Lost Terranes of Zealandia: possible development of late Paleozoic and early Mesozoic sedimentary basins at the southwest Pacific margin of Gondwana land, and their destination as terranes in southern South America

. Latesl Precambrian to Ordovician metasedimentary suecessions and Cambrian-Ordovician and Devonian Carboniferous granitoids form tbe major par! oftbe basemenl of soutbem Zealandia and adjacenl sectors of Antarctica and southeastAustralia. Uplift/cooling ages ofthese rocks, and local Devonian shallow-water caver sequences suggest tbal final consolidation oftbe basemenl occurred tbrough Late Paleozoic time. A necessary consequence oftlris process would have been contemporaneous erosion and tbe substantial developmenl of marine sedimentary basins al tbe Pacific margin of Zealandia. Tbese are fouod nowhere al tbe presenl day, suggestiog tbal tbe basins have been losl by !eclonic erosion, perhaps in a margin-parallel dextral translation similar to late Paleozoic-Mesozoic suspect terranes of New Zealand. A probable detrital zircon age pattem is asscmblcd for these lost Zealandia scdimcnts, and thcn comparcd with tbose ofpro-Jurassic (probable Triassic to Devonian) metasedimentary rocl,. in tbe Chilean archipelago. Significanl Mesoproterozoic, latest Neoproterozoic-Cambrian and Devonian-Carboniferous detrital zircon age components are common to botb, tbus supporting a possible Chilean !errane destioation for tbese 'Iost terranes of Zealandia'.


Introduction
Zealandia is that par! ofthe greater New Zealand region underlain by continental crust (Sutherland, 1999) and which fonnerly constituted a sector ofthe Pacific margin ofthe Gondwanaland supercontineot A tit of Zealandia with the adjacent continental components ofGondwanaland can be made against easternAustralia by closing the Tasman Sea, and the Ross Sea regions ofAutarctica by closing the southwest Pacific Ocean.This then reveals the original continuity of several late Precambrian and early Paleozoic fold belts as demonstrated for example by common sedimentary basins, plutonic ares and metamorphic belts (Adams, 1981;Bradshaw et al., 1983;Glen 2005).Despite a large pruportion (95%) ofZealandia being below sea-Ievel and unavailable for study, its geological history compares well with its Australian andAutarctic continental neighbours, but an important differeoce is that Silurian and Devonian sedimentary rocks are very local and limited and, apar! from a single, tiny (lOO m') seamount limestone occurrence (Jeokins and Jenkins, 1971), the Carboniferous is absent This paper examines tbis striking paucity of Silurian, Devonian and Carboniferous sedimentary rocks, and proposes that they were in fact developed in quite extensive sedimentary basins along the Zea1andia Pacitic margin, but were then trans1ated, as suspect terranes, to southem South America.
The substantial exhumation of sueh a large base-

S. Late Permian igneous rocks of tbe Median
Batbolitb or sedimentary rocks of tbe Brook Street and Murihiku Terranes (Fig. 1).Today, the crucial continental basement in southem Zealandia (Campbell Plateau) and West Antarctica (Marie Byrd Land) is hidden beneatb sea and ice.Since proven outcrop areas are extremely limited, estimates of tbe extent of sources (1 )-( 5) are very uncertain.Bradshaw el al.
The very extensive sandstone-dominated suecessions of sources (1) and (3) have plutonic, mostly granitoid, provenances, and thus, when reworked into contemporary late Paleozoic sediments, would contribute tbese characteristics.To tbis would be added similar detritos from granitoid sources (2) and (4).Thus, late Paleozoic sediments would be dominated by sands and silts witb relatively quartzose compositions, more potassic tban calcic, and witb elevated initial Sr isotopic ratios, >0.708 (Gray, 1990;Adams, 1997).There is, however, a complication closer to tbe Zealandia margin (and present-day New Zealand) where Late Permian sediment contributions from source (5), such as tbe Median Batbolith, and Brook Street and Dun Mountain-Maitai Terranes, might become important.The extension of none of tbese on to tbe Campbell Plateau is proven, altbough a linear magnetic anornaly extending 400 km soutbeast of New Zealand might be related to tbe Dun Mountain Ophiolite Belt (Sutberland 1999;Mortirner el al., 2002).Ifsource (5) was originally far larger, i.e, a large proportion has been removed by erosion, or if substantial extensions are in faet hidden on tbe Campbell Plateau and in Marie Byrd Land (Bradshaw el al., 1997), then sediments contributed from this source would be less quartzose, more volcaniclastic and with lower bulk initial "Sr/"Sr values <0.706.
An estimate of their relative extent would suggest that sources (1)-( 3) would be major zircon contributors to any Zealandia sedimentary basins, perhaps comprising 60-70% oftota!, with an important but minor contribution, perhaps 20-30%, from source (4), and an uncertain, but smallest input from source (5).This contrasts markedly with zircons delivered to the northeastemAustralian Gondwanaland margin to form the Torlesse, Waipapa and Caples sedimentary basins, which were dominated by source ( 5), ofien 40-50%, with only a minor contribution, 30-40%, from sources (1)-( 3), and very little from source (4).These pattems are exemplified in fignre 2 (upper) by zircon age distributions in Rakaia Terrane lOcks of the Torlesse Composite Terrane.Zircon age data from the Chilean complexes cited aboye are summarised in figure 2 (\ower) in a similar manner lo the equivalent data from the Permian-Triassic Rakaia Terrane ofNew Zealand displayed in figure 2 (upper).The Chilean zircon age pattems are seen to contrast with those from New Zealand.Since the Rakaia Terrane is allochthonous with respect to the southem Zealandia margin, it has zircon age data !hatdo not match the possible basement sourees there.However, the Chilean zircon data do conform lo that expected of a Zealandia margin sedimeotary basin.In particular there is a large zircon component (early Paleozoic and Precambrian) compatible with souree-types (1)-
The nearest examples of such a magmatic are would be io northern Patagonia (Augustsson et al., 2006).
was consolidated by Carboniferous times, and constituted extensive tracts latest Precambrian-Ordovician deep-marioe turbiditic metasediments, iotruded by Cambrian-Ordovician and Devonian-Carboniferous granitoids.A limited sedimentary cover comprised shollow-water Devonian rocks.There may have been local, but uncertain, extension oflhe Median Balholilh (and severol Permian-Triassic terranes) io New Zeoland soulheastwarda along Ihe Zeolandiamargin.Substantial uplift of this basement progressed through late Paleozoic time (at least Silurian lo Carboniferous), and Ihe consequent erosion and river transport should have provided volumioous sediments, of mainly granitoid compositions, to postulated marioe sedimentary basins at Ihe Pacific margin of Zeolandia.These sediments would have carried detritol zircon populations domioated by Mesoproterozoic, late Neoproterozoic-Cambrian-Ordovician, and Devonian-Carboniferous age components.The present-day absence of any such late Paleozoic sedimentary basins at the soulhem Zealandia continental margio suggests that Ihey may have been tectonicolly transported elsewhere during the Mesozoic, as suspect terranes in a dextral, margin-parallel sense, similar lo Ihat of the Easlem Province (Permian lo Crelaceous) terranes ofNew Zealand.A possible destination for Ihese terranes would be io the metamorphic complexes of Chilean Patagonia.Detrital zircon age populations in Iheir metasediments compare well wilh Ihose Ihat might be expected in Zeolandia sedimentary basins, and thus a suspect terrane interpretation is permissible.