Site 1204 is ~57.3 km (31 nmi) north of Site 1203 (see "Background and Scientific Objectives") and within ~0.5 km of Site 883, where basement rocks were recovered from Detroit Seamount during Ocean Drilling Program (ODP) Leg 145 (Rea, Basov, Janecek, Palmer-Julson, et al., 1993). At Site 883, basalt was recovered from Holes 883E (penetration = 37.8 m; recovery = 63%) and 883F (penetration = 26.7 m; recovery = 41%), which was comprised of fractured and altered plagioclase-phyric pillow basalt and massive basalt lava flows.
Two holes (1204A and 1204B), ~100 m apart, were drilled at this site. The minimum age of basement in both holes is constrained to Late Cretaceous (Campanian; 71-76 Ma) by nannofossils in the immediately overlying sediment (see "Biostratigraphy"). In Hole 1204A, basement was encountered at 819.5 mbsf (drilling depth [815.6 mbsf curated depth]) and drilled to 880.3 mbsf. Approximately 22.1 m of basement was recovered from Sections 197-1204A-6R-5 to 10R-6 (recovery = 52%). An additional 22.4 m of basement core was penetrated (Cores 197-1204A-11R to 14R), but there was no recovery because of a clogged bit. We divided the basement rocks into two units: clastic sediment (Unit 1) and basaltic lava flows (Unit 2). Subunit 1a is a diamictite, whereas Subunit 1b is a basalt lithic breccia. Unit 2 is composed of an aphyric to moderately olivine-plagioclase-phyric basalt consisting of multiple lobes (Table T5; Fig. F6).
In Hole 1204B, basement was encountered at 816 mbsf (drilling depth [814.0 mbsf curated depth]) and drilled to 954.4 mbsf. Approximately 53.2 m of basement core was recovered from a penetration of 140.5 m (recovery = 37.9%). We divided the basement into four units: Unit 1 is an aphyric basalt, Units 2a-2d consist of the sequence aphyric basalt-diabase-aphyric basalt-hyaloclastite lapilli breccia, Unit 3 is an aphyric basalt, and Unit 4 is dominantly composed of calcareous sandstone and mudstone subdivided into Subunits 4a-4d (Table T5; Fig. F6). Nannofossils in Subunit 4b show the same age range (71-76 Ma) as the oldest nannofossils in the overlying sediment (Fig. F6; see also "Biostratigraphy").
The major lithologic characteristics of Site 1204 rocks are summarized in Figure F6.
A subplanar, apparently conformable contact between Unit 1 and the overlying fine-grained, deepwater Campanian (71-76 Ma) sediment is seen in Section 197-1204A-6R-5 at 815.60 mbsf. Subunit 1a consists of a diamictite containing highly altered, aphyric vesicular basalt clasts up to 1 cm in diameter in a calcareous matrix. The upper part of Subunit 1b is composed of calcareous volcaniclastic sandstone and grades downward to breccia containing fragments of highly altered vesicular basalt and completely devitrified glass in a calcareous sandstone matrix (Fig. F7). Toward the bottom of Unit 1, the degree of abrasion and sorting decreases and the size and angularity of the clasts increase. Basalt clasts are similar in lithology to the underlying lava of Unit 2.
The contact between Subunit 1b and the underlying basalt of Unit 2 was not observed. It is inferred to lie at the bottom of Section 197-1204A-7R-1 (820.95 mbsf). Unit 2 is composed of aphyric to olivine-plagioclase-phyric basalt. The unit was divided into lobes, and the criteria for determining these subunit boundaries included changes in the phenocryst and vesicle content and the presence of glassy or aphanitic quenched lobe margins. We identified eight pahoehoe lobes (Subunits 2a-2h) that range in thickness from a few decimeters to several meters. These lobes probably represent a single lava flow. The thickest lobe (Subunit 2h) at the base of the Hole 1204A core was not fully recovered but it is >8 m thick. It has a smooth pahoehoe lobe surface, a vesicular upper crust (Sections 197-1204A-9R-4 to 10R-2), and a massive, nonvesicular lobe interior (Sections 197-1204A-10R-3 to 10R-6).
