Global MORB magma compositions support the
Phipps Morgan and Chen (1994) heat supply model:
MORB major element compositions display a general
relationship between petrologic indicators of fractionation
depth and spreading rate, with evidence for generally
higher magma crystallization pressures (Sinton and
Detrick, 1992; Michael and Cornell, 1998) and magmatic
gas equilibration pressures (Paonita and Martelli, 2007)
being more common at slow-spreading ridges. A more
recent study found very regular variations in MORB
chemistry with spreading rate and AMC depth in a large,
11,000 sample group of MORB from around the globe
(Figure 2). Here, the least differentiated MORB were
found to occur at slow-spreading rates, becoming progressively more differentiated and erupting from a shallower
AMC as spreading rate increases (Rubin and Sinton,
2007). This dichotomy, that cooler magmas generally
erupt from ridges with the highest magma supply and hotter magmas from ridges with the lowest magma supply,
demonstrates the importance of the AMC and its depth
within the crust for modulating MORB composition.
At shorter temporal and spatial scales, chemical variations in MORB within and between successive historical
eruptions at sites having an AMC show that magma chamber thermal conditions were relatively constant at these
few sites over this time interval, with just 10–20
C
of magmatic heat loss over one to two decades (e.g., at
9
50
0 N East Pacific Rise, Goss et al. 2010; 17
S East
Pacific Rise, Bergmanis et al., 2007; CoAxial segment of
the Juan de Fuca Rise, Embley et al., 2000). Many of these
same eruption deposits also preserve magma temperature
gradients along the eruptive fissures that indicate thermal
variations within the AMC along axis that are not mixed
out during eruptions (e.g., Rubin et al., 2001; Sinton
et al., 2002; Bergmanis et al., 2007; Goss et al., 2010).
Evolved magma compositions
Small volumes of high-silica lavas (andesites and dacites)
occur on mid-ocean ridges, often associated with ridge
discontinuities such as propagating rift tips (e.g., Sinton
et al., 1983), overlapping spreading centers (e.g., Wanless
et al., 2010), and ridge-transform intersections (e.g.,
Schmitt et al., 2011); however, highly differentiated lavas
can also occur in more “normal” ridge settings (e.g.,
Wanless et al., 2010) and at regions showing ridge-hotspot
interaction (Haase et al., 2005; Colman et al., 2012). Many
authors have suggested a tectonic control on the occurrence of highly differentiated lavas on mid-ocean ridges,
primarily by allowing for extensive cooling of the AMC
margins near places where magma supply might be lower.
This promotes conditions that allow for extensive fractional crystallization of the magmas. Similar highly differentiated lavas also occur in Iceland, where strong shifts in
oxygen isotopic composition relative to mantle values
indicate a major role as well for the melting of altered crust
(e.g., Martin and Sigmarsson, 2007). The bulk composition of many highly evolved mid-ocean ridge lavas is
remarkably similar, and several different occurrences have
been successfully modeled by a combination of extensive
MORB fractional crystallization plus/minus partial melting and/or assimilation of seawater-altered oceanic crust
(e.g., Haase et al., 2005; Wanless et al., 2010). Radiometric dating of zircons from dacite lava domes erupted near
the intersection of the Juan de Fuca ridge and the Blanco
transform (in the NE Pacific Ocean) indicates that this differentiation can occur rapidly (over just 10 millennia or so;
Schmitt et al., 2011).
The plutonic crust
Beneath the volcanic carapace at mid-ocean ridges lies a
thick plutonic crust. This intrusive material comprises
roughly five times the volume of the extrusive (volcanic)
section (White et al., 2006). Direct in situ observation of
this crust is difficult, so that much of what we know about
this relatively inaccessible material comes from drill hole
studies. However, rare plutonic exposures on the seafloor
occur in fracture zones and other faulting-exposed “tectonic windows,” as well as in ophiolites (pieces of young
oceanic crust obducted onto land, e.g., Cann, 1974;
Pallister and Hopson,1981). Collectively, studies of these
materials indicate a large range of lower crustal rock types,
from mafic to evolved compositions, and a range of petrologic textures, some acquired during complex cooling and
melt-reaction histories (e.g., Coogan, 2007; Coogan et al.,
2007; Keleman and Aharonov, 1998; Lissenberg and
Dick, 2008; Lissenberg et al., 2009; Suhr et al., 2008).
The extent to which magmas erupted at the surface interact
with and reflect this diversity of rocks in the lower crust is
an area of active research.
Volcanic eruptions at mid-ocean ridges
Volcanic activity occurs predominantly on the axis of the
mid-ocean ridge system with largely unknown frequency.
Our understanding of how deep submarine volcanism has
built structures on mid-ocean ridges has been greatly aided
by sonar and visual mapping of submarine volcanic edifices and by sampling and compositional analysis of rocks
there to learn the types and distributions of volcanic products that occur. There is a large literature on this subject,
much of which is reviewed in Rubin et al. (2012). In summary, mid-ocean ridges produce low-lying elongate volcanoes from the combined effects of fissure eruptions of
generally low viscosity magmas and the constant rafting
away of eruption deposits due to plate separation. Bathymetric relief is lower at higher magma supply (e.g.,
faster-spreading) ridges because volcanic construction
dominates over tectonic rifting processes (e.g., Buck
et al., 1997; Macdonald, 1998). In fact, the boundaries of
individual volcanoes and individual eruptions become
more difficult to define at the fastest-spreading ridges
because magma is erupted at relatively high frequency
(e.g., Sinton et al., 2002; Bergmanis et al., 2007) from
nearly continuous along ridge melt axial magma
chambers (e.g., Singh et al., 1998) leading to axial high
MID-OCEAN RIDGE MAGMATISM AND VOLCANISM
505
Phipps Morgan and Chen (1994) heat supply model:
MORB major element compositions display a general
relationship between petrologic indicators of fractionation
depth and spreading rate, with evidence for generally
higher magma crystallization pressures (Sinton and
Detrick, 1992; Michael and Cornell, 1998) and magmatic
gas equilibration pressures (Paonita and Martelli, 2007)
being more common at slow-spreading ridges. A more
recent study found very regular variations in MORB
chemistry with spreading rate and AMC depth in a large,
11,000 sample group of MORB from around the globe
(Figure 2). Here, the least differentiated MORB were
found to occur at slow-spreading rates, becoming progressively more differentiated and erupting from a shallower
AMC as spreading rate increases (Rubin and Sinton,
2007). This dichotomy, that cooler magmas generally
erupt from ridges with the highest magma supply and hotter magmas from ridges with the lowest magma supply,
demonstrates the importance of the AMC and its depth
within the crust for modulating MORB composition.
