Download Mobility and immobility of mid-ocean ridges and their

Mobility and immobility of mid-ocean ridges and their
implications to mantle dynamics
D.C.P. Masalu
Institute of Marine Sciences, University of Dar es Salaam, P.O. Box 668, Zanzibar, Tanzania
ABSTRACT
In the past two decades, the mobility of mid-ocean ridges relative to the
mantle (absolute migration) has been correlated with major observable
features, such as spreading asymmetry and asymmetry in the abundance
of seamounts. The mobility of mid-ocean ridges is also thought to be an
importantfactorthat influences the diversity of ridge-crest basalts. However,
this mobility has not yet been defined and mapped.The absolute migration
ofglobal mid-ocean ridges since 85 Ma has been computed and mapped.
Global mid-ocean ridges have migrated extensively at varying velocities
during that period. Presently, the fast-migrating ridges are the PacificAntarctic, Central Indian Ridge, Southeast Indian Ridge, Juan de Fuca,
Pacific-Nazca, Antarctic-Nazca, and the Australia-Antarctic ridges, migrating
at velocities of between 3.3 and 5.5 cm/yr. The slow-migrating ridges are
the Mid-Atlantic and the Southwest Indian ridges, migrating at velocities
between 0.3 and 2.0 cm/yr. Comparison of these results with mantle
tomography resultsshowthat the slow-migrating ridges have deeper depth
of origin than the fast-migrating ones, suggesting a correlation between
the absolute migration velocity and the depth of origin of ridges.
Furthermore, the Southwest Indian ridge appears to be tapping the same
portion of mantle as did the Central Indian ridge. These results have
important thermo-chemical implications, such as variations in the extent
of melting and mineralogical composition ofthe mantle beneath different
ridges, which may influence mantle dynamics.
INTRODUCTION
Mid-ocean ridges are important features of the earth for several reasons: in plate
tectonics they are boundaries of plates and play a key role in driving them; and in
geophysics they are locations along which new seafloor is created. Absolute migration
of mid-ocean ridges have beencorrelated with major observable features ofthe ridges.
For example Stein et al. (1977) correlated spreading asymmetry with migration rate,
and Davis and Karsten (1986) explained asymmetry in seamountabundance byabsolute
77
ridge migration. Ridge migration rate is also thought to be an important factor that
influences the diversity of ridge-crest lavas and the compositional uniformity of ridgecrest basalts (Davis and Karsten, 1986). Although absolute migration of mid-ocean
ridges has been linked with many features, its influence to mantle dynamics has not
yet been investigated.
In this paper, therefore, the absolute migration of global mid-ocean ridges since
85 Ma is computed and its possible influence to mantle dynamics investigated.
methods
Maqnetic lineaBons fbrni concurrendy with new seafloor on mid-ocean ridges when
passively
Magnetic domains/miner
seafloor. Fifteen ridoe
geomagnetic field,
are selected (Figure 1),
sides. Their
corresponding magnetic lineation segmenS'^K^'^
gg |via to Present are reconstructed by rotatinn ^
^ their former positions (Masalu and Tamaki 1994^^'^®^
magnetic lineations bacK w
the models of absolute moti
Clague (1985). The absolute m-g
computed.
j^iuller and Smith (1993) and n./ "^'"9
velocities for each respective ridae
ndge are then
.magnetic lineations were obtained from aCD-rqm
Locations of identified ^g ,.^gations (Cande etal., 1989). Ages were f
data of locations ofglobal magn
were ass.gn^^
70°N
60°N
30°N
30°S
eo-s
'
^
70°S
120'W
60°W
78
to identified magnetic lineations based on a recent geomagnetic polarity time scale for
the Late Cretaceous and Cenozoic time (Cande and Kent, 1992).
The assumption in the methodology is that plates have remained rigid over the
past 85 million years. This Is certainly not completely correct because both inter- and
intra-plate deformations are known to exist. Errors may also arise from incorrectly
identified magnetic lineations. There are also inherent errors from the use of models of
absolute plate motions.
