M.J., 1997. A Geophysical Study of Seamount E, Bellingshausen Sea, Antarctica. Diploma Degree of Bachelor of Arts Wesleyan University, USA. Finn, C.A., Müller, R.D., Panter, K.S., 2005. A Cenozoic diffuse alkaline magmatic province in the SW Paciﬁc without rift or plume origin. Geochemistry, Geophysics, Geosystems 6, Q02005. http://dx.doi.org/10.1029/2004GC000723. 1677 Fisher, A.T., Davis, E.E., Hutnak, M., Spiess, V., Zühlsdorff, L., Cherkaoui, A., Christiansen, L., Edwards, K., Macdonald, R., Villinger, H., Mottl, M.J., Wheat, C.G., Becker, K., 2003. Hydrothermal recharge and discharge across 50 km guided by seamounts on a young ridge ﬂank. Nature 421, 618–621. http://dx.doi.org/10.1038/nature01352. Foulger, G.R., Natland, J.H., 2003. Is “hotspot” volcanism a consequence of plate tectonics? Science 300, 921–922. http://dx.doi.org/10.1126/science.1083376. Frey, F.A., Wise, W.S., Garcia, M.O., West, H., Kwon, S.-T., Kennedy, A., 1990. Evolution of Mauna Kea Volcano, Hawaii: petrologic and geochemical constraints on postshield volcanism. Journal of Geophysical Research 95 (B2), 1271–1300. http://dx.doi.org/ 10.1029/JB095iB02p01271. Garcia, M.O., Jorgenson, B.A., Mahoney, J.J., Ito, E., Irving, A.J., 1993. An evaluation of temporal geochemical evolution of Loihi Summit Lavas: results from Alvin submersible dives. Journal of Geophysical Research 98 (B1), 537–550. http://dx.doi.org/ 10.1029/92JB01707. Geldmacher, J., Hanan, B.B., Blichert-Toft, J., Harpp, K., Hoernle, K., Hauff, F., Werner, R., und Kerr, A., 2003. Hf isotopic variations in volcanic rocks from the Caribbean Large Igneous Province and Galápagos hotspot tracks. Geochemistry, Geophysics, Geosystems 422 (7). http://dx.doi.org/10.1029/2002GC000477. Geldmacher, J., Hoernle, K., van den Bogaard, P., Duggen, S., Werner, R., 2005. New 40 Ar/39Ar age and geochemical data from seamounts in the Canary and Madeira volcanic provinces: support for the mantle plume hypothesis. Earth and Planetary Science Letters 237, 85–101. http://dx.doi.org/10.1016/j.epsl.2005.04.037. Geldmacher, J., Hoernle, K., van den Bogaard, P., Hauff, F., Klügel, A., 2008. Age and geochemistry of the Central American forearc basement (DSDP Leg 67 and 84): insights into Mesozoic arc volcanism and seamount accretion on the fringe of the Caribbean LIP. Journal of Petrology 49, 1781–1815. http://dx.doi.org/10.1093/petrology/egn046. Gohl, K., 2007. The expedition ANTARKTIS-XXIII/4 of the research vessel Polarstern in 2006. Berichte zur Polar- und Meeresforschung (Reports on Polar and Marine Research), no. 557. Alfred Wegener Institute for Polar and Marine Research, Bremerhaven (166 pp. http://epic.awi.de/26756/). Gohl, K., Nitsche, F.O., Miller, H., 1997a. Seismic and gravity data reveal Tertiary interplate subduction in the Bellingshausen Sea, Southeast Paciﬁc. Geology 25, 371–374. http://dx.doi.org/10.1130/0091-7613(1997)025b0371:SAGDRTN2.3.CO;2. Gohl, K., Nitsche, F., Vanneste, K., Miller, H., Fechner, N., Oszko, L., Hübscher, C., Weigelt, E., Lambrecht, A., 1997b. Tectonic and sedimentary architecture of the Bellingshausen and Amundsen Sea Basins, SE Paciﬁc, by seismic proﬁling. In: Ricci, C.A. (Ed.), The Antarctic Region: Geological Evolution and Processes. Terra Antartica Publication, Siena, pp. 719–723. Gohl, K., Denk, A., Wobbe, F., Eagles, G., 