0
Δ
Ray Path
ΔΔ Δ Δ Δ ΔΔ
Δ
Δ
Δ
Δ
Δ
Δ
Depth (km)
20
40
60
80
100
−150
−100
−50
0
Distance (km)
50
100
150
CSZ
CSZ
BS06
BS04
BS09
BS07
BS10
BS11
BS13 BS15
BS12
BS14
BS17
BS16
BS19
BS18
BS22
0
20
depth (km)
40
60
80
100
CSZ
50
100
150
y offset (km)
200
Evt.60.baz.148.del.39.F.0.05.0.4
BN23
0.8
BN22
0.6
BN21
0.4
BN20
0.2
0
BN19
0
20
Time (s)
40
60
BN17
BN15
4
BN14
2
BN13
0
BN09
−2
−4
BN18
BN16
6
Median Trace
Shifted traces
Station
Median Envelope
1
BN04
0
20
Time (s)
40
60
20
30
40
Time (s)
Evt.111.baz.345.del.87.F.0.05.0.4
BN23
55.4
55.2
BN20
BN19
55
Latitude
55.6
BN22
BN21
54.8
54.6
54.4
54.2
residual (s)
54
53.8
53.6
−129.5
55.2
0.3
0.2
0.1
0
−0.1
BN18
BN17
BN16
BN15
BN14
BN13
BN23
55.4
BN20
BN19
55
Latitude
55.6
BN22
BN21
0.04
BN18
BN17
BN16
BN15
BN14
BN13
54.8
54.6
54.4
0.02
0
−0.02
−0.04
errors (s)
54.2
54
53.8
BN04
−129
−128.5
Longitude
−128
−127.5
53.6
−129.5
BN04
−129
−128.5
Longitude
−128
−127.5
Evt.119.baz.22.del.47.F.0.05.0.4
55.2
BN20
BN19
55
BN22
BN21
54.6
54.4
0.2
0
54.2
54
53.8
53.6
−129.5
55.2
BN20
BN19
55
BN18
BN17
BN16
BN15
BN14
BN13
54.8
BN23
55.4
54.6
54.4
residual (s)
54.2
0.04
0.02
0
−0.02
−0.04
errors (s)
BN18
BN17
BN16
BN15
BN14
BN13
54.8
−0.2
BN22
BN21
54
53.8
BN04
−129
−128.5
Longitude
−128
−127.5
53.6
−129.5
BN04
−129
−128.5
Longitude
−128
−127.5
errors (s)
BN23
55.4
Latitude
55.6
Latitude
55.6
Evt.89.baz.221.del.92.F.0.05.0.4
BN23
55.4
55.2
BN20
BN19
55
Latitude
55.6
BN22
BN21
54.6
54.4
0.2
0
−0.2
residual (s)
54.2
54
53.8
53.6
−129.5
55.2
BN20
BN19
55
BN18
BN17
BN16
BN15
BN14
BN13
54.8
BN23
55.4
Latitude
55.6
BN22
BN21
BN18
BN17
BN16
BN15
BN14
BN13
54.8
54.6
54.4
0.05
0
−0.05
54.2
errors (s)
54
53.8
BN04
−129
−128.5
Longitude
−128
−127.5
53.6
−129.5
BN04
−129
−128.5
Longitude
−128
−127.5
Evt.60.baz.148.del.39.F.0.05.0.4
55.2
BN20
BN19
Latitude
55
54.6
54.4
54.2
0.1
0
54.6
−0.1
54.4
residual (s)
54.2
−128
−127.5
53.6
−129.5
BN22
BN21
0.02
BN18
BN17
BN16
BN15
BN14
BN13
54.8
53.8
−128.5
Longitude
BN20
BN19
0
−0.02
errors (s)
BN09
54
BN04
−129
55.2
55
BN09
54
53.6
−129.5
BN22
BN21
BN18
BN17
BN16
BN15
BN14
BN13
54.8
BN23
55.4
BN04
−129
−128.5
Longitude
−128
−127.5
errors (s)
BN23
55.4
53.8
55.6
Latitude
55.6
Subduction zone structure in Cascadia
data, seismicity and tomographic velocities. To complicate matters
further, Preston et al. (2003) proposed that the shallow-most
(<45 km) Benioff events in Cascadia, such as those beneath western Vancouver Island, may actually lie within the Juan de Fuca plate
mantle (the result of serpentine dehydration) rather than within the
subducted crust, as previously assumed. Consequently, seismicity
constraints on the position of the plate may need to be reconsidered.
The utility of seismicity in precisely locating the plate is also compromised by location errors, although approaches such as double differencing (Waldhauser & Ellsworth 2000), arrival pattern location
(Nicholson et al. 2002) and 3-D empirical traveltimes (Nicholson
et al. 2004) have recently improved location accuracy. Resolution
of the debate concerning the position of the subducted crust is of
central importance in resolving important geodynamic questions
(Rogers & Dragert 2003).
Despite uncertainty surrounding the location of the downgoing
plate, much has been learned about the geometry and character of
the E-reflectors (Clowes & Hyndman 2002). The reflectors form a
thin (<2 km) zone off the coast of central Vancouver Island (Calvert
et al. 1996; Nedimovic et al. 2003) but thicken to 12–15 km off the
coast of southern Vancouver Island (Calvert et al. 2003) and appear
to be 5 to 8 km thick inland of the western coast of Vancouver
Island (Green et al. 1986; Nedimovic et al. 2003). It is unclear at
what depth the E-reflection zone terminates. Nedimovic et al. (2003)
interpret the E-reflectors to end abruptly where the bottom of the
zone encounters the forearc mantle, at a depth of 33 ± 3 km. Calvert
et al. (2003), on the other hand, demonstrate that the E-reflectors
persist to a depth of at least 50 km. Part of this disagreement could
be explained by differences in transect position.
Kurtz et al. (1986) used magnetotelluric data to show that the
E-layer is highly conductive across most of Vancouver Island. The
E-layer and C-layer (a second, weaker, shallower zone) are also
thought to possess low density (e.g. Clowes et al. 1997) and low
P-wave velocity (e.g. Drew & Clowes 1990), although it should be
noted that recent tomographic models do not clearly demonstrate
this latter property (Ramachandran 2001; Preston et al. 2003). Receiver function studies have shown that the C-layer is less prominent
than the underlying E-layer and that both are low-S-velocity zones.
There have been several explanations proposed to explain the high
reflectivity and low S-wave velocity of the E-layer as well as the observation that it extends to mantle depths. These include interpreting
the E-layer as:
(1) the top of the subducting Juan de Fuca plate (Green et al.
1986), although this interpretation is now largely discounted;
(2) interlayered mafic and/or sedimentary rocks (Yorath et al.
1985; Clowes et al. 1987);
(3) intensely sheared sediments that trap fluids released from the
subducting plate (Calvert & Clowes 1990);
(4) thin dipping fluid-filled lenses of high porosity, with fluid
supplied by dehydration reactions within the underlying oceanic
plate (Hyndman 1988).
It has also been proposed (e.g. Nedimovic et al. 2003; Calvert
2004) that the E-reflectors may represent a zone of ongoing deformation associated with the slow slip events observed along the
Cascadia subduction zone (Rogers & Dragert 2003) with a welldefined 14-month periodicity.
