IGNEOUS
PETROLOGY AND GEOCHEMISTRY
Introduction
Drilling at Site 1141
recovered material from 72.1 m of igneous basement (see "Operations"
and "Physical Volcanology"). We did not
recover the contact with the overlying sedimentary section, but the lowermost
sedimentary rock recovered above basement is a sandy nannofossil-bearing
foraminifer limestone that contains microfossils of middle to late Eocene age
(see "Biostratigraphy" and "Lithostratigraphy").
Thus, the minimum age of basement at this site (and Site 1142; see "Lithostratigraphy")
is near the age inferred for rifting between Broken Ridge and the Kerguelen
Plateau (e.g., Mutter and Cande, 1983).
The basement section is
divided into six units (Cores 183-1141A-13R through 24R); Figure F38
contains a summary of unit boundaries, curated depths and thicknesses, and
recovery (see also Table T5). All
but the first unit are basaltic lavas that appear to have been emplaced
subaerially. We have defined the basaltic lava units on the basis of marked
differences in flow structure, oxidation, and/or vesicularity (see "Physical
Volcanology"). The most diagnostic criteria for stratigraphic
classification at Site 1141 were distinctive red or green color (e.g., the Unit
2/3 boundary) accompanied by a change from massive structure (interpreted to
correspond to flow interiors) to brecciated structure (interpreted to correspond
to flow margins) (e.g., the Unit 5/6 boundary) and/or a change in vesicularity
and vesicle shape and size (e.g., the Unit 4/5 boundary). Units 5 and 6 are also
distinguished by macroscopic differences in phenocryst content. Each lava unit
corresponds to what we provisionally interpret to be a single eruption event.
Note, however, that Unit 4 may contain an additional flow boundary and possibly
a dike (Section 183-1141A-18R-2), but a definite identification of either is
precluded by limited recovery (Table T5)
and requires detailed shore-based study.
Macroscopic
Features of Basement Lithology
Unit 1 (1.0 m curated
thickness) is represented by only three small pieces (2-3 cm) of angular,
rough-surfaced, moderately altered, medium-grained, plagioclase-clinopyroxene-olivine
gabbro, which could be originally from a dike or sill but could also have been
collected from a gravel bed. Because the rocks are likely to have been broken
during drilling, the shapes and surfaces of the pieces provide no insight into
which of these possibilities is correct.
Large portions of Units 2
and 3 are altered to clay and were disturbed considerably by drilling. Unit 2
(20.0 m curated thickness) is dominantly clay, with the exception of a few
pebble- to small cobble-sized pieces of soft, completely altered basalt near the
bottom of the unit. From these pieces, we determined that the protolith was
aphanitic to fine-grained aphyric basalt. Most of Unit 3 (8.3 m curated
thickness) is also clay; however, pieces recovered from the lower ~0.5 m of the
unit are only moderately altered. The basalt in these pieces is fine to medium
grained and sparsely plagioclase-phyric.
Unit 4 (19.6 m curated
thickness) varies from completely altered to locally slightly altered, aphyric
to moderately plagioclase- or plagioclase-olivine-phyric basalt. Alteration in
most of the massive interior portion of the unit is moderate. The basalt in
Section 183-1141A-18R-2 and in the bottom of Section 183-1141A-18R-1 is notably
fresher, finer grained, and denser than that above or below in Unit 4. This
interval could simply be a dense, better-preserved zone within the interior of
the flow. Alternatively, it could be a portion of a dike intruded into Unit 4;
however, no contacts were recovered. As with the entire basement sequence, the
magnetic polarity of this zone is normal (see "Paleomagnetism").
The top piece of Section 183-1141A-18R-1 is brecciated and could represent part
of a flow top, but, again, no contact was recovered.
Alteration in Unit 5 (8.0
m curated thickness) is high to complete in the upper portion of the unit but
decreases to moderate in the lower portion (Sections 183-1141A-21R-1 through
21R-3) (see "Alteration and Weathering").
The basalt is sparsely olivine-phyric. Most of Unit 6 (15.3 m curated thickness)
is also sparsely olivine-phyric, but a range from aphyric to moderately
olivine-plagioclase-phyric basalt is present. Alteration ranges from complete to
moderate and, in Sections 183-1141A-23R-1 and 23R-2, to locally slight.