2013. Deciphering tectonic phases of the Amundsen Sea Embayment shelf, West Antarctica, from a magnetic anomaly grid. Tectonophysics 585, 113–123. http://dx.doi.org/10.1016/j.tecto.2012.06.036. Grifﬁn, W.L., O'Reilly, S.Y., Afonso, J.C., Begg, G.C., 2009. The composition and evolution of lithospheric mantle: a re-evaluation and its tectonic implications. Journal of Petrology 50, 1185–1204. http://dx.doi.org/10.1093/petrology/egn033. Gurnis, M., Mueller, R.D., Moresi, L., 1998. Dynamics of Cretaceous vertical motion of Australia and the Australian–Antarctic discordance. Science 279, 1499–1504. http://dx.doi.org/10.1126/science.279.5356.1499. Hagedorn, B., Gersonde, R., Gohl, K., Hubberten, H.-W., 2007. Petrology, geochemistry and K–Ar age constraints of the eastern De Gerlache Seamount alkaline basalts (Bellingshausen Sea, Southeast Paciﬁc). Polarforschung 76 (3), 87–94 (http:// epic.28876.d001). Harris, R.N., Fisher, A.T., Chapman, D.S., 2004. Fluid ﬂow through seamounts and implications for global mass ﬂuxes. Geology 32, 725–728. http://dx.doi.org/10.1130/G20387.1. Hart, S.R., Hauri, E.H., Oschmann, L.A., Whitehead, J.A., 1992. Mantle plumes and entrainment: isotopic evidence. Science 256, 517–520. http://dx.doi.org/10.1126/ science.256.5056.517. Hart, S.R., Blusztajn, J., Craddock, C., 1995. Cenozoic volcanism in Antarctica: Jones Mountains and Peter I Island. Geochimica et Cosmochimica Acta 59, 3379–3388. http://dx.doi.org/10.1016/0016-7037(95)00212-I. Hart, S.R., Blusztajn, J., LeMasurier, W.E., Rex, D.C., 1997. Hobbs Coast Cenozoic volcanism, implications for the West Antarctic Rift System. Chemical Geology 139, 223–248. http://dx.doi.org/10.1016/S0009-2541(97)00037-5. Hauri, E.H., Whitehead, J.A., Hart, S.A., 1994. Fluid dynamic and geochemical aspects of entrainment in mantle plumes. Journal of Geophysical Research 99 (B12), 24275–24300. http://dx.doi.org/10.1029/94JB01257. Hein, J.R., Conrad, T.A., Staudigel, H., 2010. Seamount mineral deposits: a source of rare metals for high-technology industries. Oceanography 23, 184–189. http:// dx.doi.org/10.5670/oceanog.2010.70. Heinemann, J., Stock, J., Clayton, R., Hafner, K., Cande, S., Raymond, C., 1999. Constraints on the proposed Marie Byrd Land–Bellingshausen Plate boundary from seismic reﬂection data. Journal of Geophysical Research 104 (B11), 25321–25330. http:// dx.doi.org/10.1029/1998JB900079. Hirschmann, M.M., Stolper, E.M., 1996. A possible role for garnet pyroxenite in the origin of the “garnet signature” in MORB. Contributions to Mineralogy and Petrology 124, 185–208. http://dx.doi.org/10.1007/s004100050184. Hoernle, K., van den Bogaard, P., Werner, R., Lissina, B., Hauff, F., Alvarado, G., GarbeSchönberg, D., 2002. Missing history (16–71 Ma) of the Galápagos hotspot: implications for the tectonic and biological evolution of the Americas. Geology 30, 795–798. http:// dx.doi.org/10.1130/0091-7613(2002)030b0795:mhmotgN2.0.co;2. Hoernle, K., Hauff, F., Werner, R., Mortimer, N., 2004. New insights into the origin and evolution of the Hikurangi Oceanic Plateau (Southwest Paciﬁc) from multi-beam mapping and sampling. Eos, Transactions of the American Geophysical Union 85 (41), 401–408. http://dx.doi.org/10.1029/2004EO410001. Hoernle, K., White, J.D.L., van den