Volcanic stratigraphy and geochemistry of the Soufrière Volcanic Centre, Saint Lucia with implications for volcanic hazards

https://doi.org/10.1016/j.jvolgeores.2013.04.011Get rights and content

Highlights

  • We revise the stratigraphy of the Soufrière Volcanic Complex in Saint Lucia.

  • Combined (U–Th)/He and U–Th zircon geochronology provides new ages.

  • Pyroclastic deposits range in age from 640 to 20 ka.

  • Future activity is likely to occur from within the Qualibou depression.

Abstract

The Soufrière Volcanic Complex (SVC), Saint Lucia, represents one of the largest silicic centres in the Lesser Antilles arc. It comprises extensive pumiceous pyroclastic flow deposits, lava flows as well as Peléan-style domes and dome collapse block-and-ash-flow deposits. These deposits occur within and around the Qualibou Depression, a ~ 10-km diameter wide sector collapse structure. To date, vent locations for SVC pyroclastic deposits and their relationship to the sector collapse have been unclear because of limited stratigraphic correlation and few radiometric ages. In this study we reconstruct the geologic history of the SVC in light of new and recently published (U–Th)/He, U–Th and U–Pb zircon chronostratigraphic data, aided by mineralogical and geochemical correlation. Compositionally, SVC deposits are monotonous medium-K, calc-alkaline rocks with 61.6 to 67.7 wt.% SiO2 and display similar trace element abundances. Combined U–Th and (U–Th)/He zircon dating together with 14C ages and mineral fingerprinting reveals significant explosive eruptions at 640, 515, 265, 104, 60 and 40 ka (producing deposits previously grouped together as the “Choiseul” unit) and at 20 ka (Belfond unit). The mineralogically and geochemically distinct Belfond unit is a large, valley-filling pumiceous pyroclastic flow deposit distributed to the north, northeast, south and southeast of the Qualibou Depression that was probably deposited during a single plinian eruption. The unit previously referred to as ‘Choiseul tuff’ is much less well defined. The typical Choiseul unit comprises a series of yellowish-white, crystal-poor, non-welded pumiceous pyroclastic deposits cropping out to the north and southeast of the Qualibou depression; however its age is poorly constrained. A number of other units previously mapped as Choiseul can be distinguished based on age, and in some cases mineral and whole rock chemistry. Pyroclastic deposits at Micoud (640 ± 19 ka), Bellevue (264 ± 8 ka), Anse John (104 ± 4 ka) and La Pointe (59.8 ± 2.1 ka), Anse Noir and Piaye were all previously grouped with or associated with the Choiseul tuff (all uncertainties 1σ). We suggest that these units represent individual periods of activity spanning a range of ages, whereas Choiseul pumice at the type locality has yielded a (U–Th)/He zircon age of 515 ± 19 ka. Their overall geochemical and mineralogical similarities with the Choiseul at the type locality suggest that they might have all originated from the same centre. Morne Tabac (532 ± 21 ka) is a dome truncated by the depression escarpment, whereas Morne Bonin (273 ± 15 ka), Gros Piton and Petit Piton (71 ± 3 ka and 109 ± 4 ka, resp.), Belfond (13.6 ± 0.4 ka) and Terre Blanche (15.3 ± 0.4 ka) are domes within the Qualibou Depression. Belfond and Terre Blanche have whole rock geochemistry and mineral assemblages similar to the Belfond pyroclastic flow deposit, thus possibly representing late-erupted degassed portions of the magma that produced the Belfond pyroclastics. The geochemical characteristics and similar zircon age distributions of the silicic lava domes and pyroclastics of the SVC suggest that these share a common magma source beneath the Qualibou depression. The distribution of the pyroclastic flows and the wide range in their eruption ages makes it unlikely that these were erupted during caldera-forming activity, and we instead invoke a series of smaller-volume explosive eruptions from the area of the current depression, the earliest of which occurred from a large proto-Qualibou edifice that subsequently underwent sector collapse. Activity from this proto-Qualibou centre may have ceased sometime between 38 and 59 ka ago, it therefore seems unlikely given our present understanding that there will be another eruption from the southern central highland region. However, the young dome-forming activity in the Qualibou depression may have occurred in or close to the Holocene, and there have been dome collapse events and explosion craters formed since then. A new dome eruption or renewed activity at a dome within the depression, growing in the style of the ongoing Soufrière Hills lava dome on the nearby island of Montserrat, is possible; as is a future plinian eruption from this area. Such an eruption would not only have a devastating impact on Saint Lucia, but would also have significant regional and global impacts.