Petrography
Table T8
summarizes the thin-section mineralogy and textures of samples from Sites 1141
and 1142. All basalt samples studied from these sites have relatively uniform
mineralogical compositions and textures.
Site 1141 samples from
Units 2, 3, and most of 4 are aphyric to plagioclase-phyric basalts. Plagioclase
phenocrysts commonly exhibit sieve textures as well as overgrowths (Fig. F39).
In the lower part of Unit 4 and in Units 5 and 6, partially altered olivine is a
persistent phenocryst phase (Fig. F40).
The groundmass consists of plagioclase, clinopyroxene, titanomagnetite, and
altered glass. In the lower part of Unit 4, and throughout Units 5 and 6,
olivine is present in the groundmass. Another minor but common groundmass phase
is apatite, which is found as minute needles concentrated in interstitial glassy
regions. These two features suggest that the basalts are alkalic, an observation
that is confirmed by geochemistry.
Textures in the groundmass
range from intergranular to intersertal or trachytic, but in coarser-grained
interiors of the larger flows, subophitic to ophitic textures have developed
(Fig. F41). Samples from the
margins of Units 5 and 6 are vesicular and finer grained, but have essentially
the same mineralogy as the more slowly cooled flow interiors.
The degree of alteration
as observed in thin section is moderate to high from the top of Unit 1 to the
upper part of Unit 6. In this interval, the mafic minerals and groundmass glass
are completely replaced by fine-grained carbonate (Fig. F42),
clay, and iron oxides. Crystalline goethite is present and is an alteration
product of titanomagnetite (Fig. F43).
However, even in these rocks, a large proportion of plagioclase has escaped
alteration. In contrast, portions of the middle to lower part of Unit 6 are well
preserved and samples from this interval retain a large proportion of fresh
phenocryst and groundmass olivine (Fig. F40).
Introduction
Drilling at Site 1142
penetrated 50.9 m of basement. Four pieces of basalt, totaling 0.3 m in curated
thickness, were recovered from the bottom of the interval above the basement
sequence (interval 183-1142A-1W-1, 0-41 cm) (see "Operations").
The contact between igneous basement and the overlying sedimentary sequence was
not recovered, but Section 183-1142A-1W-1 includes two pieces of the middle to
late Eocene, sandy, nannofossil-bearing foraminifer limestone similar to that
found at Site 1141 (see "Biostratigraphy"
and "Lithostratigraphy"). Moreover,
small amounts of this limestone are cemented onto the weathered surfaces of the
basalt pieces in this section; thus, the basalt and limestone were in contact
before drilling, although the contact was not necessarily a conformable one.
Within the basement sequence, many of the volcanological features that were
useful in determining unit boundaries at the other Leg 183 sites are not present
in the material recovered at Site 1142 (see "Physical
Volcanology"); however, major lithologic changes define six
distinct basement units. The upper five units provide no evidence for submarine
emplacement; however, Unit 6 contains basalts that were possibly emplaced in
water (see "Physical Volcanology").
Table T6 presents a summary of
the unit boundaries, curated depths and thicknesses, and recovery (see also Fig.
F44).
Macroscopic
Features of Basement Lithology
Unit 1 is massive basalt
with a recovered thickness of 1.9 m and an estimated thickness of 8.8 m (Fig. F44).
There is no curated thickness for this unit. We began coring within the unit
(see Table T6; Fig. F44).
The basalt is slightly to moderately altered, fine grained, and aphyric to
sparsely plagioclase-olivine-phyric. The basalt fragments from the upper portion
of the unit (Section 183-1142A-1W-1) have prominent (~0.3 to 5 cm thick)
oxidation halos at their margins. The presence of oxidation halos and adhering
limestone suggests periods of both subaerial and submarine exposure and
weathering, possibly related to the Eocene rifting of Broken Ridge from the
Kerguelen Plateau. Basalt from the lower portion of the unit (Sections
183-1142A-2R-1 and 2R-2) lacks oxidation halos (and associated limestone).
