G.
R. Foulger
Dept.
Geological Sciences, University of Durham, Durham
DH1 3LE, U.K.
g.r.foulger@durham.ac.uk
The ideas described herein arise from several
years of brainstorming with Don Anderson, Warren Hamilton,
Anders Meibom, Jim Natland, Mike O'Hara, Dean Presnall,
Hetu Sheth, Alan Smith, Seth Stein, Peter Vogt and
Jerry Winterer.
The plume
hypothesis in perspective
Deep mantle plumes [Morgan,
1971] were proposed in 1971 to explain:
-
intraplate volcanism, and unusually
large-volume volcanism on spreading plate boundaries
(“hot spots”),
-
time-progressive volcanic chains,
and
-
the apparent relative fixity
of these “hot spots”.
Empirical associations were later made
between proposed plume volcanism and:
-
ocean-island basalt (OIB) geochemistry,
and
-
high 3He/4He
isotope ratios.
Later work modelled convection using
fluids in tanks and numerical methods and predicted
that the arrival at the surface of a new plume would
be preceded by 0 – 3 km of uplift, followed by
the eruption of a large igneous province (LIP) (representing
the “head” of a mushroom-shaped thermal
upwelling). The formation of a time-progressive volcanic
chain would follow, and was caused by the “mushroom
(plume) stem”.
Subsequent work has shown that the
volumes of magma produced are highly variable and may
be modest, e.g., at the Louisville chain. Volcanic
lineations are often not time-progressive, or are intermittently
so, e.g., the Line Islands, and “hot
spots” are not fixed relative to one another between
plates. OIB geochemistry is widespread throughout the
Pacific, occurs on normal spreading ridges and in back
arc environments, and may be entirely explained as coming
from shallow depths. High 3He/4He
isotope ratios may or may not be observed at “hot
spots”.
Evidence for uplift preceeding LIP
emplacement is variable (e.g., Deccan
and the Siberian Traps),
few LIPs have time-progressive volcanic chains emanating
from them (e.g., Ontong Java does not), and
few chains have LIPs at their oldest end (e.g.,
the Emperor-Hawaiian chain
does not).
Tectonic associations occur at individual
“hot spots” that are not predicted or explained
by the plume hypothesis, and are not persuasively attributed
to coincidence, e.g., the presence of parallel
spreading ridges near Easter island and in Iceland,
the location of the Yellowstone-Eastern Snake River
Plain province between the North American craton and
the Basin Range province, and the location of many “hot
spots” on spreading ridges or at triple junctions.
In short, there is great variability amongst “hot
spots”.
The prediction shortcomings and inability
of the plume hypothesis to explain auxiliary observations
have required immunisation of the theory against the
absence of almost every predicted primary characteristic,
and a special case for almost every individual “hot
spot”. This progressive elaboration of plume theory
has reached a point where many scientists feel that
plumes have become a data-independent a-priori
assumption rather than a valid, testable scientific
hypothesis (see also “Plumes,
Plates and Popper”). Dissatisfaction with
this, and with the attribution of remarkable tectonic
associations to coincidence, have led scientists in
recent years to question the hypothesis and investigate
alternatives.
The primary predictions of the plume
hypothesis are:
- vertical structures extending from the surface down
to the core-mantle boundary at ~ 3,000 km depth, and
- anomalously high mantle temperatures.
A plume-like thermal upwelling is expected
to rise from a layer that is heated from below - a “thermal
boundary layer”. The only such layer known in
the deep mantle is the core-mantle boundary and this
is thus expected to be the origin of plumes. However,
there is little evidence for f) (Dziewonski,
2003), though seismic anomalies extending down to
the base of the upper mantle at ~ 660 km are observed
beneath some “hot spots” [Ritsema &
Allen, 2003]. This has led to the suggestion that
plumes rise from the 660-km discontinuity. However,
this is a mineralogical phase change and there is no
evidence that it is a thermal boundary layer. Such a
source location would not anyway satisfy the requirement
for approximate relative fixity that motivated the original
sourcing of plumes in the “deep mantle”
below the level of relatively rapid convection assumed
to be associated with plate tectonics. g) is not confirmed
by heat flow measurements, and while some petrological
evidence has been cited for a hot source below some
“hot spots” (e.g., Gudfinnsson
& Presnall,
2002), most erupted material is similar to normal
mid-ocean ridge basalt (N-MORB) and provides no evidence
for a source with elevated temperature. The term “hot
spot” thus implies an underlying assumption that
is not, as yet, convincingly confirmed, and “hot
spots” are more accurately described as volcanic
anomalies.
