Multichannel seismic transects reveal an ∼2-km-thick, ∼50 × 100 km evaporite body under the shelf on the eastern margin of the Guaymas Basin, central Gulf of California (Mexico). These thick newly discovered evaporites appear to be correlated with well-known gypsum beds near Santa Rosalía to the northwest, on the Baja California peninsula. Closing the Gulf of California along kinematic flow lines suggests that the thin, scattered, ca. 7 Ma Santa Rosalía gypsum beds formed on the fringe of the much thicker evaporite deposit. This correlation, and the large volume of the Guaymas evaporates, implies that substantial marine incursions and subsequent evaporite deposition occurred during the Late Miocene and prior to lithospheric rupture. Furthermore, the shape of the Guaymas evaporite is indicative of a transtensional basin, suggesting that oblique extension existed in the central Gulf of California ca. 7 Ma.


The Gulf of California (Mexico) is part of a young rift system that occupies the boundary between the Pacific and North American plates. In the modern Gulf of California, ∼92% of Pacific–North America relative motion is accommodated at short seafloor-spreading centers offset by long transform faults (Fig. 1; DeMets and Dixon, 1999; Dixon et al., 2000). These spreading centers are perpendicular to the direction of Pacific–North America relative motion, but they are oblique, by ∼30°, to the rifted margins of the Gulf of California. This obliquity suggests that there was a fundamental change in how extension was accommodated within the region between ca. 12 Ma, when continental extension accelerated as subduction stalled offshore (Atwater and Stock, 1998), and lithospheric rupture and the onset of seafloor spreading ca. 6.3 Ma (Oskin and Stock, 2003).

Two distinct models have been proposed for how Pacific–North America relative motion was accommodated in the Gulf of California region prior to lithospheric rupture. One model argues that early strain was partitioned between NNW-trending dextral strike-slip faults west of the modern Baja California peninsula and NNW-striking normal faults within the proto-gulf (e.g., Oskin and Stock, 2003). A second model argues that, prior to rupture, extension was accommodated by transtensional deformation across the proto-gulf region, with a smaller amount of strike-slip motion west of the peninsula (Fletcher et al., 2007; Sutherland et al., 2012; Bennett et al., 2012a). In this model, basins within the proto-gulf have morphologies similar to the rhomboid-shaped pull-apart basins of the modern Salton Trough region, where Pacific–North America relative motion is accommodated by basin-scale strain partitioning and wrench tectonics (e.g., Axen and Fletcher, 1998). These two kinematic models predict different patterns of faulting in the proto-gulf, and imaging of the early extensional basins under the thick sediments of the eastern margin of the Gulf of California can improve our understanding of early rift evolution in the gulf.

We present multichannel seismic (MCS) data from the 2002 R/V Maurice Ewing cruise EW0210 (part of the PESCADOR seismic experiment) that cross the eastern Guaymas Basin (Fig. 1). These data image a unit that we interpret as an ∼2-km-thick evaporite body that formed on extended continental crust during the Late Miocene. Thick evaporites have not been previously documented in the Gulf of California. In rifts, evaporites typically form under conditions unique to the latest stages of continental rupture and the onset of seafloor spreading (Evans, 1978). Thus, the presence, age, thickness, and shape of the evaporite in the Guaymas Basin place new constraints on the history of marine incursions and early basin development during this transitional stage in the Gulf of California.


A prominent feature of stacked MCS data from the eastern Guaymas Basin is a bright undulating reflector at ∼1.5–2.0 s two-way traveltime (TWT) (Fig. 2; see the GSA Data Repository1). We interpret this reflector as the top of a thick evaporite unit, which we refer to as the east Guaymas evaporite (EGE). This interpretation is based on the similarity of features in these data to the structure and seismic character of well-studied salt bodies, such as Late Miocene Mediterranean evaporites (e.g., Fiduk, 2009). The polarity of the high-amplitude reflection from the top of salt is consistent with evaporites underlying sediments. The salt body is seismically transparent, with a base defined by an irregular reflection at ∼3 s TWT. A seismic velocity model across the Guaymas Basin (Lizarralde et al., 2007) resolves the imaged evaporite body as a region with velocities of ∼4.0–4.5 km/s, consistent with halite or gypsum (e.g., Sharma, 1997). Using these velocities, we estimate that the evaporite is ∼2 km thick.

