GEOCHEMISTRY AND PETROLOGY

Lead Isotopic Signatures

Pyle et al. (1992) showed that the near-zero–age IMM and PMM lavas define discrete fields in any of several binary combinations of the isotopes of Sr, Nd, and Pb. Relative to PMM lavas, IMM lavas have higher ⁸⁷Sr/⁸⁶Sr and lower ²⁰⁶Pb/²⁰⁴Pb, ²⁰⁷Pb/²⁰⁴Pb, and ²⁰⁸Pb/²⁰⁴Pb. More recently, Pearce et al. (1999), Hanan et al. (2000a, 2000b, submitted [N1]), and Kempton et al. (2002) have shown that binary plots of Hf and Nd isotopic ratios can also offer an effective IMM–PMM discriminant, with IMM lavas having higher ¹⁷⁶Hf/¹⁷⁷Hf and lower ¹⁴³Nd/¹⁴⁴Nd than their PMM counterparts. Here we have chosen to use Pb isotope variations, primarily ²⁰⁶Pb/²⁰⁴Pb, of fresh basaltic glass samples as our definitive discriminant of IMM–PMM sources in keeping with our previous work on the near-axis and axial SEIR data set (Pyle et al., 1992, 2000) (Fig. F7).

Pb isotope ratios for selected Leg 187 glass samples were analyzed by thermal ionization mass spectrometry (TIMS) at San Diego State University (California) in collaboration with B. Hanan and at the University of Hawaii by D. Pyle (Pyle et al., 2000; D. Pyle, pers. comm., 2004). These data are complemented by a comprehensive suite of whole-rock and glass Sr and Nd isotope ratios (Pedersen et al., this volume) and by a suite of predominantly whole-rock Hf, Pb, Nd, and Sr isotope ratios (Kempton et al., 2002). To classify the Leg 187 and other off-axis samples, we use IMM and PMM fields based on an updated compilation of near-zero–age dredge data (Fig. F7). A few zero-age samples from the isotopic boundary zone in Segment B5 plot between the IMM and PMM fields and define a TMM domain. TMM lava compositions are consistent with their derivation from a mixture of IMM and PMM source components. Several samples that we now define as TMM were designated as TPMM (for Transitional Pacific MORB mantle) during Leg 187 (Christie, Pedersen, Miller, et al., 2001), and this term was subsequently used by Kempton et al. (2002). We no longer include "Pacific" as a modifier for transitional lavas, as TMM defines a continuum between PMM and IMM.

Our Ba-Zr–based shipboard source designations (reported by Christie, Pedersen, Miller, et al., 2001) compare well to our (definitive) classification based on later shore-based isotopic determinations. The isotopic data confirm that we correctly located the Indian/Pacific boundary during Leg 187 using our Ba vs. Zr/Ba proxy (Fig. F8), despite minor differences between the Leg 187 and zero-age compositional fields. For several samples that plot close to the field boundaries or in a low-Zr/Ba extension of the Pacific field, we used an onboard TPMM designation. These included IMM lavas from Sites 1152, 1155, and 1162 and one PMM lava from Site 1158. A single IMM sample from Site 1164 was designated PMM. Although these sample designations led us to classify some sites ambiguously as transitional or mixed, no IMM site was misidentified as PMM or vice versa. The isotopic data resolve these ambiguities and greatly improve the resolution of the boundary (Figs. F2, F8).

Several sites that yielded either TMM lavas or lavas of more than one type (mixed sites) are of particular importance for our interpretation of the Indian/Pacific boundary configuration. They include the following.

Site 1157

Site 1157 is mixed IMM/PMM. This is the only Leg 187 site to include both IMM and PMM lavas. A single glass from a core catcher of Hole 1157A is IMM according to both Pb and Hf-Nd isotopic data, but three shallower whole-rock samples from this hole are PMM (Kempton et al., 2002). Three whole-rock samples from Hole 1157B, ~200 m away, are IMM. Hole 1157B extends to greater depth through more intact pillow basalts than Hole 1157A, which returned only rubble interpreted as erosional in origin. Although stratigraphic sequences cannot be reliably inferred from drilled rubble, it is probable that PMM lavas overlie IMM at this site.

Site 1158

Site 1158 is mixed TMM/PMM. A single whole-rock sample from Hole 1158A lies well within the IMM field for Hf-Nd isotopic ratios and was classified as IMM by Kempton et al. (2002) (see "Hf Isotopic Signatures"), but its ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb ratios place it within the standard TMM field (Fig. F7), although it has unusually high ²⁰⁷Pb/²⁰⁴Pb. A glass sample and a basalt whole rock from Hole 1158B and a diabase from Hole 1158C are clearly PMM. Hole 1158A, which is 270 m and 440 m south of Holes 1158B and 1158C, respectively, intersected basement at 70–80 m greater depth below the seafloor, in this case suggesting that PMM lavas may overlie TMM lavas.