Any alternative hypothesis should be
able to explain a) – e), along with the seismic,
heat flow and geothermometry observations that bear
on structure, flow and temperature beneath volcanic
anomalies. Furthermore, if it is to do better than the
plume hypothesis an alternative should also be able
to explain the variability of volcanic anomalies and
the tectonic associations currently attributed to coincidence
in the plume model.
An alternative
– Plate Tectonic Processes
Summary
The theory of Plate Tectonic Processes
(PTP) (also called “Platonics”;
see also “If
not plumes – what else?”) proposes that
volcanic anomalies are “by products” of
plate tectonics. The most important elements are:
- intraplate deformation, that results from
the non-rigidity of plates, and
- compositional variability in the upper mantle
resulting from de-homogenising processes at ridges
and subduction zones.
Simply put, volcanism occurs where
lithospheric extension allows melt to leak up to the
surface. The location of the volcanism is governed by
the stress field in the plate and the amount of melt
is governed by the fusibility of the mantle beneath.
This theory views volcanism as resulting from lithospheric
processes rather than from a heat influx from below,
at the core-mantle boundary. It predicts that volcanic
anomalies and their geochemistry are shallow sourced
and related structures do not extend very deep into
the mantle.
Intraplate deformation
Earth’s tectonic plates are in
reality, not rigid. They move coherently, but deform
internally in response to changes in stress that may
result from changes in the plate boundary configuration.
The Basin Range Province in the western USA and the
East African Rift are examples. Basin Range extension
in the western USA onset in the late Cenozoic when the
North American plate overrode the East Pacific Rise
and the plate boundary there changed from being subduction
to transform type (see animations).
The East African rift formed at ~ 30 Ma, when the eastern
boundary of the African plate changed as a result of
a ridge jump in the Indian Ocean [Burke, 1996].
Intraplate deformations are most likely to occur along
old sutures or other lines of weakness, e.g., at Yellowstone.
Compositional variability in the
upper mantle
Plate tectonics dehomogenises the mantle
by extracting fusible melt at ridges and returning it,
in the form of basalt, back into the mantle at subduction
zones, along with sub-adjacent melt-depleted abyssal
peridotite, lithospheric mantle, accumulated pelagic
and terrigenous sediments and volatiles including H2O
and CO2. Subducting slabs probably do not
sink deep into the lower mantle, which is density stratified,
and geochemical recycling is thus largely confined to
the shallow mantle. Continental collisions trap young,
hot, buoyant slabs, back-arc basins, and arc and mantle
wedge material in sutures. Subsequent continental breakup
often occurs along old sutures and the trapped material,
along with delaminated continental mantle lithosphere,
is recycled into the mantle beneath the new ocean.
Eclogite – the high-pressure
form of basalt, has a much lower solidus and liquidus
than peridotite and may be completely molten at temperatures
below the peridotite solidus. Even a 30-70 mixture of
eclogite and peridotite can produce several times more
melt than peridotite alone [Yaxley, 2000].
Subducted slabs warm up to ambient mantle temperatures
on a time-scale of tens of millions of years by conduction
of heat from the surrounding mantle and radiogenic
decay, after which time they may exist in the upper
mantle in a state of nascent melting, or even containing
a small degree of melt. Slabs may be imbricated or have
a steeply dipping aspect, which increases the melt extraction
potential at a given location. The tapping of such fertile
mantle domains can thus potentially explain the large
volumes of melt produced at some volcanic anomalies
at temperatures similar to those beneath ridges.
