Geology

doi: 10.1130/G32028.1 2011;39;911-914Geology

 Walter R. RoestDaniel Sauter, Heather Sloan, Mathilde Cannat, John Goff, Philippe Patriat, Marc Schaming and and crustal thickness?From slow to ultra-slow: How does spreading rate affect seafloor roughness  

Email alerting servicesarticles cite this article

to receive free e-mail alerts when newwww.gsapubs.org/cgi/alertsclick

Subscribe to subscribe to Geologywww.gsapubs.org/subscriptions/click

Permission request to contact GSAhttp://www.geosociety.org/pubs/copyrt.htm#gsaclick

official positions of the Society.citizenship, gender, religion, or political viewpoint. Opinions presented in this publication do not reflectpresentation of diverse opinions and positions by scientists worldwide, regardless of their race, includes a reference to the article's full citation. GSA provides this and other forums for thethe abstracts only of their articles on their own or their organization's Web site providing the posting to further education and science. This file may not be posted to any Web site, but authors may postworks and to make unlimited copies of items in GSA's journals for noncommercial use in classrooms requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in subsequenttheir employment. Individual scientists are hereby granted permission, without fees or further Copyright not claimed on content prepared wholly by U.S. government employees within scope of

Notes

© 2011 Geological Society of America

on September 13, 2011geology.gsapubs.orgDownloaded from

GEOLOGY, October 2011 911

INTRODUCTIONRecent investigations of the Southwest Indian

Ridge (SWIR) and Arctic ridges have revealed that the differences between ultra-slow and slow spreading ridges (full rate of <20 mm/yr and 20–55 mm/yr, respectively) are as great as those between slow and fast spreading ridges (Dick et al., 2003; Sauter and Cannat, 2010). The absence of volcanic activity on >100-km-long stretches of the ridge axis is one of the striking contrasts recently discovered between the SWIR and faster spreading ridges (Cannat et al., 2006; Dick et al., 2003; Sauter et al., 2004). It is also commonly inferred that both seafl oor roughness and crustal thickness change dramatically for full spreading rates of <20 mm/yr (Menard, 1967; Reid and Jackson, 1981). Many of the differences between slow and ultra-slow spreading ridges are attrib-uted to spreading rate, but how sure are we that spreading rate plays such an important role?

Some abyssal hill relief or seafl oor roughness versus spreading rate curves indicate that the greatest increase in roughness occurs as spread-ing rate decreases from slow to ultra-slow (Ehlers and Jokat, 2009; Malinverno, 1991). However, Whittaker et al. (2008), in their global analysis, found no signifi cant increase in gravity-derived seafl oor roughness with slow to ultra-slow spreading rate decrease. There is considerable

data scatter around any smooth relationship between roughness and spreading rate, as they are typically based on global compilations of data from several ridges (e.g., Ehlers and Jokat, 2009), including data from anomalously shallow or deep regions (e.g., at the SWIR; see Sauter et al., 2009). Factors other than spreading rate, such as mantle temperature, may affect the litho-spheric strength and thus the roughness. There-fore, global compilations may not provide the best characterization of seafl oor roughness varia-tion related to a spreading rate change. Until the models are tested at slow and ultra-slow spread-ing ridges with extensive data coverage, how sure are we that this ultra-slow critical limit exists?

Although seismic crustal thickness shows lit-tle or no dependency on full spreading rate down to ~20 mm/yr, below this rate it is predicted to decrease rapidly (White et al., 2001). This pre-diction is based on subaxial mantle upwelling models in which enhanced conductive cool-ing below the ultra-slow critical limit results in thickening of the lithosphere and reduced upwelling velocities (Reid and Jackson, 1981). The vertical distance over which adiabatic melting occurs is in turn reduced, diminishing the amount of melt and thereby the volume of crust produced (White et al., 2001). However, the global-scale seismic crustal thickness stud-

ies and geochemical analyses that these models conceptualize are also based upon compilations of widely spaced data (White et al., 2001) and, therefore, may produce misleading results.

