1
Appendix A. Dive Summaries
2
Dive 3963: Eleven samples were obtained during a west to east traverse of the eastern flank of
3
the ridge crest (Fig. 1, 2, Table 1) [Schouten et al., 2004]. Dive 3963 began ~2.2 km to the east
4
of the axial summit trough in a flat location of older-appearing ropey sheet flows (samples 3963-
5
1 and 3963-2). These flows transitioned to lobate flows mixed with flat and ropey sheet flows
6
covered with a similarly thin veneer of sediment. A change in morphology to pillow lavas was
7
accompanied by a change in slope, presumed to be a pillow flow front (sample 3963-3). This
8
morphological sequence was repeated up to a second flow front, from which pillow lava samples
9
3963-4 and 3963-5 were collected. Sample 3963-6 is a sheet flow from the lobate/sheet terrain at
10
the base of this second pillow flow front. The top of the flow front was covered by younger-
11
looking lobate flows with lighter sediment cover. The morphological sequence was repeated
12
again to the west, and sample 3963-7 was collected from a third flow front. The third pillow flow
13
front was surveyed using sonar Imagenex (Fig. 1, 2). Sample 3963-8 was obtained from the
14
lobate flows at the base of the flow front. Beyond the flow front, a lava channel was located
15
(identified earlier from ABE bathymetry) in a flat area of lobate and sheet flows. Sample 3963-9
16
was collected from the wall of the channel, and the area was surveyed with Imagenex. The
17
channel was followed towards the ridge axis, and terminated in lobate and sheet terrain. A few
18
hundred meters beyond the channel termination, samples 3963-10 and 3963-11 were obtained
19
from a fourth pillow flow front.
20
21
Dive 3974: Eleven more samples were acquired on the western flank during dive 3974 (Fig. 1, 2,
22
3, Table 1) [Schouten et al., 2004]. The dive traversed west-to-east along and across a series of
23
four flow fronts identified from ABE bathymetry and by their pillow morphology during dive
24
observations, starting ~1.9 km west of the axial summit trough and following the fronts back up
25
the western ridge flank towards the axial summit trough. As in dive 3963, each flow front was
26
composed of steep (10º-45º) pillow lavas located between flat (0º-4º) areas of mixed sheet and
27
lobate flows [Schouten et al., 2004]. Of the 11 samples collected, five pillow lavas were sampled
28
directly from the flow fronts, and three sheet flows and three lobate flows were sampled from
29
flat areas adjacent to the flow fronts.
30
31
In both dives, flat sheet flows sometimes appeared to form channels in the hackly sheet
32
and lobate flows [Schouten et al., 2004]. In at least one case, the flat and hackly sheet flows
33
appeared to flow down a flow front (flat sheets appeared superimposed on the adjacent pillow
34
lavas). Dive notes speculate that the pillow lavas may have been a late eruptive phase that
35
extruded from the hackly flow unit [Schouten et al., 2004]. In both dives, sediment cover along
36
the dive track was relatively constant, with slightly more sedimentation farther away from the
37
axis.
38
39
Appendix B. Inter-laboratory reproducibility of U and Th Elemental Abundances, Th/U,
40
(230Th/232Th), and (230Th/238U)
41
In Figure Appendix B1, we compare elemental abundances of U and Th as measured by
42
isotope dilution (ID), Th/U, (230Th/232Th), and (230Th/238U) for 2005-06 eruption samples (4202-
43
4, 4202-6, 4205-5, and 4205-6) analyzed in this study and for 1991-92 eruption samples (2392-9,
44
2504-1, 2372-1, 2497-1b) reported in previous studies [Lundstrom et al., 1999; Rubin et al.,
45
2005; Sims et al., 2002; Tepley et al., 2004; Turner et al., 2011]. Uranium and Th ID abundance
46
data were measured for this study at WHOI, the University of Wyoming (WILD), and the
47
University of Bristol (BIG). The WHOI, BIG and WILD measurements were made on separate
48
glass dissolutions. We also compare these Th and U abundance ID data to data measured both by
49
LA-ICP-MS at Lamont-Dohert Earth Observatory (LDEO) and reported in this study and by
50
traditional ICP-MS methods at the University of Florida and reported in Goss et al. [2010]
51
(Figure A1).
