Volcanic history and magmatic evolution of Seguam Island, Aleutian Island arc,

Brian R. Jicha† Brad S. Singer Department of Geology and Geophysics, University of Wisconsin–Madison, 1215 West Dayton Street, Madison, Wisconsin 53706, USA

ABSTRACT fractional crystallization of . Using studies, these can provide insight into magma the 40Ar/ 39Ar geochronology and estimates chamber longevity and rates of magma differ- New 40Ar/39Ar dating coupled with detailed of individual fl ow volumes, we calculated a entiation, all of which may have important roles fi eld mapping, stratigraphy, and chemical time-averaged eruptive rate at Seguam that in volcanic hazard mitigation strategies (e.g., analyses have established an eruptive chro- is similar to growth rates of other well-dated Harford et al., 2002). nology that reveals and constrains the compo- arc volcanoes in the Cascades and Chilean Despite the fact that there are well-docu- sitional and volumetric evolution of Seguam Southern volcanic zone but less than that of mented eruptive chronologies for Aleutian con- Island in the Aleutian Island arc, Alaska. Mount Katmai and Mount Mageik, which tinental arc volcanoes (Hildreth et al., 2003b; Sixty new 40Ar/39Ar ages from , domes, are located on the . The Bacon et al., 2003), very few 40Ar/39Ar data are and pyroclastic deposits were obtained using eruptive fl ux at Seguam has been highly vari- published from the historically active volcanoes furnace incremental-heating techniques on able, fl uctuating more than an order of mag- in the Aleutian Island arc (Jicha et al., 2004). replicate samples of whole-rock and ground- nitude, from 0.07 km3/k.y. during the early Seguam Island, located in the central Aleutian mass separates, and they constrain the dura- history of bimodal volcanism to 1.18 km3/k.y. Island arc, is a 79 km3, low- to medium-K, tho- tion of Pleistocene to Holocene subaerial over the past 9 k.y. leiitic complex with multiple eruptive centers. volcanism to 318 k.y. The 40Ar/39Ar plateau It is unusual relative to other Aleutian Island ages indicate that over 85% of the complex, Keywords: Aleutian island arc, 40Ar/39Ar geo- arc centers in that it has erupted a signifi cant ~68 km3 of material, was erupted almost chronology, magmatic evolution, Seguam Is- volume (~30% of the total volume erupted) of continuously between 318 ka and 9 ka. At land, eruptive rates. evolved lavas with compositions ranging from ca. 9 ka, a stratocone on the eastern half of 63% to 71% SiO2. Previous K-Ar dating of 11 the island partially collapsed producing a INTRODUCTION whole-rock samples from Seguam indicated a 4-km-wide crater. Rhyolitic dome-forming 1.07 m.y. eruptive history (Singer et al., 1992a), eruptions followed from vents in the newly For many an active , the historical although the precision of these K-Ar ages is created crater, and were likely contempora- record, as well as the isotopic and geochemi- poor (1σ errors are ±6%–70%). The goals of neous with 8.0 km3 of basaltic and basaltic cal composition of its lavas and tephras, is this study were to use new fi eld observations andesitic effusions from Pyre Peak, and a well established. Yet, high-resolution eruptive along with high-precision 40Ar/39Ar geochronol- 1.4 km3 basaltic eruption from a monogenetic histories that extend back several hundreds of ogy to: (1) build upon the prior work of Singer cone on the far eastern end of the island. Geo- thousands of years have only been determined et al. (1992a, 1992b, 1992c), (2) quantify and chemical changes over the last 318 k.y. are for very few long-lived arc volcanoes. These constrain the compositional and volumetric evo- subtle. Most notably, the earliest eruptions include Mount Adams (Hildreth and Lanphere, lution of the 79 km3 complex, (3) estimate the from 318 to 142 ka produced no andesite, 1994), Tatara–San Pedro (Singer et al., 1997), rates of magma output over the lifetime of the and basalt from this period has larger ranges Santorini (Druitt et al., 1999), Montserrat (Har- volcano, and compare them to other arc volca- in Zr/Rb and La/Yb than younger . ford et al., 2002), Mount Baker (Hildreth et al., noes worldwide, and (4) establish the precise Small volumes of dacitic to rhyolitic magma 2003a), Katmai (Hildreth et al., 2003b), and geochronologic control required to interpret U- were produced from basalt by a monotonic Ceboruco–San Pedro (Frey et al., 2004). Estab- Th isotope disequilibrium dates in terms of time crystal-liquid fractionation process that lishing a chronology and quantifying the long- scales of crystallization and magma storage over varied only slightly in successive eruptive term growth of a volcanic complex requires the past 200 k.y. (Jicha et al., 2005). phases over 318 k.y. We identifi ed minor detailed geologic mapping supported by K-Ar or Despite major advances in 40Ar/39Ar dat- geochemical changes in magma composition 40Ar/39Ar age determinations, which document ing of late Pleistocene and Holocene and during each of the three main stages of vol- major events in a volcano’s eruptive history. tephra during the last decade (e.g., Renne et canism, but overall the monotonic variations Determining the growth rate and compositional al., 1997; Singer et al., 2000, 2004), obtaining in major- and trace-element compositions evolution of an arc volcano over a protracted precise 40Ar/39Ar ages from low-K, tholeiitic of basaltic andesitic to rhyolitic lavas are period of a few hundred thousand years is criti- lavas younger than 500 ka that are susceptible consistent with an origin via closed- system cal to understanding issues such as periodicity to alteration in a humid environment is a chal- of activity and frequency of explosive episodes. lenge, due to extremely low radiogenic 40Ar* †Email: [email protected]. When coupled with petrologic and isotopic contents in the presence of large amounts of

GSA Bulletin; July/August 2006; v. 118; no. 7/8; p. 805–822; doi: 10.1130/B25861.1; 12 fi gures; 5 tables, Data Repository item 2006126.

For permission to copy, contact [email protected] 805 © 2006 Geological Society of America Jicha and Singer atmospheric 40Ar. Additionally, estimating erup- North American plate at a convergence rate of to exploration expeditions in the Aleutians have tive rates at Seguam is complicated by glacial 6.6–6.8 cm/yr (DeMets et al., 1994). The Eocene indicated that Seguam has been historically and marine erosion and by vegetation and ash seafl oor dips 10° for 100 km, and then steepens active since the late eighteenth century, with cover. In spite of these pitfalls, we determined to ~50° beneath the volcanic front (Holbrook et eight eruptions over the past 200 yr, presumably the time-averaged eruptive rate based on mini- al., 1999). A P-wave velocity model suggests from Pyre Peak (Miller et al., 1998). In 1977 mum estimates of individual fl ow volumes and that the top of the slab is ~60 km beneath the and 1992–1993 basalt and basaltic andesite 40Ar/39Ar age determinations to be 0.3 km3/k.y. active volcanoes (Holbrook et al., 1999), which erupted from a 1.0-km-long fi ssure ~2 km south Like other arc volcanoes, the eruptive fl ux has is signifi cantly shallower than previous esti- of Pyre Peak. been highly variable throughout the lifetime of mates of slab depth based on the locations of Reconnaissance mapping and sampling facili- the volcano. earthquake hypocenters (Engdahl, 1977). tated the geochemical, petrologic, and isotopic studies of Singer et al. (1992a, 1992b, 1992c), TECTONIC AND GEOLOGIC SETTING VOLCANOLOGICAL OVERVIEW who identifi ed a four-phase geologic evolution of the island on the basis of fi eld observations and Seguam Island is an ~200 km2 volcanic com- Seguam Island comprises seven vents aligned 11 whole-rock K-Ar ages. Whole-rock Sr isotope δ18 plex in the central Aleutian Island arc (Fig. 1). in an east-west orientation, including those and Oplag data from Seguam lavas combined Seismic-refl ection and seismic-refraction pro- of Pyre Peak, Wilcox volcano, and Moundhill with a comparison of the modal mineralogy to fi les from between Seguam and volcano. At least 79 km3 of basaltic to rhyolitic published low-pressure cotectic conditions sug- Islands indicate that the volcanic com- lava, tephra, scoria, and ash fl ows, along with gest that basaltic parental magmas crystallized plex sits atop 25–30 km of strongly extended several rhyolite fl ows and domes are preserved at 3–5 kbar, or ~10–15 km depth, followed by arc crust (Geist et al., 1988; Holbrook et al., on the 11 × 21 km island. The volcanic complex closed-system differentiation to dacite and rhyo- 1999; Fliedner and Klemperer, 1999). The crust was severely eroded by an extensive glacial ice lite between 1 and 2 kbar (i.e., crustal depths of maintains a similar thickness for at least 100 km cap that covered most of the ~3–6 km) (Singer et al., 1992a, 1992c). behind the arc front. P-wave velocity structures during late Wisconsin time. Although glacial ice Using detailed mapping and stratigraphy, have been interpreted to refl ect an overall mafi c was also likely present on the island during each 40Ar/39Ar age determinations, and geochemi- composition of the crust, which is believed to of the major Pleistocene glaciations, little evi- cal data, we have identifi ed a subaerial eruptive be porous or fractured extrusive and intrusive dence for earlier glaciations has been preserved history that is virtually continuous from 318 igneous rocks and volcaniclastic sediments in the Aleutians (Black, 1983). All of the lavas ± 30 ka to the present, but signifi cantly shorter in the upper 7 km, a mid-ocean-ridge basalt and tephras that have reached the shoreline have than inferred by Singer et al. (1992a). We have (MORB)–like ~6-km-thick layer in the middle been subject to marine erosion also. Glaciated subdivided the volcanic evolution into three crust, and ~10–20 km of gabbroic residua at the late Pleistocene fl ows and tuffs are capped by stages that consist of: (1) older, deeply eroded base (Holbrook et al., 1999). Forty percent of Holocene lavas, domes, and ash deposits con- lavas and domes (318–142 ka); (2) island-wide the upper crust is inferred to be silicic in com- sisting of 1.2 km3 of rhyolitic domes in the east activity (138–9 ka), which includes shallowly position (Fliedner and Klemperer, 1999). How- and 8.0 km3 of basalt fl ows and scoria beds in dipping lavas in the central and western section ever, the lacks seismic evidence for the west. Moundhill volcano, a 1.4 km3 mono- of the island, and the construction of a 20 km3 a silicic middle crust (Holbrook et al., 1999). genetic spatter cone, was constructed on the basaltic to rhyolitic stratocone on the eastern Approximately 170 km south of Seguam is far eastern end of the island sometime after the half of the island (98–9 ka) that partially col- the ~7000-m-deep Aleutian trench, where the retreat of glacial ice at ca. 12–10 ka. The obser- lapsed and produced a 0.5 km3 dacitic ignimbrite Pacifi c plate is subducting obliquely beneath the vations of geologists or other scientists attached at 9 ka; and (3) postcollapse rhyolitic activity in

174°E 176°E 178°E 180° 178°W 176°W 174°W 172°W 170°W 168°W Shishaldin North American plate 54°N Okmok Figure 1. Bathymetric map of the Aleutian Island arc showing Attu Seguam ▲

▲ the location of Seguam Island ▲ and several other well-known Kanaga ▲ 52°N islands and volcanic centers. Adak ▲ ▲ Contour interval is 400 m. The ▲ Aleutian trench lies ~170 km ▲ ▲ ▲ seaward of the volcanic axis. ▲ ▲ ▲ ▲ Subduction of the Pacifi c plate ▲ ▲ ▲ ▲ ▲ ▲ ▲ becomes increasingly oblique 50°N Al eutian Trench 6-7 cm/year in the western part of the arc.

Km 050 100 200 Pacific plate

806 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island

the crater, and subsequent formation of a western 318-142 ka accompanied by mafi c tuffs and basaltic effusions (Holocene to present) (Fig. 2).

PETROGRAPHIC AND COMPOSITIONAL OVERVIEW

Seguam lavas are characterized by an anhy- 199 236 drous phenocryst assemblage dominated by 318 plagioclase (up to 42%), with lesser amounts 211 159 of olivine, clinopyroxene, orthopyroxene, and 191 titanomagnetite (Singer et al., 1992a, 1992b). Basaltic to basaltic andesitic fl ows and dikes are porphyritic, containing 22–58 modal per- 113 281 cent phenocrysts, whereas andesites are mildly 122 142 176 phyric, containing fewer total phenocrysts 189 210 (5%–31%). Conversely, crystal poor (4%–10% 202 174 crystals) dacitic to rhyolitic lavas are often

fl ow-banded and glassy with groundmass tex- 61 62 tures ranging from pilotaxitic to vitrophyric. 120 58 138-11 ka 133 123 77 The dacitic ignimbrite that accompanied crater 32 23 32 formation at 9 ka (plag + cpx + opx + mt = 15 122 56 11 83 53 49 modal %) is unusual in that it contains abun- 78 dant apatite phenocrysts within partially welded 99 85 66 glass shards and trace amounts (<0.5 modal %) 133 98 of anhedral, strongly resorbed biotite. In Figure 3, we show the compositional range 28 of lavas and tephras erupted during the 318 k.y. subaerial history of Seguam Island. There are no primitive lavas (<49 wt% SiO , >8 wt% 42 84 93 24 2 67 85 MgO) or high-silica rhyolites (>72 wt% SiO ) 64 12 2 93 on Seguam. The elevated FeO*/MgO ratios of 122 53 Seguam lavas from basalt to rhyolite are char- 43 33 83 49 acteristic of the tholeiitic series lavas as defi ned 80 by Miyashiro (1974) and Gill (1981). Each of 117 138 the three eruptive stages contains both low-K 53 basalts and medium-K rhyolites, but the major- ity of the exposed rhyolites belongs to the post- collapse suite. Interestingly, the early history of volcanism from 318 to 142 ka is devoid of inter- 9 ka-present mediate-composition lavas. 9 METHODS 7.5 Field Studies 6.0 1.7 Access to most of the island was achieved on foot. However, several shoreline exposures were only approachable via infl atable boat due to the steep sea cliffs. Field relationships were mapped using a compilation of National Oceanic and Atmospheric Administration (NOAA) maps 49-59% SiO2 T-10322 to T-10325 (1:20,000) as a topographic > 64% SiO base. The volcanic units illustrated in the geo- 2 logic map include lava fl ows, domes, ignim- brites, and ash cover inferred to have erupted Figure 2. Digital elevation models (DEMs) of Seguam Island with shaded areas indicating from multiple vents (Fig. 4). Individual fl ows the three main stages of eruptive activity over the past 318 k.y. Dots represent sample loca- and domes were delineated on the map where tions and are shown with corresponding 40Ar/39Ar ages reported in ka. The ages of 6.0 and possible, but several of the map units are stacks 1.7 ka shown in the bottom panel represent U-Th mineral isochron ages from Jicha et al. of lava fl ows that likely shared a common vent (2005). The lower panel is further subdivided according to wt% SiO2 content.