In this paper we make an unprecedented investigation into the role of spreading rate in determining seafl oor character at slow versus ultra-slow spreading ridges by examining a well-constrained transition from slow (30 mm/yr) to ultra-slow (15 mm/yr) spreading at the SWIR. This transition occurred at magnetic anomaly C6C (ca. 24 Ma; Patriat et al., 2008) and produced only small local changes in plate boundary geometry (Baines et al., 2007). The key advance in our work is the ability to test the relationship of spreading rate with seafl oor roughness and crustal thickness within individ-ual fl ow line corridors as well as the relationship between seafl oor roughness, crustal thickness, and inferred mantle temperature variation on the fl anks parallel to the axis. We restrict our study to the 54°–67°E section of the SWIR (Fig. 1), which is the only part of the ridge with extensive data coverage, minimal sediment accumula-tion, and no hotspot infl uence. The eastern part of our study area is one of the deepest parts of the oceanic ridge system and is thus inferred to represent a colder, melt-poor end member of the ridge system (Cannat et al., 2008).

DATA AND METHODS

BathymetryWe compiled multibeam bathymetric data

from several French, Japanese, and U.S. cruises at the SWIR between 54°E and 67°E (see Table DR1 in the GSA Data Repository1). The stochastic modeling process of Goff and Jordan

Geology, October 2011; v. 39; no. 10; p. 911–914; doi:10.1130/G32028.1; 3 fi gures; Data Repository item 2011274.© 2011 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.

From slow to ultra-slow: How does spreading rate affect seafl oor roughness and crustal thickness?Daniel Sauter1, Heather Sloan2, Mathilde Cannat3, John Goff4, Philippe Patriat3, Marc Schaming1, and Walter R. Roest5

1 Institut de Physique du Globe de Strasbourg, IPGS-UMR 7516, CNRS, Université de Strasbourg/EOST, 1 rue Blessig, 67084 Strasbourg cedex, France

2 Environmental, Geographic and Geological Sciences, Lehman College, City University of New York, 250 Bedford Park Boulevard West, Bronx, New York 10468, USA

3 Equipe de Géosciences Marines, CNRS-UMR 7154, Institut de Physique du Globe de Paris, 4 place Jussieu, 75252 Paris cédex 05, France

4 Institute for Geophysics, Jackson School of Geosciences, University of Texas at Austin, 10100 Burnet Road, Austin, Texas 78758, USA5 Ifremer (Institut Français de Recherche Pour l’Exploitation de la Mer), Centre de Brest, Géosciences Marines, BP 70 29280 Plouzané, France

ABSTRACTWe examine the relationship of seafl oor roughness and gravity-derived crustal thickness to

both spreading rate and inferred mantle temperature using statistical analysis of a multibeam bathymetry and gravity data compilation of the axis and fl anks between 54°E and 67°E at the Southwest Indian Ridge (southwest Indian Ocean). Our fi ndings indicate that root mean square values of abyssal hill heights increase from 220 ± 20 m to 300 ± 20 m along fl ow line corridors that transition a well-constrained full spreading rate change from slow (30 mm/yr) to ultra-slow (15 mm/yr). Mantle Bouguer gravity anomalies, however, indicate no signifi -cant change in inferred crustal thickness at the spreading rate transition. In the axis-parallel direction, roughness of both slow and ultra-slow seafl oor increases from 54°E to 63°E while inferred crustal thickness and/or mantle temperature decrease. These fi ndings have implica-tions for the relationship between spreading rate and melt production: they suggest that man-tle temperature at slow and ultra-slow ridges may play a more important role than spreading rate in determining seafl oor morphology. The lack of evidence for signifi cant crustal thinning accompanying a change from slow to ultra-slow spreading rate lends support to focused sub-axial mantle upwelling models.