Unit 2 consists of aphyric basalt with variable amounts (0-10 modal%) of olivine microphenocrysts. Plagioclase phenocrysts are scarce but make up to 3 modal% of the rock in Section 197-1204A-9R-2. Phenocrysts are set in a fine-grained to aphanitic groundmass consisting of plagioclase, clinopyroxene, opaque oxides, and glass. Olivine microphenocrysts have been completely replaced, and their presence was inferred from pseudomorphs of carbonate and Fe oxyhydroxide. Plagioclase is typically clouded with sericite, and glass is partly to completely altered. A sharp alteration front is present in Core 197-1204A-10R, where the rock color changes from yellowish brown to green gray (see "Alteration and Weathering").
Many of the lavas are highly vesicular (up to 35 modal%) (1- to 4-mm-diameter vesicles), especially close to lobe margins. Pipe vesicles are common near lobe bases, and small vesicle cylinders (2-5 mm wide) are frequently found in the massive lobe interiors. Vesicles are variably filled with secondary minerals; calcite is the most common. Unfilled vesicles generally have an inner coating of clay minerals and Fe oxyhydroxide.
The lithologic components of the Hole 1204B rocks are listed by section in Table T6 and are summarized in Figure F6 and Table T5. The Hole 1204B lava units generally consist of multiple flow lobes that were identified using the criteria of variable vesicularity, presence of glassy (pahoehoe) lobe margins, and/or change in groundmass crystallinity and granularity (Table T6). Because of the <50% recovery (Fig. F6), we did not divide the basalt into subunits based on lobes.
Unit 1 consists of at least nine pahoehoe lobes, ~0.3 to 5.0 m thick, with variable vesicularity. They are composed of aphyric basalt with minor olivine (3%-8%) and scarce plagioclase (1%-5%) microphenocrysts (Table T6). When olivine is present, it is as either a microphenocryst or a groundmass phase. In the upper part of Unit 1 (Sections 197-1204B-1R-3 to 2R-2), the cooling cracks of the lava are filled with fine- to coarse-grained carbonate sandstone, which in places shows well-developed planar laminated bedding (Fig. F8). The lower part of Subunit 2a (i.e., Sections 197-1204B-3R-1 to 4R-1) consists of <1-m-thick lobes intercalated with decimeter-thick intervals of lapilli breccia consisting of 2- to 30-mm angular fragments of vesicular basalt glass and lava lithics embedded in fine- to medium-grained carbonate sand. The basalt lapilli breccia fragments are of the same lithology as the associated lobes. This lithofacies association shows that the lava flow interacted with a preexisting carbonate sand layer, forming the brecciated intervals in the process, most likely by quenched fragmentation. Furthermore, this occurrence suggests a significant time gap in the lava accumulation, and it is the basis of our division between Units 1 and 2.
The basalt is moderately to highly altered with olivine completely replaced by carbonate or Fe oxyhydroxide, but plagioclase and pyroxene in the groundmass are only slightly to moderately altered (see "Alteration and Weathering"). Glassy lobe margins are mostly altered to dark brown palagonite and clay, but some unaltered "islands" remain (e.g., interval 197-1204B-3R-2, 97-100 cm).
Subunit 2a consists of at least 30 variably vesicular, decimeter- to meter-thick lava flow lobes that are composed of aphyric basalt (Table T6). These lobes are moderately to highly altered basalt with petrographic characteristics that are indistinguishable from the basalt of Unit 1. A distinct lobe boundary is present at Section 197-1204B-7R-3, 30 cm. Farther downsection (at 125 cm) a sharp transition is present, although without clear indication of a lobe boundary, from highly vesicular aphanitic to fine-grained basalt to sparsely vesicular, medium-grained basalt with distinctive ophitic to subophitic (diabasic) texture. This transition marks our division between Subunits 2a and 2b (Table T5). Because this diabase appears to be a coarser-grained version of Subunit 2a, it is designated Subunit 2b rather than a new unit. At a depth of ~94 m into the basement, the grain size of the diabase begins to decrease and grades back into aphyric basalt that is virtually indistinguishable from that in Subunit 2a. A smooth-surfaced glassy (pahoehoe) lobe margin at ~918 mbsf (~104 m into basement) marks the base of Subunit 2b (Table T5).
