Ridge-hotspot interactions What Mid-Ocean Ridges Tell Us About Deep Earth Processes

major features of hotspot-ridge interactions.

S p e c i a l i S S u e F e at u r e B y J é r ô m e D y m e N t, J i a N l i N , a N D e D w a r D t. B a k e r Earth is a thermal engine that dissipates its internal heat primarily through convection. The buoyant rise of hot material transports heat to the surface from the deep interior while colder material sinks at subduction zones. Mid-ocean ridges and hotspots are major expressions of heat dissipation at Earth's surface, as evidenced by their abundant volcanic activity. Ridges and hotspots, however, could differ significantly in their origins. Ridges are linear features that wind more than 60,000 km around the globe, constituting the major diverging boundaries of Earth's tectonic plates. Hotspots, on the other hand, are localized regions of abnormally robust magmatism and distinctive geochemical anomalies ( Figure 1).
The causes of hotspots and their depths of origin are the focus of an intense debate in the scientific community. The "plume" model hypothesizes rising of buoyant mantle plumes as the primary cause of prominent hotspots such as Iceland and Hawaii (Morgan, 1971). In contrast, the "anti-plume" school argues that many of the observed "hotspot" volcanic and geochemical anomalies are simply due to melts leaking through tensional cracks in Earth's lithospheric plates-in other words, hotspots reflect only where the lithospheric plate is cracked, allowing melts to pass through, and not where the underlying mantle is hotter (see www.mantleplumes.org). A hybrid notion is that only a relatively small number of hotspots, especially those of enormous magmatic volumes, have their origin in buoyant thermal plumes rising from the deep mantle (e.g., Courtillot et al., 2003). Regardless of its specific depth of origin, however, when a hotspot is located close enough to a mid-ocean ridge, the two volcanic systems will interact, resulting in unique volcanic, geochemical, and hydrothermal features. In this paper, we discuss major features of hotspot-ridge interactions. showing that many of them are integrally connected to the global mid-ocean ridge systems (red lines) (lin, 1998). (Bottom) map of residual bathymetry of the ocean basins and 87 Sr/ 86 Sr geochemical anomalies from samples collected along the mid-ocean ridges and ocean islands (ito et al., 2003). a positive residual bathymetry marks anomalously shallow seafloor relative to the theoretical prediction of Stein and Stein (1992). circles mark rock sample locations and are colored according to 87 Sr/ 86 Sr value. Hotspots are shown by stars, and hotspots influencing mid-ocean ridges are labeled: af = afar, as = ascension, az = azores, Ba = Balleny, Bo = Bowie, Bv = Bouvet, co = cobb, cr = crozet, eS = easter/Sala y gomez, ga = galápagos, go = gough, gu = guadalupe, ice = iceland, Jm -Jan mayen, ke = kerguelen, lo = louisville, ma = marion, re = reunion,

Bathymetry
The influence of hotspots on mid-ocean ridges can be seen most clearly in unusual bathymetry, including shallowerthan-normal ridge-axis seafloor depth, underwater plateaus, or volcanic islands rising from the seafloor (Figure 2). The elevated topography near a hotspot is the direct result of thickening of the oceanic crust both by erupting magmas on top of it and intruding magmas near its base.
The active upwelling of hotter mantle plumes can also lead to the development of long-wavelength seafloor topographic swells, as observed in some hotspot-ridge systems (e.g., Sleep, 1990;Canales et al., 2002). It has also been observed that ridge segments most influenced by hot-   plained by the mixing of "normal" and hotspot mantle materials (e.g., Schilling, 1991 Ito et al., 2003). Magnetic anomalies and dating of rock-eruption ages, on the other hand, are essential for determining the ages of the oceanic crust and for reconstructing a kinematical history of interacting ridge-hotspot systems (e.g., Dyment, 1998;Müller et al., 1998Müller et al., , 2001.   revealing that regions of unusually thick and elevated oceanic crust (red and yellow) extend hundreds of kilometers from the iceland and azores hotspots along the midatlantic ridge. The two hotspots appear to have influenced nearly the entire northern segment of the mid-atlantic ridge (lin, 1998   . Note that the wolf-Darwin lineament and other volcanic features (black dashed lines) appear to connect the galápagos archipelago to the galápagos Spreading center. (Bottom) correlations between bathymetric, geophysical, and geochemical anomalies along the galápagos Spreading center . (a) measured ridgeaxis seafloor depth (solid lines) and filtered long-wavelength regional depth (dashed lines    Hart et al, 1973;Schilling, 1991;Chauvel and Hemond, 2000).

