with detailed lava flow mapping by manned submersible
and remotely operated vehicle (e.g., Soule et al, 2007) provided data to reconstruct eruption dynamics and the temporal and spatial evolution of a lava flow field. For
example, detailed study of a lava flow erupted at 9
50
0 N
EPR showed that this long fissure eruption began in summer 2005, with subsequent smaller eruptive pulses from
ever-shortening fissures, culminating in a small eruption
in mid-late January 2006 (Rubin et al., 2015), with eruption pulses being correlated in time with hydrothermal
fluid exit temperatures recorded in situ. Paleomagnetic
intensity dating of lavas (e.g., Carlut et al., 2004; Bowles
et al., 2006) and radiocarbon dating of planktonic and benthic foraminifera collected from lava flow tops (Clague
et al., 2013) are two other methods that both show promise
for developing site-specific century to millennial eruption
histories.
Mantle compositions and the spatial scales of
mantle heterogeneities
Many researchers over the past several decades have been
motivated to use MORB magmas to probe upper mantle
compositional variations because of several factors that
should simplify the connection of erupted magmas to their
mantle source, including the common occurrence of relatively undifferentiated magmas, the thin and compositionally less variable oceanic crust (compared to the
continents), and the broad spatial coverage of the upper
mantle by mid-ocean ridges, which span much of the
globe. Such studies have revealed many first-order variations in mantle compositions, as discussed above. More
recently, studies addressing the physical nature and distribution of heterogeneities have shown that a few percent of
highly fusible mantle pyroxenite veins can dominate
aggregate melt composition but not melt flux (e.g., Russo
et al, 2009) and that isotopic compositional domains along
parts of the ridge system reflect stretching, thinning, and
folding of compositionally distinct domains by mantle
convection (Graham et al., 2006). Spectral analysis of
MORB chemistry along the Mid-Atlantic Ridge has
revealed mantle heterogeneity arising from modern
hotspots as well as ancient heterogeneities preserved by
incomplete mantle mixing during convection (Agranier
et al., 2005) and perhaps different mantle-upwelling patterns beneath Indian and Atlantic mid-ocean ridges compared to those in the Pacific (Meyzen et al., 2007). New,
high-quality trace element data sets from relatively large
numbers of MORB have revealed variations in the degree
of depletion of the MORB source mantle around the world
(Arevalo and McDonough, 2010; O’Neill and Jenner,
2012) and a more refined definition of and characterization of “normal” MORB (aka N-MORB), as well as
depleted and enriched MORB variants in a geographic
framework (Gale et al., 2013).
Such studies will help researchers with their goal of
understanding the spatial scale of mantle composition
and lithology variations and how they are expressed in
erupted MORB compositions. However, over the past
decade, researchers have also come to realize that magmatic process produces significant overprinting of mantle
compositional variations in MORB (e.g., Rubin and
Sinton, 2007; Rubin et al., 2009; O’Neill and Jenner,
2012). For instance, the magnitude and temporal variability of melt supply largely affects the degree to which mantle heterogeneity is expressed in MORB magmas (Rubin
and Sinton, 2007), which among other things produces
reduced variance of mantle parameters in intermediate to
high magma supply Pacific ridges compared to other
ocean basins, an effect incorrectly assumed to reflect mantle heterogeneity or upwelling patterns in several of the
aforementioned studies. The coupling of apparent mantle
compositional homogenization with increased magma differentiation, which can be observed over a number of
length scales (Rubin et al, 2009), implies that much of this
mixing happens at shallow crustal levels, where magmas
pool and differentiate in melt-rich lenses prior to eruption.
O’Neill and Jenner (2012) argue that disruption of trace
element variations in MORB occurs in frequently
replenished, open-system magma chambers, although
other authors have also argued that magma-rock reactions
during intergranular porous flow deeper in the crust also
modify the signatures of mantle source characteristics in
the erupted MORB (Lissenberg et al., 2013). New
advances in this area will likely require more eruption
and volcano-scale studies using geological mapping, sampling, and chemical analysis to sort out the full range of
processes that operate and under which conditions they
occur.
Another way to get at the spatial scale of mantle composition and lithology variations is to look at mantle rocks
themselves where they are exposed on the seafloor or in
ophiolites. The mantle beneath mid-ocean ridges is
thought to be largely depleted by melt prior extraction
and magmatism, leading to, for instance, continental
growth. But within this depleted matrix likely also lie
veins and domains of enriched material injected into the
mantle at subduction zones and slowly mixed and spread
around by mantle convection. Studies of mantle rocks at
ridges have shown that (a) they range in composition,
from very melt depleted on ridges near hotspots to less
so at locations away from hotspots (e.g., Dick et al.,
1984), (b) a large proportion of the upwelling mantle
beneath some ridges contributes little magma to the ridge
axis because of extreme prior melt extraction (e.g., Liu
et al., 2008), and (c) the average mantle beneath ridges
may be far more isotopically depleted and compositionally more variable than inferred from erupted oceanic
basalts (e.g., Stracke et al., 2011).