1GSA Data Repository item 2011274, sources of multibeam bathymetric data, seafl oor roughness es-timation method, mantle Bouguer anomaly calcula-tion, and effects of sedimentation of seafl oor rough-ness estimates, is available online at www.geosociety.org/pubs/ft2011.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.

on September 13, 2011geology.gsapubs.orgDownloaded from

912 GEOLOGY, October 2011

(1988, 1989) produced estimates of the root mean square (RMS) height, which is the aver-age depth variation with respect to the regional depth trend, and the azimuth of abyssal hill lin-eaments. We analyzed 31 areas of seafl oor des-ignated by boxes: 10 selected between C13n.y and C8n.o where the full spreading rate was slow at 30 mm/yr (Patriat et al., 2008), and 21 between C6n.o and C3An.y where the full spreading was ultra-slow at 15 mm/yr (Patriat et al., 2008) (Fig. 1). Each box is long enough in the spreading fl ow line direction (typically ~80 km, but ranging from ~30 to ~110 km) to derive well-resolved estimates of statistical properties (Goff and Jordan, 1989). We favored segment centers, excluding nonvolcanic smooth seafl oor (Cannat et al., 2006) and seafl oor with ages younger than C3An.y, to avoid relief asso-ciated with the axial valley and ridge mountains (Fig. DR1). Seismic data within the region allow us to quantify sediment thickness, which is neg-ligible within the boxes (Fig. DR1) where sea-fl oor roughness parameters are estimated (see Figs. DR4 and DR5).

GravityWe obtained mantle Bouguer anomalies

(MBA) by removing the effect of a constant

thickness (5 km), constant density (2700 kg/m3) crust from free air anomaly data (see Fig. DR2). MBA lows correspond to thicker constant den-sity model crust and/or to lighter crust or mantle material, and MBA highs correspond to thinner constant density model crust and/or to denser material. Mean values of the MBA were calcu-lated within each of the boxes (Fig. 1) for the same time intervals (26–31 Ma and 8–13 Ma for slow and ultra-slow seafl oor, respectively) to overcome variations in MBA due to age-related subsidence between the boxes. The effect of cooling of the plates with age was calculated using the magnetic anomalies identifi ed by Patriat et al. (2008). The gravity effect of cool-ing of the plates with age was then removed from the MBA to obtain residual MBA (RMBA) values along fl ow lines.

RESULTS

Seafl oor RoughnessWe found that seafl oor roughness in our

study boxes varies not only along fl ow lines as spreading rate changes from slow to ultra-slow, but also with longitude parallel to the axis. We quantifi ed the differences between slow and ultra-slow spread seafl oor roughness properties

by calculating the mean values of the parameter estimates weighted by the proportional error estimate within slow and ultra-slow popula-tions (Goff, 1991). The mean RMS height of abyssal hills of ultra-slow spread seafl oor (300 ± 20 m) is signifi cantly larger than that of slow spread seafl oor (220 ± 20 m). The angular dif-ference (~12°) between the mean azimuths of the two populations is remarkably consistent with the change of spreading direction (~13°) obtained from kinematic analysis (Patriat et al., 2008). Figure 2A displays the variation of RMS abyssal hill height with longitude for the boxes selected for slow and ultra-slow spread seafl oor. Although there is some overlap between the two populations, they are nevertheless distinct with,

A

B

C

−100

−80

−60

−40

−20

0

20

40

54 56 58 60 62 64 66 68

−5000

−4500

−4000

−3500

−3000

−2500

−2000

54 56 58 60 62 64 66 68

50

100

150

200

250

300

350

400

450

500

54 56 58 60 62 64 66 68

RMS

heig

ht (m

)D

epth

(m)

MBA

(mG

al)

Longitude (°E)

Melville FZAtlantis II FZ

54˚E 55˚E 56˚E 57˚E 58˚E 59˚E 60˚E 61˚E 62˚E 63˚E 64˚E 65˚E 66˚E 67˚E36˚S

35˚S

34˚S

33˚S

32˚S

31˚S

30˚S

29˚S

28˚S

27˚S

26˚S

−7000 −6000 −5000 −4000 −3000 −2000 −1000 0 (m)

