M02_TARB6927_09_SE_C02.

QXD

50

1/11/07

CHAPTER 2

1:24 PM

Page 50

Plate Tectonics: A Scientific Revolution Unfolds

BOX 2.3  UNDERSTANDING EARTH

Priority in the Sciences

The priority, or credit, for a scientific idea
or discovery is usually given to the researcher, or group of researchers, who first
publish their findings in a scientific journal. However, it is not uncommon for two,
or even more researchers, to reach similar
conclusions almost simultaneously. Two
well-known examples are the independent
discoveries of organic evolution by Charles
Darwin and Alfred Wallace and development of the calculus by Isaac Newton and
Gottfried W. Leibniz. Similarly, some of
the main ideas that led to the tectonics revolution in the Earth sciences were also discovered independently by more than one
group of investigators.
Although the continental drift hypothesis is rightfully associated with the name
Alfred Wegener, he was not the first to suggest continental mobility. In fact, Francis
Bacon, in 1620, pointed out the similarities
of the outlines of Africa and South America; however, he did not develop this idea
further. Nearly three centuries later, in
1910, two years before Wegener made a
formal presentation of his ideas, American
geologist F. B. Taylor published the first
paper to outline the concept we now call
continental drift. Why, then, is Wegener
credited with this idea?
Because the papers authored by Taylor
had relatively little impact among the scientific community, Wegener was not aware of
Taylors work. Hence, it is believed that Wegener independently and simultaneously
reached the same conclusion. More important, however, Wegener made great efforts
throughout his professional life to provide a
wide range of evidence to support his hy-

pothesis. By contrast, Taylor appeared content to simply state, There are many bonds
of union which show that Africa and South
America were once joined. Further, whereas Taylor viewed continental drift as a
somewhat speculative idea, Wegener was
certain that the continents had drifted. According to H. W. Menard in his book, The
Ocean of Truth, Taylor was uncomfortable
having his ideas coupled with Wegeners
hypothesis. Menard quotes Taylor as writing, Wegener was a young professor of
meteorology. Some of his ideas are very different from mine and he went much further
in his speculation.
Another controversy concerning priority came with the development of the
seafloor spreading hypothesis. In 1960,
Harry Hess of Princeton University wrote a
paper that outlined his ideas on seafloor
spreading. Rather than rushing it to publication, Hess mailed copies of the manuscript to numerous colleagues, a common
practice among researchers. In the meantime, and apparently independently, Robert
Dietz of Scripps Institution of Oceanography published a similar paper in the
respected journal Nature (1961), titled
Continents and Ocean Basin Evolution by
Spreading of the Sea Floor. When Dietz
became aware of Hesss earlier, although
unpublished paper, he acknowledged priority for the idea of seafloor spreading to
Hess. It is interesting to note that the basic
ideas in Hesss paper actually appeared in
a textbook authored by Arthur Holmes in
1944. Therefore, priority for seafloor spreading may rightfully belong to Holmes. Nevertheless, Dietz and Hess both presented

together. Furthermore, along great faults, which he named

transform faults, plates slide past one another. In a broad
sense, Wilson had presented what would later be called the
theory of plate tectonics, a topic we will consider next.
Once the key concepts of plate tectonics had been set
forth, the hypothesis-testing phase moved forward very
quickly. Some of the evidence that these researchers uncovered to support the plate tectonics model will be presented
in this and other chapters. Much of the supporting evidence
for the plate tectonics model already existed. What this theory provided was a unified explanation for what seemed to
be numerous unrelated observations from the fields of geology, paleontology, geophysics, and oceanography among
others.
By the end of the 1960s the tide of scientific opinion had
indeed turned! However, some opposition to plate tectonics

new ideas that were influential to the development of the theory of plate tectonics.
Thus, historians associate the names Hess
and Dietz with the discovery of seafloor
spreading with occasional mention of contributions by Holmes.
Perhaps the most controversial issue of
scientific priority came in 1963 when Fred
Vine and D. H. Matthews published their
paper that linked the seafloor spreading
hypothesis with the newly discovered
data on magnetic reversals. However, nine
months earlier a similar paper by Canadian geophysicist, L. W. Morley was not accepted for publication. One reviewer of
Morleys paper commented, Such speculation makes interesting talk at cocktail
parties, but is not the sort of thing that
ought to be published under serious, scientific aegis. Morleys paper was eventually
published in 1964, but priority had already
been established, and the idea became
known as the Vine-Matthews hypothesis.
In 1971, N. D. Watkins wrote of Morleys
paper, The manuscript certainly had substantial historical interest, ranking as probably the most significant paper in the earth
sciences to ever be denied publication.
With the development of the theory of
plate tectonics came many other races for
priority by researchers from various competing institutions. Some of the new ideas
that unfolded from this body of work will
be presented in this and later chapters. Because priority for scientific ideas is complicated by the frequency of independent and
nearly simultaneous discoveries, it became
prudent for investigators to publish their
ideas as quickly as possible.

