Our data clearly show the presence of several distinct Cretaceous age populations, which can be followed through different formations up-sequence for over 30 Myr as age clusters or age peaks in cumulative probability diagrams (Figure 5b). The proportion of the sample set that produced Cretaceous provenance ages in the sediments increases from 13% in the Oligocene up to 58% in Late Miocene times. The 68-75 Ma, the 82-92 Ma and the 107-120 Ma populations are the most persistent. Ages between 93 and 106 Ma appear as transient signals in the detrital mineral age population (Figure 6). Within the persistent populations two very distinctive peaks can be traced, one at 70 Ma and the other at 120 Ma. These ages are also supported by step heating analyses, which gave plateau ages in the Rocchetta (D61b) and Cassinasco Formations (D40, D50) of 108.7 ± 1 Ma and 105.7 ± 2 Ma, 77.6 ± 1 Ma respectively (Figure 6). Also, the 107-120 Ma age population is remarkably close to 40Ar/39Ar plateau ages on white mica and Rb/Sr ages on whole rocks measured in the Internal Western Alpine Massifs and in the Sesia Lanzo zone (i.e. of rocks exposed at the surface today) (Hunziker, 1970; Oberhänsli et al., 1985; Paquette et al., 1989). Ages in the range of 70- 90 Ma have also been widely recorded by different thermochronometers (e.g. U/Pb on zircons and 40Ar/39Ar and Rb/Sr on white mica), all over the Western Alps (e.g. Chopin, 1984; Inger et al., 1996; Monié and Chopin, 1991; Oberhänsli et al., 1985; Ruffet et al., 1995; Ruffet et al., 1997).
One of the fundamental assumptions underlying the K-Ar and 40Ar/39Ar dating methods is that the system does not contain 40Ar that the time that the clock starts, i.e. at the time of crystallization of a mineral or the time of cooling through its closure temperature, and thus the calculated age can be related to the event that we intended to study. If the system contains extraneous argon at this time, then ages will be measured that are older than the true age of crystallization or cooling (Dalrymple and Lanphere, 1969). The extraneous, i.e. the non-radiogenic 40Ar component may have various sources. For the present discussion two components are relevant: excess 40Ar and inherited 40Ar.
Following Dalrymple and Lanphere (1969) and Kelley (2002) excess 40Ar is represented by parentless radiogenic argon incorporated into a mineral during crystallisation, or introduced into the mineral lattice by subsequent diffusion or occluded within fluid or melt inclusions within the mineral. Thus, excess 40Ar is controlled by diffusion and fluid advection processes transporting 40Ar during metamorphic events from areas where minerals break down to areas where new minerals crystallize. Transport distances of excess, i.e. parentless, 40Ar may range from micrometres (Harrison and McDougall, 1981) to hundreds of metres (York et al., 1981). By definition, in the case of excess 40Ar the relation between radiogenic parent isotope 40K and radiogenic daughter isotope 40Ar is broken and thus no interpretable age information can be obtained. Excess argon may be deeply or superficially bound in the crystal lattice of minerals but is also likely to reside in fluid inclusions and solid inclusion within minerals.
Inherited argon is defined as the 40Ar component, essentially radiogenic, present in a rock or mineral sample by physical contamination from older material. This definition implies that in the case of inherited argon, the relation between radioactive parent 40K and radiogenic daughter, 40Ar, is maintained (e.g. Dalrymple and Lanphere, 1969; Wijbrans and McDougall, 1986; Singer et al., 1998). In the case of inherited argon in metamorphic rocks, the apparent age is produced, in part, by the decay of 40K, before the last cooling event. Thus, anomalously old ages caused by a component of inherited argon will never exceed the protolith age of the rock.
While excess Ar leads to age spectra, which cannot be interpreted in terms of the effects of real geological events, inherited Ar will produce age spectra that are influenced by mixing components of different ages, which however, under favourable circumstances can be interpreted in terms of the compounded effects of geological events (e.g. Wijbrans and McDougall, 1986; Villa, 1998; Forster and Lister, 2003).