John W. Shervais: Research Projects

Mantle Plumes and Continental Volcanism

Our current work in the Snake River Plain of southern Idaho focuses on the role of mantle plumes ("Hot Spots") and their interactions with continental lithosphere. Mantle plumes are thought to represent the upwelling of hot, enriched material from the core-mantle boundary towards the surface. Volcanic rocks which form by passage of lithosphere over a mantle plume will display chemical and isotopic compositions which vary in response to plume dynamics and interactions between the plume and the overlying lithosphere. Our work in the Snake River Plain uses both surface samples and drill core samples from existing scientific and geothermal test wells. Our goal is to reconstruct the geochemical and geodynamical history of the Yellowstone plume and its interactions with the overlying continental lithosphere.

The Yellowstone Plume: Fact or Fiction? Petrology, Geochemistry and Isotope Geochemistry of Basalts from the Central Snake River Plain Near Twin Falls, Idaho (Collaborative Research)

The object of this project is to investigate connections between the dynamics of mantle plumes and continental basalts, and the viability of the plume model for volcanism on the Snake River Plain. The proposed connection between mantle plumes, flood basalts, and intraplate seamount chains implies that volcanic rocks which form by passage of lithosphere over a mantle plume will display chemical and isotopic compositions which vary in response to plume dynamics and interactions between the plume and the overlying lithosphere. In contrast, alternative models that involve edge effects between the western edge of North America (moving to the SW), the Farallon plate (subducting to the NE), and the convecting asthenosphere may explain a wide range of observed characteristics, including younging of volcanism to the NE (in Idaho) and to the NW (in Oregon), lithospheric structure, and the chemical and isotopic composition of lavas along the alleged plume track.

We propose to continue our investigation of these processes by examining spatial and temporal variations in the chemistry of basaltic lavas found in the central Snake River Plain in the area west, north, and northeast of Twin Falls, Idaho. This area lies between two areas that we have already studied in some detail (basalts of the Bruneau-Jarbidge eruptive center to the SW, and basalts from a deep drill core at the INEL site, to the NE). It also lies near the juncture of the western SRP graben with the eastern SRP trend. A major focus of our previous work has been the origin and evolution of the western SRP, and its role in the debate between plume-generated volcanism and "edge-effect" volcanism. This area has received little prior attention in previous studies of SRP volcanism, but it occupies a critical gap in our knowledge of chemical trends of eastern SRP volcanism.

Selected samples will be analyzed for major and trace element geochemistry by XRF spectrometry; a representative subset of these will be chosen for analysis of additional trace elements by ICP-MS and radiogenic isotopes (Pb, Sr, Nd, Hf). We also plan to date 10 lava flows using 39Ar/40Ar. These data will allow us to assess changes in the mantle source regions of the basalts, and to identify the effects of lithospheric and crustal assimilation. Sample selection will be based on existing geologic maps, mapping in progress by the Idaho Geological Survey, and new mapping that we are carrying out under grants from the USGS EDMAP program. Bill Bonnichsen of the Idaho Geological Survey will assist with our field studies and the correlation of units between areas.

The primary goal of this research is to document variations in lava chemistry and isotopic composition as a function of time and location, and to correlate these variations with those expected in a plume-generated continental basalt suite. We will contrast these results to those expected for a shallow "hot line" model in which volcanism is driven by a convective roll of asthenosphere that is shaped by the over-riding continental lithosphere. It is expected that our study will lead to an increased understanding of linear continental volcanic arrays and will help discriminate between volcanic provinces formed by mantle plume dynamics and those formed by shallow processes at the lithosphere&emdash;asthenosphere boundary.

Current Project: "Collaborative Research: Plume Tails and Continental Lithosphere: A Petrologic Traverse of the Snake River Plain"

This project documents field, petrologic, geochemical, isotopic, and geochronological relationships of Neogene basaltic volcanism in the western and central Snake River Plain. This area lies at the critical junction of the eastern Snake River Plain (which may represent the track of the Yellowstone hotspot) with the western SRP, a structural graben that connects the eastern SRP with the Columbia River Plateau flood basalt province.

Most models for the origin and evolution of the Snake River Plain (SRP) volcanic province focus on the central role of the Yellowstone hotspot and its effects on the lithosphere of North America in response to plate motions relative to this hotspot. In these models, movement of the North American plate over the Yellowstone hotspot has resulted in a linear track of volcanism that parallels this plate motion, represented today by volcanic rocks of the eastern SRP.

The western SRP structural graben is oriented at a high angle to the trace of the Yellowstone plume and to the axis of the eastern SRP. It is filled largely with lacustrine sediments related to Pliocene Lake Idaho, a large, long-lived lake system that formed first at the northwestern end of the graben (near Oregon) and extended to the southeast along with the structural graben. Lake deposits extend back into the late Miocene, and are underlain by older basalts that are best known from deep drill core. High-temperature rhyolite lavas that mark the onset of extension also become younger to the southeast. Because the western SRP lies at an acute angle to the track of North American plate motion, it cannot be related to passage of North America over a fixed hotspot in any simple way.

Basaltic volcanism in the western SRP occurred in two distinct episodes. The first episode, represented by samples from a deep drill core near Mountain Home and by older surface outcrops that sit directly on rhyolite, is characterized by ferrobasalts that are distinct from other SRP basalts. The second episode is represented by surface flows of Pleistocene age that are intercalated with or overlie lacustrine and deltaic sediments of Lake Idaho. These basalts are associated with young faults that reflect basin and range extension. The younger basalts are similar to young basalts of the ESRP, but are generally more Fe-rich; they are distinct from lavas of the Basin and Range province in Nevada and elsewhere. At the SE end of our transect, in the Bruneau-Jarbidge eruptive center, younger basalts with western SRP affinities (e.g., Salmon Falls Butte) are chemically distinct from older basalts associated with formation of the eruptive center and the eastern SRP chemical trend.

