Previous   |    Next

doi:10.2204/iodp.proc.335.203.2016

Results

Lithologic abundance

For a same run, thin sections usually show the same lithologies whatever the grain size or the fishing assembly considered (Fig. F2A). Cuttings from some runs (6, 15, 17, 18, and 20) appear relatively homogeneous in terms of abundance of the main lithologies (Fig. F2B). However, for other runs (12, 14, 19, and 21), the different thin sections display significant variations in the abundance of the main lithologies, especially basaltic lavas and granoblastic basalts.

Runs 2 and 6 recovered cuttings from an unstable section of the hole located in the lava sequence (~922 mbsf). They are thus unsurprisingly composed of more than 90% basaltic lavas (basalt and minor glass) (Fig. F2). Some dolerites (up to 7%) are also observed in Run 6, although the top of the sheeted-dike complex in the borehole is located around 1060 mbsf. This is probably related to the incomplete evacuation of cuttings from the hole during Expeditions 309/312 drilling operations.

The other runs (11–15 and 17–21) recovered cuttings from the bottom of the hole (1518–1520.2 mbsf). All the lithologies defined on board the ship are observed in these basal cuttings (Fig. F2A), confirming that they progressively accumulated in the hole since the beginning of drilling operations and thus represent an integrated sampling of the entire hole. The basal cuttings are dominated by basaltic lavas (up to 58%, Run19BSJBc-T) and granoblastic basalts (up to 73%, Run14EXJB-H) except in Runs 15 and 17 where basaltic lavas are rare (<7%) and granoblastic basalts are the predominant lithologies (68%–88%) (Fig. F2B). Dolerites and gabbroic lithologies usually have similar and relatively low abundance in each run (<20%); only Run 15 differs by the absence of doleritic grains. In most runs, the completely recrystallized granoblastic basalts are more abundant than other granoblastic basalts (Fig. F2A).

Albitites are observed in about one-third of the thin sections and, when present, they are relatively abundant (4%–13%) (Fig. F2A). The relative high proportion of albitite in the cuttings is quite surprising because no occurrence was reported in the hole during Expeditions 309 and 312 (Teagle, Alt, Umino, Miyashita, Banerjee, Wilson, and the Expedition 309/312 Scientists, 2006) and this lithology is not particularly common in the material recovered during Expedition 335 (a small piece in a core, two cobbles, and four pebbles; see the “Site 1256” chapter [Expedition 335 Scientists, 2012c]).

Taking into account all the basal cuttings studied from Runs 11–21, granoblastic basalts are the dominant lithology representing ~52.5% of the cuttings, whereas dolerites are the less abundant lithology representing ~6% of the cuttings. Basaltic lavas (i.e., basalts and glass) and gabbroic lithologies represent ~27.5% and ~10.5% of the cuttings, respectively. Within the lithologic groups, the gabbros and the completely recrystallized granoblastic basalts are the most common lithologies (~4.5% and 22% of the cuttings, respectively) (see Fig. F2). These results are very different from the lithologic abundances that we estimated from the entire hole (i.e., 64% basaltic lavas, 23% dolerites, 8% granoblastic basalts, and 5% gabbroic lithologies) and confirm that the abundance of the lithologies in the cuttings is highly dependent of their recovery rates during drilling operations (e.g., mean recovery of 41% for basaltic lavas and 9.5% for granoblastic basalts; see Fig. F18 and Table T2 in the “Expedition 335 summary” chapter [Expedition 335 Scientists, 2012a]).

These results clearly show that cuttings sample all the lithologies of the hole but overrepresent non- or poorly cored materials (i.e., albitites, granoblastic basalts).