Site 1153

All analyzed samples from this site plot within a TMM field that lies between the IMM and PMM fields similar to present-day TMM lavas from Segment B5 west (Pyle et al., 1992).

DR 10

Dredge 10 of Lanyon et al. (1995) lies slightly north of Sites 1153 and 1154 (Fig. F2) and was considered by those authors to be of Indian affinity. Although accurate age corrections are difficult to determine for these samples, by our current classification, lavas from this dredge are from a TMM source.

We interpret these four sites as lying within the Indian/Pacific isotope boundary zone. At zero age, the transition zone defined by the occurrence of TMM lavas extends <50 km along axis (Pyle et al., 1992) and it seems reasonable to interpret the presence of TMM lavas at Sites 1153 and 1158 and in DR 10 as defining a boundary zone of similar width. At Site 1158, PMM lavas occur only ~50 km east of the 127°E Fracture Zone (Fig. F2), and this distance may define the maximum width of the transition zone at this latitude. The alternation of PMM, TMM, and IMM lavas along a generally north-south transect through Zone A west; the anomalous occurrence of PMM and TMM lavas to the west of IMM lavas across the closely spaced Sites 1158, 1161, and 1162; and, especially, the probable emplacement of PMM lavas over TMM or IMM lavas at these sites all require that the mantle sources for both lava types exist in close proximity and that these distinct magmas are effectively isolated en route to the surface. We interpret these observations as resulting from a short-lived westward propagation event, comparable to the recent migration of the Indian/Pacific boundary across Segment B5. This interpretation is supported by a pair of distinctive oblique seafloor lineaments that pass close to the sites (Marks et al., 1999) (Fig. F2) and coincide roughly with the location of a propagating rift inferred from aeromagnetic data by Vogt et al. (1983).

Sr and Nd Isotopic Signatures

Pedersen et al. (this volume) focused on the Sr-Nd isotope systematics of Leg 187 glasses and especially on their spatial variability. They showed that Leg 187 isotopic data define two spatially separated intersecting linear trends in Sr-Nd and Pb-Hf isotope plots. In their Nd-Sr binary isotopic diagram (Fig. F9, lower panel), glasses from sites in Zone B, west of the 127°E Fracture Zone, plot along a well-defined negative western trend. Glasses from sites east of the 127°E Fracture Zone form a flat eastern trend with relatively invariant ¹⁴³Nd/¹⁴⁴Nd. Similar intersecting linear trends can also be discerned in the Pb-Pb isotopic binary plots discussed above (Fig. F7). The locations of individual sites or samples along these trends appear to be related to distance from the 127°E Fracture Zone and/or the depth anomaly (Pedersen et al., this volume). The development of paired linear trends suggests an unusual scenario in which both PMM and IMM end-members have mixed with a third, intermediate boundary end-member, but not with each other. The situation appears to have persisted during the time span represented by Leg 187 basalts, but not until the present.

The eastern (PMM) and western (IMM) end-member compositions lie within the much broader overall PMM and IMM compositional ranges. They are most strongly expressed at Sites 1160 and 1152, which are, respectively, the easternmost and westernmost Leg 187 sites in terms of their distance from the 127°E Fracture Zone. In Nd-Sr and Hf-Pb isotopic space, the intermediate end-member appears to be most strongly expressed at Site 1162 (Fig. F9). Site 1161, which consistently plots at the intersection of the eastern and western trends, is also strongly influenced by this end-member. Site 1157 lavas plot close to the intersection and along both trends, consistent with the mixed source affinities of lavas at this site (see "Lead Isotopic Signatures" above). Because both these sites lie close to the Indian/Pacific mantle boundary, Pedersen et al. (this volume) refer to the intermediate end-member as the "boundary component."

Mixing relationships similar to those described by Pedersen et al. (this volume) as the eastern trend were recognized by Kempton et al. (2002) (see "Hf Isotopic Signatures"), who showed that IMM lavas from Leg 187 sites in Zone A west have generally higher Hf and Nd values than those from other Leg 187 sites (Fig. F10). Because the paired compositional arrays are discernible only in the Leg 187 data set and the boundary component appears to be restricted to Zone A west, the three-component mixing that they represent was a localized and transient phenomenon that likely terminated at ~12–15 Ma, as northeast migration of the SEIR moved the 127°E Transform from west to east, first across the sublithospheric mantle locus of the depth anomaly and, 1–2 m.y. later, across the Indian/Pacific mantle boundary (see fig. 4 in Marks et al., 1999).