RESULTS
Migration of global mid-ocean ridges
Regardless of the possibility of the errors described above, the results present a first
overview of global motion of mid-ocean ridges. The results obtained indicate that
global mid-ocean ridges have migrated extensively in the past 83 million years (Table
1and Figure 2). All ridges appear to be migrating: the East Pacific Rise (EPR) appears
to have rotated clockwise for about 50° since 83Ma (Figure 2i). The South Mid Atlantic
Ridge (SMAR) shows small lateral migration but indicates extensive longitudinal migration
toward north (Figure 2ii). Similarly, the northern EPR and the Juan de Fuca ridge
segments appear to have migrated northerly at least since about 83Ma (Chron 34) to
the Present (Figure 21).
The SMAR runs between several hotspots located relatively close to it on both
sides which probably played a role in limiting its lateral migration (Uyeda and Miyashiro,
1974).
Contrary to theSMAR, the Central Mid Atlantic Ridge (CMAR) between the equator
and latitude 40° N, and the Northern Mid Atlantic Ridge (NMAR) north of the 40° N
latitude, has been migrating northwesterly since 83Ma to Present (Figure 2ii). This
suggests the existence of active deformation between the SMAR and the CMAR at least
since 83Ma. The present results may offer one explanation of the origin of tectonic
Table 1. Migration of global mid-ocean ridges
Identifier
Ridge
Distance
Time to Present
migrated (km)
(millionyears)
JUAN
Juan de Fuca
3900
PacAnt
Pacific-Antarctic
2400
PacNazca
Pacific-Nazca
Antarctic-Nazca
200
12
1010
20
South Mid-Atlantic Ridge
Central Mid-Atlantic Ridge
North Mid-Atlantic Ridge
1300
83
1200
83
1550
Southeast Indian Ridge
Central Indian Ridge
Southwest Indian Ridge
Australia-Antarctic
3900
83
83
AntNaz
SMAR
CMAR
NMAR
SEIR
QR
SWIR
AustAntar
79
83
73
3100
1250
83
1400
43
64
180°
60°
240'
210'
270°
300°
•K,
60°
%
.'''•v.
V
>17
JUAN-
30°
30°
Vil'
2«
^27
• 30
•'~t.
^33
^34
• PacNazca
-30°
-30°
PacAnl
iiiii. I
AntNaz
r
PacAnt.
-60°
-
-60°
-^.32
180°
210°
270°
300'
Figure 2 (i)
gure 2. i^igration of mid-ocean ridges for some of the selected ridge segments. Solid thick
ne segments are reconstructed zero-age paleopositions ofselected ridge segments at a Chron
age indicated by the adjacent numbers. Text with a pointing arrow is ridge identifier as in Table
. UHabove) Pacific ridges (Juan de Fuca and South Pacific ridges), (ii) (top, opposite page)
rrip" cc?o (SMAR, CMAR and NMAR), (iii) (bottom, opposite page) Indian Ocean ridges
on fh
r
Notesame
that locality
the CIRatand
SWIR times
(line connecting solid squares
tne figure,
and inset)AustAntar).
sampled the
different
80
—i
T-
/
;s\,S-5
60" -
NMAR
-H
60"
330°
340°
350°
B
-10°
C=:!5j
CMAR
1
1-
30° -
/
w
•20°
1,
340^
350"
-40* —
iivi
V
,1'', SMAR
r'
-50° •
I
30° -
ff*
300°
^ 330'
340°
Figure 2 (ii)
30°
T^,
S.
(2^
"•'C
0°
a
V
5 I
•30°
- - vV'
^
-SEIR
J^' "
SWIR—♦! »
AustAniar
\
-60'
f-Niv. r
30°
60°
90°
Figure 2 (iii)
81
120°
150°
compl^ty ,n tt,e equatonal Mantic r^ion, an area characterised by adense pattern
of mosuy m^ium to laqe offeet fracture zones as well as aseries of unusual ridges
and troughs (Mueller and Smith, 1993). These investigators have also suggested that
^the nNorth and1Atlanta
region recorded
the migration
of evidence
the plate from
bounda^
between
South Amerian
plates. Based
mainly on
fracture
zone
trends, It has been proposed that the boundary was located in the equatorial Atlantic
Cretaceous and Late Tertiary (Roest,
1987; Roest and Collette, 1986) periods.
paleopositions of
fte Central Indian Ridge (QR) between Chron 29 and Chron 25 »iat
coincide
with those of
the Southwestern Indian Ridge (SWIR) between Chron 20 and Present (Figure 2iii).
This implies that the SWIR may be tapping the same portion of the mantle as did the
CIR between Chron 29 and 24.
