|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
PETROGENESIS OF ANDESITES AND DACITES
FROM THE SOUTHERN JUAN DE FUCA RIDGE
LAURIE ANN COTSONIKA
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
Laurie A. Cotsonika
To my family.
I would like to thank my advisor, Dr. Michael Perfit, for his patience and guidance
throughout this proj ect, and Dr. Matthew Smith for the support and help he' s given me in order
to complete this. I would also like to thank the Monterey Bay Research Aquarium Institute for
providing use of their equipment and facilities and the crew and officers of the Western Flyer, as
well as the pilots of the ROV Tiburon, for the hard work they put in during the cruise to the Juan
de Fuca Ridge. I would like to thank Dr. Debra Stakes and Dr. Jim Gill for lending their support
and expertise on the cruise and with interpreting my data. Dr. W. Ian Ridley deserves my thanks
for all the help and analytical support provided in Denver and Dr. Paul Wallace for providing the
volatile data. I would also like to thank George Kamenov for the help and analytical support
given while processing my trace element data.
I would like to thank my family, especially my parents, Art and Linda, and my brother and
sister, Nick and Elizabeth, who provided much needed emotional support and understanding. I
would also like to thank all my friends in the Geological Sciences department at University of
TABLE OF CONTENTS
ACKNOWLEDGMENTS .............. ...............4.....
LIST OF TABLES .........._.... ...............7....__........
LIST OF FIGURES .............. ...............8.....
AB S TRAC T ............._. .......... ..............._ 10...
1 INTRODUCTION ................. ...............12.......... ......
Regional G eology ................. .... ... .... ........... ...............1
Previous Studies of Highly Fractionated Suites Recovered at MOR ................ ..................14
2 STUDY AREA AND SAMPLE RECOVERY ................. ...............20...............
T73 5 Dive Ob servations ................. ...............21........... ...
3 ANALYTICAL METHODS .............. ...............27....
Maj or Elements ................. ...............27........... ....
Trace Elements .............. ...............28....
4 PETROGRAPHY AND MINERAL CHEMISTRY .............. ...............32....
Basalts ................. ........... ...............32.......
Andesites and Dacites ................. ...............32........... ....
5 MAJOR AND TRACE ELEMENTS .............. ...............48....
Basalts ............... .. ....._ ...............48...
Maj or and Minor Elements ........._...... ...............48..__... ....
Trace Elements .............. ...............50....
Andesites and Dacites ........._...... ...............52...__........
Maj or and Minor Elements ........._._.. ..... .___ ...............53....
Trace Elements .............. ...............53....
6 DI SCUS SSION ........._._.. ..... .___ ...............62.....
Comparison to Other Evolved Suites .............. ...............62....
Dive T73 5 S ampl es............... ...............64
Fractional Cry stallization............... ............6
M agm a M ixing ................ .. .. ... .... ............ ...............7
Partial Melting/Assimilation Fractional Crystallization (AFC) .............. ....................7
7 CONCLUSIONS .............. ...............84....
A T73 5 DIVE LOGS ............ ..... ._ ...............86..
B PHASE CHEMISTRY FOR T735 LAVAS ...._.. ...._._._._ ......._.__. ...........0
C MAJOR AND TRACE ELEMENT DATA FOR T73 5 LAVAS ........._._ ..... ....._..........159
D PETROLOG RESULT S .............. ...............165....
LIST OF REFERENCES ............ ..... ..__ ...............178..
BIOGRAPHICAL SKETCH ............ _...... ._ ...............188...
LIST OF TABLES
1: Accepted values for maj or element standards. ................ ......... ...................3
2: ICP Trace element standards. 2005-2006 analyses of ENDV (ran as sample) together
with other MORB samples (drift 1 first sample after the standards)............... ...............3
B-1: Pyroxene compositions determined from microprobe analyses. ............. ....................10
B-2: Plagioclase compositions determined from microprobe analyses. .................. ...............132
B-3: Olivine compositions determined from microprobe analyses. ............. ......................154
C-1: Maj or and trace element data for Dive T73 5 samples ................. ............... 160..........
D-1: Results from Petrolog. Results assume QFM at 200 bars of pressure and sample T73 5-
G35 as the parent composition............... ..............16
LIST OF FIGURES
1 Map of the Juan de Fuca Ridge ................. ...............17..............
2 Overview of MBARI dives and rock cores on the southern Cleft segment since 2000. ...18
3 Overview of the dive track taken during Tiburon Dive 73 5 ................ .. ......_. ........24
4 Profile map of dacite dome morphologies and evolved sample locations............._._.. ......26
5 Examples of basaltic lava morphologies seen during dive T73 5........._.. ..........._.......3 6
6 Plain polarized and cross polar views of glomeroporphyritic texture in T73 5-G23..........36
7 Dacite dome lava morphologes............... ..............3
8 Plagioclase oikocryst surrounding several randomly oriented clinopyroxene
chadacrysts in sample T73 5-G10 ........._._.._......_.. ...............37..
9 Element map of poikilitic texture seen in T73 5-G10, in this case, a plagioclase
oikocryst surrounding low-Ca pyroxenes .............. ...............38....
10 Point compositions of pyroxene phenocrysts in T73 5-G9 and T73 5-Gl2 are plotted
to show zoning patterns. ............. ...............39.....
11 Oscillatory zoned clinopyroxene in sample T73 5-G9. ....._.._.._ ......_.._ ........._.....40
12 Skeletal clinopyroxene grain with a lacey reaction rim surrounding the crystal, this is
also indicative of rapid crystal growth. ....__ ......_____ .......__ ...........4
13 Populations of zoned pyroxenes ................ ......................... .................42
14 Olivine and plagioclase glomerophyric cluster ................. ...............43...............
15 Element map of skeletal fayalite crystal in sample T73 5-Gl2 ........._.._.. ....._.._.........43
16 Rare (<1% of the sample) euhedral zircon phenocrysts seen in samples T73 5-Gl2
and T73 5-Gl9 were discovered through microprobe analyses ................. ................ ...44
17 Plain polarized light (left) and crossed polarized light (right) views of a basaltic
xenolith in sample T73 5-Gl l....._.. ............... ...............45. ...
18 Element maps of a basaltic xenolith in sample T73 5-Gl2 .........._.._.. .......__. ..........45
19 A coarser grained xenolith, found in T73 5-G10, is composed of plagioclase, An34 -
An35, and fayalitic olivine, Fols............... ...............46..
20 Myrmekitic intergrowth of quartz and plagioclase .................... ............... 4
21 AFM Diagram comparing Smith, 1994 and Stakes, 2006 samples to dive T735
22 Maj or element plots comparing previous studies of the Cleft segment (Smith et al.,
1994; Stakes et al., 2006) to dive T735 samples............... ...............56
23 Trace element plots comparing previous studies of the Cleft segment (Smith et al.,
1994; Perfit, unpublished) to dive T735 samples. ............. ...............58.....
24 The primitive mantle normalized (McDonough and Sun, 1995) spider diagram
displays the depletion of several key trace elements in the evolved glasses compared
to the basalt compositions. .............. ...............60....
25 The primitive mantle normalized (McDonough and Sun, 1995) REE plot shows the
two distinct groupings within the sample group. .............. ...............61....
26 Comparing maj or element variations in the T73 5 lavas to other evolved suites ..............75
27 Maj or element liquid lines of descent. .............. ...............77.....__.__ ..
28 Cumulative percentage of phases plotted against melt temperature (oC). .................. .......79
29 Trace element liquid lines of descent ................. ...............81........... ..
30 Mixing models calculated using a standard mass balance equation (Langmuir et al.,
1978). ............. ...............8 2....
31 Comparison of T73 5 C1V and Cl/K ratios versus MgO. ........._._......___ ........._....83
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science
PETROGENESIS OF ANDESITES AND DACITES
FROM THE SOUTHERN JUAN DE FUCA RIDGE
Laurie A. Cotsonika
Chair: Michael R. Perfit
The Cleft segment and the ridge-transform intersection (RTI) of the Southern Juan de Fuca
Ridge have been investigated during three cruises of Monterey Bay Aquarium Research
Institutes' (MBARI) Research Vessel (R/V) Western Flyer beginning in 2000. A total of 53 rock
cores and 276 precisely located rock or glass samples were collected during sixteen dives with
the remotely operated vehicle (ROV) Tiburon. These ROV dive samples and observations allow
us to test models regarding the magmatic evolution of this segment and the relationships between
specific tectonic and morphologic features and magmatic processes. An extremely wide range of
N-type mid-ocean ridge basalt (MORB) lavas were recovered which are, on average, more
evolved (lower MgO) off-axis, away from the present neovolcanic zone, and towards the RTI.
During dive T73 5 we investigated a region of unfaulted, curved volcanic ridges that overshoot
the Blanco Transform and 39 samples of lava ranging from ferrobasalt to andesite and dacite
(SiO2 = 50.1 to 66.9 wt.%; Mg# = 49.9 to 9.7) were recovered. The highly evolved lavas were
recovered from two large constructional domes comprised of unusually large pillow flows, and
extremely blocky, vesicular flows similar to some terrestrial silicic domes. Some of the
andesite-dacite hand samples are extremely vesicular with elongate vesicles (1-10 cm) in a
glassy matrix. Mineral assemblages are dominated by microphenocrysts of ferroaugite and
ferropigeonite, with lesser amounts of sodic plagioclase and FeTi oxides. Rare zircon, fayalite
and myrmekitic intergrowths of plagioclase and quartz are present. A few of the more magnesian
phenocrysts (xenocrysts?) exhibit fine normal zoning whereas more Fe-rich crystals exhibit fine
reverse zoning. Additionally, inclusions of quenched basaltic material appear within some of the
evolved lavas. These samples represent an extensive and unique set of some of the most highly
fractionated ocean floor rocks that have ever been recovered; particularly from such a well-
documented setting. Fractional crystallization models which predict over 80% crystallization do
not adequately explain the maj or element chemistry of the silicic lavas and most incompatible
trace elements exhibit significant enrichments relative to predicted concentrations. The highly
evolved nature of the dacites, crystal zoning patterns and the presence of basaltic inclusions
suggest the lavas are the result of mixing between two crystal bearing end-members, i.e. a typical
basalt and a rhyolite likely generated by extreme amounts of fractional crystallization. The
andesitic and dacitic lavas also have elevated C1f levels that range from 4000-6000 ppm. While
the Cl-/K levels are enriched they do not indicate that significant amounts of crustal assimilation
Over 60% of the Earth' s magma flux, > 21 km3/yr, ~3 km3/yr extrusive and ~18 km3/yr
intrusive, takes place at mid-ocean ridges (MOR) (Chadwick and Embley, 1994; Perfit, 2001),
however less than 1% of the total seafloor total area has been sampled or studied in much detail
(Perfit and Chadwick, 1998) and that 1% has been concentrated upon ridge crests. Only a few
studies have focused on the broader area surrounding the immediate region of the ridge axis, and
therefore most investigations have been focused within a narrow neovolcanic zone centered at
the ridge axis. Studies centered on the ridge axis have generally not directly addressed the
processes that occur away from the ridge axis and at ridge-transform intersections (RTI). It has
recently been hypothesized that the areas 5-20 km off of the ridge axis possibly make a greater
contribution to the overall magma flux than previously thought (Goldstein et al., 1994; Graham
et al., 1996; Hekinian et a., 1999; Johnson et al., 2000; Klingelhofer et al., 2001; Zou et al.,
2002). Though published data regarding lavas erupted off-axis and at RTI's is sparse, it has been
suggested that greater amounts of fractional crystallization and assimilation can occur at these
locations. Limited magma recharge can lead to less homogenization of magmas at these
locations compared to the main axial magma body, allowing for more diverse and evolved lava
compositions to be erupted onto the oceanic floor (Christie and Sinton, 1981; Fomari et al.,
1983; Perfit and Fomari, 1983; Perfit and Fomari, 1983; Juster et al., 1989; Perfit et al.., 1994;
Perfit et al.., 1999; Stoffers et al., 2002).
In this thesis, field observations together with petrologic, major and trace element data are
utilized to determine the geochemical characteristics and petrogenesis of highly evolved lavas
recovered from the region of the intersection of the Juan de Fuca Ridge (JdFR) with Blanco
Fracture Zone (BFZ) (Figure 1). Lava samples recovered in 2004 during dive T735 of the ROV
Tiburon operated by the Monterey Bay Aquarium Research Institute (MBARI) represent an
extensive and unique set of some of the most highly fractionated oceanic rocks that have been
recovered from such a well-documented setting. This, in conjunction with the extensive
sampling of the Cleft segment of the southern JdFR and the associated geochemical database,
make it an ideal sample set to explore the magmatic processes involved in the petrogenesis of
samples from the RTI.
The JdFR in the northeast Pacific is an intermediate spreading rate ridge (56 mm/yr full
rate) that has been extensively studied beginning in the 1960s (Raff et al., 1961; Johnson and
Holmes, 1989; Embley et al., 1991; Smith, 1993; Perfit et al., 1994; Smith et al., 1994; Perfit et
al., 1998; Perfit, 2001; Karson et al., 2002; Tiemey, 2003; Stakes et al., 2006). It is located
approximately 440 km (~23 8 nm) off of the coast of Washington and Oregon and spans almost
500 km between the Blanco and Sovanco Fracture Zones.
The JdFR has been divided into seven second-order ridge segments that have distinct
morphological characteristics (Embley et al., 1991; Embley et al., 2000; Smith et al., 1994;
Chadwick et al., 2005). The Cleft segment (Figure 2) is the southernmost segment with its
northern terminus at ~45003'N and its southern terminus at 44027'N where it intersects the
Blanco Fracture Zone (BFZ) (Embley et al., 1991; Smith et al., 1994). The Cleft segment has
recently been volcanically active and has proved to be an important area for detailed
investigations of submarine volcanology, hydrothermal activity, eruption rates, and MOR
petrogenetic processes. Chadwick and Embley (1994) summarized studies of several mid-ocean
ridge basalt (MORB) lava flows believed to have erupted in 1983 and 1987 along the axis of the
Cleft segment and estimated, following Crisp, (1984) that the average extrusive output along the
whole Cleft segment is 0.003 km3/yr. Extrusive activity was determined to be bimodal with
sheet flows preceding a voluminous eruption of pillow flows (Embley and Chadwick, 1994).
