DISCUSSIONS


Ever since I became a pencil-pushing bureaucrat, much of my geological "work" consists of reading and pondering, camping trips to look at field sites in the northeastern U.S. and Canada, and maybe a little computer work (especially with the world's greatest geology software, IGPET).  Of course, working in the field and lab is not "science" until the work is published -- that is, you are not doing science unless you also communicate new data and discoveries via official channels.  Probably a web site does not count.  Prior to putting some ideas into a manuscript, however, it still seems like a reasonable idea to organize and summarize some of the basic tenets of any field of inquiry.

BASIC QUESTIONS

Where did  the name "CAMP" come from?

     Before the June 1, 1999 AGU session on this volcanic province, I suggested to Bill Hames that we canvass the participants about a name for the province that everyone could adopt, there being a mishmash of informal names in use.  Central Atlantic Magmatic Province or CAMP had recently been used by Andrea Marzoli and coauthors in his April 1999 Science article, and we agreed over lunch after the AGU session that it was the best.  How Andrea came up with it, I do not know, but  I expect CAMP will stick and become established as the "official" name.

What are the geographic limits of CAMP?

     Since the compilation of the geology of North America by P.B. King in the 1960's, we have known that in eastern North America, nearly all of the early Mesozoic dikes, as well as Mesozoic basins and lava flows, are bounded on the NW side by the highlands of the Appalachian Mountains (and virtually by the sharp gravity gradient along that orogen).  Other boundaries are not so clear, and new work keeps enlarging what we now call CAMP.  CAMP basalt is present to the west under the younger strata of Mississippi (Sundeen, 1989) and into Texas (A. Baksi, pers. comm.), and probable CAMP dikes have been sampled in the Gulf of Mexico (Schlager and others, 1984).  We know it occurs in the subsurface of Florida (Arthur, 1988).  In the south, there are extensive sills, flows and dike swarms extending through northern and central Brazil (Marzoli et al., 1999) about to the Bolivian border (Montes-Lauar et al., 1994).  There are hints of CAMP in Venezuela, and in the subsurface of the eastern Yucatan peninsula (still searching for ref's).  To the east, there are huge dike swarms in Liberia and Guinea (DuPuy et al., 1988) and great sills of CAMP basalt in Precambrian basement rocks of Guinea and Mali, and elsewhere in West Africa (Bertrand, 1991).  To the north, CAMP dikes occur through Iberia into NW France (Caroff et al., 1995), and perhaps there are CAMP basalts in Mesozoic basins even farther north.  My very rough map measurements make CAMP to be about 10 million square km in area.  It may well be larger.

Is CAMP the world's largest LIP (Large Igneous Province)?

     Certainly it is the largest in area, but judging from the exposed lava remnants, CAMP may "only" have totalled 2 to 3 million cubic kilometers of extrusive basalts (see the spreadsheet in the tables section).  The Ontong-Java Plateau, an undersea LIP, probably has much greater volume.  But there is also a large amount of CAMP tholeiite in dikes and sills, especially if we include the seaward dipping reflectors of eastern North America (about half of which may have been sub-aerial flows when formed).  Marzolli et al. (1999) report about 400,000 cubic km of basalt just in the Amazon Basin sills.  More work needs to be done on estimating CAMP dike and lava volumes.

What is the age of CAMP volcanism?

     Thanks to the fine detailed stratigraphic work of Paul Olsen and the dike/lava correlations of Tony Philpotts and his students, we know that in the northern region of CAMP, dikes and lava flows all formed within a span of about 580,000 years, and which the best dates show to be around 200 Ma (I like 201 Ma), plus or minus about 2 m.y.  Note that this plus or minus is for analytical precision, not necessarily the actual span of magmatic activity.  Stratigraphically, each lava flow event was less than 100,000 years long, probably much less, and each event covered most of the northeastern North America east of the Appalachian highlands!  The best new radiometric dates from the southern part of CAMP are similar to those from the north, although without the stratigraphic constraints.  In addition, there are some dates in South America, West Africa, and the SE USA near 196 Ma that also look good.  By way of comparison, the latest work on the Deccan Traps indicates the majority of those basalts formed at 65.6 +/- 0.3 Ma (Allegre et al., 1999).  On the other hand, the idea that volcanism continues, perhaps episodically and locally, for some millions of years appears to be true for many flood basalt provinces (such dates may be accurate for northern Brazil and adjacent Africa).