Hf Isotopic Signatures

Kempton et al. (2002) reported an extensive suite of new Nd, Pb, and Hf isotopic data for basaltic whole-rock and glass samples from Leg 187, supplemented by near-zero–age samples from the 1976 Vema dredges reported by Klein et al. (1988). Although there is almost complete overlap in the fields of published ¹⁷⁶Hf/¹⁷⁷Hf data from the Pacific and Indian Oceans, basalts derived from IMM and PMM sources in the AAD and Zone A can be effectively discriminated using Nd and Hf isotope systematics. Relative to PMM, the IMM field is systematically displaced to lower ¹⁴³Nd/¹⁴⁴Nd for a given ¹⁷⁶Hf/¹⁷⁷Hf value. Kempton et al. (2002) defined a linear discriminant separating PMM from IMM in Nd-Hf isotopic space (Fig. F10) and used this guideline to assign each Leg 187 drill site to a mantle isotopic domain. Newly available Hf isotope analyses of a much larger suite of near-zero–age dredged glasses (Hanan et al., 2000a, 2000b, submitted [N1]) generally confirm the validity of this discriminant, although three Segment A1 (PMM) dredges plot within the IMM field.

The Hf-Nd source assignments of Kempton et al. (2002) differ from the standard (Pb isotope) assignments at only two Leg 187 sites. As with the shipboard assignments, the differences are minor and have no effect on the inferred location of the Indian/Pacific mantle boundary. Sites for which the Hf isotopic designation differs from the standard are discussed briefly below.

Site 1153

Site 1153 is mixed IMM/PMM. These basalts are designated as transitional (TPMM) by Kempton et al. (2002) because the two analyzed samples are close to, but on opposite sides of, the discriminant boundary (Fig. F10).

Site 1158

Site 1158 is PMM/IMM. A whole-rock sample from Hole 1158A lies well within the IMM field for Hf-Nd isotopic ratios. Our standard designation based on Pb isotopic ratios is TMM, although it has high ²⁰⁷Pb/²⁰⁴Pb relative to other TMM samples. This is the only sample for which the two isotopic systems differ in their source designation. All other samples from this site are unequivocally PMM.

Kempton et al. (2002) recognized the importance of Zone A west lavas as a local end-member consistent with the Pedersen et al. (this volume) definition of the boundary component (see "Sr and Nd Isotopic Signatures"). IMM lavas from Zone A west have relatively high Hf and high Nd relative to all other Leg 187 sites (Fig. F10). Within the Kempton et al. data set, Zone A west lavas have higher Hf than most AAD and Zone A lavas overall, but additional data from Hanan et al. (2000a, 2000b, submitted [N1]) (Figs. F10, F13) extend the overall IMM field to higher values than those of the Zone A west Leg 187 sites. This observation has important implications for interpretations of the geodynamic origin of the AAD and the depth anomaly (see "Geochemical Evidence for Entrainment" in "Constraints on the Origin of the ADD and the AADA" in "Dynamics and Origin of the AAD").

Location of the Indian/Pacific Isotopic Domain Boundary

Perhaps the most remarkable aspect of the Leg 187 isotope systematics is the consistency of interpretation over the entire region and the absence of significant discrepancies. The isotopic data discussed in the preceding sections enable us to locate the off-axis trace of the isotopic boundary with remarkable precision, within <50 km, over much of the last ~25 m.y. Between the SEIR (~50°S) and 43°S, the boundary trace shown in Figure F2 is drawn to pass to the east of the easternmost IMM sites and west of the westernmost PMM and TMM sites. The boundary is located ~100 km east of the axis (midline) of the depth anomaly, between the –400- and –500-m depth anomaly contours. A pronounced westward deviation of the boundary trace required by the presence of PMM lavas at Sites 1158 and 1157 is interpreted as recording a transient westward propagation event very similar in scale and duration to the well-documented recent (3–4 Ma) propagation of a PMM source into a preexisting IMM source region beneath Segment B5. North of 42°S (>26 Ma), the depth anomaly broadens and deepens, sample density is sparse, and the boundary is poorly constrained. The presence of TMM lavas at Site 1153 and in DR 10 suggests that the boundary passes close to these sites, rather than turning to the east with the depth anomaly contours.

Olivine Compositions and the Origins of Indian- and Pacific-Type MORB

Sato (this volume) undertook a thorough study of mineral compositions for a suite of Leg 187 basalt samples. He identified a group of primitive basalts that can reasonably be assumed to have evolved from their primary compositions by crystal fractionation of olivine alone. For each of these samples, he used an olivine addition calculation to estimate a primary magma composition that would be in equilibrium with mantle olivine. These equilibrium mantle olivine compositions can, in turn, be used to estimate the depth at which the primary magma last equilibrated with mantle material.

