Some days ago I got into a rather enthralling conversation about weird magma types. It revolved around a particular kind of magma called nephelinite. Some of you may know about this magma variety. Or perhaps not. It is after all a word that doesn’t show up as frequently as basalt or rhyolite. If this name ever shows up here it is probably because of the somewhat disproportionate passion that a few commenters of this blog, including myself, feel towards it. Nephelinite belongs to a family of rare magmas, such as carbonatites and kimberlites. Their composition is very different from that of what is considered to be a “normal mantle”. Something that makes them even more interesting is that they make up a tiny fraction of our the Earth’s whole volume of volcanic rocks. In fact they are the rarest of them all.
Presently this rare family of magmas is being erupted mostly from the East African Rift: the nephelinite lava lake of Nyiragongo, the carbonatite lava flows of Ol Doinyo Lengai, or the kimberlite tuff cones of the Igwisi Hills. What few may know is that during the Miocene southwest Germany turned into a haven of such rare volcanism, in what is a basalt-andesite-rhyolite dominated planet. This is the story of those rare volcanoes from the past. But first let’s have the geologic context.
The European Cenozoic Rift System
Why are there volcanoes in Germany? I won’t go into this question because it is a complicated one. I could write a book about such a question. There are many similar volcanoes in unexpected locations of the planet, mysterious and unexplained. But let me at least point out that these volcanoes seem to be part of a large band of “recent volcanic activity” that runs from Romania, with the Ciomadul dome complex, then across Hungary, Slovakia, Czechia, Germany, France, Spain, to finally reach Morocco. I would place the end of this volcanic group in northern Morocco, where the possibly active Oulmes and Azrou volcanic fields are located.
These volcanoes are spatially linked to a series of rifts that together make up the so called European Cenozoic Rift System. This structure is seismically active today. The map below shows its various segments, most of which are associated to volcanic activity. The rift we are interested in is the Upper Rhine Graben, the URG.
The largest volcano of Central Europe is the Vogelsberg volcano, located within the Upper Rhine Graben. This volcano is mostly made up of basaltic lava flows, which piled up to a thickness of up to 800 meters, making a shield volcano with a diameter of ~50 km. The total erupted volume is 600 km3. The activity of Vogelsberg lasted from 10 to 20 Ma (millions of years ago), but the most intense phase was between 15 and 18 Ma. Eruptions would have been very infrequent. Dormancies may have taken centuries, thousands, or even tens of thousands of years. Each eruption would have formed a new vent and produced lava flows. The largest flows may have been long-lasting, taking years to grow, and comprised large volumes >1 km3. This description comes from comparing with other younger volcanic fields in Eastern Australia, and in Arabia, that resemble Vogelsberg, both in chemistry and morphology.
The chemistry of Vogelsberg lavas is not unusual when you look individually at the types of rocks that it has erupted. However when taken all together Vogelsberg comprises a fantastic suite of lavas spanning much of the “primitive lava spectrum”. What do I mean by primitive? One of the basic concepts relating to volcanoes is the idea of fractional crystallization. Magma can stall within intrusions during its ascent towards the surface, rather than erupting directly from the mantle. It is here that fractional crystallization happens. Magma will gradually lower its temperature and as this happens crystals will grow in the melt. The crystals will sink away. This process will take away certain elements contained in the magma, changing the composition of the residual melt, which in turn will keep moving upwards and maybe fractionating further. Magma types like rhyolite, trachyte, or phonolite are the culmination of certain fractional crystallization series.
The lavas erupted from Vogelsberg are mainly tholeiite basalts, alkali basalts and basanites. All of them are primitive. No significant fractional crystallization has taken place in their ascent. Thus primitive. As such the variety of magmas that have been supplied to the volcano is remarkable.
Tholeiites are the most common primitive lavas erupted in planet, and that includes almost all the lava that is erupted from mid-ocean ridges, where most volcanism takes place. Alkali basalts and basanites get increasingly rare. It is said that they are more alkaline. Alkalinity is an important concept when it comes to volcanism. Tholeiites would be least alkaline, followed by alkali basalts, and then basanites. The next step after basanite would be melilitite, followed by kimberlite. You can see all this magma/lava classes in the diagram below, which is called a TAS diagram.