The overall extrusive layer of the Cleft segment has an average thickness of 3 50 m and varies
from 200 to 550 m in thickness (McDonald et al., 1994) while depth to the axial magma chamber
(AMC) varies from 1.9 km under the southern Cleft hydrothermal vent systems to 2.23 km under
the northern hydrothermal fields (Canales, 2006). There is a strong correlation in the Cleft
segment between degree of fractionation and latitude, as lavas are generally more mafic to the
north and more evolved toward the southern terminus (Christie and Sinton, 1981; Smith et al.,
1994, Stakes et al.. 2006).
Since 2000, 53 rock cores and 276 rock and glass samples have been recovered from the
Cleft segment using the Remotely Operated Vehicle (ROV) Tiburon (Stakes et al., 2006; Figure
2). Overall, a wide range of normal incompatible element depleted MORB (N-MORB) lavas
were recovered from the entire length of the Cleft segment. The most highly evolved basalts and
a few high-silica lavas were recovered off-axis, away from the neovolcanic zone, and near the
RTI. In 2000, a sample of MOR, low-potassium dacite was recovered by rock core from a small
topographic dome in an area of the Cleft segment characterized by what appear to be curved
ridges and volcanic cones that overlap the westernmost part of the Blanco Transform and appear
to extend on to the Pacific plate. The extremely rare occurrence (or at least recovery) of dacite in
a MOR setting was the impetus for a ROV dive in 2004 that explored these features at the RTI,
in a common yet poorly studied MOR setting.
Previous Studies of Highly Fractionated Suites Recovered at MOR
Highly fractionated suites of atypically high silica MOR lavas have been recovered in few
other locations around the globe; these include the Galapagos Spreading Center (GSC), Iceland,
the Pacific-Antarctic Ridge and a few places on the East Pacific Rise. Hypotheses put forth to
explain the formation of these petrologically evolved suites on the oceanic floor include the
"cold edge effect" (Christie and Sinton, 1981; Fornari et al., 1983; Perfit et al., 1983; Johnson
and Holmes, 1989; Juster et al., 1989; Embley et al., 1991; Smith, 1993; Smith et al., 1994;
Juteau et al., 1995; Tiemey, 2003; Herzburg, 2004), bimodal volcanism, similar to that occurring
in intrusive suites in the western United States (Reid et al., 1982; Barbarin, 1990; Ratajeski et al.,
2001), and, more recently, crustal assimilation (Kerr et al., 1996; Bohrson and Reid, 1998;
Garcia et al., 1998; Gee et al., 1998; Hoernle, 1998; O'Hara, 1998; Weis et al., 1998; Perfit et
The eastern GSC at ~850W has been extensively studied since the late seventies (Christie
and Sinton, 1981; Anderson et al., 1982; Fornari et al., 1983; Perfit and Fornari, 1983; Perfit and
Fomari, 1983; Juster et al., 1989). The discovery of chemically fractionated (high-silica) lavas
along the GSC first led to the development of the "cold edge effect" hypothesis (Christie and
Sinton, 1981; Fornari et al., 1983; Perfit et al., 1983; Juster et al., 1989). This type of "cold edge
effect" occurs along a propagating ridge axis as the advancing ridge magmatic system intersects
older, relatively cool oceanic lithosphere located on the opposite side of a fracture zone. This
intrusion presumably leads to greater extents of cooling in magma bodies causing the magma to
undergo more extensive crystallization than would occur in typical ridge axis settings where
magma chambers are believed to be more steady-state. This hypothesis is proposed to account
for the common occurrence of MOR ferrobasalts and FeTi basalts at propagating rift tips in the
Galapagos Spreading Center (Christie and Sinton, 1981) and lavas as evolved as andesites at the
RTI at ~850W (Fomari et al., 1983; Perfit et al., 1983; Juster et al., 1989).
Silicic volcanism has also recently been documented on the Pacific Antarctic Ridge (PAR)
(Hekinian et al., 1997, 1999; Stoffers et al., 2002; Haase et al., 2005). Stoffers and others (2002)
discovered different populations of evolved lavas, some consistent with crystal fractionation at
low and high oxygen fugacities, and others resulting from the magma mixing of highly
fractionated magmas and unevolved basaltic melts. They hypothesize that extensive crystal
fractionation occurred in a solidification zone surrounding a magma chamber (e.g. Nielson and
Delong, 1992). Residual silicic magma is proposed to migrate upward along the margin of the
magma chamber and assimilate altered basaltic wall-rock, increasing its oxidation state (Stoffers
et al., 2002). Haase and others (2005), associate the silicic lavas with a ridge axial high; a
location where large volumes of magma are found and where there is associated hydrothermal
venting. They further hypothesize that extensive fractional crystallization, facilitated by cooling
from the hydrothermal vent fields, as well as assimilation of hydrothermally altered crustal
material is responsible for the petrogenesis of the andesites and dacites found on the Pacific-
Recent studies have also considered the roles of fractional crystallization and crustal
assimilation in hotspot environments such as the Galapagos Islands (Geist et al., 1998) and
Iceland where crustal assimilation has been hypothesized to, in part, contribute for the
heterogeneous nature of Icelandic volcanism (Nicholson et al., 1991; Furman et al., 1995; Gee et
Figure 1: Map of the Juan de Fuca Ridge. The Juan de Fuca medium-rate spreading ridge in
located between the Pacific Plate and Juan de Fuca. It is located approximately 440 km (~23 8
nm) off of the coast of Washington and Oregon and spans almost 500 km between the Blanco
and Sovanco Fracture Zones. The samples described in the thesis were recovered from the
southern intersection of the Juan de Fuca Ridge with the Blanco Fracture Zone, as highlighted by
the yellow box. Map created using the open source java application at www.geomapapp.org.
13?94' 130"52' 131~SO'
13WB' ~30r6' ~50114'
Figure 2: Overview of MBARI dives and rock cores on the southern Cleft segment since 2000.
53 rock cores and 276 rock and glass samples have been recovered from the Cleft segment using
the Remotely Operated Vehicle (ROV) Tiburon (Stakes et al., 2006). In 2000, a sample of dacite
was recovered by rock core from a small topographic dome in an area of the Cleft segment
characterized by what appear to be curved ridges and volcanic cones that overlap the
westernmost part of the Blanco Transform. Dive T73 5 covered the area where the dacite rock
core was taken (Stakes et al., 2006).
STUDY AREA AND SAMPLE RECOVERY
In 1998 the Cleft segment was surveyed using a hull-mounted 30 k
multibeam sonar and using 2 degree by 2 degree beam resolution, the EM300 achieved a ~30 m
lateral resolution over a 3 km swath width (Stakes et al., 2006). The mapping program was
supplemented by a series of in situ observations made from the ROV Tiburon, operated from the
Research Vessel Western Flyer, during July 2000, August 2002 and August 2004 (Stakes et al.,
2006). A total of 16 ROV dives were completed across the Cleft spreading center; Hyve dives
across the axis near the South Cleft hydrothermal Hields; Hyve dives on the southernmost part of
the segment; two dives on the northern wall of the intersection with the BFZ and dive T73 5, the
focus of this thesis, on the hooked ridges that define the western side of the nodal basin where
the ridge axis intersects the BFZ (Figure 2).
Dive T73 5 utilized the ROV Tiburon in order to observe and sample the area where a
dacitic glass was recovered by piston core in 2000 (Figure 3). The dive was dedicated to
investigating the dome-like features observed in the bathymetry and recovering samples along a
dive traverse that began south of the core location and ended to the west of the southernmost
portion of the Cleft axial valley. A detailed sampling program was carried out as part of the
ROV and surface ship operations during the dive. Lava samples were recovered with the
Tiburon manipulator and details of each sampling locality were documented by the scientists in
charge of observations and sampling during each dive. The ROV observations, contemporaneous
magnetic field measurements, digital still and video images and geologic samples were all
located with respect to the EM300 bathymetry through a real-time ArcView-based navigation
and GIS system using the EM300 bathymetric basemap (Stakes et al., 2006). The bathymetric
data, ship locations and ROV USBL (ultra-short baseline) navigation used a common GPS data-
stream with real-time depth (from the ROV) providing consistent position information within 10-
20 meters of bathymetric features on the EM300 bathymetry. Below, the bathymetric,
observational and sample data are integrated to provide a detailed account of the geology along
T735 Dive Observations
Dive T73 5 began at 14:49 Greenwich Mean Time (GMT) on a talus slope at the southern
end of the dive track (Figure 4) but soon crossed an area dominated by lightly sedimented, intact,
basaltic pillow flows, crosscut by several N-S trending fissures. Basalt samples T735-Gl, at
2187 m, through T73 5-G6, at 2216 m, were recovered from this area, before Tiburon began to
traverse upslope. At 16:05 GMT an area of lava drain-back was observed, with moderate
sediment cover in between the pillows and lobate flows. At 16: 13 an area of more blocky flows
was observed and sample T745-G7, a basalt, was recovered from the top of a small knoll (2213
m). At 16:25 GMT Tiburon dropped over the edge of a fissure and moved into an area of thick
sediment cover with isolated pillows basalts. Sample T73 5-G8 was recovered from this area at
16:39 GMT at 2232 m depth. Tiburon then moved through an area of isolated pillows with <50%
sediment cover, before reaching an area of lightly sediment covered, large striated pillows and
associated sheet flows.
Sample T73 5-G9, a dacite, was recovered at 16:45 GMT from a friable sheet flow at the
base of the first of the domes. Samples T73 5-G10 and G11i, both dacitic in composition, were
recovered from large (> Im in diameter), glassy, striated pillows lightly covered with sediment
observed in conjunction in the sheet flows that flowed down the slope to the southwest. As
Tiburon continued to traverse upslope the flows became more blocky and massive, with vapor
pockets and cavities observed between the layers of rock and very little sediment cover. T73 5-
Gl2, one of the most evolved dacitic samples, was collected at 16:59 GMT from one of the
vapor pockets observed in the area at 2211 m depth and was extremely glassy in appearance.
Tiburon then passed over an area with more local relief, in the form of big, sediment-coated
pillows with bread-crust textures surrounded by smoother pillow tubes.
At 17:06 GMT Tiburon moved into an area of fractured sheet flows where andesite
samples T735-Gl3 (17:09 GMT) and Gl4 (17:22 GMT) were collected. At 17:27 GMT,
smoother pillow forms and tube morphologies dominate with more massive flows between.
Then sample T73 5-Gl5, another andesite, was recovered from a region of highly vesicular (10-
15% vesicles), blocky flows on top of what appeared to be a constructional dome at 17:32 GMT.
The interior of the blocky flows was sampled at 17:46 GMT (T745-Gl6) at 2216 m and is dacitic
in composition. After collecting T735-Gl6 Tiburon began to traverse down-slope to the east.
Pillows with bread-crust texture were observed under moderate (<30%) sediment cover. Down
the slope of the dome at 18:06 GMT sample T735-Gl7, a dacitic pillow fragment, was recovered
from a saddle at 2221 m.
To the east, up, out of the saddle, another dome structure was encountered. At 18:16 GMT,
two small pieces of andesitic pillow crust were collected at 2206 m. Continuing upslope, smaller
pillow forms (< 1 m in diameter) were observed up to the top of the dome at 2198 m where the
morphology was dominated by flattened lobate tubes, some of which appeared to flow down-
slope to the south. Sample T73 5-Gl9, an andesite was recovered from the top of the dome
construct at 18:25 GMT.
At 18:29 GMT Tiburon began the slow descent down-slope over a structurally undisturbed
area consisting of well-formed pillows and tubes with only moderate sediment cover. At 18:36
GMT blockier, broken pillows and tubes were observed along a shallow slope. At 2234 m,
18:3 5 GMT, sample T73 5-G20, a dacite, was recovered from a small pile of talus near an in situ
pillow flow. This flow front of pillows over talus continued for a few 10's of meters before the
slope began to steepen towards the east.
Sample T73 5-G21, a large basaltic fragment from an intact pillow, was recovered from the
base of the steep slope at a depth of 2275 m at 19:01 GMT. A region of heavy sediment cover
with isolated pillows was then traversed. Sample T735-G22, was recovered at 19:16 GMT, 2256
m, from an area of extensive collapse (likely drainback) that was associated with sheet flows
under moderate sediment cover. A sample of an isolated basaltic pillow, T735-G23, was
recovered at 19:26 GMT from 2250 m in depth. Afterwards, pillows began to dominate the
landscape once more; recovery of samples G24 and G25 began at 19:37 GMT and they were
both collected from a haystack at a depth of 2249 m. The northern portion of the dive covered
much more tectonized terrain within a zone of pillow ridges separated by N-S trending fissures.
Samples T73 5-G26 through G39, all basalts, were recovered from pillow to lobate flows within
DIVE T735 (Cleft Dive #1) AUGUST 31, 2004
13P28'30" 130~'2BDD 1305t730' 13057'00'
1302B'00" 1302730" 130'27'00"
Figure 3: Overview of the dive track taken during Tiburon Dive 735. Yellow dots denote sample
locations, while yellow bulls' eyes denote the 2000 rock cores. The 2000 rock core was
recovered near the T735-Gl5 sample site. The evolved sample set is represented by samples
T73 5-G9 through T73 5-G20.
(u pd g
Figure 4:Poiempo aiedm opooie n vle apelctos rfl
esimte frmt edetofteR VBaatsmlsaedsgaebyteludi ons
andeiti sape r eintdb rng rage n aie yre icewt vr it
sample ben nubrd ag hne nth O shaigaentd
Thirty-nine lava samples were collected from the RTI of the JdFR and the BFZ during dive
T73 5 of the ROV Tiburon during the MBARI research cruise in 2004. Thin-sections were made
of a representative suite of the evolved samples and one basalt recovered during the dive.
Natural glasses were separated and coarsely crushed, and crystal-free glass separates were hand-
picked under a binocular microscope. Evolved lavas had fairly crystalline glass and in picking
those glass chips, crystal clots and phenocrysts were avoided and the least crystalline chips were
chosen for analysis. The crushed glass chips were then ultrasonically cleaned for 10-20 minutes,
in a solution made from equal parts 2.5N reagent grade HCI and 30% H202 in Order to remove
any surficial coatings, such as MnO, adhered to the sample.