Deformation in the EGE is consistent with models and observations of salt flowing seaward on a subsiding margin (e.g., Brun and Fort, 2011). To the southeast, the evaporite is overlain by an ∼30-km-wide basin, and the salt is thinned and diapiric, features indicative of salt withdrawal (Vendeville, 1992). To the northwest, the evaporite is thickened by gentle folds, indicating that the salt was compressed as it flowed downslope from beneath the withdrawal basin.

Sediments overlying the EGE can be divided into upper and lower sections, separated by an angular unconformity that marks the onset of salt deformation. In the ∼750-m-thick lower section, folds are parallel to the top of the evaporite, indicating that these sediments were deposited on a flat evaporite surface, and the salt began to flow seaward later, presumably after margin subsidence and sediment load reached a critical point. The upper sedimentary section is characterized by mostly undeformed sediments that onlap the folded lower section surface. The dip in these beds decreases upsection, indicating that deposition of the upper section was synchronous with the seaward flow of the salt. These sediments are as thick as ∼750 m in the center of the withdrawal basin and pinch out against topographic highs in the lower sediments. Where the salt is thickest (180–190 km; Fig. 2), this upper section is absent and the lower sediments are pushed up to the seafloor. At the southeastern edge of the withdrawal basin, beds are offset by normal faults and form a roll-over anticline, a feature of syntectonic sedimentation. These faults offset the seafloor, indicating that salt withdrawal is ongoing. The same thickness of sediment appears to have been deposited seaward of the salt, but these units were undisturbed by salt flow and thus lack the angular unconformity that separates the units over the evaporite.

The areal extent of the EGE is constrained by the edge of salt on four seismic lines that cross the unit and by gravity data. The imaged salt body is coincident with a low free-air anomaly observed in shipboard gravity profiles along the MCS lines (Fig. 2). A low in altimetry-derived gravity appears to be the same feature imaged by the MCS data (Fig. 3A). This ∼50 × 100 km anomaly traces a north-south–trending western margin, which correlates with the western evaporite edge revealed by the MCS data, and a northeast-southwest–trending southeastern margin. To the north and south, the anomaly and imaged evaporite terminate against the North Guaymas and Carmen Fracture Zones, respectively (Lines 26 and 1183–1184; see the Data Repository). Near its center and near the South Guaymas Fracture Zone, the basin appears to step to the southeast.

The EGE is on extended continental crust with rift-related, magmatic intrusions (Lizarralde et al., 2007). The MCS data show that the seaward (northwestern) edge of the evaporite is abrupt and steeply dipping. Seaward of this edge, the basement reflector is bright and rough, suggestive of extrusive igneous rocks. A seaward-dipping reflector sequence is present just seaward of this rough basement. Seaward-dipping reflectors are commonly attributed to subaerial lava flows near the continent-ocean transition of volcanic rifted margins (e.g., White and McKenzie, 1989). In the seismic velocity model, a lateral change in lower crustal seismic velocities from continental (intermediate) to oceanic (mafic) is centered below the seaward-dipping reflectors, suggesting that the edge of the evaporate is near the continent-ocean transition and the onset of new igneous crustal production.


Scattered gypsum beds are well documented to the northwest near Santa Rosalía, on the Baja California peninsula (Fig. 1). On the southern end of Isla San Marcos, a continuous outcrop of gypsum, ∼2 × 4 km in areal extent, is exposed by one of the world’s largest gypsum mines (Ochoa-Landín et al., 2000; Founie, 2007). In the nearby Santa Rosalía Basin, scattered gypsum beds extend over an ∼10 × 30 km region and locally reach thicknesses of as much as ∼70 m (Wilson and Rocha, 1955).

