Late Cenozoic bending of the Bolivian Andes: New paleomagnetic and kinematic constraints

New paleomagnetic measurements of essentially undeformed Late Cenozoic volcanic rocks in the Bolivian Andes, South America, constrain rigid body rotations about vertical axes during the last ∼13 Myr on both limbs of the Bolivian orocline in the Eastern Cordillera. Thermal and alternating field demagnetization was carried out on samples from 52 sites in three major volcanic complexes: (1) 13–2 Ma Los Frailes ignimbritic volcanics, outcropping at ∼19.5°S; (2) 9–5 Ma Morococala ignimbritic volcanics, outcropping at ∼18°S; and (3) 13–5 Ma shoshonitic to acidic lavas and intrusives, outcropping at ∼17.5°S. Well‐defined magnetic components were isolated, which are interpreted to represent the Earth's magnetic field at the time of volcanic activity. The mean magnetic vector for site groupings suggests regional tectonic rotations about vertical axes, with respect to stable South America, of 10° ± 8 clockwise for the 13–2 Ma Los Frailes volcanic complex (8° ± 9° clockwise for only the Miocene (13–5 Ma) Los Frailes volcanics), and 1° ± 18° anticlockwise for the combined 13–5 Ma Morococala and Eucalyptus volcanic complexes. These data are consistent with observed shortening gradients on the eastern margin of the Bolivian Andes, in the sub‐Andean zone, suggesting rotation and shortening are synchronous. A joint inversion for both tectonic rotation and the amount of shortening, assuming a linear variation in the amount of rotation along the length of the southern limb of the Bolivian orocline, shows that rotation of the Eastern Cordillera in the last 10–15 Myr, relative to stable South America, varies from zero at the oroclinal hinge, at ∼18°S, to a maximum of ∼13.5° clockwise at ∼22°S, decreasing to zero, south of 23°S, with an average rotation in the range 5° to 10° clockwise. Concomitant with this, shortening in the sub‐Andean zone decreases from a maximum of ∼86 km at the oroclinal hinge, to ∼47 km at 22°S, and then ∼33 km at 23°S. The lack of evidence for large‐scale internal faulting in the Eastern Cordillera since ∼9 Ma suggests either bending of the Eastern Cordillera was accommodated by more distributed small‐scale faulting or that relative rotation here mainly occurred prior to ∼9 Ma. The small average rotation of the Eastern Cordillera, on the southern limb of the orocline, is consistent with only very small rotation (<10°) of the forearc region in northern Chile during the last ∼20 Myr. Rotation of the northern limb is less well constrained, but it is also likely to be small (<5°) and anticlockwise.