The Fish Canyon Tuff, La Garita caldera, and newly identified precaldera lava-like rocks were the focus of an informal four-day field workshop in early September 1997, involving 18 participants from 5 countries.
Precaldera lava-like rocks
A large precaldera lava-like body (30 km across, to 0.5-1 km thick, and 200-300 km3 volume) along the recently recognized south margin of the La Garita caldera, and a small postcaldera lava flow in the northern moat, are compositionally indistinguishable from the ash-flow tuff, documenting variable eruptive processes from a large magma body. The precaldera dacite was previously interpreted as thick welded Fish Canyon Tuff (Steven and Lipman, 1976) ponded within an existing caldera, but it lacks lithic fragments and is locally flow layered. The dacite displays complex textures and structures that are transitional between typical silicic lavas and pyroclastic deposits. Initial eruptions were pumiceous, and the main dacite body contains widespread fragmental textures and distinctive "blob" breccias (rounded fragments to 2 m), indicative of initial eruption as relatively dense spatter in a poorly sorted fragmental matrix. No evidence has been found to indicate initial eruption as fluid lava, even though some proximal material is so strongly flow layered that fragmental textures have been obliterated. Gravitational spreading has locally generated large ramp structures (dips to 70°) in distal deposits. We interpret the lava-like rocks as the product of low-energy pyroclastic fountaining, rather than fluidal lava. Although petrologically diverse lavas and tuffs were erupted widely in the San Juan region before Fish Canyon eruptions, the lava-like dacite is the only compositional precursor to the voluminous Fish Canyon ash flows.
Detailed mapping of the precaldera lava-like dacite has documented that complex faulting accompanied eruption of this enigmatic body. Although not yet completely analyzed, these faults define a broad complex north-trending graben that was beheaded by subsequent subsidence of the La Garita caldera. Many of the graben faults cut the lava-like dacite but are overlain unconformably by the succeeding Fish Canyon Tuff. Along some, the dacite thins abruptly or is depositional against fault-line scarps. Along other faults, the dacite is variably disrupted, locally showing extreme brecciation or intense rheomorphic flowage. Overall, these faults define modest piecemeal block-fault subsidence associated with eruption of the dacite, the first such subsidence geometry recognized in the San Juan region. Notably, the piecemeal graben structure is associated with low-energy pyroclastic activity, rather than major ash-flow eruption and caldera subsidence.
Generalized geologic map of the La Garita caldera
La Garita caldera
The La Garita caldera is larger than previously recognized (35x75 km), more elongate, and segmented. This caldera was formerly interpreted as about 40x30 km, with its south margin concealed beneath younger caldera subsidences in the Creede area (Steven and Lipman, 1976). In the summer of 1995, however, we found the south margin of the La Garita caldera well exposed within the rugged Weminuche Wilderness, 30-35 km south of its previously inferred position. Eruptive activity, depths of subsidence, and postcollapse resurgence varied among three sectors, even though only single outflow tuff sheet is associated with the caldera. In each sector, outflow tuff is truncated along the caldera walls against which the intracaldera tuff wedges out depositionally. Both intracaldera and outflow tuffs have the same phenocryst assemblage, mineral compositions, paleomagnetic polarity, isotopic ages, and elemental compositions, providing for confident correlation, along with the stratigraphic constraints.
In the northern caldera sector, the La Garita Mountains are a resurgently uplifted block of intracaldera Fish Canyon Tuff more than 1200 m thick without exposed basal contacts (Steven and Lipman, 1976). Intracaldera tuff is strongly indurated and oxidized red-brown, in comparison to the light gray outflow, and it contains larger and more coarsely porphyritic pumice lenses (10-20 cm). Along the western side of the central sector, a tilted block that may represent another locus of resurgence exposes an incomplete section of intracaldera tuff (>350 m) resting on older volcanic units of the caldera floor. Erosional levels in the southern sector, as in the northern, expose thick intracaldera tuff (>800 m) without reaching the caldera floor. Resurgence is absent in the southern sector; instead, linear faults recurrently disrupted the caldera fill, which includes andesitic lavas (Huerto Formation) that flooded the southern sector after collapse. The "new improved" La Garita caldera completely encloses the caldera sources of the 7 major ash-flow sheets erupted from the central San Juan field during the next 1.5 m.y.
Petrologic features of the Fish Canyon Tuff and associated lava-like rocks are also more complex than previously described. Re-study of the Fish Canyon Tuff demonstrates that erupted homogeneous dacite (68- 69% SiO2) was in contact with subjacent andesitic magma, and feldspar-liquid disequilibrium is comparable to that which characterizes strikingly resorbed quartz. The latter observation explains formerly controversial thermo-barometric discrepancies, may pose problems for the utilization of Fish Canyon minerals as geochronologic standards, and places important constraints on the origin of the Fish Canyon magma body. Large poikilitic sanidines enclose plagioclase, quartz, and other minerals; grain-boundary melting along contacts with these inclusions is frequently at an advanced stage. Partial diffusive equilibration adjacent to melt pockets, and along sanidine grain margins, has produced large compositional gradients that truncate pre-existing zoning in sanidine. Rare, intact plagioclase-sanidine contacts apparently preserve up-temperature diffusion gradients superimposed on down-temperature gradients. Such features indicate that the Fish Canyon magma reservoir formed either by rapid, shallow partial fusion of an existing batholith, or by remelting of the solidified margins of a long-lived magma chamber. Quartz resorption reflects a major thermal event, not a response to decreasing pressure during magma ascent (contraction of SiO2 stability relative to feldspars). Despite uniform bulk-sample compositions of the lava-like dacite and main Fish Canyon Tuff, a mafic component was involved. Most of the precaldera dacite is homogeneous, but rare andesitic blebs (1-5 cm diameter, 58% SiO2) and concentrations of finely porphyritic mafic minerals smeared along flow layers provide evidence for mingling with andesitic magma prior to eruption. The first postcollapse volcanism around south and west margins of the caldera (voluminous lavas of the Huerto Andesite) also tapped mafic compositions.
Sparse small fragments of comagmatic granophyre in late-erupted intracaldera tuff and postcaldera lava, having mineral compositions indistinguishable from phenocrysts in the tuff and precaldera lava-like rocks, record complex events in the Fish Canyon chamber just prior to and during eruption (Lipman et al., in press). Sanidine phenocrysts in the granophyre preserve zoning evidence of mingling with andesitic magma, then shattering by decompression and volatile loss accompanying early Fish Canyon eruptions before overgrowth by granophyre. The textural and chemical disequilibria indicate that the eruption resulted from batholith-scale remobilization of a shallow subvolcanic chamber, triggered by mafic magma replenishment, and erupted before textural or chemical equilibrium were re-established, contrary to previous interpretations of magma storage and phenocryst growth in the lower crust.
Lipman, P.W., Dungan, M.A., Bachmann, O (1997) Comagmatic granophyric granite in the Fish Canyon Tuff, Colorado: implications for magma-chamber processes during a large ash-flow eruption. Geology, in press
Steven, T.A., and P.W. Lipman (1976) Calderas of the San Juan volcanic field, southwestern Colorado. U.S. Geol. Survey Prof. Paper 958, 35 p.
Peter Lipman, U.S. Geological Survey, 345 Middlefield Road, Menlo Park, California, 94025 USA ph. 415-329-5295, fax 415-329-5203
Michael Dungan and Olivier Bachmann, University of Geneva, Switzerland
last modified: Sept. 22, 1997