continued for at least a decade. Nevertheless, Wegener had

been vindicated and the revolution in geology was nearing
an end.

Plate Tectonics: The New Paradigm

Plate Tectonics
 Introduction

By 1968 the concepts of continental drift and seafloor spreading were united into a much more encompassing theory
known as plate tectonics 1tekton = to build2. Plate tectonics
can be defined as the composite of a great variety of ideas that
explain the observed motion of Earths outer shell through

M02_TARB6927_09_SE_C02.QXD

1/11/07

1:24 PM

Page 51

Plate Tectonics: The New Paradigm

FIGURE 2.18 An alternate hypothesis to continental drift was an expanding Earth. According to this model Earth was once only half its
current diameter and covered by a layer of continents. As Earth expanded
the continents split into their current configurations, while new seafloor
filled in the spaces as they drifted apart.

the mechanisms of subduction and seafloor spreading, which,

in turn, generate Earths major features, including continents,
mountains, and ocean basins. The implications of plate tectonics are so far-reaching that this theory has become the
basis for viewing most geologic processes.

Earths Major Plates

According to the plate tectonics model, the uppermost mantle, along with the overlying crust, behave as a strong, rigid
layer, known as the lithosphere 1lithos = stone, sphere =
a ball2, which is broken into pieces called plates (Figure
2.19). Lithospheric plates are thinnest in the oceans where
their thickness may vary from as little as a few kilometers at
the oceanic ridges to 100 kilometers in the deep-ocean
basins. By contrast, continental lithosphere is generally
about 100 kilometers thick but may be more than 250 kilometers thick below older portions of landmasses. The lithosphere overlies a weaker region in the mantle known as the

FIGURE 2.19

asthenosphere 1asthenos = weak, sphere = a ball2. The temperature/pressure regime in the upper asthenosphere is
such that the rocks there are very near their melting temperatures. This results in a very weak zone that permits the lithosphere to be effectively detached from the layers below.
Thus, the weak rock within the upper asthenosphere allows
Earths rigid outer shell to move.
The lithosphere is broken into numerous segments,
called lithospheric or tectonic plates, that are in motion
with respect to one another and are continually changing in
shape and size. As shown in Figure 2.20, seven major lithospheric plates are recognized. They are the North American,
South American, Pacific, African, Eurasian, Australian-Indian,
and Antarctic plates. The largest is the Pacific plate, which
encompasses a significant portion of the Pacific Ocean
basin. Notice from Figure 2.20 that most of the large plates
include an entire continent plus a large area of ocean floor
(for example, the South American plate). This is a major departure from Wegeners continental drift hypothesis, which
proposed that the continents moved through the ocean
floor, not with it. Note also that none of the plates are defined entirely by the margins of a continent.
Intermediate-sized plates include the Caribbean, Nazca,
Philippine, Arabian, Cocos, Scotia, and Juan de Fuca plates. In
addition, there are over a dozen smaller plates that have
been identified but are not shown in Figure 2.20.
One of the main tenets of the plate tectonic theory is that
plates move as coherent units relative to all other plates. As
plates move, the distance between two locations on the
same plateNew York and Denver, for exampleremains
relatively constant, whereas the distance between sites on
different plates, such as New York and London, gradually
changes. (Recently it has been shown that plates can suffer
some internal deformation, particularly oceanic lithosphere.)

Illustration of some of Earths lithospheric plates.