Bogaard, P., Hauff, F., Coombs, D.S., Werner, R., Timm, C., Garbe-Schoenberg, D., Reay, A., Cooper, A.F., 2006. Cenozoic intraplate volcanism on 1678 A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 New Zealand: upwelling induced by lithospheric removal. Earth and Planetary Science Letters 248, 350–367. http://dx.doi.org/10.1016/j.epsl.2006.06.001. Hoernle, K., Hauff, F., van den Bogaard, P., Werner, R., Mortimer, N., Geldmacher, J., Garbe-Schoenberg, D., Davy, B., 2010. Age and geochemistry of volcanic rocks from the Hikurangi and Manihiki Oceanic Plateaus. Geochimica et Cosmochimica Acta 74 (24), 7196–7219. http://dx.doi.org/10.1016/j.gca.2010.09.030. Hoernle, K., Hauff, F., Werner, R., van den Bogaard, P., Gibbons, A.D., Conrad, S., Müller, R.D., 2011. Origin of Indian Ocean Seamount Province by shallow recycling of continental lithosphere. Nature Geoscience 4, 883–887. http://dx.doi.org/10.1038/ngeo1331. Hofmann, A.W., 1988. Chemical differentiation of the Earth: the relationship between mantle, continental and oceanic crust. Earth and Planetary Science Letters 90, 297–314. http://dx.doi.org/10.1016/0012-821X(88)90132-X. Hofmann, A.W., 2003. Sampling mantle heterogeneity through oceanic basalts: isotopes and trace elements. In: Carlson, R.W. (Ed.), The Mantle and Core. Elsevier, Amsterdam, pp. 61–101. http://dx.doi.org/10.1016/B0-08-043751-6/02123-X. Hofmann, A.W., White, W.M., 1982. Mantle plumes from ancient oceanic crust. Earth and Planetary Science Letters 57, 421–436. http://dx.doi.org/10.1016/0012-821X(82) 90161-3. Hofmann, A.W., Jochum, K.-P., Seufert, M., White, W.M., 1986. Nb and Pb in oceanic basalts: new constraints on mantle evolution. Earth and Planetary Science Letters 79, 33–45. http://dx.doi.org/10.1016/0012-821X(86)90038-5. Hole, M.J., LeMasurier, W.E., 1994. Tectonic controls on the geochemical composition of Cenozoic, maﬁc alkaline volcanic rocks from West Antarctica. Contributions to Mineralogy and Petrology 117, 187–202. http://dx.doi.org/10.1007/BF00286842. Huang, J., Zhong, S., van Hunen, J., 2003. Controls on sub-lithospheric small-scale convection. Journal of Geophysical Research 108 (B8), 2405. http://dx.doi.org/10.1029/ 2003JB002456. Hutnak, M., Fisher, A.T., Harris, R., Stein, C., Wang, K., Spinelli, G., Schindler, M., Villinger, H., Silver, E., 2008. Large heat and ﬂuid ﬂuxes driven through mid-plate outcrops on ocean crust. Nature Geoscience 1, 611–614. http://dx.doi.org/10.1038/ngeo264. Irvine, T.N., Baragar, W.R.A., 1971. A guide to the chemical classiﬁcation of the common volcanic rocks. Canadian Journal of Earth Sciences 8, 523–548. http://dx.doi.org/ 10.1139/e71-055. King, S.D., Anderson, D.L., 1995. An alternative mechanism to ﬂood basalt formation. Earth and Planetary Science Letters 136, 269–279. http://dx.doi.org/10.1016/ 0012-821X(95)00205-Q. King, S.D., Anderson, D.L., 1998. Edge-driven convection. Earth and Planetary Science Letters 160, 289–296. http://dx.doi.org/10.1016/S0012-821X(98)00089-2. King, S.D., Ritsema, J., 2000. African hotspot volcanism: small-scale convection in the upper mantle beneath cratons. Science 290, 1137–1140. http://dx.doi.org/10.1126/ science.290.5494.1137. Klügel, A., Hansteen, T.H., van den Bogaard, P., Strauss, H., Hauff, F., 2011. Holocene ﬂuid venting at an extinct Cretaceous