We have achieved the principal objective of ODP Leg 187: using isotopic compositions of Leg 187 basalt glasses to map the Indian/Pacific MORB mantle domain boundary location throughout the last 25–30 m.y. The IMM/PMM boundary is an intrinsic feature of the AADA, and this geometry precludes a long-standing alternative hypothesis involving long-term, sustained rapid migration of PMM from the east. Leg 187 isotope systematics enable us to locate the off-axis trace of the isotopic boundary with remarkable precision, within ±50 km. For the last ~25 m.y., the boundary has been located ~100 km east of the midline of the residual depth anomaly, between the –400- and –500-m depth anomaly contours (Fig. F2). A pronounced westward deviation of the boundary trace near Sites 1158 and 1157 is interpreted as the record of a transient westward propagation event that culminated at ~20 Ma. This event is very similar in scale and duration to the well-documented recent (3–4 Ma) westward propagation of the PMM source beneath the SEIR in the easternmost AAD, displacing the preexisting IMM source. To the north of the Leg 187 work area (<42°S and >26 Ma), sample density is sparse and the boundary is poorly constrained.
A remarkable and unexpected outcome of Leg 187 is the observation that near-axis IMM lavas differ from their Leg 187 counterparts in important aspects of their trace and major element compositions and in the range and diversity of their isotopic compositions. Many of the elemental differences are in parameters that are most strongly influenced by depth and extent of melting, but differences in source composition are also apparent. Overall, IMM basalts are less evolved with a smaller range of generally higher MgO contents than their PMM counterparts. IMM lavas are also more variable in most other major and trace element contents at constant MgO, indicating their derivation from a more variable melting regime. In contrast, PMM lava compositions extend over a broader range of MgO values but are relatively invariant at a given MgO content, suggesting relatively uniform melting conditions and a dominant control by crystal fractionation processes. IMM lavas have consistently been derived by smaller degrees of melting than their PMM counterparts throughout the last ~30 m.y. Quantitatively, this difference shifted during the period of 14–7 Ma, for which there are no samples, with a significant apparent decrease in overall degree of melting for both groups. We infer that the decrease was related to the ~14- to 12-Ma relocation of the AADA and the IMM/PMM boundary from Zone A to Zone B as a consequence of the northeastward migration of the 127°E Fracture Zone that forms the eastern boundary of the AAD.
Dynamic plate motion reconstructions by Gurnis et al. (1998) and Gurnis and Müller (2003) suggest that the trace of the AADA through time, and the current location of the AAD, coincide with the location in a fixed mantle reference frame of a long-lived, pre-100-Ma western Pacific subduction zone. Low mantle temperatures beneath the AADA are inferred to derive from refractory subducted material accumulated in the lowermost upper mantle beneath this subduction zone. Mantle flow models incorporating this dynamic hypothesis suggest relatively recent (<20 Ma) entrainment of arc-derived material into the upwelling upper mantle. Although the relatively recent changes in apparent extent of melting beneath the AAD are qualitatively consistent with the entrainment aspects of this model, basalt glasses from Leg 187 and from young dredges do not display any of an array of possible geochemical signatures that could be associated with the subduction environment. We also argue that Hf-Nd isotopic systematics are consistent with a mantle source that is too old and too widespread to be derived from a Mesozoic subduction complex (this argument is controversial—see Kempton et al., 2002). Physically, the entrainment model is also suspect because it does not allow for realistic mantle viscosity gradients.
We conclude that the IMM/PMM boundary has been associated with the residual depth anomaly (AADA) through most or all of the separation of Australia from Antarctica. The origin of the distinctive isotopic characteristics of the two provinces remains controversial and has most recently been discussed by Hanan et al. (2000a, 2000b, submitted [N1]) and Kempton et al. (2002), but it seems clear that a significant component of the IMM source must have been derived from ancient continental material as a by-product of the breakup of Gondwana. This conclusion does not preclude an arc source for much of this material, though the need for long ingrowth times to establish observed Hf-Nd isotope ratios seems to preclude a young (Mesozoic) arc source. Prior to ~12 Ma, the AAD and the AADA were geographically separate entities that have subsequently been brought into conjunction by the northeastward absolute motion of the Australian/Antarctic plate boundary. We postulate that as this conjunction came about, cooler mantle temperatures were established beneath the AAD. This led, in turn, to reduced overall extents of melting, reduced magma supply to the intermediate-spreading SEIR, and the establishment of an extensional regime dominated by listric faulting (Christie et al., 1998). Young IMM lavas from the AAD are, therefore, most likely formed by a low-degree, relatively discontinuous melting process that allows the full diversity of the complex IMM source region to be expressed in lava geochemistry. The origin of the AAD as a morphologically anomalous section of the SEIR appears to precede the maximum influence of the depth anomaly. A clear explanation for its onset may have to await a better understanding of the dynamics of this unique region.