Seismic tomography and reflection/refraction experiments
type 2: ridges in close proximity to Hotspots For a ridge located in close proximity to a hotspot, typically a few hundred kilometers or less, a fraction of the hot mantle material (Schilling, 1991) or melt (Braun and Sohn, 2003) (Hooft et al., 2003), similar to the results obtained for the Iceland (Shen et al., 1998) and Society (Niu et al., 2002) hotspots.
Along the Galápagos Spreading Center, the seafloor is shallowest immediately north of the Galápagos Archipelago and gradually deepens to the east and west (Figure 4a)  . Basalt samples collected along the ridge axis and on the Galápagos platform exhibit systematic geochemical anomalies in K/Ti, Nb/Zr (Figure 4), 87 Sr/ 86 Sr, 3 He/ 4 He, and other elements, indicating strong plume-ridge interactions (e.g., Schilling et al., 1982;Graham et al., 1993;Detrick et al., 2002;Cushman et al., 2004). 19°S all suggest a hotspot influence (e.g., Mahoney et al., 1989). The Rodrigues Ridge is an east-west trending volcanic feature that was formed between 7 and 10 million years ago and that appears to partially connect the Reunion-Mauritius hotspot track to the present-day Central Indian Ridge axis, thus reflecting some type of distal ridge-hotspot interaction (Morgan, 1978 Figure 5. map of bathymetry showing the relationship between the reunion-mauritius hotspot track and the central indian ridge, which is located more than 1000 km away (Dyment et al., 1999). Note that a series of east-west trending volcanic ridges are located between the mauritius plateau and the central indian ridge, including the relatively large rodrigues ridge and the smaller Three magi and gasitao ridges. Full colors show multibeam bathymetric data, and pale colors the satellite-derived bathymetry   . Schematic diagram showing evolution of an interacting ridge-hotspot system. The ridge axis is assumed to migrate to the left relative to the hotspot reference frame. active volcanic structure is marked by red (ridge) and pale red (hotspot). Stage 1: The ridge approaches the hotspot. Stage 2: The ridge starts to interact with the hotspot (e.g., Foundation seamount chain and pacificantarctic ridge, Figure 2e). Stage 3: The ridge passes over the hotspot and builds an oceanic plateau (e.g., iceland hotspot and reykjanes ridge, Figures 2a and 3). Stage 4: The ridge remains in the hotspot vicinity for a while through asymmetric spreading, segment propagation, and ridge jumps (e.g., galápagos hotspot and galápagos Spreading center, Figures 2b and 4; azores hotspot and midatlantic ridge, Figure 2d; St paul-amsterdam hotspot and Southeast indian ridge, Figure 2f; marion hotspot and Southwest indian ridge, Figure 2g). Stage 5: The ridge progressively escapes the hotspot influence (e.g., reunion hotspot and central indian ridge, Figures 2c and 5). Variability in this model depends on the vigor of the hotspot, the ridge spreading rate, the ridge-hotspot relative motion, and the presence of fracture zones that tend to restrict the along-ridge extent of the hotspot influence. the interaction easier to maintain (e.g., Maia et al., 2000).

HyDrotHermal eFFectS oF riD ge-HotSpot iNter actioNS
Systematic searches for hydrothermal activity along > 7000 km of mid-ocean ridge demonstrate that the spatial density of hydrothermal activity is a robust linear function of spreading rate ( Figure 7) (Baker and German, 2004).
This trend argues that the availability of mantle heat is the first-order control on the distribution of seafloor vent fields.
The universality of this hypothesis remains to be proven, however, especially where magma supply is not a linear function of spreading rate. Ridge sections influenced by a nearby hotspot offer a unique experimental setting for such a test. For example, some crustal thermal models predict that the thicker, hotter crust associated with hotspots substantially impedes the development of convective hydrothermal cooling (e.g., Chen, 2003;Chen and Lin, 2004). Re-duced hydrothermal cooling appears to be the simplest explanation for the unusually shallow magma bodies detected along mid-ocean sections overlying the Reykjanes (Sinha et al., 1997) and Galápagos  hotspots.
About 10  . Scatter plot of incidence of hydrothermal plumes versus full spreading rate for 14 ridge sections totalling 7,000 km. least-squares regression and ±95% confidence limits are shown for ridge sections not near hotspots (blue dots). red squares show data from four hotspot-affected ridges. Data point for the gakkel ridge (in parentheses) is biased by unique hydrography and bathymetry  and is not included in the leastsquares regression. rr = reykjanes ridge (iceland); as = mid-atlantic ridge (ascension); gSc = galápagos Spreading center (galápagos); Spa = Southeast indian ridge (St. paul-amsterdam).
the hydrothermal plume incidence, p h , the fraction of ridge length overlain by hydrothermal plumes (Baker and Hammond, 1992   ( Figure 4). As at the Reykjanes Ridge, the crustal thermal model (Chen, 2003) requires weaker cooling, by perhaps as much as a factor of two, to support a magma body this shallow. Preliminary estimates of the hydrothermal plume data give a p h value of 0.1 for the Galápagos Spreading Center from 95°-89°36´W, a reduction by at least half compared to other surveyed intermediate-rate spreading ridges.
The Ascension "hotspot" may not be a mantle plume but simply the expression of a small mantle heterogeneity that supplies excess magma, but without a temperature anomaly (Bruguier et al., 2003).  ackNowleD ge meNtS