The diabase is characterized by fine- to medium-grained plagioclase laths enclosed in large (
6 mm) clinopyroxenes. As in the lava flows above, olivine microphenocrysts are completely replaced by Fe oxyhydroxide in oxidized regions and bluish-greenish gray clay in reduced regions (see "Alteration and Weathering"). Vesicles in the reduced regions are filled with the same blue-gray clay that replaces the olivine and some of the groundmass, making the original proportions of these constituents difficult to estimate.
From 918 to 935 mbsf a series of vesicular pahoehoe lobes are present that we designate as Subunit 2c (Tables T5, T6). In terms of lithology, this unit is indistinguishable from the overlying lava flows; it contains sparse plagioclase and olivine (completely replaced by Fe oxyhydroxides) microphenocrysts. Underlying Subunit 2c is ~9.7 m of matrix-supported hyaloclastite lapilli breccia, Subunit 2d, consisting of angular basalt lapilli to breccia fragments in a matrix of fine lapilli (Table T5). Large basalt clasts are also present, and some are difficult to distinguish from discrete flows within the breccia sequence. Below the hyaloclastite breccia, a >2.3-m-thick aphyric basalt unit is present; it consists of at least five pahoehoe lobes (Table T6). In terms of its volcanic architecture and petrographic characteristics, this basalt is identical to the overlying flows in Unit 2, but it has a more evolved composition (see "Geochemistry"). Therefore, this sequence of aphyric lobes is designated as Unit 3. The basal contact of Unit 3 is sharp and dips at 40° with glassy lobe margins resting directly on Subunit 4a.
Subunit 4a is a calcareous sandstone that exhibits disturbed bedding with steeply dipping (50°-60°) coarse sand and fine gravel layers that have convoluted and irregular bedding planes. The bedding in the sandstone is concordant with and roughly parallel to the basal contact of Unit 3, indicating that the disturbance is related to the emplacement of the lava flow onto unconsolidated sand. This disturbance is not found in the calcareous mudstone that makes up Subunit 4b, which may indicate that only part of Subunit 4a was recovered (Table T6). Subunit 4c is a clast-supported vitric-lithic lapilli breccia that is exceptionally matrix poor and exhibits inverse size grading. The lower contact with the underlying sandstone (Subunit 4d) is characterized by load casts indicating rapid deposition for the lapilli breccia, which may have been emplaced as a grain flow. Subunit 4d is a cross-stratified calcareous sand containing dispersed 2- to 10-mm lava lithic fragments (see "Site 1204 Core Descriptions").
We interpret the basalt units in Holes 1204A and 1204B to be pahoehoe lava flows because they consist of multiple lobes confined by smooth glassy lobe margins that are highly vesicular (Table T6). Several lobes show the threefold division into vesicular upper crust, massive lobe interior, and vesicular lower crust that is characteristic of pahoehoe lavas formed by an endogenous mode of emplacement (i.e., transport under insulating crust and growth by lava inflation) (Hon et al., 1994; Self et al., 1998). The internal architecture of the >8-m-thick pahoehoe lobe at the base of Hole 1204A and Subunit 2b in Hole 1204B is similar to that of large inflated pahoehoe sheet lobes (Thordarson and Self, 1998). Their presence within a package of smaller pahoehoe lobes is consistent with this interpretation.
We have identified the following lithofacies associations in the Hole 1204B succession: pahoehoe lava (Unit 3) resting on calcareous sediment (Unit 4), thick pahoehoe lava (Subunits 2a-2c) capping a hyaloclastite lapilli breccia, and a relatively thin pahoehoe flow containing intervals of lapilli breccia mixed with carbonate sand toward its base. These facies associations along with the pahoehoe nature of the lavas suggest that these flows originated from subaerial vents and were emplaced in a nearshore environment (see also Fig. F56 in the "Leg 197 Summary" chapter). Finally, the presence of lapilli breccia mixed with carbonate sand on the top of the pahoehoe lava flow in Hole 1204A and the striking resemblance of Hole 1204A lavas to Subunit 2b-2c lavas in Hole 1204B in terms of their morphology, petrography, and chemical composition strongly suggest that these packages may be correlated. If this interpretation is correct, then the thick pahoehoe sheet lobe at the base of Hole 1204A most likely corresponds to the sheet lobe represented by Subunit 2b in Hole 1204B.
