Abstract The establishment of accurate time scales of mineral systems is essential to construct reliable genetic models about their formation. Time scales of fossil mineral systems are directly determined through radiometric dating of different stages of development of the mineral system. In theory, porphyry systems are, among mineral systems, those whose duration can be bracketed with most accuracy and precision, because of the universal occurrence of ore and gangue minerals that can be dated with the high precision U-Pb zircon, Re-Os molybdenite, and 40 Ar/ 39 Ar dating techniques. Time scales of fossil porphyry systems reported in the literature range between <0.1 to >4 Ma. The long durations (>1 Ma) of magmatic-hydrothermal activity measured in several porphyry systems are likely the result of multiple magmatic pulses in agreement with field observations indicating that porphyry systems are associated with several intrusive events. Nonetheless, estimated long durations could also be affected by methodological problems. One methodological problem is the accuracy of the intercalibration among the three different methods. It has become evident during the last 15 years that 40 Ar/ 39 Ar dates are systematically younger compared to U-Pb dates. This has been attributed to incorrect values of the secondary standard (Fish Canyon Tuff sanidine), most commonly used to calculate 40 Ar/ 39 Ar ages, and/or of the 40 K decay constant. Systematic cross calibrations to check the consistency between Re-Os and U-Pb dates are lacking and should also be carried out. Another possible cause of erroneous long durations of porphyry systems concerns the way to determine the emplacement age of the causative intrusion. The current high precision (≤0.1%) of single zircon U-Pb dating by isotope dilution-thermal ionization mass spectrometry (ID-TIMS) shows that zircon grains extracted from a single sample of intermediate/felsic magmatic rocks do not overlap in age. This is so because zircon grains record a protracted evolution of magmas within the crust lasting several hundreds of thousands of years. Under these conditions, the emplacement age of a magmatic intrusion is best approximated by the youngest ID-TIMS age measured from a population of zircon grains. In contrast, spot ages measured with in situ techniques, due to their lower precisions (1-3%), are not able to discriminate such protracted magmatic evolution recorded by different zircon grains. This allows pooling together spot ages of different zircons, resulting in a statistically significant mean age with a low uncertainty. In reality this is a mixed age that is characteristically older (by up to a few hundreds of thousands of years) than the age of the youngest single zircon grain measured by ID-TIMS. A further problem in estimating the duration of magmatic-hydrothermal activity in porphyry systems derives from the widespread use of 40 Ar/ 39 Ar dating. Because this method does not date the crystallization of a mineral but rather its cooling below its closure temperature, 40 Ar/ 39 Ar dates may be affected by (hydro-)thermal activity that postdates the mineralization.

Abstract The Nambija gold district, southeastern Ecuador, consists of oxidized skarns developed mainly in volcaniclastic rocks of the Triassic Piuntza unit, which occurs as a 20-km-long, north-trending, contact-metamorphosed lens within the Jurassic Zamora batholith. High gold grades (10–30 g/t) are accompanied in most mines by very low Fe, Cu, Zn, and Pb sulfide contents. The skarn is constituted dominantly by massive brown garnet (mean Ad 38 ). Subordinate pyroxene-epidote skarn developed mainly at the margins of brown garnet skarn bodies. Mostly idiomorphic and more andraditic garnet (mean Ad 45 ) occurs in blue-green skarn formed as a later phase, in places with high porosity, at the transition with vugs and discontinuous dilational type I veins. The last garnet generations are mainly andraditic and occur largely as honey-yellow to red-brown clusters and cross-cutting bands (mean Ad 84 ). As typical for other skarns developed in volcaniclastic rocks, mineral zoning is poorly defined. The retrograde overprint is weakly developed, commonly fails to alter the prograde minerals, and is mainly recognized in mineral infilling of structurally controlled (N10°–60°E) vugs and up to several-centimeter-wide type I veins, as well as interstices in blue-green skarn. Retrograde minerals are milky quartz, K-feldspar, calcite, chlorite, and hematite, ±plagioclase, ±muscovite, plus minor amounts of pyrite, chalcopyrite, hematite, sphalerite, and gold. Vugs and type I veins are cut by thin (1–2-mm) throughgoing type II veins that show similar orientations and mineralogy. Native gold is associated with retrograde alteration, mainly in the irregular vugs and type I veins, and subordinately in interstitial spaces and throughgoing type II veins. It is not observed in sulfide-rich type III veins, which cut the previous vein generations. High-temperature (up to 500°C) and high-salinity (up to 60 wt % NaCl equiv) inclusions in pyroxene represent the best approximation of the fluid responsible for a significant part of the prograde skarn stage. Such a highly saline fluid is interpreted as the result of boiling of a moderately saline (~8–10 wt % NaCl equiv) magmatic fluid at temperatures of ~500°C. Moderate-to low-salinity fluid inclusions (20−2 wt % NaCl equiv) in paragenetically later garnet as well as in epidote and quartz from vugs and type I veins may represent later, slightly lower temperature (420° −350°C) trapping of similar moderately saline fluids with or without some degree of boiling and mixing. The similarity of salinities and homogenization temperatures in late garnet, epidote, and quartz fluid inclusions is consistent with the apparent continuum between the prograde and retrograde skarn stages, as illustrated by the general lack of prograde mineral alteration, even at the contacts with retrograde fillings. Gold deposition, together with that of small amounts of hematite, chalcopyrite, and pyrite, took place during fluid cooling in the retrograde skarn stages but not during the last retrograde alteration, as indicated by the absence of gold in the sulfide-rich type III veins. The abundance of gold-bearing samples with high hematite/sulfide ratios and generally low total sulfide contents suggests high oxygen fugacities during gold deposition. The northeast structural control of vugs and type I veins, compatible with regional northeast-striking structures, in part with a dilational character, suggests that skarn formation, including gold deposition in the retrograde stage, took place under conditions of tectonic stress. Minimum Re-Os ages of 145.92 ± 0.46 and 145.58 ± 0.45 Ma for molybdenite from type III veins are compatible with skarn formation and gold mineralization during Late Jurassic magmatism. A genetic relationship with felsic porphyry intrusions that cut the Jurassic Zamora batholith and crop out near several gold skarns is suggested by a published hornblende K-Ar age of 141 ± 5 Ma for a felsic porphyry in the northern part of the Nambija district. Furthermore, the minimum Re-Os ages of ~146 Ma are just slightly younger than the published K-Ar ages (154 ± 5, 157 ± 5 Ma) for the Pangui porphyry copper belt about 70 km north of Nambija.