Clean glass chips were mounted to thin sections for microprobe analysis, with glass from
UJF internal standard 2392-9, an N-MORB from the East Pacific Rise at 90 50' N, included on
each slide. Major element concentrations were determined on natural glasses using JOEL
electron microprobes at the United States Geological Survey (USGS) in Denver, CO, with the
help of Dr. Ian Ridley, and at Florida International University (FIU) by Dr. Michael Perfit.
Traditional mineral standards were used to calibrate the microprobe analyses and secondary (off-
line) normalizations were completed using the UF standard 2392-2, and the USGS standard
GSC, a synthetic andesite microbeam glass (Riehle et al., 1999). Accepted values for these
standards are given in Table 1. Operating conditions for the microprobe were an accelerating
voltage of 15 keV and a beam current of 20 nA. Probe diameter for mineral analyses was set to
<10 pm and defocused to 20 pm for glass analyses.
Typical precision was evaluated by comparing repeat analysis of UF standard glass 2392-9
(Table 1) which was mounted and analyzed on every probe slide. Percent variation in the 2392-9
analytical values was low, ranging from a one-sigma standard deviation of 6.4% in K20 to
0.13% in the SiO2. SiO2 ValUeS showed strong linear correlations with analysis total and were
corrected to using the following formula, [raw SiO2 value + ((99.7 raw total)*0.4505)], where
the raw silica value has an additional factor added to it consisting of an assumed total (99.7)
minus the raw total multiplied by the slope of the trend-line of raw SiO2 VeTSus the raw total
Trace element concentrations were determined at the University of Florida using an
Element 2 Inductively Coupled Plasma- Mass Spectrometer (ICP-MS). Dr. George Kamenov
developed the dissolution and analytical procedures (see below). Sample preparation was
performed at UF in a class 1000 clean lab facility. Standards used included the well-
characterized internal UF standards 2392-9 and ENDV, as well as the USGS standard BHVO-1,
the surface layer of a 1919 Hawaiian pahoehoe flow (Flanagan, 1976). Trace elements measured
by the UF ICP-MS have been determined to be accurate and precise to better than 15% of their
concentration (Table 2).
Phase one of the dissolution process requires that two to three standards encompassing
the likely potential range of the unknown sample concentrations be chosen; for the T735 run
AGV-1 and BCR-2 were used and compared along with two in-house standards, ENDV and
2392-9. Next, tall, clean, hex-cap Savillex Teflon vials were labeled for each sample, including
one blank. Two drops of 4x water were placed in each vial and the scale was zeroed before ~.04
g of either clean glass chips were added to each vial and the precise weight was recorded. In the
clean lab, 1 mL of optima grade HF and 2 mL of optima grade HNO3 were added to each vial
and then the vials were tightly capped and placed in a 1000C oven for approximately 48 hours.
Samples were then dried down on a hot plate (carefully rinsing the caps into the vials with 4x
water) for 12 to 24 hours.
In phase two of the dissolution process, 4 mL of an internal standard (5%HNO3, with Re
and Rh) was measured into each vial. The vials were then capped and heated on the hot plate
In order to dilute the samples for analysis 200 CLL of solution was pipetted from each vial
and transferred to a clean auto sampler tube and weighed. Four mL of the same internal standard
was added to each tube and the samples were weighed again. The final dilution factor was
Samples were introduced into the ICP-MS in one-minute uptake times followed by two-
minute washes for each sample. The specific isotopes analyzed were: Sc45, V 1, Cr52, CO59, Ni60,
Cu63, Zn66, Ga69, RbsS, Srss, Y89, Zr90 N 93, Rhl03, Bal137, Lal139, C140, Prl41, Ndl43, Sml49, Eul53,
Gd 5, Tbl59, Dyl63, Hol65, Erl66, Tml69 ybl72, Lu 7, Hf 7, Ta s, Re s, Pb208, Th232 and U238
Four runs were made with four passes per analysis in Medium Resolution mode.
Data were reduced on-line using calibration curves of the standard data acquired during
analyses and drift corrections were mathematically calculated off-line. The ENDV standard was
used for drift corrections during each analytical run; one measurement was taken at the
beginning of the sample series, one in the middle and one at the end of the series. Accuracy and
precision for the T735 analytical run was evaluated by analyzing ENDV standard as a sample
and comparing the results to recent runs of the standard using the same ICP-MS instrumentation
(Table 2). One-sigma standard deviations from the accepted values were less than or equal to
3.6% during these analyses.
Table 1: Accepted values for maj or element standards.
Internal UF Standard 2392-9 Correction Factors
SiO2 SiO2 TiO2 Al203 FeO MnO
MgO CaO Na20 K20 P205
for the averages of each group
average 50 49.49
std dev 0 0.20
% var 0 0.41
2392-9 50.04 50.04
factor 1.00 1.01
for all analyses together
average 50 49.52
std dev 0 0.29
1.28 15.41 8.99 0.15
0.02 0.19 0.05 0.01
1.57 1.22 0.50 5.13
1.31 15.48 9.38 0.18
1.03 1.00 1.04 1.18
1.28 15.41 8.99 0.15
0.04 0.20 0.10 0.02
0.04 1.32 1.07 14.73
1.31 15.48 9.38 0.18
2.62 0.10 0.11 0.004
0.03 0.01 0.01 0.001
1.33 6.43 5.59 22.79
2.56 0.09 0.12
0.98 0.92 1.07
2.62 0.10 0.11 0.004
0.06 0.01 0.03 0.003
2.39 11.77 27.06 71.99
2.56 0.09 0.12
0.98 0.92 1.07
1.01 1.01 1.03 1.00 1.04 1.17
GSC Andesitic Glass Correction Factors
SiO2 TiO2 Al203 FeO MnO MgO CaO Na20 K20 P205 C1V
average 62.11 0.01 13.71 6.27 0.02 3.77 4.99 3.58 3.63 0.04 0.01
std dev 0.45 0.02 0.21 0.10 0.01 0.07 0.05 0.33 0.05 0.02 0.00
% var 0.73 146.27 1.56 1.57 56.90 1.73 1.10 9.11 1.30 49.46 73.04
GSC 62.05 0.01 14.20 6.33 0.03 3.89 5.00 4.06 3.60
factor 1.00 0.85 1.04 1.01 1.75 1.03 1.00 1.13 0.99
Final Correction Factors Used
SiO2 TiO2 Al203 FeO MnO MgO CaO Na20 K20 P205
1.04 1.1 1.03
Table 2: ICP Trace element standards. 2005-2006 analyses of ENDV (ran as sample) together
with other MORB samples (drift 1 first sample after the standards).
ENDV Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7 Avg dev error
(ppm) (g/mL) (g/mL) (g/mL) (g/mL) (pLg/mL) (pLg/mL) (pLg/mL)
Sc 38.00 39.20 39.30 37.63 38.41 37.55 36.66 39.18 38.27 1.02 2.67
V 275.60 278.82 278.53 269.72 268.44 268.33 262.17 275.48 271.64 6.17 2.27
Cr 290.20 297.94 292.23 279.24 285.96 264.80 262.94 277.86 280.14 13.13 4.69
Co 34.40 35.06 34.82 34.32 35.28 35.06 34.74 35.84 35.02 0.48 1.36
Ni 76.20 77.93 77.24 75.36 76.13 76.13 75.21 78.85 76.69 1.36 1.77
Cu 72.40 72.86 73.02 70.64 72.69 71.41 70.83 72.38 71.98 1.00 1.39
Zn 73.15 78.03 79.20 78.95 78.93 77.45 78.75 77.78 2.13 2.74
Ga 17.20 17.19 16.76 16.91 16.73 17.07 16.86 17.02 16.93 0.17 0.99
Rb 4.38 4.54 4.40 4.37 4.53 4.24 4.26 4.38 4.39 0.12 2.68
Sr 155.80 162.21 156.59 155.15 154.02 153.99 155.42 154.84 156.03 2.87 1.84
Y 32.40 31.65 32.62 32.47 31.96 31.31 31.23 32.81 32.01 0.64 1.99
Zr 113.40 112.58 114.04 112.39 111.96 110.72 110.07 105.06 110.98 2.91 2.62
Nb 6.46 6.62 6.47 6.39 6.34 6.29 6.26 6.46 6.40 0.12 1.93
Ba 64.20 69.90 73.28 70.87 72.80 71.86 71.27 68.19 71.17 1.74 2.44
La 6.02 6.01 6.02 5.97 5.92 5.93 5.90 6.00 5.96 0.05 0.78
Ce 15.60 15.90 15.69 15.40 15.41 15.49 15.52 15.72 15.59 0.18 1.18
Pr 2.46 2.52 2.53 2.44 2.44 2.46 2.44 2.49 2.47 0.04 1.59
Nd 12.60 12.63 12.56 12.60 12.43 12.49 12.21 12.78 12.53 0.18 1.43
Sm 3.76 3.85 3.78 3.76 3.75 3.79 3.74 3.78 3.78 0.04 0.94
Eu 1.34 1.36 1.33 1.33 1.32 1.34 1.33 1.33 1.34 0.01 0.92
Gd 4.88 4.80 4.94 4.88 4.83 4.77 4.76 4.97 4.85 0.08 1.71
Tb 0.86 0.84 0.86 0.86 0.85 0.84 0.83 0.86 0.85 0.01 1.54
Dy 5.28 5.31 5.35 5.29 5.21 5.18 5.21 5.36 5.27 0.07 1.41
Ho 1.10 1.13 1.13 1.10 1.09 1.13 1.11 1.13 1.12 0.02 1.75
Er 2.98 3.13 3.08 2.97 3.03 2.99 2.99 3.02 3.03 0.06 1.84
Tm 0.48 0.48 0.50 0.48 0.49 0.48 0.48 0.49 0.48 0.01 1.51
Yb 3.12 3.12 3.14 3.11 3.15 3.10 3.05 3.12 3.11 0.03 1.03
Lu 0.48 0.49 0.47 0.49 0.49 0.46 0.46 0.48 0.48 0.01 2.83
Hf 2.88 2.85 2.95 2.88 2.83 2.79 2.78 2.63 2.81 0.10 3.56
Ta 0.38 0.37 0.37 0.39 0.37 0.38 0.38 0.39 0.38 0.01 2.34
Pb 2.00 2.06 2.05 1.97 1.97 1.97 1.99 2.07 2.01 0.05 2.26
Th 0.46 0.47 0.46 0.46 0.46 0.46 0.46 0.45 0.46 0.01 1.45
U 0.19 0.20 0.19 0.18 0.19 0.19 0.19 0.19 0.19 0.00 2.07
PETROGRAPHY AND MINERAL CHEMISTRY
Approximately two thirds of the samples recovered from dive T73 5 were basaltic in
composition, although all are moderately to highly evolved compared to previously published
Cleft MORB lava compositions. Basaltic lava morphologies ranged from pillow tubes to sheet
flows (Figure 5). Basalt hand samples have a thin glass rind with a microcrystalline interior
containing macroscopic plagioclase laths and pyroxene crystals.
A thin section was made of sample T73 5-G23 going from the surface of the hand sample
into the microcrystalline interior. The thin section average ~80% glass with ~20% quenched
microphenocrysts of plagioclase, ranging from 2 3 CL to 2 mm in length, with the average being
~ Imm in the groundmass. Larger glomeroporphyritic clusters comprised of euhedral to
subhedral, Imm long plagioclase laths, An64-An7o, with simple Carlsbad twins and stepped,
irregularly spaced twinning, and inter-grown anhedral crystals of augitic clinopyroxene (Figure
6) are present, with an Mg# of~-81.3, with the Mg# being defined as 100* [Mg/[Mg+Fe]].
Andesites and Dacites
Of the 39 lava samples collected, 12 were found to be high-silica lavas, five of which are
andesitic (52% < SiO2 <63%) and seven are dacitic (63%
lavas were recovered from two large constructional domes, approximately 34m in height, ~200
to 500 m in diameter, comprised of atypically large vesicular pillow flows, and extremely
blocky, vesicular flows (Figure 7). The flows appeared to be fresh, due to the lack of significant
sediment cover and the thinness of the manganese coating on the recovered hand samples. Some
of the andesite-dacite hand samples are extremely vesicular (10 15%) with elongate vesicles
(1-10 cm) in a glassy matrix. In hand specimen, the samples are composed of a glassy matrix
with small crystals and crystallites (~20 30%) of swallowtail plagioclase, pyroxene and
fayalitic olivine as well as larger semi-circular fragments of darker basaltic xenoliths scattered
throughout. Phase chemical data are presented in Appendix 2. Andesite and dacite assemblages
are dominated by microphenocrysts of ferroaugite and ferropigeonite, lesser amounts of sodic
plagioclase and FeTi oxides, and rare fayalite, zircon and myrmekitic intergrowths of plagioclase
and quartz. Fine-grained basaltic xenoliths are included within all of the evolved lavas.
Matrix glass is variable in color ranging from light to dark brown, nearly opaque, in
proximity to the basaltic xenoliths. Individual euhedral to rounded, subhedral plagioclase laths,
Anls An59, WeTO found scattered throughout the thin-sections and comprise ~5 10% of the
volume of the samples. Crystals have lower first order colors and simple Carlsbad twins or
stepped, irregularly spaced twins. Rare plagioclase oikocrysts with clinopyroxene chadacrysts
are present in one sample, T735-G10 (Figure 8, 9). Plagioclase crystals have two distinct
compositional groupings in the evolved samples. Plagioclase ranging from An69 to An83, OCCUT
in xenolithic clots and as large xenocrysts that appear to be in disequilibrium. More sodic
plagioclase, An38 An66, Occurs in the smaller, equilibrium crystals and in the myrmekitic clot.
Most individual clinopyroxene crystals exhibit finely zoned rims surrounding a more
massive core. Several examples of both normal and reverse zoning are observed in the
clinopyroxene grains in the evolved rocks (Figure 10). Rare clinopyroxene crystals with
oscillatory zoning, or hourglass sector zoning were also observed in these samples, a
phenomenon that has been associated with rapid crystal growth (Carpenter, 1980; Shelley, 1993).