seamount, Canary Archipelago. Geology 39, 855–858. http://dx.doi.org/10.1130/G32006.1. Koppers, A.A.P., Yamazaki, T., Geldmacher, J., Gee, J.S., Pressling, N., IODP Expedition 330 Scientiﬁc Party, 2012. Limited latitudinal mantle plume motion for the Louisville hotspot. Nature Geoscience 5, 911–917. http://dx.doi.org/10.1038/ngeo1638. Larter, R.D., Cunningham, A.P., Barker, P.F., Gohl, K., Nitsche, F.O., 2002. Tectonic evolution of the Paciﬁc margin of Antarctica — 1. Late Cretaceous tectonic reconstructions. Journal of Geophysical Research 107 (B12), 2345. http://dx.doi.org/10.1029/2000JB000052. LeMasurier, W.E., Thomson, J.W., Baker, P., Kyle, P., Rowley, P., Smellie, J., Verwoerd, W. (Eds.), 1990. Volcanoes of the Antarctic Plate and Southern Oceans. Antarctic Research Series, 48. AGU, Washington, D.C., p. 487. http://dx.doi.org/10.1029/AR048. Lowman, J.P., Jarvis, G.T., 1995. Mantle convection models of continental collision and breakup incorporating ﬁnite thickness plates. Physics of the Earth and Planetary Interiors 88, 53–68. http://dx.doi.org/10.1016/0031-9201(94)05076-A. Lowman, J.P., Jarvis, G.T., 1996. Continental collisions in wide aspect ratio and high Rayleigh number two-dimensional mantle convection models. Journal of Geophysical Research 101 (B11), 25485–25497. http://dx.doi.org/10.1029/96JB02568. McAdoo, D.C., Laxon, S., 1997. Antarctic tectonics: constraints from an ERS-1 satellite marine gravity ﬁeld. Science 276, 556–560. http://dx.doi.org/10.1126/science.276.5312.556. McCoy-West, A.J., Baker, J.A., Faure, K., Wysoczanski, R., 2010. Petrogenesis and origins of mid-Cretaceous continental intraplate volcanism in Marlborough, New Zealand: implications for the long-lived HIMU magmatic mega-province of the SW Paciﬁc. Journal of Petrology 51, 2003–2045. http://dx.doi.org/10.1093/petrology/egq046. McDonough, W.F., Sun, S.-S., 1995. The composition of the earth. Chemical Geology 120, 223–253. http://dx.doi.org/10.1016/0009-2541(94)00140-4. McMurtry, G.M., Fryer, G.J., Tappin, D.R., Wilkinson, I.P., Williams, M., Fietzke, J., GarbeSchönberg, D., Watts, P., 2004. Megatsunami deposits on Kohala Volcano, Hawaii, from ﬂank collapse of Mauna Loa. Geology 32, 741–744. http://dx.doi.org/10.1130/G20642.1. Miller, H., Grobe, H., 1996. The expedition ANTARKTIS-XI/3 of RV ‘Polarstern’ in 1994. Berichte zur Polarforschung, no. 188 (http://epic.10189.d001). Montelli, R., Nolet, G., Dahlen, F.A., Masters, G., 2006. A catalogue of deep mantle plumes: new results from ﬁnite-frequency tomography. Geochemistry, Geophysics, Geosystems 7, Q11007. http://dx.doi.org/10.1029/2006GC001248. Morgan, W.J., 1971. Convection plumes in the lower mantle. Nature 230, 42–43. http:// dx.doi.org/10.1038/230042a0. Müller, R.D., Gaina, C., Tikku, A., Mihut, D., Cande, S.C., Stock, J.M., 2000. Mesozoic/Cenozoic tectonic events around Australia. Geophysical Monograph 121, 161–188. http://dx.doi.org/10.1029/GM121p0161. Müller, R.D., Gohl, K., Cande, S.C., Goncharov, A., Golynsky, A.V., 2007. Eocene to Miocene geometry of the West Antarctic Rift System. Australian Journal of Earth Sciences 54, 1033–1045. http://dx.doi.org/10.1080/08120090701615691. Mutter, J.C., Buck, W.R., Zehnder, C.M., 1988. Convective partial melting 1. A model for the formation of thick basaltic