North
American plate

Eurasian
plate
North
American
plate

Eurasian
plate
Arabian
plate

African
plate

Caribbean
plate
Cocos
plate

Philippine
plate
Pacific
plate

South
American
plate
Nazca
plate

Scota plate
Antarctic plate

51

Australian-Indian
plate

Antarctic
plate

M02_TARB6927_09_SE_C02.QXD

54

1/11/07

CHAPTER 2

1:24 PM

Page 54

Plate Tectonics: A Scientific Revolution Unfolds

Lithospheric plates move relative to each other at a very

slow but continuous rate that averages about 5 centimeters
(2 inches) per year. This movement is ultimately driven by
the unequal distribution of heat within Earth. Hot material
found deep in the mantle moves slowly upward and serves
as one part of our planets internal convection system. Concurrently, cooler, denser slabs of oceanic lithosphere descend into the mantle, setting Earths rigid outer shell into
motion. Ultimately, the titanic, grinding movements of
Earths lithospheric plates generate earthquakes, create volcanoes, and deform large masses of rock into mountains.

Plate Boundaries
Tectonic plates move as coherent units relative to all other
plates. Although the interiors of plates may experience
some deformation, all major interactions among individual
plates (and therefore most deformation) occur along their
boundaries. In fact, plate boundaries were first established by
plotting the locations of earthquakes. Moreover, plates are
bounded by three distinct types of boundaries, which are
differentiated by the type of movement they exhibit. These
boundaries are depicted at the bottom of Figure 2.20 and are
briefly described here:
1. Divergent boundaries (constructive margins)where
two plates move apart, resulting in upwelling of material from the mantle to create new seafloor (Figure
2.20A).
2. Convergent boundaries (destructive margins)where
two plates move together, resulting in oceanic lithosphere descending beneath an overriding plate, eventually to be reabsorbed into the mantle, or possibly in
the collision of two continental blocks to create a
mountain system (Figure 2.20B).
3. Transform fault boundaries (conservative margins)
where two plates grind past each other without the
production or destruction of lithosphere (Figure
2.20C).
Each plate is bounded by a combination of these three
types of plate margins. For example, the Juan de Fuca plate
has a divergent zone on the west, a convergent boundary on
the east, and numerous transform faults, which offset segments of the oceanic ridge (Figure 2.20). Although the total
surface area of Earth does not change, individual plates may
diminish or grow in area depending on any imbalance between the growth rate at divergent boundaries and the rate
at which lithosphere is destroyed at convergent boundaries.
The Antarctic and African plates are almost entirely bounded by divergent boundaries and hence are growing larger
by adding new lithosphere to their margins. By contrast, the
Pacific plate is being consumed into the mantle along its
northern and western flanks faster than it is being replaced,
and therefore is diminishing in size.
It is also important to note that plate boundaries are not
fixed but move about. For example, the westward drift of
the South American plate is causing it to override the Nazca

plate. As a result, the boundary that separates these plates

is gradually being displaced as well. Moreover, since the
Antarctic plate is surrounded by constructive margins and
is growing larger, the divergent boundaries are migrating
away from the continent of Antarctica.
New plate boundaries can be created in response to
changes in the forces acting on these rigid slabs. For example, a relatively new divergent boundary is located in the
Red Sea. Less than 20 million years ago the Arabian Penninsula began to rift away from Africa. At other locations,
plates carrying continental crust are presently moving toward one another. Eventually these continents may collide
and be sutured together. In this case, the boundary that once
separated two plates disappears as the plates become one.
The result of such a continental collision is a majestic mountain range such as the Himalayas.
In the following sections we will briefly summarize the
nature of the three types of plate boundaries.

Divergent Boundaries
Plate Tectonics
 Divergent Boundaries

Most divergent 1di = apart, vergere = to move2 boundaries are located along the crests of oceanic ridges and can
be thought of as constructive plate margins since this is where
new oceanic lithosphere is generated (Figure 2.21). Divergent boundaries are also called spreading centers, because
seafloor spreading occurs at these boundaries. Here, as the
plates move away from the ridge axis, the fractures that
form are filled with molten rock that wells up from the hot
mantle below. Gradually, this magma cools to produce new
slivers of seafloor. In a continuous manner, adjacent plates
spread apart and new oceanic lithosphere forms between
them. As we shall see later, divergent boundaries are
not confined to the ocean floor but can also form on the
continents.

