Page 66 - Caldera Volcanism Analysis, Modelling and Response

Residence Times of Silicic Magmas Associated with Calderas 41

one unit fall in the time evolution of the main reservoir suggesting that either it was
derived from the still remaining melt or that the crystals are recycled from an
already crystallised part of the same reservoir.

4.1.2. Residence times recording previous magmatic episodes (erupted or not)
The long residence compared to the solidification times of some small eruptions,
as well as their positions with respect to the caldera cycle above, open the question
about the sources of these crystals. At least two possibilities have been proposed:
they are recycled from previous magmatic episodes that have been erupted; or
where the crystals’ ages have no correspondence with the erupted record, they are
derived from plutonic magmas or rocks.
Recycling of crystals from well identified magmatic activity has been
documented in Yellowstone, where the ages of the crystals in some young and
small post-caldera lavas (o0.6 Ma) are the same as those of the Huckleberry Tuff
(Bindeman et al., 2001) or the same as other post-caldera products (Vazquez and
Reid, 2002). Recycling of crystals from non-erupted magmas has been invoked
in many other systems. For example, zircons in deposits o45 ka of Taupo volcanic
zone display two main peaks at ca. 30 and 100 ka but such ages have no
correspondence with the erupted record (Charlier et al., 2005). This is also found in
post-caldera lavas of the Long Valley system, where several units contain zircons
with ages between 200 and 300 ka, which corresponds to a gap in the post-caldera
volcanism at Long Valley (e.g., Hildreth, 2004). Once crystal inheritance has been
recognised, the plutons could be from the same magmatic system (or even reservoir)
as the magmas that carried the crystals to the surface or not. Evidence for the
former is found in the Taupo system (see above) and another is from Crater Lake.
Bacon et al. (2000) and Bacon and Lowenstern (2005) showed that the zircon ages
from granodiorite xenoliths from the caldera forming deposits are the same as those
of pre-caldera rhyodacite lavas.

4.2. Time scales of magmatic processes
Other time information obtained from caldera-related magmas is the duration and
rates of magmatic processes (Table 2). Residence times establish a crystallisation
episode or history of the phases used to derive the age. Geochronological studies
that use Rb–Sr isotope systematics of main phenocrysts could quantify dif-
ferentiation rates of silicic magmas. Unfortunately, the systems that have been
studied in detail show that the role of open system process is so important that the
time information is difficult to be interpreted (e.g, Long Valley, Valles-Toledo). The
most reliable results appear to be those obtained from accessory minerals. Although,
it is not clear how these may relate to a significant event of magma differentiation
their time information seems more robust. For example, the study of Vazquez and
Reid (2004) of allanites from the Toba Tuff quantified fractionation rates of silicic
magmas to be 1–4% every 10 ky.
Time information on magmatic processes and rates is also derived from
relaxation of chemical zoning in minerals by diffusion or from thermal models