3 D AT A A N D P R O C E S S I N G
In the period 2002 April–2002 November, 26 broad-band seismometers were deployed by the POLARIS consortium across the northern
C
2005 RAS, GJI, 161, 849–859
851
Figure 2. Distribution of broad-band, three-component POLARIS-BC stations over northern Cascadia (red and white triangles). Red triangles indicate
stations used in Section 4.1. The dashed line shows the 2-D profile used in
Section 5, while the dotted region extending from PFB in the west to DURW
in the east shows the region of seismicity used in Fig. 8.
Figure 3. Distribution of events with sufficiently high signal-to-noise ratio. The region west of the array is well covered, along with central and
southern America. Note that this distribution is reasonably good for a global
teleseismic study.
Cascadia subduction zone to supplement five permanent broad-band
stations operated by the Geological Survey of Canada (see Fig. 2).
Together these 31 stations form the POLARIS-BC array—an approximately linear array spanning southern Vancouver Island, the Strait
of Georgia, coastal Washington and the British Columbia Fraser
Valley. The array operated until 2004 December.
As of 2004 January, 71 teleseismic events of magnitude 6.0 or
greater and with high signal-to-noise ratio had been recorded by the
POLARIS-BC stations (see Fig. 3). Azimuthal coverage is relatively
good since almost the entire Pacific, central and South America
lie within the range of teleseismic P (30◦ –100◦ ). The azimuthal
coverage is poor only to the south and northeast of the study region.
Subduction zone structure in Cascadia
853
Figure 5. (a) Source-normalized scattered S waves for all the data used in the multistation neighbourhood inversion. (b) Synthetic scattered S waves generated
using the best-fit model (see Table 1). Stations are ordered by position along the profile, with western stations first. Note that the main features of the data are
well reproduced by the synthetics. Also note that the correlated signal fades to the east of SSI.
Table 1. Model parameters for best-fit plate model across Vancouver Island.
Parameters for layers 1–3, 5 and 7 are held fixed in the inversion, as is the P
velocity and density for all layers.
Layer Thickness (km) V P (m s−1 ) V S (m s−1 ) ρ (kg m−3 ) Strike Dip
1
2
3
4
5
6
7
8
9
1.0
5.0
1.0
4.0
11.0
4.5
3.0
5.0
90.0
5000
6000
6700
6350
7100
6350
7750
6800
7750
2890
3460
3870
3790
4100
2990
4470
4280
4470
2800
2800
2930
2800
3140
2800
3030
2880
3200
0
0
0
0
0
0
−48
9
0
0
−48 16.5
0
0
−48 17
0
0
region (Calvert et al. 1996; Nedimovic et al. 2003). Owing to the
use of data from an array of stations, the strike (N48◦ E) is likely
to be better constrained than that determined from previous singlestation receiver function analysis. The S-wave velocity contrasts, of
C
2005 RAS, GJI, 161, 849–859
1.1 km s−1 at the top and 1.45 km s−1 at the bottom of the E-layer,
are consistent with the results of Frederiksen et al. (2003), although
they are larger than those proposed by Cassidy (1995).
The other two low-velocity layers (C and F) are significantly
weaker than the E-layer. In particular, we note that the F-layer arises
from the presence of slightly larger-amplitude signals within the
E-layer P-s-s window than expected (see Fig. 5), implying interference with an F-layer P-p-s phase. There is little or no visual indication of an F-layer P-s-s phase arriving at later times, and removing
the layer entirely only changes the fit to the data by a small amount
(less than 1 per cent), consistent with the results of Frederiksen et al.
(2003).
4.2 Single-station inversions
Inverting for the structure beneath individual stations allows the
depth and material properties of layering to be determined locally near each station. Thus variations in V S and deviations from
Friday Feb. 24 meeting agenda (10 minute presentations)
1. Let’s decide Friday and Sunday morning: a) student projects b)
data sharing formats c) publications (active source time offset).
2. Ken: Neighborhood mapping of S-SKS posteriori probability density
functions for 5 simplest anisotropy models from Billings array data
3. John: Grid search of Lodore array Pds signals to constrain velocity
steps amplitude and sharpness
4. Steve: construction of minimum phase spectrally shaped Greens’
functions from teleseismic P coda.
5. Josh: Imaging of Yellowstone mid-crustal sill using seism, doubleseism, and infra-seismic ‘hum’ sources.
6. Arizona 1
7. Arizona 2
8. Arizona 3
Batholith natural source research sub-projects
1. Migration of p-sv-sh green fcts for image
2. SKS & S shear wave anisotropy (standard processing)
3. S and SKS shear wave anisotropy using NA PPD mapping and 5model testing.
4. P and S-wave teleseismic body wave tomography.
5. Modeling p-sv-sh azimuthal response for crustal anisotropy/dipping
boundaries.
6. Surface waves:
a) ‘In-line’ mostly broadside ballistic surface wave imaging.
b) Surfaces waves imaging between the two lines
c) Diffusive wave imaging between stations and lines.
7. local and regional moment tensor analysis for stress state.
8. regional wave crustal modelling
9. deep disconitinuity imaging for LAB, 410, 660, etc arrivals
10. construction of the master model (all results on SAME spatial
overlay).
11. Inline Pn velocity analysis
Data sharing formats (both: us and rest of group)
1. antelope dbase tables (north and south individual dbases
and one-master dbase): site-sitechan-instrument-sensornetwork-wfdisc-arrival-assoc. CORRECT, (time dependent)
sitechan table of orientations.
2. telesesimic travel-times (css3.0 residual table and a
matlab matrix)
3. teleseismic tomography images (matlab matrix).
4. migrated p-sv-sh response (matlab matrix?)
5. anisotropy (add anisotropy table to css3.0 schema)
Batholith experiment visit schedule
1. march-april
2. june-july
3. september-october (demobilization)
Evt.111.baz.345.del.87.F.0.05.0.4
0.8
BN23
0.6
BN22
BN21
0.4
BN20
0.2
0
BN19
0
20
Time (s)
40
60
4
Median Trace
Aligned traces
Station
Median Envelope
1
BN18
BN17
BN16
2
BN15
BN14
0
BN13
−2
−4
BN04
0
20
Time (s)
40
60
20
30
Time (s)
40
Evt.119.baz.22.del.47.F.0.05.0.4
BN23
1
BN22
BN21
0.5
0
BN20
BN19
0
20
Time (s)
40
60
3
BN18
BN17
BN16
2
Median Trace
Aligned traces
Station
Median Envelope
1.5
BN15
1
BN14
0
−1
BN13
−2
BN04
−3
0
20
Time (s)
40
60
20
30
Time (s)
40
29
Figure 4. Comparison of VP and VP/VS for ACCRETE rock samples (Table 1) with seismic
measurements (Figures 2a-b). Laboratory velocities were corrected for high crustal temperature
under the CMB, using a temperature/pressure gradient of 0.8°C/MPa [Christensen, 1979].