Oceanic Ridges and Seafloor Spreading

Along well-developed divergent plate boundaries, the
seafloor is elevated, forming the oceanic ridge. The interconnected oceanic ridge system is the longest topographic feature on Earths surface, exceeding 70,000 kilometers
(43,000 miles) in length. Representing 20 percent of Earths
surface, the oceanic ridge system winds through all major
ocean basins like the seam on a baseball. Although the
crest of the oceanic ridge is commonly 2 to 3 kilometers
higher than the adjacent ocean basins, the term ridge
may be misleading because this feature is not narrow but
has widths from 1000 to 4000 kilometers. Further, along the
axis of some ridge segments is a deep down-faulted structure called a rift valley.
The mechanism that operates along the oceanic ridge
system to create new seafloor is appropriately called
seafloor spreading. Typical rates of spreading average around

M02_TARB6927_09_SE_C02.QXD

1/11/07

1:24 PM

Page 55

Convergent Boundaries

55

Continental Rifting
Divergent plate boundaries can also
develop within a continent, in which
case the landmass may split into two or
more smaller segments, as Alfred Wegener had proposed for the breakup of
Lithosphere
Pangaea. The splitting of a continent is
Magma
chamber
thought to begin with the formation of
an elongated depression called a conAsthenosphere
tinental rift. A modern example of a
continental rift is the East African Rift.
Whether this rift will develop into a
North
full-fledged spreading center and evenEurope
America
e
g
tually split the continent of Africa is a
d
Ri
matter of much speculation.
Nevertheless, the East African Rift
represents the initial stage in the breakup of a continent (see Figure 13.20,
Africa
page 366). Here tensional forces have
stretched and thinned the continental
Lith
osp
crust. As a result, molten rock ascends
here
Asth
Upwelling
eno
from the asthensophere and initiates
sph
ere
volcanic activity at the surface (Figure
2.22A). Large volcanic mountains such
as
Kilimanjaro and Mount Kenya exFIGURE 2.21 Most divergent plate boundaries are situated along the crests of oceanic ridges.
emplify the extensive volcanic activity
that accompanies continental rifting.
Research suggests that if tensional forces
5 centimeters (2 inches) per year. This is roughly the same
are maintained, the rift valley will lengthen and deepen,
rate at which human fingernails grow. Comparatively slow
eventually extending out to the margin of the plate, splitting
spreading rates of 2 centimeters per year are found along
it in two (Figure 2.22C). At this point, the rift becomes a narthe Mid-Atlantic Ridge, whereas spreading rates exceeding
row sea with an outlet to the ocean, similar to the Red Sea.
15 centimeters (6 inches) have been measured along secThe Red Sea formed when the Arabian Peninsula rifted from
tions of the East Pacific Rise. Although these rates of lithoAfrica, an event that began about 20 million years ago. Conspheric production are slow on a human time scale, they
sequently, the Red Sea provides oceanographers with a view
are nevertheless rapid enough to have generated all of
of how the Atlantic Ocean may have looked in its infancy.
Earths ocean basins within the last 200 million years. In
fact, none of the ocean floor that has been dated exceeds 180
million years in age.
Convergent Boundaries
The primary reason for the elevated position of the
oceanic ridge is that newly created oceanic crust is hot, and
Plate Tectonics
occupies more volume, which makes it less dense than
 Convergent Boundaries
cooler rocks. As new lithosphere is formed along the oceanic ridge, it is slowly yet continually displaced away from
Although new lithosphere is constantly being produced at
the zone of upwelling along the ridge axis. Thus, it begins
the oceanic ridges, our planet is not growing largerits
to cool and contract, thereby increasing in density. This
total surface area remains constant. To balance the addition
thermal contraction accounts for the greater ocean depths
of newly created lithosphere, older, denser portions of
that exist away from the ridge crest. It takes about 80 miloceanic lithosphere descend into the mantle along converlion years before the cooling and contracting cease comgent 1con = together, vergere = to move2 boundaries. Bepletely. By this time, rock that was once part of the elevated
cause lithosphere is destroyed at convergent boundaries,
oceanic ridge system is located in the deep-ocean basin,
they are also called destructive plate margins (Figure 2.23A).
where it may be buried by substantial accumulations of
Convergent plate margins occur where two plates move
sediment. In addition, cooling causes the mantle rocks
toward each other and the leading edge of one is bent downbelow the oceanic crust to strengthen, thereby adding to the
ward, allowing it to slide beneath the other. The surface explates thickness. Stated another way, the thickness of
pression produced by the descending plate is a deep-ocean
oceanic lithosphere is age-dependent. The older (cooler) it
trench, such as the PeruChile trench (see Figure 13.9, page
is, the greater its thickness.
357). Trenches formed in this manner may be thousands of
Mi

dAt
lan
tic

Rift
valleys

M02_TARB6927_09_SE_C02.QXD

56

1/11/07

CHAPTER 2

1:24 PM

Page 56

Plate Tectonics: A Scientific Revolution Unfolds

Upwarping
Continental crust

A.
Continental rift

B.