SWIR

RTJ

SEIR

CIR

Indian Ocean

Africa

India

s

Gal

lieni

FZ

Nov

ara

FZ

Mel

ville

FZ

Triple

junction tr

aceA

tlan

tis

II F

Z

Axis

Magnetic

anomaly

C5n.o

C6n.o

C8n.o

C13n.y

Ν

Figure 1. Multibeam bathymetric compilation map of Southwest Indian Ridge (SWIR) between Gallieni Fracture Zone and 67°E with location map inset. Black boxes—selected ultra-slow spreading areas for roughness estimations; white boxes—selected slow spreading areas for roughness estimations; dashed red line—SWIR axis; black lines—transform fault and fracture zone; solid red lines—location of gravity profi les shown in Figure 3. Magnetic anomaly identifi -cations C6n.o, C8n.o, and C13n.y are after Patriat et al. (2008) and Sauter et al. (2008); C5n.o is after Lemaux et al. (2002). See Figure DR1 (see footnote 1) for more detail. FZ—fracture zone; SEIR—Southeast Indian Ridge; CIR—Central Indian Ridge; RTJ—Rodriguez Triple Junction.

Figure 2. Results of stochastic modeling of abyssal hill statistical parameters and man-tle Bouguer anomaly (MBA) analysis with longitude for boxes of ultra-slow spreading seafl oor (red) and slow spreading seafl oor (blue). A: Root mean square (RMS) abyssal hill height (meters). B: Mean depth (meters). C: Mean mantle Bouguer anomalies (mgal). Squares—ultra-slow spreading lithosphere, diamonds—slow spreading lithosphere with 1σ error bars. Lines show second-order polynomial fi t. FZ—fracture zone.

on September 13, 2011geology.gsapubs.orgDownloaded from

GEOLOGY, October 2011 913

for the most part, values for ultra-slow spread seafl oor greater than for slow spread seafl oor at a given longitude. West of the Atlantis II Frac-ture Zone, all but one box of ultra-slow spread seafl oor show RMS heights of <200 m, the same or less than the RMS height estimated for slow spread seafl oor (Fig. 2A). From the Atlantis II Fracture Zone to 63°E, both slow and ultra-slow populations display signifi cant increase in RMS height (Fig. 2A). East of 63°E RMS heights for slow and ultra-slow populations decrease to val-ues approaching the mean value for each of the respective populations.

Mean Seafl oor Depth and Gravity SignatureMean MBA and seafl oor depth in our study

boxes display the same regional trend as observed on axial data (Cannat et al., 1999; Georgen et al., 2001). The lowest MBA values for both slow and ultra-slow spreading crust are found east of the Atlantis II Fracture Zone and indicate thicker crust and/or lower densities. Between the Atlantis II and Melville Fracture Zones, mean MBA values and mean seafl oor depth increase from west to east for slow and particularly for ultra-slow spread seafl oor (Fig. 2), indicating an eastward thinning of the crust and/or an increase of densities. East of the Melville Fracture Zone, the mean MBA for ultra-slow spread seafl oor decreases moderately eastward while the mean depth remains relatively constant.

Figure 3 displays RMBA variations along fl ow lines within three segments that main-tained stable axial geometries since C13n.y (ca. 33 Ma). The three profi les are located west of the Atlantis II Fracture Zone, between the Atlantis II and Melville Fracture Zones and east of the Melville Fracture Zone, in three regional domains that have contrasting seafl oor depth and MBA (Fig. 1; see the Data Repository). To the west of the Atlantis II Fracture Zone, there is a progressive decrease in RMBA as age decreases, with no visible variation associ-ated with the change in spreading rate 24 Ma. Between the Atlantis II and Melville Fracture Zones, and east of the Melville Fracture Zone, the RMBA remains relatively constant on aver-age over the entire investigated period (mean RMBA −15 ± 6 mGal and 10 ± 5 mGal, respec-tively; Fig. 3). To the east of the Melville Frac-ture Zone, variations around this mean RMBA value have larger amplitudes and wavelengths for ultra-slow spread seafl oor than for slow spread seafl oor. This is in agreement with the more detailed analysis of RMBA performed in the eastern region (Cannat et al., 2006).