52
With the exception of sample 4202-6, for which BIG data are significantly higher than
53
WHOI and WILD data for both Th (~20%) and U (~15%), and sample 4205-6, for which U and
54
Th data are ~4-5% lower than WHOI and WILD data, sample U and Th ID concentrations
55
measured at WHOI are within 1.3% and 2.9% of the U and Th ID concentrations determined by
56
BIG, thus overlapping within average analytical uncertainties of ~±1.5% (Fig. Appendix B1a, b).
57
Uranium and Th ID data from WILD are all within 1% of data from the WHOI lab. WHOI and
58
WILD used the same 229Th and 233U spikes. The averages of WHOI, WILD, and BIG U and Th
59
ID abundance data for 2005-06 samples (excepting 4202-6 and 4205-6, for which only WHOI
60
and WILD data are averaged) are consistently different from LDEO LA-ICP-MS measurements:
61
the average ID Th measurements are 11.1% low to 9.3% high; U is 7.8% low to 9.1% high; but,
62
these values are also within the larger estimated uncertainties of the LA-ICP-MS method (±10%
63
for Th; ±20% for U, RSD (2)). Compared with the University of Florida data, the average ID
64
measurements are ~7.8-13% lower for U and ~3.9-10% lower for Th. These values are outside of
65
the analytical uncertainties reported by Goss et al. [2010] for U (<5%, 2; Th reproducibility is
66
not reported explicitly), suggesting either poorer accuracy or underestimation of analytical
67
uncertainties. Overall, with the anomalous exception of sample 4205-6, the U and Th ID data for
68
the 2005-06 samples appear to be remarkably consistent, and even inter-laboratory uncertainties
69
appear to be within the range of estimated within-lab external reproducibility (<±1.5%). Uranium
70
and Th ID data from 1991-1992 eruption samples (including measurement of 2392-9 by BIG) are
71
similarly consistent. Laser ablation and traditional ICP-MS methods are consistent with the ID
72
data given external reproducibility of ~±10% (2). Thus, ID methods have larger precision-
73
normalized variability than LA-ICP-MS and ICP-MS methods.
74
As with Th and U elemental abundances, Th/U calculated using ID data from WHOI,
75
WILD, and BIG are indistinguishable within propagated analytical uncertainties of ±2.1% (2)
76
(Appendix Fig. B1c). Even Th/U for sample 4202-6, for which the BIG Th and U abundances
77
are significantly different, is in agreement with the WHOI and WILD data within analytical
78
uncertainties. Ratios of Th/U calculated using data from LDEO and the University of Florida are
79
also in agreement with WHOI and BIG ID data within analytical uncertainties, but uncertainties
80
for LDEO data are much greater (~12%, 2). Measured Th isotopic ratios and (230Th/238U) from
81
WHOI and BIG are indistinguishable within analytical uncertainties (±0.8-1.2%, 2) for all
82
samples except 4202-6, for which the (230Th/232Th) values from WHOI are higher than BIG
83
(Appendix Fig. B1d, e).
84
85
Appendix C. Measured Data for Putative 2005-06 Eruption Sample NH-D1a
86
As part of this study, we also measured major and trace element, Nd and Hf isotope, and
87
238
U-230Th-226Ra isotope compositions for sample NH-D1a (Appendix Table C1), which was
88
dredged from the northern off-axis part of the 2005-06 flow at 9º54.5’N [cf., Goss et al., 2010].