Linear sea

C.
Mid-ocean ridge

Rift valley
Continental
crust
Oceanic crust

OceanicContinental
Convergence

D.

Whenever the leading edge of a plate capped

with continental crust converges with a slab
of oceanic lithosphere, the buoyant continental block remains floating, while the
denser oceanic slab sinks into the mantle
(Figure 2.23A). When a descending oceanic
slab reaches a depth of about 100 kilometers,
melting is triggered within the wedge of hot
asthenosphere that lies above it. But how does the subduction of a cool slab of oceanic lithosphere cause mantle rock
to melt? The answer lies in the fact that volatiles (mainly
water) act like salt does to melt ice. That is, wet rock, in a
high-pressure environment, melts at substantially lower
temperatures than dry rock of the same composition.
Sediments and oceanic crust contain a large amount of
water which is carried to great depths by a subducting plate.
As the plate plunges downward, water is squeezed from
the pore spaces as confining pressure increases. At even
greater depths, heat and pressure drive water from hydrated (water-rich) minerals such as the amphiboles. At a depth
of roughly 100 kilometers, the mantle is sufficiently hot that

FIGURE 2.22 Continental rifting and the formation of a new ocean basin. A. Continental rifting
is thought to occur where tensional forces stretch and thin the crust. As a result, molten rock
ascends from the asthenosphere and initiates volcanic activity at the surface. B. As the crust is
pulled apart, large slabs of rock sink, generating a rift valley. C. Further spreading generates a
narrow sea. D. Eventually, an expansive ocean basin and ridge system are created.

kilometers long, 8 to 12 kilometers deep, and between 50 and

100 kilometers wide.
Convergent boundaries are also called subduction
zones, because they are sites where lithosphere is descending (being subducted) into the mantle. Subduction occurs
because the density of the descending tectonic plate is
greater than the density of the underlying asthenosphere. In
general, oceanic lithosphere is more dense than the asthenosphere whereas continental lithosphere is less dense
and resists subduction. As a consequence, it is always
oceanic lithosphere that is subducted.
Slabs of oceanic lithosphere descend into the mantle at
angles that vary from a few degrees to nearly vertical (90 de-

grees), but average about 45 degrees. The

angle at which oceanic lithosphere descends
depends largely on its density. For example,
when a spreading center is located near a
subduction zone, the lithosphere is young
and, therefore, warm and buoyant. Hence,
the angle of descent is small. This is the situation along parts of the PeruChile trench.
Low dip angles usually result in considerable interaction between the descending slab
and the overriding plate. Consequently,
these regions experience great earthquakes.
As oceanic lithosphere ages (gets farther
from the spreading center), it gradually cools,
which causes it to thicken and increase in density. Once oceanic lithosphere is about 15 million years old, it becomes more dense than the
supporting asthenosphere and will sink when
given the opportunity. In parts of the western
Pacific, some oceanic lithosphere is more than
180 million years old. This is the thickest and
most dense in todays oceans. The subducting
slabs in this region typically plunge into the
mantle at angles approaching 90 degrees.
Although all convergent zones have the
same basic characteristics, they are highly
variable features. Each is controlled by the
type of crustal material involved and the tectonic setting. Convergent boundaries can
form between two oceanic plates, one oceanic and one continental plate, or two continental plates. All three situations are illustrated
in Figure 2.23.