DISCUSSIONSeafl oor roughness increases along fl ow lines

on the fl anks of the SWIR between 54°E and 67°E, where the full spreading rate decreased from 30 mm/yr to 15 mm/yr ca. 24 Ma, but

we fi nd no evidence for signifi cant reduction in crustal thickness as a result of this spread-ing rate change. In fact, more negative RMBA values suggest that crustal thickness increases locally in ultra-slow spread seafl oor (e.g., west of the Atlantis II Fracture Zone; Fig. 3A). The mean depth, seafl oor roughness, and mean MBA within the ridge fl ank study boxes (Fig. 1) increase from west to east for both slow and ultra-slow spread seafl oor between 54° and ~63°E (Fig. 2). These fi ndings challenge some aspects of currently accepted models of sea-fl oor spreading, in particular the relationships between spreading rate and melt production.

The SWIR is a unique example of an ultra-slow spreading ridge with little sediment accu-mulation and extensive survey coverage on axis and off axis, thus meeting the criteria for cal-culating stochastic estimates of seafl oor rough-ness. Larger RMS abyssal hill height values (265–584 m) obtained from isolated seismic refl ection profi les in the ultra-slow spreading Arctic Basin (Ehlers and Jokat, 2009) may not adequately account for regional morphology variability. RMS abyssal hill heights of slow spread SWIR seafl oor (220 ± 20 m) are close to those estimated using the same method at the slow spread Mid-Atlantic Ridge (220–240 m) (Goff et al., 1995; Neumann and Forsyth, 1995). The thickness of the lithosphere near ridge axes

increases with decreasing spreading rate (Huang and Solomon, 1988). Greater elastic strength at ultra-slow spreading ridges can support forma-tion of larger topographic features. Accordingly, RMS abyssal hill height values are signifi cantly larger for ultra-slow spread seafl oor than for slow spread seafl oor, suggesting that variation in roughness along fl ow lines at the SWIR is likely due to the change in spreading rate. This is further supported by the coherence of SWIR roughness estimates with global estimates and their relationship to spreading rate. However, the variability of seafl oor roughness with longitude in our study area is also correlated, independent of spreading rate, with variations of mean depth and MBA (Fig. 2). This suggests that spreading rate is not the sole controlling factor of seafl oor roughness.

These longitudinal variations of mean depth and MBA (Fig. 2) suggest that there is an axis-parallel, along-isochron variability in the den-sity structure of the off-axis lithosphere, consis-tent with a gradual thinning of the low-density crustal layer from west to east, and/or with a colder mantle to the east. Such an eastward decrease of crustal thickness and/or mantle temperature was also inferred from gravity, sea-fl oor depth, and basalt sodium content data col-lected along the SWIR axis in the same region (Cannat et al., 1999, 2008). Thinner crust in the easternmost part of the SWIR is confi rmed by seismic data (Minshull et al., 2006; Muller et al., 1999). Colder mantle temperatures leading to lower degrees of mantle melting and a lesser melt production could affect the strength of the axial lithosphere (Behn and Ito, 2008; Shaw and Lin, 1996), and may play an important role in determining seafl oor roughness. A similar infl u-ence of mantle temperature on seafl oor rough-ness may be observed at the Southeast Indian Ridge between the Saint Paul and Amsterdam hotspot and the Australian-Antarctic discor-dance (e.g., Goff et al., 1997).

The trend of increasing seafl oor roughness in our study boxes (Fig. 2) from the Atlantis II Fracture Zone to 63°E suggests that the strength of the axial lithosphere at the time this seafl oor was formed increased with longitude. In the boxes west of the Atlantis II Fracture Zone and to the east of 63°E, we observe some variance in this trend that may be attributed to enhanced melt production (Figs. 1 and 2). Cold mantle tem-perature regimes may produce not only thicker lithosphere, but also narrower zones of mantle upwelling beneath the axis. This focusing of mantle upwelling may result in locally enhanced melt production (Bown and White, 1994; Can-nat et al., 2008). Additional factors such as ridge obliquity and variations of melt supply with time at a given on-axis location may also affect the rheology of the axial lithosphere, explaining local variability of seafl oor roughness.