It is unknown how much melt could pond
beneath the lithosphere prior to eruption but the huge
volumes (up to ~ 6 x 107 km3)
and magma production rates (possibly exceeding 6 km3/yr
for millions of years) [Coffin & Eldholm,
1994; Gladczenko et al., 1997] of some LIPs
are difficult to explain by any mechanism without appealing
to significant ponding. Such source materials can potentially
explain OIB geochemistry [e.g., Hofmann
& White, 1982]. Compared with MORB, OIB geochemistry
is characterised by enrichment in light-rare-earth-elements
and incompatible elements such as U and Sr, and isotope
ratios indicative of an older source. These features
could result from recycling aged subducted oceanic crust
or continental lithosphere with their diverse lithologies,
associated sediments and mantle wedge material. The
wide distribution of OIB geochemistry throughout the
Pacfic and its occasional occurrence at ridges, suggests
that it may also be widespread at the lithosphere-asthenosphere
boundary, where it has been termed the “perisphere”
[Anderson,
1995]. Recycling in the shallow mantle is less astonishing
than the requirement of plume theory that fusible subducted
crust is transported, unmelted, all the way to the 3,000-km-deep
core-mantle boundary, where it is swept up in plumes
and restored to the surface. High 3He/4He
isotope ratios may arise from helium that has been stored
in an environment low in U+Th relative to [He]. Old
dunite-rich cumulates in subducted crust, or mantle
residuum could provide such an environment (see Helium
Fundamentals page).
Predictions
of PTP theory
The apparent fixity of volcanic
anomalies
How does this model account for the
apparent fixity of volcanic anomalies? First, what are
the observations? Few long volcanic chains are well
dated, the Emperor-Hawaiian chain being perhaps the
best. Many other chains are constrained by only a few
dates or none. For example, the Iceland volcanic anomaly
has no time-progressive track. The available data show
that the apparent relative fixity of volcanic anomalies
is, to a first order, confined to individual plates
or plate boundaries. Some Pacific volcanic anomalies
move coherently relative to some on the African plate
or its boundary at a rate of ~ 3 cm/year [Raymond
et al., 2000; Tarduno & Gee, 1995].
The requirement for this is readily understandable from
the fact that a disproportionate number of volcanic
anomalies lie on or very close to plate boundaries (e.g.,
Iceland, Azores, Tristan, Easter, Samoa, Galapagos and
Afar), and plate boundaries must migrate with respect
to one another since some plates are growing and others
shrinking.
PTP theory predicts that volcanic anomalies
occur where stress in the lithosphere is extensional.
Extensional regions may exist in the interiors of plates
as a result of their boundary configurations, e.g.,
encircling subduction zones or differential lithospheric
cooling. They are expected to remain quasi-fixed relative
to the overall plate stress field, which will remain
relatively constant while the boundary configuration
remains stable. In this way, intraplate volcanic anomalies
in a single plate may remain approximately fixed relative
to one another through their stablilty relative to the
overall plate stress field. Time progression then results
from motion of the plate relative to its boundary as
a whole.
This model does not require that a
set of volcanic anomalies on one plate remains fixed
relative to a set on another plate. It predicts, however,
re-organisations of volcanic anomalies when plate boundary
configurations change radically, e.g., when
continents collide, subduction zones change into transform
faults with changes in plate motion direction, and ridges
are subducted. Global plate boundary reorganisation
occurred, for example, at the end of the Mesozoic. Gradual
changes in plate boundary configuration are predicted
to result in gradual evolution of the intraplate stress
field and gradual migration of loci of extension. This
would be manifest by melt anomalies migrating at speeds
different from the plate rate. Stresses induced by local
loading may be responsible for local volcanic anomalies
[e.g., Hieronymus & Bercovici,
1999] that do not exhibit time progression, large volumes,
longevity or fixity, e.g., the Pukapuka
ridge. The common occurrence of volcanic anomalies
on or close to spreading ridges is predicted, since
these are zones of extension.
Size distribution of volcanic anomalies PTP theory predicts that many aspects
of volcanic anomalies will be distributed over a wide
spectrum, e.g., the volumes of melt produced, the spatial
and temporal distributions of volcanism and the lateral
and vertical extents of the sources and underlying structures
detectable by seismic tomography. This contrasts with
the bimodal distribution predicted by the plume hypothesis
which attributes volcanism to two separate, unrelated
sources – shallow plate tectonics on the one hand
and deep mantle plumes on the other.
