The Devil is in the Details: The Interpretation of Multistage U-Pb Data, with an Example from St. Mary’s Islands

The U-Pb method has become one of the most prominent approaches for dating geological events. A number of U-bearing minerals can be used, depending on the problem addressed and the type of geological material available, but zircon is the most useful and commonly analysed mineral. In many cases, this mineral provides high-precision dates for one or a sequence of stages in the evolution of a rock unit, including the timing of magmatic and metamorphic events. In detail, the information provided by the U-Pb method can be affected by a number of complicating factors, which can be difficult to detect for low-precision analyses, but become increasingly evident as the degree of precision increases. The most common effect observed is due to partial Pb loss that affects mostly old and U-rich domains of zircon due to alteration effects, diffusion or leaching. Loss of Pb also occurs when zircon undergoes metamorphic recrystallisation or resorption. Another type of complication arises when multiple growth generations are present in a population and/or in individual grains. Initial isotopic disequilibrium with excess 230Th or 231Pa, or loss of gaseous 222Rn, can all cause discordance. High precision geochronology can resolve multistage zircon growth during different stages of single magmatic cycles, with antecrysts first developed in magma chambers at depth and subsequent zircon domains or whole grains forming in the final stages of magma crystallisation, or just before eruption for volcanic rocks. One such example is reported for a volcanic rock from St. Mary’s Islands on the western coast of India, previously dated at 91.2 ± 0.2 Ma. Zircon grains from our new sample provide a classical U-Pb data array reflecting the presence of antecrysts besides zircon formed shortly before eruption. The array spans some 3 m.y., and the youngest date of 91.8 Ma is close to the published age.