sequences during the initiation of spreading. Journal of Geophysical Research 93 (B2), 1031–1048. http://dx.doi.org/10.1029/JB093iB02p01031. Nardini, I., Armienti, P., Rocchi, S., Dallai, L., Harrison, D., 2009. Sr–Nd–Pb–He–O isotope and geochemical constraints on the genesis of Cenozoic magmas from the West Antarctic Rift. Journal of Petrology 50 (7), 1359–1375. http://dx.doi.org/10.1093/ petrology/egn082. Natland, J.H., Winterer, E.L., 2005. Fissure control on volcanic action in the Paciﬁc. In: Foulger, G.R., Natland, J.H., Presnall, D.C., Anderson, D.L. (Eds.), Plumes, Plates and Paradigms. Geological Society of America, Boulder CO, pp. 687–710. http:// dx.doi.org/10.1130/0-8137-2388-4.687. Panter, K.S., Kyle, P.R., Smellie, J.L., 1997. Petrogenesis of a phonolite–trachyte succession at Mount Sidley, Marie Byrd Land, Antarctica. Journal of Petrology 38 (9), 1225–1253. http://dx.doi.org/10.1093/petroj/38.9.1225. Panter, K.S., Hart, S.R., Kyle, P., Blusztanjn, J., Wilch, T., 2000. Geochemistry of Late Cenozoic basalts from the Crary Mountains: characterization of mantle sources in Marie Byrd Land, Antarctica. Chemical Geology 165, 215–241. http://dx.doi.org/ 10.1016/S0009-2541(99)00171-0. Panter, K.S., Blusztanjn, J., Hart, S.R., Kyle, P.R., Esser, R., McIntash, W.C., 2006. The origin of HIMU in the SW Paciﬁc: evidence from intraplate volcanism in southern New Zealand and Subantarctic islands. Journal of Petrology 47, 1673–1704. http://dx.doi.org/10.1093/petrology/egl024. Portnyagin, M., Savelyev, D., Hoernle, K., Hauff, F., Garbe-Schönberg, D., 2008. MidCretaceous Hawaiian tholeiites preserved in Kamchatka. Geology 36, 903–906. http://dx.doi.org/10.1130/g25171a.1. Prestvik, T., Duncan, R.A., 1991. The geology and age of Peter I Øy, Antarctica. Polar Research 9, 89–98. http://dx.doi.org/10.1111/j.1751-8369.1991.tb00404.x. Prestvik, T., Barnes, C.G., Sundvoll, B., Duncan, R.A., 1990. Petrology of Peter I Øy (Peter I Island), West Antarctica. Journal of Volcanology and Geothermal Research 44, 315–338. http://dx.doi.org/10.1016/0377-0273(90)90025-B. Rocchi, S., Armienti, P., D'Orazio, M., Tonarini, S., Wijbrans, J., Di Vincenzo, G., 2002a. Cenozoic magmatism in the western Ross Embayment: role of mantle plume versus plate dynamics in the development of the West Antarctic Rift System. Journal of Geophysical Research 107 (B9), 2195. http://dx.doi.org/10.1029/2001JB000515. Rocchi, S., LeMasurier, W.E., Di Vincenzo, G., 2002b. Uplift and erosion history in Marie Byrd Land as a key to possible mid-Cenozoic plate motion between East and West Antarctica. Geological Society of America Abstracts with Programs 34 (6), 238. Rocholl, A., Stein, M., Molzahn, M., Hart, S.R., Wörner, G., 1995. Geochemical evolution of rift magmas by progressive tapping of a stratiﬁed mantle source beneath the Ross Sea Rift, Northern Victoria Land, Antarctica. Earth and Planetary Science Letters 131, 207–224. http://dx.doi.org/10.1016/0012-821X(95)00024-7. Rudnick, R.L., 1995. Making continental crust. Nature 378, 571–578. http://dx.doi.org/ 10.1038/378571a0. Safonova, I.Y., 2009. Intraplate magmatism and oceanic plate stratigraphy of the PaleoAsian and Paleo-Paciﬁc Oceans from 600 to 140 Ma. Ore Geology Reviews 35, 137–154. http://dx.doi.org/10.1016/j.oregeorev.2008.09.002. Safonova, I.Y., Santosh, M., 