A simplifying
theory of mantle convection
PTP theory views magmatic fecundity
as being a function of lithospheric extension and mantle
fertility. This amounts to a continuum view of Earth
in contrast to a bimodal one. Instead of viewing Earth
as being surfaced by rigid plates that move relative
to one another only at plate boundaries, deformation
is recognised to be a continuum with intraplate deformation
sometimes exceeding the rates of motion at slow plate
boundaries. Instead of assuming Earth’s mantle
to be essentially homogenous, it is viewed as having
varying physical properties, including composition,
fusibility, fertility, temperature and state. Earth
is maintained in this state by stresses imposed on the
interiors of plates, for example, by plate boundaries
whose configurations may change temporally, and by processes
at plate boundaries that continually de-homogenise the
upper mantle.
A criticism leveled at PTP theory is
that it is more complex than simple, elegant plume theory.
However, volcanic anomalies are not all simple and alike
– there is great variability between them. Hawaii,
Yellowstone and Iceland,
for example, are fundamentally different in many respects.
This variability is problematic to explain in plume
theory. PTP theory offers a unifying theory
of volcanism and convection in Earth’s mantle.
In the framework of plume theory, two separate, independent
modes of convection are proposed – plate tectonics
on the one hand, accounting for volcanism at subduction
zones and spreading plate boundaries, and plumes on
the other, to which intraplate volcanism is attributed.
On-ridge volcanic anomalies, with their spectrum of
variability from large-volume, long-lived features to
small-volume, short-lived phenomena, or even mere isolated
ridge segments with slightly more OIB-like geochemistry,
sit uneasily in between, along with the plethora of
mid-ocean seamounts, many unassociated with time-progressive
lineations but nevertheless also capped with OIB. The
present theory suggests that there is only one mode
of convection in Earth – that driven by plate
tectonics, and that this can explain most volcanism
on Earth, including that currently attributed to plumes.
The Plate Hypothesis applied to various melting anomalies
Yellowstone: Yellowstone time-progressive volcanism results from time-progressive extension
-
-
Burke, K., The African plate,
S. Afr. J. Geol., 99,
341-409, 1996.
-
Coffin, M.F., and O. Eldholm,
Large igneous provinces: crustal structure, dimensions
and external consequences, Rev. Geophys.,
32, 1-36, 1994.
-
Gladczenko, T.P., M.F. Cofin,
and O. Eldholm, Crustal structure of the Ontong
Java Plateau: Modeling of new gravity and existing
seismic data, J. geophys. Res., 102,
22,711-22,729, 1997.
- Gudfinnsson,
G.H., and D.C. Presnall, The minimum potential temperature
of the Hawaiian mantle is about 1420°C, EOS
Trans. AGU Fall Meet. Suppl. Abstract, 83,
V52D-06, 2002.
-
Hieronymus, C.F., and D. Bercovici,
Discrete alternating hotspot islands formed by interaction
of magma transport and lithospheric flexure, Nature,
397, 604-607, 1999.
-
Hofmann, A.W., and W.M. White,
Mantle plumes from ancient oceanic crust, Earth
planet. Sci. Lett., 57, 421-436,
1982.
- Meibom, A.,
and D.L. Anderson, Statistical upper mantle assemblage,
Earth planet. Sci. Lett., 217,
123-139, 2003.
-
Morgan, W.J., Convection plumes
in the lower mantle, Nature, 230,
42-43, 1971.
-
Raymond, C.A., J.M. Stock, and
S.C. Cande, Fast Paleogene motion of the Pacific
hotspots from revised global plate circuit constraints,
in History and Dynamics of Plate Motions,
edited by M.A. Richards, R.G. Gordon, and R.D. van
der Hilst, pp. 359-375, 2000.
-
Ritsema, J., and R.M. Allen,
The elusive mantle plume, Earth planet. Sci.
Lett., 207, 1-12, 2003.
-
Tarduno, J.A., and J. Gee, Large-scale
motion between Pacific and Atlantic hotspots, Nature,
378, 477-480, 1995.
-
Yaxley, G.M., Experimental study
of the phase and melting relations of homogeneous
basalt + peridotite mixtures and implication for
the petrogenesis of flood basalts, Cont. Min.
Pet., 139, 326-338, 2000.
last modified 9th October, 2016 |