M02_TARB6927_09_SE_C02.QXD

1/11/07

1:24 PM

Page 57

Convergent Boundaries
Continental
volcanic arc

Trench
Oceanic crust
Continental crust
Subduct
ing
oce
a
100 km

nic
l

it h

Continental
lithosphere

os
ph
e

re

Asthenosphere

Melting

200 km

A.
Volcanic island arc
Trench

Oceanic crust
Continental crust
Oceanic lithosphere
Melting

100 km

Asthenosphere
S

n
cti
du
ub

here
osp
lith
c
ni
ea
oc

200 km

B.
Collision mountains

Continental
lithosphere

Suture

Continental
lithosphere

57

ment, these mantle-derived magmas may

ascend through the crust and give rise to a
volcanic eruption. However, much of this
molten rock never reaches the surface;
rather, it solidifies at depth where it acts to
thicken the crust.
Partial melting of mantle rock generates
molten rock that has a basaltic composition,
similar to what erupts on the Island of
Hawaii. In a continental setting, however,
basaltic magma typically melts and assimilates some of the crustal rocks through which
it ascends. The result is the formation of a silica-rich (SiO2) magma. On occasions when
silica-rich magmas reach the surface, they
often erupt explosively, generating large
columns of volcanic arc and gases. A classic
example of such an eruption was the 1980
eruption of Mount St. Helens. You will learn
more about the formation of magma and its
influence on the explosiveness of volcanic
eruptions in Chapters 4 and 5.
The volcanoes of the towering Andes are
the product of magma generated by the subduction of the Nazca plate beneath the South
American continent (see Figure 2.20). Mountains such as the Andes, which are produced
in part by volcanic activity associated with
the subduction of oceanic lithosphere, are
called continental volcanic arcs. The Cascade Range of Washington, Oregon, and California is another volcanic arc which consists
of several well-known volcanic mountains,
including Mount Rainier, Mount Shasta, and
Mount St. Helens (see Figure 5.19, p. 142).
(This active volcanic arc also extends into
Canada, where it includes Mount Garibaldi,
Mount Silverthrone, and others.)

OceanicOceanic Convergence

An oceanicoceanic convergent boundary has

many features in common with oceaniccontinental plate margins. The differences are
sp
he
mainly attributable to the nature of the crust
r
e
Asthenosphere
capping the overriding plate. Where two
200 km
oceanic slabs converge, one descends beneath
C.
the other, initiating volcanic activity by the
same mechanism that operates at oceanic
FIGURE 2.23 Zones of plate convergence. A. Oceaniccontinental B. Oceanicoceanic
C. Continentalcontinental.
continental plate boundaries. Water squeezed
from the subducting slab of oceanic lithosphere triggers melting in the hot wedge of
mantle rock that lies above. In this setting, volcanoes grow up
the introduction of water leads to some melting. This
from the ocean floor, rather than upon a continental platform.
process, called partial melting, generates as little as 10 perWhen subduction is sustained, it will eventually build a chain
cent molten material, which is intermixed with unmelted
of volcanic structures that emerge as islands. The volcanic
mantle rock. Being less dense than the surrounding mantle,
islands are spaced about 80 kilometers apart and are built
this hot mobile material gradually rises toward the surface
upon submerged ridges of volcanic material a few hundred
as a teardrop-shaped structure. Depending on the environ100 km

Subductin
g oc
ean
ic
lith
o

M02_TARB6927_09_SE_C02.QXD

58

1/11/07

CHAPTER 2

1:24 PM

Page 58

Plate Tectonics: A Scientific Revolution Unfolds

kilometers wide. This newly formed land consisting of an arcshaped chain of small volcanic islands is called a volcanic island arc, or simply an island arc (Figure 2.23B).
The Aleutian, Mariana, and Tonga islands are examples
of volcanic island arcs. Island arcs such as these are generally located 100 to 300 kilometers (60 to 200 miles) from a
deep-ocean trench. Located adjacent to the island arcs just
mentioned are the Aleutian trench, the Mariana trench, and
the Tonga trench (see Figure 1.17).
Most volcanic island arcs are located in the western Pacific. Only two volcanic island arcs are located in the Atlanticthe Lesser Antilles arc adjacent to the Caribbean Sea
and the Sandwich Islands in the South Atlantic. The Lesser
Antilles are a product of the subduction of the Atlantic beneath the Caribbean plate. Located within this arc is the island of Martinique, where Mount Pele erupted in 1902,
destroying the town of St. Pierre and killing an estimated
28,000 people; and the island of Montserrat, where volcanic
activity has occurred very recently.*
Relatively young island arcs are fairly simple structures
that are underlain by deformed oceanic crust that is generally
less than 20 kilometers (12 miles) thick. Examples include the
arcs of Tonga, the Aleutians, and the Lesser Antilles. By contrast, older island arcs are more complex and are underlain by
crust that ranges in thickness from 20 to 35 kilometers. Examples include the Japanese and Indonesian arcs, which are
built upon material generated by earlier episodes of subduction or sometimes on a small piece of continental crust.

