Continental crust is very long-lived but continents themselves are ephemeral. Earth's inventory of continental crust changes slowly but how this material is partitioned into separate aggregations is always changing. We call these distinct aggregations of buoyant, felsic crust “continents” and now there are 6 of them: North America, South America, Antarctica, Eurasia, Africa, and Australia. There are many smaller minicontinents like Arabia, Madagascar, and Greenland. From the human perspective, these names have great significance as the sites of our nations, civilizations, and cultures, but the names of continents are not useful in the long run because continents gain and lose mass, such that over time their sizes and shapes change greatly. Heraclitus noted that you cannot step twice into the same river, for new waters are ever flowing around you. In this sense, a continent is like a river, forever changing.

Of course, continents do not change as fast as a river flows, but their changes are far less predictable. Rivers flow downstream, generally getting larger as they move to the sea, but continents are just as likely to shrink as to grow with time. This is because continental shape-shifting manifests plate tectonic processes, which separate as well as amalgamate crust. Continents mostly change by rifting and colliding. These processes reflect the approach of a continent to a subduction zone (collision) or the formation of a new ocean (rifting) in the midst of a continent. These changes affect the continents in many ways. Continental aggrandizement and diminishment change surface elevations as well as continental outline, forming mountains and basins. Drainage networks must constantly adjust, not only to changing relief due to interactions on the margins (and formation of new margins), but also to the drift of the continents into different climate regions.

Continents mostly grow by colliding and shrink by rifting. Eurasia today is the largest continent and is growing rapidly. With the recent addition of India, imminent addition of Arabia and then Australia, Eurasia is well on its way to becoming the next supercontinent. At the other end of the spectrum, Africa is the only shrinking continent, riven with rifts and surrounded by spreading ridges and continental fragments, geomemories of the much larger continental assembly, Gondwana. Other processes also change continental size and shape, such as strike-slip (transform) faulting, although such geo-whittling doesn't affect continental configurations as profoundly as colliding and rifting do. Interactions with mantle melting anomalies (plumes) or subduction of a spreading ridge beneath a continent also act as modest rearrangers of continental crust.

“The Supercontinent Cycle” (Nance et al., 1988) captures the most important aspects of continental reconfiguration, albeit greatly simplified. The splitting apart and coming together of continental fragments is stochastic and cannot be characterized as truly cyclic. Continents split up and come together because of stresses imposed by plate interactions in the vicinity, not because of global pulsations. With this understood, the supercontinent cycle is useful for emphasizing the transitory nature of continents and how reconfigurations cause changes in sea level, climate, and macroevolution.

This themed issue of Geosphere, “Making the Southern Margin of Laurentia,” looks at how the margins of part of our continent were shaped. A couple of questions come immediately to mind: What is Laurentia, and how is it different than the North American continent? Where is its southern margin? Laurentia is named after the Laurentian craton, another name for the North American craton. The Laurentian craton is named after the St. Lawrence River, which flows over it. This craton is exclusively Precambrian crust, distinct in this regard from the North American continent, which includes significant tracts of accreted Phanerozoic crust (Fig. 1). In contrast to the more ancient crust of northern Laurentia, which mostly formed >1.8 Ga, that of southern Laurentia formed 1.8–1.3 Ga (Karlstrom et al., 2001). Southern Laurentian crust was modified by a wide range of processes, including two episodes of continental collision (the Grenville orogeny ∼1.1 Ga accompanying formation of supercontinent Rodinia and the Ouachita orogeny ∼300 Ma leading to the formation of Pangea) and two episodes of rifting to form new ocean basins (Cambrian rifting ∼530 Ma to form the Rheic ocean and Jurassic rifting ∼165 Ma to form the Gulf of Mexico). Southern Laurentia faces the Pacific on its west, and subduction of Pacific seafloor has been an intermittently important aspect of its tectonic evolution since Pangea formed ∼300 Ma. To the east, southern Laurentia felt the extensional stresses of Tethys, opening within Pangea. All these episodes played roles in shaping the southern margin of Laurentia, and left scraps of crust that are part of North America, but not part of Laurentia. Many of these exotic terranes in the SW are accreted fragments of Pacific arcs or plateaux, whereas exotic terranes in the SE are orphans left by Gondwana when Pangea broke up.

Where does Laurentia terminate in the south? Some show Laurentia ending near the U.S.-Mexico border, but this is an oversimplification (Fig. 1). A recent estimate of lithospheric thickness based on seismic-wave receiver function data indicate that the “stable” part of the North American continent is bounded to the west by the Laramide deformation/Rocky Mountain front and the Ouachita and Appalachian fronts to the south and east (Yuan and Romanowicz, 2010). However, Grenville-aged crust of Laurentia extends southward in Mexico, far beyond the limit of the seismically defined stable continent. The suture between Laurentia and Gondwana is the Ouachita orogen, so Coahuila is a Gondwanan orphan, as is stretched basement below coastal Texas, Louisiana, and Mississippi (Sabine, Wiggins, and Monroe uplifts). Florida is also an accreted peri-Gondwanan terrane. Along strike to the southwest in Mexico, the crust is all accreted subduction complexes or arcs, which originated in the Pacific realm. West of the Ouachita front, Laurentian basement exists in northern Sonora and Chihuahua but it does not extend past the California-Coahuila transform (Mojave-Sonora megashear overprinted by the Mesozoic magmatic arc). In this region, the Laurentia-Gondwana join (at the Ouachita orogen) reaches south through Chihuahua but butts against Phanerozoic accreted terranes somewhere near the southern border of Chihuahua, at the southern edge of the Mesozoic arc. Unfortunately, critical relationships are hidden beneath Cretaceous and Tertiary cover, so no one knows exactly where the triple join is between Laurentia (in the north), Gondwana (eastern Mexico), and post-Ouachita accreted terranes (to the southwest).

This special issue is launched as Earthscope's Magnetotelluric Transportable Array of seismometers moves slowly across the region, which is providing an unparalleled look into the lithospheric and mantle structure of the region. This themed issue draws attention to this region, in the expectation that the next few years will provide useful new information about the lithospheric and upper mantle structure of the U.S. part of Laurentia.