Unit 2 (1.2 m curated
thickness) consists of a single section (Section 183-1142A-3R-1) containing 20
moderately to completely altered cobble-sized pieces of genetically unrelated
rock types. With the exception of material in the core-catcher of this section,
these rocks represent all that was recovered of a 9.6-m interval. The rocks
include volcanic breccia (see "Physical Volcanology"
for description), clinopyroxene-phyric basalt, plagioclase-phyric basalt,
olivine-phyric basalt, aphyric basalt, and feldspar- and feldspar-quartz-phyric
felsic volcanic rocks. Approximately half of the pieces are round to subround,
with abraded, slightly weathered surfaces; this feature, together with the
mixture of rock types present, may indicate that we sampled some type of
immature, uncemented sedimentary deposit, perhaps a near-source debris flow or a
talus pile. The felsic rocks demonstrate that a wider range of igneous
lithologies is present in the upper levels of basement in this region of Broken
Ridge than represented by the basaltic lavas drilled at either Site 1141 or
1142.
Unit 3 (19.3 m curated
thickness) is a completely altered brecciated aphanitic basalt; the upper
portions are aphyric (Sections 183-1142A-3R-CC through 5R-1) and the lower
portion (Sections 183-1142A-5R-2 through 6R-1) is aphyric to moderately
olivine-plagioclase-phyric. Much of the unit is highly disturbed by drilling,
but the lower portion contains more massive and coherent breccia and thin (~0.1
m) lava lobes. A description of the textural features used to subdivide the unit
into three subunits is in "Physical Volcanology".
Unit 4 (4.3 m curated
thickness) is a well-indurated, normally graded claystone or mudstone. The red
clay matrix supports very coarse sand-sized to small-granule-sized clasts of
very highly to completely altered lithic fragments and crystals of quartz and
altered feldspar. This unit is interpreted as a mudflow deposit (see detailed
description and discussion in "Physical Volcanology").
Unit 5 (8.8 m curated
thickness) is composed of aphanitic, aphyric volcanic breccia. Alteration is
nearly complete, except near the base of the unit (Section 183-1142A-8R-1),
where some very highly altered material is preserved (see "Physical
Volcanology" and "Alteration and
Weathering"). The only hint of the original rock type is provided
by very rare quartz crystals in Section 183-1142A-7R-2, suggesting an evolved
composition.
Evidence of possible
eruption into water is provided by Unit 6. This unit is composed of 
8.5
m of nonvesicular, massive, and possibly pillowed basalt. The portion we
recovered contains at least two possible glassy margins, now altered to clay.
The basalt is aphyric and fine-grained to aphanitic. Rare plagioclase
phenocrysts are present in Sections 183-1142A-9R-2 through 9R-4. Alteration
varies from high to moderate in most of the unit, although portions of Section
183-1142A-10R-1 are only slightly altered. In addition to the presence of
possible glassy margins (Figs. F36,
F37), the pattern of
veining (see "Alteration and Weathering")
is consistent with that of pillow lavas. The near absence of vesicles (see "Physical
Volcanology") suggests eruption of the unit in relatively deep
water. Distinctive, curved, and commonly roughly concentric brown to orange
oxidation bands are abundant in all but the lowermost section of the unit
(Section 183-1142A-10R-3); these bands are similar to those in weathered pillow
lava lobes. However, such bands are not typical of submarine alteration and
probably represent subaerial weathering during the period of uplift and rifting
of Broken Ridge from the Kerguelen Plateau.
Petrography
The general petrographic
characteristics of the basement units from Site 1142 are presented in Table T8.
Unit 1 of Site 1142 is sparsely olivine-phyric basalt with an intergranular
texture. Although relatively fresh, with large proportions of all silicate
minerals including unaltered olivine (Fig. F45),
it contains large patches of carbonate in the groundmass. Thin sections from the
best-preserved parts of the underlying Units 2, 3, and 6 all show moderate to
high alteration. Of the primary phases, only a portion of the plagioclase and
magnetite has not been replaced by secondary minerals. Textures, however, are
well preserved. Unit 3 is a moderately olivine-plagioclase-phyric basalt with
plagioclase that exhibits sieve textures and olivine that is completely altered
to clay (Fig. F46). A Unit 6
sample from a possible pillow margin (183-1142A-9R-4, 3-7 cm) contains an
apparent outer glassy rind. The rind is brecciated, replaced by
cryptocrystalline goethite and clay, and intruded by multiple calcite veins;
despite this, it retains the appearance of relict glass (Fig. F47).