Is CAMP responsible for the mass extinction at the Triassic-Jurassic boundary?

    There are several lines of thought about this question.  Here are some pros and cons.

PRO:  There was an enormous amount of volcanism related to CAMP, and new dates show that much of it probably occurred during a very brief time -- perhaps as several big events that each lasted a few thousand years or so.  Some dates appear to overlap the age of the Tr-J boundary and extinction event.  Calculated amounts of volatiles emitted during eruption, such as CO2, SO2, and perhaps F and Cl, are likewise huge.  The magmatism included both intrusive and subarial volcanism that straddled the paleo-equator, thus contaminating both the northern and southern hemispheric atmospheres.  We know from the Laki, Iceland eruption of 1783 that even much smaller fissure eruptions can cause major environmental problems.  There is a very good time correlation of large flood basalt events with all of the big mass extinctions (including the K-T event), while hardly any of the numerous dated meteor impacts coincide with extinctions.  No large meteor craters of Tr-J boundary age are known, and a search for shocked quartz and iridium anomaly (impact effects) at the Tr-J boundary in Nova Scotia was negative (Mossman et al., 1998).  There is good paleobotanical evidence for a great increase in CO2 in the atmosphere at the Tr-J boundary (McElwain et al., 1999), as predicted from CAMP volcanism.

CON: Because of the excellent evidence that a very large bolide (meteor or comet) impact coincided exactly with the K-T extinction event, with wide-spread effects, there is a reasonable expectation that mass extinctions are caused by impacts.  There is some doubt that fissure eruptions push most of their volatile emissions high enough into the atmosphere for worldwide circulation.  In northeastern North America, the only locations where detailed stratigraphy of both lavas and the extinction horizon have been studied, show that the first CAMP lavas at those sites probably postdate the extinction by some thousands of years.  Many other CAMP dates show somewhat younger, not older, ages for volcanism outside of the detailed stratigraphic basins.  If CAMP dikes with the correct age exist elsewhere, their stratigraphic record as lava producers is poor.  Most of the required surface basalt is missing, we don't see where it could have eroded to, and there is, as yet, scanty evidence for volcanic ash, pollutants, and chemical changes in boundary sediments.  In the northeastern USA, fossils such as fish scales and dinosaur tracks are locally common in sediments deposited between CAMP lava flows (although the variety of species represented is not what was found before the extinction event).  Shocked quartz was described at the Tr-J boundary in Italy (Bice et al., 1992).

My thoughts: Flood basalt volcanism is a feasible cause for mass extinction, and it deserves more study as a CAMP hypothesis.  One reason why the initial lava in northeastern North America is separated by several meters of sedimentary strata from the extinction horizon is because lava flows are not the cause of extinction -- it is the eruptions at the fissures that really pollute the air.   Some CAMP lavas traveled great distances from their sources (dikes today, volcanic fissures then), perhaps 400 km or more, allowing considerable time to elapse between the initial volcanism/mass extinction and the arrival of a resultant lava flow.  See the interesting work by Stephen Self on the inflation mechanism for (slow) flood basalt flows.
 

What are the basalt groups that comprise CAMP?

