Product has been added to the basket

Don Anderson joins the Caribbean Debate

Prof. Don L AndersonOrigin of the Caribbean Plate (Written December 2009)

Geoscientist Online 2 February 2010

Prof. Don L Anderson writes: Where did the Caribbean Plate come from? Did it originate in the Pacific Ocean from a deep thermal “starting plume” rising under the Galapagos, as most geodynamicists and some geologists believe, or does it have a tectonic explanation?

A fluid dynamic, non-geological idea has long dominated discussions of the Caribbean and other oceanic plateaux. Based on fluid dynamic injection experiments large igneous provinces (LIPs) are thought, by some, to be the result of large spherical plume heads rising from the core-mantle boundary (CMB), by their own buoyancy, through a cold (subsolidus) homogeneous mantle. Away from these hot plumes the otherwise homogeneous mantle only exceeds the melting point at ridges and arcs. The Caribbean crust has been attributed to a decapitated Galapagos mantle plume. The Parana flood basalts have likewise been attributed to a decapitated Atlantic plume. These may, however, be natural consequences of plate and arc tectonics, lithospheric and crustal processes, and magmatism from a heterogeneous refertilized upper mantle1. There is no need to forsake well-established tectonic mechanisms to explain fertile mantle and the locations of magma egress from the mantle, particularly in regions of fracture zones, multiple arcs and arc polarity flips. Caribbean geology does not reflect a problem with tectonic theory and does not require a deep hot upwelling that is independent of plate tectonics. Geological debates should not require a tooth fairy.

Continental crust has been built up over time but it is also destroyed by erosion and delamination; abandoned mantle wedges, and delaminated lower continental and island arc crust are potent sources for large rapidly produced oceanic plateaux2.

Thermal anomalies take a long time to form and to dissipate; the short duration of magmatism is an argument for stress control, or dyke physics, and a finite shallow fertile source. Delamination and recycling are rapid and shallow processes, and lithospheric stress valves can be turned on and off rapidly. The isotope composition of oceanic plateaux does not imply either high absolute temperature or deep sources. The rates at which arc crust delaminates and young unsubductable oceanic crust enters subduction zones exceeds the rate of so-called midplate magmatism2. When arcs collide or when subduction polarity flips, the fertile mantle wedge becomes a shallow fertile source.

We do not know, and cannot know, from geochemical and isotopic investigations, that the Caribbean plateau could only have formed by the impingement of a hot mantle plume on the lithosphere. Isotopic data cannot constrain the temperatures, depths or sizes of mantle reservoirs. It is the underlying assumptions (ambient mantle is homogeneous, essentially isothermal and subsolidus) that dictate depths and temperatures; the geochemical data is permissive of passive, geological, tectonic and shallow explanations for midplate magmatism.

The plume idea was introduced into Earth sciences because of the belief that plate tectonics and normal mantle forms of melting and convection could not explain volcanic chains, hotspots and large oceanic plateaux. The evidence for plumes included fixity, non-MORB chemistry, thinned (and thickened) crust, and large erupted volumes over short periods of time. The underlying assumptions included a homogeneous isothermal subsolidus mantle. Neither mantle convection nor plate tectonics can operate under such conditions and plate tectonics itself creates variations in melting point and temperature, even in the absence of plumes. Additionally, there are tectonic explanations for volcanic chains, and petrological explanations for melting anomalies and low seismic velocities.

Plate tectonics is a top-down theory that goes far beyond kinematics. Plate processes cool, enrich, deplete and fertilize the mantle, displace or suck-up warmer mantle, and introduce solidus-lowering volatiles. The mantle wedges associated with backarc basins are large fertile and shallow sources that are launched into the mantle during collisions and breakup. The mantle is heated from within as well as at the core-mantle boundary. In spite of the failure of all the predictions of the plume hypothesis, the implausibilty of the assumptions in a plate tectonic world, and the existence of plate tectonic explanations, there are those that argue that one should not throw out the plume baby with the discredited bathwater. But what is left of the baby except the assertion that oceanic plateaux and thick crust regions require deep mantle plumes rather than a heterogeneous mantle? Rapid emplacement rate is not a prediction of plume theory or fluid dynamic experiments but an after-the-fact postdiction based on the geochronology of “known” plume eruptions. Fluid dynamic theory and experiment predict long drawn out uplift and magmatism associated with starting plumes. Tectonic and dyke theory predict rapid release of ponded or underplated magmas. Plume theory does not predict the increase of magmatism with time along the Hawaiian chain and its relation to the Molokai Fracture Zone.

Plate tectonics is a much more powerful concept than generally acknowledged1,2. It involves more than the creation and destruction of oceanic crust. Oceanic crust is not the only low-melting component introduced into the mantle by plate tectonics and it need not be returned to the surface only in hot mantle plumes from the lower mantle. Oceanic plates do, however, displace warm material out of the transition region that can be mistaken for thermal plumes. Delaminated continental crust melts at ambient mantle temperature; high absolute temperatures are not required by the geology, geochemistry, tectonics or geophysics of midplate magmatism2,3. The assumption that midocean ridge basalts represent the composition and temperature of ambient Depleted Upper Mantle (the DUM hypothesis) underlies the assertion that Caribbean crust requires a high temperature lower mantle plume for its formation.

Good geological practice does not require the imposition of upwellings from the core-mantle boundary whenever the going gets tough. This is just another version of the deus ex machina, procrustean bed and tooth fairy approaches. Multiple-working-hypotheses is the cornerstone of good geologic practice; an a priori assignment of cause is not.

We do not know that the Caribbean plate could only have formed by the rapid impingement of a hot mantle plume beneath the lithosphere. Occam's razor does not require that the Caribbean plate be the result of subduction to the CMB, extraction of heat from the core, rebirth as a Pacific mantle plume over the Galapagos hotspot, decapitation and transferal to the American plate. This requires turning the coldest lower mantle components, slabs, into the hottest parts, plumes. If eclogites and pyroxenites are used to lower the melting point, to avoid this paradox, then thermal plumes and core heat are not required. Ambient mantle will do.


  1. Anderson, D. L., 2006, Speculations on the nature and cause of mantle heterogeneity: Tectonophysics v. 416, p. 7-22
  2. Anderson, D.L., 2007, The Eclogite engine: Chemical geodynamics as a Galileo thermometer, in Foulger, G.R., and Jurdy, D.M., eds., Plates, plumes, and planetary processes: Geological Society of America Special Paper 430, p. 47–64, doi: 10.1130/2007.2430(03).
  3. Anderson, Don L., 2007, New Theory of the Earth 2nd Edition; (ISBN-13: 9780521849593) DOI: 10.2277/0521849594. 408 pp