At shorter temporal and spatial scales, chemical variations in MORB within and between successive historical
eruptions at sites having an AMC show that magma chamber thermal conditions were relatively constant at these
few sites over this time interval, with just 10–20
C
of magmatic heat loss over one to two decades (e.g., at
9
50
0 N East Pacific Rise, Goss et al. 2010; 17
S East
Pacific Rise, Bergmanis et al., 2007; CoAxial segment of
the Juan de Fuca Rise, Embley et al., 2000). Many of these
same eruption deposits also preserve magma temperature
gradients along the eruptive fissures that indicate thermal
variations within the AMC along axis that are not mixed
out during eruptions (e.g., Rubin et al., 2001; Sinton
et al., 2002; Bergmanis et al., 2007; Goss et al., 2010).
Evolved magma compositions
Small volumes of high-silica lavas (andesites and dacites)
occur on mid-ocean ridges, often associated with ridge
discontinuities such as propagating rift tips (e.g., Sinton
et al., 1983), overlapping spreading centers (e.g., Wanless
et al., 2010), and ridge-transform intersections (e.g.,
Schmitt et al., 2011); however, highly differentiated lavas
can also occur in more “normal” ridge settings (e.g.,
Wanless et al., 2010) and at regions showing ridge-hotspot
interaction (Haase et al., 2005; Colman et al., 2012). Many
authors have suggested a tectonic control on the occurrence of highly differentiated lavas on mid-ocean ridges,
primarily by allowing for extensive cooling of the AMC
margins near places where magma supply might be lower.
This promotes conditions that allow for extensive fractional crystallization of the magmas. Similar highly differentiated lavas also occur in Iceland, where strong shifts in
oxygen isotopic composition relative to mantle values
indicate a major role as well for the melting of altered crust
(e.g., Martin and Sigmarsson, 2007). The bulk composition of many highly evolved mid-ocean ridge lavas is
remarkably similar, and several different occurrences have
been successfully modeled by a combination of extensive
MORB fractional crystallization plus/minus partial melting and/or assimilation of seawater-altered oceanic crust
(e.g., Haase et al., 2005; Wanless et al., 2010). Radiometric dating of zircons from dacite lava domes erupted near
the intersection of the Juan de Fuca ridge and the Blanco
transform (in the NE Pacific Ocean) indicates that this differentiation can occur rapidly (over just 10 millennia or so;
Schmitt et al., 2011).
The plutonic crust
Beneath the volcanic carapace at mid-ocean ridges lies a
thick plutonic crust. This intrusive material comprises
roughly five times the volume of the extrusive (volcanic)
section (White et al., 2006). Direct in situ observation of
this crust is difficult, so that much of what we know about
this relatively inaccessible material comes from drill hole
studies. However, rare plutonic exposures on the seafloor
occur in fracture zones and other faulting-exposed “tectonic windows,” as well as in ophiolites (pieces of young
oceanic crust obducted onto land, e.g., Cann, 1974;
Pallister and Hopson,1981). Collectively, studies of these
materials indicate a large range of lower crustal rock types,
from mafic to evolved compositions, and a range of petrologic textures, some acquired during complex cooling and
melt-reaction histories (e.g., Coogan, 2007; Coogan et al.,
2007; Keleman and Aharonov, 1998; Lissenberg and
Dick, 2008; Lissenberg et al., 2009; Suhr et al., 2008).
The extent to which magmas erupted at the surface interact
with and reflect this diversity of rocks in the lower crust is
an area of active research.
Volcanic eruptions at mid-ocean ridges
Volcanic activity occurs predominantly on the axis of the
mid-ocean ridge system with largely unknown frequency.
Our understanding of how deep submarine volcanism has
built structures on mid-ocean ridges has been greatly aided
by sonar and visual mapping of submarine volcanic edifices and by sampling and compositional analysis of rocks
there to learn the types and distributions of volcanic products that occur. There is a large literature on this subject,
much of which is reviewed in Rubin et al. (2012). In summary, mid-ocean ridges produce low-lying elongate volcanoes from the combined effects of fissure eruptions of
generally low viscosity magmas and the constant rafting
away of eruption deposits due to plate separation. Bathymetric relief is lower at higher magma supply (e.g.,
faster-spreading) ridges because volcanic construction
dominates over tectonic rifting processes (e.g., Buck
et al., 1997; Macdonald, 1998). In fact, the boundaries of
individual volcanoes and individual eruptions become
more difficult to define at the fastest-spreading ridges
because magma is erupted at relatively high frequency
(e.g., Sinton et al., 2002; Bergmanis et al., 2007) from
nearly continuous along ridge melt axial magma
chambers (e.g., Singh et al., 1998) leading to axial high
MID-OCEAN RIDGE MAGMATISM AND VOLCANISM
505