89
Although Goss et al. [2010] report that NH-D1a is “210Po-dated as having erupted during the
90
2005-06 event (K.H. Rubin et al., unpublished data, 2008)”, they also describe the sample as
91
“chemically anomalous,” highlight major and trace element compositional differences between
92
NH-D1a and other 9º50’N EPR lavas, and suggest that it “is unlikely to be directly
93
petrogenetically related to any of the other lavas erupted in 2005-06.” We agree with this
94
statement, but based on observations of anomalous Hf and Nd isotope composition of NH-D1a
95
(Hf=13.41, Nd= 10.02), which does not resemble 9º-10ºN EPR lavas, the low 226Ra ~22 fg/g and
96
(226Ra/230Th) ~1.11, and hydrothermal staining and alteration of the glass in hand sample and
97
under binocular microscope, we wonder if this sample, which we note again was dredged, is
98
truly representative of this event as compared with all of the other lavas here. Therefore, we do
99
not otherwise consider sample NH-D1a in this study.
100
101
Appendix Figure B1. Inter-laboratory comparison of (a) Th and (b) U abundances, (c) Th/U,
102
(d) (230Th/232Th), and (e) (230Th/238U) for 2005-06 samples measured at Woods Hole
103
Oceanographic Institution (WHOI), the University of Wyoming (WILD), the University of
104
Bristol Isotope Group (BIG), Lamont-Doherty Earth Observatory (LDEO), and the University of
105
Florida [Goss et al., 2010]. Uranium and Th abundances were measured by isotope dilution at
106
WHOI, WILD, and BIG. Additional U and Th ID data and U and Th isotope compositions are
107
compiled for 1991-92 eruption samples from [Lundstrom et al., 1999; Rubin et al., 2005; Sims et
108
al., 2002; Tepley et al., 2004; Turner et al., 2011].
109
110
111
112
113
114
115
116
117
118
119
120
121
References
Goss, A.R., M.R. Perfit, W.I. Ridley, K.H. Rubin, G.D. Kamenov, S.A. Soule, and A. Fundis
(2010), Geochemistry of lavas from the 2005-2006 eruption at the East Pacific Rise,
9º46’N-9º56’N: Implications for ridge crest plumbing and decadal changes in magma
chamber compositions, Geochem., Geophys., Geosyst., 11(5), 1-35, doi:
10.1029/2009GC002977.
Lundstrom, C.C., D.E. Sampson, M.R. Perfit, J. Gill, and Q. Williams (1999), Insights into midocean ridge basalt petrogenesis: U-series disequilibria from the Siqueiros Transform,
Lamont Seamounts, and East Pacific Rise, J. Geophys. Res., 104, 13,035-13,048.
Rubin, K.H., I. van der Zander, M.C. Smith, and E.C. Bergmanis (2005), Minimum speed limit
for ocean ridge magmatism from 210Pb-226Ra-230Th disequilibria, Nature, 437, 534-538.
122
123
124
125
126
127
128
129
130
131
132
133
134
Schouten, H., M. Tivey, D.J. Fornari, W. Seyfried, and Ship Board Science and Technical Teams
(2004), Central Anomaly Magnetic High and Volcanic Processes on Fast-spreading MidOcean Ridges, Cruise Report R/V Atlantis Voyage 11 Leg 7, 1/26-2/24/2004.
Sims, K.W.W., S.J. Goldstein, J. Blichert-Toft, M.R. Perfit, P. Kelemen, D.J. Fornari, P.
Michael, M.T. Murrell, S.R. Hart, D.J. DePaolo, G. Layne, and M. Jull (2002), Chemical
and isotopic constraints on the generation and transport of melt beneath the East Pacific
Rise, Geochim. Cosmochim. Acta, 66, 3481-3504.
Tepley, F.J. III, C.C. Lundstrom, K.W.W. Sims, R. Hekinian (2004), U-series disequilibria in
MORB from the Garrett Transform and implications for mantle melting, Earth Planet.
Sci. Lett., 223, 79-97.
Turner, S., C. Beier, Y. Niu, C. Cook (2011), U-Th-Ra disequilibria and the extent of off-axis
volcanism across the East Pacific Rise at 9º30’N, 10º30’N, and 11º20’N, Geochem.
Geophys. Geosyst., 12(7), doi:10.1029/2010GC003403.