Age (Ma)Age (Ma)

Man

tle B

ougu

er a

nom

aly

(mG

al)

Man

tle B

ougu

er a

nom

aly

(mG

al)

−80−60

−40−20

02040

0 10 20 30 40

−80−60−40

−200

2040

0 10 20 30 40

−80−60

−40

−200

2040

0 10 20 30 40

West of Atlantis II FZ

East of Melville FZ

Between Atlantis II and Melville FZs

SlowUltra-slow

Figure 3. Variation of mantle Bouguer anoma-lies (MBA) (thin gray lines) and residual MBA (thick black lines) along fl ow lines as func-tion of age (Ma) within three segments that maintained stable axial geometries since C13n.y (ca. 33 Ma). Locations in Figure 1. Vertical thick gray bar indicates spreading rate change ca. 24 Ma from slow (30 mm/yr) to ultra-slow (15 mm/yr). FZ—fracture zones.

on September 13, 2011geology.gsapubs.orgDownloaded from

914 GEOLOGY, October 2011

The 15 mm/yr rate change from slow to ultra-slow spreading that occurred at the SWIR transitions the critical range of rates for which mantle rising beneath the ridge at corner fl ow velocities similar to the half-spreading rate would depart from adiabatic decompression (see White et al., 2001, their Fig. 24). Our obser-vations show that there is no evidence for a thin-ner crust as spreading rates decrease from slow to ultra-slow in our study area. This is an argu-ment in favor of a more complex relationship between spreading rate and melt production in the subaxial mantle. In the model of Bown and White (1994), which calls for corner fl ow with a spreading rate–dependent lithosphere wedge angle, mantle upwelling is focused, and therefore accelerated, beneath the ridge. While the half-spreading rate upwelling velocity cor-ner fl ow model of mantle upwelling predicts a strong decrease of melt thickness (>2 km) as the full spreading rate decreases from 30 mm/yr to 15 mm/yr (e.g., Dick et al., 2003), the focused and accelerated mantle upwelling model pre-dicts only a slight reduction (~0.5 km) for the same spreading rate decrease (Bown and White, 1994). The absence of a signifi cant reduction in crustal thickness as spreading rate decreases at the SWIR confi rms that, as proposed by Can-nat et al. (2008) and Sauter and Cannat (2010) on the basis of compared gravity, seafl oor depth, and basalt chemistry, enhanced melt production due to focused and accelerated mantle upwell-ing may be a likely mechanism for the genera-tion of magma at ultra-slow spreading ridges.

ACKNOWLEDGMENTSWe thank M. Maia, C. Hemond, and J. Dyment for

providing unpublished data from the PLURIEL and GEISEIR cruises, and H. Kinoshita, S. Arai, T. Matsu-moto, and H. Fujimoto, who agreed to provide YK98-08, KR00-06, YK01-14, and YK98-07 (Indoyo) cruise data to us. We also thank X. Morin and B. Ollivier of IPEV (Institut Polaire Français), R. Iwase of JAM-STEC (Japan Agency for Marine-Earth Science and Technology), and B. Loubrieu of Ifremer (Institut Français de Recherche Pour l’Exploitation de la Mer), who helped us to gather the data, and two anonymous reviewers and Henry Dick for their comments.

REFERENCES CITEDBaines, A.G., Cheadle, M.J., Dick, H.J.B., Scheirer,

A.H., John, B.E., Kusznir, N.J., and Matsu-moto, T., 2007, Evolution of the Southwest Indian Ridge from 55°45′E to 62°E: Changes in plate-boundary geometry since 26 Ma: Geochemistry Geophysics Geosystems, v. 8, Q06022, doi:10.1029/2006GC001559.

Behn, M.D., and Ito, G., 2008, Magmatic and tec-tonic extension at mid-ocean ridges: 1. Con-trols on fault characteristics: Geochemis-try Geophysics Geosystems, v. 9, Q08O10, doi:10.1029/2008GC001965.

Bown, J.W., and White, R.S., 1994, Variation with spreading rate of oceanic crustal thickness and geochemistry: Earth and Planetary Science Letters, v. 121, p. 435–449, doi:10.1016/0012-821X(94)90082-5.

