Excursion survey by: Paul Bogaard
The southern half of La Palma is formed by the so-called Dorsal series. This is the youngest, still active part of the island. The Dorsal Series consist mostly of primitive (basanitic) cinder cones and lava flows. These lavas generally have a more primitive character then the rest of the island. They become gradually younger to the south, and are still enlarging La Palma southward. Today we visited two of the youngest volcanoes, the San Antonio (1677) and the Teneguia (1971).
San Antonio:
The San Antonio is a typical cinder cone (a large heap of rubble with a hole in the middle). Two eruptive stages can be recognized: In the east a well sorted, coarse deposit with chaotic, hardly recognizable layering crops out. This is a typical strombolianic scoria breccia. On the western flank this deposit is overlain by a much finer, badly sorted but well layered deposit that can be identified as a phreatomagmatic surge deposit. Both deposits have a basanitic composition.
Teneguia:
Teneguia volcano, that was formed in 1971, is the youngest volcano in La Palma, and on all the Canary Islands. It is basanitic in composition. The volcano is a large cinder cone, that erupted mostly extremely vesicular lava. The lava has very bright, oily colours on the surface of vesicles. These colours are caused by finely divided ore minerals. A lava flow extends southward to the La Palma coast. A small fumarole on top of the volcano gives evidence of its recent activity.
Xenoliths:
The phreato-magmatic deposit of San Antonio volcano carries large amounts of xenoliths. It is a well known fact that xenoliths are often concentrated in phreatomagmatic eruptions. Two different explanations can be given for this phenomenon:
1.) Xenoliths generally have a higher density than their host magma, and are in chemical disequilibrium with this magma. Xenoliths transported in magma will therefore usually either stay behind in the crust or mantle, or dissolve in the magma. Magma that does transport xenoliths to the surface must have travelled through the crust very fast (much faster then the falling rate and then the dissolution rate of the xenoliths). When a magma that travels so fast passes through a groundwater layer close to the surface, it will be able to mix very effectively with this water, and thereby cause a phreatomagmatic explosion.
2.) magma containing dense mantle xenoliths is, of course, more dense then magma that doesn't carry xenoliths. Therefore it will not usually reach the surface. Only through later mixing with groundwater, and the consequential phreatomagmatic explosion, the xenolith-rich magma will see the daylight.
It seems that 1. is the better explanation, It is hard to see why xenolith-rich magmas would be able to reach levels close to the surface (where there is groundwater), but not the surface itself.
The xenoliths have rounded edges and rims. They are clearly pieces of solid rock that were broken off by rising magma, and were then rounded during transport. Rare large xenoliths are much more rounded. Rocks of this size can only just be transported by the rising magma, and spent more time in the magma to reach the surface.
The eruption carried mantle as well as crustal xenoliths. The crustal rocks (95%) display a large variety in mineralogy:
olivine + pyroxene
pyroxene + olivine
mostly pyroxene
pyroxene + plagioclase
plagioclase + pyroxene
Overall, (clino)pyroxene is the dominant mineral in these rocks. Alignment of elongated, euhedral pyroxenes indicates that all these rocks are cumulates, most probably from magma chambers located close to the crust-mantle boundary.
The mantle xenoliths (5%) are very olivine rich, with smaller amounts of pyroxene (Cr-diopside and orthopyroxene) and spinel. Several macroscopic features indicate that these are indeed mantle rocks, rather then olivine-rich cumulates:
- The olivines are green, indicating that they are rich in MgO. Typical mantle olivines have forsterite contents of 88-90. Basaltic melts coexisting with such olivines have much lower Mg/Fe ratios. Olivines crystallising from these melts are again Mg-richer than the melt, but Fo contents will be 80-82 at most. More Fa-rich olivines are more yellow in colour, as seen in phenocrysts in the San Antonio lavas.
- The olivines are equigranular. They are idiometric, but not idiomorph. That is, minerals are little 'balls' with flat surfaces that are not crystal planes. Spinels sit in niches between 3 or 4 olivines. These features suggest that the rocks are well equilibrated.
Geochemical studies of mantle xenoliths can give a lot of information about the (geodynamical) history of the upper mantle below the volcanic region that transported the xenoliths. An example of such a study is given below.