Mineral composition

Pyroxenes

Our database for pyroxenes consists of 1537 analyses (1053 for cpx and 484 for opx). Clinopyroxenes in basaltic lavas are mainly Mg-augites with a mean composition of En₄₉Wo₃₅Fs₁₅ and show scattered composition toward subcalcic augites and pigeonites (Wo < 30%) and more ferrous augites (Fs > 20%) (Fig. F3A). Microphenocrysts are mainly concentrated in the Mg-augite field; the more ferrous and less calcic pyroxenes likely reflect groundmass crystals and/or phenocryst rims (Perfit and Fornari, 1983; Dziony et al., 2008; Umino, 2007). Clinopyroxenes in dolerites are predominantly augites with variable magnesium content (26% < En < 56%) probably reflecting core/rim variations (Sano et al., 2011) (Fig. F3A). Some diopsides (Wo > 45%) and rare pigeonites (Wo_11–13) are also observed. Clinopyroxenes in gabbroic lithologies and granoblastic basalts have more focused composition toward diopsides and Mg- and Ca-rich augites (Wo > 35%) (Fig. F3B, F3C). Rare pigeonites (Wo_5–11) and Ca-poor augites are observed. Orthopyroxenes in these lithologies are iron-rich enstatites with En contents ranging from 56% to 68%. Clinopyroxenes in albitites have homogeneous Mg- and Ca-rich compositions (En_42–44Wo_42–47Fs_9–15) similar to cpx from gabbroic lithologies and granoblastic basalts (Fig. F3B).

Clinopyroxenes in basaltic lavas and dolerites show a wide range of composition with Mg# ranging from 44% to 86% (Fig. F4). However, cpx microphenocrysts in basalts usually show primitive compositions (Mg# > 69%). On the contrary, cpx in gabbroic lithologies and granoblastic basalts show a limited range of composition with Mg# varying from 62% to 85% (mean Mg# is ~ ">72%). In these lithologies, opx also describe a limited range of composition (58% < Mg# < 71%) (see RESULTS in “Supplementary material”). Clinopyroxenes in albitites have homogeneous primitive compositions with Mg# ranging from 73.5% to 83.0%.

From cpx compositions, two lithologic groups can be easily defined: a basaltic lavas-dolerites group and a gabbroic lithologies-granoblastic basalts group (Figs. F3, F4). For a given differentiation degree, cpx in gabbroic lithologies and granoblastic basalts are more depleted in Al₂O₃, Cr₂O₃, and TiO₂ than cpx in basaltic lavas and dolerites (Fig. F4A–F4C). However, cpx in gabbroic lithologies are on average slightly more depleted in Al₂O₃ and TiO₂ than cpx in granoblastic basalts (0.8% Al₂O₃ and 0.2% TiO₂ in gabbroic lithologies and 1.1% Al₂O₃ and 0.4% TiO₂ in granoblastic basalts) (Fig. F4B, F4C). Clinopyroxenes in albitites have compositions similar to cpx in the gabbroic lithologies-granoblastic basalts group, especially the gabbroic lithologies (Fig. F4B, F4C). However, they are on average richer in sodium (mean Na₂O is ~0.40% for albitites and ~0.23% for gabbroic lithologies-granoblastic basalts) (Fig. F4D).

No significant chemical variation is observed in clinopyroxenes between the different kinds of granoblastic basalt and between the different gabbroic lithologies (Figs. F3C, F4; see RESULTS in “Supplementary material”). However, cpx in one gabbro grain (Run14EXJB-T-G15) clearly differ from cpx in other gabbroic grains by their high Mg# (80.5%–85.5%) and enstatite content (En_47–49) and their strong enrichment in Cr₂O₃ (0.6%–0.9%) and Al₂O₃ (2.4%–3.0%) for a given Mg# (Figs. F3, F4A, F4C).

Feldspar

Our database for feldspar consists of 1552 analyses. All feldspars are plagioclases (Or < 10%). In most lithologic groups, plagioclases show a wide range of composition from albite (An < 10%) to bytownite (An > 70%) (Fig. F5A). These strong chemical variations are related to chemical zoning and to the coexistence of primary and secondary (i.e., resulting from alteration or recrystallization) plagioclases in all lithologies (Alt et al., 2010; Sano et al., 2011; Koepke et al., 2008).