At the same time as Vogelsberg was erupting, a small volcanic field developed 235 kilometres to the south. 350 small volcanoes erupted here between 16-17 Ma. The town of Bad Urach stands nowadays within the center of this ancient volcanic field. Because of this the volcano is known as Urach. It might seem like a small non-important volcano at first sight, but if you look at the lavas you will see something different. Urach erupts melilitite. Melilitites are very rare, alkaline, silica-undersaturated lavas. Actually normally I would say Urach erupts nephelinite, but since I’m writing an article I should be technically correct. Nephelinite and Melilitite might reside within the same area in a TAS diagram but I’m sure there are important mineralogical differences between the two.
“Rocks in the alkaline magma series are distinguished from rocks in the subalkaline tholeiitic and calc-alkaline magma series by their high content of alkali metal oxides (K2O plus Na2O) relative to silica (SiO2).”
This is a quote taken from Wikipedia. I think it is a poor definition of what alkalinity really means, not what definitions say, but what you actually see when you plot the data yourself. A primitive magma gets more alkaline when it has less silicon and aluminium, and more of practically every other element which is present in the magma (there are a lot of them). Each different type of primitive magma is going to give rise to a different fractional crytallization series. The crystallization series of basanite could be said to be more alkaline than the crystallization series of basalt, because basanite is more alkaline than basalt. The following are examples of crystallization series:
Tholeiite basalt > Basaltic andesite > Andesite > Dacite > Rhyolite
Basanite > Tephrite > Tephriphonolite > Phonotephrite > Phonolite
But some magma types are so rare they don’t ever seem to make a crystallization series. It was about 10,000 years ago, in a remote corner of Tanzania, that a dike rose to the surface making three small tuff cones: the Igwisi Hills. The eruption was tiny. It probably didn’t last more than a few days. And yet it ejected what is probably the single-most weirdest lava composition ever to have erupted in the entire Quaternary period, at least as far as we know. The most silica-undersaturated kind of lava, a kimberlite.
This plot shows Vogelsberg, Urach, and Igwisi Hills. Vogelsberg erupts what corresponds roughly to the entire “typical” range of primitive lavas, from basalts to basanites. Urach has an extreme lava composition. Igwisi Hills is the extreme among the extreme. The following plot shows the average of magnesium oxide+calcium oxide, versus the average of silica+aluminium oxide. The more magnesium and more calcium a lava has, which relates to it being more alkaline, the less silica and aluminium it has.
So how would the eruption of such silica undersaturated magmas look like? One may recall that less silica means more fluidity. Within molten rock the atoms of silicon will bond with oxygen making silica tetrahedra, aluminium will behave similarly, occupying the same place as silicon in the tetrahedra. These silica tetrahedra will link up to make polymers. More polymers means higher viscosity. That is why aluminium and silicon control the viscosity of magmas. Here are the average silica (SiO2), and aluminum oxide (Al2O3) contents in lavas from the volcanoes discussed:
Vogelsberg (tholeiite end member): 54 wt% SiO2, 14.3 wt% Al2O3.
Vogelsberg (basanite end member): 40.5 wt% SiO2, 12.8 wt% Al2O3.
Urach melilitite: 35.8 wt% SiO2, 8.1 wt% Al2O3.
Igwisi Hills kimberlite: 20.9 wt% SiO2, 3.2 wt% Al2O3.
It is usually claimed that tholeiitic basalts like those from Hawaii, or Galapagos, or the Mid-Atlantic Ridge, are silica-poor and thus fluid. In reality they are near the silica-rich end of primitive lavas; they are more or less are in the middle of the whole silica spectrum of lava compositions. There are many myths in geology textbooks. Of course textbooks like to put things simple. I like things complicated. A lava from Urach should be more fluid than lavas from Hawaii, as long as other factors such as temperature stay similar, which I don’t think are too different, particularly with La Palma’s 2021 basanite eruption having had the same temperature as Kilauea’s tholeiite. Mid-ocean ridges might be somewhat hotter though, if certain numbers from Fagradalsfjall I’ve seen claimed are to be given credibility.
