ridge segments. However, the lavas from the ITSCs in the Siqueiros and Garrett
Transform Faults have lower concentrations of incompatible trace elements com-
pared to the lavas from the adjacent ridge segments, showing the opposite signa-
tures expected for a transform fault effect (Perfit et al. 1996 ; Wendt et al. 1999 ).
Formation of ITSCs may be the result of a “leaky transform” phenomenon.
Segmentation of transform faults is thought to be the result of transtensional forces
imposed on the transform domain by plate motion reorganization (Pockalny et al.
1997 ). The change in spreading direction enables mantle upwelling beneath the
transform domain and results in formation of ITSCs (Pockalny et al. 1997 ). The
above geochemical characteristics are explained by the recent numerical modeling
that considers brittle weakening of the lithosphere along a transform fault. The
model generates a region of enhanced mantle upwelling and elevated temperatures
at its center relative to the adjacent ridge segments (Behn et al. 2007 ). Elevated
temperatures near the center of the transform fault may in turn promote the devel-
opment of ITSCs during changes in plate motion. The leaky-type ITSCs are there-
fore robustly magmatic and not likely to host OCCs. It should be noted that the
peridotites from the Garrett Transform Fault are considered to be exposed at a
ridge-transform intersection, not at an ITSC (Constantin 1999 ). The Garrett perido-
tite is thus manifesting a classic transform fault effect.
Recent numerical modeling for detachment faulting at spreading ridge (assumed
spreading rate = 5 cm/year full-rate) indicates that the right amount of melts is
necessary to facilitate long-lived detachment faults and the resultant OCCs
(Tucholke et al. 2008 ). They showed that long-lived detachment faults form when
30-50% of total extension is accommodated by magmatic accretion, and do not
form or persist for long when magmatism is extremely limited. This indicates that
long-lived detachment faults appear to form only within a limited window of the
condition of melt supply where there is neither too much nor too little melt, but
instead just right amount of melt. Tucholke et al. ( 2008 ) therefore viewed their
hypothesis as the “magmatic Goldilocks hypothesis”.
We have proposed three possible mechanisms that can explain the unusual tec-
tono-magmatic characteristics of the rapid intermediate-spread PVB. In order to
test these ideas, we have examined the global mid-ocean ridges that have the similar
features to the PVB in terms of occurrence of assumed cold and/or refractory
mantle domain, multiple fracture zones, and OCCs in intermediate-spreading rate.
The unusual tectono-magmatic characteristics of the PVB may be manifesting the
characteristic features predicted by Tucholke et al.'s model. In the western PVB, a
cold and/or refractory mantle domain inhibited a large amount of mantle melting
within an intermediate-spreading ridge, attaining the Tucholke et al's limited win-
dow of the condition of magma supply in an otherwise robust magmatic environ-
ment. In the central PVB, a transform sandwich effect and/or declining spreading