Velocities between lithostatic pressures of 400 and 1000 MPa are shown, with filled symbols
corresponding to 800 MPa and arrows pointing in the direction of increasing pressure and
temperature. Note that high crustal temperature leads to velocities that are nearly constant or even
Figure 3. Apatite He ages as function of elevation for three vertical
profiles located in Figure 2, with 2 s uncertainties. Regression
through Foch Mountains profile yields exhumation rate of 0.22 mm/
yr, consistent with other two profiles as well.
Figure 2. Douglas Channel study area. Symbols denote sea-level
sampling locations and apatite (U-Th)/He ages. Three vertical sampling profiles are labeled. Bearing of line A-A9 is N40E.
basin subsidence (Hyndman and Hamilton, 1991; Rohr and Dietrich,
1991; Rohr and Currie, 1997). In the study area (Figs. 1 and 2), the
Coast Mountains are composed of high-grade metamorphic rocks and
Mesozoic and Tertiary plutons (Roddick, 1970). Samples for this study
are tonalite and granodiorite with several metamorphosed granitoid
rocks.1
A deep fjord, Douglas Channel, provides excellent rock exposure
for sampling of the range both in the vertical and horizontal dimensions. It crosses the northwest trend of the Coast Mountains adjacent
to the Queen Charlotte basin and so should reflect the effects of lithospheric-scale deformation proposed by Rohr and Currie (1997). We
analyzed 33 samples from an ;85-km-long sea-level traverse to identify possible fault offsets, block tilting, and paleotopographic effects
on He age patterns. In addition, 22 samples on 3 vertical transects were
analyzed from the southwest, central, and northeast portions of the
traverse (Mount Jenkinson, Foch Mountains, and Mount Elizabeth profiles, respectively, Fig. 2). These three profiles provide between 1 and
1.6 km of vertical coverage, each sampled over a few kilometers lateral
distance.
RESULTS
Apatite He age data are reported in the Appendix (see footnote
1). He ages on the vertical transects range from 4.5 to 19.4 Ma and
correlate strongly with sample elevation (Fig. 3). Over the ;1100 m
from sea level to the summit of Mount Jenkinson, He ages increase
from 7 to 11 Ma. The Foch Mountains transect is similar: He ages
increase from 4.5 Ma at sea level to 11.6 Ma at 1400 m. At Mount
1GSA Data Repository item 200115, Appendix 1, Methods and data, is
available on request from Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301-9140, editing@geosociety.org, or at www.geosociety.org/
pubs/ft2001.htm.
100
Elizabeth, He ages increase from 5.7 Ma at 600 m to 9.5 Ma at 1450
m; above this elevation ages increase much more rapidly, to 19.4 Ma
at 1800 m. A striking feature of these data is the linear age-elevation
segment in all three transects. A regression line through the most complete profile (Foch Mountains) yields a slope of 0.22 6 0.01 mm/yr;
this slope is very similar to those on the Mount Jenkinson (0.28 6
0.03) and lower segment of the Mount Elizabeth (0.24 6 0.03) profiles.
Although they share essentially the same slope, the three profiles are
shifted from each other, such that the He age found at sea level on
Mount Jenkinson is found at ;500 m in the Foch Mountains and ;900
m at Mount Elizabeth. The vertical displacement between these profiles
is similar to the relative mean elevation of the sampling locations,
which for the Mount Jenkinson, Foch Mountains, and Mount Elizabeth
regions increases from 401 m to 634 m to 1051 m.
Sea-level apatites along Douglas Channel yield He ages that range
from 2.8 to 14 Ma; the youngest ages are near the center of the traverse
and the oldest ages are at the southwest end (Fig. 2). When projected
onto a line crossing the range (A-A9, Figs. 2 and 4A), most He ages
along the traverse define a shallow asymmetric v-shape. Except for two
samples at the southwest end of the traverse, ages decline from ca. 9
Ma in the southwest to ca. 4 Ma at the center, then increase to ca. 5.5
Ma in the northeast. As in the vertical profiles, ages on the sea-level
traverse correspond with local mean elevation (10 3 10 km grid) measured beside Douglas Channel: samples from regions of higher local
mean elevation have lower He ages (Fig. 4B). The two anomalously
old samples at the southwest end of the traverse are interpreted as the
result of local faulting (Rusmore et al., 2000).
INTERPRETATION
Apatite He ages throughout this part of the Coast Mountains are
young, requiring exhumation from temperatures of .70 8C since the
middle Miocene. Based on a geothermal gradient of ;28 8C/km (Lewis
et al., 1991; Parrish, 1983), this corresponds to at least 2.2 km of
exhumation. The age-elevation profiles in Figure 3 document three distinct aspects of this exhumation history. The shallowly sloped portion
of the Mount Elizabeth profile defines an exhumed He partial retention
zone established during a period of approximate thermal stability from
before 20 Ma to 10 Ma, during which exhumation was ,0.03 mm/yr.
He age modeling (Wolf et al., 1998) indicates that the shallow slope
GEOLOGY, February 2001
Figure 4. A: Sea-level He ages, projected on line A-A9 in Figure 2.
Two circled samples are likely on single downthrown fault block,
and were excluded from B. B: Position of 8 Ma He age surface
(stepped line, dashed where inferred) compared to mean elevation
(smooth curve). Age surface was constructed by interpolation from
three vertical profiles (triangles) and extrapolation of observed sealevel ages using age-elevation slope of 0.22 mm/yr (open symbols).
Mean elevations were computed in 10 3 10 km squares along length
of profile using digital elevation model (DEM). Because DEM excludes all surfaces covered by seawater, mean elevations were computed for squares abutting Douglas Channel. Elevations of age surface were averaged in 10 km intervals to be compatible with
averaging used on topography.
of this profile is best fit by thermal stability extending to .30 Ma. The
break in slope at ;1500 m elevation indicates the onset of more rapid
exhumation at 10 Ma. After this time all three profiles document relatively slow and nearly constant exhumation of ;0.22 mm/yr.
The data also indirectly define the more recent exhumation history.
A 28 8C/km geothermal gradient and a local mean annual temperature
of 6 8C (Environment Canada, 2000) imply that the apatite closure isotherm (68 8C, Farley, 2000, and Appendix [see footnote 1]) is located
;2.3 km below the Earth’s free cooling surface, or 1.9 km, 1.7 km, and
1.2 km below sea level at Mount Jenkinson, the Foch Mountains, and
Mount Elizabeth, respectively. However, extrapolation of the age-elevation profiles in Figure 3 indicates that in the absence of a change in
exhumation rate, the closure isotherm would be at ;1.4 km, 1.0 km,
and 0.6 km below sea level at these sites. This discrepancy suggests
enhanced recent exhumation of ;0.6 km. The timing of this increase is
best defined at the Foch Mountains, where it must have occurred after
4 Ma. Compared to the exhumation rate from 10 to 4 Ma, the post-4
Ma rate is at least 70% higher (i.e., 0.37 mm/yr if the change occurred
at 4 Ma, and greater if the change occurred later). While large changes
in mean annual temperature may be erroneously interpreted as a change
in exhumation rate (O’Sullivan and Brown, 1999), existing estimates of
the degree of post-Miocene cooling in this region (Mathews and Rouse,
1963) are insufficient to account for the inferred 0.6 km. Unless the
geothermal gradient is substantially higher than estimated, the data require a significant acceleration of exhumation after 4 Ma.