It is likely that the Santa Rosalía evaporites formed at the same time and place as the EGE. These gypsum beds reconstruct to the southern edge of the EGE basin if the Guaymas rift segments and the Carmen segment are closed by 280 ± 20 km along an azimuth of 306°, parallel to the modern spreading direction and bounding transform faults and fracture zones (Fig. 3). The ±20 km uncertainty is the half-width of the EGE border along the Carmen Fracture Zone. This reconstruction places the relatively thin northwestern evaporites at the edge of the more extensive, ∼2-km-thick EGE. The relative thickness of the two units makes sense if the Santa Rosalía evaporites formed along the fringe of the larger EGE basin, as suggested by the reconstruction.

The ∼280 km displacement vector implied by the correlation of the EGE with gypsum units near Santa Rosalía is in close agreement to two independent estimates of post-rupture extension, implying that the reconstruction represents a time just before lithospheric rupture and the onset of seafloor spreading. Using seismic refraction data, Lizarralde et al. (2007) estimated ∼280 km of new igneous crust formation along the modern spreading direction in the northern Guaymas rift segment. Oskin and Stock (2003) correlated and dated volcanic tuffs in the northern Gulf of California and estimated 276 ± 13 km of post-rupture separation along an azimuth of 315° across the Upper Delfín Basin–Tiburón Basin rift segment since ca. 6.3 Ma. Similar magnitude, post-rupture displacements of ∼275–310 km have been suggested for the southern Gulf of California (Fig. 1; Fletcher et al., 2007; Sutherland et al., 2012).

The Santa Rosalía gypsums are generally underlain by and locally interbedded with limestones (Wilson and Rocha, 1955) that contain Late Miocene (11.2–5.3 Ma) marine fauna (Ortlieb and Colletta, 1984). Holt et al. (2000) concluded that these marine rocks were deposited by 6.93–7.09 Ma, based on an 40Ar-39Ar age of 6.76 ± 0.9 Ma for the Cinta Colorada, an andesitic tuff that overlies the limestone and gypsum units, and magnetostratigraphy. This well-constrained age for the Santa Rosalía gypsums suggests that the EGE also formed during the Late Miocene, beginning ca. 7 Ma.


Evaporite volume is a function of source-water salinity, evaporation efficiency, and deposition duration. Complete evaporation of a 100-m-deep column of typical ocean water (∼3.5 wt% dissolved salt) produces a total evaporite thickness of ∼1.7 m (Warren, 2006). Thus, ∼115 km of seawater is needed to produce the ∼2-km-thick EGE. This is a substantial amount of seawater, and the required amount would be even greater if the brine water was less saline than seawater. For this reason, marine-fed brines have been invoked as the source of salt deposits of similar thickness (Warren, 2006), such as those in the Mediterranean (e.g., Hsü et al., 1973) and the Atlantic (e.g., Evans, 1978; Jackson and Cramez, 2000). The observed 2 km of salt could have been deposited in as few as ∼57–115 k.y., assuming an initial brine with the salinity of seawater and modern subtropical net evaporation rates of 1–2 m/yr (e.g., Schanze et al., 2010). We thus infer that the EGE resulted from substantial, repeated marine incursions, and, based on the correlation with the Santa Rosalía gypsum units, we infer that these incursions began just prior to ca. 7 Ma, ∼0.5–1.0 m.y. before the earliest marine incursions in the northern Gulf of California (e.g., McDougall, 2008; Dorsey et al., 2011; Bennett et al., 2012b).

The Late Miocene rocks of the Santa Rosalía Basin inform us of conditions during deposition of the associated EGE. This basin is floored by scattered marine limestone and gypsum units that unconformably overlie heavily faulted pre-rift Comondu arc volcanic rocks (Wilson and Rocha, 1955), indicating that these units represent the first deposition in proto-gulf extensional basins. Color banding in the gypsum units suggests that brine chemistry varied through time, and the gypsum is interbedded with thin clastic horizons (Ochoa-Landín et al., 2000), suggesting that sediment supply also varied. Both the basal marine limestone and gypsum are overlain by a sequence of prograding fan-delta deposits that formed during repeated periods of basin-floor subsidence, a result of synrift extensional faulting (Wilson and Rocha, 1955; Ochoa-Landín et al., 2000). These sequences suggest that a series of marine incursions flooded the Santa Rosalía Basin through the end of the Late Miocene into the Pleistocene (Ochoa-Landín et al., 2000). Large positive values of δ34S in the gypsum units (Ortlieb and Colletta, 1984) have been interpreted as evidence that they formed by evaporation of seawater in isolated basins (Ochoa-Landín et al., 2000). A similar setting has been invoked to explain the deposition of thin gypsum deposits in Late Miocene basins of the northern Gulf of California (Escalona-Alcázar et al., 2001). These observations and interpretations suggest that the EGE formed in an isolated basin supplied by a series of marine flooding events.