But there is yet another textbook misconception I want to destroy. Fluid magmas are less explosive right? In some ways yes, but in others not. Below are amounts of water and carbon dioxide, the most abundant magmatic gasses, from different magmas in order of increasing alkalinity as reported in various scientific articles. All of the articles are quoted near the end of this post. Volatile contents are difficult to measure so there is a substantial spread in data:
East Pacific Rise tholeiite (Siqueiros fault): 0.037–0.122 wt% H2O, 0.0044-0.0244 wt% CO2.
Mauna Loa, Hawaii, tholeiite: 0.09-0.87 wt% H2O, 0-0.021 wt% CO2.
El Hierro basanite: 0.4–3.0 wt.% H2O, 0.006-0.34 wt CO2.
Ol Doinyo Lengai nephelinite: 0.7-10.1 wt% H2O, 2.7-8.7 wt% CO2
You may observe that gas content increases substantially as magmas grow more alkaline. Tholeiites are very poor in volatiles. That makes tholeiites relatively unexplosive, producing little tephra when they erupt. However the gas content increases by about two orders of magnitude when reaching highly silica-undersaturated magmas.
Melilitites and kimberlites seem to almost always be associated with diatremes/maars. This is also the case of Urach. Maars consist of large explosion craters, sometimes as much as 3 km across, surrounded by a low ring of ejecta. Sometimes the crater is inundated with water. It has long been thought that maars were formed because magma interacted with groundwater leading to large steam-driven explosions. But now there is evidence that melilitite has enough magmatic gas of its own to blow open a maar without any help from groundwater. These are the so-called dry maars. Such dry maars have been recently documented around Ol Doinyo Lengai, and in the Calatrava Volcanic Field of Spain. This has been a known mechanism of kimberlites. Now it appears that a lot of maars previously attributed to groundwater interaction may actually have to do with magmatic-gas-driven explosions of melilitite magmas. In fact I think it might eventually turn out that the majority of maars are “dry maars”.
Highly silica-undersaturated magmas are very explosive, although they lack the huge scale of silicic magmas like rhyolite or dacite. Viscous magmas such as rhyolite require very wide conduits to erupt, due to their high viscosity, say a 100 meter wide pipe, at least. There will be a lot more magma that can come up a 100 meter rhyolite pipe than through a 1 meter basalt, basanite, or melilitite pipe. Because of this, rhyolites, and other highly silicic magmas, are capable of erupting a lot in a very short amount of time. It leads to spectacular vulcanian explosions, lateral blasts, long distance pyroclastic flows, and powerful plinian eruptions. However rhyolites might actually fall behind fluid highly silica-undersaturated magmas in terms of gas content. That gas is key in shaping melilitite eruptions like those of Urach.
The Urach volcanic field is largely eroded, but one can imagine how it would have looked like when it was active, by looking at present analogues, like those in the Albertine Rift, or Tanzania. Eruptions must have been short and small, mostly lasting from hours to a few days. VEI numbers would have been 2-3. Some weaker eruptions may have produced columns of ash and small tuff cones. More powerful eruptions would have blown away the bedrock, making craters up to more than 100 meters deep. Ash surges would have swept out from these craters destroying vegetation and killing animals. The very high magma gas content would put any witnesses in a particularly severe risk to die from suffocation.
The closest present day analogues to Urach are probably the Katwe-Kikorongo and Bunyaruguru volcanic fields on the Albertine Rift. If one had stood at Urach, 17 Ma ago, the landscape may have been similar. Craters on top of craters. Mainly explosive tuff on the ground. Numerous vents would be flooded with water, making maars.
While Vogelsberg stood at the northern end of the Upper Rhine Graben, and Urach was an off-rift volcano, yet a third volcano was also erupting at the same time from the southern end of the UPR. It was an stratovolcano, known as Kaiserstuhl. The activity of the volcano lasted 18-16 Ma. It was contemporaneous with the peak activity of Vogelsberg, and the formation of Urach. Kaisertuhl was the European version of Ol Doinyo Lengai. It erupted a great variety of evolved lavas, including nephelinites, tephrites, phonotephrites, and phonolites, as well as rare carbonatite magmas.