Introduction

Eruptions of silicic magma have produced some of the largest and most explosive volcanic events in Earth's history. Although relatively infrequent compared to more mafic eruptions, large silicic eruptions have the potential to produce severe socio-economic impacts on both regional and global scales. Most of the recent work on understanding evolution and hazards of large silicic magma bodies has focussed on areas where this magmatism is most common: continental arcs such as the Taupo Volcanic Zone (e.g. Wilson et al., 2009) or long-lived intra-continental settings such as Long Valley and Yellowstone (e.g. Christiansen, 2001, Reid, 2008). Less attention has been given to silicic magmas in island arc settings, although recent work on the Izu–Bonin, Scotia, Aegean and Tonga–Kermadec arcs (e.g. Tamura and Tatsumi, 2002, Leat et al., 2003, Mortazavi and Sparks, 2004, Smith et al., 2006, Bachmann et al., 2012) has shown that silicic magmatism can be a significant component of island arc systems.

The island of Saint Lucia in the Lesser Antilles provides an opportunity to study the eruptive history and potential hazards of silicic systems in an archetypal, well-established island arc setting. Saint Lucia and its neighbour Dominica have erupted dominantly medium-K calc-alkaline andesite and dacite, in contrast to islands to the north and south in the arc, which have produced dominantly more mafic magmas (Macdonald et al., 2000). The largest eruptions of such magmas in Saint Lucia are represented by a series of extensive pyroclastic deposits considered by previous workers to have erupted between 40 and 20 ka ago (Wright et al., 1984, Wohletz et al., 1986). This silicic pyroclastic activity is not typical for the Lesser Antilles arc, the only comparable deposits being the Pleistocene–Holocene Roseau tuff, Grand Savanne ignimbrite, Layou Tuff and Grand Bay ignimbrite and associated domes on Dominica (Sparks et al., 1980a, Sparks et al., 1980b, Lindsay et al., 2003, Lindsay et al., 2005a, Tinnin et al., 2006). These deposits in Saint Lucia and Dominica, which represent the largest known explosive eruptions in the Lesser Antilles arc, can be considered a flare up of silicic activity in an otherwise dominantly basaltic-andesite to andesitic portion of the arc. Interestingly, several of these silicic deposits are associated temporally and geographically with major sector collapse and debris avalanche events, yet the relationship between the collapses and the eruptions is unclear (e.g. Lindsay et al., 2005a, Lindsay et al., 2005b).

The pyroclastic deposits on Saint Lucia, together with associated lava flows, domes, block-and-ash-flow deposits and explosion craters, form the Soufrière Volcanic Centre (SVC) in the southwest of the island (Fig. 1). The SVC is located within a major collapse feature called the Qualibou depression, and is associated with vigorous geothermal activity at the Sulphur Springs geothermal field (Fig. 1). There has been no historical magmatic activity in Saint Lucia, however there was a series of phreatic eruptions in the SVC in 1766 (Lindsay, 2005). Because of its vigorous geothermal activity, recent intense seismic activity (in 2000–2001) and potential for violent explosive eruptions, the SVC is the focus of volcanic monitoring and hazard assessment in Saint Lucia by researchers at the Seismic Research Centre, University of the West Indies, Trinidad (Lindsay et al., 2002, Lindsay et al., 2005b, Joseph et al., 2013).

Previous workers developed contrasting theories for the source and significance of the pyroclastic units in southern Saint Lucia. Wright et al. (1984) proposed that they were derived from andesitic stratovolcanoes in the central highlands, while Wohletz et al. (1986) concluded that they were sourced from within the Qualibou depression during caldera collapse. This distinction has important implications for the assessment of volcanic hazards related to the Qualibou depression, and has motivated the renewed study of the volcanism reported here. One major difficulty in understanding the volcanic history of this region has been the lack of reliable correlation criteria from the various widespread pyroclastic units. Recently, Schmitt et al. (2010) used combined (U–Th)/He and U–Th zircon geochronology to establish a new chronostratigraphic framework for the SVC. In our study, we refine the stratigraphic relations within the SVC in light of this and other new data, and by applying new petrographic and geochemical analysis from a suite of lava flows, lava domes, and pyroclastic flow deposits that include the stratigraphically oldest and youngest units of the SVC. The aim is to reconstruct the volcanologic history of the SVC, in particular that of the silicic pyroclastic units associated with the Qualibou depression, and to discuss the implications for volcanic hazards.

