Union College, Schenectady NY
Recommended citation: Shoemaker, S.J., 2000, The exhumation history of the Pyrenees using detrital fission track thermochronology; Union College, Schenectady NY, Unpublished BSc thesis, 103 p.
This study focuses on the use of detrital zircon fission track thermochronology to determine the exhumation history of the Pyrenees mountain chain (Figure 1). Detrital zircon fission track thermochronology finds the cooling ages of individual zircon grains, in this case collected from either side of the Pyrenean orogen. Once the cooling ages of the zircon grains are known, the “lag time”, or the difference between the time it takes the grain to travel from the depth at which it was cooled to the time at which the grain is deposited, can be determined (Figure 2). The timing and patterns of exhumation help to explain the structural history of the orogen.
To determine the exhumation of the Pyrenees, samples were collected from the Ariege River in France, draining to the north of the orogen, and from the Segre River in Spain, draining to the south of the orogen. Four stratigraphic samples were also taken from the southern border of the Aquitaine foreland basin in France, whose deposition times covered from ~50 Ma to ~30 Ma, the main phase of Pyrenean activity. Using cooling ages gained from these samples, this study attempts to define the exhumation patterns in the Pyrenees.
HISTORY OF EXHUMATION STUDIES
Mountain chains expose rocks generally formed at very deep levels compared to surrounding areas. For a long time, geologists have wondered how much “cover” rock has been removed to expose the deeply exhumed rock, and how was this cover rock removed. Exhumation is “the unroofing history of a rock, as caused by tectonic and/or surficial processes” (Ring et al., 1999). As exhumation proceeds, isostatic forces cause readjustment and rocks are brought up from depth. Exhumation can occur just about anywhere and in any tectonic setting, especially at continental convergent collision zones, such as that of the Pyrenees (Ring et al., 1999). It is generally agreed that exhumation is accomplished by three processes: normal faulting, ductile thinning (crustal extension), and erosion (Ring et al., 1999). Erosion is a surficial process that depends heavily on climate and topography of a region. In general, humid regions experience faster rates of erosion than do arid regions, vegetated areas experience less removal of material than do barren areas, and mountainous, tectonically active regions experience more erosion than areas of low relief (Ring et al., 1999). Normal faulting, responsible for unroofing mid-crustal rocks, and ductile thinning, a contributor of the unroofing of metamorphic rocks, are both tectonic processes associated with the exhumation of orogenic systems (Ring et al., 1999). The exhumation of surface rocks, either by surficial processes or by tectonic thinning of layers, is a major factor in mountain building and the formation of topography. As rocks are stripped away, the lower rock rises in isostatic response, maintaining mountainous topography.
Exhumation of orogenic belts can be studied in a variety of ways, but none of the currently employed techniques provide a complete picture of the evolution of an orogenic belt. Some studies examine the sediment flux from an area, or analyze the volume changes in sedimentary basins over time, and then relate the changes back to the processes that were active during exhumation. These studies attempt to determine the amount of sediment removed during exhumation by evaluating a sediment budget. This technique contains considerable error associated with the loss of sediment to adjacent basins (Ring et al., 1999). Another approach is to examine the cooling ages of bedrock in the orogenic system. This method only provides ages for one piece of the geologic time frame, and makes it difficult to know the entire history. This technique is also limited to active orogenic systems. It is also possible to examine the cooling ages of detrital sediments that have been exhumed and deposited in adjacent basins. These sediments document the long-term record of orogenic development. Cooling ages of the detrital sediments allow most of the orogenic episode to be studied, and from this sedimentary record the exhumation history can be reassembled to provide a picture of the cooling history of the source area. One advantage of this technique is that it allows the evaluation of ancient orogenic systems that have long since eroded away. However, one disadvantage is that the evidence of exhumation is fragmentary and needs to be reassembled. Detrital cooling ages represent the time since closure and therefore they reflect the rate that rocks moved through the system to get to the surface. By determining the “lag time” between the cooling of the rocks and the deposition of the individual mineral grains, the rate of exhumation can be determined.
Fission track studies concentrate on the cooling ages of rocks, and relate those ages to exhumation and exhumation processes. As the rocks are exhumed, they eventually reach a certain depth and temperature, depending on the mineral, at which the fission tracks themselves are essentially retained within that mineral, and the FT clock starts. The zone through which the mineral retains fission tracks is known as the Partial Annealing Zone (PAZ) (Wagner and van den Haute, 1994), the zone of depth/temperature conditions that allow 10%-90% of tracks to be retained. For zircon, the closure temperature is ~240° C, and the depth is ~7 km (Brandon et al., 1998).
In general, continental collision zones show the deepest levels of exhumation, sometimes exhuming rocks from depths of up to 100 km (Ring et al., 1999). In more typical settings, mid-crustal rocks are unroofed, and the hanging wall is stripped away. The removal of the hanging wall has the effect of setting the footwall rocks to a common isotopic age due to their rapid cooling (Ring et al., 1999). It is this common isotopic age of exhumed minerals of progressive depths that FT studies depend on. Examining these exhumed minerals for their thermal record of time introduces two major concerns. One is the temperature of closure, or the closure depth, at which the fission tracks are retained in the chosen thermochronometer. The second concern is that of the rate of exhumation. It is important to know about how long it took to remove the layer of rock above the closure depth. The syn-orogenic rock is below the closure depth prior to orogenesis, and therefore records the time at which orogenesis forced those rocks up through the closure depth. Therefore, these FT thermochronometers provide a manner to estimate the time taken for rocks to travel from the depth of closure to the surface and subsequent estimation of the rates of exhumation in the system can be made.
The Pyrenees stretch east-west between France and Spain, along the northern boundary of the Iberian peninsula (Figure 1). The Pyrenees are flanked by the Ebro foreland basin in the south (Spain) and the Aquitaine foreland basin in the north (France), as well as several smaller basins within the orogen (Verges et al., 1995; Meigs et al., 1996). The Pyrenean Orogen was formed by a collision between the Iberian plate and the European plate, and subsequent partial subduction of the Iberian plate underneath the European plate (Verges et al., 1995). This tectonic activity lasted from the Late Cretaceous through the Oligocene-Early Miocene (Verges et al., 1995, Teixell, 1998). After a time of tectonic quiescence at the end of the Oligocene deformation, exhumation of the orogen was reinitiated to its present relief, probably in the last 6 Myr (Fitzgerald et al., 1999). The Pyrenees are one of few young mountain ranges where there has been no granitic plutonism or migmatization (Choukroune et al., 1989). Because the tectonic activity ceased in the Early Miocene, the mountains are generally considered to be inactive.