Nearby Holes 883E and 883F (Fig. F1) also recovered lava flows with multiple lobes that we infer are analogous to the pahoehoe lava flows in Holes 1204A and 1204B. Specifically, 27 basement units were identified in Hole 883E and 16 in Hole 883F; these units were defined on the basis of glassy margins and carbonate sediment filling voids between adjacent lobe bases of glassy margins and carbonate (Rea, Basov, Janecek, Palmer-Julson, et al., 1993). The basement units defined by Rea, Basov, Janecek, Palmer-Julson, et al. (1993) range in thickness from 0.2 to 2.7 m, and these units are equivalent to our definition of lobes. In their overall architecture and lobe structure the lava flows from Holes 883E and 883F and Holes 1204A and 1204B are identical. Therefore, we conclude that at Sites 883 and 1204, the basement consists of compound pahoehoe lavas erupted in a similar environment.
Lavas from Holes 1204A and 1204B are typically aphyric and range in grain size from basalt to diabase; diabase is present only in Hole 1204B. Plagioclase phenocrysts (0%-2%) can be distinguished by their more equant form, compared to those in the groundmass, and the presence of well-defined oscillatory zoning and fracturing (Fig. F9). Olivine, now pseudomorphed by Fe oxyhydroxide, iddingsite, and calcite (Fig. F10), is euhedral and typically of a similar size to the groundmass (0.8 mm), although its morphology may imply it is a microphenocrystic phase. Unaltered olivine is present in one section (interval 197-1204B-3R-2, 97-100 cm) in a glassy lobe margin (Figs. F11, F12). These olivines also contain melt inclusions (Fig. F13).
The groundmass of the basalt and diabase typically consists of skeletal or elongate plagioclase surrounded by clinopyroxene oikocrysts in a subophitic texture (Fig. F14). Clinopyroxenes range up to 5 mm in size and typically display a dusty pink color and weak pleochroism (Fig. F15), indicative of elevated Ti contents and suggesting an alkaline affinity for the lavas.
Glass and titanomagnetite are also present in the groundmass. Glass forms up to 30% and titanomagnetite up to 8% of the groundmass in lobe interiors. Glass is completely altered to green and brown clay and Fe oxyhydroxide, whereas the titanomagnetite shows partial to complete alteration to maghemite. In segregated areas adjacent to vesicles, clinopyroxene and titanomagnetite are concentrated, each displaying an acicular form (dendritic for titanomagnetite) indicative of a quench texture (Fig. F16). Clinopyroxene in segregated material found in vesicle cylinders (e.g., Sample 197-1204B-15R-1, 15-18 cm) (Fig. F17) and lining vesicles exhibits more pronounced pleochroism than it does in the groundmass.
Vesicles are present (0%-15%) and are round to irregular in form, are <1.5 mm in size, and are typically lined with green clay and filled with calcite, although some remain unfilled. The abundance of calcite is apparent in the chemical analyses. For example, Sample 197-1204A-7R-3, 36-37 cm, has the highest loss on ignition (LOI) (9.5 wt%) and the highest CaO contents (19.4 wt%) of the Site 1204 samples analyzed (Table T7).
No petrographic differences exist between the lavas of Holes 1204A and 1204B, apart from a coarsening in grain size in the Subunit 2b diabase in Hole 1204B. In contrast, most of the lavas at Site 1203 contain more plagioclase and olivine phenocrysts, both present as glomerocrysts.
We note that the opaque mineralogy of the basalt and diabase units from Holes 1204A and 1204B is dominated by oxide and oxyhydroxide minerals (Table T8). The change in color from yellowish orange brown to greenish blue gray noted in Cores 197-1204A-10R and 197-1204B-9R and 13R conform with a change in alteration conditions from oxidizing to reducing (see "Alteration and Weathering"). Sulfide, dominantly secondary pyrite, is present only in these greenish blue-gray, reduced regions (e.g., Fig. F18). Primary sulfide is extremely rare and is present only as blebs of micrometer-sized unidentifiable sulfide included in the primary silicate and oxide minerals.
