Sources of bias in detrital zircon geochronology: Discordance, concealed lead loss and common lead correction

• Published in: Earth-science reviews 2019-10, Vol.197, p.102899, Article 102899
• Main Authors: Andersen, Tom, Elburg, Marlina A., Magwaza, Boniswa N.
• Format: Article
• Language: English
• Subjects: Common lead; Detrital zircon; Lead loss; Lu-Hf; Provenance; U-Pb
• ISSN: 0012-8252; 1872-6828
• DOI: 10.1016/j.earscirev.2019.102899

The distribution pattern of U-Pb ages of detrital zircon in a sedimentary rock is commonly assumed to reflect the ages of igneous or metamorphic processes in rocks that have contributed material to the sedimentary basin (i.e. the protosources), directly or through recycling of older sedimentary rocks. If the Pb isotopic composition of detrital zircon is modified by processes after crystallization, or influenced by unintended effects of data treatment such as discordance filters and common lead correction, the value of detrital zircon as geological indicator is compromised. Discordance filters will identify zircons having suffered recent lead loss, but significant amounts of ancient lead loss may pass undetected. Lead loss events after the zircon's primary crystallization can be induced by regional or contact metamorphism, but also by low temperature processes during diagenesis and weathering, and can be coupled to uptake of a mixture of common lead and unsupported radiogenic lead, which cannot be properly corrected by common‑lead correction routines. Concealed ancient lead loss and overcorrection for common 207Pb may cause bias towards lower ages, while remaining within acceptable discordance limits. This creates spurious age fractions that may give false indications of sedimentary provenance, invalidate estimates of maximum limits for the age of deposition, and cause problems for comparison and correlation studies based on detrital zircon age data. Careful scrutiny of all U-Pb analyses in combination with Hf isotope analysis may help identifying these effects in detrital zircon data, but will not provide a universal guarantee against biased age spectra.