2012. Accretionary complexes in the Asia–Paciﬁc region: tracing archives of ocean plate stratigraphy and tracking mantle plumes. Gondwana Research 25, 126–158. Salters, V.J.M., Stracke, A., 2004. Composition of the depleted mantle. Geochemistry, Geophysics, Geosystems 5, Q05B07. http://dx.doi.org/10.1029/2003GC000597. Shank, T.M., 2010. Seamounts: deep-ocean laboratories of faunal connectivity, evolution, and endemism. Oceanography 23, 108–122. http://dx.doi.org/10.5670/ oceanog.2010.65. Smith, W.H.F., Sandwell, D.T., 1997. Global sea ﬂoor topography from satellite altimetry and ship depth soundings. Science 277, 1956–1962. http://dx.doi.org/10.1126/ science.277.5334.1956. Soager, N., Holm, P.M., Llambias, E.J., 2013. Payenia volcanic province, southern Mendoza, Argentina: OIB mantle upwelling in a backarc environment. Chemical Geology 349–350, 36–53. http://dx.doi.org/10.1016/j.chemgeo.2013.04.007. Stock, J.M., 1997. Geophysical secrets beneath Antarctic waters. Engineering Sciences 60 (3), 18–27. Storey, B.C., Leat, P.T., Weaver, S.D., Pankhurst, R.J., Bradshaw, J.D., Kelley, S., 1999. Mantle plumes and Antarctica–New Zealand rifting; evidence from MidCretaceous maﬁc dykes. Journal of the Geological Society 156, 659–671. http:// dx.doi.org/10.1144/gsjgs.156.4.0659. Stracke, A., 2012. Earth's heterogeneous mantle: a product of convection-driven interaction between crust and mantle. Chemical Geology 330–331, 274–299. http:// dx.doi.org/10.1016/j.chemgeo.2012.08.007. Suetsugu, D., Isse, T., Tanaka, S., Obayashi, M., Shiobara, H., Sugioka, H., Kanazawa, T., Fukao, Y., Barruol, G., Reymond, D., 2009. South Paciﬁc mantle plumes imaged by seismic observation on islands and seaﬂoor. Geochemistry, Geophysics, Geosystems 10, Q11014. http://dx.doi.org/10.1029/2009GC002533. Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A.D., Norry, M.J. (Eds.), Magmatism in the Ocean Basins. Geological Society Special Publications, 42, pp. 313–345. http://dx.doi.org/10.1144/GSL.SP.1989.042.01.19. Sutherland, R., Spasojevic, S., Gurnis, M., 2010. Mantle upwelling after Gondwana subduction death explains anomalous topography and subsidence history of eastern New Zealand and West Antarctic. Geology 38, 155–158. http://dx.doi.org/10.1130/G30613.1. Tang, Y.-J., Zhang, H.-F., Ying, J.-F., Su, B.-X., 2013. Widespread refertilization of cratonic and circum-cratonic lithospheric mantle. Earth-Science Reviews 118, 45–68. http://dx.doi.org/10.1016/j.earscirev.2013.01.004. Tappenden, V.E., 2003. Magmatic Response to the Evolving New Zealand Margin of Gondwana During the Mid–Late Cretaceous. (PhD thesis) University of Canterbury, Christchurch, New Zealand. Tarduno, J.A., Duncan, R.A., Scholl, D.W., Cottrell, R.D., Steinberger, B., Thordason, T., Kerr, B.C., Neal, C.R., Frey, F.A., Torii, M., Carvallo, C., 2003. The Emperor Seamounts: A. Kipf et al. / Gondwana Research 25 (2014) 1660–1679 southward motion of the Hawaiian hotspot plume in earth's mantle. Science 301, 1064–1069. http://dx.doi.org/10.1126/science.1086442. Taylor, B., 2006. The single largest oceanic plateau: Ontong Java–Manihiki–Hikurangi. Earth and Planetary Science Letters 241, 372–380. http://dx.doi.org/10.1016/ j.epsl.2005.11.049. Timm, C., Hoernle, K., Werner, R., Hauff, F., van den