The opaque mineralogy in the majority of the lavas consists of primary titanomagnetite with secondary maghemite, goethite, and Fe oxyhydroxide. Note that here, as in other Leg 197 site reports, we do not distinguish between maghemite and titanomaghemite (or Ti-bearing maghemite), as they are indistinguishable using reflected-light microscopy. Fe oxyhydroxide is a term used for a group of Fe-rich secondary minerals that have poor crystal form and include goethite when it is amorphous (
-FeOOH), akaganeite (ß-FeOOH), lepidocrocite (
-FeOOH), feroxyhyte (
-FeOOH), and ferrihydrite (5Fe2O3·9H2O) (see Waychunas, 1991). As is evident from Table T8, goethite in the igneous basement rocks from Site 1204 is generally well developed; where it exhibits a crystalline form and can be polished (e.g., Fig. F19), it is noted as a separate, identifiable mineral from Fe oxyhydroxide.
Titanomagnetite (Fe2+1 + xFe3+2 - 2xTi4+xO4) is present throughout the igneous basement at Site 1204. It is converted to cation-deficient titanomagnetite and eventually to maghemite (or titanomaghemite: [(Fe3+0.96[]0.04)(Fe2+0.23Fe3+0.99Ti4+0.42[ ]0.37O4)] after Collyer et al., 1988, where [ ] represents cation vacancies) by oxidation at temperatures below ~250°C (Collyer et al., 1988; Goss, 1988; Banerjee, 1991). The oxidation process requires the addition of oxygen to the crystal boundary, as cations are not being removed; only their valence state is raised (Lindsley, 1976; Waychunas, 1991). Maghemite is metastable with respect to hematite, and the structure may require bonded water or H+ ions for stabilization (Waychunas, 1991). Structurally, maghemite is a defect spinel with incomplete cation site occupancy.
In the basement sequence from both holes at Site 1204, where the alteration environment has been reducing (i.e., in the greenish blue-gray regions of the cores), titanomagnetite has not been visually altered (Fig. F18). However, where alteration has been oxidizing, titanomagnetite shows slight (Fig. F20A, F20B, F20C, F20D, F20E, F20F) to complete (Fig. F21A, F21B) alteration to maghemite (see also Table T7). Generally, maghemite engulfs titanomagnetite from the rim to the center of the crystal (e.g., Fig. F22A, F22B), but in some cases, maghemite growth has exploited cleavage planes, leaving only relict strips of the titanomagnetite (Fig. F23A, F23B). Maghemite exhibits exsolution and breakdown in the more highly oxidized samples. For example, in Sample 197-1204B-10R-3, 25-27 cm, it has exsolved a darker gray-brown phase along cleavage planes (e.g., Fig. F24A, F24B). We tentatively suggest that this texture is caused by the Ti content of the maghemite reaching a concentration where it can not longer be accommodated, and it combines with the remaining Fe2+ to form ulvöspinel or ilmenite. Evidence that maghemite has become unstable is seen in Figure F25, where the rim of maghemite is degraded before complete replacement of the titanomagnetite has been achieved. We interpret this to indicate the continued presence of circulating water under low temperatures (150°C?) altering the maghemite to Fe oxyhydroxide rather than completing the transition to hematite.
Major and trace element abundances were determined by ICP-AES (see "Igneous Petrology" in "Physical Volcanology and Igneous Petrology" in the "Explanatory Notes" chapter) for five samples from Hole 1204A and sixteen samples from Hole 1204B (Table T7). All samples from Site 1204 are alkalic basalt. The composition of lava flows from Hole 1204A and Subunits 2a and 2b in Hole 1204B overlap those from Site 883 (Fig. F26). Samples from Units 1, 2a, 2c, and 3 in Hole 1204B have the highest total alkali contents. Samples from Subunit 2a in Hole 1204B have relatively low SiO2 contents (Fig. F26). Samples of Unit 1 and Subunit 2a lava from Hole 1204B and three samples from Site 883 have the lowest MgO contents of lavas recovered from Detroit Seamount (Table T7; Fig. F27). The Subunit 2a sample with the highest MgO content (Sample 197-1204B-7R-3, 42-44 cm) (Table T7) is the lowermost Subunit 2a sample analyzed. It is similar in composition to Subunit 2b samples (e.g., Fig. F27). It is possible that this sample is from the vesicular flow top of Subunit 2b (Table T6).