Southern Pyrenean Thrust System
Generally, the Pyrenees are divided into two parts, referred to as the Southern Pyrenean thrust system, with vergence to the south, and the Northern Pyrenean thrust system, with vergence to the North (Teixell, 1998; Meigs et al., 1996; Verges et al., 1995). The Southern Pyrenean thrust system is mainly composed of post-Hercynian Triassic shale and gypsum, followed by a south-tapering prism of Upper Cretaceous carbonate rocks, shale, and sandstone (Teixell, 1998). Imbricate thrust sheets composed of basement and cover rocks, displaying a southern vergence, form the Southern Pyrenean Belt (Teixell 1998). There is a piggyback sequence of deformation manifested in the southern belt from the Late Cretaceous to the Early Miocene (Teixell, 1998). The Southern Pyrenean Belt also contains the Axial Zone of the Pyrenean orogen, manifested in the east by the antiformal stack of thrust sheets of basement rock (Teixell, 1998). Total shortening in the Southern Pyrenean Thrust system is estimated to be near 55 km (Teixell, 1998).
The Axial Zone of the orogen is composed of up to 10 km of Cambrian to Carboniferous rocks that underwent metamorphism during the Hercynian orogeny (Verges et al., 1995). These older rocks were deformed by Hercynian compression in the Late Carboniferous and then intruded by crust-derived granitic rocks during the Early Permian (Verges et al., 1995). Also present in the Axial Zone are ~ 2.5 km of post-Hercynian rocks of continental sediments and volcanic rocks dating from the Late Carboniferous to the Early Triassic, which unconformably overlie deformed Paleozoic rocks (Verges et al., 1995). The late deformation in the Axial Zone formed synchronously with thrusting in the southern fold and thrust belt (Fitzgerald et al., 1999). The thrust sheets of the Axial Zone are arranged in an antiformal stack (Teixell, 1998). In the west, these thrust sheets display less overlapping, and constitute crustal ramp anticlines, with the basal thrust faults forming low-angle ramps flattening to bedding-parallel décollements at the base of the cover, within the Triassic or Upper Cretaceous rocks (Teixell, 1998). The basal décollements root distinct upper level thrust systems, whose relationships with the synorogenic sediments indicate a piggyback sequence of deformation from the Eocene or possibly the Latest Cretaceous to Early Miocene times (Teixell, 1998).
Northern Pyrenean Thrust System
The Northern thrust system developed on the European plate and represents only about 20-25 km of shortening, much less than the southern system (Verges et al., 1995, Teixell, 1998). A thick Mesozoic succession in the Northern Pyrenean thrust system is comprised of Triassic shales and evaporites, Jurassic to mid Cretaceous (Aptian-Albian) carbonates, overlain by several thousand meters of Upper Cretaceous deep-water shales and turbidites (Teixell, 1998). The margin of the Cretaceous flysch basin is now observed near the Axial Zone in the Albian and Upper Cretaceous conglomerates and breccias that directly overlap the Paleozoic basement (Teixell, 1998). After Late Cretaceous extension, compressional regimes started in the Northern Pyrenean thrust system during the early Late Cretaceous (~85-80 Ma), and occupied progressively more and more external domains during the Early Tertiary (Choukroune, 1992). This thrusting imbricated basement rocks now exposed in the Axial Zone and some parts of the northern thrust system. These metamorphic rocks include metasediments, metavolcanics, and high-grade metamorphic and intrusive rocks associated with the Late Carboniferous to Early Permian Hercynian Orogeny (Verges et al 1995). Throughout the Pyrenees, where basement rocks are exposed, there is a low- to high-grade trend in the metamorphic sequences (Wickham, 1987).
The Northern Pyrenean Belt contains an imbricate system of north-directed thrust faults involving Hercynian basement rocks and Mesozoic to Lower Eocene cover rocks (Verges et al., 1995; Teixell, 1998). The Upper Cretaceous flysch north of the divergence zone that acts as a boundary between the northern and southern belts is deformed to the north in a fold and thrust system (Teixell, 1998). The two belts are mostly separated in the east by the major thrust fault of the orogen, known as the North Pyrenean Fault (NPF) (Teixell 1998). However, the North Pyrenean Fault dies out to the west, where the southern boundary of the North Pyrenean Belt is the axis of vergence between the two belts. The North Pyrenean Fault overrides Eocene sediment of the Aquitanian foreland in France (Teixell, 1998). The compressional deformation north of the divergence zone is Late Cretaceous to the Late Eocene or Early Oligocene (Teixell, 1998). There are two groups of ages obtained in the North Pyrnenan Thrust system, that of syntectonic metamorphism, at ~90 Ma (K/Ar) and that of Upper Eocene considered to characterize the Pyrenean orogenic event (Choukroune, 1989). The difference in timing between the Northern and Southern belts suggests that deformation in the southern belt continued after deformation ceased in the northern belt (Teixell, 1998; Fitzgerald et al., 1999; Yelland, 1990).
Ebro Foreland Basin (Spain)
The Ebro foreland basin consists of Paleocene red bed deposits, Lower and Middle Eocene platform carbonate deposition, and Middle Eocene evaporites, overlain by a thick succession of Middle and Upper Eocene marls that correspond to distal parts of alluvial fans and deltas related to the south-vergent thrust front in the Pyrenees (Verges et al., 1995). The marked change from marine deposition to widespread continental sedimentation in the Ebro basin occurred at ~37 Ma, during the Early Priabonian. The youngest sediments are late Early Oligocene (Verges et al., 1995). The Ebro Foreland basin closed due to Late Eocene to Oligocene tectonism along its margins, allowing synorogenic detritus, derived primarily from the Pyrenees, to fill the basin and continue to backfill across the actively deforming South Pyrenean Thrust Belt (Fitzgerald et al., 1999). This backfilling resulted in the burial of the southern Pyrenean range with 2-3 km of continentally derived conglomerates topped by a graded depositional surface which merged with high-level erosion surfaces into the Axial Zone (Fitzgerald et al., 1999).
