Evolution of shield-building and rejuvenescent volcanism of Mauritius Jacob Moore a, 1 , William M. White a, ⁎, Debajyoti Paul b, 2 , Robert A. Duncan c , Wafa Abouchami d , Stephen J.G. Galer d a Department of Earth and Atmospheric Sciences, Cornell University, Snee Hall, Ithaca, NY 14583, USA b Department of Earth and Environmental Science, The University of Texas San Antonio, One UTSA Circle, San Antonio, TX 78249, USA c College of Oceanic and Atmospheric Sciences, Oregon State University, Ocean Administration Building 104, Corvallis, OR 97331, USA d Max-Planck-Institut für Chemie, Abteilung Biogeochemie, Postfach 3060, D-55020 Mainz, Germany abstract article info Article history: Received 30 November 2010 Accepted 16 July 2011 Available online 30 July 2011 We report chemical and isotopic analyses of 68 samples and 40 Ar/ 39 Ar ages of 47 samples from Mauritius undertaken to understand the compositional evolution of the volcano and its causes through time. New 40 Ar/ 39 Ar ages show that construction of the Mauritius shield was well underway by 8.9 Ma, 1.1 m.y. earlier than previously thought and that the hiatus between the Intermediate and Younger Series was shorter than previously thought, as eruption of the rejuvenescent Intermediate Series continued through at least 1.66 Ma and Younger Series volcanism began by at least 1.0 Ma. Eruption frequency over the last 50 ka has been rather typical of Younger Series volcanism over the last 400 ka and future eruptions are possible. Although outcrops of the Intermediate Series lavas are confined to the Southwest, Intermediate Series are present beneath Younger Series flows in drill cores throughout the island. We estimate the total volume of rejuvenescent lavas at ~35 km 3 or about 0.05% of the volume of the volcano, similar to rejuvenescent volume fractions on Hawaiian volcanoes. As earlier studies found, Older Series lavas, which on average are slightly normatively silica-saturated, are somewhat more incompatible-element enriched than are the Intermediate and Younger Series, which are both slightly silica-undersaturated on average. Mean Sr and Nd isotope ratios of the Intermediate and Younger Series are nearly identical, but mean Pb isotope ratios, La/Sm, Nb/Y, and Nb/Zr of the Intermediate Series are higher than in the Younger Series. New high precision Pb isotope data, which shows considerably less scatter than previously published data, plot between the Older Series and basalts of the Central Indian Ridge, allowing the possibility that the source of the rejuvenescent lavas is a mixture of this plume and depleted mantle. We propose two possible explanations for the composition of rejuvenescent lavas. The first is that plume-derived melts reacted with deep lithosphere to form pyroxenite veins during the early shield-building stage. Later, these veins melted as a consequence of conductive heating of the lithosphere by the plume to produce the rejuvenescent lavas. Alternatively, rejuvenescent lavas may be derived from a sheath of thermally entrained mantle that surrounds the plume and is a mixture of plume material and depleted upper mantle. © 2011 Elsevier B.V. All rights reserved. 1. Introduction and regional setting “You gather the idea that Mauritius was made first, and then heaven; and that heaven was copied after Mauritius.” –Mark Twain, “Following the Equator”, 1897 Oceanic island volcanoes are thought to be produced as convective columns of hot, rising mantle approach the Earth's surface and melt (Morgan, 1971). As the lithospheric plate upon which it rides moves steadily away from the mantle plume, the volcano's connection to it is broken and the volcano becomes extinct as a new volcano begins to grow just upstream (Wilson, 1963). This progressive growth and death of volcanoes produces the volcanic chains that decorate the ocean floor like strings of pearls. Best known are the island chains of the Pacific such as Hawaii and Samoa, but they occur in the other oceans as well. Mauritius is the penultimate island of the Réunion hotspot track (e.g., Morgan, 1981), inferred to have erupted from magmas of the deep-seated Réunion mantle plume (Courtillot et al., 2003; Montelli et al., 2004) as the Indian plate moved northeastward over the stable plume. In all three of these examples – Hawaii, Samoa, and Mauritius – this simple mantle plume model fails in an important respect: volcanic activity Journal of Volcanology and Geothermal Research 207 (2011) 47–66 ⁎ Corresponding author. Tel.: + 1 607 255 7466; fax: + 1 607 254 4780. E-mail address: wmw4@cornell.edu (W.M. White). 1 Now at Ellington & Associates, Inc., 1414 Lumpkin Road, Houston, TX 77043, USA. 2 Now at Department of Civil Engineering (Geosciences), Indian Institute of Technology Kanpur, 208016 (UP), Kanpur, India. 0377-0273/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jvolgeores.2011.07.005 Contents lists available at ScienceDirect Journal of Volcanology and Geothermal Research journal homepage: www.elsevier.com/locate/jvolgeores