Additionally, the observed sector zoning is commonly combined with fine concentric zoning of
crystal rims (Figure 11). Skeletal clinopyroxene crystals, also indicative of rapid crystal growth,
exhibit lacey, almost spongy, reaction rims (Figure 12). The cores of the pyroxene crystals fall
into two distinct groupings, with one group exhibiting cores that range from Mg# of 56.0 58.6,
and the other range in core Mg# from 8.24 12.6. In addition to these two groupings, there is
also a population of individual pyroxenes, Mg# ranging from 29.3 30.21, whose compositions
correspond with those of the rims in the zoned pyroxenes (Figure 13).
Rare fayalite crystals, Fos Fols, were found in several of the evolved lavas. In plain
polarized light (PPL) these crystals are rounded or embayed and have a deep green color (Figure
14). In crossed polars (XPL) the fayalite crystals displayed high second order colors and a low
2V angle of ~500. Skeletal fayalite crystals have lacey to spongy reaction rims similar to those
seen on the quenched clinopyroxene crystals (Figure 15).
Rare (occurring <1% by volume of the sample) subhedral zircon crystals were found in
several of the dacitic to rhyodacitic samples (Figure 16). Zircons were first discovered through
microprobe analyses and later identified by their high relief and very high birefringence in thin
section. Zircon grains occur almost exclusively as individual grains with no apparent connection
to xenoliths or crystal clots.
Two populations of xenoliths were found in the evolved samples. Rounded xenoliths of
basaltic composition with chilled margins and microcrystalline interiors (Figure 17; Fiigure 18)
and coarse-grained xenoliths composed entirely of large crystals, > 1mm, with rough, irregular
shaped edges. The basaltic xenoliths are composed of plagioclase, An32 An71, and pyroxene,
Mg# ranging from 56.7 to 80.9, and range in size from <1 mm to several mm in diameter.
Larger crystals (>0.5 mm in length) are usually confined to the center of the xenoliths and can
become subophitic in texture with large pyroxene crystals partially enclosing smaller plagioclase
laths. Basaltic xenolithic inclusions have very dark, glassy to microcrystalline interior edges that
extend around the circumference of the inclusions. An optically clearer (PPL) and less
crystalline and darker zone, ~. 1-.5 mm thick, surrounds many of the xenoliths.
The coarser grained xenolith, found in T73 5-G10, an andesite, is composed of plagioclase,
An34 An35, and fayalitic olivine, Fols (Figure 19). It exhibits no mineralogical reactions along
the margin, but the edges of the xenolith are ragged and look like they were ripped from the
A single, resorbed myrmekite of plagioclase, An27 An30, and quartz was found in sample
T735-Gl9 (Figure 20). This is typically a plutonic texture that occurs primarily through grain-
boundary reactions in slowly changing conditions (Shelley, 1993). The myrmekite is rounded
and ~1 mm in diameter, with worm-like intertwining of plagioclase and quartz.
Figure 5: Examples of basaltic lava morphologies seen during dive T73 5. Left: Lobate flows at
the top of a fissure. Lobates usually form at the edges of sheet flows when the lava flow slows
enough to form a fluid core under a solid crust. Right: Pillows, ~lm or less in diameter, and
pillow buds under moderate sediment cover. Pillows form when the extrusion rate of the lava is
slow enough to form a thick outer crust around the erupting lava.
Figure 6: Plain polarized and cross polar views of glomeroporphyritic texture in T73 5-G23.
Thin sections average ~80% glass with quenched microphenocrysts of plagioclase scattered
throughout the groundmass. The larger glomeroporphyritic clusters were comprised of euhedral
to subhedral, Imm long plagioclase laths, An64-An70, and inter-grown anhedral crystals of augitic
clinopyroxene, with an Mg# of ~81.3.
Figure 7: Dacite dome lava morphologies. The dacite domes display much different types of lava
morphologies then those seen where basaltic samples were recovered. Left: Blocky, vesicular
flow at the edge of one of the dacite domes. The large amount of vesicles seen is a by-product of
the degassing lava. Right: Extremely large pillow tubes (>1.5m in diameter) were also noted
cascading down-slope on the domes.
Figure 8: Plagioclase oikocryst surrounding several randomly oriented clinopyroxene
chadacrysts in sample T73 5-G10. This poikilitic texture can be indicative of plutonic origins and
it indicates that the oikocryst mineral had a much more rapid growth rate then the enclosed
chadacrysts (Shelley, 1983; Higgins, 1998).
Figure 9: Element map of poikilitic texture seen in T73 5-G10, in this case, a plagioclase
oikocryst surrounding low-Ca pyroxenes. Element map were completed using X-ray imaging on
the JOEL microprobe. Element labels are under each picture and count scales are located on the
right-hand side of the images. Brighter colors indicate higher concentrations of a particular
element. The gray-scale image is the backscatter image in which elements with higher mass are
0 0.2 0.4 0.6 0.8 1 1.2
Relative Distance from Core to Rim
Figure 10: Point compositions of pyroxene phenocrysts in T735S-G9 and T735S-Gl2 are plotted to
show zoning patterns. Oscillatory (points 1-4; blue line) and reverse (points Sand 6; pink line)
zoning can be seen in the pyroxene from T735S-Gl2 (right). The pyroxene from sample T735S-G9
(left) also demonstrates compositional zoning through its hourglass pattern.
Figure 11: Oscillatory zoned clinopyroxene in sample T735-G9. This crystal texture is indicative
of rapid crystal growth. Note the fine concentric zoning on the edges of the crystal and the
reversely zoned pyroxene to the upper left of the large crystal. Mapping details in Figure 9.
Figure 12: Skeletal clinopyroxene grain with a lacey reaction rim surrounding the crystal, this is
also indicative of rapid crystal growth. Mapping details in Figure 9.
0 0.2 0.4 0.6 0.8 1 1.2
Relative Distance from Core to Rim
Figure 13: Populations of zoned pyroxenes. There are two distinct populations of zoned
pyroxenes. One group exhibits cores that range from Mg# of 56.0 58.6, and the other range in
core Mg# from 8.24 12.6. In addition to these two groupings, there is also a population of
individual pyroxenes, Mg# ranging from 29.3 30.21, whose compositions correspond with
those of the rims in the zoned pyroxenes.
Relative Distance from Core to Rim
Figure 14: Olivine and plagioclase glomerophyric cluster. Olivine is fayalitic in composition.
Figure 15: Element map of skeletal fayalite crystal in sample T73 5-Gl2. Note the lacey reaction
rim. Mapping details in Figure 9.
Figure 16: Rare (<1% of the sample) euhedral zircon phenocrysts seen in samples T735-Gl2 and
T73 5-Gl9 were discovered through microprobe analyses. Zirconium concentrations in the
evolved samples range from 397 to 672 ppm. They were later able to be recognized in thin
section due to their high relief and birefringence.
;"I (gFi~hq" ~ *:i~Cr Q1;
,2 \ \-~51''57~
-r ~;PI LS~
_X;I Y:~1 r
r ; '1/+
Figure 17. PPL (left) and XPL (right) views of a basaltic xenolith in sample T73 5-G1 i. Basaltic
xenoliths typically are rounded with slightly coarse grained interiors and chilled margins.
Figure 18: Element maps of a basaltic xenolith in sample T73 5-Gl2. Note the high Mg content,
indicating the basaltic nature of the xenolith. Mapping details in Figure 9.
Figure 19: A coarser grained xenolith, found in T73 5-G10, is composed of plagioclase, An34 -
An3 5, and fayalitic olivine, Fols. The edges of the coarse grained xenolith are rough and there
are no signs of mineralogical reactions. Mapping details in Figure 9.
Figure 20: Myrmekitic intergrowth of quartz and plagioclase. The PPL view of myrmekitic
intergrowth of plagioclase, An27 SO, and quartz found in sample T73 5-Gl9. Element maps
of the myrmekite display the wormy intergrown texture of the quartz and sodic plagioclase.
Mapping details in Figure 9.
MAJOR AND TRACE ELEMENTS
Mid-ocean rdge basalts recovered from the Cleft segment exhibit geochemical
characteristics consistent with typical normal, incompatible element-depleted MORB (N-
MORB). Representative basaltic glass compositions are given in Table 3, with the complete data
presented in Appendix 3.
Major and Minor Elements
Basalt compositions are fairly evolved overall with nineteen of the twenty-three basalts
recovered being ferrobasalts (FeOT > 12.0 wt.%). Basalt compositions are tightly grouped with
MgO ranging from 7.55 weight % in the most primitive basalts to 6.29 wt% in the most evolved
basalt recovered (Table 3). The Al203 ValUeS decrease from 13.8 wt%, in the most primitive
basalt sampled, to 12.8 wt% in the most evolved. The CaO concentrations also fall, with
abundances decreasing from 11.8 wt% to 10.9 with decreasing MgO. The FeOT and TiO2 ValUeS
increase with decreasing MgO from 10.8 to 12.9 wt% and 1.61 to 2.26 wt% respectively.
Concentrations of Na20, K20 and P20s also increase with decreasing MgO, rising from 2.45 to
2.69 wt%, 0.12 to 0.20 wt% and 0. 14 to 0.25 wt% respectively. Lastly, MnO shows a loose
overall decline, from 0.24 to 0.18 wt% with decreasing MgO, the trend is scattered due to the
low concentration of the oxide in the samples. The maj or elements within the recovered basalt
suite all plot smoothly relative to MgO (and each other) with no maj or inflection points.
The T735 basalts have compositions similar to other southern JdF basalts but have limited
variation compared to the all of Cleft segment basalts, of which two hundred and eighty-six
samples were analyzed for major elements (Smith, 1994; Stakes et al., 2006; Figure 21).
Overall, Cleft segment MORB show relatively little maj or element chemical variation at any
given MgO content, forming well-defined geochemical trends consistent with differentiation via
low-pressure fractional crystallization of similar parental lavas (Smith, 1994; Figure 22).
Previously published data for Cleft segment lavas range in MgO from 10.5 wt%/ in the most
primitive basalts, to 4.41 wt% in the most evolved samples and reported Mg#'s range from 70.8
to 32.6. However, the maj ority of reported data exhibit MgO contents greater than 6 wt%, with
sixty samples extending the compositional range to the more evolved compositions. The maj or
elements for the T73 5 basalts all plot near the center of the overall range of MgO values (Figure
The Al203 ValUeS of the T73 5 basalts are lower at a given MgO than in the existing Cleft
data set which ranges from 17.7 to 12.2 wt% with falling MgO (Smith, 1994; Stakes et al.,
2006). For example, at an MgO of ~7 wt% Cleft basalts range from 13.5 to 14.9 wt%, whereas
T735 samples range from 13.1 to 13.3 wt%. Similarly, FeOT COncentrations of T735 samples,
which range from 12.9 to 10.8 wt%, form a trend within the Cleft data trend which ranges form
7.92 to 16.3 wt% with falling MgO (Smith, 1994; Stakes et al., 2006). At an MgO of ~7 wt%
FeO ranges from 11.5 to 13.1 wt% in Cleft basalts, while T735 samples have a range of 11.07 to
1 1.22 wt%. Within the overall Cleft segment data, a group of 2 basalts, 8 ferrobasalts and one
andesite form an apparent subsidiary trend towards lower FeOT COncentrations, from 13.4 to 11.6
wt%, at values of MgO, ranging from 5.98 to 4.46 wt. % (Figure 22) while most of the Cleft data
exhibits an increase in FeO to~-16 wt%. It is important to note that the group of low FeO
samples exhibits a trend with decreasing MgO towards the compositions of the andesite and
dacite samples described below.
The TiO2 ValUeS for dive T73 5 basalts and ferrobasalts are relatively high particularly the
more evolved end of the ferrobasalts, which have values from 1.60 to 1.99 wt% TiO2 at a MgO
value of ~7 wt% and ranges from 2.02 to 2. 18 in the T73 5 dive samples. In both the Cleft
basalts, which range from 0.89 to 3.15 wt% TiO2 with falling MgO, and the T73 5 samples the
TiO2 trend in the basalts is one of increasing concentration with decreasing MgO. As with the
FeO data discussed above, there are several Cleft samples in the ferro-basalt/andesite range that
display relatively low TiO2 concentrations trending towards the compositions of andesites and
dacites recovered in dive T735. The CaO and Na20 data for T735 basalts and ferrobasalts fall
in the center of the data trends established by the Cleft samples, which range from 12.6 to 8.47
and 2. 11 to 3.38, respectively, with decreasing MgO. The observed trends of decreasing CaO and
increasing Na20 with decreasing MgO are also smooth, without inflection points.
Overall, the maj or element trends observed in the data support fractional crystallization of
plagioclase and olivine, indicated by the trend of decreasing Al203, with decreasing MgO
coupled with a trend of increasing CaO/Al203 with decreasing MgO observed in the most mafic
samples. It is likely that clinopyroxene enters as a crystallizing phase when the magma reached
approximately 7.9 wt% MgO, as indicated by the inflection point in CaO/Al203 (Figure 22)
which steadily decreases thereafter.
Water and CO2 COncentrations for two of the T73 5 basaltic lavas, T73 5-G23, and T73 5-
G35, the most mafic sample recovered, are 0.27 and 0. 17 wt. % and 118 and 101 ppm,
respectively. Chlorine concentrations for the T735 basalts range from 90 to 410 ppm with
Trace element concentrations in the T735 basaltic lavas lie within the compositional ranges
of other Cleft basalts (Figure 23). While the T73 5 basalts display little to no trend on their own,
when included with the reported Cleft data overall trends are apparent. Zirconium increases in a
tight, smooth pattern with decreasing MgO from 102 to 142 ppm in the T735 basalts and from
51.4 to 607 ppm in the Cleft samples (Smith, 1994; Perfit, unpublished). Yttrium also displays
an increasing concentration with decreasing MgO ranging from 37.6 to 53.4 ppm in the T73 5
samples and from 21 to 220 in the Cleft samples. Tantalum values plot in a tight group ranging
from 0. 15 to 0.29 ppm in the T735 data with a slight, but poorly defined, increasing trend vs.
decreasing MgO while the Cleft samples range in Ta concentration from 0. 11 to 0.90 ppm, with
most samples plotting within ~.1 ppm at given MgO contents. Lanthanum, Sm and Lu all show
well defined increasing trends plotted against decreasing MgO when included with the Cleft
samples. Lanthanum ranges from 3.40 to 5.24 ppm in the T735 samples and from 1.9 to 21 ppm
in the Cleft samples, respectively, Samarium ranges from 3.49 to 4.87 ppm and from 1.8 to 17.6
ppm respectively, and Lu ranges from 0.58 to 0.81 ppm in the T735 basaltic samples and from
0.30 to 3.44 ppm in the Cleft samples (Smith, 1994; Perfit, unpublished). Rubidium and Nb
show increasing trends that are less well-defined and show greater scatter than the trace elements
discussed above (Figure 23) when plotted against decreasing MgO. Rubidium ranges from 0.73
to 1.3 1 pmm in the T73 5 basaltic samples, while ranging from 0.24 to 5.8 ppm in the Cleft
samples and vary by ~2 pmm at ~7 wt% MgO ~2.7 ppm at an MgO of ~5 wt%. Niobium ranges
from 2.88 to 4.76 ppm in the T735 basalts and from 1.4 to 18.7 ppm in the Cleft samples (Smith,
1994; Perfit, unpublished).