Adjacent to the rind, the rock consists of fine, well-oriented plagioclase laths
and lesser amounts of granular clinopyroxene and titanomagnetite, all altered to
clay and secondary iron oxide. As in the region farther from the margin (e.g.,
Sample 183-1142A-10R-1, 38-40 cm), which has similar texture and mineralogy,
this portion of the rock is sparsely plagioclase-phyric, with plagioclase
phenocrysts that exhibit sieve-textured rims (Fig. F48).
Sites
1141 and 1142 Geochemistry
We report X-ray
fluorescence (XRF) analyses for eleven lavas, nine from Site 1141 and two from
Site 1142, in Table T9. The
samples are generally olivine-normative basalts, although three samples (two
from Site 1141 and one from Site 1142) are quartz normative. Compositions are
generally slightly alkalic (Fig. F49),
with data for all but two samples (one from Site 1141, a trachybasalt, and one
from Site 1142, a basaltic andesite) plotting in the basalt field. For this
discussion, all samples will be referred to as basalts.
The samples exhibit
moderate compositional variations (Site 1141: SiO2 = 45.3-51.6 wt%;
MgO = 3.16-6.45 wt%; Mg#, assuming 80% of the total iron is FeO, = 0.38-0.55.
Site 1142: SiO2 = 46.7-53.9 wt%; MgO = 2.77-6.56 wt%; Mg# =
0.39-0.55). At Site 1141, the three units sampled generally exhibit limited
variations in TiO2, Al2O3, Na2O, and
K2O from Unit 4 to Unit 6 (Table T9;
Fig. F50), except for the
uppermost sample from Unit 4 (183-1141A-17R-1, 27-31 cm), which contains notably
higher abundances of these elements. Silica exhibits little variation in Units 4
and 5 (47.2-48.2 wt%; Table T9),
but in Unit 6, it shows a range from 45.3 to 51.6 wt%. This may reflect variable
alteration but the samples with different SiO2 have similar Na2O
and K2O contents (Fig. F50).
The MgO contents steadily increase from the top of Unit 4 to the base of Unit 5,
then markedly decrease at the top of Unit 6 before steadily increasing down
through this unit. This pattern is generally paralleled by that of CaO. Fe2O3
varies more in Unit 4 (9.54-12.9 wt%), with a maximum value in the center of the
flow, than in the other two units (10.8-12.3 wt%).
Minor and trace elements
show similar downhole patterns to those of the major elements, with P2O5,
Nb, Ba, Rb, and Zr (not shown) contents decreasing downhole markedly from the
uppermost Unit 4 sample to similar abundances throughout the remainder of Unit 4
and Units 5 and 6 (Fig. F51).
Chromium increases from 125 ppm near the top of Unit 4 to 292 ppm at the base
but varies less through Units 5 and 6 (296-344 ppm). Primitive mantle-normalized
(Nb/Zr)N varies little in Unit 6 (~1.7-1.8) compared to Units 4 and
5, where it decreases from ~2.5 at the top of Unit 4 to ~1.7 in the middle of
the unit, then increases downward to ~1.9 within Unit 5. Variation in (Zr/Y)N
broadly parallels and (Zr/Ti)N is opposite to that displayed by (Nb/Zr)N.
In terms of the
geochemistry, no compositional differences in Unit 4 correlate with the
composite nature of this unit observed in the cores. The downhole compositional
variation can be explained by igneous fractionation and alteration processes.
For example, Samples 183-1141A-22R-3, 94-96 cm, and 22R-3, 142-146 cm, of Unit 6
are quartz normative (Table T9)
yet contain olivine as phenocrysts and as a groundmass phase. Their
quartz-normative character may be explained by the secondary silica present as
vesicle fill (see "Alteration and Weathering").