Primary Pacific-type MORB compositions determined in this way are derived from greater depths (15 kbar and 45 km) than primary Indian-type MORB compositions (10 kbar and 30 km). Consequently, even the most primitive Pacific-type lavas must have undergone more fractionation prior to eruption than their Indian-type counterparts. This conclusion is consistent with the presence of cooler mantle beneath the AAD.

Major and Trace Element Constraints

A remarkable and unexpected outcome of Leg 187 is the observation that near-axis (0–7 Ma) IMM lavas differ from their Leg 187 counterparts in important aspects of their trace and major element compositions. Many of these differences are in parameters that are most strongly influenced by depth and extent of melting (C.J. Russo et al., pers. comm., 2004), but differences in source composition are also apparent.

As a group, the IMM lava population is distinct from the PMM group in having significantly more variable major, minor, and trace element contents at a given MgO content. IMM lavas are characterized by lower FeO, CaO, and TiO₂ and higher K₂O, Al₂O₃, Na₂O, and SiO₂ than PMM lavas, suggesting a wider and more variable range of primary compositions (Fig. F11A) (Pyle, 1994; C.J. Russo et al., pers. comm. 2004). PMM lava compositions are relatively invariant at a given MgO content and the near-axis PMM population extends over a broader range to significantly lower MgO values, suggesting a dominant control by crystal fractionation processes, but this distinction is barely apparent in the Leg 187 samples.

One key compositional difference consistent with the derivation of IMM magmas by lower extents of melting from shallower depths in a cooler mantle regime is expressed in fractionation-corrected Na₂O and FeO contents, expressed as Na8 and Fe8 (Klein and Langmuir, 1987). Near-axis (0–7 Ma) IMM lavas have distinctly higher Na8 and lower Fe8 values than all other AAD and Zone A lavas (Fig. F12). They constitute one of the highest Na8 populations from major mid-ocean spreading centers, consistent with deep axial depths and low extents of melting (Klein and Langmuir, 1987). All PMM lavas have lower Na8 and higher Fe8, consistent with greater mean extents and deeper mean depths of melting. Leg 187 IMM lavas do not have the high Na₂O contents of their near-axis counterparts. Their Na8-Fe8 values overlap a high-Na8 subset of the near-axis PMM field, whereas Leg 187 PMM lavas overlap a different, lower-Na8 subfield. Other characteristics of IMM lavas that are consistent with lower extents and/or shallower depths of melting include higher SiO₂ and Al₂O₃, lower CaO, and generally lower CaO/Al₂O₃ than PMM lavas of the same age group. In all these parameters, the differences between Leg 187 IMM and PMM lavas are similar to those between the near-axis populations, but the boundary is shifted toward, or into, the on-axis PMM field (Fig. F11).

Major and minor element parameters that most strongly reflect source composition include TiO₂ and K₂O, in addition to Na₂O (Fig. F11A). Near-axis IMM lavas have lower TiO₂ at a given MgO content (Fig. F11A) than all other AAD region lavas. This distinction is inconsistent with lower extents of melting, suggesting a difference in source composition. In contrast to this apparent source depletion, K₂O and K/Ti values for near-axis IMM lavas are higher and more variable than expected for a simple decrease in extent of melting. This variability suggests a variable and complex source composition. Leg 187 lavas also fall into separate IMM and PMM fields for some but not all elements. These fields differ from each other in the same sense as the near-axis IMM and PMM fields but with field boundaries displaced toward and partially or completely overlapping the near-axis PMM fields (Figs. F8, F11). In terms of incompatible trace element contents, near-axis IMM lavas have high Ba/La, Ba/Zr, and La/Sm ratios relative to all other AAD region lavas (Fig. F11B). Leg 187 IMM lavas have generally lower values of these ratios, but both IMM groups define distinct fields that have little or no overlap with PMM lavas. These differences are consistent in direction with lower extents of melting for IMM lavas, but their magnitude is too great to be solely attributable to extent of melting.

Overall, the contrasts between near-axis IMM lavas, Leg 187 IMM lavas, and all PMM lavas seem to be primarily related to lower extents of melting in the Indian domain that have diminished significantly through time. In parallel with this apparent decrease, there appears to have been an overall increase in incompatible element contents, reflecting greater contributions to individual lavas from an enriched source component. Whether this source component has become more abundant in recent time or whether it is simply contributing a greater fraction of the diminishing overall magma production is not yet clear.