Absolute ridge migration velocities and mantle tomography
The absolute migration velocities of all ridges investigated In this study are shown in
Figure 3. The ridges divide into two groups: the slow-migrating ridge group includina
the CMAR, SMAR, NMAR and the SWIR ridges, and the fast migrating ridge group
including all other remaining ridges. Presently, slow-migrating ridges are migrating at
velocities between 0.3-2.0cm/yr whereas fast-migrating ridges are migrating at velocities
between 3.3-5.5cm/yr.
The results were compared with seismic velocity anomalies beneath global midocean. Su et al. (1992) shows that seismic velocity anomalies associated with mid12
—AustAntar
—JUAN
— NMAR
—CIR
-CMAR
-r
O
8
-
SEIR
—SMAR
—SWIR
- -PacAnt
—PacNazca
—AntNaz
o
O
i_r
:
. t""
:
^
__
n
•ii
1
,'-l •-"I"
-i-l .I-n
•L-J ^
--
,
1
r
•
.-3—:
• r-
iliniliii.i liii
Age (Ma)
Figure 3. Variation of migration velocities of global mid-ocean ridges. Text identifiers as in
Table 1
82
- -- - ^
ocean ridges appear as continuous features to 300km deptti. For the NIMAR SI^IAR,
Pacific-Antarctic ridge (PacAnt), SWIR, and the Carlsberg ridges, seismic velocities
remain slower than normal, on average down to 400l<m depth, and this situation persists
to 600l<m depth. Except for the PacAnt, the 'deep-rooted' ridges are the slow-migrating
ridges (Figure 3). As for the PacAnt, though failing in the same group with fast-migrating
ridges, it is the slowest ridge in the group. These results suggest the existence of a
correlation between absolute migration velocities and the depth of origin of mid-ocean
ridges, whereby fast-migrating ridges have shallow depth of origin and slow-migrating
ones have deeper depth of origin. Intuitively, this may be explained as follows: mantle
from the same vertical locality beneath a ridge have more time for passive upwelling
for slow-migrating ridges and less time for fast-migrating ridges. Slow-migrating ridges
allow the development of stable and deep-rooted mantle convection cells beneath
them, whereas fast-migrating ones probably cause some disturbances to mantle
convection cells beneath them, and thus allow only the development ofshallow-rooted
convection cells.
DISCUSSION
The observations that the SWIR may be tapping the same portion of the mantle as did
the CIR, and the correlation between absolute migration velocity and the depth of
origin of mid-ocean ridges has far-reaching thermal and chemical implications. One of
the most important dynamic processes in the earth's interior is thermal convection in
the mantle. Currently, two methods are used to study temperature variations in the
upper mantle: the study of major-element chemistry of basalts erupted at mid-ocean
ridgeswhich is directly influenced bythe temperature ofthe mantle beneath (e.g. Klein
and Langmuir, 1987), and seismic velocity studies (e.g. Su et al., 1992).
Klein and Langmuir (1987,1989) studied the chemical systematicsof global midocean ridges and proposed a model to explain the observed chemical systematics of
global MORB (Mid-Ocean Ridge Basalts) byvariations amongdifferent melting columns.
According to their model, a hotter parcel ofthe mantle (from deeper depths) intersects
the solidus at greater depths and produces a taller melting column, leading to greater
mean pressures and extent ofmelting, whereas a cooler parcel ofmantle (from shallow
depth) intersects the solidus at shallower depths and produces a shorter melting column,
leading to lower mean pressures and extent of melting. Their model can becombined
directly with the results of this study whereby fast-migrating ridges correspond to
ridges with cooler sources of parcels of mantle, and slow-migrating ridges correspond
to ridges with hotter sources of parcels of mantle. The combined result predicts that
the fast-migrating ridges (e.g. theSEIR) will have deeper water depths and their basalts
will have relatively high Nag,, and SiO^ with low Fe^p and CaO/AiPj, and that slow
migrating ridges (e.g. the IMMAR) will beshallow in water depth and their basalts would
have relatively low Nag j, and SiO^, but relatively high Feg ^and CaO/AiPj. Additionally,
83
these results suggest that the mantle beneath fast-migrating ridges has undergone
smaller extents ofmelting (melt extraction) whereas thatbeneath slow-migrating ridges
has undergone greater extents ofmelting (e.g. Figure 4 ofKlein and l_angmuir, 1989)
which may influence mantle convection. Furtherstudy is needed from different branches
ofgeosciences to investigate whether the 'deeper origin' ridges (Su et al., 1992) have
any peculiar characteristics from other ridges.