Cannat, M., Rommevaux-Jestin, C., Sauter, D., Deplus, C., and Mendel, V., 1999, Formation of the axial relief at the very slow spreading Southwest Indian Ridge (49°–69°E): Journal of Geophysical Research, v. 104, p. 22,825–22,843, doi:10.1029/1999JB900195.

Cannat, M., Sauter, D., Mendel, V., Ruellan, E., Okino, K., Escartin, J., Combier, V., and Baala, M., 2006, Modes of seafl oor generation at a melt-poor ultra-slow-spreading ridge: Geol-ogy, v. 34, p. 605–608, doi:10.1130/G22486.1.

Cannat, M., Sauter, D., Bezos, A., Meyzen, C., Hum-ler, E., and Le Rigoleur, M., 2008, Spreading rate, spreading obliquity, and melt supply at the ultra-slow spreading Southwest Indian Ridge: Geochemistry Geophysics Geosystems, v. 9, Q04002, doi:10.1029/2007GC001676.

Dick, H.J.B., Lin, J., and Schouten, H., 2003, An ul-traslow-spreading class of ocean ridge: Nature, v. 426, p. 405–412, doi:10.1038/nature02128.

Ehlers, B.-M., and Jokat, W., 2009, Subsidence and crustal roughness of ultra-slow spreading ridges in the northern North Atlantic and the Arctic Ocean: Geophysical Journal Interna-tional, v. 177, p. 451–462, doi:10.1111/j.1365-246X.2009.04078.x.

Georgen, J.E., Lin, J., and Dick, H.J.B., 2001, Evi-dence from gravity anomalies for interactions of the Marion and Bouvet hotspots with the Southwest Indian Ridge: Effects of transform offsets: Earth and Planetary Science Letters, v. 187, p. 283–300, doi:10.1016/S0012-821X(01)00293-X.

Goff, J.A., 1991, A global and regional stochastic analysis of near-ridge abyssal hill morphol-ogy: Journal of Geophysical Research, v. 96, p. 21713–21737, doi:10.1029/91JB02275.

Goff, J., and Jordan, T., 1988, Stochastic modeling of seafl oor morphology: Inversion of Sea Beam data for second-order statistics: Journal of Geophysical Research, v. 93, B11, p. 13589–13608, doi: 10.1029/JB093iB11p13589.

Goff, J.A., and Jordan, T.H., 1989, Stochastic mod-eling of seafl oor morphology: Resolution of topographic parameters by Sea Beam data: IEEE Journal of Oceanic Engineering, v. 14, p. 326–337, doi:10.1109/48.35983.

Goff, J.A., Tucholke, B.E., Lin, J., Jaroslow, G.E., and Kleinrock, M.C., 1995, Quantitative anal-ysis of abyssal hills in the Atlantic Ocean: A correlation between axis crustal thickness and extensional faulting: Journal of Geophysical Research, v. 100, p. 22509–22522, doi:10.1029/95JB02510.

Goff, J.A., Ma, Y., Shah, A., Cochran, J.R., and Sem-péré, J.-C., 1997, Stochastic analysis of sea-fl oor morphology on the fl ank of the Southeast Indian Ridge: The infl uence of ridge morphol-ogy on the formation of abyssal hills: Journal of Geophysical Research, v. 102, p. 15521–15534, doi:10.1029/97JB00781.

Huang, P., and Solomon, S.C., 1988, Centroid depths of Mid-Ocean ridge earthquakes: Dependence on spreading rate: Journal of Geophysical Re-search, v. 93, p. 13445–13477, doi:10.1029/JB093iB11p13445.

Lemaux, J., Gordon, R.G., and Royer, J.Y., 2002, Location of the Nubia-Somalia boundary along the Southwest Indian Ridge: Geology, v. 30, p. 339–342, doi:10.1130/0091-7613(2002)030<0339:LOTNSB>2.0.CO;2.

