Large Eruptions and Extinction Events
One of the things I have been fascinated with is the connection between large volcanic eruptions and their impact on the environment and in turn their impact on life itself. We are familiar with what happens in the vicinity of an eruption to the population of humans, wildlife and plant life. Generally, the larger the eruption or event, the larger and more persistent the impact.
Some historic eruptions like Tambora in 1812 or Krakatau in 1883 have global impacts in terms of lives lost, global temperature changes for a few years, loss of growing seasons, and coverage of thousands of km2 with ash and other eruption products. Tambora produced around 100 km3 of material over the course of a few days.
Interestingly enough, some large historic eruptions like Baedeku which also produced around 100 km3 of material did not appear to have a global impact. The question is why. Some suspect the location of the eruption is also a factor in its global impact.
Still larger individual eruptions in the VEI 7 – 8 range are popularly referred to as super eruptions. Here in the US, Yellowstone, Long Valley and Valles Caldera are the most recent. These eruptions produce thousands of km3 of material over the course of an eruption and directly impact tens to hundreds of thousands of km2 of surrounding lands and oceans with products of eruptions.
The largest sequence of eruptions creates things called Large Igneous Provinces (LIPs). These are not individual eruptions. Rather they are massive outpourings of primarily basalts over the course of a few million years.
The most recent flood basalt is the Columbia River Basalt Group. It covered over 210,000 km2 in Oregon, Washington and Idaho between 16.7 – 5.5 Ma. There are more than 350 lava flows that contributed to this LIP. 93% of the flood basalts were produced in a mere 1.1 Ma centered around 16 Ma. Total volume of the Columbia is estimated to be over 174,000 km3. Do the math, and you have a nearly 500 km3 eruption every 3,400 years over the 1.1 million years of highest activity.
LIPs are often associated with continental rifting. They may also be tied to hot spots and mantle plumes, deep sources from the mantle that move hot, new material from deep in the mantle to the crust.
When a LIP gets large enough, and is situated in the right place at the right time, you end up with a major extinction event, something which kills an appreciable percentage of life on earth. The two largest and most recent of these are the Cretaceous – Tertiary (K – T) event some 65 Ma, tied to the Deccan Traps, and the Permian – Triassic (P – T) event some 252 Ma, tied to the Siberian Traps. Total volume of the Deccan Traps is 512,000 km3. Total volume of Siberian Traps is estimated to be 3 – 5,000,000 km3 or 6 – 10 times the volume of the Deccan.
The K – T event killed off the dinosaurs (not all of them, as many survived as birds) along with about 75% of all species on earth at the time. The P – T event was much worse and killed off 96% of all marine species, and 75% of all land vertebrates.
The K – T event is also associated with a massive impact that created a crater 180 km in diameter centered on the coastline of the Yucatan Peninsula of Mexico. There are ongoing arguments on the kill mechanism associated with the Siberian Traps, with at least one possible impact structure identified.
Some 200 Ma at the end of the Triassic and beginning of the Jurassic, yet another extinction event. This one wiped out killed over 50% of all species on earth. It is associated with the Central Atlantic Magmatic Province (CAMP) which was active at that time. I have found no second mechanism associated with this die-off.
The end of the Eocene some 35 Ma was a minor extinction event. This one has an associated LIP in the Afar region and at least three identified impact craters, a 100 km diameter crater in Siberia, an 85 km diameter crater in the Chesapeake Bay region of the US, and a third smaller one offshore New Jersey. It also is toward the end of an ignimbrite outbreak in the Central US – Colorado, Utah and Nevada.
Over the years, I have concluded that life on this planet is well adapted to volcanic events large and small, as they are always present at some level and have been since life began. The LIP Commission has a database with over 300 LIPs dating back over 3 Ga. As most of these are not associated with an extinction event, I am wondering if two massive events near simultaneously are needed to trigger an extinction event.
For instance, current description of both the K – T and P – T events has the associated volcanism taking place over a few millions of years, with the impact event for the K – T taking place during the volcanic activity.
Perhaps the LIPs do stress life and a second large insult (large impact, gamma ray burst, close supernovae, etc.) to the biosphere is enough to trigger serious extinction of large numbers of species.