Plagioclases in basaltic lavas show the widest range of composition from An_0.3 to An_85.4, but the most common composition is An_60–70 (Fig. F5A). Microcrysts in glass grains and microphenocrysts usually have anorthite contents >60% and show the most calcic compositions (An > 80%) (Fig. F5B). The composition of plagioclases in dolerites mainly ranges from An₄₂ to An₈₃ (Fig. F5A; see RESULTS in “Supplementary material”). The most common composition is An_60–70, and sodium-rich plagioclases (An < 30%) are rare. Most plagioclases in gabbroic lithologies have anorthite contents in the range of 50%–60% (Fig. F5A). Sodium-rich compositions (An < 30%) are essentially observed in plagioclases from gabbros (especially quartz-gabbros) and diorites, and ol-bearing lithologies usually have calcium-rich plagioclases (An_63–76) (Fig. F5C; see RESULTS in “Supplementary material”). Plagioclases in granoblastic basalts show relatively homogeneous compositions with 70% of the analyses falling in the range An_50–60 (Fig. F5A). Plagioclases from the different kinds of granoblastic basalt display similar compositions (Fig. F5D). As previously, two lithologic groups can be identified from plagioclase compositions: a basaltic lavas-dolerites group (An_60–70) and a gabbroic lithologies-granoblastic basalts group (An_50–60). Albitites clearly differ from other lithologies by the sodium-rich composition of their plagioclases; >90% of the analyses fall in the range An_0–20 (Fig. F5A).

Plagioclases in basaltic lavas (microcrysts and microphenocrysts) and dolerites show a wide range of MgO contents from 0% to 0.8% with a mean value around 0.2% (Fig. F6). On the contrary, plagioclases from gabbroic lithologies and granoblastic basalts mainly have very low MgO contents below or close to the detection limit of the microprobe (~0.05%). Only some analyses show higher MgO values (up to 0.6%). Plagioclases in albitites are devoid of magnesium. This characteristic associated with their sodium-rich compositions shows that plagioclases in albitites are secondary phases related to alteration processes (Alt et al., 2010).

Fe-Ti oxides

Our database for Fe-Ti oxides consists of 1011 analyses. Magnetite and ilmenite solid solutions (595 and 416 analyses, respectively) are observed in all lithologic groups and coexist in about 50% of the grains (Fig. F7A). Magnetite alone is usually found in grains of basalt, dolerite, and granoblastic basalt, whereas ilmenite alone is usually found in gabbroic and albitite grains (see RESULTS in “Supplementary material”).

The titanomagnetites of basaltic lavas clearly differ from the magnetites of gabbroic lithologies in having high TiO₂ content (16.6%–22.6% vs. 0.7%–7.2%) and low Cr₂O₃ content (0.06% vs. 0.44% on average) (Fig. F7). Magnetites in granoblastic basalts have compositions similar to those in gabbroic lithologies but are sometimes more titaniferous (up to 16% TiO₂). Dolerites show either titanomagnetites with compositions similar to those from basaltic lavas (15%–24% TiO₂, Cr₂O₃ < 0.2%) or magnetites with compositions similar to those from gabbroic lithologies and granoblastic basalts (1.5%–4.5% TiO₂, 0.02%–0.92% Cr₂O₃). Albitites usually display low-Ti and high-Cr magnetites (0.3%–5.0% TiO₂ and 0.5%–1.2% Cr₂O₃), which also points to a hydrothermal origin and is in accordance with plagioclase chemistry (e.g., Abzalov, 1998). Again, a clear distinction exists between the titanomagnetites of basaltic lavas-dolerites and the magnetites of gabbroic lithologies-granoblastic basalts.

Contrary to magnetites, ilmenites do not show significant chemical variations between the different lithologies (Fig. F7A). They have quite homogeneous compositions with TiO₂ and MgO contents ranging from 41.5% to 53.1% and 0% to 0.6%, respectively, and plot on the ilmenite-rich side of the solid solution. All ilmenites are relatively manganiferous (MnO ranges from 0.5% to 5.8%, see RESULTS in “Supplementary material”).

No significant chemical variation is observed in Fe-Ti oxides between the different kinds of granoblastic basalt and between the different gabbroic lithologies (Fig. F7; see RESULTS in “Supplementary material”).

Olivine

Only seven analyses of olivine are included in our database because of the rarity of ol-bearing grains. Olivines are exclusively observed in gabbroic grains. They have relatively evolved compositions with forsterite (Fo) contents ranging from 63.7% to 67.5% and have similar compositions in ol-gabbros and ol-gabbronorites (see RESULTS in “Supplementary material”)

Amphibole

Our database for amphiboles consists of 103 analyses. They were analyzed in all lithologic groups except albitites (Fig. F8). All amphiboles are calcic with compositions ranging from Mg-hornblende to actinolite and minor ferro-actinolite (Fig. F8A). Ferrohornblendes and tremolites are also observed in a few gabbros. Only two edenites were analyzed in basalt and gabbro grains (Fig. F8B). These amphiboles are mainly secondary phases resulting from the alteration of clinopyroxenes. However, the geothermometry study shows that some of them are most likely magmatic phases (see “Temperature and redox conditions”).