Metasomatic enrichment by melts: Upper Mantle Xenoliths from La Palma
Oceanic islands, such as the Canary Islands, generally form above regions with anomalously hot asthenosphere. Oceanic Island Basalts (OIBs) are characterized by enrichment of highly incompatible over moderately incompatible elements. The hot asthenosphere is thought to be the source of this enriched material. Lithospheric xenoliths in OIBs, however, have enriched trace element signatures as well. Since Sr and Nd isotopic compositions usually do not reflect this enrichment, it is presumably directly related to the magmatic activity.
Geophysical and geochemical investigations suggest that at least the western Canary Islands are underlain by old oceanic peridotite. Spinel harzburgites en lherzolites from La Palma show strong similarities with 'normal' oceanic peridotites with respect to mineralogy and major element chemistry. But they also show abundant evidence of pervasive metasomatism; both modal (widespread occurrence of phlogopite) and cryptic (enrichment of highly incompatible elements in clinopyroxene). Several processes were suggested in the literature (e.g. Menzies & Hawkesworth, 1987) to explain such cryptic metasomatism:
- Infiltration of fluids (either H2O-rich or CO2-rich or intermediate)
- Infiltration of (asthenospheric) basaltic melts
- Infiltration of carbonatitic melts
Veined xenoliths from La Palma give important clues to the processes acting in the sub-Canaries mantle.
Although refractory spinel harzburgite xenoliths are the most common by far on the Canary Islands, metasomatism has led to a large variety of other rock types: spinel wehrlites, amphibolites, pyroxenites and spinel dunites. Wulff-Pedersen et al. (1996) studied two veined xenoliths, a spinel harzburgite and a spinel dunite, in detail. The xenoliths are cut by phlogopite-amphibole rich, glass bearing veins. Narrower veins and veinlets branch of from these and penetrate the peridotite fragments (Fig. 1). The mineralogy and chemistry of these veins changes gradually and systematically with vein width:
Fig. 1 Drawing of a phlog-amph vein cutting spinel harzburgite. The central parts of the vein consist of coarse grained phlogopite, amphibole, augite and spinel. The grain size decreases towards the contact with peridotite fragments, where there is a reaction zone consisting of a fine-grained intergrowth of olivine + phlogopite + glass ± clinopyroxene ± spinel along the contacts with olivine in peridotite fragments. (from Wulff-Pedersen et al. (1996)).
The mineralogy changes from amphibole dominated in broad veins to phlogopite dominated in small veinlets. In broad veins, amphibole content and grain size decrease and phlogopite content increases towards the contact with peridotite fragments. Glass inclusions are brown in broad amphibole veins, pale brown in intermediate amphibole-phlogopite veins and colourless in phlogopite veinlets. In broad and intermediate veins, reaction zones consisting of phlog-ol-cpx needles are formed at the contact with peridotite. In phlogopite veinlets the contacts between glass and peridotite are sharp. The c7temisal composition of vein minerals shows a trend of decreasing Ti and Fe and increasing Mg and Cr with decreasing vein width. The Ti-Fe and Mg-Cr concentration of minerals in veinlets, and in reaction zones along broad veins, is similar to that in unveined peridotite xenoliths. The glass composition changes from basaltic in broad veins to Si-K-Na rich in phlogopite veinlets.
Textural relations and phase assemblages suggest that the gradual change in melt composition resulted from reaction between peridotitic host rock and infiltrating basaltic melt (for details of reactions see Wulff-Pedersen et al. (1996)). Geochemical modelling of the observed change in glass composition roughly agrees with the observed reactions (Fig. 2).
Fig. 2 Compositional variations among glasses in a veined spinel harzburgite and a veined spinel dunite from La Palma. The compositional ranges among mineral phases in these two xenoliths are shown by dark grey fields. The chemical ranges of glass inclusions in olivine and orthopyroxene in unveined spinel harzburgites / Iherzolites and spinel dunites from La Palma are shown as light grey fields marked harz/lherz and dunite, respectively. Samples marked by numbers or letters indicate glasses used as parent and daughter compositions in petrographic mixing calculations. Dashed arrows indicate reactions modelled by petrographic mixing calculations. Dotted arrows indicate fractional crystallization in the middle of broad veins. (from Wulff-Pedersen et al. (1996)).