Section snippets

Regional setting

Currently about one million people in the Caribbean region are threatened by the direct effects of volcanic eruptions. The ongoing Soufrière Hills eruption on Montserrat has had a major impact on that island's population since it began in 1995 (Sparks and Young, 2002), and is a constant reminder of the potential for volcanic activity in other islands. The island of Saint Lucia (620 km2 area) has a population of 163,267 (2001 census), all of whom would be affected should there be renewed volcanic

Geology of Saint Lucia

Saint Lucia consists almost entirely of volcanic rocks (Fig. 1), which can be categorised into three main groups based on age and distribution (Lindsay, 2005). From older to younger these are: (1) eroded basalt and andesite centres, (2) dissected andesite centres, and (3) the Soufrière Volcanic Centre. The oldest radiometric ages obtained from basalts in Group 1 are 15–18 Ma (Briden et al., 1979, Le-Guen de Kerneizon et al., 1983), although shallow marine volcaniclastic deposits in the northern

Stratigraphy and dating of the Soufrière Volcanic Centre

The dense vegetation, extreme weathering of outcrops, and high degree of leaching of organic material that could be used for dating have made it difficult to constrain the volcanic stratigraphy of the SVC. Early studies of the units associated with the SVC (Tomblin, 1964, Wright et al., 1984, Wohletz et al., 1986) lead to conflicting interpretations. More recent dating studies (e.g. Samper et al., 2008, Schmitt et al., 2010) have revised older chronostratigraphy for several volcanic features

Sampling and analytical methods

For this study we sampled all units associated with the SVC, and analysed a total of 58 pumice and lava samples for whole-rock geochemistry, which represent pyroclastic deposits (36 samples) as well as dacitic domes and associated lava flows (13 samples). The dissected andesite centres of the central highlands are the source of the numerous andesitic volcaniclastic units in and around the SVC. These centres are densely forested and for the most part inaccessible; our coverage is limited to a

Petrography and mineral chemistry

There is significant variation in textures and modal abundances among the various pumiceous pyroclastic units in the SVC, particularly between the Belfond pyroclastic flow deposit and units that were previously referred to collectively as the Choiseul tuff (e.g. Wright et al., 1984). Schmitt et al. (2010) showed that the quartz-poor pumice deposits collectively mapped as the ca. 40 ka old Choiseul tuff include products of at least three other eruptions at ca. 264 ka (Bellevue), 104 ka (Anse John)

Geochronology

In order to close key gaps in the chronology of the Soufrière Volcanic Complex we carried out age determinations on (1) pumice from Micoud pyroclastic flow deposit to date deposits previously mapped as Choiseul tuff on the east coast of Saint Lucia; (2) pumice from the Choiseul type-location to test the previously determined 14C age of 40 ka; and (3) lava from near the summit of the pre-Qualibou-collapse Morne Tabac, in order to further constrain the age of the collapse (Fig. 1).

The new

Overview

Mineral assemblage and compositional data were combined with age data to identify key silicic units and their correlations within the SVC (Table 2) as a basis for grouping samples in the geochemical descriptions given here. This was done according to the following scheme. Firstly, units for which a unique age is known are identified in all plots. This includes the Belfond, La Pointe, Anse John, Bellevue and Micoud pyroclastic deposits and the Terre Blanche and Belfond domes. Secondly,

Re-definition of pyroclastic units and relationship to domes

The geochronological, mineralogical and geochemical constraints discussed here permit a revised definition of the SVC pyroclastic flow deposits. The Belfond unit is confirmed as a large, valley-filling pumiceous pyroclastic flow deposit; with deposit thicknesses ranging from 1 to over 100 m (it occurs in the Piaye river valley where thick inaccessible cliffs visible from the Londonderry village appear to be > 100 m high). It is distributed to the north, northeast, south and southeast of the

Conclusions

The geological history of the Soufrière Volcanic Complex (SVC) in Saint Lucia, has been reconstructed in light of new and recently published (U–Th)/He, U–Th and U–Pb zircon chronostratigraphic data, aided by mineralogical and geochemical correlations. The main conclusions of this work are:

  • Lava and pumice from domes and pyroclastic flows of the SVC are monotonous medium-K, calc-alkaline andesites and dacites with 61.6 to 67.7 wt.% SiO2, and with similar trace element abundances.