     There are more than a thousand good chemical analyses published for CAMP dikes and basalts, and they can be readily studied with computer programs such as IGPET99 by Michael Carr (Terra Softa, Inc).  The analyses that are well pinned to specific dikes and basalt flows are especially helpful.  Some confusion exists because crystal fractionation and crustal contamination of the dikes and lavas create both linear trends and scatter in chemical diagrams, and groups may overlap.  A few groups of dike analyses are very uniform and make tight and distinct clusters on the diagrams, while other dikes and most lavas plots are somewhat smeared out.  We know how crystal fractionation has made changes in the surface lavas in relation to their dike magma sources.  We can only conjecture on the role of crystal fractionation in hypothetical magma chambers in the crust or upper mantle, and how such models may (or may not) fit geographic patterns of the proposed parent and daughter magmas.  There are many important references on grouping the magmas -- see my bibliography -- but here is my summary.
     In the northeastern USA and Atlantic Canada, there are four distinct magma types in dikes, three of which produced lava flows now exposed in continental basins.  We know the sequence of ages of the three lavas and thus the sequence of intrusion of three of the dike magmas.
     In the southeastern USA, the same three lava-magma types appear in dikes, but there is another major dike type - olivine tholeiite - which is not found anywhere else in the CAMP province.  It has its own NW-SE dike trend preference, and its dikes probably fed olivine basalt lavas that are preserved in the subsurface South Georgia rift terrane.  Because a large olivine basalt dike occurs in the Culpeper basin, but only the three northeastern quartz tholeiite lavas are now found in that basin, I infer that the olivine type is slightly younger and any Culpeper olivine basalt lava has been removed by erosion of the Jurassic strata in that basin.  This is rather conjectural of course.
     In addition, there is a pesky group of good-looking radiometric analyses in the eastern USA near 196 Ma, which could represent magma of the N-S quartz tholeiite dike group of the Carolinas and Virginia.  Chemically, these are hard to distinguish from some NE tholeiites, but because of their distinct dike trend group and the (very few) younger ages, they may form a separate magma type (or at least a separate generation).
     In west Africa, northern South America, and possibly Florida, a group of basalts with very high titanium (3 to 2 weight percent or so) is found.  These analyses tend to be scattered along a vague Ti-Fe trend on diagrams, and their relationship with the other types is obscure.  But because they might be highly fractionated and perhaps can be compared with high-Ti basalts that appear late in the magma sequence of other flood basalt provinces, I would provisionally assume them to be the youngest type.

     Thus from oldest (bottom) to youngest (top), I propose:

Type    Dike (example)      Lava (example)          Possible age

7.        high-Ti diabase       high-Ti basalt             196?
6.        (Pageland)                ?                             196
5.        low-Ti diabase        olivine basalt              199?
4.        (Shelburne)              ?                              200?
3.        (Bridgeport)           (Hampden)                201
2.        (Buttress)               (Holyoke)                  201
1.        (Higganum)            (Talcott)                     201 Ma (types 1, 2, and 3 span 580 K years)

     Type 1 magma is the oldest and most widespread, occurring everywhere except for the southern USA.  Therefore it has great significance in geodynamic and petrologic models.  Type 5 occurs in a central location in the province, and is probably also a "primary" magma, and so also has special importance.  The other types may represent specific mantle magmas, but because they occur later and are often not very Mg rich, they may be best modeled as differentiates from other primary melts.

     It is highly likely the above grouping will be modified in the future, as better data arrive, and as better geologists make sense of the data.
 

Is CAMP formed from a very large plume (mega-plume), or is a different geodynamic model indicated?

     My thoughts on this are expressed in my Tectonophysics paper, a version of which you can get to from the index page of this website.  In addition, read the abstracts from the AGU meeting, linked on the index page, for other recent opinions.  There are also pertinent abstracts that assume or promote plume models available from the Penrose 2000 conference, linked on the index page.
 

To be addressed:

How is CAMP related to rifting events expressed by the Mesozoic basins?

How is CAMP related to the seaward-dipping reflectors of eastern North America, and to the initial production of the Atlantic seafloor?

What mechanism could produce CAMP basalts over such a large area in a brief timespan?

What controlled the compositions of CAMP magma types and their geographic distribution?

Did CAMP dikes flow vertically or horizontally?