Glass

Our database for basaltic glasses is composed of 153 analyses (76 for glass grains and 77 for groundmass glass). Fresh glasses have XMg (Mg/[Mg + Fe^T]) ranging from 0.51 to 0.60 and a mean composition of 52% SiO₂, 1.1% TiO₂%, 13.9% Al₂O₃, 10.9% FeO, 0.2% MnO, 7.5% MgO, 11.4% CaO, 2.2% Na₂O, and 0.09% K₂O (Fig. F9; see RESULTS in “Supplementary material”). Only one grain of fresh glass (Run18BSJB-G20, 5 analyses) differs from other grains, with a less siliceous and more titaniferous and alkaline composition (49.6% SiO₂, 1.9% TiO₂, 2.7% Na₂O, and 0.2% K₂O) (Fig. F9C, F9D). These compositions are within the range observed in mid-ocean-ridge basalt (MORB) glasses from the Pacific Ocean (Bézos and Humler, 2005). Devitrified glasses have compositions relatively close to fresh glasses but show wider chemical variations with, for example, XMg and CaO content ranging from 0.45 to 0.75 and from 9.9% to 13.1%, respectively (Fig. F9A). Glass from basalt and dolerite groundmasses show a wide range of differentiation with XMg and SiO₂ content ranging from 0.55 to 0.75 and from 49.9% to 54.2%, respectively (Fig. F9C, F9D). They have on average higher XMg (~0.66) and higher CaO and Al₂O₃ contents (~12.2% and ~15.7%, respectively) than fresh and devitrified glasses (Fig. F9A, F9B).

Temperature and redox conditions

Temperature conditions

The single-clinopyroxene (cpx) thermometer and the 2-pyroxene (px) thermometer of Brey and Kohler (1990) give similar equilibrium temperatures for all lithologies (difference of 20°C–30°C on average), whereas the QUILF 2-px thermometer of Andersen et al. (1993) usually gives estimated temperatures ~100°C higher than the other pyroxene thermometers (Table T2). The pyroxene data for the different kinds of granoblastic basalt give similar ranges of temperatures (833°C–861°C for the single-cpx thermometer and 966°C ± 52°C to 985°C ± 52°C for the QUILF 2-px thermometer), and the poorly recrystallized samples (Degrees 1 and 2) give the lowest temperatures. The single-cpx and 2-px temperatures estimated for gabbros and gabbronorites are closer but slightly lower than those calculated for the granoblastic basalts (~810°C for the single-cpx thermometer and 958°C ± 55°C for the QUILF 2-px thermometer). Diorites and albitites display the lowest temperatures calculated with the single-cpx thermometer (780°C–799°C), whereas dolerites show the highest temperatures (1017°C). The Run14EXJB-T-G15 gabbro for which the cpx composition clearly differs from that of the other gabbroic lithologies (see “Mineral composition”) displays a single-cpx temperature higher than that estimated for other gabbros (996°C). Moreover, the only grain of ol-gabbronorite studied shows single-cpx and 2-px temperatures higher than those estimated for gabbronorites (895°C and 1040°C ± 131°C for single-cpx and QUILF 2-px thermometers, respectively).

Temperatures obtained for coexisting hornblende and plagioclase in dolerites and granoblastic basalts are usually significantly lower than those estimated from pyroxenes (from 707°C in one doleritic grain to 776°C in Degree 5 granoblastic basalts). Only one amphibole-plagioclase pair in a Degree 3 granoblastic basalt gives an equilibrium temperature of 953°C (see RESULTS in “Supplementary material”) leading to an average temperature for this lithology closer to that estimated from pyroxenes (Table T2). Amphibole-plagioclase pairs in gabbros and diorites have average temperatures closer to or higher than those estimated from pyroxenes (817°C–887°C). Among this, one edenite-plagioclase pair in a dioritic grain has an equilibrium temperature of 1005°C (see RESULTS in “Supplementary material”) that is consistent with a magmatic origin.

