The San Jacinto right-lateral strike-slip fault zone is crucial for understanding plate-boundary dynamics, regional slip partitioning, and seismic hazards within the San Andreas fault system of southern California, yet its age of initiation and long-term average slip rate are controversial. This synthesis of prior and new detailed studies in the western Salton Trough documents initiation of structural segments of the San Jacinto fault zone at or slightly before the 1.07-Ma base of the Jaramillo subchron. The dextral faults changed again after ca. 0.5–0.6 Ma with creation of new fault segments and folds. There were major and widespread basinal changes in the early Pleistocene when these new faults cut across the older West Salton detachment fault. We mapped and analyzed the complex fault mesh, identified structural segment boundaries along the Clark, Coyote Creek, and San Felipe fault zones, documented linkages between the major dextral faults, identified previously unknown active strands of the Coyote Creek fault 5 and 8 km NE and SW of its central strands, and showed that prior analyses of these fault zones oversimplify their complexity. The Clark fault is a zone of widely distributed faulting and folding SE of the Santa Rosa Mountains and unequivocally continues 20–25 km SE of its previously inferred termination point to the San Felipe Hills. There the Clark fault zone has been deforming basinal deposits at an average dextral slip rate of ≥10.2 +6.9/−3.3 mm/yr for ~0.5–0.6 m.y. Five new estimates of displacement are developed here using offset successions of crystalline rocks, distinctive marker beds in the late Cenozoic basin fill, analysis of strike-slip–related fault-bend folds, quantification of strain in folds at the tips of dextral faults, and gravity, magnetic, and geomorphic data sets. Together these show far greater right slip across the Clark fault than across either the San Felipe or Coyote Creek faults, despite the Clark fault becoming “hidden” in basinal deposits at its SE end as strain disperses onto a myriad of smaller faults, strike-slip ramps and flats, transrotational systems of cross faults with strongly domain patterns, and a variety of fault-fold sets. Together the Clark and Buck Ridge–Santa Rosa faults accumulated ~16.8 +3.7/−6.0 km of right separation in their lifetime near Clark Lake. The Coyote Ridge segment of the Coyote Creek fault accumulated ~3.5 ± 1.3 km since roughly 0.8–0.9 Ma. The San Felipe fault accumulated between 4 and 12.4 km (~6.5 km preferred) of right slip on its central strands in the past 1.1–1.3 Ma at Yaqui and Pinyon ridges. Combining the estimates of displacement with ages of fault initiation indicates a lifetime geologic slip rate of 20.1 +6.4/−9.8 mm/yr across the San Jacinto fault zone (sum of Clark, Buck Ridge, and Coyote Creek faults) and about ~5.4 +5.9/−1.4 mm/yr across the San Felipe fault zone at Yaqui and Pinyon ridges. The NW Coyote Creek fault has a lifetime slip rate of ~4.1 +1.9/−2.1 mm/yr, which is a quarter of that across the Clark fault (16.0 +4.5/−9.8 mm/yr) nearby. The San Felipe fault zone is not generally regarded as an active fault in the region, yet its lifetime slip rate exceeds those of the central and southern Elsinore and the Coyote Creek fault zones. The apparent lower slip rates across the San Felipe fault in the Holocene may reflect the transfer of strain to adjacent faults in order to bypass a contractional bend and step at Yaqui Ridge. The San Felipe, Coyote Creek, and Clark faults all show evidence of major structural adjustments after ca. 0.6–0.5 Ma, and redistribution of strain onto new right- and left-lateral faults and folds far removed from the older central fault strands. Active faults shifted their locus and main central strands by as much as 13 km in the middle Pleistocene. These changes modify the entire upper crust and were not localized in the thin sedimentary basin fill, which is only a few kilometers thick in most of the western Salton Trough. Steep microseismic alignments are well developed beneath most of the larger active faults and penetrate basement to the base of the seismogenic crust at 10–14 km. We hypothesize that the major structural and kinematic adjustments at ca. 0.5–0.6 Ma resulted in major changes in slip rate within the San Jacinto and San Felipe fault zones that are likely to explain the inconsistent slip rates determined from geologic (1–0.5 m.y.; this study), paleoseismic, and geodetic studies over different time intervals. The natural evolution of complex fault zones, cross faults, block rotation, and interactions within their broad damage zones might explain all the documented and implied temporal and spatial variation in slip rates. Co-variation of slip rates among the San Jacinto, San Felipe, and San Andreas faults, while possible, is not required by the available data. Together the San Jacinto and San Felipe fault zones have accommodated ~25.5 mm/yr since their inception in early Pleistocene time, and were therefore slightly faster than the southern San Andreas fault during the same time interval. If the westward transfer of plate motion continues in southern California, the southern San Andreas fault in the Salton Trough may change from being the main plate boundary fault to defining the eastern margin of the growing Sierra Nevada microplate, as implied by other workers.

The majority of the San Andreas fault zone is convergently oblique to relative plate motion. The commonness of transpression makes it significant for understanding deformation of the continental lithosphere. We have quantified the distribution of transpressional deformation along the San Andreas fault zone with respect to variations in boundary conditions along its length and distance from the fault zone itself. Rock uplift was used as a proxy for transpressional deformation. The pattern of exhumation along the fault was synthesized based on previously determined apatite fission-track and (U-Th)/He ages from 210 locations within 40 km of the fault trace. Patterns of mean elevation and slope in swaths along the fault were used as rough proxies of surface uplift and erosion. Relatively higher exhumation rates and mean elevations occur most commonly along the most oblique sections of the fault, such as in the Transverse Ranges. The highest rates of exhumation (>0.5 mm/yr) and highest and steepest topography also occur almost exclusively in the near field (i.e., within ∼10 km) of the fault trace. These trends are consistent with the strain-partitioning model of transpression, in which distributed deformation is concentrated in the fault zone and the degree of partitioning between simple and pure shear is a function of obliquity. However, the pattern of rock uplift also exhibits considerable variability. Neither the degree of obliquity nor the distance to the fault trace is enough to predict where high exhumation or mean elevation will occur. This suggests that heterogeneity in boundary conditions, including mechanical weaknesses and variations in erodibility, is equally important for controlling the pattern of transpressional deformation.