Strontium and Cr both decrease in concentration with decreasing MgO (Figure 23).
Strontium shows an overall decrease with decreasing MgO, ranging from 118 to 99.7 ppm in the
T735 basalts and from 150 to 82.4 ppm in the Cleft samples (Smith, 1994; Perfit, unpublished).
The spread of the main body of data is fairly wide with the Sr ranging from 98 to 128 ppm at an
MgO of ~7.2 wt. percent. Chromium shows a clearer trend ranging from 287 to 98.3 ppm with
decreasing MgO in the T735 samples and from 584 to 11 ppm in the Cleft samples (Smith, 1994;
Vanadium shows an increasing trend in the Cleft data until ~6 wt. percent MgO (full range:
8.69 to 3.41 wt. percent MgO) when there is an inflection point and a downward trend toward the
more evolved compositions. Vanadium ranges from 395 to 330 ppm in the T735 basalts and
from 569 to 185 in the Cleft samples (Smith, 1994; Perfit, unpublished). The inflection point
seen is consistent with the introduction of titano-magnetite as a crystallizing phase.
On a trace element primitive mantle-normalized plot (McDonough and Sun, 1995; Figure
24), the samples show highly incompatible element depletions and negative Sr (4.73 to 5.6 times
primitive mantle) and Pb (6.44 to 9.26 times primitive mantle) anomalies. Concentrations of the
most highly incompatible trace elements are at 10 times primitive mantle or below.
Primitive mantle-normalized Rare Earth Elements (REE) diagrams (Figure 25),
(McDonough and Sun, 1995), show that the samples have Light Rare Earth Element depleted
smooth patterns with concentrations at about 5 to 10 times primitive mantle. The light REE
show a slight depletion in relation to the heavy REE, and slight Eu anomalies, attributed to
plagioclase crystallization. Lanthanum/Samarium ratios range from 0.63 to 0.74 in the MORB
samples recovered, while Ce/Yb ranges from 0.78 to 0.89. Overall, there is an increase in total
REE with decreasing MgO content.
Andesites and Dacites
The more evolved rocks create an extension of the tholeiitic differentiation trend seen in
the recovered MORB. Representative evolved glass compositions are given in Table 3, with the
complete data presented in Appendix 3.
Major and Minor Elements
In the chemically fractionated lava suite, SiO2 TangeS from 62.0 to 66.9 wt%, over an
MgO wt% range from 1.94 to 0.6, with an Mg # range of 15.9 to 6.1; significantly lower than the
nearby basalts and all of the previously recovered samples from the Cleft ridge axis. Aluminum
oxide values range from 13.6 to 12.2 wt% with decreasing MgO and are slightly scattered
(Figure 22). Titanium oxide values range from 1.4 to 0.81 wt% and FeOT TangeS from 7.28 to
10.3 wt%, both oxides falling in loosely constrained trends with MgO content. The offshoots
from the FeTi-basalts found in the overall Cleft dataset, are most pronounced in the TiO2 and
FeOT variation diagrams as they trend towards the more evolved compositions. Calcium oxide
falls from 5.4 to 3.4 wt%/ with MgO, in a very tight trend along the same line as the basalts. The
Na20 also has a tightly constrained grouping with a range in weight percent from 4.36 to 5.07
that plots very smoothly along the same trend line as the basalt values. The K20 values increase
from 1.02 to 1.30 wt% with falling MgO concentration.
H20 concentrations in the evolved samples range from 1.5 to 2.0 wt% while CO2
concentrations were below detectable limits. Chlorine concentrations range from 4100 to 6100
ppm with decreasing MgO.
Incompatible trace elements increase in concentration in the evolved lava samples as
MgO concentrations fall and form a trend wherein the single dacite glass, that was recovered in a
rock core taken in 2000 (RC 10) is comparable and is therefore included in the discussion of the
evolved samples from Dive T735. Zirconium rises from 397 to 737 ppm in evolved samples
with decreasing MgO. Yttrium rises from 158 to 240 ppm and Ta concentrations rise from 0.927
to 1.6 ppm with falling MgO in the evolved samples. Lanthanum concentrations rise with falling
MgO from 22.4 to 33.8 ppm in the evolved lavas. The concentration of Sm ranges from 13.8 to
21.1 ppm with falling MgO and Lu ranges from 2.38 to 3.65 ppm in the evolved lavas. The Rb
and Nb concentrations increase as well with falling MgO in the evolved samples with Rb rising
from 9.17 to 12.7 ppm and Nb rising from 14.0 to 21.2 ppm in the samples.
Strontium and Cr concentrations both decrease with decreasing MgO in the T735 samples
(Figure 23). Strontium falls from 92 to 67.4 in a loosely defined trend and Cr falls from 54.1 to
1.84 ppm in the evolved lavas, overall the trend is much tighter and more distinct then seen in the
Sr. Scandium and V both decrease in concentration in well-defined trends and extend the trends
in the Cleft samples after the inflection point at ~6 wt. percent MgO. Scandium falls from 23.4
to 15.9 ppm in the T73 5 samples, and the V falls from 178 to 47.6 ppm.
Trace element concentrations show much more variation in the mantle-normalized and
REE diagrams than observed in the T73 5 basalts, consistent with certain phases differentiating
out of the melt (Figure 24). Large negative anomalies in Sr (3.19 to 4. 12 times primitive
mantle), P (5.13 to 19.3 times primitive mantle) and Ti (3.62 to 6.21 times primitive mantle) are
consistent with the crystallization of plagioclase, apatite and titanomagnetite, respectively, from
the melt. All of these phases are present in the evolved samples.
The primitive mantle-normalized REE diagram displays an overall smooth, flat pattern
within the evolved samples (Figure 25). Lanthanum/Samarium ratios range from 0.98 to 1.12,
while Ce/Y ranges from 1.0 to 1.1. Europium displays an increasing negative anomaly (7.03 to
9. 16 times primitive mantle) as MgO decreases which is consistent with the continued
crystallization of plagioclase out of the magma.
*Smith, 1994; Stakes, 2006
O Dive T735
0 10 20 30 40 50 60 70 80 90 100 MgO
Figure 21: AFM Diagram comparing Smith, 1994 and Stakes, 2006 samples to dive T735
samples. Basalts follow a typical tholeiitic differentiation trend. The evolved samples from dive
T73 5 form an extension of the tholeiitic trend taken to >80% crystallization, after a gap of no
data. The evolved samples have lower concentrations of FeO due to the crystallization of FeTi
oxides out of the melt.
0 Smith, 1994
+ Elve 735
0 2 6 8 10 1:
S2 4 6 8 0 1
+'i I f
S2 6 8 10 1:
Figure 22: Maj or element plots comparing previous studies of the Cleft segment (Smith et al.,
1994; Stakes et al., 2006) to dive T735 samples.
2 6 8 10 12
2 6 8 10 1;
) 2 4 6 8 10 12
+ Elve 735
12 -- -- --
1 0 .. .. .. .. ..
Figure 22: Continued.
E~mith,. 1994; Perfi unpublished
D 2 6 8 10 12
D 2 6 8 10 12
i 2 6 8 ID 12
] 2 6 B 10 12
D 2 6 B 10 12
Figure 23: Trace element plots comparing previous studies of the Cleft segment (Smith et al.,
1994; Perfit, unpublished) to dive T735 samples.
1~~~ ....BII I
32 6 8 10 12
D 2 6 B 109 12
8 Io 12
Figure 23: Continued.
Smith, 194; Perfit unpublished
Rb Ba Th U K Nb Ta La Ce Pb Sr Nd P Hf 1 Sen Ti Tb Y MD
Figure 24: The primitive mantle normalized (McDonough and Sun, 1995) spider diagram
displays the depletion of several key trace elements in the evolved glasses compared to the basalt
compositions. Strontium is incorporated into the structure of plagioclase crystals, while the Ti
depletion is due to the crystallization of titanomagnetite. The P depletion is due to the
crystallization of apatite in the evolved samples. Sample 99RC99 (-) is an andesite recovered
from the coaxial segment and shows the same trend as the evolved samples recovered from the
C~' .. -==cc-
iu --~ rjj
?7; 11 _rri' --i~jjl~~
3'-13 i~3j I~ i-3ii
I~ -- i7 ~i i~~~ J
i c i-
I La 1.e Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Y
Figure 25: The primitive mantle normalized (McDonough and Sun, 1995) REE plot shows the
two distinct groupings within the sample group. The basalt samples show the typical flat REE
pattern while the evolved glasses have a characteristic Eu depletion from plagioclase
crystallization. The black outline displays the REE range of the Smith, 1994, samples. The
LREE in the evolved glasses are slightly enriched compared to the basalt glasses. Sample
99RC99 shows a similar enrichment trend.
The lavas recovered during dive T73 5 represent one of the most complete and evolved
suites recovered from a MOR environment. The atypical composition of the T73 5 suite provides
an opportunity to investigate petrogenetic processes of MORB evolution that have been well
addressed by traditional magma chamber models created for on-axis volcanism that do not fully
explain the petrogenesis of these andesites and dacites. Here we explore several hypotheses in an
attempt to constrain the potential petrogenetic history of this unique suite.
One of the prevailing hypotheses for the petrogenesis of evolved MOR lavas is that they
are formed by extreme fractional crystallization of basaltic magmas, in part due to the "cold edge
effect", an effect associated with MOR lavas erupted close to an oceanic transform fault or an
area where a ridge is propagating into older oceanic crust (Christie and Sinton, 1981; Fornari et
al., 1983; Perfit et al., 1983; Johnson and Holmes, 1989; Juster et al., 1989; Embley et al., 1991;
Smith, 1993; Smith et al., 1994; Juteau et al., 1995; Tierney, 2003; Herzburg, 2004). The ridge-
transform intersection (RTI) of the JdFR and the BFZ is an area where the cold edge effect is
likely to affect the petrogenesis of MOR lavas (Stakes et al., 2006). The Blanco Transform is
approximately ~345 km long from the southern terminus of the JdFR to the northern terminus of
the Gorda Ridge (Chadwick et al., 1998) that results in the axis of the southern Cleft segment
abutting old and cold lithosphere that is approximately 6.3 million years old; hence providing an
environment for extensive cooling.
Comparison to Other Evolved Suites
Although relatively rare in MOR environments, highly evolved lavas and plutonic rocks
(generally known as plagiogranites) have been found in a number of diverse environments that
include fast, intermediate and slow spreading ridges, back-arc basins, and quite a few ophiolites.
A wide range of samples from these environments are compared to the T73 5 suite in Figure 26.
These samples include lavas from the Galapagos Rift (Fornari et al., 1983) volcanic samples
from the Mid-Atlantic Ridge (Hekinian et al., 1997) back arc basin (BAB) samples from the
Southwest Pacific (Nakada et al., 1994), the Lau Basin (Falloon et al., 1992) and the western
Pacific (Bloomer, Smithsonian Institution Volcanic Glass Individual Analysis File, VG no# 9772
9777). Geochemical comparisons have also been made with evolved plutonic samples from
supra-subduction zone ophiolites from California (Beard, 1998), Greece (Tsikouras and
Hatzipanagiotou, 1998; Bebien, 1991), Norway (Pedersen and Malpas, 1984), Newfoundland
(Malpas, 1979), Oregon (Phelps and Ave Lallemant, 1980), Chile (Saunders et al., 1979), Crete
(Koepke, 1986), and Canada (Flagler and Spray, 1991). Four experimental plagiograntic
residual melt compositions were also compared to the T73 5 samples; residual melts from a
gabbro heated to 900 OC and 940 oC (Koepke et al., 2004) the partial melt of a MORB protolith
heated to 955 OC (Dixon-Spulber and Rutherford, 1983) and the partial melt from a hydrous
MORB heated to 950 OC (Berndt, 2002).
SiO2 ValUeS for the T73 5 lavas are generally lower those of plagiogranite compositions at
similar MgO concentrations (Figure 26). SiO2 ValUeS of the evolved T73 5 suite range from 62.0
to 66.9 wt% in samples containing less than 2 wt% MgO compared to plagiogranite samples that
have values that mostly range from 64.5 to 68.13 wt% SiO2. Experimental partial melts of wet
and dry MORB protoliths (~ 9000 9550C) have quite variable SiO2 ValUeS, (60.9 to 65.8 wt%
SiO2) but they all generally display higher MgO then the T735 evolved lavas, with the lowest
MgO value, 1.29 wt%, in the partial melt generated from the dry MORB parent.
In general the T73 5 evolved lavas have lower concentrations of Al203 COmpared to the
maj ority of evolved samples. The T73 5 evolved lavas have a very limited range of Al203 frOm
13.6 to 12.23 wt% compared the global volcanic suite that has values ranging 16.0 to 11.4 wt%,
including some subvolcanic rocks having the highest values at 16.0 and 15.83 wt%. Al203 in the
plutonic suite ranges from 18.1 to 10.4 wt%/. The experimentally derived melts all have
comparatively high concentrations of Al203 Tanging fTOm 20. 1 (gabbro protolith) to 15.41 (dry
MORB protolith) wt% Al203.