In the Ebro Foreland Basin, there is a variation in thickness of the Middle and Upper Eocene marine strata below the originally horizontal Cardona salt level that shows evidence for a double-wedge basin geometry, likely related to the effect of the Pyrenean thrust load to the north and the added load of the Catalan Ranges to the south-east (Verges et al., 1995). Both basins show a marine to continental transition and are interpreted to have formed by flexure in response to loading. The Jaca basin, a small basin located in the southern thrust system, is comprised of several thousand meters of deformed Eocene to Lower Oligocene flysch and molasse that are inferred to represent the major phase of orogeny (Teixell 1998).
Aquitaine Foreland Basin (France)
The Aquitaine foreland basin is the result of flexure of the upper plate due to the loading of the North Pyrenean thrust sheets (Verges et al., 1995). During the Late Cretaceous, synorogenic sedimentary infilling began with turbidite deposition (Verges et al., 1995). Early Eocene shelfal carbonate deposits overlie Paleocene red beds and are overlain by several thousand meters of alluvial deposits (Verges et al., 1995). The oldest syntectonic flysch in the northern foredeep is Turonian (~91-88 Ma) and youngs progressively to the north (Desegaulx et al., 1990). Syn-tectonic sediments in the Aquitaine basin are affected by the North Pyrenean thrust system, but Eocene deposits are deformed only by the front of the Sub-Pyrenean thrust wedge (Verges et al., 1995).
Existing Fission Track Analysis in the Pyrenees
Apparent fission track ages for apatite samples from the Northern Pyrenean zone are between 48.5+9.5 Ma and 44.5+11.5 Ma, and are interpreted to represent rapid cooling during the Paleocene to Eocene (Yelland, 1990). Apparent fission track ages (apatite) for samples from the Southern Pyrenean zone are between 27.3+5.5 Ma and 18+6 Ma (Oligocene to Miocene)(Yelland, 1990). These apatite ages suggest that many of the fission tracks formed in the partial annealing zone (PAZ) during a protracted period of time (Yelland, 1990). Due to the long residence time in the PAZ, it is likely that tracks were shortened or perhaps lost entirely, reducing the apparent age. Verges et al.(1995) report ages from basement units of the Axial Zone to be decreasing from ~50 Ma in the north to ~17 Ma in the south, using fission track zircon/apatite dating.
Fitzgerald et al. (1999) obtained three vertical profiles of apatite FT ages, allowing them to use data from a thermal reference to determine the timing, amount, and rate of exhumation. The profile taken at the border between the Axial Zone and the North Pyrenean thrust system records the onset of rapid exhumation at ~50 Ma (Fitzgerald et al., 1999). The profile taken from near the center of the Axial Zone records relatively slow exhumation rates of ~173 m/Myr during ~44-36 Ma, followed by extremely rapid exhumation starting at ~35 Ma in the Early Oligocene, concentrated on the southern flank of the Axial Zone (Fitzgerald et al., 1999). At ~32-30 Ma, the southernmost profile records a dramatic slowing or even cessation of exhumation, interpreted to mark the end of major tectonic activity in the Central Pyrenees.
Track-length analysis reveals two distinct styles of cooling in the Pyrenees. The high mean track length and low standard deviation in apatites from the Northern Pyrenean zone imply shorter residence times in the partial annealing zone. Yelland (1990) suggests two possible explanations for this pattern. One is the possibility of rapid uplift from temperatures or structural depths at which there is zero track retention, and the other is the possibility of overprinting by a previous strong thermal event and subsequent rapid cooling to ambient temperatures. In contrast, the low mean track length and high standard deviation in apatites of the Southern Pyrenean zone indicate uniform cooling in the zone of zero track retention, through the partial annealing zone down to “stable” ambient temperatures (Yelland, 1990).
I. Hercynian Orogeny
The tectonic evolution of the Pyrenees begins with the Hercynian Orogeny at
~ 300 Ma (Carboniferous) (Yelland, 1990). During this time, the region of the Pyrenees was most likely the site of continental rifting associated with strike-slip deformation and widespread high temperature/low pressure metamorphism (Yelland, 1990; Wickham, 1986, 1987). Also accompanying the Hercynian Orogeny was generation of magma and emplacement of plutonic rocks (Yelland, 1990). The Hercynian metamorphism took place between 340-280 Ma, which would indicate that sedimentation and erosion occurred at the surface while metamorphism and plutonism occurred at depth (Wickham, 1986).
II. Opening of the Bay of Biscay
Following the Hercynian Orogeny, contraction in the Pyrenees is inferred to be associated with the opening of the Bay of Biscay and the rotation of the Iberian plate 15-40 degrees relative to the European plate (Yelland, 1990). The opening of the Bay of Biscay marks a time of crustal thinning along the southern margin of the European plate, and the formation of pull-apart basins, with synchronous high T-low P metamorphism (~90 Ma) (Desegaulx et al., 1990). The end of plate rotation which led to the opening of the Bay of Biscay corresponds to initial shortening within the fold belt, during the Upper Cretaceous metamorphism and deformation was restricted to the narrow North Pyrenean Zone. Rifting was apparently accompanied by strike-slip movement on the North Pyrenean Fault, synchronous with locally intense high temperature/low pressure metamorphism at 100+10 Ma (Yelland, 1990). During the Upper Cretaceous to lower Tertiary, northwest movement of the Iberian plate towards the more or less “stable” European plate marked a change from sub-vertical strike-slip tectonics to collisional, sub-horizontal thrust tectonics (Yelland, 1990). The Cretaceous cooling signal overprinted part of the Hercynian signal, and together the Hercynian-cooled and Cretaceous-cooled rocks comprise the orogenically “dead layer”, or cover layer, of the Pyrenees (Figure 3).
III. Pyrenean Orogeny
The major Pyrenean compressional phase occurred during Latest Cretaceous to the Eocene, with the resultant development of the southward-verging imbricate thrusts in the Southern Pyrenean thrust system and northward-verging imbricate thrusts in the Northern Pyrenean thrust system (Yelland, 1990). The observed Eocene-Oligocene exhumation patterns result from a change in the tectonic style of the Pyrenean double-wedge from inversion-dominated thrust stacking to crustal wedging and internal deformation (Fitzgerald et al., 1999).