The chemical change takes place over very short distances upon deeper penetration of the peridotitic host rock. The residual melt is enriched in Si K and Na, and depleted in Ti and P, suggesting enrichment of highly incompatible elements relative to moderately incompatible elements. Vesicles in phlogopite veinlets may indicate the presence (and later unmixing) of large amounts of volatiles (CO2, H20, Cl, F) in the silicic melts. Percolation of silicic magma, and possibly unmixed fluids, may cause cryptic metasomatism, even relatively far away from basaltic conduits (silicic glass inclusions similar in composition to phlogopite veinlets were found in xenoliths up to 25 cm in diameter). Although the composition of melt residue changes strongly during this process, the amount of melt hardly decreases. Therefore this process is capable of creating large amounts of highly silicic melts and has a great potential to cause widespread cryptic metasomatism.
The Canaries in a regional perspective: The European Asthenospheric Reservoir
The Canary Islands are part of a large "province" of Cenozoic volcanism, that also includes the Madeira Islands, the Western Mediterranean and the Central European Volcanic Province (CEVP, including the Eifel, Kaiserstuhl, Vogelsberg, Hessian Depression, Massif Central, Pannonian Basin) (Fig. 3). These volcanics were formed in a variety of geodynamic environments:
Fig. 3 Map view showing the extent of the low (less then global mean) S-wave velocity anomaly (LVA) at depths of 100, 300 and 500 km. The LVA becomes narrower at depth, and dips to the WNW. Volcanic provinces are denoted by shading (from Hoernle et al. (1995)).
The Madeira and western Canary Islands were formed on old oceanic lithosphere (see above). The Eastern Canary Islands erupted above the African Continental Shelf (transitional between oceanic and continental lithosphere). The Central European Volcanics formed as a result of continental rifting, through (thinned) continental lithosphere (possibly affected by ancient subduction). The Mediterranean Volcanics erupted in an area influenced by recent subduction.
The various volcanic regions produced a large variety of rocks, ranging from extremely undersaturated leucites and nephelinites to tholeiitic basalts. All regions display a large variety of Sr, Nd and Pb isotopic compositions. These compositions converge, however, on a relatively restricted isotopic ,component', with 87Sr/86Sr = 0.7030-0.7034; 143Nd/l44Nd = 0.51282-0.51294; 206Pb/204Pb = 19.9-20.1; 207Pb/204Pb = 15.62-15.68 (Fig. 4). This component is distinct from the Depleted MORB Mantle (DMM) component, and is not in end member in Cenozoic volcanic rocks surrounding the province (e.g. Iceland, Scotland).
Fig. 4 Plots of 206Pb/204Pb against 207Pb/204Pb (a), 87Sr/86Sr (b) and 143Nd/144Nd (c) for Cenozoic volcanic rocks from four geographical regions. The isotope data from these domains form fan-shaped patterns. The data array for each region converges on a common composition with 206Pb/204Pb » 19.9-20.1; 207Pb/204Pb » l5,62-l5,68; 87Sr/86Sr » 0.7030-0 ,7034 and 143Nd/144Nd » 0.51282-0.51294. This component is called LVC (Hoernle et al. (1995)) or EAR (Granet et al. (1995).
The different volcanic regions form trends on isotopic variation diagrams, that can be interpreted as mixing trends between a sublithospheric ,European Asthenospheric Reservoir' (EAR) and a local (lithospheric?) component. For the Western Canaries and Madeira this local component is MORB-like. The Eastern Canaries were influenced by a component with less radiogenic Pb and more radiogenic Sr than the LVC, probably located in the upper lithosphere. The second component for Central Europe is similar in trace element and isotopic composition to local granulite and spinel lherzolite xenoliths. Recently subducted crust, sediments and possibly a MORB-like component formed a more complicated trend in the Mediterranean volcanics.