Eruption detection and response
Multidisciplinary studies at several focus sites around the
world illustrate the important interplays between
magmatism, hydrothermal activity, and biological processes at active and dormant mid-ocean ridge volcanoes
MID-OCEAN RIDGE MAGMATISM AND VOLCANISM
509
and remotely operated vehicle (e.g., Soule et al, 2007) provided data to reconstruct eruption dynamics and the temporal and spatial evolution of a lava flow field. For
example, detailed study of a lava flow erupted at 9
50
0 N
EPR showed that this long fissure eruption began in summer 2005, with subsequent smaller eruptive pulses from
ever-shortening fissures, culminating in a small eruption
in mid-late January 2006 (Rubin et al., 2015), with eruption pulses being correlated in time with hydrothermal
fluid exit temperatures recorded in situ. Paleomagnetic
intensity dating of lavas (e.g., Carlut et al., 2004; Bowles
et al., 2006) and radiocarbon dating of planktonic and benthic foraminifera collected from lava flow tops (Clague
et al., 2013) are two other methods that both show promise
for developing site-specific century to millennial eruption
histories.
Mantle compositions and the spatial scales of
mantle heterogeneities
Many researchers over the past several decades have been
motivated to use MORB magmas to probe upper mantle
compositional variations because of several factors that
should simplify the connection of erupted magmas to their
mantle source, including the common occurrence of relatively undifferentiated magmas, the thin and compositionally less variable oceanic crust (compared to the
continents), and the broad spatial coverage of the upper
mantle by mid-ocean ridges, which span much of the
globe. Such studies have revealed many first-order variations in mantle compositions, as discussed above. More
recently, studies addressing the physical nature and distribution of heterogeneities have shown that a few percent of
highly fusible mantle pyroxenite veins can dominate
aggregate melt composition but not melt flux (e.g., Russo
et al, 2009) and that isotopic compositional domains along
parts of the ridge system reflect stretching, thinning, and
folding of compositionally distinct domains by mantle
convection (Graham et al., 2006). Spectral analysis of
MORB chemistry along the Mid-Atlantic Ridge has
revealed mantle heterogeneity arising from modern
hotspots as well as ancient heterogeneities preserved by
incomplete mantle mixing during convection (Agranier
et al., 2005) and perhaps different mantle-upwelling patterns beneath Indian and Atlantic mid-ocean ridges compared to those in the Pacific (Meyzen et al., 2007). New,
high-quality trace element data sets from relatively large
numbers of MORB have revealed variations in the degree
of depletion of the MORB source mantle around the world
(Arevalo and McDonough, 2010; O’Neill and Jenner,
2012) and a more refined definition of and characterization of “normal” MORB (aka N-MORB), as well as
depleted and enriched MORB variants in a geographic
framework (Gale et al., 2013).
Such studies will help researchers with their goal of
understanding the spatial scale of mantle composition
and lithology variations and how they are expressed in
erupted MORB compositions. However, over the past
decade, researchers have also come to realize that magmatic process produces significant overprinting of mantle
compositional variations in MORB (e.g., Rubin and
Sinton, 2007; Rubin et al., 2009; O’Neill and Jenner,
2012). For instance, the magnitude and temporal variability of melt supply largely affects the degree to which mantle heterogeneity is expressed in MORB magmas (Rubin
and Sinton, 2007), which among other things produces
reduced variance of mantle parameters in intermediate to
high magma supply Pacific ridges compared to other
ocean basins, an effect incorrectly assumed to reflect mantle heterogeneity or upwelling patterns in several of the
aforementioned studies. The coupling of apparent mantle
compositional homogenization with increased magma differentiation, which can be observed over a number of
length scales (Rubin et al, 2009), implies that much of this
mixing happens at shallow crustal levels, where magmas
pool and differentiate in melt-rich lenses prior to eruption.
O’Neill and Jenner (2012) argue that disruption of trace
element variations in MORB occurs in frequently
replenished, open-system magma chambers, although
other authors have also argued that magma-rock reactions
during intergranular porous flow deeper in the crust also
modify the signatures of mantle source characteristics in
the erupted MORB (Lissenberg et al., 2013). New
advances in this area will likely require more eruption
and volcano-scale studies using geological mapping, sampling, and chemical analysis to sort out the full range of
processes that operate and under which conditions they
occur.
Another way to get at the spatial scale of mantle composition and lithology variations is to look at mantle rocks
themselves where they are exposed on the seafloor or in
ophiolites. The mantle beneath mid-ocean ridges is
thought to be largely depleted by melt prior extraction
and magmatism, leading to, for instance, continental
growth. But within this depleted matrix likely also lie
veins and domains of enriched material injected into the
mantle at subduction zones and slowly mixed and spread
around by mantle convection. Studies of mantle rocks at
ridges have shown that (a) they range in composition,
from very melt depleted on ridges near hotspots to less
so at locations away from hotspots (e.g., Dick et al.,
1984), (b) a large proportion of the upwelling mantle
beneath some ridges contributes little magma to the ridge
axis because of extreme prior melt extraction (e.g., Liu
et al., 2008), and (c) the average mantle beneath ridges
may be far more isotopically depleted and compositionally more variable than inferred from erupted oceanic
basalts (e.g., Stracke et al., 2011).
Eruption detection and response
Multidisciplinary studies at several focus sites around the
world illustrate the important interplays between
magmatism, hydrothermal activity, and biological processes at active and dormant mid-ocean ridge volcanoes
MID-OCEAN RIDGE MAGMATISM AND VOLCANISM
509