Geological Society of America Bulletin, July/August 2006 807 Jicha and Singer

8 2.5 Postcollapse lavas (Holocene-historic) 7 Wilcox volcano lavas (98-9 ka) High-K Shoreline lavas (138-11 ka) 2.0 6 Older eroded volcanics (318-142 ka)

5 1.5 Medium-K O 4 2 Tholeiitic K 1.0 3 FeO*/MgO

2 0.5 Low-K 1 Calc-Alkaline Basaltic Basalt Andesite Dacite Rhyolite 0 Andesite 0 48 52 56 60 64 68 72 48 52 56 60 64 68 72 SiO SiO2 2

Figure 3. FeO*/MgO and K2O variation versus SiO2 (wt%) for 141 samples from Seguam Island. Analyses by inductively coupled plasma– mass spectrometry (ICP-MS) at Actlabs in Ontario, Canada (n = 51), X-ray fl uorescence (XRF) at the Universität Göttingen, Germany (n = 40), and XRF from Singer et al. (1992a, 1992b) (n = 50). Samples have been subdivided according to fi eld relations and age constraints (Table 1). Compositional subdivisions of fi eld boundaries are from Miyashiro (1974).

and are mineralogically and chemically similar. Irradiation Tube (CLICIT), where they received ing each sample. All blanks were atmospheric in Most units have a narrow compositional range, fast neutron doses of 1.2–5.4 × 1015 n/cm2. composition and one to two orders of magnitude although a few include lavas that span a large J values for the samples were determined by smaller than the sample signals; thus, the impact range in SiO2 content. Ash cover and extensive interpolating laterally across the Al disks. The of blanks on age uncertainty was minimal. Mass erosion caused by repeated glacial advances precision of the J values was between ±0.19 and discrimination was monitored using an auto- have resulted in complicated stratigraphic rela- 0.50% (1σ), which was based on the weighted mated air pipette and varied between 1.0022 tions. Individual map units were often correlated average of 12–15 laser fusion analyses of single ± 0.002 and 1.0043 ± 0.002 per atomic mass based on a combination of 40Ar/39Ar age, major- sanidine crystals from each Al disk. Based on unit (a.m.u.) during the analytical periods. element geochemistry, and petrography. previous experiments, corrections for undesir- Even though single incremental-heating able nucleogenic reactions on 40K and 40Ca are: experiments often yielded precise ages, most 40 39 40 39 36 37 Ar/ Ar Geochronology [ Ar/ Ar]K = 0.00086; [ Ar/ Ar]Ca = 0.000264; samples required one to four replicate experi- 39 37 σ [ Ar/ Ar]Ca = 0.000673. ments on subsamples to achieve 2 analytical Holocrystalline groundmass separates were At the University of Wisconsin Rare Gas precisions of ±10%–15% and ±2%–3%, for prepared from porphyritic lava samples by Geochronology Laboratory, the groundmass basalts and rhyolites, respectively (Fig. 5). For crushing, sieving, magnetic sorting, and hand- packets and mini-cores were incrementally four samples, incremental-heating experiments picking under a binocular microscope to remove heated in a double-vacuum resistance furnace were performed on whole-rock mini-cores and olivine, pyroxene, and plagioclase phenocrysts attached to a 300 cm3 gas clean-up line. Prior groundmass separates from the same rock and and minimize the potential for xenocrystic con- to each incremental-heating experiment, sam- yielded nearly identical results (Table 1; GSA tamination. Most groundmass separates were ples were heated to 500–675 °C and pumped Data Repository Table DR11). Because iso- 250–500 μm. For a few samples, this size frac- to remove potentially large amounts of atmo- chron regressions (York, 1969) agreed with tion contained numerous microphenocrysts of spheric argon and water (Baksi, 1974). Fully plateau ages and did not reveal evidence that olivine and pyroxene, therefore the 180–250 μm automated experiments consisted of 4–10 steps excess argon was present in any of the lavas (i.e., fraction was used. Groundmass plagioclase var- from 675–1450 °C; each step included a 2 min 40Ar/36Ar intercepts that are indistinguishable ied between 20 and 90 μm. Whole-rock mini- increase to the desired temperature that was from 295.5), we consider the plateau ages to give cores, which were 5 mm in diameter and ranged maintained for 10–15 min. During the heating the best estimate of the time elapsed since erup- from 160 to 450 mg in weight, were drilled time, and for an additional 5–8 min afterward, tion (Table 1). All ages were calculated using the from aphyric lava fl ows. Groundmass separates the sample gas was exposed to three SAES C50 decay constants of Steiger and Jäger (1977). (~100–375 mg) were weighed and then wrapped Zr-Al getters. Isotopic measurements and data in 99.99% copper foil packets and, along with reduction followed the procedures of Singer the mini-cores, were placed into 2.5-cm-diam- et al. (2000, 2004). These measurements were 1GSA Data Repository item 2006126, Table DR1, eter Al disks with 1.194 Ma sanidine from the critically dependent on characterizing the blank the complete 40Ar/ 39Ar incremental heating results Alder Creek rhyolite (Renne et al., 1998) as a levels in the analytical system and the mass dis- from Seguam Island; and Table DR2, whole-rock major and trace element compositions of 40Ar/ 39Ar neutron fl uence monitor. The Al disks were irra- crimination of the mass spectrometer. Blanks dated samples, is available on the Web at http://www. diated for 20–90 min at the Oregon State Uni- were measured over a range of temperatures geosociety.org/pubs/ft2006.htm. Requests may also versity reactor in the Cadmium-Lined In-Core between 700 and 1350 °C prior to and follow- be sent to [email protected].

808 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island x Point 52°15’N 52°20’N Moundhill 15 mvl x X mvl 1925 28 x Moundhill volcano Ar age (ka) 39 wvl 172°20’W Ar/ Bowling ball basalt (1977 eruption)Bowling deposit Dacitic pumice airfall Mossy basalt Alluvium Colluvium/vegetation Basaltic ash & scoria Right angle basalt basalt Brown basalt fan Western of base camp Basalt west Andesite in amphitheater Andesitic ignimbrite Moundhill volcano lavas Moundhill volcano an 40 olc o 56 (NOAA) (1972) maps T-10322, T-10323, T-10323, T-10322, (NOAA) (1972) maps v Moundhill volcano x and lavas Pyre Peak wvl 33 49 bbb dpa Qal bas rab brb wfb bwb aam aig

o mvl msb Qcv msb basaltic lavas (Holocene) basaltic lavas

c 98 il (Holocene-historic)tephras 66 W X 85 daf X wvl 191 53 2779 rcc eld Cove Finch 211 32 Qcv 23 faf srf rdf Point Finch 12

ssb 400

77 0

Qcv

600

800 1000

rcv Lava Point

120

anking the Wilcox volcano are in purple. Red units are syncol- in purple. Red units are volcano are Wilcox anking the 24 1400

l lpd rfc a Peak and of of andesitic ignimbrite north Pyre ow directions 1710 str Q dig X 7.5

lpr 0 0

58 2

ows fl ows 0 x lpr Spiny rhyolite flow (6 ka) flow rhyolite Spiny rhyolitic cone in crater rhyolitic dome 1710 rhyolite (7.5 ka) valley in crater flow Rhyolite 1710 rhyolite dome 1710 rhyolite Long Canyon dacite Long Canyon rhyodacite Point Lava Dacitic ignimbrite (9 ka)

8.4

0

0 afp 0 0 0

172°25’W 1

0 8 4 600 84 85 62 93 Postcollapse dacites Postcollapse Stratocone collapse (9 ka) Stratocone and rhyolites (< 8.4 ka) and rhyolites lpr lcd lpr Qcv str Qcv srf rcv rcc dig bas 32 ssl Lava rdf Cove lcd Qal awf 53 61 bas dig bwb dig Y 64 bas 67 aig bas aig Qcv Z 123 rab 122 Qal bas 120 83 bwb msb South shore basaltic andesite (12 ka) tuff (23 ka) ash-flow Finch Cove dacite (24 ka) Point Lava (32 ka) rhyodacite Finch Cove andesite (58 ka) Finch Point (61 ka) of Finch Point Andesite west (77 ka) rhyolite Finch Cove (98-49 ka) lavas volcano Wilcox 133 Wilcox volcano lavas lavas volcano Wilcox (98-9 ka) and tephras 122 bbb 3458 172°30’W caldera Western X brb

bbb

aig faf rdf afp rfc lpd awf ssb wvl 80 0

Pyre Peak

260 33 2400 2200 49 11 2000 0

1800 2200 160 0 Qcv

1400 Turf

Point 83 2000 120 0 100 Qcv

800 600 1800 wfb 78 200 aig 400 ows, and tuffs. Basaltic lavas, scoria, and ignimbrites from Pyre Peak are green to light gray. Alluvium, beach sand, glacial to light gray. green Peak are Pyre ows, and tuffs. Basaltic lavas, scoria, ignimbrites from

138

dpa wfb 1600 tpr 142

174

1400 117

0 bas 99 120 aml

aam

210

1000 800 eld 93

10 x 600 281 dpa

aml 133

400 x X 200 53 Point 43 nsl Saddleridge 10 172°35’W Scale 1:20,000 eld Qcv 199 236 318 ssl Contour interval 200 feet 189 Qcv x X 0 1 2 3 4 km nsr 159 eld N Compiled from N.O.A.A. Maps T-10322-325 (1972) T-10322-325 Maps Compiled from N.O.A.A. 202 10 Andesitic to dacitic amphitheater lavas (< 93 ka) Andesitic to dacitic amphitheater lavas (318-174 ka) & dikes Deeply eroded lavas Basaltic to andesitic north shore lavas (133-11 ka) Basaltic to andesitic north shore lavas (138-33 ka) south shore lavas Basaltic to rhyolitic dome (159 ka) North shore rhyodacite Turf Point rhyodacite dome (142 ka) rhyodacite Point Turf Shoreline lavas (138-11 ka) Shoreline lavas 42 nsl Seguam Island Volcanic Complex Volcanic Seguam Island 113 Older eroded volcanics (318-142 ka) Older eroded volcanics tpr aml eld nsl ssl nsr 52°20’N ows from Moundhill volcano. Dashed line around Pyre Peak marks inferred rim of Holocene caldera. Peak marks inferred Pyre Moundhill volcano. Dashed line around ows from 176 122 Figure 4. Geologic map of Seguam Island. Topographic base map is compiled from National Oceanic and Atmospheric Administration Atmospheric National Oceanic and base map is compiled from Figure Topographic 4. Geologic map of Seguam Island. Lava fl orange, and gray. shown in shades of brown, The oldest of the 34 units are T-10325. and T-10324, lapse and postcollapse dacitic to rhyolitic domes, fl indicate fl given in feet. Black arrows Elevations are all shown in shades of yellow. are till, and tundra cover thin basaltic fl

Geological Society of America Bulletin, July/August 2006 809 Jicha and Singer

100 0.0040 SEG 03 32 rhyolite whole rock σ 75 weighted mean plateau 7.5 ± 2.0 ka (2 ) 0.0035 3 separate experiments Ar

50 40 0.0030 SEG 03 32 σ Ar/ 6.1 ± 3.5 ka (2 ) 36 40Ar/36Ar = 296.5 ± 2.4 25 0.0025 i MSWD = 0.05 n = 15 of 16 Apparent age (ka) 0 0.0020

240 0.0040 SEG 04 24 andesite groundmass σ 180 weighted mean plateau 41.9 ± 4.1 ka (2 ) 0.0035 2 separate experiments Ar 120 40 0.0030 SEG 04 24

Ar/ 38.8 ± 8.5 ka (2σ) 36 40 36 60 0.0025 Ar/ Ari = 296.9 ± 3.4 MSWD = 0.32

Apparent age (ka) Apparent age n = 15 of 15 0 0.0020

240 0.0040 SEG 04 35 andesite whole rock SEG 04 35 σ 80.7 ± 5.5 ka (2σ) 180 weighted mean plateau 82.9 ± 3.6 ka (2 ) 0.0035 40Ar/36Ar = 296.8 ± 2.45 2 separate experiments i Ar MSWD = 0.08 120 40 0.0030 n = 13 of 13 Ar/ 36 60 0.0025 Apparent age (ka) Apparent age 0 0.0020

400 0.0040 SEG 03 34 basaltic andesite groundmass SEG 03 34 weighted mean plateau 174.3 ± 4.8 ka (2σ) σ 300 0.0035 173.0 ± 14.0 ka (2 ) 3 separate experiments 40 36 Ar/ Ari = 295.8 ± 2.8 Ar

40 MSWD = 0.67 200 0.0030 n = 16 of 16 Ar/ 36 100 0.0025 Apparent age (ka) 0 0.0020

400 0.0040 SEG 04 42 basaltic andesite groundmass SEG 04 42 σ weighted mean plateau 202.3 ± 8.7 ka (2 ) 207.0 ± 12.4 ka (2σ) 300 2 separate experiments 0.0035 40 36 Ar/ Ari = 294.5 ± 1.8 Ar MSWD = 0.89 40 200 0.0030 n = 14 of 15 Ar/ 36 100 0.0025 Apparent age (ka) Apparent age 0 0.0020

1000 0.0040 SEG 04 31 dacite whole rock SEG 04 31 weighted mean plateau 318.3 ± 30.1 ka (2σ) 306.4 ± 54.6 ka (2σ) 750 0.0035 1 experiment 40 39 Ar Ar/ Ari = 296.1 ± 2.3 40 MSWD = 0.33 500 0.0030 Ar/ n = 6 of 7 36

250 0.0025 Apparent age (ka) Apparent age 0 0.0020 0 20406080100 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Cumulative % 39Ar released 39Ar/40Ar Figure 5. Representative 40Ar/39Ar age spectrum and inverse isochron diagrams for fi ve Seguam lavas and one ignimbrite. The preferred ages and ±2σ errors in bold are given by the plateau steps in the age spectra.

810 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island

TABLE 1. SUMMARY OF 40Ar/39Ar INCREMENTAL-HEATING EXPERIMENTS

Sample Map SiO2 Latitude Longitude Material No. of Age spectrum Inverse isochron analysis

39 ‡ 40 36 unit (%) (N) (W) expts. Plateau age Increments Ar N MSWD Total fusion Ar/ Ari Age (ka) MSWD (ka)† ±2σ used % age (ka) ±2σ ±2σ ±2σ (°C)

Postcollapse activity (younger than 9 ka) SEG 03 32§ rcv 70.8 52°21.07′ 172°25.02′ gm 3 7.5 ± 2.0 725–1450 99.8 15 of 16 0.04 8.4 ± 3.2 296.5 ± 2.4 6.1 ± 3.5 0.05 SEG 03 44§ dig 64.5 52°21.07′ 172°25.02′ wr/gm 4 8.4 ± 1.5 725–1275 100.0 19 of 19 0.14 8.9 ± 2.1 298.2 ± 3.7 7.2 ± 3.2 0.21 Island-wide activity (138–9 ka) SEG 04 34 nsl 58.4 52°21.61′ 172°31.45′ wr 1 11.1 ± 3.5 750–1240 100.0 7 of 7 0.39 12.7 ± 6.7 296.7 ± 3.1 9.7 ± 4.7 0.34 SEG 03 67 ssb 55.2 52°17.20′ 172°24.00′ gm 2 12.1 ± 5.1 840–1300 99.2 10 of 11 0.07 12.4 ± 6.4 296.0 ± 2.5 10.3 ± 9.4 0.23 SEG 03 02 fcf 62.0 52°22.66′ 172°23.50′ gm 2 22.8 ± 5.1 850–1300 100.0 9 of 9 0.75 21.6 ± 6.3 295.0 ± 4.3 27.0 ± 14.0 0.22 SEG 03 66§ lpd 62.8 52°17.00′ 172°24.20′ wr 2 23.5 ± 5.8 875–1230 100.0 9 of 9 0.32 24.6 ± 8.5 287.0 ± 69.0 36.0 ± 34.0 0.55 SEG 04 05 wvl 66.9 52°18.68′ 172°19.61′ wr 1 27.5 ± 1.7 865–1275 100.0 7 of 7 0.14 27.4 ± 2.1 295.6 ± 8.4 27.5 ± 2.0 0.17 SEG 04 07 wvl 64.4 52°18.68′ 172°19.61′ gm 1 27.9 ± 7.7 800–1310 100.0 10 of 10 0.27 33.3 ± 17.7 296.7 ± 3.1 24.2 ± 11.8 0.24 SEG 03 03§ rdf 67.9 52°22.50′ 172°23.35′ wr 2 31.7 ± 1.2 900–1375 100.0 10 of 10 0.72 31.8 ± 1.4 295.6 ± 3.6 31.8 ± 2.1 0.80 SB88–3 rdf 67.1 52°22.35′ 172°25.90′ wr 2 31.7 ± 2.0 850–1375 100.0 11 of 11 0.84 31.6 ± 2.5 294.7 ± 8.7 32.1 ± 4.1 0.01 SB87–56# ssl 70.1 52°16.01′ 172°31.33′ wr/gm 4 33.2 ± 0.9 900–1300 96.6 25 of 27 0.34 31.9 ± 1.1 295.0 ± 4.2 34.0 ± 1.9 0.80 SEG 04 24 nsl 60.8 52°18.11′ 172°37.13′ gm 2 41.9 ± 4.1 780–1260 100.0 15 of 15 0.22 43.3 ± 4.9 296.9 ± 3.4 38.8 ± 8.5 0.32 SEG 03 40 aml 61.8 52°15.95′ 172°35.18′ wr/gm 3 43.3 ± 5.7 675–1300 100.0 15 of 15 0.01 43.4 ± 6.2 293.5 ± 3.5 52.0 ± 16.0 0.32 SB87–9# ssl 53.5 52°15.90′ 172°31.34′ gm 2 48.9 ± 7.9 800–1260 100.0 12 of 12 0.49 54.0 ± 11.0 296.7 ± 2.4 40.7 ± 16.2 0.01 SB88–23# wvl 54.1 52°21.35′ 172°20.80′ gm 4 49.2 ± 7.6 825–1200 96.3 17 of 20 0.56 50.1 ± 8.6 295.9 ± 2.1 47.1 ± 13.3 0.26 SB87–63# ssl 53.3 52°15.95′ 172°35.18′ wr/gm 3 52.9 ± 13.7 850–1100 81.7 13 of 18 0.82 55.0 ± 16.0 294.2 ± 2.0 82.7 ± 37.7 0.75 SEG 03 45§ wvl 62.4 52°20.91′ 172°22.51′ wr 2 53.0 ± 1.3 850–1300 100.0 11 of 11 1.15 53.1 ± 2.3 294.9 ± 4.1 53.2 ± 2.1 0.42 SEG 03 25 ssl 60.5 52°17.20′ 172°27.02′ wr 2 53.4 ± 3.8 875–1300 100.0 10 of 10 0.02 53.3 ± 4.3 294.8 ± 3.6 54.1 ± 5.2 0.13 SB88–25 wvl 63.2 52°21.35′ 172°20.80′ wr 2 56.2 ± 1.4 850–1320 98.5 9 of 10 0.62 58.7 ± 2.5 296.1 ± 4.7 55.5 ± 3.9 0.87 SB88–18# afp 63.4 52°22.70′ 172°24.80′ gm 2 57.7 ± 5.3 900–1250 96.9 8 of 9 0.82 63.5 ± 8.9 295.4 ± 1.2 58.6 ± 9.5 0.15 SB88–16# awf 55.6 52°23.40′ 172°26.80′ gm 4 61.4 ± 5.9 920–1240 91.3 18 of 23 1.53 58.1 ± 5.1 295.2 ± 1.4 63.4 ± 10.4 1.40 SEG 03 04 afp 57.6 52°23.20′ 172°25.60′ wr 2 62.0 ± 6.5 875–1300 100.0 11 of 11 0.05 61.2 ± 8.2 294.8 ± 3.1 66.0 ± 17.0 0.02 SEG 03 64 ssl 58.8 52°16.71′ 172°28.27′ wr 2 64.4 ± 2.2 875–1330 97.2 11 of 12 0.86 65.4 ± 2.4 294.0 ± 10.7 65.3 ± 4.2 0.35 SEG 03 48§ wvl 52.2 52°20.54′ 172°21.68′ gm 3 65.6 ± 13.7 850–1275 100.0 16 of 16 0.06 72.6 ± 16.6 296.7 ± 3.1 40.0 ± 29.0 1.16 SEG 03 68 ssl 52.1 52°16.98′ 172°27.91′ gm 3 67.0 ± 12.6 900–1300 99.9 12 of 13 0.16 67.9 ± 13.7 296.3 ± 9.1 63.6 ± 34.9 0.44 SEG 03 01§ rfc 70.4 52°22.83′ 172°23.81′ wr 2 76.8 ± 1.1 875–1375 100.0 10 of 10 1.20 76.8 ± 1.2 295.4 ± 5.5 76.8 ± 1.5 0.55 SEG 04 36 nsl 56.4 52°21.44′ 172°32.83′ wr 2 78.2 ± 9.6 820–1150 94.0 14 of 16 0.05 90.5 ± 16.5 296.0 ± 1.4 73.6 ± 16.3 0.57 SEG 03 12 ssl 54.7 52°15.69′ 172°31.08′ gm 3 79.7 ± 6.8 725–1300 100.0 16 of 16 0.05 79.7 ± 8.3 294.1 ± 3.2 86.3 ± 17.1 0.22 SB87–4 ssl 58.5 52°16.55′ 172°29.60′ wr 2 82.8 ± 2.4 900–1350 99.7 10 of 11 0.02 83.1 ± 5.1 295.3 ± 3.1 82.9 ± 3.6 0.40 SEG 04 35 nsl 52.9 52°21.59′ 172°32.27′ wr 2 82.9 ± 3.6 835–1270 100.0 13 of 13 0.65 84.8 ± 5.2 296.8 ± 2.5 80.7 ± 5.5 0.08 SEG 03 21 ssl 64.1 52°17.33′ 172°26.04′ wr 2 83.6 ± 1.6 825–1275 100.0 10 of 10 0.99 83.6 ± 2.3 295.3 ± 14.3 84.0 ± 2.4 1.40 SEG 03 49§ wvl 52.4 52°20.58′ 172°21.78′ gm 4 84.6 ± 14.2 875–1250 92.6 16 of 17 0.35 79.1 ± 17.2 295.3 ± 3.3 82.0 ± 45.0 0.23 SEG 03 23 ssl 64.0 52°17.30′ 172°26.26′ gm 2 84.7 ± 3.6 850–1325 100.0 12 of 12 0.33 85.0 ± 4.8 294.7 ± 4.6 85.9 ± 9.3 0.21 SJ87–47# ssl 68.6 52°17.80′ 172°25.75′ gm 2 92.8 ± 2.8 875–1325 100.0 10 of 10 0.85 93.3 ± 3.6 296.1 ± 1.8 90.4 ± 7.3 0.16 SB87–49# aml 56.8 52°15.95′ 172°35.18′ gm 3 93.1 ± 9.5 880–1125 71.7 16 of 24 1.03 74.7 ± 12.4 296.3 ± 1.6 75.6 ± 27.0 0.70 SEG 03 50§ wvl 52.5 52°20.04′ 172°21.55′ gm 3 98.1 ± 18.5 940–1250 100.0 12 of 12 0.24 105.1 ± 24.1 296.7 ± 4.9 84.0 ± 71.0 0.03 SEG 04 38 nsl 62.8 52°21.26′ 172°33.53′ wr 1 98.8 ± 3.5 750–1250 100.0 8 of 8 0.55 99.4 ± 4.0 295.7 ± 3.7 98.4 ± 6.9 0.64 SEG 04 20 ssl 54.8 52°16.65′ 172°37.75′ wr 2 112.9 ± 15.9 785–1200 90.2 14 of 15 0.06 102.3 ± 22.1 295.7 ± 2.3 111.0 ± 37.0 0.22 SEG 03 35§ ssl 54.7 52°15.43′ 172°33.10′ gm 2 116.5 ± 13.6 875–1325 100.0 11 of 11 0.26 115.3 ± 16.0 296.3 ± 18.2 111.7 ± 84.3 0.30 SEG 04 17 nsl - 52°29.16′ 172°33.10′ gm 1 120.1 ± 9.8 840–1220 87.9 9 of 10 1.18 106.5 ± 9.5 296.7 ± 5.2 115.7 ± 21.5 1.31 SEG 04 15 nsl - 52°22.07′ 172°29.39′ wr 1 121.5 ± 6.1 800–1250 100.0 6 of 6 0.23 120.4 ± 9.9 294.8 ± 4.3 123.3 ± 12.6 0.26 SEG 03 16 ssl 67.5 52°16.24′ 172°30.26′ wr 2 121.6 ± 1.3 875–1225 97.0 10 of 12 0.94 121.1 ± 1.6 284.0 ± 26.0 122.9 ± 2.6 0.00 SEG 04 21 ssl 61.4 52°16.65′ 172°37.75′ gm 2 121.8 ± 12.7 825–1320 100.0 17 of 17 0.01 125.5 ± 29.2 296.1 ± 1.3 114.8 ± 20.3 0.90 SEG 04 16 nsl 58.6 52°22.10′ 172°29.42′ wr 1 123.2 ± 2.7 800–1250 100.0 7 of 7 0.09 123.2 ± 3.3 294.4 ± 12.2 123.5 ± 4.4 0.10 SEG 04 33 nsl 61.7 52°20.11′ 172°34.49′ wr 1 132.5 ± 3.7 790–1280 100.0 8 of 8 0.15 132.1 ± 4.9 294.9 ± 6.1 133.0 ± 6.5 0.17 SEG 04 18 nsl 54.2 52°22.16′ 172°29.83′ gm 1 133.3 ± 8.4 790–1170 87.7 7 of 8 0.25 145.4 ± 11.5 296.7 ± 2.5 127.9 ± 13.7 0.09 SEG 03 36 ssl 67.9 52°15.43′ 172°33.10′ wr 2 138.4 ± 1.3 875–1325 100.0 12 of 12 0.37 138.2 ± 1.5 296.5 ± 8.4 138.3 ± 2.7 1.30 Older, eroded volcanics (318–142 ka) SEG 03 43§ tpr 68.7 52°15.43′ 172°33.10′ wr 2 141.9 ± 2.2 950–1250 96.2 10 of 12 0.00 144.1 ± 2.7 293.2 ± 9.5 143.3 ± 6.2 0.36 SEG 04 27 nsr 67.3 52°18.36′ 172°36.32′ wr 1 159.2 ± 2.2 780–1280 100.0 8 of 8 0.91 160.5 ± 2.9 297.6 ± 2.3 158.0 ± 2.6 0.55 SEG 03 34 eld 55.1 52°15.37′ 172°33.11′ gm 3 174.3 ± 4.8 715–1320 100.0 16 of 16 1.08 173.0 ± 5.9 295.8 ± 2.8 173.0 ± 14.0 0.67 SEG 04 41 eld 54.4 52°15.73′ 172°37.78′ gm 2 175.7 ± 9.2 775–1245 100.0 19 of 19 0.11 190.6 ± 31.6 295.7 ± 1.4 173.3 ± 15.4 0.44 SEG 04 43 eld 51.7 52°15.16′ 172°36.40′ gm 2 188.5 ± 11.6 760–1235 100.0 14 of 14 0.41 188.6 ± 15.5 295.8 ± 2.1 186.8 ± 17.6 0.17 SEG 03 74 eld 50.5 52°18.25′ 172°22.50′ gm 5 190.6 ± 38.5 825–1340 91.4 23 of 27 0.45 176.0 ± 40.0 294.7 ± 2.3 224.0 ± 100.0 0.49 SEG 04 29 eld 66.7 52°19.19′ 172°35.38′ wr 1 199.1 ± 1.8 790–1300 100.0 8 of 8 0.15 199.3 ± 2.3 297.0 ± 8.0 198.7 ± 2.6 0.15 SEG 04 42 eld 53.6 52°15.16′ 172°36.40′ gm 2 202.3 ± 8.7 760–1250 99.8 14 of 15 0.18 197.7 ± 11.8 294.5 ± 1.8 207.0 ± 12.4 0.89 SB87–59 eld 50.5 52°15.95′ 172°33.97′ gm 3 209.5 ± 55.7 975–1250 77.8 11 of 13 1.50 174.9 ± 51.2 296.0 ± 2.0 161.0 ± 84.0 0.70 SB87–39 eld 54.9 52°18.45′ 172°22.25′ gm 2 210.6 ± 42.4 940–1270 96.6 9 of 11 0.00 198.8 ± 51.9 293.1 ± 9.1 344.0 ± 290.0 1.30 SEG 04 30 eld 52.8 52°19.19′ 172°35.38′ wr 1 235.8 ± 3.8 785–1285 100.0 8 of 8 0.43 234.4 ± 4.8 292.3 ± 5.6 238.3 ± 5.8 0.29 SEG 03 41 eld 52.6 52°16.13′ 172°34.55′ wr 4 281.1 ± 33.0 725–1250 100.0 19 of 19 0.46 299.5 ± 34.5 299.0 ± 10.0 168.0 ± 160.0 0.05 SEG 04 31 eld 66.5 52°19.19′ 172°35.38′ wr 1 318.3 ± 30.1 800–1150 95.8 6 of 7 0.32 295.9 ± 101.2 296.1 ± 2.3 306.4 ± 54.6 0.33

Note: Abbreviations: gm—groundmass; wr—whole rock. †Ages calculated relative to 1.194 Ma Alder Creek Rhyolite sanidine (Renne et al., 1998); uncertainties reported at 2σ precision. ‡N = number of plateau/isochron steps used in regression. §Data from Jicha et al. (2005). #Data from Jicha et al. (2004).

Geological Society of America Bulletin, July/August 2006 811 Jicha and Singer

TABLE 2. REPRESENTATIVE WHOLE-ROCK MAJOR- AND TRACE-ELEMENT COMPOSITIONS OF SEGUAM ISLAND LAVAS Older Dacitic Postcollapse Moundhill eroded lavas Shoreline lavas Wilcox volcano ignimbrite evolved lavas Pyre Peak lavas volcano Sample 03 74 04 42 04 35 03 64 04 38 03 50 03 47 03 45 03 44 03 30 03 31 03 29 03 19 04 32 04 02 04 09 (wt%)

SiO2 50.54 53.60 52.90 58.81 62.80 52.50 57.21 62.36 64.51 69.83 71.39 51.51 53.54 59.80 51.80 51.60

TiO2 0.63 0.86 0.77 1.12 1.04 0.69 1.11 0.90 0.75 0.53 0.48 0.65 0.79 0.82 0.75 0.73

Al2O3 20.19 18.30 18.40 15.51 15.40 20.89 15.69 15.16 15.54 14.77 13.73 18.66 19.52 16.60 20.40 20.40 FeO* 7.87 9.28 9.39 9.50 8.39 7.12 9.79 6.66 5.53 3.90 3.48 8.84 8.44 7.07 8.68 8.55 MnO 0.13 0.16 0.16 0.15 0.16 0.13 0.16 0.14 0.12 0.11 0.10 0.15 0.14 0.14 0.14 0.14 MgO 6.48 4.74 5.68 3.16 1.75 3.62 3.30 2.08 1.40 0.66 0.52 7.04 4.06 3.06 5.26 5.33 CaO 11.71 9.32 9.97 6.89 4.71 10.48 7.53 4.73 3.94 2.52 2.12 10.42 10.51 6.07 10.42 10.35

Na2O 2.29 3.18 2.71 3.51 4.69 2.79 3.50 4.33 4.59 5.17 4.97 2.42 2.92 4.14 2.64 2.60

K2O 0.30 0.54 0.53 1.15 1.33 0.65 0.87 1.53 1.80 2.13 2.13 0.43 0.62 1.37 0.47 0.47

P2O5 0.08 0.13 0.10 0.15 0.21 0.12 0.14 0.19 0.17 0.12 0.10 0.08 0.09 0.18 0.09 0.10 Total 100.08 100.11 100.61 99.65 100.48 100.22 99.18 98.60 99.18 99.91 99.16 99.73 100.23 99.24 100.65 100.28 (ppm) Cs 0.54 0.29 0.35 0.87 2.09 0.45 2.05 1.62 2.58 4.21 4.00 0.85 1.13 2.40 0.24 0.77 Sc 33 39 32 35 24 27 35 23 19 14 13 33 36 24 31 30 V 220 270 274 313 99 281 334 91 71 12 12 229 277 133 245 236 Cr – 34 66 – 24 29 – – – – – – – 31 38 35 Co 26 28 30 19 10 17 20 10 8 3 2 30 19 12 32 29 Ni 61 33 53 5 85 15 6 8 4 4 3 57 22 30 41 33 Cu 83 72 110 103 24 66 122 42 28 21 20 98 80 45 49 65 Zn 38 79 73 66 91 45 64 68 64 67 68 47 51 78 66 67 Ga 17 20 17 19 19 18 19 18 18 18 17 17 18 17 22 19 Rb710133032142038425551111433910 Sr 315 291 298 252 257 376 303 223 224 165 132 306 323 265 358 354 Zr 26 57 52 95 121 45 67 134 144 177 179 33 43 119 46 48 Nb 1.10 1.00 1.00 2.19 4.00 1.42 1.85 2.97 3.26 3.62 3.47 0.95 1.07 5.00 – – Ba 159 238 202 412 465 258 324 551 646 738 727 175 219 459 214 206 Y 11191723331419322835341114311314 La 3.17 5.41 4.00 7.77 9.46 5.40 7.06 12.14 11.14 14.90 14.38 3.56 4.41 9.40 4.24 4.12 Ce 7.37 11.80 9.42 17.73 22.10 12.14 15.89 27.25 24.73 33.22 32.60 8.16 10.15 21.70 9.65 9.51 Pr 1.05 1.85 1.33 2.47 3.12 1.63 2.16 3.69 3.16 4.28 4.17 1.14 1.39 2.98 1.35 1.32 Nd 5.35 9.27 6.61 11.79 15.20 7.94 10.64 17.51 14.72 19.75 18.79 5.72 6.84 13.90 6.41 6.29 Sm 1.64 2.69 1.97 3.52 4.25 2.25 3.01 4.94 4.03 5.23 5.14 1.72 2.06 3.79 1.79 1.77 Eu 0.69 0.97 0.74 1.12 1.31 0.86 1.07 1.37 1.37 1.32 1.27 0.69 0.82 1.13 0.73 0.71 Gd 1.88 3.23 2.38 3.83 4.87 2.37 3.41 5.19 4.30 5.47 5.38 1.98 2.34 4.31 2.08 2.06 Tb 0.35 0.53 0.39 0.70 0.79 0.42 0.60 0.95 0.82 1.02 1.03 0.35 0.43 0.69 0.34 0.33 Dy 2.17 3.62 2.74 4.29 5.45 2.53 3.66 5.73 4.99 6.14 6.15 2.20 2.69 4.74 2.34 2.28 Ho 0.45 0.76 0.57 0.91 1.14 0.53 0.76 1.19 1.05 1.31 1.30 0.48 0.57 0.99 0.48 0.48 Er 1.38 2.19 1.67 2.79 3.35 1.59 2.36 3.70 3.33 4.12 4.06 1.46 1.75 2.91 1.44 1.39 Tm 0.21 0.31 0.25 0.43 0.49 0.24 0.36 0.57 0.52 0.63 0.61 0.22 0.26 0.43 0.20 0.20 Yb 1.33 2.10 1.64 2.71 3.25 1.57 2.23 3.66 3.50 4.13 4.09 1.35 1.62 2.87 1.35 1.35 Lu 0.19 0.32 0.25 0.38 0.51 0.22 0.32 0.53 0.50 0.61 0.61 0.20 0.24 0.45 0.21 0.21 Hf 1.09 1.53 1.48 3.02 3.46 1.31 2.20 3.91 4.12 4.97 5.07 1.18 1.40 3.38 1.23 1.24 Ta – 0.07 0.09 0.09 0.09 0.05 0.08 0.17 0.20 0.25 0.23 0.02 0.04 0.24 0.10 0.11 Tl – 0.04 0.05 0.15 0.17 – 0.13 0.25 0.41 0.43 0.49 0.07 0.09 0.19 0.04 0.04 Pb – 5.51 4.79 5.26 12.30 9.08 3.87 3.14 9.22 12.74 13.00 3.29 – 11.00 4.02 4.51 Th 0.65 0.88 1.10 2.52 2.72 1.14 1.63 3.12 3.49 4.23 4.16 0.80 1.03 2.63 0.96 0.96 U 0.34 0.50 0.63 1.46 1.58 0.62 0.90 1.80 2.01 2.42 2.33 0.46 0.59 1.51 0.50 0.51 Note: All samples have “SEG” prefi x. Major- and trace-element concentrations were determined on SEG 03 samples following the procedures described in Jicha et al. (2004); analyses of SEG 04 samples were performed at the Universität Göttingen, Germany, following the procedures of Hinners et al. (1998). Precision of the inductively coupled plasma–mass spectrometry (ICP-MS) data is described in Jicha et al. (2004). (–) denotes analysis was performed at or below detection limit.

812 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island 1 NE km 0 7.5 ka dome (rcv) Pyre Peak (1054 masl) 83 ka flow (ssl) olidated, crystal-rich, dacitic matrix (location Holocene basalt Pyre Peak (1054 masl) vent 1977 &1993 amme. This photograph was taken along the northern amme. vent attened fi 1977 &1993 6kaflow(srf) of the ows and domes in the west, shallowly dipping lavas center 122 ka flow (ssl) D 33 ka flow (ssl) 49 ka flow (ssl) 3 0.6 km rhyolitic cone (rcc) 80 ka flow (ssl) ows, and domes in the foreground, and 1054-m-high Pyre Peak and the 1977 and 1992–1993 vent crater in the background. in the background. Peak and the 1977 1992–1993 vent crater and 1054-m-high Pyre ows, and domes in the foreground, ows extending southward from Pyre Peak. (B) View from the northeast rim of Wilcox volcano (location X in Fig. 4) looking west- Wilcox the northeast rim of from View Peak. (B) Pyre ows extending southward from 1 142 ka dome (tpr) km 174 ka flow (eld) South rim of Wilcox volcano 0 Turf Point 202-189 ka flows (eld) 66 ka flow (wvl) 85 ka flow (wvl) 98 ka flow (wvl) C A B SW Figure 6. (A) Panorama of the southwest shoreline of the island including Turf Point. Note the older fl Point. Note the older Figure 6. (A) Turf of the island including Panorama of the southwest shoreline photo, and the Holocene to historic basalt fl southwestward, showing Holocene rhyolitic cone, fl cm light-gray dacitic pumice clasts in a poorly cons ka dacitic ignimbrite showing 10–25 9 The uppermost section of the ca. (C) welded and contains fl ka dacitic ignimbrite is moderately to strongly 9 The base of the ca. 4). (D) in Fig. Y 4, ~5 km northwest of image in C. at location Z in Figure shoreline

Geological Society of America Bulletin, July/August 2006 813 Jicha and Singer

ERUPTIVE CHRONOLOGY cut these lavas gave ages of 281, 210, and 189 ka; the northern and southern shorelines, although a the 281 and 210 ka dikes crosscut older, undated few deeply glaciated remnants have been found The 34 map units shown in Figure 4 are lavas in this sequence, whereas the 189 ka dike within the interior of the island. Lavas along grouped into three main stages of volcanism cuts through the basaltic fl ow dated at 202 ± 9 ka both shorelines dip gently away from the center based on 40Ar/39Ar age, inferred eruptive vent, (Figs. 2 and 4; Table 1). of the island, but their vent areas are unknown. and clearly defi ned superposition relationships. Deeply eroded, subhorizontal lavas and inter- Individual fl ow lobes average 15–20 m in thick- The 40Ar/39Ar geochronology agrees with stra- bedded pyroclastic rocks are exposed along ness, and the thickness of the entire sequence tigraphy at 16 different locations on the island; 2 km of the southeastern shore northeast of Lava of lavas is ~300 m. North shore lavas (nsl) vary nowhere do the age determinations violate the Point (Table 2; Figs. 4 and 6). This sequence from basaltic to dacitic compositions and give stratigraphy. These relationships as well as a consists of moderately altered basaltic and 40Ar/39Ar plateau ages from 133 ± 8 to 11 ± 3 ka, detailed description of each eruptive stage are basaltic andesitic lavas and dikes and several whereas the south shore lavas (ssl) range from discussed below. Note that the younger than small dacitic fl ows. A basalt and basaltic andes- basaltic to rhyolitic and give ages from 138 ± 1 9 ka postcollapse period of activity includes ite from this section were dated at 191 ± 39 and to 33 ± 1 ka (Figs. 2 and 4; Tables 1 and 2). potentially coeval eruptions from several vents 211 ± 42 ka, respectively. Figure 6A illustrates the agreement between the on the island, and therefore is subdivided into A recently discovered sequence of older 40Ar/39Ar geochronology and the stratigraphic three phases based on geographic location. lavas and domes crops out 2 km southwest of relationships among the south shore lavas. Saddleridge Point (Fig. 4). Several 50-m-thick Approximately 3 km northwest of Turf Point, Older, Eroded Volcanics (318–142 ka) basaltic and dacitic fl ows are capped by several the north and south shore lavas are overlain by ~10-m-thick basalt fl ows. A clast from the brecci- a 200-m-thick sequence of nearly horizontal The oldest record of subaerial volcanism ated base of one of the 50-m-thick fl ows yielded andesites and dacites. The remnants of these is preserved in three distinct locations on the a 40Ar/39Ar age of 318 ± 30 ka, the oldest material lavas (aml) have been deeply eroded into the island. It includes deeply eroded lavas and dikes we found on the island. The other fl ow lobes gave shape of an amphitheater. The lowermost lava in (eld), the north shore rhyodacite dome (nsr), and ages of 199 ± 2 and 236 ± 4 ka (Table 1). This this section gave a 40Ar/39Ar age of 93 ± 9 ka. the Turf Point rhyodacite dome (tpr). Two of the entire sequence of lavas was intruded by a mas- three localities (e.g., at the western end and along sive, 60-m-thick rhyodacite dome (nsr), which Wilcox Volcano (98–9 ka) the southeastern shore) were recognized as pre- was characterized by columnar jointing that radi- After 200 k.y. of primarily effusive eruptions serving relatively old sequences by Singer et al. ated from the core of the dome. The hexagonal of sheet-like subhorizontal lavas, the eruptive (1992a). The preserved volume at the western end columns averaged 15 cm in diameter, and were fl ux at Seguam increased and the focus of volca- of the island is composed of tens of 1–2-m-thick likely formed due to emplacement of the dome nism shifted to the east, and a 20 km3 stratocone basaltic lavas that fl owed northward from a pre- into ice. A single experiment from the dacite was constructed. Wilcox volcano consists of tens existing center likely located offshore between dome gave a 40Ar/39Ar plateau age of 159 ± 2 ka. of basaltic to rhyolitic fl ows (wvl) of variable the southwest corner of the island and Turf Point. thickness that dip 30–40° radially away from the The north-dipping fl ows are cut by several 2-m- Island-Wide Activity (138–9 ka) former central vent now occupied by a 0.6 km3 wide basaltic to andesitic dikes. Near Turf Point, rhyolitic cone (Table 2). Several of the andesitic a 0.2 km3 rhyodacite dome (tpr) intrudes nota- Shoreline Lavas (138–11 ka) to rhyolitic lavas making up the northern fl ank bly thicker (15–30 m) basaltic lavas, which also These lavas represent 46% of the total vol- (rdf, awf, afp, rfc) spread out laterally as they dip gently northward. The 40Ar/39Ar plateau age ume erupted and cover more than half the island, approached the shoreline (Fig. 4). It is unclear from the rhyodacite dome is 142 ± 2 ka, whereas ~110 km2, including the entire area between the how far the lavas extended to the south because the basaltic lavas from several different locali- eroded volcanics on the western end and the intense wave action has eroded away much of ties within this eroded section yield ages ranging southeastern shore. The majority of these gently the southern fl ank of the volcano. from 202 ± 9 to 174 ± 5 ka. The dikes that cross- dipping basaltic to rhyolitic lavas crop out along The lowest exposures of the stratocone consist of basalt fl ows intercalated with thick pyroclas- tic breccias and subvertical dikes. Several of the TABLE 3. COMPARISON OF K-Ar AND 40Ar/39Ar AGES OF SEGUAM ISLAND LAVAS fl ows in this 200-m-thick section are moderately Description K-Ar data†40Ar/39Ar data (this study) oxidized, contain chlorite and serpentinized Sample Age % 40Ar* Sample Plateau age % 40Ar*‡ Total fusion age olivine, and are weakly hydrothermally altered. (ka) ±2σ (ka) ±2σ (ka) ±2σ Nonetheless, 40Ar/39Ar ages of 98 ± 18, 85 ± 14, Basaltic andesite fl ow B87–63 1070 ± 320 2.7 SB87–63 52.9 ± 13.7 0.6 55.0 ± 16.0 and 66 ± 14 ka were obtained from stratigraphi- Basalt fl ow B87–32 1050 ± 180 4.7 SEG 03 74 190.6 ± 38.5 1.5 176.0 ± 40.0 Basalt fl ow B87–34 220 ± 100 1.8 SB87–34 §§ § cally successive basaltic lavas (Table 1). These Rhyodacite dome J87–61 200 ± 40 6.8 SEG 03 43 141.9 ± 2.2 41.1 144.1 ± 2.7 basal units are capped by numerous basaltic Dacite fl ow B87–67 180 ± 20 11.6 SEG 03 16 121.6 ± 1.3 80.0 121.1 ± 1.6 andesitic to rhyolitic fl ows that erupted from 62 Basaltic andesite fl ow J87–43 170 ± 180 0.5 SEG 03 73 ## # to 27 ka (Table 2). At 23 ± 5 ka, the eruption of Basaltic andesite fl ow B87–59 100 ± 140 1.6 SB87–59 209.5 ± 55.7 1.5 174.9 ± 51.2 ≥ 3 a 0.02 km andesitic (62.0 wt% SiO2) ignim- Dacite fl ow J87–48 80 ± 20 8.0 SEG 03 21 83.6 ± 1.6 68.1 83.6 ± 2.3 brite fi lled a valley between the Finch Cove rhy- Basaltic andesite fl ow J87–59 70 ± 80 1.0 SEG 03 12 79.7 ± 46.8 5.3 79.7 ± 8.3 Basaltic andesite fl ow B88–23 50 ± 40 5.0 SB88–23 49.2 ± 7.6 2.6 50.1 ± 8.6 olite (rfc) and Finch Cove rhyodacite (rdf). The Andesite fl ow B88–16 30 ± 20 0.9 SB88–16 61.4 ± 5.9 2.8 58.1 ± 5.1 ~25-m-thick ignimbrite is moderately welded †Data from Singer et al. (1992a). throughout and contains ~20% phenocrysts ‡% 40Ar* was calculated from all steps comprising the plateau; % 40Ar* of each step was weighted by the (12% plagioclase; 4% clinopyroxene; 2% oliv- fraction of 39Ar released. §40Ar/ 39Ar analysis not attempted on this sample. ine; 2% oxides), abundant lithics (up to 20 cm), #Sample did not yield measurable radiogenic 40Ar. and pumice (15–40 cm). Some of the last activ-

814 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island

ity at Wilcox volcano prior to the sector collapse 100 Sample SiO Postcollapse lavas at ca. 9 ka included the eruption of dacitic to 2 rhyolitic lavas, which formed Moundhill Point SEG 03 32 70.8 (Holocene-historic) and the adjacent region to the south. SEG 03 55 63.2 At 9 ka, Wilcox volcano partially collapsed forming a 4-km-wide crater. The sector col- SEG 03 11 58.6 lapse was accompanied by an explosive lat- eral blast, which deposited a 0.45 km3 dacitic SEG 03 19 53.5 ignimbrite (dig) to the northwest of the edi- SEG 03 08 51.2 fi ce. This sequence of events was likely simi-

lar to those during the May 1980 eruption of Sample/Chondrite 10 Mount St. Helens (Christiansen and Peterson, 1981) and the October 1902 eruption of Santa A María volcano in Guatemala (Williams and Self, 1983). The dacitic ignimbrite retains a 100 Wilcox volcano lavas (98-9 ka) relatively uniform thickness of 80–100 m for 4 km until it reaches the northern shoreline. Sample SiO2 Because offshore bathymetry is not available, SEG 03 03 67.9 the distal extent of the ignimbrite is unknown, but it is clearly recognizable 10 km north SEG 03 47 57.2 of the coastline in U.S. Geological Survey SEG 03 50 52.5 (USGS) GLORIA sidescan images. A small outcrop is also preserved along the southeast- SJ-88-6 51.7 ern rim (Fig. 4). The deposit is characterized

by 10–25 cm, dark-gray dacitic pumice clasts Sample/Chondrite 10

or fi amme (64.0 wt% SiO2) hosted in a lighter- gray, fi ne ash supported matrix of similar com- B position (64.9 wt% SiO2). The upper 50–70 m of the ignimbrite is nonwelded and poorly 100 consolidated, but the base is strongly welded Shoreline lavas (138-11 ka) Sample SiO2 and contains fl attened fi amme (Fig. 6). Four SB-87-56 70.0 incremental-heating experiments on whole- rock samples and a phenocryst-free separate SEG 03 40 61.8 of the poorly consolidated facies of the ignim- SEG 03 64 58.8 brite gave a weighted mean 40Ar/39Ar age of 8.4 SEG 03 12 54.7 ± 1.5 ka. A U-Th mineral isochron age for this SEG 04 35 52.9 sample (defi ned by cpx + mt + plag + glass + wr) gave an age of 10.1 ± 1.3 ka (Jicha et al.,

2005). Thus, we infer that emplacement of this Sample/Chondrite 10 ignimbrite and collapse of the occurred at 9 ± 1 ka. C Postcollapse Activity (Younger Than 9 ka) 100 Older eroded volcanics (318-142 ka) Sample SiO Dacitic to Rhyolitic Domes and Flows 2 (since 7.5 ka) SEG 04 27 67.3 Two rhyolitic fl ows, a small dome, and a rhy- SEG 04 28 67.1 olitic cone were erupted from the crater, which was created in response to partial sector collapse of the stratovolcano. In addition, dacitic and rhyodacitic lavas erupted from fl ank vents to the south of Wilcox volcano at 335 m above sea SEG 03 74 50.6 3 level (Fig. 4). The 0.6 km cone (rcc) is com- Sample/Chondrite 10 SJ-87-79 50.6 posed of vesicular lavas of similar composition, and is mantled by coarse, silicic pumice blocks D up to 50 cm in diameter. An older, tundra-cov- ered 0.08 km3 rhyolite fl ow (rcv) crops out to La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Dy Tm Yb Lu the west of the rcc cone. This plagioclase-phyric Figure 7. Chondrite-normalized rare earth element (REE) patterns for (A) postcollapse lavas, lava dips 20° to the northwest and is strongly (B) Wilcox volcano lavas, (C) shoreline lavas, and (D) older, eroded volcanics. Chondrite val- fl ow banded with alternating layers of red rub- ues are from Anders and Grevesse (1989).

Geological Society of America Bulletin, July/August 2006 815 Jicha and Singer ble and black, glassy rhyolite. Three incremen- from this dome gave a U-Th mineral isochron (Fig. 4). Its southeastern fl ank is built upon tal-heating experiments from the glassy rhyolite age of 1.7 ± 0.5 ka (Jicha et al., 2005). To the 15–20 m of andesitic lavas of unknown age, but yielded a weighted mean plateau age of 7.5 south of the dome lies a small, undated dacite the remainder of the cone rises from sea level ± 2.0 ka (Table 1; Fig. 5). The western fl ank of fl ow (lcd), which fi lled a long, narrow canyon up to 590 m. The entire cone is composed of the cone and southern margin of the rcv fl ow are and an adjacent valley. East of the dacite is the numerous 1–3-m-thick, sheet-like, chemically overlain by a small, gray, vitrophyric lava fl ow ~50-m-thick, Lava Point rhyodacite (lpr), which monotonous basalt fl ows (mvl) distinguished (srf), the most evolved on the island (71.4 wt% erupted along a 3-km-wide, east-west–trending by unusually large (up to 0.7 cm) and abundant 40 39 SiO2) (Table 2; Fig. 4). The Ar/ Ar geochro- fi ssure and fi lled three fl uvial valleys until the plagioclase (40 modal %), clinopyroxene (9%), nology was unsuccessful on this material, but fl ows merged together to form the headlands at and olivine (6%) phenocrysts in a glassy matrix whole-rock, plagioclase, clinopyroxene, mag- Lava Point (Fig. 4). The 40Ar/39Ar geochronol- (Table 2). Each fl ow exhibits pahoehoe struc- netite, and glass yielded a U-Th isochron age ogy experiments did not yield any radiogenic ture and extends from near the summit crater of 6 ± 4 ka (Jicha et al., 2005). The rcv and srf 40Ar* in this lava fl ow because it is likely late all the way to the coast. Levees on several fl ows fl ows likely erupted from the same vent located Holocene in age. descending from the west side of the crater sug- near the base of the cone. The 1710 dome (str), gest that the initial fl ow direction was westward, the westernmost outcrop of postcollapse rhyo- Moundhill Volcano (Late Holocene) but then shifted to the north or the south around lites, abuts the southern margin of the crater rim. Moundhill volcano is a postglacial monoge- the fl anks of the stratovolcano (Fig. 4). It is pos- Glass + magnetite + orthopyroxene + whole rock netic cone located at the eastern end of the island sible that the entire cone formed during a single, long-lasting eruption. Several attempts to obtain 40Ar/39Ar ages from Moundhill volcano lavas were unsuccessful due to very low radiogenic A 10 40Ar* yields. Post collapse lavas (Holocene-historic) 8 Wilcox volcano lavas (98-9 ka) Pyre Peak Lavas and Tephras (Holocene to Shoreline lavas (138-11 ka) Present) Older eroded volcanics (318-142 ka) 6 The initial phase of this late Holocene activ- ity was explosive. A vent-clearing eruption resulted in the formation of an explosion caldera Zr/Rb 4 3 × 4 km wide and produced a 3.2 km3 andesitic (58.62 wt% SiO ) ignimbrite that blankets the 2 Fractional crystallization 2 Pleistocene lavas over more than 30% of the 0 island (Fig. 4). The majority of this pyroclastic 030405010 20 60 deposit has been covered by subsequent erup- tions and effusions of basalt and basaltic andes- Rb (ppm) ite from Pyre Peak, a 0.25 km3 basaltic scoria cone, which is the highest point on the island B 5 (1054 m). Pyre Peak lavas, which partially r Fractional crystallization infi lled the caldera, fl owed primarily to the south go 4 and west and cascaded over the eroded cliffs of

n

Increased Pleistocene lavas along the southern shoreline 3 rtial meltinheterogeneity (Figs. 4 and 6). Because the Pyre Peak lavas are pa presumably only hundreds to a few thousands source 40 39

(La/Yb) 2 of years old, they cannot be dated by Ar/ Ar methods. However, the relative stratigraphic 1 order of several lavas has been determined based on fi eld observations. From oldest to youngest, 0 they are the mossy basalt (msb), basalt west 015183 6 9 12 of base camp (bwb), western fan basalt (wfb), La (ppm) and brown basalt (brb). The msb, bwb, and brb units are individual lava fl ows with a volume of Figure 8. (A) Plot of Zr/Rb versus Rb (ppm). Basalts, mainly from the earliest eruptive phase, <1 km3. The western fan basalt (wfb), the most with <12 ppm Rb display a signifi cant variability in Zr/Rb ratios, which likely refl ects small voluminous of these units, consists of 2.6 km3 variations in the degrees of partial melting or extraction and an origin from weakly het- of basaltic andesitic lavas fl ows (52.9–53.1 wt% erogeneous mantle-wedge sources with different Zr/Rb ratios. In contrast, the Zr/Rb ratios SiO2) that likely erupted from a vent near Pyre of lavas with >12 ppm Rb are remarkably constant, suggesting a fractional crystallization– Peak over a relatively short period of time. The controlled origin. Signifi cant crustal contamination is unlikely to have occurred. (B) Plot of most recent activity includes the eruption of chondrite-normalized La/Yb versus La. Seguam basalts and basaltic andesites with <6 ppm basalt and basaltic andesite in 1977 and 1992– La defi ne an inclined array, which, similar to the Zr/Rb ratios, likely refl ects various degrees 1993 (Miller et al., 1998; Masterlark and Lu, of partial melting, or melting of a heterogeneous source. The nearly constant linear trend of 2004; Price, 2004; Jicha et al., 2004), respec-

(La/Yb)n at ~3 displayed by all lavas with >6 ppm La implies that crystal-liquid fractionation tively, from a vent located 2 km to the southwest has been a common process creating rhyolite at Seguam for at least 320 k.y. of Pyre Peak (Figs. 4 and 6).

816 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island

DISCUSSION 1.5 t = ka 0 20 40 60 80 100 120 140 Crater formation K-Ar versus 40Ar/39Ar Ages 1.4

1977

o 1.3 The ages of ten of the samples determined by Undi 40 39 1993 Ar/ Ar analyses were originally determined stu using the conventional K-Ar method on large 1.2 rbed Th) (11.6–25.7 g) whole-rock samples by Singer et m agma st

232 1.1 al. (1992a). Nine of these samples yielded repro- ducible 40Ar/39Ar plateau ages ranging from 210 ora 1.0 ge to 49 ka, but one basaltic andesite lava fl ow (sam- Th/ ple J87-43) failed to yield measurable radiogenic

230 0.9 40

Ar. Incremental-heating experiments on hand- ( picked groundmass separates from the basaltic 0.8 lavas, which gave K-Ar ages of 1.07 ± 0.32 and 1.05 ± 0.18 Ma, yielded mean 40Ar/39Ar plateau 0.7 ages of 53 ± 14 and 191 ± 39 ka, respectively (Table 3). Two dacitic to rhyodacitic lavas gave 0.6 40Ar/39Ar ages that were slightly younger than 1.0 1.5 2.0 2.5 3.0 3.5 the K-Ar ages, whereas 40Ar/39Ar ages from four eλt lavas were indistinguishable at 2σ from the K- 230 232 Ar ages (Table 3). One possible explanation for Figure 9. Evolution of ( Th/ Th)0 ratios measured in Seguam Island lavas as a function of λ the younger 40Ar/39Ar ages of the two evolved eruptive age (expressed as e t) (after Jicha et al., 2005). These ratios were obtained from the lavas is that these glassy samples were affected intercepts of mineral-glass whole-rock isochrons with the equiline. The lavas that erupted 39 230 232 by Ar recoil. between 142 and 9 ka display a monotonic increase in ( Th/ Th)0 ratios, which has been The % 40Ar* determined from 40Ar/39Ar incre- interpreted to refl ect a derivation from a thermally buffered, basaltic melt lens that had mental-heating experiments on six of the nine been stored in the lower crust for 130 k.y. The abrupt change in magma composition at 9 ka 230 232 lavas was equal to or greater than those from the most likely refl ects the introduction of new basaltic magma, which had a lower ( Th/ Th)0 K-Ar experiments (Table 3). Interestingly, the % ratio, into the plumbing system. 40Ar* yields of three basaltic to basaltic andes- itic lavas dated by 40Ar/39Ar methods were lower than those reported by Singer et al. (1992a). We 80 suspect that the older K-Ar ages refl ect either inaccurate measurement of 40Ar* or K content Cumulative volume of the extremely large, coarse (500–1000 μm) Erupted (km3) whole-rock samples melted for the K-Ar analy- ses. Moreover, K-Ar ages were determined in 60 duplicate on samples J87-43, B88-23, and B88- 16 by Singer et al. (1992a), but the resulting % 40 Ar* yields and ages were not equivalent in Total each experiment. This poor level of reproduc- ibility most likely refl ected the fact that these 40 / samples, which were 50–100 times larger than alt esie Bas nd 40 39 ic A the samples used for Ar/ Ar analyses, were alt Bas internally heterogeneous with respect to K2O concentration and 40Ar content. The internal 20 reproducibility and stratigraphic consistency of Dacite 40 39 e the 60 new Ar/ Ar incremental-heating ages Andesit presented here imply that the subaerial eruptive Rhyolite history of Seguam Island is one-third as old as 0 previously inferred on the basis of the eleven K- 400 300 200 100 0 Ar ages of Singer et al. (1992a). Age (ka) Petrologic Evolution Figure 10. Cumulative eruptive volume versus time for the Seguam Island volcanic com- Singer et al. (1992a, 1992c) suggested that plex showing the total volume of each magma type erupted over the past 318 k.y. The total the monotonic variation in major- and trace- volume of erupted material is denoted by the black line. The oldest volcanism is primarily element abundances and narrow range of Sr, basaltic; however, >30% of the total eruptive volume has been dacitic to rhyolitic. If a small Nd, Pb, and O isotope compositions of the portion of rhyolitic fl ows at 318 and 200 ka are ignored, eruptions have become more com- diverse suite of magmas (49.7–71.4 wt% SiO2) positionally diverse over the past 200 k.y.

Geological Society of America Bulletin, July/August 2006 817 Jicha and Singer

75 further supporting a fractional crystalliza- Postcollapse lavas (Holocene-historic) tion–controlled origin for 320 k.y. (Fig. 8). Wilcox volcano lavas (98-9 ka) U-Th isotope data from minerals, glass, and 70 Shoreline lavas (138-11 ka) whole rocks, interpreted in light of 40Ar/39Ar Older eroded volcanics (318-142 ka) ages, and limited Sr isotope variability sug- gest rapid decompression-driven crystalliza- 65 tion and differentiation of ascending magmas following isolated storage of the parent basalt in the lower crust (Fig. 9; Jicha et al., 2005). 60 Conversely, Seguam basalts, particularly those

(wt%) of the earliest stage of volcanism from 318 to 2 142 ka, display signifi cant variability in Zr/Rb 55 and La/Yb ratios, which likely refl ects either

SiO a slightly greater range in degrees of partial melting, or derivation from a heterogeneous mantle-wedge source at this time. 50 Island-wide activity Older, eroded volcanics (318-142 ka) (138-11 ka) Compositional Trends through Time Crater formation 45 Basalts and basaltic andesites (49–56 wt% 400 350 300 250200 150 100 50 0 SiO2) make up >50% of the cumulative volume Age (ka) estimated and account for >90% of the erupted products over the initial 220 k.y. of subaerial activity. From 98 to 66 ka, basaltic eruptions Figure 11. SiO2 versus age and compositional evolution of Seguam lavas and tuffs. Error bars represent 2σ uncertainties. Although there is no overall apparent correlation between formed the base of Wilcox volcano and emanated from vents in the center of the island. Holocene age and SiO2 silica content, several important patterns emerge. Older lavas between 318 and 142 ka include virtually no andesite, whereas andesite is voluminous in the shoreline basaltic activity has been restricted to eruptions and Wilcox stratovolcano lavas of 138–9 ka. With exception of two andesitic lavas dated at from Pyre Peak and Moundhill volcano. Andesites (56–62 wt% SiO ) are far less 12 and 11 ka, lavas comprising the shoreline phase and Wilcox stratovolcano lavas tend to 2 voluminous than the basalts and basaltic andes- become more SiO2-rich with time. Compositional diversity is most extreme, 51%–71% SiO2, in the postcollapse phase. Gray bars represent the boundaries between the three main erup- ites and represent only ~16% of the total vol- tive phrases discussed in the text. Formation of the 4-km-wide crater is noted at ca. 9 ka. ume erupted (Fig. 10). Nearly all of these low- Symbols are the same as Figure 3. to medium-K lavas erupted between 133 and 22 ka. Examples include the 10–20-m-thick fl ow lobes exposed along the northern and southern shorelines (nsl and ssl), and those intercalated erupted at Seguam are consistent with an ori- rhyolites have slightly light (L) REE-enriched with stacks of basaltic and dacitic lavas on the gin via closed-system fractional crystallization patterns and large, negative Eu anomalies, fl anks of the eastern stratovolcano. Only three of basalt. Model calculations suggest that ~80 refl ecting plagioclase fractionation. In con- andesitic eruptions have occurred over the past wt% crystallization of an anhydrous mineral trast, lavas erupted from 138 to 11 ka, includ- 22 k.y., with the most notable of those being the assemblage from basalt can produce a 70 wt% ing those from Wilcox volcano and those from explosive 3.2 km3 vent-clearing eruption (58.6

SiO2 rhyolite (Singer et al., 1992a). However, vents in the center of the island, display slight wt% SiO2) associated with caldera formation on the observations of Singer et al. (1992a, 1992b, LREE-enriched patterns and only weak Eu the western half of the island.

1992c) were based primarily on the geochemical anomalies. Postcollapse lavas have the great- Dacites (62–69 wt% SiO2) are the sec- and isotopic compositions of the older, eroded est LREE enrichment of all Seguam lavas and ond most abundant lava type found on the lavas and the 138–33 ka lavas along the south- show strong positive and negative Eu anomalies island (23% of the total eruptive volume). The ern shoreline. In light of the 60 new 40Ar/39Ar similar to those of the older eroded volcanics. 40Ar/39Ar geochronology suggests that dacitic ages, and detailed fi eld mapping and sampling, The parallel nature of the REE patterns within activity extended as far back as 159 ka, when a plus new inductively coupled plasma–mass each group of lavas provides compelling evi- dacitic dome intruded older, basaltic lavas and spectrometry (ICP-MS) major- and trace-ele- dence for crystallization-dominated evolution pyroclastic rocks. Interestingly, most dacites, ment analyses (Table 2; electronic Table DR2 (Singer et al., 1992a). The subtle changes in the like the andesites, erupted between 133 and [see footnote 1]), we are now able to investigate slope of the REE patterns for each of the three 20 ka; a period in which all eruptions were tap- geochemical and isotopic variations throughout stages of volcanism suggest that a new basaltic ping a basaltic melt lens located in the lower the entire suite of erupted lavas. parent magma became available as a parent to crust (Figs. 10 and 11; Jicha et al., 2005). How- Rare earth element (REE) patterns of basal- the crystal-liquid fractionation processes at the ever, several km3 of dacite were generated over tic to rhyolitic lavas are distinctly different dur- beginning of each of these periods. the past 20 k.y., including eruptions from the ing each of the main stages of volcanism over Incompatible trace-element ratios (i.e., stratovolcano that fl owed to the southwest of the past 320 k.y. (Fig. 7). From 318 to 142 ka, Zr/Rb and La/Yb) of Seguam basaltic andes- Moundhill Point, the crater-forming eruption at basalts have nearly fl at REE patterns and strong ites to rhyolites are remarkably constant over 9 ka, and the postcollapse fl ank eruption of the positive Eu anomalies, whereas dacites and large ranges in SiO2, La, and Rb contents, Long Canyon dacite (lcd).

818 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island

3 Rhyolites (>69 wt% SiO2) amount to 6 km , TABLE 4. VOLUME ESTIMATES FOR SEGUAM LAVAS AND TEPHRAS <8% of the total volume of products erupted Stage/eruptive unit Map label SiO2 range of Age range of Present volume Total volume over the past 318 k.y. From 133 to 33 ka, several products activity (km3) (%) small fl ows of rhyolite crop out as part of the Moundhill volcano thick sequence of widely scattered lavas along Moundhill volcano lavas mvl 51.6–51.8 Holocene 1.4 2 the southern shoreline, and the 0.4 km3 Finch Cove rhyolite erupted from the Wilcox strato- Pyre Peak lavas and tephras volcano at 77 ka. Rhyolitic output increased at Basaltic ash and scoria bas 51.7–53.6 Holocene-historic 0.7 Bowling ball basalt bbb 51.5–51.7 1977 eruption 0.06 ca. 30 ka as eruptions occurred near Turf Point Dacitic pumice air fall dpa 64.0–64.2 Holocene-historic 0.06 and rhyolites descended from the stratocone Andesitic ignimbrite aig 58.6–59.8 Holocene-historic 3.2 forming Moundhill Point (Fig. 4). Over the past Right-angle basalt rab 51.9–52.1 Holocene-historic 0.01 9 k.y., a 0.6 km3 rhyolitic cone was constructed, Brown basalt brb 52.4–53.6 Holocene-historic 0.01 and several rhyolitic lavas erupted from inside Western fan basalt wfb 52.9–53.1 Holocene-historic 2.6 the crater and along the southern fl anks of Basalt west of base camp bwb 51.7–51.9 Holocene-historic 0.6 Mossy basalt msb 51.6–52.7 Holocene-historic 0.7 the volcano. Most of the rhyolitic activity has Andesite in amphitheater aam 59.0–59.3 Holocene 0.01 occurred over the past 30 k.y., and ignoring a 8.0 10 pair a small volume fl ows at 318 and 200 ka, Postcollapse dacites and rhyolites eruptions have become more compositionally Long Canyon dacite lcd 62.9–66.4 <7.5 <0.01 diverse over the past 200 k.y. (Fig. 11). Lavas Lava Point rhyodacite lpr 68.5–69.5 <7.5 0.4 erupted following creation of the crater span 1710 rhyolite dome str 69.8–69.9 1.7 ± 0.5 0.06 Spiny rhyolite srf 71.4–71.6 6 ± 4 0.01 a larger compositional range, 51%–71% SiO2, than any previous phase of activity (Fig. 11). Rhyolitic cone in caldera rcc 69.8–69.9 <7.5 0.6 Rhyolite fl ow in caldera valley rcv 69.9–70.8 7.5 ± 2.0 0.08 1.2 <1 Time-Volume Relationships Sector collapse/crater formation Dacitic ignimbrite dig 64.5–67.9 8.4 ± 1.5 0.45 <1 Eruptive rates over periods of thousands of Wilcox volcano years can be estimated using a detailed chronol- South shore basaltic andesite ssb 55.2–55.5 12.1 ± 5.1 0.2 ogy combined with volumetric estimates based Finch Cove ash-fl ow tuff faf 62.0–66.6 22.8 ± 5.1 0.02 on the thickness and lateral extent of the erupted Lava Point dacite lpd 62.7–62.8 23.5 ± 5.8 0.2 Finch Cove rhyodacite rdf 67.9–68.3 31.7 ± 2.0 0.6 units (Table 4). The preserved subaerial volume Finch Point andesite afp 57.6–58.2 57.7 ± 5.3 0.9 at Seguam represents only a fraction of the total Andesite west of Finch Point awf 55.6–57.7 61.4 ± 5.9 0.6 extruded volume because a signifi cant percent- Finch Cove rhyolite rfc 70.2–70.4 76.8 ± 1.1 0.4 age of the erupted products has been deposited Wilcox volcano lavas wvl 52.1–69.9 98.1–27.5 17.4 in the sea or been removed via marine erosion, 20.3 26 which peaked during interglacial highstands Shoreline lavas at ca. 10, 120, 200, and 330 ka (Shackleton Amphitheater lavas aml 54.1–65.9 <93.1 0.2 South shore lavas ssl 52.1–70.1 138.4–33.2 19.6 et al., 1990). Moreover, glacial ice and explo- North shore lavas nsl 52.9–65.2 133.3–11.1 16.0 sive events have carried additional material 35.8 46 away from the island. Therefore, the calculated Older eroded volcanics magma output rates for Seguam discussed next Turf Point rhyodacite dome tpr 68.6–68.7 141.9 ± 2.2 0.2 represent minimum estimates. North shore rhyodacite dome nsr 66.7–67.3 159.2 ± 2.2 0.1 The time-averaged eruptive rate for the 79 km3 Deeply eroded lavas and dikes eld 51.7–55.1 318.3–174.3 11.8 complex is 0.25 km3/k.y., although the eruptive 12.1 15 Total volume 79.2 fl ux has been highly variable throughout the lifetime of the volcano. Basalts and basaltic andesites and a few dacitic domes erupted from 318 to 142 ka comprise 15% of the total volume of the complex, which corresponds to an erup- the removal of volcanic material during The The time-averaged eruptive rate for this period tive rate of 0.07 km3/k.y. This sequence of lavas, Last Glacial Maximum in the Aleutians from is 1.2 km3/k.y., which represents the highest dikes, and domes is deeply eroded; therefore, ca. 25 to 12 ka (Fig. 10; Black, 1983; Shackle- magma output rate during the 318 k.y. subaerial the actual eruptive rate during this time period ton et al., 1990). history of the volcano. Given that most of the may have been signifi cantly higher. From 100 Postcollapse (younger than 9 ka) activity is basaltic activity from Pyre Peak is younger than to 25 ka, the eruptive rate increased to ~0.6 km3/ fairly well preserved, allowing for more accurate the silicic lavas and domes that erupted from 9 k.y., corresponding to stratovolcano growth volume estimates, but several of the lavas are to 1.7 ka, the eruptive rate for Pyre Peak lavas coupled with the continued eruptions of basaltic covered by recent eruptions of basaltic ash and could be as high as 4.7 km3/k.y. to rhyolitic lavas from vents in the center of the scoria. The total volume for all of the evolved island. For 15 k.y. prior to collapse of the east- (e.g., dacitic and rhyolitic) lavas and domes Comparison to Other Arc Volcanoes ern stratovolcano, activity waned slightly, which plus all of the basalts and basaltic andesites may represent a period of quiescence. However, erupted from Pyre Peak and Moundhill volcano The long-term eruptive rates at Seguam can this likely refl ects a lack of preservation due to is 10.6 km3, or 13% of the total volume erupted. be compared to those of other well-dated arc

Geological Society of America Bulletin, July/August 2006 819 Jicha and Singer

100 volcanoes (i.e., Mount Adams, Tatara–San Pedro, Santorini, Montserrat, Mount Baker, Kat-

iation Katmai ximum mai, and Ceboruco–San Pedro volcanic fi eld) (Fig. 12). Because several of the volcanic com-

lacial ma

age 6 glac

)

St Stage 8 glaciation plexes in these detailed studies cover large areas

3 80 Stage 10 glaciation 2

Last g (400–1600 km ) encompassing a stratocone(s) and peripheral lavas, we can make comparisons to the eruptive rates of the individual stratocones as well as to the entire volcanic fi eld. 60 The time-averaged eruptive rate at Seguam (0.25 km3/k.y.) is similar to 90–20 ka Volcán Seguam Tatara in the Tatara–San Pedro complex in the Chilean Southern volcanic zone (Singer et al., 1997) and the 940 ka Mount Adams volca- 40 nic fi eld (Hildreth and Lanphere, 1994) in the Cascades, but slower than those estimated for Santorini in the South Aegean arc (Druitt et al., 1999), Mount Baker and Mount Adams strato- cones in the Cascades (Hildreth et al., 2003a; Cumulative volume (km volume Cumulative 20 Mt. Baker Hildreth and Lanphere, 1994), and Mount Kat- Mt. Adams Ceboruco mai and Mount Mageik on the Alaska Peninsula (Hildreth et al., 2003b) (Table 5). The reasons Tatara-San Pedro for these differences are not well understood. 0 Several factors may contribute to the contrast- 400 300 200 100 0 ing eruptive rates, including the variable crustal thicknesses beneath each of the volcanoes or Age (ka) differences in the magma production rate for each arc. The average eruptive rate at Seguam Figure 12. Minimum cumulative volume versus time for Seguam Island and other well- over the past 9 k.y. was 1.2 km3/k.y., with pos- dated Pleistocene to Holocene arc volcanoes. Data are restricted to the volume erupted at sible rates as high as 4.7 km3/k.y. for the period each volcano over the last 400 k.y. Data from Katmai volcanic cluster encompass seven fron- of basaltic volcanism from Pyre Peak. Although tal arc volcanoes and fi ve rear-arc monogenetic cones along 100 km of the Alaska Peninsula not representative of the entire 318 k.y. sub- (Hildreth et al., 2003a). Other volcanoes include Mount Adams (Hildreth and Lanphere, aerial record, these more rapid rates of magma 1994), Tatara–San Pedro (Singer et al., 1997), Mount Baker (Hildreth et al., 2003b), and output in the Holocene at Seguam are compa- Ceboruco–San Pedro (Frey et al., 2004). Gray bars indicate the timing and duration of rable to (1) the average Aleutian eruptive rates major glaciations that have occurred over the past 400 k.y. (Shackleton et al., 1990). estimated by Crisp (1984) (2.1 km3/k.y.) and Marsh (1982) (1.6 km3/k.y.), (2) the main pulses of stratocone growth at Mount Adams (2–5 km3/ TABLE 5. SUMMARY OF TIME-AVERAGED ERUPTIVE RATES AT LONG-LIVED, WELL-DATED ARC 3 VOLCANOES k.y.), and (3) the construction of 79 km at Volcano Arc Crustal Volume Duration Average Reference Mount St. Helens, which has only been active thickness (km3) (k.y.) eruptive rate for ~40 k.y. (~2 km3/k.y.; Sherrod and Smith, (km) (km3/k.y.) 1990). Even though the magma output rate from Seguam Aleutians 25–30 79 318 0.25 This study Pyre Peak is comparable to that of several other Mt. Katmai Aleutians 30–36 70 89 0.79 Hildreth et al. (2003b) arc volcanoes, it is almost an order of magnitude Mt. Mageik Aleutians 30–36 30 93 0.32 Hildreth et al. (2003b) less than the eruptive fl ux at mid-ocean-ridge Entire Katmai cluster Aleutians 30–36 179 292 0.61 Hildreth et al. (2003b) 3 Tatara–San Pedro SVZ, Andes 30–35 55 930 0.06 Singer et al. (1997) spreading centers (24 km /k.y. per 400 km of Volcán Tatara SVZ, Andes 30–35 22 91 0.24 Singer et al. (1997) ridge segment) (Crisp, 1984), and the average Mt. Adams volcanic fi eld Cascades 40–45 231 940 0.25 Hildreth and Lanphere eruptive rate of the 8–10 ka Klyuchevskoy vol- (1994) cano (32–22 km3/k.y.), which is often cited as Mt. Adams stratocone Cascades 40–45 200 520 0.38 Hildreth and Lanphere (1994) the most active island arc volcano in the world Mt. Baker volcanic fi eld Cascades 40–45 105 1300 0.08 Hildreth et al. (2003a) (Fedotov et al., 1987). Mt. Baker stratocone Cascades 40–45 15 43 0.35 Hildreth et al. (2003a) A striking feature of Figure 12 is the infl ection Santorini South Aegean 20–32 300 650 0.46 Druitt et al. (1999) of the growth rate curves for several volcanoes Montserrat Lesser Antilles 30–40 26 174 0.15 Harford et al. (2002) corresponding with the penultimate (oxygen iso- Ceboruco–San Pedro Trans-Mexican 35–40 81 819 0.10 Frey et al. (2004) tope stage 6) and last (oxygen isotope stage 2) Average arc output: Aleutians 50–70 7350 3500 2.10 Crisp (1984) Aleutians 50–70 4700 3000 1.57 Marsh (1982) major global glaciation maxima. For example, Seguam, Katmai, and Mount Adams all show Lesser Antilles 30–40 285 100 2.85 Crisp (1984) comparable large increases in growth rate fol- Note: Eruptive volumes are minimum estimates. Average eruptive rates were calculated by dividing the lowing the penultimate glaciation ca. 125 ka, erupted volume by the duration of volcanism. SVZ—Southern volcanic zone. whereas these three volcanoes, plus Mount

820 Geological Society of America Bulletin, July/August 2006 Volcanic and magmatic evolution of Seguam island

DeMets, C., Gordon, R.G., Argus, D.F., and Stein, S., 1994, Baker, Ceboruco, and Tatara–San Pedro have that of other well-dated arc volcanoes in the Effect of recent revisions of the geomagnetic reversal pronounced infl ections either corresponding Lesser Antilles arc, Chilean Southern volcanic timescale on estimates of current plate motions: Geo- to, or immediately following, the Last Glacial zone, and Trans-Mexican volcanic belt. More- physical Research Letters, v. 21, p. 2191–2194, doi: 10.1029/94GL02118. Maximum ca. 25 ka. These data suggest that the over, the average growth rate of Seguam Island Druitt, T.H., Edwards, L.E., Mellors, R.M., Pyle, D.M., repeated infl uence of highly erosive glaciers on is 40% less than that of Mount Katmai and Sparks, R.S.J., Lanphere, M.A., Davies, M., and Bar- reirio, B., 1999, Santorini volcano: Geological Society large, long-lived arc stratovolcanoes (Singer et Mount. Mageik, which are located on the Alaska of London Memoir 19, 165 p. al., 1997) may be widespread. If true, one must Peninsula, suggesting that the eruptive fl ux may Engdahl, E.R., 1977, Seismicity and plate subduction in be extremely cautious in using the preserved be higher in the continental sector of the Aleu- the central Aleutians, in Talwni, M., and Pitman, W.C., eds., Islands arcs, deep sea trenches and back- volumes of the portions of these volcanoes older tian arc. This along-arc variation was originally arc basins: Washington, D.C., American Geophysical than ca. 125 ka to constrain long-term eruptive observed by Marsh (1982), who noted that erup- Union, p. 259–271. rates in subduction zones. tive output decreased systematically from east Fedotov, S.A., Khrenov, A.P., and Jarinov, N.A., 1987, Kly- uchevskoy volcano, its activity and evolution, 1932– to west along the Aleutian arc owing to increas- 1986: Vulkanologiya i Seysmologiya, v. 4, p. 3–16. CONCLUSIONS ingly oblique subduction of the Pacifi c plate Fliedner, M., and Klemperer, S.L., 1999, Structure of an island arc: Wide-angle seismic studies in the eastern Aleu- toward the west (Fig. 1). Further work is needed tian islands, Alaska: Journal of Geophysical Research, The new chronostratigraphy of Seguam Island to quantify magma output rates at several other v. 104, p. 10,667–10,694, doi: 10.1029/98JB01499. demonstrates that the 40Ar/39Ar furnace incre- Aleutian arc volcanoes in order to quantify the Frey, H.M., Lange, R.A., Hall, C.M., and Delgado-Granados, H., 2004, Magma eruption rates constrained by 40Ar/39Ar mental-heating technique can provide precise eruptive fl ux in the oceanic and continental sec- chronology and GIS for the Ceboruco–San Pedro volca- ages for latest Pleistocene to Holocene, low-K, tors of the arc. nic fi eld, western Mexico: Geological Society of America tholeiitic volcanic rocks. Replicate analyses of Bulletin, v. 116, p. 259–276, doi: 10.1130/B25321.1. Geist, E.L., Childs, J.R., and Scholl, D.W., 1988, The origin ACKNOWLEDGMENTS subsamples of each lava or tuff yielded 2σ ana- of summit basins of the Aleutian ridge: Implications lytical precisions of as low as ±2%–3%. Sixty for block rotation of an arc massif: Tectonics, v. 7, 40 39 Our most sincere thanks go to Captain Kevin Bell p. 327–341. new Ar/ Ar ages, including four age determi- and the seasoned crew of the U.S. Fish and Wildlife Gill, J.B., 1981, Orogenic andesites and plate tectonics: Ber- nations of 12 ka or younger, indicate that sub- Service M/V Tiglax for support during two of our lin, Springer-Verlag, 390 p. aerial volcanic activity began as early as 318 ka, fi eld campaigns, Lee Powell and Xifan Zhang for their Harford, C.L., Pringle, M.S., Sparks, R.S.J., and Young, expertise and quality-control assistance in the Rare S.R., 2002, The volcanic evolution of Montserrat using not 1.07 Ma, as suggested by the K-Ar ages of 40Ar/39Ar geochronology, in Druitt, T.H., et al., eds., Singer et al. (1992a). Detailed mapping, strati- Gas Laboratory, and staff at the Oregon State Univer- The eruption of Soufriere Hills volcano, Montserrat sity reactor for performing the irradiations. We also from 1995 to 1999: Geological Society of London graphic relationships, and geochemical correla- would like to thank Charlie Bacon and an anonymous Memoir 21, p. 93–113. tions along with the new age determinations pro- reviewer for their careful reviews and many thought- Hildreth, W., and Lanphere, M.A., 1994, Potassium-argon vide insights into the spatial, compositional, and ful comments that helped us improve the paper. Our geochronology of a basalt-andesite-dacite arc system— The Mount Adams volcanic fi eld, Cascade Range of volumetric evolution of Seguam Island. Volca- work was supported by U.S. National Science Founda- tion (NSF) grants EAR-0114055 and EAR-0337667, southern Washington: Geological Society of America Bulletin, v. 106, p. 1413–1429, doi: 10.1130/0016- nism was focused beneath multiple vents aligned a Geological Society of America Bruce L. “Biff” Reed in an east-west orientation, but shifted across the 7606(1994)106<1413:PAGOAB>2.3.CO;2. grant to Jicha, and a University of Wisconsin–Madi- Hildreth, W., Fierstein, J., and Lanphere, M.A., 2003a, Erup- island in a nonsystematic way over time. Erup- son Graduate School Research Award to Singer. With tive history and geochronology of the Mount Baker tions of andesite through rhyolite were coeval his permission, we have named the prominent volcano volcanic fi eld, Washington: Geological Society of with basaltic activity, often emanating from dominating the eastern half of Seguam after Ray E. America Bulletin, v. 115, p. 729–764, doi: 10.1130/0016- Wilcox, a University of Wisconsin alumnus and U.S. 7606(2003)115<0729:EHAGOT>2.0.CO;2. vents on opposites ends of the island. Petrologic Geological Survey scientist, known best for his pio- Hildreth, W., Lanphere, M.A., and Fierstein, J., 2003b, and isotopic evidence suggests that the compo- neering studies of Paricutín volcano in Mexico, and Geochronology and eruptive history of the Katmai volcanic cluster, Alaska Peninsula: Earth and Plan- sitional spectrum of lavas erupted at Seguam is the in the Aleutians. etary Science Letters, v. 214, p. 93–114, doi: 10.1016/ the result of rapid, closed-system crystal-liquid S0012-821X(03)00321-2. fractionation of magma following isolated stor- REFERENCES CITED Hinners, T.A., Hughes, R.H., Outridge, P.M., Davis, W.J., Simon, K., and Wollard, D.R., 1998, Interlaboratory age of basaltic melt in the lower to middle crust Anders, E., and Grevesse, N., 1989, Abundances of ele- comparison of mass spectrometric methods for lead for tens of thousands of years, and this has prob- ments: Meteoric and solar: Geochimica et Cosmo- isotopes and trace elements in NIST SRM 1400 bone ably persisted for more than 320 k.y. Variations chimica Acta, v. 53, p. 197–214, doi: 10.1016/0016- ash: Journal of Analytical Atomic Spectrometry, v. 13, 7037(89)90286-X. p. 963–970, doi: 10.1039/a803373k. in the proportion of fractionating minerals and Bacon, C.R., Calvert, A.T., Nye, C.J., and Sisson, T.W., Holbrook, W.S., Lizarralde, D., McGeary, S., Bangs, N., and the extent of fractionation were subtle over the 2003, History and eruptive style of Mount Veniaminof, Diebold, J., 1999, Structure and composition of the Aleutian island arc and implications for continental crust past 320 k.y., but REE patterns and U-Th iso- a huge Alaskan basalt-to-dacite volcano with Pleisto- cene and Holocene caldera-forming eruptions: Eos growth: Geology, v. 27, p. 31–34, doi: 10.1130/0091- topes indicate that new basaltic parent magma (Transactions, American Geophysical Union), v. 84, 7613(1999)027<0031:SACOTA>2.3.CO;2. replenished the crustal plumbing system at least p. F1570. Jicha, B.R., Singer, B.S., Brophy, J.G., Fournelle, J.H., Johnson, C.M., Beard, B.L., Lapen, T.J., and Mahlen, four times during the past 318 k.y. Baksi, A.K., 1974, Isotopic fractionation of a loosely held atmospheric argon component in the Picture Gorge N.J., 2004, Variable impact of the subducted slab on Like other arc volcanoes, the eruptive fl ux at basalts: Earth and Planetary Science Letters, v. 21, Aleutian Island arc magma sources: Evidence from Seguam is highly variable through time, with p. 431–438, doi: 10.1016/0012-821X(74)90183-6. Sr, Nd, Pb, and Hf isotopes and trace element abun- Black, R.F., 1983, Glacial chronology of the Aleutian dances: Journal of Petrology, v. 45, p. 1845–1875, doi: periods of elevated activity occurring during Islands, in Thorsen, R.M., and Hamilton, T.D., eds., 10.1093/petrology/egh036. the construction of a stratovolcano from 100 to Glaciation in Alaska: Fairbanks, University of Alaska Jicha, B.R., Singer, B.S., Beard, B.L., and Johnson, C.M., 2005, Contrasting timescales of crystallization and 30 ka. Glacial ice and coastal erosion removed Museum Occasional Paper, v. 2, p. 5–8. Christiansen, R.L., and Petersen, D.W., 1981, Chronology of magma storage beneath the Aleutian Island arc: Earth a signifi cant percentage of the erupted prod- the 1980 eruptive activity, in Lipman, P.W., and Mul- and Planetary Science Letters, v. 236, p. 195–210, doi: ucts, allowing for minimum estimates of the lineaux, D.R., eds., The 1980 eruptions of Mount St. 10.1016/j.epsl.2005.05.002. Marsh, B.D., 1982, The Aleutians, in Thorpe, R.S., ed., cumulative eruptive volume. Despite this short- Helens: U.S. Geological Survey Professional Paper 1250, p. 17–30. Andesites; orogenic andesites and related rocks: coming, the calculated time-averaged eruptive Crisp, J.A., 1984, Rates of magma emplacement and volcanic Chichester, UK, John Wiley & Sons, p. 99–114. rate at Seguam is similar to that of the 940 ka output: Journal of Volcanology and Geothermal Research, Masterlark, T., and Lu, Z., 2004, Transient volcano defor- v. 20, p. 177–211, doi: 10.1016/0377-0273(84)90039-8. mation sources imaged with interferometric synthetic Mount Adams volcanic fi eld, and greater than aperture radar: Application to Seguam Island, Alaska:

Geological Society of America Bulletin, July/August 2006 821 Jicha and Singer

Journal of Geophysical Research, v. 109, B01401, doi: the Royal Society of Edinburgh, Earth Sciences, v. 81, America Bulletin, v. 109, p. 127–142, doi: 10.1130/0016- 10.1029/2003JB002568. p. 251–261. 7606(1997)109<0127:VAEDTP>2.3.CO;2. Miller, T.P., McGimsey, R.G., Richter, D.H., Riehle, J.R., Sherrod, D.R., and Smith, J.G., 1990, Quaternary extrusion Singer, B.S., Hildreth, W., and Vincze, Y., 2000, 40Ar/39Ar Nye, C.J., Yount, M.E., and Dumoulin, J.A., 1998, Cat- rates of the Cascade Range, northwestern evidence for early deglaciation of the central Chilean alog of the historically active volcanoes of Alaska: U.S. and southern British Columbia: Journal of Geophysical Andes: Geophysical Research Letters, v. 27, p. 1663– Geological Survey Open-File Report 98-582, 104 p. Research, v. 95, p. 19,465–19,474. 1666, doi: 10.1029/1999GL011065. Miyashiro, A., 1974, Volcanic rock series in island arcs and Singer, B.S., Myers, J.D., and Frost, C.D., 1992a, Mid-Pleis- Singer, B.S., Ackert, R.P., and Guillou, H., 2004, 40Ar/39Ar active continental margins: American Journal of Sci- tocene lavas from the Seguam volcanic center, central and K-Ar chronology of Pleistocene glaciations in ence, v. 274, p. 321–355. Aleutian arc: Closed-system fractional crystallization Patagonia: Geological Society of America Bulletin, Price, E., 2004, Dynamic deformation of Seguam Island, of a basalt to rhyodacite eruptive suite: Contributions v. 116, p. 434–450, doi: 10.1130/B25177.1. Aleutians, Alaska, 1993–2000: Implications for mag- to Mineralogy and Petrology, v. 110, p. 87–112, doi: Steiger, R.H., and Jäger, E., 1977, Subcommission on geo- matic and hydrothermal processes: Journal of Geo- 10.1007/BF00310884. chronology; convention on the use of decay constants physical Research, v. 109, B04202, doi: 10.1029/ Singer, B.S., Myers, J.D., and Frost, C.D., 1992b, Mid-Pleis- in geo- and cosmochronology: Earth and Planetary 2003JD004400. tocene basalt from the Seguam volcanic center, central Science Letters, v. 36, p. 359–362, doi: 10.1016/0012- Renne, P.R., Sharp, W.D., Deino, A.L., Orsi, G., and Aleutian arc, Alaska: Local lithospheric structures and 821X(77)90060-7. Civetta, L., 1997, 40Ar/ 39Ar dating into the historical source variability in the Aleutian arc: Journal of Geo- Williams, S.N., and Self, S., 1983, The October 1902 Plinian realm; calibration against Pliny the Younger: Science, physical Research, v. 97, p. 4561–4578. eruption of Santa María volcano, Guatemala: Journal v. 277, p. 1279–1280, doi: 10.1126/science.277.5330. Singer, B.S., O’Neil, J.R., and Brophy, J.G., 1992c, of Volcanology and Geothermal Research, v. 16, p. 33– 1279. Oxygen isotope constraints on the petrogen- 56, doi: 10.1016/0377-0273(83)90083-5. Renne, P.R., Swisher, C.C., Deino, A.L., Karner, D.B., esis of Aleutian arc magmas: Geology, v. 20, York, D., 1969, Least squares fi tting of a straight line with cor- Owens, T.L., and DePaolo, D.J., 1998, Intercalibra- p. 367–370, doi: 10.1130/0091-7613(1992)020<0367: related errors: Earth and Planetary Science Letters, v. 5, tion of standards, absolute ages, and uncertainties in OICOTP>2.3.CO;2. p. 320–324, doi: 10.1016/S0012-821X(68)80059-7. 40Ar/ 39Ar dating: Chemical Geology, v. 145, p. 117– Singer, B.S., Thompson, R.A., Dungan, M.A., Feeley, T.C., 152, doi: 10.1016/S0009-2541(97)00159-9. Nelson, S.T., Pickens, J.C., Brown, L.L., Wulff, A.W., MANUSCRIPT RECEIVED 23 JUNE 2005 REVISED MANUSCRIPT RECEIVED 19 DECEMBER 2005 Shackleton, N.J., Berger, A., and Peltier, W.R., 1990, An Davidson, J.P., and Metzger, J., 1997, Volcanism MANUSCRIPT ACCEPTED 25 FEBRUARY 2006 alternative astronomical calibration of the lower Pleisto- and erosion during the past 930 ky at the Tatara–San cene timescale based on ODP Site 677: Transactions of Pedro complex, Chilean Andes: Geological Society of Printed in the USA

822 Geological Society of America Bulletin, July/August 2006