There seems to be a consensus as to the general tectonic evolution of the Pyrenees. Thrusting in the Pyrenees resulted in significant crustal thickening and the formation of many of the structures now seen in the present-day Pyrenean Range. Differences in the thrusting patterns and ages between the Northern and Southern zones of the Pyrenees require a tectonic model capable of explaining such differences. From apatite fission track ages, Fitzgerald et al. (1999) documents the prevailing southward vergence in the Pyrenees, and attributes it to the asymmetric distribution of inherited intracrustal discontinuities above the northward subducting Iberian lower crust. He relates the rock uplift and formation of the Axial Zone antiformal stack to underthrusting, and suggests that rock uplift north of the North Pyrenean Fault is caused by wedging of the European crust (Fitzgerald et al., 1999). Many proposed models are a result of analysis of the ECORS deep seismic reflection profile taken across the Pyrenees in combination with a number of other studies. Yelland (1990) contends that cooling ages support northward verging A-type subduction of the Iberian plate beneath the European plate evidenced in the ECORS profile, and that the ECORS profile also shows evidence of “flake tectonics”, or evidence of the lower middle Iberian crust detached/decoupled by the rigid European plate indentor. In a similar context, Zoetemeijer et al. (1990) interpret the ECORS profile to show the partial subduction of the Iberian continental lithosphere underneath the European continent and image a drastic increase in crustal thickness. Based on the study of surface geology and the ECORS-Arzacq profile (in the western Pyrenees), Teixell (1998) favors a double or stacked wedge geometry as a structural model for the tectonic evolution of the Pyrenees. This model defines the Pyrenean orogenic prism to be bivergent in the upper crust, consisting of “a system of south directed thrust sheets accreted piggyback from the upper (Iberian) wedge, and a smaller, delaminating retrovergent belt”.
In the field, modern river sand from rivers draining either side of the orogen was collected, as well as bedrock samples. The sand was collected from the sand bars of the Ariege River in France, and the Segre River in Spain. The rivers were chosen because they symmetrically drain either side of the orogen and have headwaters that originate at the drainage divide. River sand was panned on sandbars which were chosen based on the accessibility of the sandbar and grainsize of the sand. The heavy minerals in the sand were concentrated by panning. The bedrock was collected on the northern (French) side only, in roughly 4 kg quantities from well-mapped bedrock deposited during the Sannoisian, Lutetian, Bartonian, and Oligocene.
Laboratory Separation Methods
The purpose of the lab separation is to segregate the zircon minerals from other sand grains in each sample. Because zircon is one of the most dense and least magnetic minerals, all separation steps relate to the density of the sand grains or the degree to which they are magnetic. Once in the lab, the bedrock samples were put through the rock crusher, one sample at a time. The crusher was thoroughly cleaned after each sample to avoid sample contamination, as was each piece of equipment after every step. Samples were then sieved with (–60, +200) mesh (a size split of 74-250 mm), and the portion of the sample that did not go through the mesh was pulverized. The samples were then run through the Rogers’ Table, using water to separate the sample into bins based on density. Each density group of the sample was then dried, and the densest portion of the sample was set aside for heavy liquid separation. The river samples, if very coarse sand, were sieved using (–60, +200) mesh, with the finest sand collected and set aside for heavy liquid separation. If medium to fine grained, the sand was set aside for heavy liquid separation without intervening steps.
From this point on each step was repeated for each sample, keeping each sample separate at all times and thoroughly cleaning each piece of apparatus between samples to avoid sample contamination. During all periods of exposure to the toxic heavy liquids, work was done under a hood, and latex gloves were worn to avoid contact with the skin. In a large flask, Tetrabromethane (TB) was added to each sample. The volume of sample in the flask was ~500 mL, for efficiency in the separation process. Enough TB was added to saturate the entire sample, with a small volume left sitting on top of the wet sample. The flask was sealed and the sample and TB were swirled together, so that all of the sample was floating in TB and none remained in the bottom or on the sides. The flask was then allowed to sit, and the denser heavy minerals fell to the bottom of the flask through the TB, while the less-dense minerals floated in the TB, allowing for relatively easy mineral separation. The heavy minerals were released through the bottom of the flask using a stopcock, into a funnel lined with a filter paper cone and taking care that none of the lighter minerals were also released. The process of swirling the sample with TB was repeated two or three times to ensure a complete separation, and all the heavy minerals were caught into the filter paper, where they were rinsed very thoroughly with acetone. The minerals were rinsed by washing them multiple times into several filter paper cones in succession, keeping the heavy minerals separate from the less dense, so that none of the TB remained anywhere near the minerals. The heavy minerals were then completely dried and set aside for magnetic separation.
The completely dried, dense mineral sample was then run through a Franz magnetic separator. The minerals were run through the separator multiple times, with an increase in the intensity of the magnet by ~0.1 to 0.5 amp increments to a final intensity of ~1.5 Amps. After each run, the magnetic minerals were run through the magnetic separator again, and the non-magnetic minerals were put aside, labeled non-magnetic, and saved. The magnetic minerals at the end of the final run through the magnetic separator were saved and labeled as magnetic, ready to go on to the next step of the separation process.
The final step of the mineral separation process is Methelyne Iodide (MI). The magnetic minerals were poured into a smaller flask, and MI was added. As with the TB, enough MI was added to saturate the sample, with a little extra volume on top. The process of MI is similar to that of TB. The sample and MI were swirled together after the flask had been sealed, and the dense minerals sunk to the bottom of the flask, where they were released using a stopcock. The lighter minerals floated, and the process was repeated at least twice to ensure a complete separation. The denser minerals were rinsed in the same manner as with the TB, using acetone and several filter-paper cones. The lighter minerals were thoroughly rinsed with acetone and saved, and the flasks were rinsed with acetone as well. As with TB, all work was done under a hood and latex gloves were worn so that contact with the skin was avoided.
The last few processes, prior to counting, involve preparing the samples for counting at the microscope. First, the samples were mounted in small Teflon squares at ~330° C, using two glass slides for mounting (Garver et al., 1999). At least two mounts were made of each sample. Each mount was cut with 600 grit sandpaper to expose the grains, and then hand-polished using a 9 mm diamond paste, and then a 1 mm diamond paste. Mounts were polished until most cracks and scratches were visibly reduced or disappeared as viewed under a microscope at ~500x.
The etching process takes advantage of this exposure of zircon surfaces to etch out, open up and make more visible, the fission tracks in each grain. Etching solution is a 7:5 NaOH:KOH eutectic and all etching is done in covered TeflonÒ dishes at exactly 228° C (+ 1) in a laboratory oven (Garver et al., 1999). The mounts were given a 1 hour pre-etch to rid the sample of any contaminants such as iron sulfides, as well as to force the disintegration of metamict grains (Garver et al., 1999). After this pre-etch, the remainder of the etch was done using a fresh etchant. Mounts for each sample were etched for an additional 9 or 23 hours, so that each sample had a 10 and a 24 hour total etch time. After all mounts had been etched, they were prepared for irradiation.