ContinentalContinental Convergence
As you saw earlier, when an oceanic plate is subducted beneath continental lithosphere, an Andean-type volcanic arc
develops along the margin of the continent. However, if the
subducting plate also contains continental lithosphere, continued subduction eventually brings the two continental
blocks together (Figure 2.23C). Whereas oceanic lithosphere
is relatively dense and sinks into the asthenosphere, continental lithosphere is buoyant, which prevents it from being
subducted to any great depth. The result is a collision between the two continental fragments (Figure 2.23C).
Such a collision occurred when the subcontinent of India
rammed into Asia, producing the Himalayasthe most
spectacular mountain range on Earth (Figure 2.24). During
this collision, the continental crust buckled, fractured, and
was generally shortened and thickened. In addition to the
Himalayas, several other major mountain systems, including the Alps, Appalachians, and Urals, formed during continental collisions.
Prior to a continental collision, the landmasses involved
were separated by an ocean basin (Figure 2.24A). As the continental blocks converge, the intervening seafloor is subducted beneath one of the plates. Subduction initiates partial
melting in the overlying mantle, which in turn results in the
growth of a volcanic arc. Depending on the location of the
subduction zone, the volcanic arc could develop on either of
*More on these volcanic events is found in Chapter 5.

Students Sometimes Ask . . .

Someday will the continents come back together and form a
single landmass?
Yes, it is very likely that the continents will eventually come
back together, but not anytime soon. Since all of the continents
are on the same planetary body, there is only so far a continent
can travel before it collides with another landmass. Recent research suggests that a supercontinent may form about once
every 500 million years or so. Since it has been about 200 million
years since Pangaea broke up, we have only about 300 million
years to go before the next supercontinent is assembled.

the converging landmasses, or if the subduction zone developed several hundred kilometers seaward from the coast, a
volcanic island arc would form. Eventually, as the intervening seafloor is consumed, these continental masses collide
(Figure 2.24B). This folds and deforms the accumulation of
sediments and sedimentary rocks along the continental margin as if they had been placed in a gigantic vise. The result is
the formation of a new mountain range composed of deformed and metamorphosed sedimentary rocks, fragments
of the volcanic arc, and often slivers of oceanic crust.

Transform Fault Boundaries

Plate Tectonics
 Transform Fault Boundaries

The third type of plate boundary is the transform 1trans =

across, forma = form2 fault, where plates slide horizontally
past one another without the production or destruction of
lithosphere (conservative plate margins). The nature of transform faults was discovered in 1965 by J. Tuzo Wilson of the
University of Toronto. Wilson suggested that these large
faults connect the global active belts (convergent boundaries, divergent boundaries, and other transform faults) into
a continuous network that divides Earths outer shell into
several rigid plates. Thus, Wilson became the first to suggest
that Earth was made of individual plates, while at the same
time identifying the faults along which relative motion between the plates is made possible.
Most transform faults join two segments of an oceanic
ridge (Figure 2.25). Here, they are part of prominent linear
breaks in the oceanic crust known as fracture zones, which include both the active transform faults as well as their inactive
extentions into the plate interior. These fracture zones are
present approximately every 100 kilometers along the trend of
a ridge axis. As shown in Figure 2.25, active transform faults
lie only between the two offset ridge segments. Here seafloor
produced at one ridge axis moves in the opposite direction as
seafloor produced at an opposing ridge segment. Thus, between the ridge segments these adjacent slabs of oceanic crust
are grinding past each other along the fault. Beyond the ridge

M02_TARB6927_09_SE_C02.QXD

1/11/07

1:24 PM

Page 59

Transform Fault Boundaries

59

Continental volcanic arc

India

Continental
shelf
deposits

Tibet

Developing
accretionary
wedge

Ocean basin
Continental
crust
Subducting oceanic

A.

litho
sph
ere

Melting

Asthenosphere
B.

Himalayas
India
today

Ganges
Plain
10 million
years ago

India
Tibetan
Plateau
38 million
years ago

Suture
55 million
years ago
71 million
years ago

Asthenosphere
C.
FIGURE 2.24 The ongoing collision of India and Asia, starting about 45 million years ago,
produced the majestic Himalayas. A. Converging plates generated a subduction zone, while partial
melting triggered by the subducting oceanic slab produced a continental volcanic arc. Sediments
scraped from the subducting plate were added to the accretionary wedge. B. Position of India in
relation to Eurasia at various times. (Modified after Peter Molnar) C. Eventually the two landmasses
collided, deforming and elevating the sediments that had been deposited along the continental
margins. In addition, slices of the Indian crust were thrust up onto the Indian plate.