The T73 5 evolved lavas are enriched in TiO2 relative to the bulk of the compared evolved
samples. The T735 evolved lavas range from 1.39 to 0.81 wt%/ TiO2 while within the volcanic
suites, the only one with samples of higher value, range from 2.44 to 0.34 wt%. The plutonic
suite displays very low TiO2 ValUeS, which range from 0.84 to 0.30 wt%. Such low TiO2 ValUeS
could be indicative of an island arc signal given that many of these rocks were believed have
been formed in supra-subduction zone environments. The evolved liquid from the MORB
protoliths have TiO2 ValUeS of 1.25 and 1.17 wt%, which lie within the field outlined by the T735
evolved lavas. The gabbroic protoliths both have very low TiO2 ValUeS, 0.36 and 0.12 wt% TiO2.
K20 is extremely enriched in the T73 5 evolved lavas relative to any of the plagiogranite
comparison suites, ranging from 1.02 to 1.30 wt%. Only one sample, a volcanic sample from the
MAR (Hekinian et al., 1997), lies within the field created by the T73 5 samples, with its value
being 1.09 wt% K20.
Dive T735 Samples
During dive T73 5, we documented the presence of two constructive volcanic mounds that
are comprised of massive pillows to blocky, vesicular, silicic lavas. Morphology ranges from
large (>2 m in diameter) pillow tubes, to extremely vesicular (~20% elongate vesicles), blocky
flows, a drastic change in morphology from the small (~lm in diameter), rounded non-vesicular
pillows observed in the surrounding areas where only basalts were recovered.
The petrography and phase chemistry of the evolved samples recovered are indicative of a
complex petrogenetic process. Two populations of phenocrysts were found within the evolved
samples, each exhibiting chemical zoning. Iron-rich pyroxenes, that have core Mg# from 8.24 to
12.6, were reversely zoned while those with Mg-rich cores, ~Mg# of 57.2, exhibit normal
zoning. The rims of both populations approach similar compositions with Mg# ranging from
27.7 to 37.5. Significantly, the rims approach the microphenocryst compositions in the glass
matrix. This petrographic evidence suggests two things;
Mixing between a relatively mafic basalticc) magma and a highly evolved magma, likely
rhyodacitic, each of which crystallized pyroxene before mixing occurred (Barbarin, 1990;
Furman et al., 1995; Perfit et al., 1999; Grove, 2000; Ratajeski et al., 2001). The silicic magma
was relatively dry when it erupted, indicated by the presence of fayalite and Fe-rich pyroxene in
the recovered rocks rather than amphibole (Dixon-Spulber and Rutherford, 1983; Koepke et al.,
2004; Berndt et al., 2005).
Several xenoliths with coarser-grained textures, indicative of slower, crystallization
conditions, are also present in the andesites and dacites. These include rare plagioclase
oikocrysts containing low-Ca pyroxene chadacrysts as well as a coarse-grained xenolith
comprised of plagioclase (An34 An35) and fayalitic olivine, (Fols), present in sample T73 5-G10.
A quartz-plagioclase myrmekitic intergrowth is present in sample T735-Gl9. The presence of
these textural features indicates that that the erupted lava interacted with an evolved plutonic
body or mush zone.
Rounded basalt xenoliths are also present in the evolved lavas. They range in size from <1
mm to several mm in diameter, and have fine-grained to opaque rims. These presence and
texture of these xenoliths basaltic magma intruded/recharged an evolving magma body. The
fresh, mafic material would have been at a much higher temperature then the evolved, viscous
melt it intruded into/mixed with. The high viscosity of the evolved melts would have inhibited
large scale mixing with the fresh mafic material and the temperature difference between the
melts basalticc >1200 oC; dacitic <1000 oC) could explain the fine-grained textures of the
basaltic xenoliths and chilled margins.
Maj or and trace element variations observed in the entire T73 5 suite indicate a bimodal
sample set, that has no samples of intermediate composition (Figure 22; Figure 23) although it
should be noted that most of the basaltic samples recovered around the andesite/dacite domes
have fairly fractionated compositions with many being ferrobasalts (FeO >12 wt.%).
There are a number of different ways that these kinds of rock suites have been
hypothesized to form;
* the extreme fractional crystallization of a MOR magma (Fornari et al., 1983; Perfit et al.,
1983; Juster et al., 1989; Geist et al., 1998; Perfit et al., 1999; Grove, 2000);
* magma mixing between an extremely evolved, possibly rhyodacitic end-member and a
ferrobasalt (Barbarin, 1990; Perfit et al., 1983; Furman et al., 1995; Hekinian et al., 1999;
Perfit et al., 1999; Grove, 2000; Ratajeski et al., 2001);
* partial melting of the hydrated oceanic crust (Dixon et al., 1995; Brandriss et al., 1999;
Berndt et al., 2005);
* assimilation of oceanic crust and assimilation fractional crystallization (AFC) (Bohrson et
al., 1998; Garcia et al., 1998; Gee et al., 1998; Hornle et al., 1998; Grove, 2000).
Silica rich rocks found in oceanic environments may be formed by extreme amounts of
fractional crystallization (Perfit et al., 1983, Juster et al.., 1989). The Galapagos are a prime
example of an area where the evolved lavas recovered fit the extreme fractional crystallization
model well. Juster et al., 1989, used experimentally determined phase boundaries to model the
fractionation of lavas recovered at 850 W. The results of the experiments determined that the
range of compositions found there were the result of shallow-level differentiation processes and
the higher levels of fractionation and silica enrichment were due to the higher fO2, which allowed
for a titanomagnetite bearing assemblage to begin to crystallize earlier, driving the
concentrations of SiO2 up and FeO and TiO2 down.
To test a fractional crystallization model for the Cleft suite, liquid lines of decent (LLD)
(Appendix 4) were calculated using the program Petrolog (Danyushevsky, 2001) under various
starting conditions (e.g. pressure, oxygen fugacity, different mineral assemblages). Sample
T73 5-G3 5 was chosen as the composition of the parent melt because it is the most mafic sample
recovered in the area were the evolved samples were recovered. The mineral-melt models of
Danyushevsky (1999) were chosen for calculating the compositions of olivine, plagioclase and
clinopyroxene; Ariskin (1993) for orthopyroxene and Ariskin and Barmina (1999) were used for
magnetite compositions. Melt oxidation states along the QFM buffer were determined using the
conditions of Borisov and Shapkin, 1990. Calculations were run at 200 bars and 1 kb of pressure,
in order to simulate the pressure at the seafloor and within the shallow crust (~3 km). All
calculations were executed assuming perfect fractional crystallization in 0.1 wt% incremental
crystallization steps and were typically terminated ~88% total crystallization when certain
components were expended. Using the 200 bar model, which started at 1185 oC and was run at
QFM, the calculated maj or element trends (Figure 27) are fairly smooth with a maj or inflection
point seen in SiO2, Al203, TiO2, and FeO at an MgO of ~3 wt% when titanomagnetite enters as a
crystallizing phase, after ~65% crystallization of the parental melt. Oxidation state affects the
temperature and composition at which titanomagnetite (and ilmenite) crystallizes; higher oxygen
fugacity causes earlier (higher T, less total crystallization) stabilization of oxide phases.
Consequently, the modeled inflection point occurs earlier in the chemical evolution when using
an oxygen buffer higher than QFM. Figure 28 shows the relationship between temperature and
percentages of mineral phases in the crystallizing assemblage from a successful model.
The modeled order of crystallization along a QFM buffer at 200 bars is olivine -olivine +
clinopyroxene -olivine + plagioclase + clinopyroxene -plagioclase + clinopyroxene +
orthopyroxene -plagioclase + clinopyroxene + orthopyroxene + magnetite (Figure 28). The
modeled LLD at QFM corresponds well to the interpreted crystallization order based on
petrographic examination of sample thin sections. The order of crystallization is olivine +
plagioclase -plagioclase + clinopyroxene 1 olivine -plagioclase + clinopyroxene + FeTi
oxides & olivine -plagioclase + pigeonite (or orthopyroxene) + FeTi oxides.
Dacite compositions are predicted to form after approximately 80% crystallization, even
though the phase equilibria used in the Petrolog program are not very well constrained in such
evolved compositions. The results of the models generally agree with the maj or element trends
observed in the dacites. SiO2 ValUeS in the evolved rocks are slightly higher than model
predictions and K20 is over-enriched, while TiO2, Al203 and P20s are lower relative to the
calculated abundances (Figure 27).
Variations in the trace element abundances were modeled for selected elements using the
Raleigh fractionation equation, phase proportions predicted by the maj or element models and
published trace element Kd's (Bougault and Hekinian, 1974; Shimizu and Kushiro, 1975; Duke,
1976; Matsui et al., 1977; Mysen, 1978; Kravuchuk, 1981; Villemant et al., 1981; Colson et al.,
1988; Kloeck and Palme, 1988; Agee, 1990; McKenzie and O'Nions, 1991; Keleman and Dunn,
1992; Hart and Dunn, 1993; Hauri et al., 1994; Nikogosian and Sobolev, 1997; Bindeman et al.,
1998) and also by assuming Kd's of zero for the most incompatible elements. Calculated trends
for several of the incompatible trace elements also seem to follow calculated fractional
crystallization trends (Figure 29) although in order to reach the concentration levels of the most
evolved compositions greater than ~90% crystallization is required. The enrichments observed
in some of the most incompatible trace elements could not be modeled even if the bulk
distribution coefficient was assumed to be zero.
Trace element variations generated with published Kd's are only slightly below the actual
trends of the lavas recovered (Figure 29). Zr, Y and Sm concentrations are well-modeled using
the Rayleigh fractionation equation with a D of zero and the proportions of minerals determined
from the Petrolog models. The La, K and Rb values of the evolved lavas, on the other hand, are
significantly higher than the calculated abundances even using an assumed D of zero, which
would give the maximum enrichment. Modeled variations for Cr, Ni and V compare favorably to
the actual evolved compositions, but the Cr and Ni concentrations of some intermediate to
evolved basalt samples are lower than those in the calculated model. Kd's for Ni and Cr are not
well-constrained for basaltic systems though. Sr and Sc are significantly lower in the evolved
samples when compared to the modeled concentrations.
In general, fractional crystallization seems to adequately (though not entirely) explain the
elemental trends, although models suggest extreme amounts of crystallization (>80%) of an
already evolved basalt parent in order to duplicate the maj or element compositions of the dacites.
With such extreme amounts of crystallization needed in order to model the evolved
compositions, there would be a problem with a crystal-laden melt being able to erupt, as the melt
has progressed well past the solidification front predicted at 40% crystallization (Marsh, 2000).
Petrographic evidence supports magma mixing was involved in the petrogenesis of the
evolved suite, and might help to explain some of the discrepancies in the elemental trends as well
as the compositions of some basaltic andesites recovered along the axis of the Cleft segment
north of this study area. Clearly the bimodal pyroxene compositions and extensive amounts of
zoning, as well as chilled basaltic xenoliths observed in all of the evolved samples, point to some
type of magma mixing event or events. The phase chemical variations suggest there was mixing
between a basaltic end-member, possibly as evolved as a ferrobasalt, and a very evolved magma,
possibly as evolved as a rhyodacite or rhyolite, to create the andesitic and dacitic lava samples
recovered. There is no evidence of evolved magma mixing with mafic basaltic liquids.
In order to evaluate the potential role of magma mixing, mass balance mixing calculations
were performed between likely mixing end-member compositions. As the most evolved silicic
end-member(s) is unknown, all mixing lines were calculated using the composition of dacite
sample T73 5-Gl2 (Mg# of 6. 14), the most evolved lava recovered. The three basaltic end-
members chosen were T73 5-G3 5 (Mg# of 37.5, MgO = 7.55 wt%), the most mafic basalt
recovered during the dive, T73 5-G7 (Mg# of 32.3, MgO = 6.79 wt%), a moderately evolved
basalt and T73 5-G32 (Mg# of 29.8, MgO = 6.29 wt%), the most evolved basalt recovered on the
dive. Increments of mixing were calculated for every 10% portion (i.e. 10% of A and 90% of B,
20% of A and 80% of B, etc.) using a standard mass balance equation (Langmuir et al., 1978).
The calculated mixing lines (Figure 30) represent a few of the potential mixing models that are
Results of the mixing calculations show that the compositions of some of the outlying
samples on the major element plots that did not fit the fractional crystallization models are better
explained by magma mixing. In part, the group of 2 basalts, 8 ferrobasalts and one andesite from
the entire S. Cleft suite that show a trend of decreasing FeO concentrations with decreasing MgO
fall along the mixing line from T73 5-G7 to T73 5-Gl2. Mixing lines calculated using the two
more mafic basalt samples as end-members don't seem to include any outlying basaltic and
andesitic samples in their trends. It could be interpreted that, if mixing was involved in creating
these intermediate outliers, and the evolved compositions they trend towards, that one of the end-
members involved in the mixing would have to be a fairly evolved ferrobasalt or FeTi-basalt,
since the most evolved sample recovered provides the best end-member to explain elemental
trends in the full dataset.
The mixing calculations suggest a mix of 30 40% of the ferrobasalt (T73 5-G32) with 60
- 70% of the dacite (T73 5-Gl2) would be required in order to create the observed evolved
compositions in the T735 suite. The moderately evolved andesites observed in the larger Cleft
dataset would require only about 30% of the dacitic end-member to be mixed with the
ferrobasaltic liquid. The petrology of the samples, and the presence of the myrmekite, zircons,
fayalite as well as sodic plagioclase (Anls) and iron-rich pyroxene (Mg# of 8.24) crystals, as
well as the fine magnesian rims surround Fe-rich pyroxene (Figures 11 and 12) and fayalite
(Figure 15) cores, strongly suggest that there is the possibility of an even more evolved end-
member then T73 5-Gl2 was involved that was mixed back to create the compositions recovered.
Based upon the traj ectory of the mixing lines drawn between the evolved samples and possible
basaltic end-members, the proj ected evolved end member might be expected to have a
composition similar to this: ~67 wt% SiO2, ~12 wt% Al203, 0.75 wt% TiO2, 7.5 wt% FeOT, ~3.0
wt% CaO, ~5.2 wt% Na20 and K20 ~1.4 wt%. All these values were estimated by extrapolating
the current mixing model to an MgO of almost zero.
The trace element compositions also seem to be more supportive of a mixing model, as
crystal fractionation cannot account for over-enrichments seen in highly incompatible elements
such as La, Sm, Zr, and Y where a Kd of zero is needed to approximate the observed values in
the most fractionated lavas (Figure 29). This would however, require that the evolved end-
member was formed by extreme amounts of fractional crystallization (>90%).