Mounts flattened by pressing them in a plate glass sandwich at 230° C for fifteen minutes were then cleaned in a dilute Hydrofluoric (HF) acid for 15 minutes. Mica flakes were affixed with tape to the top of each mount, and the samples were sent to the Oregon State Nuclear Reactor for a nominal fluence of 2 x 1015 neutrons/cm2. The samples were irradiated with age standards of Buluk Tuff and the Fish Canyon Tuff and CN-5 glass dosimeter at each end of the sample package.
Upon return from irradiation, the mica flakes were etched in 48% HF at room temperatures for twenty minutes. Teflon mounts and their mica pairs are both mounted onto one petrographic glass slide with nail polish so that that the mica is affixed as a mirror image of the mount. The slides are aligned on the microscope using an automated stage and zircon grains are marked for counting. Using low reflected light, bright grains (i.e. those with a well-polished surface), were marked for counting at 125x magnification. Under transmitted light, the marked grains were examined under 625x or 1250x to determine their countability. To obtain an unbiased data set, every effort was made to count every grain. However, this is nearly impossible due to single-grain imperfections. To select countable grains the following procedure was used: (1) The etch quality was examined first, and those grains that were over-etched, under-etched, or markedly unevenly etched (judged by the size of the tracks) were rejected. (2) The grains with well-etched tracks were further rejected based on: (a) the lack of a countable surface area; (b) The alignment of the grain on an axis other than the C-axis; (c) The presence of too many cracks and scratches on the grain; (d) The presence of too many inclusions.
Fission tracks were counted at 1250x dry magnification (100x dry objective, 1.25x tube factor, 10x ocular), and where possible 25 grains were counted per mount (2 mounts per sample). Some mounts had very few grains so in these cases, a minimum of at least 5 grains were counted per mount, yielding at least ten grains counted per sample. As discussed in the results section, the lack of countable grains reflects the fact that the zircons had considerable radiation damage and that they were rather old. Standards of Buluk and Fish Canyon Tuffs were treated the same as the unknown samples counted, and a total of six standards from two different irradiations were counted. Glasses from the top and bottom of the irradiation tube were counted to determine the fluence throughout the package. Fluence values were extrapolated for each sample position from bottom to top. In general the fluence variation from bottom to top is less than ~8%. The individual data for each mount were combined and run through analytical programs (Brandon, 1996). Probability plots of all the data were made using Sigma Plot.
Fission track data for each sample were run through a peak-fitting program to determine the ages of the grains in that sample. An optimal solution was found for each sample (Brandon, 1996) (Table 1). Sand from the Ariege River was collected ~7 km north of the town of Foix (N 43°01.767, E 01°36.410). The sand was collected from gravel bars with minor sand components. The pebbles are dominated by plutonic rocks and contain foliated plutonic clasts (Granitics), schist (mica), and micaceous sandstone. From this sample, four peaks were obtained from the 50 dated grains. The first peak, containing 7.9% of the grains, is 30.6 (-4.8/+5.8) Ma. The second, at 49.0 (-4.7/+5.1) Ma contained 20% of the grains. The third peak has an age of 84.6 (–5.6/+6.0) Ma and contained 62% of the grains, and the fourth peak contained 10% of the grains and showed an age of 174.9 (-34.1/+42.2) Ma.
Unit and Number Binomial Peak-Fit ages
Ariege River, France 30.6(-4.8/+5.8) 49.0(-4.7/+5.1) 84.6(-5.6/+6.0) 174.9(-34.1/+42.2)
99-26 50 Nf = 7.9% Nf = 20% Nf = 62% Nf = 10%
W = 30% W = 22% W = 25% W = 32%
Segre River, Spain 27.9(-2.6/+2.8) 43.9(-3.0/+3.2) 94.6(-11.1/+12.6) 209.8(-18.8/+20.6)
99-31 50 Nf = 24.8% Nf = 42.9% Nf = 10.8% Nf = 21.6%
W = 25% W = 23% W = 24% W = 32%
G2 Eocene 62.4(-6.2/+6.9) 139.4(-14.3/+15.9) 220.7(-21.7/+24.0)
99-30 41 Nf = 19.9% Nf = 39.4% Nf = 40.7%
(~52-50 Ma) W = 28% W = 28% W = 30%
E6 Bartonian 118.6(-10.3/+11.3) 290.4(-27.8/+30.7) n.d n.d.
99-27 20 Nf = 37.9% Nf = 62.1%
(42.1-38.6 Ma) W = 23% W = 33%
E7 Ludian 29 78.7(-4.4/+4.6) 190.4(-17.5/+19.2) n.d. n.d.
99-28 Nf = 71% Nf = 28.5%
(~38-35 Ma) W = 20% W = 24%
G1 Sannoisian 95.7(-21.6/+27.9) 174.1(-13.5/+14.6) n.d. n.d.
99-29 16 Nf = 13.9% Nf = 86.1%
(~33-30 Ma) W = 33% W = 27%
Ariege River, France 30.6(-4.8/+5.8) 49.0(-4.7/+5.1) 84.6(-5.6/+6.0) 174.9(-34.1/+42.2)
99-26, 50 Nf = 7.9% Nf = 20% Nf = 62% Nf = 10%
(0 Ma) W = 30% W = 22% W = 25% W = 32%
Note: Nt = total number of grains counted; Nf = calculated number of grains in a specific peak or fraction; W = estimated relative standard deviation for a peak, expressed in percentage of peak age; uncertainties cited at ±1 se. Zircons were dated using standard methods for FT dating using an external detector. Zircons were extracted using standard separation procedures (Naesar, 1979). All samples were crushed, pulverized, and then passed over a GeminiÒ or Wilfly table, then processed in heavy liquids and a magnetic seperator. Zircon grains were mounted in 2 x 2 cm2 squares of PFA TeflonTM. During polishing, each mount was first cut with 800 grit wet sandpaper, and then polished successively on 1 mm, 9 mm diamond paste, and then finished using a 0.3 mm Al2O3 paste. Mounts were etched in a eutectic NaOH-KOH mixture at 228°C for 24-36 hours. Etch times were for A1 to A7 varied (15 to 32 hr) due to poor etching efficiency, probably due to disintegration of old metamict grains which subsequently affected the quality of the chemical etchant. Etch times were for T1 to T11, and LU1 to LU9 were 15 and 27 hr and 15 and 30 hr respectively which includes a 1 hour pre-etch. After etching, mounts were covered with a low-uranium mica detector, and irradiated with thermal neutrons at Oregon State University with a nominal fluence of 2 x 1015 n/cm2, along with zircon age standards (Buluk Tuff and Fish Canyon Tuff) and a reference glass dosimeter CN-5. For A1-A7, samples were counted at 1250x in oil (100x objective, 1.0 tube factor, 12.5 oculars) using a zeta (CN-5) of 365.84±12.03 (MEB). For T1-T11, samples were counted at 1250x (100x objective, 1.0 tube factor, 12.5 oculars) using a zeta (CN-1) factor of 133.58±1.1 (GX). For LU1-LU9 samples were counted at 1652x dry (100x objective, 1.25 tube factor, 12.5 oculars) using a zeta (CN-5) of 305.01±6.91.