Our data suggest that the western SRP graben represents an aulocogen-like structure formed in response to thermal tumescence above the Yellowstone plume head (?) as it rose under eastern Oregon and Washington, during and after eruption of the Columbia River Plateau and Steens Mountain basalts. The plume head was deflected northwards either by subduction of the Farallon plate (Geist and Richards, 1993) or by impingment of North American plate lithosphere (Camp, 1995). Basaltic volcanism in the western SRP may be related to the flow of depleted plume-source mantle along a sublithospheric conduit beneath the western SRP graben from the Columbia River Plateau toward the plume track. The basalts would form by pressure-release melting of this previously depleted material, along with the overlying mantle lithosphere. The younger volcanic episode apparently formed in response to basin and range extension, in a fashion analogous to young basaltic volcanism in the eastern SRP. The source of these basalts is uncertain, but may be plume-modified subcontinental lithosphere.

Well-characterized basalts from the Snake River Plain (SRP) were analyzed for Pb, Sr, and Nd isotopes. The purpose was to test the proposed connection between the Yellowstone Plume and the volcanic rocks of the SRP. Such a connection implies that the SRP basalts should display chemical and isotopic compositions that vary in response to plume dynamics and interactions with the overlying sub continental mantle lithosphere (SCML). The spatial and temporal variation of 28 basalts from INEL WO-2 core site in the northern SRP, 19 basalts from the Bruneau-Jarbridge and 5 basalts from the King Hill areas (B-J) in the central SRP, and 16 basalts from the Steens Mountain area (located between the Columbia River Basalt (CRB) province and McDermitt caldera in southwest Oregon, in collaboration with J. Johnson and P. Hooper) have been analyzed for Pb, Sr, and Nd isotopes. The Pb, Nd, Sr isotope relationships for the SRP basalts can be interpreted to define a mixing array between SCLM sources and a mantle source similar to OIB/plume mantle sources (Fig.1). The high 3He/4He ratios found along the plume track at Yellowstone and in the CRB and SRP support the interpretation that the SRP basalts represent regularly varying interaction of the Yellowstone plume with the SCML (Craig et al., 1978; Kennedy et al., 1985; Poreda and Cerling, 1992; Dodson et al., 1997). Spatial (and temporal) isotope variation along the SRP from the B-J area to Yellowstone is characterized by decreasing 206Pb/204Pb and 143Nd/144Nd and increasing 87Sr/86Sr ratios. The average Pb, Sr, and Nd isotope ratios for locations along the SRP between the B-J area and Yellowstone are linearly correlated with the square of the distance from the Yellowstone volcanic field. This suggests that the proportion of SCLM to plume source increases from west to east. The Pb isotopes for the Steens Mt. Basalts are consistent an oceanic-like mantle source.


Publications Resulting From This Project

  • Abstracts
  • Geologic Mapping

    Eight USGS 7.5' quadrangles , all located in the Mountain Home area of Idaho, were mapped as part of this study:
  • Mountain Home North
  • Mountain Home South
  • Teapot Dome
  • Crater Rings
  • Crater Rings SW
  • Crater Rings SE
  • Reverse
  • Cinder Cone Butte
  • Six quadrangles were mapped in the Shoshone area in 1998 and 1999 with USGS EDMAP funding (Dietrich, Dietrich Butte, Owinza, Owinza Butte, Star Lake, and Shoshone SW).

    In the Mountain Home area, the resulting composite map spans 30 of longitude (W115û30 to W116û0) and 15 of latitude (N43û0 to N43û15), encompassing about 450 square miles (1,170 square kilometers). All of these map products have been prepared digitally using the GIS system MapInfo, which is fully compatible with ArcInfo and complies with Federal GIS standards. All of the completed maps will be assembled for publication by the Idaho Geological Survey as Technical Reports at 1/24,000 scale; the Mountain Home 30x15 maps will also be compiled for publication as a Map Series map at approximately 1/50,000 scale.

    In addition, portions of quadrangles were mapped in the Mount Bennett Hills area as part of the Centenary College RUI effort: Deer Heaven Mountain 7.5, Dempsey Meadows 7.5, King Hill 7.5, Hog Creek 7.5, Davis Mountain SW 7.5.


    Graduate Student Research:
    Gaurav Shroff Petrology and Geochemistry of basaltic volcanism, Mountain Home, Idaho; Master's Thesis, University of South Carolina, In Progress.
    Scott Matthews Petrology and Geochemistry of basaltic volcanism Near Shosone, Central Snake River Plain, Idaho; Master's Thesis, University of South Carolina, 2000.
    Matt Cooke Petrology, Geochemistry, and Hydrology of basaltic volcanism, Central Snake River Plain, Idaho; Master's Thesis, University of South Carolina, 1999.

    Senior Thesis and/or Directed Independent Studies: Centenary College, Louisiana
     Scott Matthews The Geology of the Western Mount Bennett Hills, King Hill Area, Southern Idaho, May 1998, Departmental honors (Centenary College, Louisiana)
    Ryan Lanclos GIS of Geochemical Data for the Western Snake River Plain; Directed Independent Study, In progress.
     Heather Mahaney Digitizing the Mount Bennett Hills 7.5 quads and producing a 3-D image of the area., In progress
    Clay Hargett Basaltic Volcanism of Salmon Butte, SW Idaho: Implications for the Western Snake River Plain, In progress.
     Peter Guarisco Mapping of the Dempsey Meadows Quad, Mount Bennett Hills, In progress.