The 2-oxide oxythermobarometers of Sauerzapf et al. (2008) and Andersen et al. (1993; QUILF program) provide similar equilibrium temperatures for all lithologies (difference of 50°C–60°C on average) except for doleritic and basaltic grains (Table T2). For all lithologies, these geothermometers give the lowest temperatures. Coexisting ilmenite and magnetite in granoblastic basalts, albitites, and gabbroic lithologies have similar equilibrium temperatures, ranging from 611°C to 644°C for the Sauerzapf et al. (2008) 2-oxide thermometer and from 677°C ± 12°C to 697°C ± 119°C for the QUILF 2-oxide thermometer (Table T2). For one basaltic grain, the magnetite-ilmenite pairs display equilibrium temperatures of 798°C and 689°C ± 103°C for the Sauerzapf et al. (2008) and Andersen et al. (1993; QUILF program) 2-oxide thermometers, respectively. These low temperatures most likely reflect oxide reequilibration during subsolidus cooling or hydrothermalism. In dolerites, the average temperature estimated from the Sauerzapf et al. (2008) 2-oxide geothermometer is 902°C, close to magmatic values, but temperatures estimated from the QUILF 2-oxide thermometer is much lower (716°C ± 97°C).

These results are in agreement with previous thermometry studies conducted on cores of granoblastic basalt, gabbro, and lava from Hole 1256D (France et al., 2009; Alt et al., 2010; Koepke et al., 2011; Dziony et al., 2008).

Redox conditions

The Andersen et al. (1993; QUILF program) 2-oxide oxybarometer usually shows more reducing values than those calculated with the Sauerzapf et al. (2008) oxybarometer (0.8–1.2 log units more reduced considering that ΔNNO ≈ ΔFMQ – 0.7; Jugo et al., 2005) except for dolerites (Table T2). Redox estimates indicate highly oxidizing conditions for granoblastic basalts, gabbroic lithologies, and albitites with ΔNNO values ranging from +1.97 to +2.66 (according to Sauerzapf et al., 2008) and ΔFMQ values ranging from +1.49 to +2.57 (according to Andersen et al., 1993). The most oxidizing conditions are recorded in albitites, and among granoblastic basalts, the most oxidizing conditions are found in the poorly recrystallized rocks (Degrees 1 and 2). These oxidizing values significantly exceed magmatic values. Coexisting Fe-Ti oxides in dolerites and in one grain of basalt give more reducing ΔNNO values of –0.80 and –1.26 (ΔFMQ of 0.91 and –1.44), respectively. These values are in agreement with those obtained for coexisting magnetite and ilmenite in one sample of the Hole 1256D lava pond (Dziony et al., 2008; Koepke et al., 2008) and are within the range observed in MORB glasses (Christie et al., 1986; Bézos and Humler, 2005; Cottrell and Kelley, 2011).

The clinopyroxene-plagioclase oxybarometer of France et al. (2010a) shows lower ΔNNO and ΔFMQ values compared to those derived from the 2-oxide oxybarometers except for dolerites (Table T2). The different temperatures used in the calculation have little influence on the estimated ΔNNO and ΔFMQ values. Redox estimates for granoblastic basalts and gabbroic lithologies indicate similar and relatively oxidizing conditions with ΔNNO values ranging from –0.08 to +0.58 and ΔFMQ values ranging from +0.66 to +1.35, the most oxidizing conditions being observed in Degree 6 granoblastic basalts. These values are still higher than magmatic values. Coexisting clinopyroxene and plagioclase in dolerites have crystallized under highly oxidizing conditions (ΔNNO of +1.22 and ΔFMQ of +1.95) that clearly differ from the oxidation state estimated from the 2-oxide oxybarometers. This could be partly related to the crystallization of plagioclase and clinopyroxene under nonequilibrium conditions in some grains (see RESULTS in “Supplementary material”). The Run14EXJB-T-G15 gabbro, which differs from other gabbroic lithologies in terms of cpx composition and single-cpx equilibrium temperatures, is also characterized by the crystallization of coexisting plagioclase and clinopyroxene under higher oxidizing conditions (ΔNNO of +2.28 and ΔFMQ of +3.02) (Table T2). It is thus clear that this gabbro has suffered from peculiar processes compared to other gabbroic lithologies. The clinopyroxene-plagioclase pairs in the ol-gabbronorite also recorded high oxidizing conditions compared to the mean oxidation state of gabbronorites.

Top of page   |    Previous   |    Next