  • Combined U–Th

Acknowledgements

Rudolf Naumann and Daniel Goode are thanked for their assistance in XRF analyses at the GFZ-Potsdam, Ian Smith for LA-ICP-MS analytical work at the Research School of Earth Science, Australian National University, and Roman Kislitsyn for the assistance with the (U–Th)/He dating. We thank Daniel Goode, Richard Robertson, and Roman Kislitsyn for their assistance with field work. This work was supported by the National Geographic Society grant 8517-08. The ion microprobe facility at UCLA is partly

References (56)

  • A. Samper et al.

    Radiometric dating of three large volume flank collapses in the Lesser Antilles Arc

    Journal of Volcanology and Geothermal Research

    (2008)
  • A.K. Schmitt et al.

    The Geysers–Cobb Mountain Magma System, California (Part 1): U–Pb zircon ages of volcanic rocks, conditions of zircon crystallization and magma residence times

    Geochimica et Cosmochimica Acta

    (2003)
  • A.K. Schmitt et al.

    Episodic growth and homogenization of plutonic roots in arc volcanoes from combined U–Th and (U–Th)/He zircon dating

    Earth and Planetary Science Letters

    (2010)
  • I.E.M. Smith et al.

    Petrogenesis of dacite in an oceanic subduction environment: Raoul Island, Kermadec arc

    Journal of Volcanology and Geothermal Research

    (2006)
  • R.S.J. Sparks et al.

    The entrance of pyroclastic flows into the sea, I. Oceanographic and geologic evidence from Dominica, Lesser Antilles

    Journal of Volcanology and Geothermal Research

    (1980)
  • R.S.J. Sparks et al.

    The entrance of pyroclastic flows into the sea, II. Theoretical considerations on subaqueous emplacement and welding

    Journal of Volcanology and Geothermal Research

    (1980)
  • S. Turner et al.

    U-series isotopes and destructive plate margin magma genesis in the Lesser Antilles

    Earth and Planetary Science Letters

    (1996)
  • K. Wohletz et al.

    The Qualibou Caldera, St. Lucia, West Indies

    Journal of Volcanology and Geothermal Research

    (1986)
  • Aquater SpA

    Exploration of St. Lucia's Geothermal Resources, Annex A — Geological Survey

    (1982)
  • R.J. Arculus et al.

    The petrology of plutonic blocks and inclusions from the Lesser Antilles island arc

    Journal of Petrology

    (1980)
  • O. Bachmann et al.

    Evolution of silicic magmas in the Kos-Nisyros volcanic center, Greece: a petrological cycle associated with caldera collapse

    Contributions to Mineralogy and Petrology

    (2012)
  • R. Bezard et al.

    Assimilation of sediments embedded in the oceanic arc crust: myth or reality?

    Geology

    (2013)
  • G. Boudon et al.

    Volcano flank instability in the Lesser Antilles arc: diversity of scale, processes and temporal recurrence

    Journal of Geophysical Research

    (2007)
  • J.C. Briden et al.

    K–Ar geochronology and paleomagnetism of volcanic rocks in the Lesser Antilles island arc

    Philosophical Transactions of the Royal Society of London

    (1979)
  • R.L. Christiansen
  • J.P. Davidson

    Isotopic and trace elements constraints on the petrogenesis of subduction-related lavas from Martinique, Lesser Antilles

    Journal of Geophysical Research

    (1986)
  • C. DeMets et al.

    GPS geodetic constraints on Caribbean–North America plate motion

    Geophysical Research Letters

    (2000)
  • E. Heath et al.

    Magmagenesis at Soufriere Volcano, St Vincent, Lesser Antilles Arc

    Journal of Petrology

    (1998)
  • Cited by (16)

    • Guidelines for digital geological maps of Pliocene-Holocene composite volcanoes: A contribution from Colombia

      2021, Journal of South American Earth Sciences
      Citation Excerpt :