I have mentioned Ol Doinyo Lengai several times already. This is an inordinate one because it is the only presently erupting carbonatite volcano on the planet (possibly there might be other occasional carbonatite eruptors in the EARS, I haven’t looked up at every single one of them, yet).
Carbonatites are odd magmas. At Ol Doinyo Lengai such lavas are black during the day, they glow deep-red at night, and turn white when solidifying. Carbonatites can remain molten to temperatures as low as 650ºC, so they loose the daytime glow, and are more fluid than silicate lavas.
There is much evidence that carbonatite is formed from nephelinite. As nephelinite cools and crystallizes in upper level storage it increases in carbon dioxide until reaching saturation, then the melt partitions into two liquids, a carbonate liquid, and a silicate liquid, each being immiscible with the other. Some elements like calcium or phosphorus go mainly into the carbonatite melt. Other elements like silicon, aluminium or iron stay mainly in the silicate melt.
The Kaiserstuhl stratovolcano, now heavily eroded, would have originally stood more than 1 kilometre above its surroundings as a symmetrical cone of ash and lava. The main magma fractionation series was basanite>tephrite>tephriphonolite>phonotephrite>phonolite, all of which were erupted by Kaiserstuhl. This series constructed most of the volcano. Likely it was formed through a combination of lava flows and subplinian eruptions emitted from a long-lived conduit system in the center of the volcano. It’s way most stratovolcanoes are built. Minor magmas include nephelinite/melilitites, and carbonatites, probably belonging to a separate series of more silica undersaturated magmas.
In this sense it is similar to Ol Doinyo Lengai which is known to have gone through three phases of distinct composition. Ol Doinyo Lengai grew to its present height from phonolite tephra and lava. One side of the cone later collapsed and filled with a younger cone of nephelinite. Presently it erupts both carbonatite and nephelinite. This third phase only constitutes a minor mantling on top of the cone. During historical times it has either produced low spattering of carbonatite in its summit crater, with occasional flows reaching outside the crater and sweeping down the flanks. Or it has erupted nephelinite explosively, sometimes sending black ash plumes to an altitude of 15-17 km and generating pyroclastic flows down the vertiginous slopes of the cone.
Kaiserstuhl would as far as we know have erupted similarly. It contains beds of teardrop shaped carbonatite. Teardrops, known as Pele’s tears, would have formed from carbonatites being sprayed into the air as fountains. The heart of Kaiserstuhl makes up a large cylindrical carbonatite intrusion that covers an area of 1 km2. I find this large intrusion enigmatic. The way it was formed is enigmatic. Somehow the magma ate away the walls of the central conduit and turned it into an enormous mass of carbonatite. This is not rare for stratovolcanoes though. One wonders however if it was all molten at once or if it was formed incrementally. Carbonatite also made dikes ranging from less than 1 centimeter in width up to 1 meter. Those dikes likely fed outbursts of carbonatite lava from the flanks of the stratovolcano.
Taken together, Vogelsberg and Urach display much of the spectrum of primitive lavas that can be erupted on Earth. Urach has rare silica-undersaturated lavas that possibly led to a style of volcanism somewhat linked to kimberlites, or as close as you can get without erupting an actual kimberlite. Kaiserstuhl is one of the best studied examples of carbonatite volcanism in the world. These three volcanoes, that were contemporaneously active, within and near the Upper Rhine Graben, make up an unusual volcanic episode that is interesting to study.
It will be hard to match the peculiarities of the Rhine rift volcanoes, but I will be writing a few other upcoming articles about European volcanoes. You may find that the European continent has some surprising examples of volcanic activity. Both extinct and active.
Ol Doinyo Lengai:
Discussion on maars that are driven by magmatic gasses
Sources of magma volatile data:
VC articles mentioning the general regions, but covering different topics