Bogaard, P., White, J., Mortimer, N., Garbe-Schoenberg, D., 2010. Temporal and geochemical evolution of the Cenozoic intraplate volcanism of Zealandia. Earth-Science Reviews 98, 38–64. http:// dx.doi.org/10.1016/j.earscirev.2009.10.002. Timm, C., Hoernle, K., Werner, R., Hauff, F., van den Bogaard, P., Michael, P., Cofﬁn, M., Koppers, A., 2011. Age and geochemistry of the oceanic Manihiki Plateau, SW Pacific: new evidence for a plume origin. Earth and Planetary Science Letters 304 (1–2), 135–146. http://dx.doi.org/10.1016/j.epsl.2011.01.025. Udintsev, G.B., Kurentsova, N.A., Teterin, D.E., Roshchina, I.A., 2007. Petrology of the Hubert Miller Seamount, Marie Byrd Seamounts Province, West Antarctic, Southern Ocean. Doklady Earth Sciences 415A (6), 895–900. http://dx.doi.org/10.1134/ S1028334X07060141. Uenzelmann-Neben, G., Gohl, K., 2012. Amundsen Sea sediment drifts: archives of modiﬁcations in oceanographic and climatic conditions. Marine Geology 299–302, 51–62. http://dx.doi.org/10.1016/j.margeo.2011.12.007. Watts, A.B., Koppers, A.A.P., Robinson, D.P., 2010. Seamount subduction and earthquakes. Oceanography 23 (1), 166–173. http://dx.doi.org/10.5670/ oceanog.2010.68#sthash.6fn0xSSg.dpuf. Weaver, S.D., Pankhurst, R.J., 1991. A precise Rb–Sr age for the Mandamus Igneous Complex, North Canterbury, and regional tectonic implications. New Zealand Journal of Geology and Geophysics 34, 341–345. http://dx.doi.org/10.1080/ 00288306.1991.9514472. 1679 Weaver, S.D., Storey, B.C., Pankhurst, R.J., Mukasa, S.B., DiVenere, V.J., Bradshaw, J.D., 1994. Antarctica–New Zealand rifting and Marie Byrd Land lithospheric magmatism linked to ridge subduction and mantle plume activity. Geology 22, 811–814. http://dx.doi.org/10.1130/0091-7613(1994)022b0811: ANZRAMN2.3.CO;2. Wessel, P., Sandwell, D.T., Kim, S.-S., 2010. The global seamount census. Oceanography 23, 24–33. http://dx.doi.org/10.5670/oceanog.2010.60. White, W.M., 2010. Oceanic island basalts and mantle plumes: the geochemical perspective. Annual Review of Earth and Planetary Sciences 38, 133–160. http:// dx.doi.org/10.1146/annurev-earth-040809-152450. Willbold, M., Stracke, A., 2010. Formation of enriched mantle components by recycling of upper and lower continental crust. Chemical Geology 276, 188–197. http:// dx.doi.org/10.1016/j.chemgeo.2010.06.005. Wilson, J.T., 1963. Evidence from islands on the spreading of the ocean ﬂoor. Nature 197, 536–538. http://dx.doi.org/10.1038/197536a0. Winterer, E.L., Sandwell, D.T., 1987. Evidence from en-echelon cross-grain ridges for tensional cracks in the Paciﬁc Plate. Nature 329, 534–537. http://dx.doi.org/ 10.1038/329534a0. Wobbe, F., Gohl, K., Chambord, A., Sutherland, R., 2012. Structure and breakup history of the rifted margin of West Antarctica in relation to Cretaceous separation from Zealandia and Bellingshausen Plate motion. Geochemistry, Geophysics, Geosystems 13, Q04W12. http://dx.doi.org/10.1029/2011GC003742. Workman, R.K., Hart, S.R., 2005. Major and trace element composition of the depleted MORB mantle (DMM). Earth and Planetary Science Letters 231, 53–72. http:// dx.doi.org/10.1016/j.epsl.2004.12.005. Zindler, A., Hart, S.R., 1986. Chemical geodynamics. Annual Review of Earth and Planetary Sciences 14, 493–571. http://dx.doi.org/10.1146/annurev.ea.14.050186.002425.