Variations in TiO2, MgO, and CaO, (Fig. F50),
along with P2O5, Ni, and Cr (Fig. F51),
are consistent with two inputs of fresh, comparatively primitive magma at the
beginning of the Unit 5 and Unit 6 eruptions. For example, at the top of Unit 6,
MgO has decreased to 3.19 wt%, from 6.45 wt% at the base. At the base of Unit 5,
MgO is again more than 6 wt%. We suggest these changes represent pulses of new,
more primitive magma, rather than crystal accumulation at the base of the flows.
We observed no increase in the proportion of phenocrysts at the bases of flow
units in thin section. From the Al2O3 and CaO variations,
it appears that olivine was joined by clinopyroxene, rather than plagioclase, as
a fractionating phase because CaO increases then decreases upward (i.e., as each
magma evolved) (Fig. F50),
whereas Al2O3 generally increases up the core (Fig. F50).
Also, clinopyroxene fractionation can affect incompatible element ratios (i.e.,
30% fractional crystallization of clinopyroxene results in an increase of 1 unit
in [Zr/Y]N, it results in an increase of <0.08 for [Nb/Zr]N).
However, clinopyroxene is not observed as a phenocryst phase in thin section.
Rather, sieve textured plagioclase phenocrysts are present. Resorption of
plagioclase could be the result of the influx of more primitive magma. The range
of incompatible element ratios may reflect open-system fractional
crystallization and/or magma mixing.
At Site 1142, the two
samples analyzed demonstrate compositional differences between Units 1 (basalt)
and 6 (basaltic andesite). Compositional variations between the samples from
Site 1142 suggest that the two units represented are not related by simple
closed-system fractional crystallization. For example, although the sample with
the highest MgO content (interval 183-1142A-2R-1, 50-53 cm, of Unit 1, with 6.56
wt% vs. interval 183-1142A-10R-1, 37-40 cm, of Unit 6, with 2.77 wt%) has the
highest Ni and Cr abundances (184 vs. 34 ppm and 303 vs. 21 ppm, respectively),
it also contains the highest TiO2, Nb, and Zr abundances (2.77 vs.
2.46 wt%, 26.3 vs. 16.1 ppm and 233 vs. 220 ppm, respectively). In addition,
primitive mantle-normalized incompatible element ratios differ significantly
between the Unit 1 and Unit 6 samples, which have, respectively, (Nb/Zr)N
= 1.8 and 1.2, (Zr/Ti)N = 1.6 and 1.8, and (Zr/Y)N = 3.2
and 2.3. As with Site 1141 basalts, clinopyroxene fractionation can affect these
ratios, but it is not observed as a phenocryst phase.
As Sites 1141 and 1142 are
only ~800 m apart, establishing similarities between the two basement sections
was difficult on the basis of core descriptions or petrography. Whole-rock
geochemistry demonstrates interval 183-1142A-2R-1, 50-53 cm, of Unit 1 has a
similar composition to those from Unit 6 from Site 1141 (Figs. F52,
F53, F54;
Table T9). This suggests that
Site 1142 Unit 6 is from beneath the recovered sequence at Site 1141. The
analyzed sample from Unit 6 (interval 183-1142A-10R-1, 37-40 cm) has a
composition similar to other Kerguelen Plateau basalts and it is tholeiitic
(Fig. F49).
Comparison
with Other Broken Ridge Samples
Mahoney et al. (1995)
reported geochemical data for dredge samples from the central and eastern Broken
Ridge and concluded that at least some of these samples contained evidence of a
continental lithospheric, possibly crustal, influence in their genesis
especially for the eastern part of Broken Ridge. These data are presented on
primitive mantle-normalized diagrams in Fig. F55,
where they are compared with those from Sites 1141 and 1142. It is evident that
the dredge and drill-core samples generally have different compositions. For
example, the Leg 183 samples have subparallel profiles with (Nb/Ce)N
ratios generally >1, distinct from the dredge samples, which have (Nb/Ce)N
<1. Only interval 183-1142A-10R-1, 37-40 cm, has a (Nb/Ce)N ratio
< 1. This sample appears to be similar to the dredge samples from Broken
Ridge (Fig. F55). Overall, the
other Sites 1141 and 1142 samples are enriched in the incompatible minor and
trace elements relative to the dredge samples and lack evidence of any obvious
continental influence. General distinctions between the Broken Ridge drill and
dredge samples are further emphasized in Figures F52,
F53, and F54.
