Structural and metamorphic analysis (Ring et al. 1999a, 2001, 2007; Ring and Layer 2003) shows that deformation can generally be divided into four main stages:
(1) Eocene and earliest Oligocene ~ESE-WNW-oriented nappe stacking (D1 and D2) associated with blueschist- and transitional blueschist/greenschist-facies metamorphism (M1 and M2). Maximum high-pressure assemblages in the Cycladic blueschist unit developed during the first deformational event, D1, and are therefore referred to as M1. The associated S1 foliation was porphyroblastically overgrown by glaucophane, chloritoid and kyanite during a static growth event. Internal imbrication of the nappes of the Cycladic blueschist unit under M1 peak high-pressure metamorphism occurred at ~44-37 Ma and followed and also was followed by high-pressure mineral growth (Ring et al. 1999a; Ring and Layer 2003). D2 caused emplacement of the Cycladic blueschist unit onto the Kerketas nappe, which started at ~35-30 Ma and eventually caused high-pressure metamorphism in the latter at 24-21 Ma (Ring et al. 2001). Maximum pressure in the Kerketas nappe occurred during the D2 deformation and therefore the mildly blueschist-facies event in the Kerketas nappe is regarded as M2. D2 thrusting was out-of-sequence and occurred during decompression bringing 18-20 kbar rocks of the Cycladic blueschist unit on top of 8-10 kbar rocks of the Kerketas nappe. During D2, the M1 high-pressure assemblages in the Cycladic blueschist unit were replaced by M2 transitional blueschist/greenschist facies assemblages. Data reported by Ring et al. (2007) showed that the deformation-related greenschist facies overprint in the Selçuk normal shear zone at the top of the Selçuk nappe occurred at the same time as the basal nappes of the Cycladic blueschist unit were thrust onto the foreland. As mentioned above, these data constrain that greenschist facies metamorphism in the uppermost Selçuk nappe can generally be grouped into the M2 event and occurred before 32 Ma.
(2) A subsequent history of Miocene horizontal crustal extension (D3). In the Ampelos nappe there is evidence that D3 proceeded before and after a greenschist facies metamorphism (M3). This greenschist facies metamorphic overprint was characterized by the prograde formation of garnet and more rarely by biotite in metapelite of the Agios Nikolaos and Ampelos nappes (Chen 1995). Chen (1995) estimated ~6-7 kbar and 450-490°C for M3 with slightly higher temperatures in the western than in the eastern part of the island. The data show that M3 occurred during further decompression but increasing temperature. This M3 greenschist facies event can not be related to the above mentioned >~32 Ma greenschist facies metamorphic event in the uppermost Selçuk nappe and must be a second greenschist facies event that can also be seen as a post high-pressure metamorphic event in the Kerketas nappe (characterized by the reaction diaspore to corundum during increasing temperature and decreasing pressure) and must therefore be younger than 21 Ma. Ductile flow during D3 was characterized by a high degree of coaxial deformation but in general caused displacement of upper units towards the ENE. Fission-track dating by Stephanie Brichau (Brichau 2004) shows that ductile top-ENE extensional reactivation at the base of the Selçuk nappe occurred in the Early Miocene as indicated by zircon fission track ages of 20-18 Ma. The zircon fission-track ages consistently young northeastward in the direction of hangingwall slip (Kumerics et al. 2005). One zircon fission track age of 14.1±0.8 Ma from a conglomerate at the southern slopes of Kerkis Mountain suggests that the Kerketas extensional system is slightly younger than the extensional fault at the base of the Selçuk nappe. The pattern of fission-track ages indicates that both extensional fault systems are unrelated to each other.
Late-stage D3 emplacement of the Kallithea nappe had a top-NW/NNW sense of shear (Ring et al. 1999a). Inception of the Kallithea detachment is fairly well dated at ~10 Ma.
(3) A short period of brittle E-W crustal shortening (D4) occurred between >8.6 to ~9 Ma. D4 shortening caused numerous W-vergent folds and reverse faults and only affected the lower sequence of the Neogene sediments below the unconformity.
(4) A phase of N-S-directed normal faulting (D5, 8.6 Ma to Recent). A granodiorite dike in the footwall of the Kallithea detachment yielded a zircon fission track age of 7.3±0.6 Ma (Brichau 2004) indicating that the Kallithea detachment continued to operate or was reactivated during D5 extension.
The cause for the short-lived D4 shortening event between 8.6 and 9 Ma remains enigmatic. It is also not fully clear whether E-W shortening during D4 was coeval with N-S extension; however, the general absence of NW-striking sinistral and NE-striking dextral strike-slip faults suggests that E-W shortening was not coeval with N-S extension. However, it might be that the extensional emplacement of the Kallithea nappe continued during the short-lived shortening event.