The origin and nature of this reservoir is hotly debated in the literature. A tomographic study by Hoernle et al. (1995) revealed a large area with anomalously low S-wave velocity, indicative of relatively hot (possibly volatile-enriched) mantle material, below the Cenozoic province. This Low Velocity Anomaly (LVA) has a sheet-like form that is narrow and deep (up to 500 km) off the west coast of Africa, and outstretched and shallow (2100 km) below Central Europe and the Mediterranean (Fig. 3). The LVA covers an area of 2500 x 4000 km, which is much larger than expected for a typical large mantle plume. Also the peculiar ,inclined sheet' form doesn't correspond with a typical mushroom shaped plume head. Hoernle et al. (1995) interpreted this feature as an area of upwelling mantle material, forming a well mixed, homogeneous reservoir for the European Volcanics.
However, a more detailed tomographic study in the Massif Central (Granet et al., 1995) showed a much more localized region of anomalously hot mantle that extended downward to at least 400 km depth, and possibly to the 670 km boundary. This suggests that the EAR is a much deeper feature, from which small, relatively closely spaced 'plumes' rise up to form the separate volcanic fields.
Literature:
Granet, M.; M. Wilson; U. Achauer (1995). Imaging a mantle plume beneath the French Massif Central; Earth Planet. Sci. Lett. 136: 281-296
Wulff-Pedersen, E.; E.-R. Neumann; B. B. Jensen (1996). The upper mantle under La Palma, Canary Islands: formation of Si-K-Na-rich melt and its importance as a metasomatic agent. Contrib. Mineral. Petrol. 125: 113-139.
Hoernle, K.; Y.-S. Zhang; D. Graham (1995). Seismic and geochemical evidence for large-scale mantle upwelling beneath the eastern Atlantic and central Europe. Nature 374: 34-39.
Menzies, M. A.; C. J. Hawkesworth, eds. (1987). Mantel Metasomatism. Academic Press Geology Series. pp. 460.
Wilson, M. (1989). Igneous Petrogenesis (a global tectonic approach). Chapman & Hall. pp. 466.
Lithospheric flexure below oceanic islands
Volcanic islands and seamounts such as the Canary Islands form a load on the underlying lithosphere. As a result of this loading, the lithosphere bends. The magnitude of this flexure gives information on the elastic properties of the lithospheric plate.
The response to loading of a plate is described by the flexure equation, that results from a balance of forces and moments on the bending plate:
where D is the flexural rigidity, w is the deflection, x is distance from the load, P is the horizontal force working on the flexing element and q(x) is the downward force per unit area. In the case of a lithospheric plate overlying the asthenosphere, the flexure is countered by an isostatic force:
with rm and rw = the density of mantel and water respectively and g = the gravity constant. In the case of a plate bending under a line load (e.g. Hawaiian volcanic chain), these equations combine to:
The general solution of this equation is:
where the constants cl, c2, c3 and c4 are determined by the boundary conditions and
The parameter a is called the flexural parameter. The form of the bending plate is shown in figure 5: the plate bends deeply directly below the load, but on both sides it flexes upward above the original seafloor level (peripheral bulge, Fig. 5). It can be shown that:
The peak of the peripheral bulge xb can be measured directly from bathymetric profiles over the bending lithosphere (Fig. 6). This allows calculation of the flexural parameter a, and via equation 1 of the thickness of the elastic lithosphere, which is that part of the lithosphere that does not relax elastic stresses on geological timescales. World wide studies of bending oceanic lithosphere under loads showed that the thickness of the elastic lithosphere depends strongly on age, and therefore on temperature distribution. Generally, the lower boundary of the elastic lithosphere falls in the area between the 300 °C and 600 °C isotherms.
A thorough discussion of this and many other mathematical problems in the geosciences can be found in:
Turcotte & Schubert (1982): Geodynamics; applications of continuum physics to geological problems. Wiley & Sons, 450 pp.
Fig. 5 Half of the theoretical deflection profile for a floating elastic plate supporting a line load. the location of x0 (where the absolute deflection is 0) and xb (top of the periferal bulge) depend on the load and on the elastic properties of the plate.
Fig. 6 Bathymetry anomalies along a north-south line centred on the Hawaiian island of Oahu. The volcanic islands act as a load on the Pacific Plate bending it downwards and resulting in both a bathymetry and gravity anomaly (after Watts and Daly (1981)).