Partial Melting/Assimilation Fractional Crystallization (AFC)
Another hypothesis for generating highly evolved melts is to partially melt the basaltic
crust (Petford et al., 2001; Castillo et al., 2002; Coogan et al., 2003) or assimilate country rock
into the melt (Nicholson et al., 1991; Bohrson et al., 1998; Garcia et al., 1998; Gee, 1998;
Hoernle, 1998; O'Hara et al., 1998; Weis et al., 1998; Grove, 2000).
Several studies have shown it possible to produce silicic melts from partially melting
hydrous mafic protoliths (Holloway and Burnham, 1972; Helz, 1973; Beard and Lofgren, 1991;
Kawamoto, 1996, Koepke et al., 2003; Koepke et al., 2004). There is also evidence of anatexis in
plagiogranitic rocks found in ophiolite sequences (Malpas, 1979; Pederson and Malpas, 1984;
Flagler and Spray, 1991) as well young oceanic crst (Mevel, 1988) although the exact
compositions of the protoliths in these studies has not been well constrained and may include
gabbros and sheeted dikes that may have been altered due to hydrothermal activity. Water is
usually assumed to be a component of the melting process due to the presence of amphibole
found in samples of felsic oceanic crust (Bebien, 1991; Beard, 1998; Tsikouras and
Hatzipanagiotou, 1998; Koepke et al., 2002), often in poikilitic textures, suggesting a magmatic
origin (Koepke, 1986).
Phase chemical data from melting experiments (Koepke,. 2004) also demonstrated an
increase in olivine Fo content in the restite due to increasing temperature and the influence of
water. Residual plagioclase were also more An-rich in the experiments due to the effect of water
(Sisson and Grove, 1993; Berndt, 2002) and amphibole was present in all systems at
temperatures <~-980 oC.
Assimilation of crustal material has also been hypothesized as a method to create evolved
compositions. Assimilation in a MOR regime most likely involves oceanic crust that has been
hydrothermally altered and this has been substantiated by direct field observations of xenolithic
basaltic material in ophiolites (Castillo, 2002). The addition of altered material into a MORB
magma results in an over-enrichment of chlorine, relative to other incompatible elements such as
K20 and TiO2 (Jambon et al., 1985; Michael and Schilling, 1989; Michael and Cornell, 1998;
Castillo, 2002; Coogan, 2003).
When compared with the phase chemistry of the resulting melts from the Koepke, 2004,
experiments, the phase chemistries of evolved lavas from dive T73 5 are conspicuously free of
any signal of water activity. No evidence of amphibole is present in the T73 5 lavas, nor is there
evidence that amphibole was present in the residue of melting due to the absence of a LREE to
HREE enrichment in the REE pattern of the T73 5 glasses, as the Kd values for REE in
amphiboles decrease in value for the lighter elements (McKenzie et al., 1991). The phase
compositions of the T73 5 lavas and the absence of evidence of hydrous phases would seem to
preclude the notion of the silicic end-member being the result of a partial melt of hydrous
oceanic crust. While small batches of xenolith material are present in the T73 5 evolved lavas
none contain amphibole. In addition, there is no elevated chlorine signal (Cl/K ratios remain
fairly constant; Figure 31), precluding large amounts of assimilation of altered crust taking place
in the petrogenesis of the evolved lavas.
Strontium and oxygen isotope values for the evolved T735 are identical to those in MORB
from the Cleft segment, consistent with a lack of any seawater alteration or contamination
(Perfit, personal communication). If high-temperature altered crust were melted to form dacitic
partial melts, the values of sSr/86Sr would be higher than fresh basalts and oxygen isotopic
values would be lower. In both cases this is not true instead values well within the range of JdFR
fresh, unaltered lavas.
62 x + Volcanic
60~ m Putonic
m T735 samples
0 1 2 3 4 5 6 7 8
Comparison of TiO2 Values
I ~ I T735 evolved lavas
0 0.5 1 1 .5 2 2.5 3 3.5
Com prison of SiO2 ValUsS
Figure 26: Comparing maj or element variations in the T73 5 lavas to other evolved suites. Other
suites are from varied locations and petrogenetic origins, including lavas from the Galapagos Rift
(Fornari et al., 1983) and Mid-Atlantic Ridge (Hekinian et al., 1997), back arc basin (BAB)
samples from the Southwest Pacific (Nakada et al., 1994), the Lau Basin (Falloon et al., 1992)
and the western Pacific (Bloomer, Smithsonian Institution Volcanic Glass Individual Analysis
File, VG no# 9772 9777), plutonic samples from supra-subduction zone ophiolites in California
(Beard, 1998), Greece (Tsikouras and Hatzipanagiotou, 1998; Bebien, 1991), Norway (Pedersen
and Malpas, 1984), Newfoundland (Malpas, 1979), Oregon (Phelps and Ave Lallemant, 1980),
Chile (Saunders et al., 1979), Crete (Koepke, 1986), and Canada (Flagler and Spray, 1991). The
experimental plagiograntic residual melt compositions compared were from a gabbro taken to
900 oC and 940 oC (Koepke et al., 2004) the partial melt of a MORB protolith taken to 955 OC
(Dixon-Spulber and Rutherford, 1983) and the partial melt from a hydrous MORB heated to 950
oC (Berndt, 2002). SiO2 values were plotted with the recovered T73 5 basalt as well as the
evolved samples in order to demonstrate the amount of differentiation.
Comparison of K20 values
O0.81 m Plutonic
0.6 m1 m &perimental
m T735 evohred lavas
0 0.5 1 1 .5 2 2.5 3 3.5
S0.5 1 1.5 2 2.5 3 35
Figure 26: Continued.
Comparison of~ AIO, valus
* T735-evolved lave
D 2 4 6
8 10 12
Figure 27: Maj or element liquid lines of descent. Liquid lines of decent are modeled using
Petrolog (Danyushevsky, 2001) using a QFM buffer at 200 bars of pressure and were run to
~88% crystallization. Sample T73 5-G3 5 was used as the starting composition, as it was the most
mafic sample recovered during the dive. Sio2 ValUeS in the evolved rocks are slightly higher than
model predictions and K20 is over-enriched, while TiO2, Al203 and P20s are lower relative to
the calculated abundances.
0 2 4 6 8 10 12
a 2 4 6 8 10 12
D 2 4 B 8 10 12
a Srnith, 1994
+ Eive T735
x Petr~log LLD
0 2 4 6 8 10 12
Figure 27: Continued.
O + : +;
BOO 850 900 950 1000 1050 1100 1150 1200
Figure 28: Cumulative percentage of phases plotted against melt temperature (oC). This figure
illustrates the changing mineral assemblage as the modeled parent composition cools. Model
assumes QFM and 200 bars of pressure.
Figure 29: Trace element liquid lines of descent. Liquid lines of decent are modeled using
calculated Kd's and the results of a Petrolog model run using a QFM buffer at 200 bars of
pressure, run to ~88% crystallization. Sample T735-G35 was used as the starting composition, as
it was the most mafic sample recovered during the dive. While most trace elements follow the
calculated trends, La and Lu show over-enrichments, even when LLD's are calculated using a
Kd of zero. Zr and Sm also show slight enrichments relative to the calculated Kd' s, but can be
predicted using a D of zero.
Cl 2 46 8 10 12
8 10 12
D 2 4 6
B ID 12
2 6 g 10 12
SSmith, 1994; Perit, nulse
* Eiv T735
El Calculated Kd's
D 2 + 9
2 6 8 10 12
Figure 29: Cotinued.
o Sm ith, 1994
+- Stakes, 2006
* Dive T735
x G35to 012
7 G32 to 012
= G7 inG1 2
0 24 6
8 10 12
0 2 46
8 10 12
0 .5 i
OM O M O ~
Figure 30: Mxing modelscaluae sigasadadms alneeutin(aguita.
197).Figre26:Miinglies er cacuate t T735--Gl2, th ---e most evolve end-membe
reoerdduigthe die h he ai ndmmescoe eeT75G5(lc ') h
EPR 9 10N (le Rour at al.)
SS. Cleft (StakeE, eat .)
.. ....... .. .......... .. T7 35 R TI
8 GSC BSiW Perrit etal.)
0 2 4 6 8 10~ 1;
* EPR 9-10N (le Roux atal.)
S S. Cleft (Stakes 81 l.)
+ 735 RTI
O SC Bli W (Fedit aet al.}
S2 4 6
Figure 3 1: Comparison of T73 5 Cl~ and Cl/K ratios versus MgO. The Cl/K ratio of entire T73 5
suite remains fairly constant over the full range of MgO. While the Cl- concentration increases
with decreasing MgO the ratio of Cl-/K in the T735 samples (green diamonds) does not show the
same rate of increase seen in the Southern Cleft samples (Stakes et al., 2006), the GSC (Perfit et
al., 1999) and the EPR (le Roux et al., 2006) where assimilation takes on a much larger role.
-e- ----a-- -- -- ---- -- -- -- --- -
* Low pressure fractional crystallization models adequately reproduce the observed trends in
major elements and some of the minor/trace elements. There are slight discrepancies
between the predicted and observed compositions though, most notably with K20 being
very over enriched, while TiO2, Al203 and P20s are lower relative to the predicted
abundances (Figure 20). Calculated trace element LLDs for La, Rb and K, even assuming
Kd's of zero, do not fully predict the concentrations seen in the evolved samples, which are
over-enriched to models values.
* The calculated models require fractional crystallization of greater than 80% in order to
duplicate the evolved lava compositions recovered on Dive T73 5; well past the amount of
crystallization required to form an impermeable solidification front predicted by Marsh,
* Petrographic evidence strongly supports mixing as a significant petrogenetic process that
created the evolved lavas. Chilled basaltic xenoliths as well as disequilibrium crystals are
prolific in all the evolved samples, as well as complex normal and reverse zoning seen in
pyroxene and plagioclase crystals. The petrography and chemistry suggest that that a
possible recharge event between and evolved, rhyodacitic melt and a ferrobasalt created
the andesitic and dacitic lavas recovered.
* Large amounts of assimilation or partial melting of altered oceanic crust do not seem to be
involved in the petrogenesis of these evolved lavas. Signals expected from the
assimilation/incorporation of hydrothermally altered oceanic crust, such as over
enrichment of C1F, due to the interaction of seawater with the crust (Coogan2 et al. 1, 2003),
or a sloped REE pattern, due to the lower Kd' s of the LREE in amphiboles causing an
enrichment of them in the melt (M~cKenzie et al., 1991), are not seen in the T73 5 evolved
The evolved lavas recovered from dive T73 5, at the southern terminus of the JdFR, have
had an extremely complex petrogenesis. While large scale amounts of assimilation and/or partial
melting of hydrous crust can be ruled out as methods of creating this particular suite of rocks
magma mixing between two distinct magma types that formed by different amounts of fractional
crystallization is certain. The ferrobasaltic xenoliths provide a good estimate as to what the more
mafic end-member composition might be, but the extremely evolved end-member is much more
difficult to constrain.
These highly evolved melts would have been extremely viscous and had very limited
mobility. Due to the fact they the lavas were located at the RTI, propagating dikes, composed of
much hotter, ferrobasaltic material, from the larger magma bodies found up-ridge could have
provided the necessary heat to mobilize the dacitic melts, allowing them to exsolve H20 and CO2
through decompression, lowering the density enough to allow them to erupt on the ocean floor.
