Normal faults in thrust sheets: pre-orogenic extension, post- orogenic extension, or both? Enrico Tavarnelli Centro di Geodinamica, Universita Á della Basilicata, Via Anzio, 85100 Potenza, Italy Received 28 January 1998; accepted 18 January 1999 Abstract In fold-and-thrust belts that experienced both pre-orogenic and post-orogenic extension, it may be dicult to establish whether observed normal faults pre-dated, post-dated, or were synchronous with thrusting. Geometrical structural patterns may be insucient to constrain the relative chronology of extensional and contractional deformations. The systematic use of kinematic criteria makes it possible to unequivocally de®ne the timing relationships of reverse and normal fault development, and hence to correctly unravel complex structural evolutions. Kinematic analysis in the southernmost Umbria±Marche Apennines of Italy, where both normal and thrust faults are present, revealed a history of repeated tectonic inversion, characterised by two distinct stages of extension separated by an episode of folding and thrusting. Structural overprinting relationships observed at thrust±normal fault intersections were useful for: (i) removing sequentially younger deformations; and hence (ii) separating and quantifying the eects of orogenic contraction from those of both pre-orogenic and post-orogenic extension. # 1999 Elsevier Science Ltd. All rights reserved. 1. Introduction Inversion tectonics applies to regions that experi- enced a switch in deformation regime from extension to compression (positive inversion) or from com- pression to extension (negative inversion: Williams et al., 1989). A common characteristic of inverted regions is the modi®cation of their previous architecture by newly formed structures (Butler, 1989; Hayward and Graham, 1989). The degree of modi®cation may vary greatly depending upon scale, strain rate and intensity of deformation. Should deformation occur under brittle to semi-brittle conditions, pre-existing fault sur- faces may either be truncated by, or reactivated as, younger faults (Welbon, 1988; Faccenna et al., 1995; McClay, 1995). Because of their relatively long history, orogenic belts commonly bear the signature of at least one epi- sode of positive inversion, from early passive-margin or continental rift extension to late collision (e.g. the French Western Alps: Gillcrist et al., 1987; the Spanish Pyrenees: Bond and McClay, 1995). Orogenic belts emplaced in passive-margins may also experience post- orogenic extension (e.g. the Northern Apennines: Carmignani et al., 1994; the North American Cordillera: Constenius, 1996), so that it is often di- cult to establish whether observed normal faults either pre-dated or post-dated the thrusts (O'Dea and Lister, 1995). Yet, the eects of pre-thrust extension must be separated from those of post-orogenic extension, because a correct de®nition of the age of normal faults relative to thrusts can provide important constraints for quantitative estimates in cross-section restoration. Reliable criteria are needed to enable us to do so. In multiply inverted settings, thrust sheets are often aected by hinterland-dipping normal faults, which terminate downwards against the thrust surfaces. In the absence of stratigraphic data, this geometrical pat- tern alone does not constrain the relative chronology of thrusting and normal faulting, as two distinct in- terpretations can be made: (i) thrust faults are older and are partly reactivated as low-angle normal faults (Fig. 1a); or (ii) thrust faults are younger, and propa- Journal of Structural Geology 21 (1999) 1011±1018 0191-8141/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S0191-8141(99)00034-6 www.elsevier.nl/locate/jstrugeo E-mail address: tavarnelli@unibas.it (E. Tavarnelli)