The samples are all altered, and LOI ranges from 1.93 to 9.48 wt% (Table T7; see also "Alteration and Weathering"). The three Site 1204 samples (197-1204A-7R-3, 36-37 cm; 197-1204B-6R-4, 21-24 cm; and 197-1204B-7R-3, 140-142 cm) with the highest LOI have anomalously high CaO contents (15.4-19.4 wt%) (Table T7; Fig. F28). The sample with the highest LOI (Sample 197-1204A-7R-3, 36-37 cm) also has a low SiO2 abundance (Table T7; Fig. F26). After normalization to a volatile-free basis, the other major and trace element abundances, including total alkali content, in this sample are comparable with those in the other four samples analyzed from Hole 1204A (Figs. F26, F27, F29). Apparently, this sample contains ~20 modal% carbonate, probably a result of abundant carbonate-filled vesicles. We infer that all samples with >15 wt% CaO (Table T7; Fig. F28) contain a secondary carbonate component, commonly introduced as vesicle fillings.
At a given MgO content, the Site 1204 basalt has TiO2 contents intermediate between basalt from Site 884 and the alkalic lavas from Site 1203 (Fig. F27). In this TiO2 vs. MgO plot, basalt compositions from Sites 883 and 1204 overlap and generally have slightly higher TiO2 at a given MgO than the Site 1203 transitional to tholeiitic basalt.
Because Ti and Zr are incompatible and relatively immobile during alteration, we use a Ti vs. Zr plot to further investigate the petrogenesis of lavas forming Detroit Seamount (Fig. F29). Except for lavas from Units 23 and 26 at Site 1203, all of the Detroit Seamount lavas define a near-linear trend with Ti/Zr = 92 ± 7, very close to the primitive mantle estimate of 116 (Sun and McDonough, 1989). Basalt from Site 884 and the olivine-rich basalt (picrites) from Site 1203 have the lowest Ti and Zr abundances. For Site 884 basalt, a depleted source has been inferred (Keller et al., 1995, 2000; M. Regelous et al., unpubl. data), and for the Site 1203 picrites, the low abundances are the result of olivine accumulation. The highest Ti and Zr abundances are in the alkalic lavas from Site 1203. Since these have relatively high MgO contents (Table T7), their high Ti and Zr abundances are unlikely to be a result of extensive crystal fractionation.
Basalt from Site 883, Hole 1204A, and Subunits 2b and 2c from Hole 1204B have similar Ti contents, but these Site 1204 samples have slightly higher Zr contents than Site 883 lavas (Fig. F29). Since different analytical techniques were used for analyzing Site 883 lavas (X-ray fluorescence and inductively coupled plasma-mass spectroscopy]) (Keller et al., 1995; M. Regelous et al., unpubl. data) and Site 1204 lavas (ICP-AES) (see "Igneous Petrology" in "Physical Volcanology and Igneous Petrology" in the "Explanatory Notes" chapter), this difference in Zr content must be evaluated with shore-based analyses. We note, however, that most of the Site 1204 basalt ranges to higher Y contents than Site 883 basalt (Fig. F30), so it is possible that there are small but significant geochemical differences between lava flows from Sites 883 and 1204.
The low MgO lavas from Unit 1 and Subunit 2a in Hole 1204B have higher Ti and Zr abundances than other Site 1204 lavas, and this result is likely to be a result of extensive crystal fractionation. Some caution is required in interpreting the petrogenesis of Unit 1 and Subunit 2a basalt because two of the four samples from these units are anomalously enriched in P and Ba (Table T7; Fig. F30). These anomalies are likely to be the result of secondary apatite (see "Alteration and Weathering"). Apatite in these samples is not obvious in thin section, but it may occur in the groundmass and as vesicle linings. The sample with the highest P content (Sample 197-1204B-4R-3, 29-31 cm) is also elevated in Y (Fig. F30), an element enriched in apatite, but these samples do not have anomalous Ti and Zr contents (Table T7; Fig. F29).
Detroit Seamount has been sampled by six drill holes (Holes 883E, 883F, 884E, 1203A, 1204A, and 1204B, with basement penetrations of 37.8, 26.7, 87.0, 452.6, 47.3, and 140.5 m, respectively). Site 884 was in relatively deep water (3824 m) on the eastern flank of the seamount, whereas Sites 883, 1203, and 1204 were drilled in shallower water (2370-2593 m) on the plateau of the seamount (Fig. F1). Important results include the following:
