Two principal metamorphic belts or petrotectonic zones have been recognized in the metamorphic complex
(Fig.3) by Yen (1963). The Tailuko belt on the west consists mainly of pelitic schist, gneiss, migmatite,
metamorphosed limestone, greenschist, siliceous schist, and amphibolite. Most of the rocks in this belt have
been subjected to high-grade greenschist fades metamorphism. More intense physical conditions of
metamorphism produced higher metamorphic grades in the northeastern part of the Tailuko belt than the rest
of the metamorphic terrain. This is attested by the presence in this area of amphibolite, paragneiss,
granitic intrusion, migmatite, and local occurrence of sillimanite (Chu, 1981). The exposure of these
features in northeastern Taiwan may reflect deeper burial during the Plio-Pleistocence convergence, as this
area was in the site of first impact during the collision (Ernst, 1982). The Tailuko belt has been intruded
locally in the north by granitic rocks, which were later metamorphosed into gneiss. The pegmatite in the
granite yields the oldest radiometric age of 86 m.y. (Yen and others, 1964). This Late Cretaceous age date
may represent an episode of crustal deformation and late-stage magmatic activities. The main period of
granitic intrusion could have been earlier in the Early Mesozoic.
The Yuli belt on the east is much smaller than the Tailuko belt. It is composed of a monotonous series of
black pelitic schist with minor intercalations of greenschist. Most of the mafic tectonic fragments or
exotic blocks of metamorphosed oceanic rocks are included in the Yuli belt. Other significant lithologic
differences between these two belts are the conspicuous occurrence of carbonate strata and granitoid rocks
in the Tailuko belt, and the dominance of meta-pelite and presence of glaucophane schist in the Yuli belt.
Early study indicated that these two metamorphic belts are separated by a longitudinal fault of great
magnitude (Yen, 1963), the so-called Shoufeng fault. However, the location, nature and extension of this
fault are poorly defined in the literature. Later workers in this metamorphic terrain have not located this
boundary fault unequivocally in the field. Thus tectonic contact between these two belts cannot be clearly
defined. Yang (1981) reports a fault in Juisui and believes this could be the trace of the Shoufeng fault.
Stanley and others (1981) indicated the possible presence of this fault on the southern cross-island highway.
Here the fault was reported to be a low-angle thrust.
The immediate juxtaposition of two suites of highly contrasting lithologies in these two metamorphic belts
is suggestive of a significant crustal displacement, and a fault boundary can be accepted. However, defining
and locating the exact trace of this fault in the field is rather difficult. Therefore the boundary between
PM3 and PM4 (representing these two metamorphic belts) on the revised map is rather arbitrarily determined.
Based on the occurrence of glaucophane, Yen (1963) ascribed the Yuli belt to the blueschist fades of high
P/T conditions and the Tailuko belt to the greenschist fades of low P/T conditions. These two metamorphic
belts thus constitute paired metamorphic belts following Miyashiro's (1961) observation on circum-Pacific
mountain belts. In the plate tectonic context, Biq (1971c) suggested a fossil subduction zone to explain
these two belts. The Yuli belt is interpreted to be a subduction zone beneath Mesozoic Taiwan and the Tailuko
belt is postulated to be an old magmatic arc on the continental side of the trench. Comparable to this
hypothesis, Ernst (1983) postulated that these two metamorphic belts possess some of the contrasts
characteristic of circum-Pacific volcanic/plutonic arcs and outboard subduction complexes. He indicated
further that the Tailuko terrain may represent a pre-existing rifted margin onto which was built an island
or continental margin volcanic/plutonic arc. The Yuli belt may be a volcanogenic argillite trench (?)
deposit. The age of early deformation and recrystallization is mainly Mesozoic. As previously stated, Liou
(198 la) suggested that the Yuli belt may include a melange complex, compatible with the fossil subduction
zone advocated by Biq (1971c) who also compared the Yuli belt with the Franciscan melange complex in