In the proto-gulf, extensional tectonics and volcanic ridges would have created elongate basins with restricted access to inundating seawater, conditions favorable for sustaining hypersaline brines throughout the deposition of thick evaporites (e.g., Warren, 2006). A modern analogue of such a setting is the Danakil Depression (Afar triangle, Africa; Orszag-Sperber et al., 1998; Jackson and Cramez, 2000), where marine water from the Red Sea seeps through and/or periodically spills over an emerging horst, supplying brine water to fault- and lava flow-bounded basins (Manighetti et al., 1997; Jackson and Cramez, 2000). In these isolated basins, the resupply of saline water is outpaced by evaporation, enabling brines to maintain saturation and precipitate thick salt deposits (Friedman, 1972). These Afar evaporites are forming directly on proto-oceanic crust in the latest stages of a transition from continental extension to seafloor spreading, similar to the rift stage suggested for formation of the EGE by the proximity of seaward-dipping reflectors (Fig. 2) and fast lower crustal velocities. In the Afar triangle, heat supplied by rift-related magmatism aides in evaporation from brines, as may have been the case during magma-rich rifting of the Guaymas Basin (Lizarralde et al., 2007).


The margins of the evaporite likely follow the orientation of basin-bounding faults, providing constraints on models of strain partitioning during the earliest stages of rifting. The interpreted western edge of the EGE basin trends north-south, implying that, without regional rotation, early basin-forming faults along this edge also trended approximately north-south, and that some component of proto-gulf extension was oriented east-west (Fig. 3). The southeastern edge of the evaporite trends northeast-southwest. This edge is colocated with a 5-km-deep escarpment in the crustal velocity model (see the Data Repository). This escarpment also appears to trend northeast-southwest in the gravity data, implying that a significant component of early extension was oriented parallel to the modern, northwest-southeast spreading direction. Furthermore, the EGE steps to the southeast near the South Guaymas Fracture Zone, creating an S-shaped outline that is similar to the shape of a salt deposit in the transtensional Laguna Salada Basin of the southern Salton Trough (e.g., Axen and Fletcher, 1998). Together, these observations suggest that early extension, as it localized within the Gulf of California ca. 7 Ma, was partitioned onto both north-south– and northeast-southwest–striking normal faults within transtensional basins. This conclusion is consistent with the model Bennett et al. (2012a) proposed to explain transtensional deformation in coastal Sonora that accelerated ca. 6.5 Ma.


Seismic data reveal a large evaporate unit that formed during repeated marine incursions into the Late Miocene proto-Gulf of California. Closing the gulf along northwest-striking fracture zones places evaporites near Santa Rosalía along the southern edge of the EGE, suggesting that these gypsum units formed along the fringe of the much larger EGE and that the evaporites were deposited prior to lithospheric rupture. The shape and size of the basin, and our inferred age for the evaporite, support a kinematic model in which a significant portion of Pacific–North America relative motion is accommodated by transtensional shearing along the eastern proto-Gulf of California ca. 7 Ma.

We are grateful to Cathy Busby and Paul Umhoeffer for discussions about the Santa Rosalía gypsum units and to three anonymous reviewers for their critiques of this paper. This work was funded by a grant from the U.S. National Science Foundation MARGINS program.

1GSA Data Repository item 2013070, Figures DR1–DR4 (high-resolution plots and interpretations of seismic data), Figure DR5 (cross section and ca. 7 Ma reconstruction), and Table DR1 (locations and ages of earliest marine sediments), is available online at www.geosociety.org/pubs/ft2013.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.