The sand from the Segre River in Spain was collected at Organya and the crossover to Figols d’ Organya (N 42°12.808, E 01°20.406). The sand is a dark lithic sand with abundant white mica. Pebbles in the river are dominated by plutonic rocks, limestone, arkosic sandstone, basalt/greenstone, biotite schist, listwanite, and quartz-pebble conglomerate clasts. FT ages of detrital zircon from this sample, also of 50 grains, are fit by four peaks. The first peak contains 12.4% of the grains and occurs at 27.9 (-2.6/+2.8) Ma. The second peak contains 21.5% of the grains and occurs at 43.9 (-3.0/+3.2) Ma. The third and fourth peaks occur at 94.6 (-11.1/+12.6) Ma and 209.8 (-18.8/+20.6) Ma and contain 5.4% and 10.8% of the grains, respectively.
The FT data collected from zircon grains from sandstones of the stratigraphy on the north side of the Pyrenees (France) is sparse because the yield was low and concentration in the field could not be done. The Lower Lutetian (Lower Eocene) sandstone sample was collected at Menay, France (N 43°04.881, E 01°26.289), from a roadside outcrop of thick bold sandstone and conglomerate. The coarse-grained sandstone is grey-white weathering, friable, and is dominated by quartz but also contains lithics and white mica, with few scattered pebbles. The three peaks of the Lower Lutetian data, comprised of 41 grains, occurr at 62.4 (-6.2/+6.9) Ma, 139.4 (-14.3/+15.9) Ma, and 220.7 (-21.7/+24.0) Ma, and contain 19.9%, 39.4%, and 40.7% of the grains, respectively.
The Bartonian (Middle-Upper Eocene) sandstone sample was collected from a roadcut south-east of the town of Crampanga, France (N 43°01.473, E 01°36.182). The bedrock was comprised of a massive clast supported pebble to cobble conglomerate with minor interbedded quartz-rich sandstone. The clasts in the conglomerate are dominated by sedimentary rock fragments (60% gray limestone and 40% buff/yellow sandstone), and the sand is very well cemented with abundant rust-weathered clasts. Ages of FT grain-age populations are at 118.6 (-10.3/+11.3) Ma, with 37.9% of the 20 grains, and 290.4 (-27.8/+30.7) Ma, containing 62.1% of the 20 grains.
The Ludian (Upper Eocene) sandstone samples were taken from a roadcut west of the town of Crampagne, France (N 43°01.910, E 01°36.208). The rock is comprised of a cobble-pebble conglomerate with quartz and mica-rich interbedded sandstone. The clasts are rich in sandstone, limestone, schist, and plutonics (red porphyry, yellow mica sandstone, intermediate plutonics, Nummulitic limestone, crystalline limestone, biotite schist, chloritic schist, fine-grained sandstone, medium-grained arkose). The FT grain ages, (consisting of 29 grains), have peaks at 78.7 (-4.4/+4.6) Ma, containing 71% of the grains, and 190.4 (-17.5/+19.2) Ma, containing 28.5% of the grains.
The Sannoisian (Lower Oligocene; Rupelian) sandstone samples were taken from an outcrop north-west of Calzen, France (N 43°02.563, E 01°44.442). The roadcut was composed of a pebble conglomerate layer overlying interbedded siltstone and thin-bedded sandstone. The sandstone sampled is a coarse-grained sandstone, though much of the sandstone was fine-grained. The FT grain age data, (comprised of only 16 counted grains), yield peaks at 95.7 (-21.6/+27.9) Ma (13.9%) and at 174.1 (-13.5/+14.6) Ma (86.1%).
The Ariege River in France shows its most significant peak at 84.6 Ma, with lesser peaks at 49.0 Ma and 30.6 Ma, and one small older peak at 174.9 Ma (Figure 4). These data suggest that the majority of detrital zircons in the river receives today is 84.6 Myr old. The peaks at ~49 Ma and ~31 Ma are within the window of orogenesis and therefore represent rock cooled below zircon closure during the collision. The peak at 84.6 Ma may represent early Pyrenean activity, and the peak at 174.9 Ma clearly represents the dead layer being stripped off the orogen, cooled well before the Pyrenean collision.
The Segre River in Spain shows its most siginificant peak at 43.9 Ma, much younger than that of the Ariege. The Segre has older peaks, one at 94.6 Ma and a larger peak at 209.8 Ma. The Segre also shows a young peak at 27.9 Ma (Figure 5). These data suggest that the Segre’s main pulse of sediment is ~ 44 Myr old, which is synorogenic with the Pyrenean Orogeny, as is the young peak at 27.9 Ma. The peaks at 94.6 Ma and at 209.8 Ma are representative of grains cooled before the Pyrenean Orogeny.
Stratigraphy on the North Flanks of the Pyrenees (France)
The four samples of north-flanking stratigraphy show varying peak ages that change upsection. The Lower-Middle Eocene (Lutetian) strata, deposited during ~52-50 Ma, displays three major peaks in grain ages. The youngest and smallest peak occurs at 62.4 Ma (Paleocene) and probably represents early initiation of Pyrenean activity. The other two peaks, of about equal significance, occur at 139.4 Ma (Early Cretaceous) and 220.7 Ma (Late Triassic) (Figure 6), and represent cooling ages prior to orogenesis.
The Bartonian, deposited at 42.1-38.6 Ma (Middle-Late Eocene) shows two significant peaks of nearly equal size, one at 118.6 Ma (Early Cretaceous) and one at 290.4 Ma (Late Carboniferous) (Figure 7), but these data are rather limited due to poorly etched grains.