The observation that the SWIR may be tapping the same portion of the mantle as
did the CIR, suggests that different portions of the mantle have undergone different
phases/history ofrecycling, and thus offers one possible mechanism oforigin of lateral
heterogeneities In the mantle. Astudy which will involve entire ridges may reveal more
of such observations (Masalu and Tamaki, 1994).
CONCLUSIONS
The absolute migration of global mid-ocean ridges for the past 85 million years and its
potential role in mantle dynamics have been investigated. Absolute migration velocities
and depth of origin of ridges appear to correlate, whereby slow-migrating ridges
(migrating at 0.3-2.0cm/yr) have deeper depth of origin than fast-migrating ones
(migrating at 3.3-5.5cm/yr). The SWIR appears to be tapping the same portion of
mantle as did the CIR, a phenomenon that may explain lateral mantle heterogeneity.
Active deformation which appears to be taking place between the SMAR and CI^ar at
least since 85 Ma is one possible explanation of the origin of tectonic complexity in the
equatorial Atlantic.
ACKNOWLEDGEMENTS
Ithank Dr K. Tamaki who is my mentor at the Ocean Research Institute (ORI), University
of Tokyo, Dr S. Uyeda and Dr W. W. Sager both from Texas A&M University, for their
constructive critical reviews of another paper which is still under preparation from
which the main ideas for the present paper originated. This work was partly done when
I was a JSPS Fellow at ORI.
REFERENCES
Cande, S.C. &Kent, D.V. 1992. Anew geomagnetic polarity time scale for the Late Cretaceous
and Cenozoic. Journal of Geophysical Research 97\ 13917-13951.
Cande, S.C., LaBrecque, J.L, Larson, R.L., Pitman III, W.C., Golovchenko, X. & Haxby, w F
1989. Magnetic Lineations oftfie World's Ocean Basins. American Association of Petroleurri
Geologists.
Davis, E.E. &Karsten, J.L. 1986. On the cause of the asymmetric distribution of seamounts
about the Juan de Fuca ridge: Ridge-crest migration over a heterogeneous
asthenosphere. Earth and Planetary Science Letters 79: 385-396.
Duncan, R.A. &Clague, D.A. 1985. Pacific plate motion recorded by linear volcanic chains pp
89-1211. In Nairn, A.E.M. etal., (eds). The Ocean Basins andMargins yo\. 7A. New York:
Piihlkhina.
Plenumn Publishing
84
Klein, E.M. 8tLangmuir, C.H. 1987. Global correlations of ocean ridge basaltchemistry with axial
depth and crustal thiclcness. Journal of Geophysical Research'yi-. 8089-8115.
Klein, E.M. &Langmuir, C.H. 1989. Local versus global variations in ocean ridge basalt composition:
A reply. Journal of Geophysical Research 94: 4241-4252.
Masalu, D.C.P. & Tamaki, K. 1994. Highlights on global absolute motion and tectonics of the
mid-ocean ridge system from 85 Ma topresent. Abstract in 1994 Japan Earth and Planetary
Science Joint Meeting, p. 159.
Muller, R.D. &Smith,W.H.F. 1993. Deformation of the oceanic crust between the North American
and South American plates. Journal of Geophysical Research'iZ-. 8275-8291.
Roest, W.R. 1987. Seafloor spreading pattern of the North Atlantic between 10? and 40?N.
GeoL Ultraiectinal
121.
Roest, W.R. & Collette, B.J. 1986. The fifteen twenty fracture zone and the North AmericanSouth American plate boundary. Journal of the Geological Society of London 143- 833843.
Stein, S., Melosh, H. J. &Minster, J.B. 1977. Ridge migration andasymmetric sea-floor spreading
Earti) and Planetary Science Letters 36: 51-62.
Su, W., Woodward, R.L & Dziewonski, A.M. 1992. Deep origin of mid-ocean ridge seismic
velocity anomalies. Nature 360: 149-152.
Uyeda, S. &Miyashiro, A. 1974. Plate tectonics and theJapanese islands: Asynthesis. Geological
Society of America Bulletin 85: 1159-1170.
85