Malinverno, A., 1991, Inverse square-root depen-dence of mid-ocean-ridge fl ank roughness on spreading rate: Nature, v. 352, p. 58–60, doi:10.1038/352058a0.

Menard, H.W., 1967, Sea fl oor spreading, topography, and the second layer: Science, v. 157, p. 923–924, doi:10.1126/science.157.3791.923.

Minshull, T.A., Muller, M.R., and White, R.S., 2006, Crustal structure of the Southwest Indian Ridge at 66°E: Seismic constraints: Geophysi-cal Journal International, v. 166, p. 135–147, doi:10.1111/j.1365-246X.2006.03001.x.

Muller, M.R., Minshull, T.A., and White, R.S., 1999, Segmentation and melt supply at the South-west Indian Ridge: Geology, v. 27, p. 867–870, doi:10.1130/0091-7613(1999)027<0867:SAMSAT>2.3.CO;2.

Neumann, G.A., and Forsyth, D.W., 1995, High res-olution statistical estimation of seafl oor mor-phology: Oblique and orthogonal fabric on the fl anks of the Mid-Atlantic Ridge, 34°–35.5°S: Marine Geophysical Researches, v. 17, p. 221–250, doi:10.1007/BF01203464.

Patriat, P., Sloan, H., and Sauter, D., 2008, From slow to ultra-slow: A previously undetected event at the Southwest Indian Ridge ca. 24 Ma: Geology, v. 36, p. 207–210, doi:10.1130/G24270A.1.

Reid, I., and Jackson, H.R., 1981, Oceanic spreading rate and crustal thickness: Marine Geophysi-cal Researches, v. 5, p. 165–172, doi:10.1007/BF00163477.

Sauter, D., and Cannat, M., 2010, The ultraslow-spreading Southwest Indian Ridge, in Rona, P., et al., eds., Diversity of hydrothermal systems on slow-spreading ocean ridges: American Geo-physical Union Geophysical Monograph 188, p. 153–173.

Sauter, D., Mendel, V., Rommevaux-Jestin, C., Par-son, L.M., Fujimoto, H., Mével, C., Cannat, M., and Tamaki, K., 2004, Focused magma-tism versus amagmatic spreading along the ultra-slow spreading Southwest Indian Ridge: Evidence from TOBI side scan sonar imag-ery: Geochemistry Geophysics Geosystems, v. 5, Q10K09, doi:10.1029/2004GC000738.

Sauter, D., Cannat, M., and Mendel, V., 2008, Mag-netization of 0–26.5 Ma seafl oor at the ultra-slow spreading Southwest Indian Ridge 61–67°E: Geochemistry Geophysics Geosystems, v. 9, Q04023, doi:10.1029/2007GC001764.

Sauter, D., Cannat, M., Meyzen, C., Bezos, A., Patriat, P., Humler, E., and Debayle, E., 2009, Propaga-tion of a melting anomaly along the ultra-slow Southwest Indian Ridge between 46°E and 52°20′E: interaction with the Crozet hot-spot?: Geophysical Journal International, v. 179, p. 687–699, doi:10.1111/j.1365-246X.2009.04308.x.

Shaw, W.J., and Lin, J., 1996, Models of ocean ridge lithospheric deformation: Dependence on crustal thickness, spreading rate, and segmenta-tion: Journal of Geophysical Research, v. 101, p. 17977–17993, doi:10.1029/96JB00949.

White, R.S., Minshull, T.A., Bickle, M.J., and Rob-inson, C.J., 2001, Melt generation at very slow-spreading oceanic ridges: Constraints from geochemical and geophysical data: Journal of Petrology, v. 42, p. 1171–1196, doi:10.1093/petrology/42.6.1171.

Whittaker, J.M., Muller, R.D., Roest, W.R., Wessel, P., and Smith, W.H.F., 2008, How superconti-nents and superoceans affect seafl oor rough-ness: Nature, v. 456, p. 938–941, doi:10.1038/nature07573.

Manuscript received 22 December 2010Revised manuscript received 15 April 2011Manuscript accepted 27 April 2011

Printed in USA

on September 13, 2011geology.gsapubs.orgDownloaded from