The similarity of the three vertical profiles coupled with the
smooth pattern of ages on the horizontal traverse indicate that the entire
GEOLOGY, February 2001
Douglas Channel region underwent a broadly similar thermal history
from 10 to ca. 4 Ma. Although there is abundant structural evidence
for faulting in the area, some of which corresponds to deviations in
the He age pattern (including the old ages at the southwest end of the
traverse, Rusmore et al., 2000), the He ages in general limit localized
post-10 Ma faulting in the region to less than a few hundred meters of
vertical motion.
As shown in Figure 4B, the elevation of contours of equal He age
exhibit to ;1 km of long-wavelength relief that corresponds with local
mean elevation. This observation is consistent with two alternative
models of the geomorphic evolution of the Coast Mountains. Because
near-surface isotherms will in general mimic topography, He ages of
apatites exhumed through such a temperature field carry a record of
topography present during cooling. A correlation between topography
and He ages in the Sierra Nevada was interpreted by House et al.
(1998) to be the record of long-wavelength paleotopography. If we
apply this interpretation to the Coast Mountains, then the correspondence in Figure 4B documents the topographic expression of the Coast
Mountains back to at least 10 Ma. Based on the present topographic
wavelength along our traverse (;40 km), relief on the closure isotherm
would be ;75% of the topographic relief (House et al., 1998), so the
approximate correspondence in Figure 4B suggests relief similar to or
slightly greater than at present. In other words, this interpretation requires the range to have looked essentially like it does today for the
past 10 m.y. despite more than 2 km of exhumation and a major transition in climate and erosional style following the onset of continental
glaciation ca. 3 Ma. Given the unambiguous evidence for intense glacial incision producing Douglas Channel and tributary fiords, such
long-term geomorphic stasis seems implausible.
Our preferred interpretation is that the range developed exclusively after 4 Ma. In this scenario, differential rock uplift after 4 Ma vertically deflected both the local Earth’s surface and originally near-horizontal He age marker horizons to yield the correlation observed in
Figure 4. This hypothesis suggests a broad upwarp centered on the
Foch Mountains region, and a second and larger uplifted region near
Mount Elizabeth. The magnitude of post-4 Ma differential uplift and
relief development implied by our results is ;1 km. This finding is
compatible with the suggestion by Parrish (1983) that part of the relief
in the southern Coast Mountains developed in the late Neogene. A
significant corollary of this hypothesis is that long-wavelength topography across the range must have been small enough to maintain nearhorizontal isotherms prior to differential uplift. More specifically, the
region must have had subdued long-wavelength relief (less than a few
hundred meters with wavelength .20 km) across the 85 km investigated here, starting at 10 Ma and extending at least until Pliocene time.
On the basis of palynologic and geomorphic studies of sites on the
plateau east of the Coast Mountains, Rouse and Matthews (1979) and
Matthews (1991) similarly inferred a lack of topographic barrier in
middle Miocene time.
DISCUSSION
Apatite fission-track age-elevation profiles led Parrish (1983) and
O’Sullivan and Parrish (1995) to propose two major episodes of exhumation in the southern Coast Mountains, one in the late Eocene (35–
40 Ma) and a second in the late Miocene (5–8 Ma) separated by a
period of relative stability. The fission-track results are complicated by
strong compositional effects on annealing kinetics, but the 30–5 Ma
portion of this history is generally consistent with our He ages in the
northern Coast Mountains. Parrish (1983) also proposed a late Miocene
to Holocene period of exhumation; this may be equivalent to the enhanced post-4 Ma exhumation we propose for Douglas Channel.
The exhumation history we document has implications for both
101
Isotopic evidence of a near surface history for the source rocks of the
central Coast Plutonic Complex, British Columbia, Canada
Paul H. Wetmore#*, Kelley N. Stair#, and Mihai Ducea#
UNIVERSITY OF
SOUTH FLORIDA
#Department of Geosciences, University of Arizona, Tucson, AZ 85721
*Department of Geology, University of South Florida, Tampa, FL 33620
Western Early K
18.4
Western Jurassics
GJP
Pb
204
Pb isotopic compositions all plot within the “mature arc primitive” field of Zartman
and Doe (1981). Ratios become more radiogenic and exhibit less scatter through time.
The intrusions with the most radiogenic Pb are located nearest the Coast Shear Zone
(CSZ).
17.8
17.0
17.5
18.0
18.5
206
Pb/204 Pb i
206
East
19.0
19.5
20.0
-100
20.5
-75
-50
-25
0
25
50
75
Distance from CSZ (km)
Pb/ 204Pb
19.2
15.68
19.0
15.66
15.64
18.8
15.62
18.6
15.60
18.4
15.58
18.2
15.56
18.0
1
-8
15.54
West
17.8
East
100
150
200
-100
-75
-50
-25
0
25
50
75
Distance from CSZ
O Isotopic data
80
PGJ
0.710
0.709
0.706
0.706
0.705
9
-7
7
PGJ
GJP
a
5
9. 7 M
87
5 1.
l
ana +
Sr/ 86 Sr i
0.707
8
l
ba
West
0.703
East
0.702
0.702
0
25
50
75
100
125
150
175
-100
200
-80
-60
-40
7
-7 Ma
GJP 0.7
+
.6
Coast Shear Zone Intrusive
Suite
-20
0
20
40
60
INTERPRETATIONS AND CONCLUSIONS
80
Distance from CSZ
Age (Ma)
55
Douglas Channel
Transect
Ecstall Intrusive Suite
latest Early to earliest Late
Cretaceous Intrusives
12
12
11
11
10
10
Kilometers
25
Paleozoic/Mesozoic
Metasediments and Volcanics
Coast Shear Zone
T
OR
AF
SE
7
7
6
6
-100
West
0
50
100
G
100.0 + 2.3 JP
-4
0
Key
Bella
Bella
-60
-40
-20
0
20
40
Distance from CSZ (Km)
18
G
JP
G
JP 36
-3
5
G
G JPJP 3
-3 3
4
Sr vs. d O
12
G
JP
-2
3
82.8 + 2.8
latest Paleocene and Eocene
intrusives.
Coast Shear Zone Intrusive
Suite
Eastern Late Cretaceous
intrusives.
182.4 + 4.7
61.5 + 1.7
G
JP
-2
Ecstall Intrusive Suite
DEA
G
JP
-3
G
JP
-3
1
N CH
AN N
G
JP
G
-2
JP
8
-2
78.8 + 1.8 9
EL
0
G
JP
-2
1
G
D
G JPG JP- 25
JP 2
-2 6
7
2
C
EAN
HA
11
EL
NN 157.1 + 3.2
10
L
18
G
JP
-1
9
G
JP
-2
BURK
E
9.3 + 0.4 G
JP
-
CHAN
17
G
JP
-1
4
0
NEL
G 52.7 + 1.5
JP
-1
6
G
JP
-1
ACKNOWLEDGMENTS
8
This research was funded through the Continental Dynamics project,
BATHOLITHS. The data and conclusions presented here were made possible through
the direct assistance of many, many colleagues. Specifically, we would like to thank
George Gehrels, P. Jonathan Patchett, and Mark Barton at the University of Arizona,
Jeffrey Ryan at the University of South Florida, Margaret E. Rusmore at Occidental
College, and Glenn J. Woodsworth with the Geological Survey of Canada.
5
G
JP GJ
-1 P98.4 + 1.8 3 12
7
58.4 + 1.9
SO
UN
H
UG
H
FI
TZ
G
JP
-1
1
H
RT 77.3 + 1.7
O
N
BE
G
N
G
JP
-1
JP
-0
9
0
TIN
G
JP
CK
-0
8
G
JP
-0
7
6
MORB
Bella
Coola
5
0.702
0.703
Mariana Arc
0.704
0.705
0.706
87
H
SOUT
G
JP
-0
3
G
G JPJ
G P- 0 4
JP 0
-0 5
6
144.5 + 2.9 GJ
P0
1
The isotopic data from CPC samples reported herein indicate that the composition of
the source rocks for these magmas are not simple mantle-derived diorites that ponded in
the lower crust. While Pb and Sr data are consistent with a relatively primitive source
composition (e.g. mantle-derived diorites) the oxygen data are far to elevated to
represent simple mantle derivation. The heavy oxygen compositions require that
source rocks interacted with meteoric waters at low temperatures. This implies that
prior to being melted to form the CPC magmas, the source rocks were at the surface in a
subaerial environment.
In addition to a period of residence in a subaerial environment interacting with
meteoric waters the rocks that comprise the source(s) of the CPC magmas must have
been geographically isolated from any continent or continental terrane. This follows
from the fact that Pb and Sr isotopic data are not sympathetically enriched with oxygen.
If such rocks were spatially associated with these source rocks it is expected that the
meteoric waters would have acquired the more evolved radiogenic isotopic
compositions that characterize continental rocks.
Japan Arcs
D
BU
E
K
R
A
H
C
E
NN
G
JP
-
Ch.
Coast Shear Zone.
9
n
a
R
s
e
g
18
Paleozoic/Mesozoic
Metasediments and Volcanics.
d O
E
ER
CH
Jurassic Intrusives.
r
a
l
u
ns lith
i
n o
e
P ath
B
G
JP
-2
4
U
LABO
G 92.1 + 2.1
JP
-3
2
Jurassic volcanics.
-80
Age (Ma)
King Island Syenite.
latest Early to earliest Late
Cretaceous Intrusives
Jurassic metasediments.
200
L
20
G107.5 + 1.5
G JPJP 3
-3 7
8
150
East
NE
10
Kilometers
107.6 + 3.0
G
JP
-3
9
Miocene volcanics.
BENTINCK
Powders of each sample were dissolved using traditional HF-HNO3 methods. Sr
and Pb were isolated using Sr-spec resins in disposable columns. No spikes were
added. Isotopic concentrations of both Sr and Pb were determined on a
multicollector Inductively Coupled Mass Spectrometer (GVI Isoprobe) at the Arizona
LaserChron Center, University of Ariznoa. Measured Sr ratios were normalized to a
87
Sr/86Sr = 0.7102 4 value for NBS 987. NBS 987 was analyzed 145 times during the
87
86
207
206
study with an average Sr/ Sr = 0.71027. Lead data were normalized to Pb/ Pb =
207
206
0.91475 for NBS 981. The average Pb/ Pb value for 112 analyses of NBS 981
during this study was 0.91460. Clean, hand picked quartz separates were
decomposed in Nickle bombs in the presence of BrF5 and converted to CO2 gas in the
of heated graphite in Stable Isotope lab of Dr. Mark Barton at the University of
Arizona. Oxygen isotopic values were normalized using in-house standards (Barton
Lab Brazilian Quartzite).
8
9
AN
CH
0
ANALYTICAL METHODS
8
H
Dean-Burk Channel
Transect
9
d 18 O
Paleozoic Intrusive
d 18 O
Central Gneiss Complex
Western Jurassic Intrusives
The d O values range from 6.9 to 11.3 %0 with an average value of 8.8 %0. Samples of
intrusions of all ages yield an ~2%0 range in compositions with the western Late Jurassic
18
yielding the heaviest oxygen values. However, for nearly all samples the d O is
significantly elevated over those of typical island arc or MORB derived lavas.
0.704
0.704
latest Paleocene and Eocene
intrusives.
S r/ 8 6 S r I
0.708
0.708
rC
dne
Gar
ss
ce
in l
Pr ya
Ro land
Is
a
M
60 .7
P- + 1
GJ 8.4
14
nal
50
0.710
Jurassic volcanics.
INTRODUCTION
The average Sr isotopic composition of all samples analyzed is 0.7042. With the
exception of the King Island Syenite, compositions overlap those of most island arc lavas
(e.g. Mariana Arc). The wester Late Jurassic plutons exhibit the largest scatter of initial
ratios (0.7035 to 0.706).
18
Key
N
Sr Isotopic Data
15.52
0
Age (Ma)
nel
Ursula Chan
18.0
87
Prin
Cha cipe
nne
l
Is.
pani
a
Pb Isotopic Data
Pb/204 Pb
7
West
The isotopic data from this study are displayed in the plots to the left. Here we
enumerate some of what we consider to be the significant insights the data provide.
18.2
207
-8
Fields derived from
Zartman and Doe (1981)
MORB
15.10
16.5
GJP
a
Gardner C
1
le
a
h
W
l
e
n
n
a
h
C
Pb/
15.50
18.6
206
Pb/204Pb
207
1
-9 a
GJP22M
3
88
PGJ
nne
l
Ecstall Equivalent
a
M
6
Cam
E Late Jr Intrusives
15.60
18.8
15.30
86 a
P- .1 M
GJ+ 1
.9
Cha
at
0
-9 a
GJP 52M
d
slan ks
I
c
i
c
Ro
ean
Oc canic
Vol
E Late K Intrusives
15.40
81
7
84 0 .
P- +
GJ 60.9
-6
10.23
v
as
n
tio
a
t
el
n
an
Ch
KEY ASPECTS OF THE DATA
19.0
CSZ Intrusives
15.20
0
Generation and evolution of continental arcs remains a fundamental problem in the
Earth sciences. Specifically, the dominant magmas from the mantle in all tectonic
settings are basaltic, yet the bulk composition of continental crust is andesitic. An
increasingly accepted view of Cordilleran arc magma generation is the anatexis of
mantle derived diorites that have ponded in the lower crust. In large part this model
stems from the observation that relatively silicic magmas possess juvenile or primitive
radiogenic isotopic compositions. While this model may, in fact, be generally applicable
to some portions of the North American Cordillera we argue here that at least some
portions of this extensive Mesozoic and early Cenozoic arc system do not necessarily
conform to this view of batholith generation.
The Coast Plutonic Complex (CPC) of western British Columbia, Canada is one of
the northern most segments of the Mesozoic and Tertiary batholiths that help to define
the North American Cordillera. As part of a Continental Dynamics Project,
BATHOLITHS, we have generated a regional isotopic database on the numerous coastparallel age-equivalent plutonic belts that comprise the CPC. Approximately 60
plutonic samples were collected along two transects of the CPC and Pb and Sr isotopic
compositions were determined each. When combined with major and trace elemental
data the CPC rocks are characterized as being both relatively silicic and isotopically
primitive suggesting possible derivation from the melting of mantle-derived diorites
ponded in the lower crust. However, oxygen isotopic data determined for 22 of the
samples suggest that at least some of the source rocks must have experienced near
surface residence prior to being melted.
Mature Arc
Primitive
~Eocene Intrusives
15.70
a
M
7 a
0.
+ .8 M
1
.5
49 .8 +
82 1
P- 3 8
GJ JP-8
G
Squ
ally
tim
Ki
15.80
65
P- a
GJ.4 M
1
+
38.154
9
0.
3
-6
15.567
62 a
P- .1 M
GJ + 1
.5
18.591
94
0.70474
-6
142.5 to 188
GJP
Western Late J Intrusives
De
10
GJP
8.65
64
P- Ma
GJ1.7
+
7
7.
38.239
8.54
10
15.568
l
Gi land
Is
18.745
38.353
8.00
9
0.70434
15.593
37.966
5
PGJ
94.5 to 107.7
18.902
15.559
a
M
Western Middle K Intrusives
0.70414
18.375
8
-5 Ma
GJP1.5
+
2
4.
78.8 to 92.1
0.70334
14
Ecstall Intrusives
144.5 to 182.4
GJP
4
Eastern Late J Intrusives
57 a
P- 7 M
GJ+ 1.
7
6.
9.31
55 .5
P- + 1
GJ7.8
14
38.259
6
-5 Ma
GJP 1.9
+
5
2.
15.573
8.10
14
14
18.735
38.209
3
0.70363
15.588
5
PGJ
61.5 to 81.9
18.857
y
ne
wd
De
Is
Eastern Late K Intrusives
0.70438
-5
58.4 to 60.9
GJP
CSZ Intrusives
D
a
M
8 9
1. P-8
GJ
9.54
Douglas Channel
5
-8
GJP a
M
59
38.425
7
15.597
6
PGJ
18.959
2
0.70437
5
PGJ
a
9M
49.5 to 52.7
+
a
M
1
1.
+ -68
.0
97 GJP
14
ch
ut
Tr land
Is
Paleocene-Eocene Intrusives
as
l
g
u
o
2
0.
7.35
69
P- a
GJ .2 M
1
+
38.468
.9
1
15.595
91
-5
19.034
7
2. Ma
15 1.9
+
0.70602
0
9.3
5
PGJ
King Is. Syenite
l
e
n
n
a
h
C
0
8. Ma
18 2.6
+ -7 6
GJP
 O
19.2
King Is. Syenite
13
18
74
P- 5
GJ -7
GJP
73
PGJ 2
7
PGJ
Pb/
204
Pbi
1
-7 a
M
GJP0.9
+
208
70
P- Ma
GJ110
94
Pb/
204
Pbi
.6
207
GJP
Pb/
204
Pbi
82
206
G
r
e
n
v
i
l
le Chan
ne l
Sr/86Sr
initial
tt
Pi land
Is
87
15.90
za
ta
is
Ar land
Is
Radiogenic isotopic data from intrusive suites have recently been the key to
interpretations that many of the Cordilleran batholiths are the direct products of
partial melting of mantle derived diorites that ponded within the lower crust. However,
results from a regional study of the Coast Plutonic Complex of British Columbia,
Canada, part of the BATHOLITHS Continental Dynamics project, conflict with this
interpretation and indicate that a substantial portion of the source rocks resided at near
surface levels at some time prior to burial and partial melting.
Radiogenic isotopic data, including Sr, and Pb, completed on more than 60 plutonic
samples ranging in age from 322 Ma to 9 Ma indicate that the source regions for these
melts were indeed very primitive. Initial Strontium 87/86 ratios range from 0.7032 up to
0.7062 with a mean of 0.7042. Similarly, lead isotopic data range from 19.675 to 19.520
206
204
207
204
208
204
for Pb/ Pb, 15.550 to 15.662 for Pb/ Pb, and 38.155 to 39.589 for Pb/ Pb. In
contrast to these relatively primitive isotopic data, oxygen data indicate a somewhat
more evolved component to or characteristic of the deep crustal source regions for these
melts. Delta 18O values for quartz separates determined for more than 30 of the
samples range from 7.1 %0 up to 9.8 %0. Such high oxygen values clear preclude the
possibility that these melts were generated from mantle derived diorites just as the Sr
and Pb data preclude significant contamination from an evolved crustal source.
We argue that these data indicate that the rocks that formed the source region for
these melts were originally derived from the mantle but experienced a period of near
surface residence after initial crystallization and before they were again melted to form
the igneous bodies of the Coast Plutonic Complex. During this interval these highly
primitive rocks would have interacted with meteoric waters at low temperatures in
order to change their oxygen isotopic composition away from their mantle signatures.
Such a process of depression and remelting of once epizonal rocks is also consistent with
the common Pb data which tend plot within the Mature Primitive Arc field of Zartman
and Doe (1981).
Pb/238U zircon
age (Ma)
s
nk
Ba land
Is
ABSTRACT
206
N
Average Isotopic Compositions
USF
0.707
0.708
0.709
0.710
86
Sr/ Sr
Modified from Taylor & Scharf, 1986
Isotopic evidence of a near surface history for the source rocks of the
central Coast Plutonic Complex, British Columbia, Canada
Paul H. Wetmore#*, Kelley N. Stair#, and Mihai Ducea#
UNIVERSITY OF
SOUTH FLORIDA
#Department of Geosciences, University of Arizona, Tucson, AZ 85721
*Department of Geology, University of South Florida, Tampa, FL 33620
Western Early K
18.4
Western Jurassics
GJP
Pb
204
Pb isotopic compositions all plot within the “mature arc primitive” field of Zartman
and Doe (1981). Ratios become more radiogenic and exhibit less scatter through time.
The intrusions with the most radiogenic Pb are located nearest the Coast Shear Zone
(CSZ).
17.8
17.0
17.5
18.0
18.5
206
Pb/204 Pb i
206
East
19.0
19.5
20.0
-100
20.5
-75
-50
-25
0
25
50
75
Distance from CSZ (km)
Pb/ 204Pb
19.2
15.68
19.0
15.66
15.64
18.8
15.62
18.6
15.60
18.4
15.58
18.2
15.56
18.0
1
-8
15.54
West
17.8
East
100
150
200
-100
-75
-50
-25
0
25
50
75
Distance from CSZ
O Isotopic data
80
PGJ
0.710
0.709
0.706
0.706
0.705
9
-7
7
PGJ
GJP
a
5
9. 7 M
87
5 1.
l
ana +
Sr/ 86 Sr i
0.707
8
l
ba
West
0.703
East
0.702
0.702
0
25
50
75
100
125
150
175
-100
200
-80
-60
-40
7
-7 Ma
GJP 0.7
+
.6
Coast Shear Zone Intrusive
Suite
-20
0
20
40
60
INTERPRETATIONS AND CONCLUSIONS
80
Distance from CSZ
Age (Ma)
55
Douglas Channel
Transect
Ecstall Intrusive Suite
latest Early to earliest Late
Cretaceous Intrusives
12
12
11
11
10
10
Kilometers
25
Paleozoic/Mesozoic
Metasediments and Volcanics
Coast Shear Zone
T
OR
AF
SE
7
7
6
6
-100
West
0
50
100
G
100.0 + 2.3 JP
-4
0
Key
Bella
Bella
-60
-40
-20
0
20
40
Distance from CSZ (Km)
18
G
JP
G
JP 36
-3
5
G
G JPJP 3
-3 3
4
Sr vs. d O
12
G
JP
-2
3
82.8 + 2.8
latest Paleocene and Eocene
intrusives.
Coast Shear Zone Intrusive
Suite
Eastern Late Cretaceous
intrusives.
182.4 + 4.7
61.5 + 1.7
G
JP
-2
Ecstall Intrusive Suite
DEA
G
JP
-3
G
JP
-3
1
N CH
AN N
G
JP
G
-2
JP
8
-2
78.8 + 1.8 9
EL
0
G
JP
-2
1
G
D
G JPG JP- 25
JP 2
-2 6
7
2
C
EAN
HA
11
EL
NN 157.1 + 3.2
10
L
18
G
JP
-1
9
G
JP
-2
BURK
E
9.3 + 0.4 G
JP
-
CHAN
17
G
JP
-1
4
0
NEL
G 52.7 + 1.5
JP
-1
6
G
JP
-1
ACKNOWLEDGMENTS
8
This research was funded through the Continental Dynamics project,
BATHOLITHS. The data and conclusions presented here were made possible through
the direct assistance of many, many colleagues. Specifically, we would like to thank
George Gehrels, P. Jonathan Patchett, and Mark Barton at the University of Arizona,
Jeffrey Ryan at the University of South Florida, Margaret E. Rusmore at Occidental
College, and Glenn J. Woodsworth with the Geological Survey of Canada.
5
G
JP GJ
-1 P98.4 + 1.8 3 12
7
58.4 + 1.9
SO
UN
H
UG
H
FI
TZ
G
JP
-1
1
H
RT 77.3 + 1.7
O
N
BE
G
N
G
JP
-1
JP
-0
9
0
TIN
G
JP
CK
-0
8
G
JP
-0
7
6
MORB
Bella
Coola
5
0.702
0.703
Mariana Arc
0.704
0.705
0.706
87
H
SOUT
G
JP
-0
3
G
G JPJ
G P- 0 4
JP 0
-0 5
6
144.5 + 2.9 GJ
P0
1
The isotopic data from CPC samples reported herein indicate that the composition of
the source rocks for these magmas are not simple mantle-derived diorites that ponded in
the lower crust. While Pb and Sr data are consistent with a relatively primitive source
composition (e.g. mantle-derived diorites) the oxygen data are far to elevated to
represent simple mantle derivation. The heavy oxygen compositions require that
source rocks interacted with meteoric waters at low temperatures. This implies that
prior to being melted to form the CPC magmas, the source rocks were at the surface in a
subaerial environment.
In addition to a period of residence in a subaerial environment interacting with
meteoric waters the rocks that comprise the source(s) of the CPC magmas must have
been geographically isolated from any continent or continental terrane. This follows
from the fact that Pb and Sr isotopic data are not sympathetically enriched with oxygen.
If such rocks were spatially associated with these source rocks it is expected that the
meteoric waters would have acquired the more evolved radiogenic isotopic
compositions that characterize continental rocks.
Japan Arcs
D
BU
E
K
R
A
H
C
E
NN
G
JP
-
Ch.
Coast Shear Zone.
9
n
a
R
s
e
g
18
Paleozoic/Mesozoic
Metasediments and Volcanics.
d O
E
ER
CH
Jurassic Intrusives.
r
a
l
u
ns lith
i
n o
e
P ath
B
G
JP
-2
4
U
LABO
G 92.1 + 2.1
JP
-3
2
Jurassic volcanics.
-80
Age (Ma)
King Island Syenite.
latest Early to earliest Late
Cretaceous Intrusives
Jurassic metasediments.
200
L
20
G107.5 + 1.5
G JPJP 3
-3 7
8
150
East
NE
10
Kilometers
107.6 + 3.0
G
JP
-3
9
Miocene volcanics.
BENTINCK
Powders of each sample were dissolved using traditional HF-HNO3 methods. Sr
and Pb were isolated using Sr-spec resins in disposable columns. No spikes were
added. Isotopic concentrations of both Sr and Pb were determined on a
multicollector Inductively Coupled Mass Spectrometer (GVI Isoprobe) at the Arizona
LaserChron Center, University of Ariznoa. Measured Sr ratios were normalized to a
87
Sr/86Sr = 0.7102 4 value for NBS 987. NBS 987 was analyzed 145 times during the
87
86
207
206
study with an average Sr/ Sr = 0.71027. Lead data were normalized to Pb/ Pb =
207
206
0.91475 for NBS 981. The average Pb/ Pb value for 112 analyses of NBS 981
during this study was 0.91460. Clean, hand picked quartz separates were
decomposed in Nickle bombs in the presence of BrF5 and converted to CO2 gas in the
of heated graphite in Stable Isotope lab of Dr. Mark Barton at the University of
Arizona. Oxygen isotopic values were normalized using in-house standards (Barton
Lab Brazilian Quartzite).
8
9
AN
CH
0
ANALYTICAL METHODS
8
H
Dean-Burk Channel
Transect
9
d 18 O
Paleozoic Intrusive
d 18 O
Central Gneiss Complex
Western Jurassic Intrusives
The d O values range from 6.9 to 11.3 %0 with an average value of 8.8 %0. Samples of
intrusions of all ages yield an ~2%0 range in compositions with the western Late Jurassic
18
yielding the heaviest oxygen values. However, for nearly all samples the d O is
significantly elevated over those of typical island arc or MORB derived lavas.
0.704
0.704
latest Paleocene and Eocene
intrusives.
S r/ 8 6 S r I
0.708
0.708
rC
dne
Gar
ss
ce
in l
Pr ya
Ro land
Is
a
M
60 .7
P- + 1
GJ 8.4
14
nal
50
0.710
Jurassic volcanics.
INTRODUCTION
The average Sr isotopic composition of all samples analyzed is 0.7042. With the
exception of the King Island Syenite, compositions overlap those of most island arc lavas
(e.g. Mariana Arc). The wester Late Jurassic plutons exhibit the largest scatter of initial
ratios (0.7035 to 0.706).
18
Key
N
Sr Isotopic Data
15.52
0
Age (Ma)
nel
Ursula Chan
18.0
87
Prin
Cha cipe
nne
l
Is.
pani
a
Pb Isotopic Data
Pb/204 Pb
7
West
The isotopic data from this study are displayed in the plots to the left. Here we
enumerate some of what we consider to be the significant insights the data provide.
18.2
207
-8
Fields derived from
Zartman and Doe (1981)
MORB
15.10
16.5
GJP
a
Gardner C
1
le
a
h
W
l
e
n
n
a
h
C
Pb/
15.50
18.6
206
Pb/204Pb
207
1
-9 a
GJP22M
3
88
PGJ
nne
l
Ecstall Equivalent
a
M
6
Cam
E Late Jr Intrusives
15.60
18.8
15.30
86 a
P- .1 M
GJ+ 1
.9
Cha
at
0
-9 a
GJP 52M
d
slan ks
I
c
i
c
Ro
ean
Oc canic
Vol
E Late K Intrusives
15.40
81
7
84 0 .
P- +
GJ 60.9
-6
10.23
v
as
n
tio
a
t
el
n
an
Ch
KEY ASPECTS OF THE DATA
19.0
CSZ Intrusives
15.20
0
Generation and evolution of continental arcs remains a fundamental problem in the
Earth sciences. Specifically, the dominant magmas from the mantle in all tectonic
settings are basaltic, yet the bulk composition of continental crust is andesitic. An
increasingly accepted view of Cordilleran arc magma generation is the anatexis of
mantle derived diorites that have ponded in the lower crust. In large part this model
stems from the observation that relatively silicic magmas possess juvenile or primitive
radiogenic isotopic compositions. While this model may, in fact, be generally applicable
to some portions of the North American Cordillera we argue here that at least some
portions of this extensive Mesozoic and early Cenozoic arc system do not necessarily
conform to this view of batholith generation.
The Coast Plutonic Complex (CPC) of western British Columbia, Canada is one of
the northern most segments of the Mesozoic and Tertiary batholiths that help to define
the North American Cordillera. As part of a Continental Dynamics Project,
BATHOLITHS, we have generated a regional isotopic database on the numerous coastparallel age-equivalent plutonic belts that comprise the CPC. Approximately 60
plutonic samples were collected along two transects of the CPC and Pb and Sr isotopic
compositions were determined each. When combined with major and trace elemental
data the CPC rocks are characterized as being both relatively silicic and isotopically
primitive suggesting possible derivation from the melting of mantle-derived diorites
ponded in the lower crust. However, oxygen isotopic data determined for 22 of the
samples suggest that at least some of the source rocks must have experienced near
surface residence prior to being melted.
Mature Arc
Primitive
~Eocene Intrusives
15.70
a
M
7 a
0.
+ .8 M
1
.5
49 .8 +
82 1
P- 3 8
GJ JP-8
G
Squ
ally
tim
Ki
15.80
65
P- a
GJ.4 M
1
+
38.154
9
0.
3
-6
15.567
62 a
P- .1 M
GJ + 1
.5
18.591
94
0.70474
-6
142.5 to 188
GJP
Western Late J Intrusives
De
10
GJP
8.65
64
P- Ma
GJ1.7
+
7
7.
38.239
8.54
10
15.568
l
Gi land
Is
18.745
38.353
8.00
9
0.70434
15.593
37.966
5
PGJ
94.5 to 107.7
18.902
15.559
a
M
Western Middle K Intrusives
0.70414
18.375
8
-5 Ma
GJP1.5
+
2
4.
78.8 to 92.1
0.70334
14
Ecstall Intrusives
144.5 to 182.4
GJP
4
Eastern Late J Intrusives
57 a
P- 7 M
GJ+ 1.
7
6.
9.31
55 .5
P- + 1
GJ7.8
14
38.259
6
-5 Ma
GJP 1.9
+
5
2.
15.573
8.10
14
14
18.735
38.209
3
0.70363
15.588
5
PGJ
61.5 to 81.9
18.857
y
ne
wd
De
Is
Eastern Late K Intrusives
0.70438
-5
58.4 to 60.9
GJP
CSZ Intrusives
D
a
M
8 9
1. P-8
GJ
9.54
Douglas Channel
5
-8
GJP a
M
59
38.425
7
15.597
6
PGJ
18.959
2
0.70437
5
PGJ
a
9M
49.5 to 52.7
+
a
M
1
1.
+ -68
.0
97 GJP
14
ch
ut
Tr land
Is
Paleocene-Eocene Intrusives
as
l
g
u
o
2
0.
7.35
69
P- a
GJ .2 M
1
+
38.468
.9
1
15.595
91
-5
19.034
7
2. Ma
15 1.9
+
0.70602
0
9.3
5
PGJ
King Is. Syenite
l
e
n
n
a
h
C
0
8. Ma
18 2.6
+ -7 6
GJP
 O
19.2
King Is. Syenite
13
18
74
P- 5
GJ -7
GJP
73
PGJ 2
7
PGJ
Pb/
204
Pbi
1
-7 a
M
GJP0.9
+
208
70
P- Ma
GJ110
94
Pb/
204
Pbi
.6
207
GJP
Pb/
204
Pbi
82
206
G
r
e
n
v
i
l
le Chan
ne l
Sr/86Sr
initial
tt
Pi land
Is
87
15.90
za
ta
is
Ar land
Is
Radiogenic isotopic data from intrusive suites have recently been the key to
interpretations that many of the Cordilleran batholiths are the direct products of
partial melting of mantle derived diorites that ponded within the lower crust. However,
results from a regional study of the Coast Plutonic Complex of British Columbia,
Canada, part of the BATHOLITHS Continental Dynamics project, conflict with this
interpretation and indicate that a substantial portion of the source rocks resided at near
surface levels at some time prior to burial and partial melting.
Radiogenic isotopic data, including Sr, and Pb, completed on more than 60 plutonic
samples ranging in age from 322 Ma to 9 Ma indicate that the source regions for these
melts were indeed very primitive. Initial Strontium 87/86 ratios range from 0.7032 up to
0.7062 with a mean of 0.7042. Similarly, lead isotopic data range from 19.675 to 19.520
206
204
207
204
208
204
for Pb/ Pb, 15.550 to 15.662 for Pb/ Pb, and 38.155 to 39.589 for Pb/ Pb. In
contrast to these relatively primitive isotopic data, oxygen data indicate a somewhat
more evolved component to or characteristic of the deep crustal source regions for these
melts. Delta 18O values for quartz separates determined for more than 30 of the
samples range from 7.1 %0 up to 9.8 %0. Such high oxygen values clear preclude the
possibility that these melts were generated from mantle derived diorites just as the Sr
and Pb data preclude significant contamination from an evolved crustal source.
We argue that these data indicate that the rocks that formed the source region for
these melts were originally derived from the mantle but experienced a period of near
surface residence after initial crystallization and before they were again melted to form
the igneous bodies of the Coast Plutonic Complex. During this interval these highly
primitive rocks would have interacted with meteoric waters at low temperatures in
order to change their oxygen isotopic composition away from their mantle signatures.
Such a process of depression and remelting of once epizonal rocks is also consistent with
the common Pb data which tend plot within the Mature Primitive Arc field of Zartman
and Doe (1981).
Pb/238U zircon
age (Ma)
s
nk
Ba land
Is
ABSTRACT
206
N
Average Isotopic Compositions
USF
0.707
0.708
0.709
0.710
86
Sr/ Sr
Modified from Taylor & Scharf, 1986
2Dmodes
http://w3.uwyo.edu/~seismic/accrete/Images/2Dmodes.htm (1 of 2)2/14/2006 11:09:14 AM
Evt.89.baz.221.del.92.F.0.05.0.4
BN23
1
BN22
BN21
0.5
0
BN20
BN19
0
20
Time (s)
40
60
4
Median Trace
Shifted traces
Station
Median Envelope
1.5
BN18
BN17
BN16
2
BN15
BN14
0
BN13
−2
−4
BN04
0
20
Time (s)
40
60
20
25
30 35
Time (s)
40