The Ludian strata (~38-35 Ma) shows two major peaks, the greater of those at 78.7 Ma (Late Cretaceous) and a smaller peak at 190.4 Ma (Early Jurassic) (Figure 8). The large peak at 78.7 Ma indicates the early phases of Pyrenean activity, represented by the flood of basement lithologies present in the bedding. The smaller peak at 190.4 Ma is representative of rocks cooled prior to the Pyrenean event.
The Sannoisian ( ~33-30 Ma-Early Oligocene) shows a major peak at 174.1 Ma (Middle-Late Jurassic) and a smaller peak at 95.7 Ma (Late Cretaceous) (Figure 9), but only 20 grains were datable from the sample, therefore the peak ages are uncertain.
The majority of sediment that the modern Ariege (0 Ma) receives today is 84.6 Myr old. The river also still receives sediment at least as old as ~175 Ma, and receives sediment in smaller pulses from ~49 Ma and ~31 Ma.
Pyrenean Orogeny-Main Phase (~50-30 Ma)
Today, the drainage divide in the Pyrenees runs along the spine of the range, and it is for this reason that rivers running off each side of the orogen were sampled. The zircons from these rivers provide an indication of the grain-ages delivered to the foreland and hinterland basins at the present. The grain ages of the zircons shed off the north side of the orogen, collected from the Ariege River in France, show that about 28% are either 49 Ma or 31 Ma (main-phase Pyrenean) (Figure 10). The grain ages of the zircons collected from the Segre River in Spain, draining off the south side of the range, show that about 66% are from similar ages, either 44 Ma or 23 Ma. These young age signals in the modern record from both sides of the orogen suggest that rock cooled during the main phase of the Pyrenean event are being exhumed today, with minimum lag times of ~20 Ma and rates of exhumation of ~300 m/Myr (see Figure 11 from Garver et al., 1999).
Therefore, the ages shed off either side of the orogen are similar, but a greater percentage of syn-orogenic cooling ages are shed to the Iberian foreland as opposed to the French hinterland, indicating a slightly asymmetric exhumation pattern. There is almost no signal of this true Pyrenean age in any of the older strata sampled from the northern flanks of the orogen. Though the structure and exhumation patterns of the orogen are unknown for the past, the present-day asymmetry shown by these peaks is in agreement with the asymmetry suggested by the apatite bedrock data (Yelland, 1990; Fitzgerald et al., 1999). It is likely that these ages reflect the development of the orogen, especially the beginning and termination of shortening.
Pyrenean Orogeny-Early Phase (~90-50 Ma)
The earliest syntectonic flysch exposed in the north is inferred to represent the onset of a compressional regime. The timing is controversial, having commenced at ~91-88 Ma (Desegaulx et al., 1990), 88-66 Ma (Choukroune, 1992), or in the latest Cretaceous (i.e. >65 Ma, see Teixell 1998; Choukroune 1992).
In the modern Ariege River sample, there is a peak population of ~85 Ma. In the north-flanking bedrock sampled, the lower Lutetian (Lower-Middle Eocene) sample displays at peak at ~62 Ma, the Bartonian strata contains no peak of this age, the Ludian (Upper Eocene) strata displays a peak at ~79 Ma, and the Sannoisian strata also displays no peak. The lack of this peak age may be of little significance for the Bartonian and Sannoisian strata, due to the lack of datable grains in each sample. These ages are all within the initiation ages suggested above and fall within the early phase of the Pyrenean event.
Taken together with the Ariege River data, a trend of backwards moving peaks is seen upsection from the lower Lutetian to the modern sediments, with peak ages moving upsection from ~62 Ma (lower Lutetian), ~79 Ma (Ludian), and ~85 Ma (Modern) (Figure 12). There are a couple of possible explanations for this phenomenon. One is that the signal being picked up in this population is derived from grains previously deposited in a basin (Garver and Brandon, 1994; Garver et al., 1999). Another possible explanation is that the signal is due to the downward erosion of a tilted crustal section. Back-thrusting could have caused imbrication of thrust sheets, which might produce the phenomenon of a backwards-moving peak. It seems that the northern side of the orogen has experienced erosional “burrowing” down through the crustal imbricates since the inception of movement.
One very likely explanation for the trend in the peak ages lies in varying rates of exhumation. The ~62 Ma signal present in the Lutetian strata suggests that rock with the cooling signal of the initiation phases of the Pyrenean event was being exposed ~10-12 Ma later, indicating a short lag time and high rates of exhumation (~800 m/Myr). The ~79 Ma signal present in the Ludian strata represents a lag time of ~40 Ma, indicating slower rates of exhumation (~ 200 m/Myr), in agreement with the interpretation of Fitzgerald et al. (1999).
Rifting Associated with the Opening of the Bay of Biscay
The opening of the Bay of Biscay is generally associated with middle-Late Cretaceous rifting (~100 Ma) and the formation of pull-apart basins (Desegaulx et al., 1990). The signal for this time frame is seen in the modern river samples only in the Segre River, at a peak age of ~95 Ma, but is absent from the Ariege River signals. In the north-flanking samples, there appears to be a progressively younging trend upsection of this peak age. The lower Lutetian has a peak at ~139 Ma, the Bartonian contains a peak at ~118 Ma, there is no signal in the Ludian, and the Sannoisian shows a peak at ~96 Ma. The younging upsection reflects a normal exhumation trend of erosion down into the sample section, progressively stripping away material.
The presence of two opposite trends in the peak ages between the early phase of the Pyrenean Orogeny and the rifting associated with the Bay of Biscay is complicated. The data are sparse and therefore in some cases the peak ages are poorly approximated. The migrating trend in peak ages could possibly reflect changes in the development of the basin, such as a change in the source area from which sediment is derived. More likely, the different trend in peak ages is a reflection in the different rates of exhumation through time.
The ages for the Hercynian Orogeny are ~340-280 Ma, with clear widespread cooling at ~290 Ma and younger (Wickham and Oxburgh, 1987). The modern river sediments both contain clear Hercynian cooling signals, the Ariege at ~175 Ma and the Segre at ~210 Ma. The north-flanking strata sampled also all contain a Hercynian signal. The lower Lutetian shows a signal at ~221 Ma, the Bartonian at ~290 Ma, the Ludian at ~190 Ma, the Sannoisian at ~174 Ma, and the modern at ~175 Ma (Ariege). There seems to be no trend to these data on the north-flanking side, suggesting derivation from a number of crustal blocks that record this slow post-orogenic cooling.