crests are the inactive zones, where the fractures are preserved
as linear topographic scars. The trend of these fracture zones
roughly parallels the direction of plate motion at the time of
their formation. Thus, these structures can be used to map the
direction of plate motion in the geologic past.
In another role, transform faults provide the means by
which the oceanic crust created at ridge crests can be transported to a site of destruction: the deep-ocean trenches.
Figure 2.26 illustrates this situation. Notice that the Juan de
Fuca plate moves in a southeasterly direction, eventually
being subducted under the west coast of the United States.
The southern end of this plate is bounded by the Mendocino

fault. This transform fault boundary connects the Juan de

Fuca ridge to the Cascadia subduction zone (Figure 2.26).
Therefore, it facilitates the movement of the crustal material
created at the ridge crest to its destination beneath the North
American continent (Figure 2.26).
Although most transform faults are located within the
ocean basins, a few cut through continental crust. Two examples are the earthquake-prone San Andreas Fault of California and the Alpine Fault of New Zealand. Notice in Figure
2.26 that the San Andreas Fault connects a spreading center
located in the Gulf of California to the Cascadia subduction
zone and the Mendocino fault located along the northwest

M02_TARB6927_09_SE_C02.QXD

60

1/11/07

CHAPTER 2

1:24 PM

Page 60

Plate Tectonics: A Scientific Revolution Unfolds

Fracture zone

Inactive
zone

Transform fault
(active)

Inactive
zone

tal in solidifying the support for this

new idea follows. Note that much of this
evidence was not new; rather, it was
new interpretations of already existing
data that swayed the tide of opinion.

Evidence from Ocean

Drilling
Some of the most convincing evidence
confirming seafloor spreading has come
from drilling directly into the ocean
floor. From 1968 until 1983, the source
Oceanic crust
of these important data was the Deep
Sea Drilling Project, an international
Oceanic
lithosphere
program sponsored by several major
oceanographic institutions and the National Science Foundation. The primary
Asthenosphere
goal was to gather firsthand information about the age of the ocean basins
Africa
and processes that formed them. To accomplish this, a new drilling ship, the
Glomar Challenger, was built.
Operations began in August 1968 in
M
idthe South Atlantic. At several sites holes
Atl
were drilled through the entire thickan
tic
ness of sediments to the basaltic rock
Ri
dg
below. An important objective was to
e
gather samples of sediment from just
above the igneous crust as a means of
dating the seafloor at each site.* Because sedimentation begins immediateKEY
ly after the oceanic crust forms, remains
Spreading centers
South
of microorganisms found in the oldest
America
Fracture zones
sedimentsthose resting directly on
Transform faults
the crustcan be used to date the ocean
floor at that site.
FIGURE 2.25 Diagram illustrating a transform fault boundary offsetting segments of the MidWhen the oldest sediment from each
Atlantic Ridge.
drill site was plotted against its distance
from the ridge crest, the plot demonstrated that the age of the sediment increased
with increasing distance from the ridge. This finding supportcoast of the United States. Along the San Andreas Fault, the
ed the seafloor-spreading hypothesis, which predicted the
Pacific plate is moving toward the northwest, past the North
youngest oceanic crust would be found at the ridge crest and
American plate. If this movement continues, that part of Calthe oldest oceanic crust would be at the continental margins.
ifornia west of the fault zone, including the Baja Peninsula,
The data from the Deep Sea Drilling Project also reinwill eventually become an island off the West Coast of the
forced the idea that the ocean basins are geologically youthUnited States and Canada. It could eventually reach Alaska.
ful because no seafloor with an age in excess of 180 million
However, a more immediate concern is the earthquake activyears was found. By comparison, continental crust that exity triggered by movements along this fault system.
ceeds 4 billion years in age has been dated.
The thickness of ocean-floor sediments provided additional verification of seafloor spreading. Drill cores from the
Testing the Plate Tectonics Model
Glomar Challenger revealed that sediments are almost entirely absent on the ridge crest and the sediment thickness with
With the development of the theory of plate tectonics, reincreasing distance from the ridge. Because the ridge crest is
searchers from all of the Earth sciences began testing this new
model of how Earth works. Some of the evidence supporting
continental drift and seafloor spreading has already been presented. In addition, some of the evidence that was instrumen-

*Radiometric dates of the ocean crust itself are unreliable because of the alteration of
basalt by seawater.