T73 5 DIVE LOGS
Dive objectives: Explore RTI near southern Cleft
13:31 Om rov entered water, start dive T735
14:51 2192m on bottom for past 2 minutes, talus slope
14:55 setting down to sample, pillow fragments, some microbial growth on fragments
14:57 2190m sample T73 5-Gl, from pile of pillow debris, placed in S3 in rov drawer, slightly
15:00 2190 looking around, some possibly intset pilows present
15:01 2187m starting traverse along curved ridge
15:03 2187m pillow talus, crossing dome summit
15:06 2187m can see glassy remains on many of the fragments
15:08 2193m larger pillows here, no convincing in-place pillows
15:11 2200m possible in-place mound, looking for place to sample
15:12 2202m small pillow fragment, T735-G2, brownish, placed in S3
15:13 see flows pointing downhill, pillows appear vesicular, small hornito in area, turning
around to look at hornito
15:16 2201m spatter mound-tube thing, or vapor escape tube??,
15:19 trying to grab sample, fractured in place, highly altered, glassy
15:22 2202m grab samplefrom tube thing, T73 5-G3, placed in S3, broken off from top of tube
15:23 underway again, some flows appear inplace, heading 341
15:24 2205m crossing partially covered fissure by pillows
15:25 contact to almost all intact flows
15:26 2204m intact pillow flows, slight sed covered, flows appear to be draping topography,
15:27 2202m collapse features, mod sed cover in interstices of flows
15:28 2203m stop to sample
15:31 2205m looking for piece to pickup
15:33 2205m collected sample of crust, T735-G4, in S3 again thin piece from lobate flow
15:36 2201m stepping down to west going to look at rigde in sonar, crossing faultsteps,
15:38 2201m fault scarp, razorback ridge series separate by crevasses slightly wider than the
rov, flows on top are intact, heading 336, pie shape wedge of crust between fissures, flattened
lobates, this is the wall from the sonar
15:42 2200m fissures seemed to j oin, its much wider here, about as deep as it is wide, fissure
wall has clean surface, appears pulled apart
15:43 2199m two Hissure walls are closing in according to the sonar
15:51 2220m stopping for sample from lobates
15:54 2221m smal pillow wedge collected, sample T735-G5, also in S3
15:57 2217m another fissure, buried by overlying pillows, hard to tell if flowing in or out
15:59 2215m another fissure, pillows neatly broken along edge
16:03 2216m collect sample of pillow crust, intact?, T73 5-G6, placed in S3, blocky reddish
16:05 2214m drainback, area of pillows, mod sed between pillows, very plastic flows covering
16:06 2210m crack fissure starting here, pillows are knobby, older terrain appeared to be large
16:09 2215m still intact flows, fissure off to port
16:13 2212m stopping for sample, different looking stuff, more blocky
16: 18 2213m collected grab sample T73 5-G7, from flat area at near top of knoll, placed in front
16:20 2209m intact pillows, mod sed cover between
16:22 2207m appears more constructional, fissure to west side, sharp drop, bottom a few meters
down, approx. 5m wide
16:23 2207m dropping down over edge of fissure, top of wall is intact, base has lots of debris,
appears sed covered even on talus and pillows, appears old
16:25 2214m thick sed cover, few pillows sticking up from sed
16:27 2216m mixture of intact pillows and broken pilow debris
16:34 2225m 50-60% sed cover, good sed cover on pillows too
16:36 2231Im more pillows, tubular, somewhat smaller than previous, stopping for sample
16:39 2232m sample T73 5-G8, blocky pillow frag, placed in S4
16:41 2229m scattered pillows, <50%s sed cover
16:42 2231Im tube and sheet flow down to SW, stopping to sample
16:44 2232m sampling sheet, too friable need to put in biotube, must be glassy
16:45 2232m sample sheet flow?, T73 5-G9, fragments placed in BT5
16:48 2227m pillow tubes appear different, very large well formed tubes, striations
16:50 2226m stopped to sample
16:52 2226m collected sample T73 5-G1 1 from large pillows, very glassy, placed in S4
16:54 2219m very well formed pillows flowing downslope, very little sed on top, very glassy
16:58 2212m, vapor pockets and cavities between layers of basalt
16:59 2211Im, picking sample of vapor pocket, T73 5-Gl2
17:02- 2213m, sample looked glassy, still passing over breadcrust texture, sed coating on all
17:03 2211m, lots of local relief, big pillows surrounding by smoother tubes
17:06 2211m, twrining on to next line, sheetier appearance
17:07 2209m, large flat broken up sheet flow, perhaps a silicic constructional dome
17:08 2210m, trying to sample at top of constructional feature, slabby flow
17:09 2210m, collected sample of striated top of slabby flow, in BT4, T73 5-Gl3
17: 12 2211m, was broken from beneath, top is flat but sides are tilted
17:13 2214m, still same flow, very straited, blocky iand angular n some places, plus few big
17:16 lost my comment
17:18 2214m, turned to the left, and are back in the straited, sheety thick flow
17:19 2214m, collecting another piece of striated flow
17:20 2214m, sample of sheet flow, won't fit into bio tube
17:22 2214m, sample was collected in between the P/S boxes, T73 5 Gl4, another piece
collected at same time
17:24 2213m, finished collecting another sample of what is hopefully a dacite flow
17:26 2214m, nav jumpy because we're tunring, lots of pillows
17:27 2216m, large broken pillows, tubular
17:28 2209m, between the tubes and pillows still getting massive flows (layered)
17:30 2209m, pillow texture is smoother
17:3 1 2212m, stopping to collect another sample, lots of gas cavities
17:32 2212m, sample going into biotube 3, T735 Gl5
17:34 2211Im, back on the line, hdg 62, this area had lots of vesicles, and dome-like feature had
both angular blocky rocks and pillows
17:38 2213m, tilted slab block, probably from eruption
17:40 2219m, still in the same general area of geology and morphology, still looking at dome-
like structure, probably not basalt
17:44 2215m, going to sample in this area
17:45 2216m, sample will be taken from the interior, more broken
17:46 2216m, sample T735-Gl6, from interior of flow, S5, should be smallest piece in S5
17:50 2215m, coming down slope of construct, hdg back to line
17:52 2216m, still at pillows with breadcrust texture, moderate sed cover
17:53 2218m, very evolved rocks, broken up large pillow, no tectonic fissures on sonar
17:57 octopod, red Dumbo
17:58 2216m, large striated pillow, broken down the middle
18:00- 2216m, heading back to line, irregualr terrain
18:01- 2216m, irregular terrain, with broken up blocky lava
18:04- 2221Im, collect a sample of broken pillow, to see if it's the same as the previous samples
or it it's less viscous
18:06 2221m, taking a pillow sample
18:06 2221m, S1 pillow sample with lots of vesicles, T735-Gl7
18:09 2221m, at bottom of saddle, will soon be heading upslope
18:10 2214m, coming up constructional slope, pillows flowing downhill radially
18:12 2207m, attempting to sample the crust of these pillows, looking for small piece
18:14 2205m, continued to move upslope looking for a breakable pillow crust
18:16 2206m, collected two pieces of pillow crust, small, in S1, T73 5-Gl8,
18:17 2192m, moving upslope, smaller pillows, reached a bench
18:19 2199m, moving up feature, more relief than dacite mound
18:21 2198m, tublar, flattened pillows, probably at top
18:22 2198m, much less blocky than last dome, collecting sample
18:25 2198m, picking up sample previously dropped, near top of slope, PS, T73 5-Gl9
18:26 2198m, this may have been sample site, nav jump
18:28 2199m, tubes flowing off to right, south
18:29 2201m, crabbing downslope, very little tectonic features
18:31 2208m, well-formed pillows and tubes, heading downslope to south
18:33 2205m, stepping downslope, mod sediment on pillows
18:34 2206m, elongate tubes look like they're flowing south
18:36 2199m, tubes and broken pillows, pillows are more fragmented
18:40 2203m, still see some blocky, angular flows, on shallow slope
18:41 2204m, slow progress on hdg because of current
18:46 2216m, heading downslope, flattened tubular pillows
18:49 2216m, still elongate tubes
18:50 2234m, looking at in place pillows, attempting to sample, also smaller rubble
18:52 223 5m, collected rock sample from talus near in place pillows
18:35 -2234m, T735- G20, P5
18:57 2258m flow fron talus, large bocky boulders
19:01 2275m collect large fragment of intact pillow, T73 5-G21, from base of steep slope, slope
continues down to east, skinny pillows
19:07 2263m heavy sed cover in gully between two knolls, isolated exposed pillows
19:08 2258m heading 93 degrees
19: 10 2256m much less sed cover, pillows sticking out of mod sed cover
19:13 2254m pillows with heavy sed cover
19:14 2252m collapse feature showing sheet flows beneath pillows
19: 16 2256m collect sheet flow sample T73 5-G22 from collapse feature
19: 18 2251m large pillow with inner drain features, pillows more bulbous, mod sed
19:21 2249m picking of fragment of pillow from top of sed, too big!
19:24 2249m looking for sample to collect, pillows in thick sed cover
19:26 2250m collect large pillow fragment from top of sed, T73 5-G23, put in P5
19:28 2247m constructional pillows in sed cover
19:29 2246m veered off course to NE, heading back S to top of knoll
19:31 2249m closer pillows, everything appears intact, possibly big hornito, all pillows flow
downhill from top of mound
19:37 2249m collect sample T735-G24 in S2, greenish-white looking inside
19:41 2250m collect basalt sample from same area, T735-G25
19:44 2248m heading down mound to north, bulbous pillows poking up through sed
19:45 2249m Hield of unsedimented pillow surfaces, slightly more gentle slope than the west
19:49 2252m mostly jumbled rubble, some intact pillows heading downslope to saddle between
knolls, 50% sed cover with equal amounts of pillow rubble
19:53 2254m crossing Eissure almost N-S several meters deep, broken pillow pieces, appears
tectonic, sed on walls and at base, all pillows exposed in fissure wall
19:56 2260m following same fissure to north along trackline, starting to shallow out
20:03 2266m collect grab sample from more massive unit, T735-G26, in S5
20:05 2265m massive flow overlain by pillow, lowest most exposed unit, well fissured with
pillows on top, pillows appears to possibly flow inside indicating pillows post-date fissure, N-S
20:07 2264m crossing massive flow, large fissure also to righ of ROV (east), possibly moving
into the transition from hooked ridge to normal ridge morphology
2008 2265m following fissure, series of fissures N-S cutting massive flow, columnar talus in
20:09 2265m some evidence of uplift along this section of fissure
20:13 2258m very heavy sed cover
20:16 2265m more pillows, less sed
20:22 2271m thick sed cover, large pilows sticking up
20:24 stopped for sample, cow patty looking pillow
20:26 2271m sample from cowpatty pillow, crust, T735-G27, in P2
20:31 2274m drifting to port towards large depression, some broken pillows with spill-outs,
20:32 2275m fissure heading N-S, shows well on sonar, looks deep, tectonic
20:35 2268m, mound in depression near fisure, constructional
20:38 2275m small hornito, drippy flows, ~lm relief on structure
20:39 2271m appears like old fissure, sed filled, some truncated pillows???, narrow
20:42 2270m crossing fissure, just past linear series of small constructs, ~lm in relief, dropped
down crossing fissure to west
20:45 2272m shallow fissure, sed covering talus, orientation ~340
20:46 2271m another fissure, slight downdrop to east
20:48 2269m possible emuptive fissure, N-S, east side dropped down
20:50 2269m sample from pillow along fissure, T735-G28, pie shaped, in P2
20:57 2268m, hdg to east, 100% sediment cover
20:58 2267m, crossed a fissure trending N-S
20:59 2267m, still in region of pillow and heavy sediment
21:01 2270m, sparse pillow/lobate outcrops with heavy sed between
21:02 lost comment
21:05 2272m, moved over small fissure trace, pillows
21:07 2274m, typical deep sea biology
21:09 2268m, many more outcrops right now, small ridge of broken pillows
21:11 2264m, many more broken pillow outcrops
21:12 2259m, slope is shallowing to the north
21:15 2261m, on edge of slope with pillows outcropping, slope steepens to S
21:17 2261m, on the edge of large escarpment
21:18 2263m, fissure runs N-S, regional fissure system
2: 120 2266m, lots of broken pillows exposed here, much less sediment
21:22 2264m, back to isloated pillows and heavy sediment, flat
21:24 2259m, climbing toward top of dome, pillows loo tectonized, but broken in place
21:25 2258m, shallow depression filled in partially
21:27 2260m, at 2257m, maybe reached the summit of this feature, as bathy dropped off
21:29 2254m, pillow tubes, many going downslope to N(?)
21:30 2253m, crossing over smaller N-S fissure, can see both sides, sedimented on inside
21:32 2249m, continuing to move upward, most pillow tubes in place, plus rubble
21:33 2248m, in place pillows and a small rubble-filled fissure
21:35 2249m, collecting pieces of pillow crust near top of dome
21:36 2246m, T73 5-G29 will be in P3, looks old and oxidized
21:39 2242m, at top of mound, several fissures dissecting the top, running NS
21:42 2246m, running along a small N-S fissure, many on sonar, hdg is now 345ish
21:43 2251m, inside fissure, blocky and talus inside fissure, 20-30m wide
21:45 2250m, flying along the fissure, walls on both sides, small fissures wihtin main wall
21:49 2257m, irregular, tectonized terrain, lots of rubble, some intact pillows on edge of small
21:50 2264m, mod sediment covered, more fractured terrain, filled in fissures
21:53 2270m, vehicle depth has been consistently 10m shallow than bathy
21:54 collection a sample for biotube 2, T73 5-G30, hopefully at intersection bewteen regional
fabric and local valley fabric
21:57 2272m, dropped first sample, looking for another piece of pillow crust
21:58 got another sample T73 5-G30, triangular piece of pillow crust, into biottube2
22:01 2272m, back to heavy sediment, flat, small isolated outcrops
22:03- 2270m, 40m to E is a fissure, flat-lying, 100% sed cover
22:05 2272m, 100% sed cover, isolated pillow outcrop
22:07 2276m, more of the same
22:09 2285m, more exposure of small bulbous pillows, still lots of sed
22:11 2293m, continuing to get deeper, still heavily sed
22: 17 2294 heading to the ridge parallel wall, sediments 100%
22:19 2294 sed. continue, tape change, type 8
22:24 2295 pilow ridges, about to get sample
22:24 same depth, sample T73 5-G3 1 taken from the pillow ridge
22:29 2296, sed. again, some pillow ridges again
22:29 2296, sed. again, some pillow ridges again
22:34 2296 still 90% sed. some flow tos broken
22:36 2290 fissure 337 orientation, talus on the sides on the fissure,
22:39 2287, more pillow terrane, heavily fissured
22:41 2289, still fissures, heavily sed. between issues
22:43 2280, compl. tectonized big pile of debrii pillows
22:46 2291 old fault (fault sliver) heavily tectonised
22:47 2290 approaching the opposite wall, covered with talus, another fault sliver
22:49 2287 wedging out fault sliver, two intersecting faults
22:51 2294 still old tectonized terrane
22:53 2297 materail changes, more collapsed pillows
22:54 2298 about to get sample
22:56 2298 sample G32 from the base of the wall
22:59 2296 ROV loosing main comp. going toward the wall
23:01 2291 base of the wall flat sheets, tectonised ridge
23:03 2291 small pillows with buds on the side base of the wall
23:04 2294 moderate sed. cover, lobates, pillows with buds, still intact, gradually going uphill
23:05 2293, fissure parallel to the wall
23:07 2293 first ridge parellel tectonised zone
23:08 2290 into the real fault, big step up
23:10 2288 intact collapse pit, not consistent wall
23:12 2281 change headings to 300 up the fault
23:14 2276 intact pillows and lobate flows downhill from us,
23:15 2274 remarkably little fractionation on pillows, constractional part of the wall?
23:17 2269 tectonised area
23:19 2263 heavily tectonized zone, small fragments nothing intact
23:20 2260 coming to relative. flat artea, sedimented
23:23 2257 continue flat area, covered with sheety flow debri
23:25 2257 sample from the sheety glassy flow -G33
23:26 2257 sample G33 very altered glassy, taken another piece
23:31 2248 intact flows and lobates, going down parallel to the wall
23:33 2244 half-way up the wall, flows align down
23:34 2242 sedim. area, covered with thin sheeted flows, contiune up the wall
23:36 2239 approaching steep wall..
23:3 8 2239 escarpment pillow debrii, series of fault stpes, broken sheets
23:40 2233 moving into area of intact lobate flows
23:40 2232 still intact flow
23:41 2225 alteranting talus and flows
23:42 2226 samples from the flow sheet and lobate flows
23:45 2228 sample G34 from lobate flow
23:45 2228 second sample G35 from the same area
23:48 2230 near the top of the wall coherent flows
23:50 2231 sample G36
23:53 2231 G37 and G 38 samples
23:55 2231 G 39 sample
23:59 ROV is coming up
01:26 Om end dive T735
PHASE CHEMISTRY FOR T73 5 LAVAS