      In particular, the FGDC Digital Cartographic Standard for Geologic Map Symbolization (Federal geographic data committee [prepared for the federal geographic data committee by the United States geological survey]. (FGDC-USGS), 2006) presents a useful and comprehensive practical guide showing the symbology used for geological features, distinguishing the variable lithological settings. We integrate this information with the one retrieved from the maps published by Barsch and Liedtke (1980), Lara et al. (2004), Nonnotte et al. (2008), Madeira et al. (2010), Martí et al. (2010), Tibaldi (2010), Branca et al. (2011), Lindsay et al. (2013), the compilation of the Aeolian Islands (Lucchi et al., 2013, and references therein), Marroni et al. (2015), Báez et al. (2015) and Carrasco-Núñez et al. (2017) to design a new catalogue. These examples consistently follow the International Stratigraphic Guide (Salvador, 1994) and the accepted international terminology for volcanic lithologies (e.g. Fisher, 1966; Schmid, 1981; Fisher and Schmincke, 1984; Cas and Wright, 1987; Le Maitre et al., 2002; White and Houghton, 2006; Nèmeth and Martin, 2007).

    • Extreme alteration in an acid-sulphate geothermal field: Sulphur Springs, Saint Lucia

      2018, Chemical Geology
      Citation Excerpt :

      The oxygen was converted to CO2 over red-hot graphite for isotopic measurement using a Micromass Optima dual-inlet, stable-isotope-ratio mass-spectrometer. Hydroxyl-group hydrogen was extracted from silicate whole-rock samples using a method modified after Godfrey (1962), Kyser and O'Neil (1984) and Vennemann and O'Neil (1993). The H2 was analysed using a Thermo Scientific™ DeltaPlusXL or Delta VPlus dual-inlet stable-isotope-ratio mass-spectrometer.

    • The onset of the volcanism in the Ciomadul Volcanic Dome Complex (Eastern Carpathians): Eruption chronology and magma type variation

      2018, Journal of Volcanology and Geothermal Research
      Citation Excerpt :

      The effect of isotope disequilibrium has to be considered, however, only for Quaternary samples, because for older zircons the uncertainty related to the (U-Th)/He-method is greater than the secular disequilibrium effect on the ages (Farley et al., 2002). This correction can be performed based on the crystallization ages of analyzed zircon crystals obtained by U-Th or U-Pb zircon dating (e.g., Schmitt et al., 2006, 2010; Schmitt, 2011; Danišík et al., 2012, 2017; Lindsay et al., 2013). If crystallization ages cannot be obtained, or are heterogeneous, the “secular equilibrium assumed” (U-Th)/He age and the “full disequilibrium assumed” (U-Th)/He age will give the upper and lower age limits for the eruption age, respectively.

    • Application of combined U-Th-disequilibrium/U-Pb and (U-Th)/He zircon dating to tephrochronology

      2017, Quaternary Geochronology
      Citation Excerpt :

      Because the (U-Th)/He ages corrected for disequilibrium represent a multi-component dataset that may display complexities such as anomalous “outliers” that may potentially skew the final result, it is strongly recommended to critically analyse and evaluate the data and not use MCHeCalc simply as a black-box. The ZDD method was introduced ten years ago and, although its major focus of application has been in studying the dynamics of young magmatic systems (Schmitt et al., 2006, 2010a, b, 2011; 2013, 2014a, b; Lindsay et al., 2013; Gebauer et al., 2014; Howe et al., 2015), the potential of the ZDD method for tephrochronology is undisputable. The following are summaries of several studies that illustrate the application of ZDD in tephrochronology and demonstrate its capability to produce accurate ages that agree with ages obtained from other dating techniques.

    • Assimilation of sediments embedded in the oceanic arc crust: Myth or reality?

      2014, Earth and Planetary Science Letters
      Citation Excerpt :

      The lavas from this single island encompass almost the whole range of isotopic variation observed in the arc making it the perfect location to study the influence of crustal components (Fig. 2). The more continental crust-like compositions are found in andesites (3 Ma to ca. 250 ka) and dacites (100 ka to present) of the Soufriere Volcanic Complex (SVC), while more mantle-like compositions are observed in the more mafic Pre-SVC lavas dominated by basalt, basaltic andesite and andesite (18 Ma to 1.1 Ma; Briden et al., 1979; De Kerneizon et al., 1983; Lindsay et al., 2013; Samper et al., 2008; Schmitt et al., 2010). Except for 7 samples in which Sr and Nd isotope ratios were measured at the Geochemical Analysis Unit at Macquarie University (Aus), whole rock powders were analysed at the Arthur Holmes Isotope Geology Laboratory (AHIGL) which is part of Durham Geochemistry Centre (DGC) at Durham University (UK).

    View all citing articles on Scopus
    View full text