Comparison
with Kerguelen Plateau Sites
The basalts from Sites
1141 and 1142 are compared with those from sites on the Kerguelen Plateau, as
well as the dredge samples from Broken Ridge in Figures F52
and F53. Except for lavas from
Site 1139, samples from Sites 1141 and 1142 have lower MgO and higher Al2O3,
TiO2, P2O5, Zr, Nb, Ti, and Ni contents
relative to lavas from other sites. The generally alkaline composition of Site
1141 and 1142 basalts is supported by the presence of apatite in the groundmass,
which is consistent with their elevated P2O5 contents
relative to most Kerguelen Plateau and other Broken Ridge samples (Figs. F52,
F54). Data for one of the two
basalts analyzed from Site 1142 (183-1142A-2R-1, 50-53 cm, the sample with the
higher MgO content) fall within the range of the Site 1141 data, whereas the
other sample plots with the majority of Kerguelen Plateau and Broken Ridge
basalts (Fig. F53). In terms of
Zr, Ti, and Nb, the Leg 183 Broken Ridge basalts are similar to those from Site
1139 and the alkali basalt from Site 748, but they contain lower Y contents for
a given Zr abundance, having similar (Zr/Y)N ratios to those of Site
1137 basalts (Fig. F54). In
terms of Ni, basalts from Sites 1141 and 1142 are distinct (Figs. F53,
F54). The elevated Ni contents are
consistent with the presence of olivine both as a phenocryst and groundmass
phase in some of the Site 1141 basalts. The only other samples exhibiting such
elevated Ni abundances are from the Site 748 alkali basalt. Curiously, lavas
with hot Ni content at Sites 748, 1141, and 1142 have high (Nb/Zr)N
(Fig. F54).
At Sites 1141 and 1142,
six basement units were identified. On the basis of macroscopic and microscopic
features, these units cannot be correlated, even though Site 1142 is only ~800 m
to the south of Site 1141. From geochemical evidence, however, we suggest that
Unit 1 of Site 1142 is equivalent to Unit 6 of Site 1141. This is consistent
with seismic data that suggest basement in this area has a persistent northerly
dip (see "Background and Objectives").
The observed differences could also be a result of paleotopographic control on
the deposition of the lavas and other basement units, or locally different
erosion rates.
Compared to most basement
lavas from the Kerguelen Plateau and Broken Ridge, basement lavas at Sites 1141
and 1142 have higher abundances of Ni and incompatible elements. The relatively
high Zr and Nb contents result in (Nb/Zr)N and (Zr/Y)N
ratios similar to Site 1139 basement lavas; only samples from the Site 748
alkali basalt have higher values (Fig. F54).
We infer that Site 1141 and 1142 basalts were generated by lower degrees of
partial melting than basement lavas from Leg 183 Sites 1136, 1137, 1138, and
1140. The elevated Ni contents, however, indicate that they experienced lower
amounts of olivine fractional crystallization. The ranges in incompatible
element ratios suggest fractional crystallization did not occur in a closed
system (especially as clinopyroxene is not observed as a phenocryst phase).
Sieve-textured plagioclase phenocrysts suggest that fractional crystallization
was accompanied by magma mixing.
Given that drilling at
Sites 1141 and 1142 penetrated only the uppermost few tens of meters of basement
from this many kilometers thick edifice, the rocks recovered clearly represent
the terminal stages of volcanism on this portion of Broken Ridge. The igneous
activity recorded in the drill cores was alkalic and may reflect lower degrees
of partial melting under the influence of a waning heat source. Only the base of
Site 1142 penetrated beneath this terminal alkalic sequence of lava flows into
the general tholeiitic sequence. This lowermost unit at Site 1142 is
compositionally distinct from other Site 1141 and 1142 basalts in that it
exhibits (Nb/Ce)N < 1, similar to the dredge samples from Broken
Ridge (Fig F55), a result that
is consistent with contamination by a continental component (Mahoney et al.,
1995). All other Site 1141 or 1142 lavas do not exhibit evidence of this
contamination (i.e., they have [Nb/Ce]N > 1).
