The Formation of Volcanic Islands
Earth's surface is not made up of a single sheet of rock that forms a crust but rather a number of "tectonic plates" that fit closely, like the pieces of a giant jigsaw puzzle. Some plates carry islands or continents; others form the seafloor. All are slowly moving because the plates float on a denser semiliquid mantle, the layer between the crust and Earth's core. The plates have edges that are spreading ridges (where two plates are moving apart and new seafloor is being created), subduction zones (where two plates collide and one plunges beneath the other), or transform faults (where two plates neither converge nor diverge but merely move past one another). It is at the boundaries between plates that most of Earth's volcanism and earthquake activity occur.
Generally speaking, the interiors of plates are geologically uneventful. However, there are exceptions. A glance at a map of the Pacific Ocean reveals that there are many islands far out at sea that are actually volcanoes-many no longer active, some overgrown with coral-that originated from activity at points in the interior of the Pacific Plate that forms the Pacific seafloor.
How can volcanic activity occur so far from a plate boundary? The Hawaiian Islands provide a very instructive answer. Like many other island groups, they form a chain. The Hawaiian Island Chain extends northwest from the island of Hawaii. In the 1840s American geologist James Daly observed that the different Hawaiian Islands seem to share a similar geologic evolution but are progressively more eroded, and therefore probably older, toward the northwest. Then in 1963, in the early days of the development of the theory of plate tectonics, Canadian geophysicist Tuzo Wilson realized that this age progression could result if the islands were formed on a surface plate moving over a fixed volcanic source in the interior. Wilson suggested that the long chain of volcanoes stretching northwest from Hawaii is simply the surface expression of a long-lived volcanic source located beneath the tectonic plate in the mantle. Today's most northwestern island would have been the first to form. Then, as the plate moved slowly northwest, new volcanic islands would have formed as the plate moved over the volcanic source. The most recent island, Hawaii, would be at the end of the chain and is now over the volcanic source.
Although this idea was not immediately accepted, the dating of lavas in the Hawaiian (and other) chains showed that their ages increase away from the presently active volcano, just as Daly had suggested. Wilson's analysis of these data is now a central part of plate tectonics. Most volcanoes that occur in the interiors of plates are believed to be produced by mantle plumes, columns of molten rock that rise from deep within the mantle. A volcano remains an active "hot spot" as long as it is over the plume. The plumes apparently originate at great depths, perhaps as deep as the boundary between the core and the mantle, and many have been active for a very long time. The oldest volcanoes in the Hawaiian hot-spot trail have ages close to 80 million years. Other islands, including Tahiti and Easter Island in the Pacific, Reunion and Mauritius in the Indian Ocean, and indeed most of the large islands in the world's oceans, owe their existence to mantle plumes.
The oceanic volcanic islands and their hot-spot trails are thus especially useful for geologists because they record the past locations of the plate over a fixed source. They therefore permit the reconstruction of the process of seafloor spreading, and consequently of the geography of continents and of ocean basins in the past. For example, given the current position of the Pacific Plate, Hawaii is above the Pacific Ocean hot spot. So the position of the Pacific Plate 50 million years ago can be determined by moving it such that a 50-million-year-old volcano in the hot-spot trail sits at the location of Hawaii today. However, because the ocean basins really are short-lived features on geologic time scales, reconstructing the world's geography by backtracking along the hot-spot trail works only for the last 5 percent or so of geologic time.