Unroofing the Orogen
The cooling ages reflected in the zircons were set at a depth of around ~7 km, indicating that at least ~7 km of sediment had to be removed before the main pulse of tectonic activity recorded in the cooling ages could have occurred in sediments. This main phase of tectonism, in the case of the Pyrenees, occurred between ~50 and ~30 Ma. The stratigraphic samples represent ~20 Ma to exhume 7000 m of rock. The pulse of cooling ages reflecting the main stages of Pyrenean tectonism does not show up in any of the north-flanking stratigraphic age populations, indicating that rates of exhumation must have been less than ~350 m/Myr (Garver et al., 1999). However, signals of early-phase Pyrenean activity do appear in these age populations. The “lag time” of this signal ranges from ~10-12 Ma in the Lutetian to ~ 40 Ma in the Ludian, suggesting decreasing exhumation rates. The north-flanking stratigraphic ages also contain signals of Hercynian and Cretaceous cooling, showing the stripping off of the cover layer so that rocks of a Pyrenean signal can be exposed (Figure 13).
The modern river samples contain young peaks that reflect the main phase of orogenesis in the Pyrenees, 44 Ma and 23 Ma in the south and 49 Ma and 31 Ma in the north. These ages indicate the modern day removal of rocks carrying a Pyrenean cooling signal, perhaps supporting Fitzgerald’s contention that there has been post-orogenic reactivation of exhumation recently (Fitzgerald et al., 1999).
Through the use of detrital zircon fission track studies in the Pyrenees, grain-age populations have been interpreted to define the exhumation history of the area. The presence of rocks cooled during the post-Hercynian cooling event and Cretaceous rifting in the north-flanking stratigraphic samples and in the modern rivers indicates that during the ~20 Ma of main-phase tectonism, the cover layer to the system was not fully removed. The presence of Pyrenean cooling ages in the modern rivers suggests the possibility that exhumation of the orogen is still taking place. The greater percentage of Pyrenean ages being shed to the south of the orogen in comparison with those being shed to the north suggests a slightly asymmetric nature to the orogen and deeper exhumation on the southern foreland side. Lag times suggest variation in the rate of exhumation over time, from initial rates on the order of ~800 m/Myr to rates of ~200 m/Myr during the main phase, and ~300 m/Myr today.
While these data are useful and much information has been obtained, there are many things that could be focused on in future studies. The most important of these is to obtain samples from the south-flanking stratigraphy. Exhumation rates in the north could then be compared with those in the south to determine a more complete exhumation history of the whole orogen. Etch times should be adjusted to minimize grain damage incurred from the etching process, which would consequently increase the volume of countable grains. Ideally, a minimum of 50 grains per sample should be counted, and so a greater volume of stratigraphic samples should be obtained to increase the amount of zircons in each mount, which would also increase the volume of countable grains.
Burbank, D.W.; Verges, J.; Muñoz, J.A.; Bentham, P. Coeval hindward-and forward- imbricating thrusting in the south-central Pyrenees, Spain: timing and rates of shortening and deposition. Geological Society of America Bulletin, v. 104, n 1, p 3-17,1992.
Choukroune, P. Tectonic Evolution of the Pyrenees. Annual Review of Earth and planetary sciences, v 20, p 143-158, 1992.
Choukroune, P. et al. The ECORS Pyrenean deep seismic profile reflection data and the overall structure of an orogenic belt. Tectonics, v 8, n 1, p 23-39, 1989.
Desegaulx, P.; Roure, F; Villein, A. Structural evolution of the Pyrenees: tectonic inheritance and flexural behavior in the continental crust. Tectonophysics, v 182, n 3-4, p 211-225, 1990.
Fitzgerald, P.G. et al., Asymmetric exhumation across the Pyrenean orogen: implications for the tectonic evolution of a collisional orogen. Earth and Planetary Science Letters, v 173, p 157-170, 1999.
Garver, John I. et al. Detrital zircon fission-track thermochronology: practical considerations and examples. To be published in Memorie di Scienze Geologiche, v 51, 1999.
Garver, John I. et al. Exhumation history oforogenic highlnds determined by detrital fission-track thermochronology. In Ring et al. (eds). Exhumation Processes: Normal Faulting, Ductile Flow and Erosion. Geological Society, London, Special Publications, v 154, p 283-304, 1999.
Meigs. A. J.; Verges, J.; Burbank, D.W. Ten-million-year history of a thrust sheet. GSA Bulletin, v 108, n 12, p 1608-1625, 1996.
Ring et al. Exhumation processes. In Ring et al. (eds). Exhumation Processes: Normal Faulting, Ductile Flow and Erosion. Geological Society, London, Special Publications, v 154, p 1-27, 1999.
Teixell, Antonio. Crustal structure and orogenic material budget in the west central Pyrenees. Tectonics, v 17, n 3, p 395-406, 1998.
Verges et al. Eastern Pyrenees and realted freland basins: pre-, syn-, and post-collisional crustal-scale cross-sections. Marine and Petroleum Geology, v 12, n 8, p 903-915, 1995.
Wickham, Stephen M.; Oxburgh, E. Ronald. A rifted tectonic setting for Hercynian high-thermal gradient metamorphism in the Pyrenees. Tectonophysics, v 129, n 1-4, p 53-69, 1996.
Wickham, Stephen M.; Oxburgh, E. R. Low-pressure regional metamorphism in the Pyrenees and its implications for the thermal evolution of rifted continental crust. Philosophical Transactions-Royal Society of London, Series A, v 321, n 1557, p 219-242, 1987.
Yelland, A.J. Fission track thermotectonics in the Pyrenean orogen. Nuclear Tracks and Radiation Measures, v 17, n 3, p 293-299, 1990.
Back to Union Fission Track Home page
Back to Home page for J.I. Garver
Back to Union Geology Department
This document can be located from http://idol.union.edu/~garverj/FT/FThome.html© Geology Department, Union College, Schenectady N.Y. 12308-3107.
All rights reserved. No part of the document can be copied and/or redistributed, electronically or otherwise, without written permission from: J.I.Garver, Geology Department, Union College, Schenectady NY, 12308-2311, USA.
Last Revised: 26 January 2003
Some of this material is based upon work supported by the US National Science Foundation under Grant No. 9614730. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.