Ontong – Java Plateau
With that introduction in hand, I came across a LIP that was mostly emplaced around 120 Ma. This is the combined Ontong – Java Plateau (OJP), Manihiki Plateau and Hikurangi Plateau region. Primary volcanism was completely subsurface and took place around 120 Ma. There was a smaller magmatic pulse around 90 Ma.
The Ontong – Java Plateau (OJP) is huge, covering some 1.5 million km2 of ocean floor, equivalent to the area of Alaska or western Europe. Total volume is estimated to be in the vicinity of 80,000,000 km3, some 16 times larger than the Siberian Traps. This is the largest LIP of the last 200 Ma. Crustal thickness beneath the plateau is some 35 km, approaching that of continental thickness.
The early and most prolific period of emplacement coincides with an oceanic anoxic event some 124 – 122 Ma. The thinking is that the event was caused by gasses contained in the magma as it met the ocean. Most of the lavas were emplaced at a depth between 2,500 – 1,100 m. There is no evidence that lava was emplaced above the ocean’s surface, though there is some evidence of phreatomagmatic debris meaning emplacement was relatively close to the surface at least a few places on the plateau.
Over the millions of years, rifting separated the Manihiki and Hikurangi Plateaus from the Ontong – Java. Today, the Manihiki Plateau is some 2,500 km to the east of the Ontong – Java and the Hikurangi is some 3,000 km south of the Manihiki. Today they are thought to be related because of some chemical similarities in their lavas. Primary formation for all three is around the same time. Problem is that the Manihiki and Hikurangi Plateau deep lavas are still relatively unknown.
The region is located on the Pacific Plate and was governed for a very long time by spreading centers on the Pacific Plate itself. Some researchers believe that the original weak point in the plate that allowed all the magma to reach the surface was a failed spreading center in the area covered by the OJP.
There are three places where the OJP is breaking the surface due to a couple kilometers of uplift. The largest is the Ontong Java atoll, alternately named Luangiua or Lord Howe atoll. It is near Nukumanu atoll and the Roncador Reef. The Takuu atoll lies some distance to the west. The Ontong Java atoll is huge, measuring some 1,400 km2 and hosts a population of around 2,000. Inhabitants are Polynesian and have occupied the atoll for perhaps 2,000 years.
The Cretaceous was a very long time ago. Since then, a lot has changed tectonically. For instance, at that time, Antarctica, India, Australia were in the midst of separating. The south end of Africa and South America had separated from the larger Antarctica mega-continent; The Atlantic was just starting to form, with North American separating from the western portion of Africa and opening both the Atlantic and Caribbean. Africa has not yet impacted Europe and free passage on the Tethys Sea may have been possible worldwide.
The area we are interested in was in the western portion of the greater Pacific Ocean.
Over time, the Atlantic continued to open, India impacted the southern portion of Eurasia, and the Australian Plate started it quick movement north. It is the interaction of the Australian Plate with the Pacific Plate that has created the complex region of platelets, blocks, trenches, and uplift that marks the region from Indonesia to north of New Zealand.
The actual shape of the Pacific Plate over the last 120 Ma is unknown, especially at its complex active southern boundary now impacted by the Australian Plate. What we do know is the area of interest from the western end of New Guinea east to the eastern end of the Solomon Islands.
On a gross level, the Pacific Plate is generally moving NW and the Australian Plate is generally NNE. There are a series of trenches on the north side of New Guinea stretching west to east along the northern side of the Solomon Islands. These trenches include the New Guinea Trench, the Melanesian Trench, the North Solomon Trench, and the Vitaz Trench.
The complex region has a number of smaller plates and platelets jostling between the larger gross motion of the impacting Australian Plate from the south and the generally NE-ward motion of the Pacific Plate. From west to east these include the Caroline Plate (wholly in the Pacific Plate), the Bismark Sea Plate (north of New Guinea), the Solomon Sea Plate (south of the Solomon Islands), the Woodlark Plate (south of the Solomon Sea Plate), and the Solomon Islands Plate. The interaction of these plates and platelets drives much of the volcanic and tectonic activity from New Guinea eastward to the Solomon Islands.
