This contrasts with our results, which show a distinctively heterogeneous lithosphere Fig. From periods of 30—60 s Fig. S1 in Supporting Information , fast velocity zones are observed in central-western Baja and beneath the central-southern Gulf.
In general at 50 s Fig. This wedge-shaped low is present at all periods and is particularly prominent at longer periods. At 50 s, the plate boundary in the northern Gulf is relatively slow, while its segment between the Guaymas and Pescadero basins appears fast. The map at 50 s in Fig. Going deeper into the mantle 60— s in Fig. In all cases, it is closely associated with the BTF, the longest active transform fault in the Gulf.
Low velocities are also observed at these periods beneath northern Baja near the southern California border. The fast velocity region beneath western Baja is again evident at 90 s and longer periods. We conclude that the major features at a period of 30 s are similar, but in the range 30—50 s our results Fig. In order to give an estimate of the spatial resolution of dispersion maps in Fig.
We created a non-standard checkerboard with rectangular 1. The cell size is 0. This creates a pattern that is not aligned with the Gulf coastline and provides a reliable test of the lateral resolution of our model both parallel to and perpendicular to the coastline.
Results of the tomographic inversion Fig. Checkerboard resolution test for Rayleigh wave group velocities at 15, 30 and 50 s middle panels corresponding to the input model shown in the top panel.
The ray path distribution at corresponding periods is shown in the bottom panels. The initial crustal model as well as the thickness of sediments varies laterally depending on whether the crust is continental or oceanic. The inversion procedure is based on taking the starting velocity model, calculating the predicted dispersion curve, and comparing it to the observed dispersion.
The starting model is then iteratively refined until it fits the observations. Iterations are controlled to avoid artificial features such as spurious low-velocity zones. We invert for both the layer thickness and velocity and each layer is equally weighted. We did not consider any physical dispersion corrections. The least-squares inversion includes a damping factor that limits the range of variations in the model between two subsequent iterations.
The damping factor affects how fast the model converges and minimizes residuals between the observed and predicted values. In this study, we found that a damping factor of 3 gives the most stable results with a reasonable number of iterations 20 maximum when inverting the Rayleigh wave group velocities. The errors associated with the group velocity tomographic maps described in Section 3. In this inversion, we neglect the effects of anisotropy see Section 3.
In order to reduce this trade-off, the group velocity measurements need to be accurate, systematic errors over a broad range of periods have a larger effect than random errors that vary rapidly with period Lebedev et al. In addition, a priori constraints on crustal and mantle structure, and a large number of measurements at short and long period across the study region, will reduce the velocity—depth trade-off Lebedev et al.
In order to show that shear wave velocity maps do not depend on the initial velocity model, we plot in Fig. Although some discrepancy is observed in the crustal thickness, our resulting S -wave velocity models solid lines in the left-hand panels of Fig. Discrepancies observed at crustal depths between the two models could also be due to the fact that the CRUST2. In the left-hand panel, the obtained solid shear wave velocities versus depth are plotted with respect to the initial models dashed.
The velocity model consists of 39 layers each 5-km thick over a half-space and the damping in all the 30 iterations was 3.
Along the Gulf, the S -wave velocities increase from north to south and reach the highest values at the mouth of the Gulf and beneath the EPR.
This pattern of high velocities beneath the Gulf could likely reflect the differences in the sediment thickness in the northern basins with respect to the southern ones. Note the difference in colour scale for each panel.
Black triangles are the seismic stations and the black line is the plate boundary. In the depth interval 20—30 km Fig. Although in the Gulf generally shear velocities are lower than at shallower depths, two main low-velocity areas are noted: the northernmost one, which underlies the northern Gulf Wagner Basin—Salton Trough , but is also located along the Cerro Prieto Fault and east of the Agua Blanca Fault, a shear zone that extends across the northern peninsula towards the Pacific Fig.
This may indicate hotter or weaker crust north of the Guaymas Basin. The question mark in the NW corner of the interpreted fossil slab shaded in blue indicates that in our study we cannot establish the lateral continuity of the slab, as well as whether or not it extends past the peninsula into the Gulf.
Low shear wave velocity regions are shaded in red. Other main tectonic features are labelled as in the caption of Fig. Black triangles are the seismic stations. NW—SE cross-section dashed red segment in Fig. Contours are at 3. A moving average is applied across 5 gridpoints for smoothing. Tick marks on the x -axis refer to longitudes crossed along the profile. There are three distinct elongated velocity highs located in the southern basin and Range of Sonora at depths of 30—40 km left-central panel in Fig.
These N—NW trending features are also present in the 15 and 30 s group velocity maps Fig. These slow patches are also shown in the E—W cross-sections of Fig. Previous workers Wang et al. Their anomalies were however, located beneath the Wagner, Delfin and Guaymas basins, with no coverage in the southern Gulf, and depending on the depth, the anomalies were off the rift axis.
We find a similar anomaly for the Wagner Basin and beneath the northern part of Delfin Basin at 20—30 and 50—km depths. Our central Gulf anomaly is centred more on the southern segment of the BTF than on the Guaymas Basin in the depth range 20—40 km.
Average shear wave velocities along east—west profiles dashed blue lines in Fig. The average is taken over three grid cells in and out of the plane of the profile. Region boundaries are marked along the top of each profile. More precisely, it does not reach the tip of the peninsula and the Gulf, and is located between the two northernmost slow patches discussed above.
A relatively high-velocity anomaly was also found by Zhang et al. Considering the distribution of volcanic rocks in Baja, Wang et al. The fast velocity zone is also evident in the east—west profiles in Fig. We also want to point out the prominent high-velocity region Fig.
Based on our coverage, this feature is not continuous with the fossil slab we interpret to the north, it is deeper than that and smaller in lateral extent. While we do not have good resolution as far west as Brothers et al. For a better understanding of the significance of the observed anomalies in the study region, it is important to take into account the distribution of the volcanism.
The localized low-velocity patches in Baja Fig. The larger low shear wave velocity regions in central-eastern Baja and the Gulf near the northern portion of the BTF Fig. S2 , although the reliability of those results is limited due to the reduced number of ray paths at periods longer than s. The large low-velocity zones beneath the Gulf likely reflect the upwelling of asthenospheric material in the upper mantle Lizarralde et al. Numerical models of continental break-up Corti et al.
The more anomalous volcanic rocks shown in yellow and green in Fig. Previous studies Zhang et al. In our study, this feature is not as laterally continuous, nor does it cover as large an area as that observed by Wang et al. The lateral continuity of the slab, as well as whether or not it extends past the peninsula into the Gulf is not unambiguous in our model hence a question mark in Fig.
We however note the extent of this high-velocity feature in the E—W profiles shown in Fig. In addition, numerical models show that shearing in a mantle wedge above the slab beneath the peninsula could produce the unique compositional variation in post-subduction volcanism found in Baja Negrete-Aranda et al.
More importantly, the timing of melt production from these numerical models matches geochemical observations. Calmus et al. Tholeiites and alkali basalts of subslab origin rose through this window, while the adakites were derived from the partial melting of its upper lip, close to the trench.
Pallares et al. Castillo instead proposes that the Pacific asthenosphere was the direct source for post-subduction magmas that erupted in Baja California, considering the chronology of magmatism and the isotopic composition of lavas.
The three velocity highs in Sonora at depths of 30—40 km Fig. We note that the lower resolution of the Magsat data — km relative to our shear wave velocity model does not allow all components of the lows and highs in that data set to be distinguished.
In addition, there is only very limited constraint on crustal thicknesses in this region. Two are associated with the northernmost velocity highs in northern Sonora at 30—km depth Fig.
Noteworthy is the MSM correlates roughly with the boundary between a high and low in the second vertical derivative of the Magsat data Campos-Enriquez et al. Station NE80 Fig. At depths of 20—40 km, we also interpret the southernmost velocity high in Sonora as associated with the Seri Terrane or southern segment of the Caborca Block.
It should, however, be noted that station NE82 located at the southern end of this anomaly has a Moho depth of In this case, the deeper shear wave high would instead be associated with the upper mantle. The low velocities extending from the eastern edge of the Gulf across Sinaloa left-central panel in Fig. In the Gulf, we interpret this low-velocity feature as a region of mantle upwelling as previously described above.
We have shown that Rayleigh wave group velocity measurements allow us to derive important information on the Earth structure in the tectonically complex Gulf of California region. We have found heterogeneous lithosphere in the group velocity maps at 30 s, but complex tectonic features are also mapped at lithospheric and asthenospheric depths periods greater than 60 s. This study differs from previous studies in the same area because we use group instead of phase velocity data, and these are more sensitive to shallow structure than phase velocity at corresponding periods.
Since variations in crustal structure greatly influence the propagation of regional phases and consequently earthquake locations, it is very important to have accurate information on crustal shear wave velocity. Our study contributes a reliable velocity model for the Gulf of California region.
Complexities in the lithospheric structure beneath the study region are pointed out in our shear wave velocity maps as well as their association with the main types of volcanic lavas in the area.
We would like to thank J. Contreras, J. Gaherty, R. Negrete-Aranda, M. Pasyanos, J. Lin for the very constructive comments on different aspects of the manuscript. We also thank F. Pereira for his help in developing the tomographic code. We are grateful to B. Savage supplying his velocity model. Comments from two anonymous reviewers and the Editor W. Friederich contributed to improve the final version of the manuscript.
Google Scholar. Google Preview. Additional Supporting Information may be found in the online version of this article:. Figure S1. Tomographic dispersion maps for Rayleigh waves at periods from 60 to s. Corresponding periods are labelled in the bottom left corner of each panel. Note the different colour scale for different periods.
Figure S2. Shear wave velocity at several depth intervals as indicated at the top of each panel. Supplementary Data. Please note: Oxford University Press is not responsible for the content or functionality of any supporting materials supplied by the authors.
Any queries other than missing material should be directed to the corresponding author for the article. Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Search Menu. Article Navigation.
Close mobile search navigation Article Navigation. Volume Article Contents Abstract. Seismic structure beneath the Gulf of California: a contribution from group velocity measurements.
Di Luccio , F. Di Luccio. E-mail: francesca. Oxford Academic. Revision received:. Cite Cite F. Select Format Select format. Permissions Icon Permissions. Abstract Rayleigh wave group velocity dispersion measurements from local and regional earthquakes are used to interpret the lithospheric structure in the Gulf of California region.
Surface waves and free oscillations , Seismic tomography , Dynamics of lithosphere and mantle , Crustal structure. Open in new tab Download slide. The along-path group velocity measurements for multiple periods are converted into tomographic images using kernels which vary in off-path width as a function of the average velocity and of the period.
Off-path rays are weighted according to a cosine function, with a maximum along the master path and a null weight at the boundaries of the kernel band. Because the bandwidth is period-dependent, off-path rays become increasingly effective at longer periods. This spacing was chosen based on the minimum overall bandwidth, which was measured at the s period.
For each period, the tomographic maps shown in Fig. Following Rodi et al. Testing the Mojave-Sonora megashear hypothesis: evidence from Paleoproterozoic igneous rocks and deformed Mesozoic strata in Sonora, Mexico.
Google Scholar Crossref. Search ADS. The current limits of resolution for surface wave tomography in North America. How sediment promotes narrow rifting: applications to the Gulf of California. Farallon slab detachment and deformation of the Magdalena shelf, southern Baja California. The uppermost marine rocks Tcmu and superadjacent nonmarine rocks Tcnm are only slightly deformed and overlie older, deformed strata across an angular unconformity.
The capping Pliocene volcanic units and Quaternary sedimentary units appear to be undeformed. The La Cruz fault is a major northwest-striking, dextral strike-slip fault that parallels the southwestern coastline of the island Fig.
Although the La Cruz fault is primarily a subvertical strike-slip fault, it also has a history of vertical offset, with evidence for both transpressional and transtensional deformation. Overall displacement across the fault in the study area is dextral oblique normal, with down-to-the-northeast vertical offset.
Crystalline basement rocks are exposed southwest of the fault and are not exposed northeast of the fault in the study area Fig. The La Cruz fault cuts unit Tsmt and all older marine and nonmarine units. Adjacent faults and folds do not appear to deform the La Cruz fault. From this we surmise that slip on these secondary structures was contemporaneous with slip on the La Cruz fault in a dextral-wrench shear zone.
However, less dextral offset may be required if down-to-the-northeast displacement across the fault has obscured correlative units. Younger arc-related volcanic rocks northeast of the fault may conceal older arc-related units that are correlative to those now exposed southwest of the fault. Many map units are also deformed by folds and reverse faults.
The traces of fold axial planes are oriented approximately west-northwest—east-southeast, slightly more westward than the northwest-striking La Cruz fault Fig. Displacement on these faults is minor, with tens to hundreds of meters of total slip. Folds and reverse faults only deform unit Tsmt and older units. In general, map units on both sides of the La Cruz fault dip north, except locally in south-dipping fold limbs. This is expressed on the geologic map Fig. Normal faults also are present in the study area Figs.
Mapped normal faults are oriented approximately north-northwest to north-northeast with moderate to steep dips. Displacement on these normal faults is minor, with tens of meters of total displacement. Normal faults deform Tsmt and older units, and cut folds and reverse faults.
The dike system Tcsd that fed the rhyodacite of Cerro Starship Tcsf is oriented approximately north-northeast across SWIT, subparallel to the strike of many mapped normal faults in the study area Fig.
This orientation is similar to the expected orientation of extensional structures e. The Tcsd dike system likely formed under a transtensional strain field. Deformation associated with the La Cruz fault and adjacent faults and folds likely initiated sometime after the emplacement of the youngest arc-related volcanic rocks ca. Deformation was ongoing during sedimentation in the SWIT marine basin. The oldest marine units, Tsm, Tcml, and Tcsh, are faulted and folded in a manner similar to underlying arc-related volcanic rocks.
Intermediate-age marine units Ttoa, Tsmt, and Tcmm are the youngest map units folded or cut by these faults. Younger units, including the Gilbert delta system foreset deposits Tcmu and laterally equivalent nonmarine conglomerate Tcnm , appear to be undeformed and cap all faults in the SWIT study area Fig.
We conclude that deformation associated with the La Cruz fault likely ceased sometime after emplacement of the 6. The results of our new detailed geologic mapping, geochronology, and stratigraphic analysis support the interpretation that marine rocks on SWIT were deposited during latest Miocene to Pliocene time Oskin and Stock, a. Based on these results, we address and resolve long-standing controversies over the age and depositional setting of the SWIT marine basin and the age and emplacement mechanism of several volcanic units that provide key constraints on the timing of marine deposition.
Oskin and Stock a suggested that this breccia outcrop belongs to older volcaniclastic strata stratigraphically below the marine deposits but their mapping was insufficient to verify this notion. The andesite breccia sample locality Fig. On the hill south of the arroyo exposure, Tcml, Tbx, and older units are all deformed by gentle folds Fig. A gently north-dipping fold limb is mapped on the north-facing dip slope of this hill, with beds dipping toward the arroyo exposure.
In the arroyo floor, the distinctive landslide breccia Tbx is in contact stratigraphically above the andesitic breccia Tbxv , and nearby, is stratigraphically below the marine conglomerate. Thus, as suggested in Oskin and Stock a , the andesite breccia at this location is exposed in an erosional window carved through and surrounded by gently dipping marine conglomerate and landslide breccia that unconformably overlie the andesite breccia Fig. In fact, we mapped multiple erosional windows such as this within 0.
These relationships demonstrate that the isotopically dated andesitic breccia is stratigraphically below, rather than within, the marine deposits and is part of the older arc-related volcanic rocks Tbxv. In this study we provide new age constraints that bracket the marine deposits on SWIT.
The lithologically distinctive landslide breccia marker Tbx is exposed almost continuously at the base of the marine section. In general, this breccia directly overlies, in angular unconformity, pre-marine, early rift volcanic rocks e. The age of this distinctive marker unit, and thus the base of the marine deposits, is bracketed between the 6. The youngest SWIT marine deposits predate emplacement of the 3. However, this biostratigraphic age constraint was not incorporated into the conclusions of the original study Gastil et al.
That interpretation is untenable because it assumes a middle Miocene age for the base of the marine section on SWIT, which we show here is incorrect. The results of our study agree with previously published biostratigraphic evidence for a latest Miocene to early Pliocene age of earliest marine sedimentation and conclusively resolve previous age controversies for the SWIT basin.
Of more regional concern, paleontological studies of molluscan assemblages e. Marine deposits at some of these locations have subsequently been shown to be late Miocene in age e. Future paleontological studies in the Gulf of California region should discontinue use of the erroneous middle Miocene interpretation for the age of marine strata and fossil assemblages on SWIT. Our detailed structural and stratigraphic results require a reinterpretation of the controls on SWIT basin geometry, subsidence, and depositional setting.
Most of the subsidence that formed the SWIT basin took place northeast of the La Cruz fault in response to down-to-the-northeast separation. The exposed marine deposits fill an elongate, 1. This trough continues to the southeast for at least an additional 5 km, and is occupied by nonmarine strata that we infer to be laterally equivalent to the marine section Fig. For most of its length, the basin is restricted to the northeast side of the La Cruz fault, supporting the idea that it formed above the down-dropped, northeastern side of this dextral oblique fault zone.
We interpret the map units Tsmt, Tcmm, Tcmu, and Tcnm to represent a linked fluvial-marine Gilbert-type fan delta system e. The transport direction and architecture of the Gilbert-delta system was controlled by the latest stages of deformation on the La Cruz fault; only the lower units of the delta deposits e.
Nonmarine topset deposits Tcnm both truncate Figs. Gently dipping marine tuffaceous sandstone deposits Tsmt are interpreted to be marine turbidites formed in distal Gilbert-delta bottomsets. In places, coarser marine conglomerate and sandstone Tcmm erosionally overlie Tsmt, possibly recording a temporary fall in relative sea level. Marine conglomerate deposits Tcmu overlie Tsmt and Tcmm.
We interpret Tcmu as a system of prograding marine foresets. Sediments were sourced from an updip fluvial system to the southeast, and prograded to the northwest across older, deformed marine units e. The fan delta deposits appear to have filled a marine embayment that formed as a result of subsidence along the La Cruz fault in an area that previously hosted a more heterogeneous, isolated set of fault-bounded marine subbasins.
However, marine units are of variable thickness across the study area and in no single section are all observed at their maximum thickness. Neuhaus and Gastil et al. Our observations indicate that none of these igneous units crosscut or overlie marine strata, but rather are within the stratigraphic section, where they are faulted and folded similar to underlying and overlying map units. A rhyolite from arroyo 1 was interpreted as either a flow that overlies SWIT marine conglomerate Neuhaus, ; unit Mr or as a crosscutting dike Gastil et al.
Their petrographic descriptions are very similar to our description of the tuff of Hast Pitzcal Tthp , which crops out at their sample location on our geologic map Fig.
The previous K-Ar age of 5. In this area, Tthp is tilted and faulted and unequivocally within the stratigraphic section, unconformably beneath marine deposits Figs. Locally, in a side canyon on the western side of arroyo 1, the upper contact of Tthp diapirically intrudes overlying stratified ash and pumice beds of unit Ttua Fig. We observe similar diapiric relationships in outcrops just west of a coastal cove, between arroyo 0 and Hast Pitzcal Fig.
This contact relationship may have occurred as a result of the rapid deposition and overburden pressure of these breccia deposits onto older, unconsolidated deposits of water-saturated tuff. Such local diapric flow contact relationships may have led to previous interpretations that the emplacement of the tuff of Hast Pitzcal occurred as an intrusion. Neuhaus mapped another local rhyolite unit Mry from an outcrop in arroyo 4, in the southeastern corner of the study area Fig. Neuhaus interpreted this unit as a rhyolite flow that overlies SWIT marine conglomerate.
The sample location and petrographic description of Neuhaus are both similar to the tuff of Ensenada Blanca Tteb on our geologic map Fig. Because Tteb is overlain by nonmarine conglomerate Tcnm , a unit coeval with the youngest SWIT marine deposits Tcmu , and is gently folded similar to the oldest marine deposits Tcml , Tteb is likely to predate or be coeval with marine conditions on SWIT.
They interpreted this unit as a rhyolite dike crosscutting marine conglomerate. Our mapping indicates that these outcrops are moderately dipping deposits of the distinctive landslide breccia unit Tbx Fig. Breccia clasts of Tbx are identical to primary tuff outcrops of tuff of Ensenada Blanca Tteb. We also observe angular to subrounded cobbles and boulders of cemented Tbx reworked into overlying Tcml marine conglomerate beds Figs. If we take the ages from these units reported by Gastil et al.
Conflicting interpretations Gastil and Krummenacher, b ; Gastil et al. The basis for this conflict is in the disparity between the original K-Ar ages obtained by Gastil and Krummenacher b for the flow A younger K-Ar age of 4. This interpretation was rejected in Oskin and Stock a and it was instead proposed that the K-Ar age of We revisited this capping rhyodacite flow and nearby dikes to assess their relative age and emplacement mechanisms. The 3. The same dike is exposed along strike to the southwest, across arroyo 1, and continues up the eastern flank of Hast Pitzcal.
Here the dike thickens to 20—40 m wide, crosscuts marine and nonmarine conglomerate beds, and crops out continuously up to the base of the rhyodacite flow Figs.
The dike does not cut through the rhyodacite flow. Instead, we observe the subhorizontal rhyodacite flow as continuous and intact across this dike. In some exposures the dike exhibits continuous foliation with the overlying flow. Our observations confirm that the base of the flow is underlain by thin and discontinuous deposits of pyroclastic material unit Tcsp.
The majority of the rhyodacite of Cerro Starship consists of a lava flow. This is supported by the observation that both the matrix of the basal vitrophyre and that of the main flow deposit do not consist of ash or other pyroclastic material e. We also revisited the southern end of the main, elongate flow body Fig. We do not observe dikes exposed along the continuous, undulatory upper flow surface Fig.
In summary, our observations concur with the interpretation in Oskin and Stock a that the dike Tcsd fed the capping rhyodacite flow Tcsf. As additional support of this hypothesis, we find that basal pyroclastic deposits Tcsp are limited to an area immediately south of the point where this dike intersects the flow Fig.
To resolve disparate interpretations of the age of the rhyodacite, we dated a sample from basal Tcsf, away from any mapped dikes, where its sharp subhorizontal basal contact is well exposed Fig. Our ages of 3. Altogether, these data confirm that the K-Ar age of A latest Miocene to Pliocene interpretation for the age of the SWIT marine basin has important regional implications for the tectonic evolution of the Gulf of California rift and marine incursion into it.
Below, we discuss these implications and develop a paleogeographic reconstruction of a narrow marine embayment that formed during the earliest stages of oblique opening of the northern Gulf of California rift.
This integrated analysis illustrates the relationship between marine basin formation and the geodynamic evolution of this portion of the Pacific—North America plate boundary.
Throughout this region Fig. This late Miocene marine incursion Fig. Marine incursion into the northern Gulf of California and Salton Trough appears to have shortly followed the ca. The question of whether latest Miocene marine incursion was preceded by an earlier, middle Miocene marine seaway remains controversial.
At most of these localities, the middle Miocene microfossils were considered to be reworked and not in situ specimens. For example, both middle Miocene and Cretaceous microfossils were observed in Cerro Prieto geothermal wells Cotton and Vonder Haar, , suggesting that the deposits hosting these specimens likely postdate arrival of the Colorado River, which is dated as 5.
However, Winker discounted the in situ status of these specimens due to the unavailability of samples and necessary documentation. In addition, Gastil and Krummenacher b, p. No such marine rocks occur in mountain ranges that flank the modern-day Sonora coastline, despite evidence for significant faulting, tilting, and basin formation ca.
This contradicts other published studies that found that transtensional marine basins in the western Gulf rift began opening ca. This interpretation is at odds with onshore evidence that significant rift-related faulting in the region did not begin until after The existence of a middle Miocene marine basin of this extent also contradicts onshore timing constraints for the opening of the northern Gulf of California, that restore the modern shorelines and outcrops of several extensive late Miocene ash-flow tuffs to close proximity ca.
Correlation of these tuffs across the Gulf of California is based on their unique paleomagnetic remanence vector directions Lewis and Stock, ; Stock et al. Their original proximity was not based on paleolatitude interpretations from paleomagnetic data, but rather the thickness and outcrop distribution of multiple tuff markers and correlation of internal cooling units Oskin and Stock, b.
This cross-gulf tuff correlation is robust and must be considered in any discussion regarding the tectonic evolution of, or marine seaway incursion into, the northern Gulf of California. A middle Miocene interpretation for basal sediments in the offshore PEMEX wells is also at odds with abundant evidence for a late Miocene to Pliocene age of earliest marine deposits in numerous locations around the northern Gulf of California e. To reconcile the paleontological results with onshore geologic constraints, we suggest that the specimens asserted to be in situ by Helenes et al.
Age determinations of sediments bearing fossils of mixed ages must always be based on the youngest fossils present, because of the possibility that the sediments contain older reworked fossils.
Application of this age to basal marine deposits unit A that also contain this species in the Altar basin Pacheco et al. Moreover, unit A where documented by Pacheco et al. Although the first appearance of S. Honoring age constraints from correlative tuffs Oskin et al. In order to avoid substantial overlap of relatively thick continental crust i. In addition, an exploration well drilled on the outboard edge of the Puertecitos shelf location L in Fig.
Although we contend that in situ middle Miocene microfossils have not been convincingly documented from marine deposits in or adjacent to the Gulf of California, the presence of reworked middle Miocene marine microfossils in the Gulf of California and Salton Trough McDougall, suggests that marine conditions existed somewhere in the region during middle Miocene time.
Here we explore three potential origins of these reworked microfossils, and suggest that they may have been eroded from middle Miocene strata either within or beyond the Gulf of California rift. Such a seaway likely would have been considerably smaller and more discontinuous than the well-documented late Miocene seaway blue squares in Fig.
A marine basin in this location could have connected to the Pacific Ocean via a seaway that transected the Baja California peninsula e. However, upon restoration of the Baja California peninsula e.
A southern connection to the Pacific is even farther, and also unlikely because marine deposition in the southern Gulf of California occurred later, ca. If a middle Miocene basin or embayment existed, portions of it that contained middle Miocene microfossils could have been exposed and eroded during subsequent rift-related faulting, contributing specimens to the late Miocene basins. However, geologic constraints that tie the margins of the Gulf of California into close proximity at late Miocene time Oskin et al.
Another possible source may have been a shallow seaway that may have existed in an area of middle Miocene backarc extension located northeast of and parallel to the present-day Gulf of California, in Sonora and Sinaloa Fig.
Such a middle Miocene marine backarc basin would have existed northeast of a northwest-southeast chain of middle Miocene volcanic centers, in the wake of the southwestward-migrating volcanic arc Karig and Jensky, ; Hausback, ; Dorsey and Burns, ; Ferrari et al.
A marine backarc basin in this location would require similar connections to the Pacific Ocean, as discussed herein. Subsequent uplift and exposure of these marine deposits could have eroded and transported middle Miocene microfossils into the latest Miocene marine basins in the northern Gulf of California via roughly southwest-directed drainages Fig.
It is important that no outcrops of middle Miocene marine strata are known anywhere in western Mexico, despite the documented presence of nonmarine strata of the appropriate age Herman and Gans, ; Darin, ; Bennett et al.
This does not support the concept of a middle Miocene seaway, whether located beneath the modern-day Gulf of California or in a middle Miocene backarc basin. Regionally extensive outcrops of the A third possible source for reworked middle Miocene microfossils is the continental shelf on the Pacific side of the southernmost Baja California peninsula, adjacent to the mouth of the Gulf of California Fig.
Middle Miocene marine strata in this area were exposed to wave-base erosion after deposition Brothers et al. Further work is required to support this long-distance source, such as documentation of microfossils species at this possible source location.
Synchronous late Miocene marine incursion along the Pacific—North America plate boundary is a critical event in the tectonic and paleogeographic evolution of the Gulf of California rift. A major geodynamic change must have occurred for this region to evolve from a subaerial volcanic arc at the end of middle Miocene time Hausback, ; Dorsey and Burns, ; Umhoefer et al. Such a substantial modification of regional topography implies significant crustal thinning and related subsidence.
The Gulf of California shear zone created a series of fault-bounded nonmarine transtensional basins ca. Flooding of marine waters into the northern Gulf of California may have been delayed by as much as 1 m. Thus the formation of a proto—Gulf of California seaway likely was intimately linked, in time and space, to the onset of focused oblique rifting, which became localized in the northern Gulf of California region near the end of Miocene time Oskin et al.
Post—6 Ma Pacific—North America relative plate motion subsequently localized into the core of the Gulf of California shear zone, rifted apart these regions of nonmarine transtensional basins, and widened the Gulf of California seaway e. Isotopic and biostratigraphic data conclusively show that marine strata on SWIT range in age from late Miocene to early Pliocene. The timing of earliest marine deposition is constrained by the 6. Marine sedimentation in the study area ceased prior to subaerial eruption of the 3.
These age constraints are consistent with the microfossil assemblage from the same section that independently yielded an age of 6. New detailed geologic mapping and modern geochronologic data demonstrate that previous interpretations for the age and depositional environment of marine deposits on SWIT were based on incorrect K-Ar results and misinterpretation of field relationships, in part due to the structural complexity of the area.
The previously proposed upper age constraint of The Our results are at odds with a study that proposed widespread middle Miocene deposits in offshore marine basins Helenes et al. A late Miocene age for the SWIT marine basin is consistent with regional marine incursion into the northern proto—Gulf of California no earlier than 6. Middle Miocene microfossils documented at some locations in the Gulf of California and Salton Trough region are unlikely to be in situ.
Rather, these specimens probably are reworked into late Miocene strata from older deposits, as documented in several onshore exposures e. The locations of sources for middle Miocene microfossils remain unclear and the subject of ongoing speculation. We suggest that the most likely sources for reworked middle Miocene microfossils are 1 a marine extensional backarc basin in the northwestern Mexican Basin and Range province for which outcrops are not known, or 2 middle Miocene marine deposits along the western continental shelf of southernmost Baja California that were exposed to wave-base erosion during late Miocene time Brothers et al.
Late Miocene marine incursion into the northern Gulf of California resulted from focused oblique extension, crustal thinning, and resulting tectonic subsidence. This narrow, late Miocene seaway was colocated with, and flooded immediately following the development of, the Gulf of California shear zone, a narrow belt of localized strike-slip faulting, clockwise block rotation, and formation of pull-apart basins Bennett and Oskin, Resolution of the controversy over the age of oldest marine deposits on SWIT provides a critical step forward in a long-standing debate over the age of marine sedimentation in the northern Gulf of California.
Improved age constraints on marine deposition provide important new insights into late Miocene focused deformation related to localization of the Pacific—North America transtensional plate boundary in the northern Gulf of California and Salton Trough region.
Tom Donovan, Gregory Smart, and the Prescott College Kino Bay Center for Cultural and Ecological Studies staff provided incredible logistical support and comfortable accommodations while conducting field work. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.
Active faults and spreading centers are red; inactive are green. Active pull-apart basins are dark gray; inactive are light gray. Pliocene—Quaternary hybrid oceanic crust and sediments are solid purple. Basaltic ocean crust is purple with gray normal polarity magnetic seafloor anomalies. SAF—San Andreas fault. Cross-gulf tie points: P—Poway conglomerate Abbott and Smith, ; F—fusulinid-rich clast conglomerate Gastil et al. Geology was compiled from Gastil and Krummenacher, a ; Oskin, ; this study.
New and previously published geochronologic sample locations and isotopic ages shown. SS—sandstone; congl. S2 or the full-text article on www. Geochronologic and biostratigraphic ages are in Ma.
See text for descriptions and isotopic ages of units. A Typical exposure of monolithologic volcanic breccia Tbxv. Photograph was taken in the upper reaches of arroyo 4; hammer is 38 cm long. B Exposure of volcanic breccia Tbxv in erosional window in arroyo 3, with print out of field photograph provided by J.
Matching clasts verify that this is the identical exposure visited by Smith et al. Note the matching caliche precipitate on vertical face of clast in upper left corner. Arrow on card is 10 cm long. See geologic map Fig. C View of andesitic breccia clast sampled by Smith et al. Relatively fresh, angular surfaces of clast in foreground are from sample collection by Smith et al. After photograph was taken, a portion of this same breccia clast was sampled, for which we report an Area of Figure 4B is shown in red rectangle.
Tera-Wasserburg concordia diagrams are on left and age spectrum diagrams are on right. MSWD—mean square of weighted deviates. B Tuff of Oyster Amphitheater Ttoa interbedded within the marine basin.
C Tuffs of Hipat Mesa Tthm that cap the marine basin. D Rhyodacite of Cerro Starship Tcsf that caps the marine basin. Zircon crystals omitted from mean age calculation are gray squares with gray, dashed error ellipses. Relatively younger zircons have high uranium concentrations Supplemental Table 2 [see footnote 2 ] and are omitted from mean age due to possible lead loss. Relatively older zircons are omitted due to potential inheritance.
See Supplemental Table 2 see footnote 2 for analysis data. View of angular unconformity between pre-basin volcanic units Tthp, Ttua, Tbx and overlying marine deposits Tcmu , nonmarine sediments Tcnm , and volcanic units Tcsp, Tcsf see text.
Photograph taken looking northwest along eastern flank of Hast Pitzcal, west of arroyo 1. Circled geologists for scale. See Supplemental Table 3 see footnote 3 for analysis data. Gray shaded letters indicate steps excluded from the isochron age determination.
See Supplemental Table 4 see footnote 4 for analysis data. A Distinctive landslide breccia Tbx in sharp contact above unnamed airfall tuffs Ttua see text. Hammer for scale. Photograph from Neuhaus B Typical outcrop texture of distinctive landslide breccia deposit Tbx. Eutaxitic foliation of flattened pumice fiamme is visible in large breccia clasts, similar to intact source outcrops of the tuff of Ensenada Blanca Tteb. C Typical outcrop of polylithologic sedimentary breccia Tbxs that discontinuously underlies marine deposits in arroyos 1 and 4.
Photograph taken in the upper reaches of arroyo 4. Hammer is 38 cm long. Photograph is from Keogh A Panoramic view of the cove outcrops where the floor of the marine basin is well exposed circled person for scale. B Annotated interpretation of panorama in A.
Marine conglomerate deposits Tcml containing discontinuous exposures of the tuff of Arroyo Sauzal Ttas unconformably overlie the landslide breccia Tbx and pre-basin volcaniclastic conglomerate Tvc see text.
Three previously published geochronologic ages have been reported from this cove outcrop. Hammer in each photo is 38 cm long A Moderately to well-stratified lower conglomerate Tcml outcrops along coastline, northwest of Hast Pitzcal. B Well-stratified lower conglomerate Tcml outcrops in the cove Fig. C Fossiliferous lower conglomerate Tcml outcrops on the western flank of Hast Pitzcal. Black rectangle shows enlargement in F. D Poorly stratified lower conglomerate Tcml outcrops on the western bank of arroyo 2.
Clast composition is dominated by locally derived arc-related volcanic rocks with occasional clasts of the distinctive landslide breccia deposit Tbx. E Boulder in lower conglomerate Tcml encrusted with possible remnants of late Miocene barnacles. F Enlargement of rectangle area in C showing a pair of oyster fossils. G Outcrop of tuffaceous marine sandstone Tsmt deposits on the eastern bank of arroyo 2.
Lower massive bed contains occasional layers of conglomerate stringers and marine fossils. Upper, well-stratified part contains fossiliferous sandstone and sandy conglomerate beds. Hammer is circled. H Large late Miocene gastropod fossil in Tsmt conglomerate bed.
I Subvertical burrow arrows in tectonically inclined Tsmt sandstone and conglomerate beds. Hammer for scale in B and C is 38 cm long. A An inverse graded Tcmu conglomerate bed on the western bank of arroyo 4. Bed inclination is primary depositional dip. B Inclined Tcmu conglomerate beds on the western bank of arroyo 4 containing multiple trace fossils. Black rectangle shows enlargement in C. C Enlargement of rectangle area in B showing multiple burrows in Tcmu sandy conglomerate bed. Subvertical burrows black arrows are oriented orthogonal to modern-day horizontal, not orthogonal to bedding.
D Panoramic view of Tcnm nonmarine conglomerate and sandstone deposits and Tcmu outcrops in arroyo 4. Looking west at exposure where subhorizontal, nonmarine topset beds Tcnm laterally grade into inclined, marine foreset beds Tcmu. Circled person for scale. E Annotated interpretation of panorama in D.
Hammer for scale in A and B is 38 cm long. A Moderately stratified exposure of Tcnm. B Tcnm contains subrounded to angular clasts of locally derived, middle Miocene arc-related volcanic rocks and late Miocene welded tuffs.
C Panoramic view of northeastern portion of study area, looking northeast at arroyo 5 where Tcnm deposits are perched upon and banked against paleotopographic highs of pre-basin, arc-related volcanic rocks. The top of the basin is visible as a subhorizontal geomorphic bench that dips gently toward the Gulf of California. D Annotated interpretation of panorama in C cong. Hammer for scale in B and D is 38 cm long.
A Looking northwest at Hast Pitzcal. B Ash- and pumice-rich pyroclastic deposits Tcsp locally underlie the capping rhyodacite of Cerro Starship Tcsf lava flow. C Looking north along undulatory upper surface of Hast Pitzcal. No crosscutting dikes are observed along crest of hilltop. Circled geologist for scale. D Centimeter-scale flow banding in porphyritic rhyodacite. Maroon object in center of photo is loose clast. E Flow banding in rhyodacite outcrops Tcsf shows highly variable patterns of foliation, suggesting viscous flow prior to cooling.
F Looking southwest on the eastern flank of Hast Pitzcal at subvertical foliation in feeder dikes Tcsd. Foliation in Tcsd is parallel to outcrop pattern of dike outcrops. Dike leads directly to base of capping rhyodacite of Cerro Starship Tcsf lava flow. G Looking north from southwestern corner of study area. H Looking north at southernmost tip of main flow body of Tcsf. Base of rhyodacite flow consists of 2—4-m-thick black glassy vitrophyre.
Geologist is standing on middle Miocene arc-related basalt flows. Basal contact of Tcsf white arrows is continuous and sharp. No crosscutting dikes are observed. White region in upper left portion of photo is recent landslide scar. Map unit colors and other symbols as in Figure 2. Units shaded lighter are above the ground surface. Dip values from structural measurements used to construct cross sections are shown near ground surface.
0コメント