Alternate theories abound about the existence of these smaller plates.
Collision of the OJP with the Solomon Trench slowed subduction of the Pacific Plate and began subduction of the Australian Plate below the Solomon Island arc. Both subduction zones may have been active between 10 – 3 Ma until Pacific Plate subduction broke down due to the thickness of the OJP. After that, the province was uplifted and deformed.
Volcanic and tectonic activity of the Solomon Islands is due to subduction of the Solomon Plate under the Pacific Plate which in turn has uplifted the OJP to the point where some of the higher points are now atolls.
Subsurface oceanic plateaus are today thought to be the oceanic equivalent of continental flood basalt emplacements. As such, their cause has been hotly debated over the years. The next section discusses an impact mechanism for creation of oceanic plateaus.
A March 2004 piece by Engle & Coffin in Mantle Plumes makes the case that a large impact event as the cause of the Ontong – Java Plateau formation along with four neighboring plateaus. These include the related Manihiki to the east, and the Nauru, East Mariana and Pigafetta Plateaus to the north. The piece makes the following arguments against a mantle plume source for the OJP:
- Should a mantle plume be involved, various models predict seafloor uplift of 1 – 4 km depending on arrival of a 1,000 – 2,000 km diameter plume head, with dynamic models predicting 3 – 4 km uplift and an isostatic model predicting 1 – 3 km. Actual uplift of the crust falls far below predictions of all three models. This is a problem for the mantle plume theory.
- Emplacement of the volume of lavas capable of turning a 7 km thick oceanic crust into a 35 km thick plateau should have significant aerial volcanic activity. Other than some near surface phreatomagmatic activity, the majority of eruptions were deep below the surface.
- Subsidence of oceanic structures over the course of 120 Ma is in the neighborhood of 3 – 3.8 km. Subsidence of the new OJP crust was only 1 – 2.8 km.
- There is no connection to any hot spot track or current hot spot based on the geochemistry of erupted lavas. The Louisville hot spot was originally proposed as a starting source.
- While there is a 1,200 km by 300 km deep low velocity zone beneath the OJP, its volume is too large to be explained by mantle material remaining after extracting the massive eruptive products needed to form the OJP.
They then go on to propose an impact by a 20 km diameter body. This would create a crater some 200 km in diameter, with a penetration depth of 60 km. Such an impact would trigger massive decompression melting to a minimum depth of 300 km. It is this region of melt that then supplies the source for OJP and neighboring magmas. The fossilized low velocity zone would be the right size and shape. The impact would also explain the anoxic event and some marine fauna extinctions related to the anoxia event.
The problem is that this impact is half again as large as the K – T Chicxulub impactor and there is no associated widespread terrestrial extinctions known.
On the other hand, with over 70% of the earth’s surface covered with oceans, craters in oceanic crust are few and far between. Theoretically, the oceans should have at least twice as many craters in them as the continents do. Yet they don’t. Does this mean that oceanic crust for whatever reason tends to heal itself? Are oceanic plateaus the calling cards of large impact events? There is already speculation that oceanic hot spots are caused by antipodal impacts.
Sometimes the important, massive things are completely hidden. Such is the case with the OJP. It was created by the most massive outpouring of lavas over the last 200 Ma. It was emplaced relatively quickly in geologic time. While pouring out 16 times as much lava as the Siberian Traps did, there was relatively little environmental impact at least from the perspective of extinction events.
While it is volcanically inactive today, it is not tectonically inactive as the ongoing collision between the Australian Plate and Pacific Plate is busily uplifting the newly created crust to sea level and above.
There is so much we don’t know about this event. Where did the magmas come from? Why did they breach the surface at that particular place? How did it start? What ended it? Are the two related plateaus really part of the original? What moved them as far as they moved? Humanity has not been on this planet long enough to see an actual flood basalt. Just because we haven’t seen oune yet doesn’t mean there won’t be another one sometime in the future.
As always, more study is necessary to characterize exactly what we are seeing with the OJP. The more we find, the more questions arise.