Introduction: Regional Setting of the Galápagos Islands

Tectonic Setting

This field trip was designed to traverse the Galápagos islands from the oldest end of the archipelago to the youngest end then back to the old part. The theme of the conference, how ocean island volcanoes change with time, is one of the great remaining problems of Galápagos geology: we are uncertain how much of the volcanologic and petrologic variation is due to evolution of the individual systems as they are carried across the hotspot and how much is attributable to magma generation and transport in different tectonic environments within the archipelago.

The Galápagos lie on the Nazca Plate but just south of an active mid-ocean ridge, the Galápagos Spreading Center (Hey, 1977). The absolute motion of the Nazca plate is about 37 mm/y at an azimuth of about 91^o. Owing to relatively fast motion of the GSC to the northwest, the absolute motion of the Nazca plate is nearly parallel to the ridge, almost perpendicular to the relative motion. Another important implication of the fast absolute motion of the GSC is that it probably lay over the Galápagos hotspot about 8 million years ago.Figure 1: Tectonic setting of the Galápagos Islands, astride the Galápagos Spreading Center. Magnetic stripes from Wilson and Hey (1995).

The Galápagos is a hotspot, but many features of the archipelago are complicated by the proximity of the ridge. For example, many Galápagos lavas are compositionally and isotopically similar to MORBs. Also, a 100-km transform offset at 91^o W separates thick and strong lithosphere to the west from thin and weak lithosphere to the east (Feighner and Richards, 1994). Likewise, the hotspot has clearly affected the ridge: along the axis of the ridge, the ridge shallows and there are pronounce gravity, isotopic, and geochemical anomalies (e.g. elevated ⁸⁷Sr/⁸⁶Sr, La/Sm, K) towards the archipelago, peaking at about 91^o W (Ito and Lin, 1995; Verma et al., 1983).

Regional Geology

One of the most curious aspects of the Galápagos is that volcanism is relatively long-lived. The western volcanoes are more historically active and no rocks older than about 0.2 Ma have been found there, as the plate motions would indicate. But as we shall see, easternmost San Cristobal is largely surfaced with Holocene lavas. These are not particularly distinct from the older lavas, so no comparison to Hawaiian alkaline or post-erosional volcanism is evident. Also, both young and old volcanoes are strung out to the north and south, perpendicular to plate motion. Darwin (1844) noticed a distinct alignment of the Galápagos volcanoes (which are now referred to as the "Darwinian" trends) trending roughly N30W and N60E. No one has yet come up with a good explanation for these alignments; they bear no simple relation to plate motion, local transforms, or likely directions of plume flow.

McBirney and William's geologic exploration of the Galápagos revealed that the archipelago is separated into several geologic subprovinces, on the basis of the ages of the volcanoes, their geomorphic forms, and the petrology of their lavas (McBirney and Williams, 1969).

The "old" subprovince comprises Espanola (Hall, 1983), Santa Fe (Geist et al., 1985), and Baltra Islands (Bow, 1979). These islands are the strongly-faulted remnants of ancient subaerial volcanoes that were active from about 3 to 1 Ma. The compositions of the lavas of these older volcanoes are not distinguishable from lavas of the surrounding, younger volcanoes. Curiously, these islands are aligned to about N30W, parallel to the regional alignment currently in the western Galápagos.

The central subprovince is made up of San Cristobal (Geist et al., 1986), Santa Cruz (Bow, 1979), and Santiago Islands (Swanson et al., 1974). San Cristobal has been active for more than 2 million years. The volcanoes have relatively shallow slopes, lack a caldera, and each has aligned systems of satellite vents. The lavas are dominated by primitive, high-MgO basalts that have rather extreme trace element variation; each of the islands has erupted an unusual suite of lavas depleted in incompatible elements (such as potassium) as well as alkali-olivine basalts that are strongly enriched in incompatible elements. Floreana is unique and does not naturally fit in the other subdivisions (Bow and Geist, 1992). It has a subdued form, late parasitic cones, and long-lived history, similar to the volcanoes of the central subprovince. Floreana's lavas are distinctly more alkaline than elsewhere in the archipelago, and many of them contain abundant ultramafic inclusions.

The western subprovince is made up of the historically active volcanoes of Isabela and Fernandina islands and Roca Redonda volcano. These are the classic Galápagos shields, with roughly symmetric forms, steep slopes, and proportionately enormous calderas. At one time, these volcanoes were thought to be made up of petrologically monotonous tholeiites. To some extent, this is true, but detailed exploration and study has revealed suites of lavas that are much more diverse. Cerro Azul has erupted tholeiitic and alkali-olivine basalts that are related to one another by high-pressure differentiation (Naumann et al., in prep.). Fernandina (Allan, in prep.), Sierra Negra (Reynolds and Geist, 1995), and Wolf volcanoes (Geist et al., in prep.) have erupted exceedingly monotonous basalts with MgO ca. 6%. Roca Redonda has erupted picrites and alkali-olivine basalts (Standish et al., 1998). Volcan Ecuador likewise is somewhat alkaline but with a very complex magmatic plumbing system. Alcedo has erupted a bimodal suite of basalts and rhyolites (Geist et al., 1995).

The northern subprovince is exceedingly diverse petrologically: basically, the archipelago's entire range isotopic and trace element values are encompassed by the 5 islands Wolf, Darwin, Genovesa, Marchena, and Pinta. Likewise, each has a distinct form, and Marchena and Genovesa have calderas and the others do not.

Regional Geochemistry

One of the most intriguing aspects of the Galápagos is that the volcanologic, petrologic, and geochemical features are distinctly related in space, not so much in time (so far as we can tell). The most distinctive manifestation of this is the isotopic horseshoe, where the Sr, Nd, and Pb isotopes indicate the increasing influence of depleted mantle in the center of the archipelago, trailing off to the east (Geist et al., 1988; White et al., 1993; Harpp, 1995).

As with every other ocean island on the planet, isotope correlation diagrams have been interpreted as resulting from mixing between a relatively enriched plume and depleted upper

Figure 2: Sr and Nd isotopic data from White et al., 1993 and Harpp, 1995. Note the broad overlap of Galápagos island lavas with the East Pacific Rise field. Also, it is apparent that there are at least 3 enriched endmembers defined by Floreana, Sierra Negra, and Pinta islands.

mantle. The fine details of the correlations and inclusion of ³He data indicate that the picture is much more complex, involving either more than two isotopic reservoirs, or segregation of the Sr, Nd, Pb, and He by melting and mixing processes (Graham et al., 1993; Kurz and Geist, in prep.

Figure 3: Isotopic and geochemical features of the Galápagos volcanoes have a clear spatial relationship. Lavas from the central part of the archipelago are depleted, whereas those from the west, north, and south are enriched. We refer to this as the "isotopic horseshoe", and it may be attributable to either plume-asthenosphere mixing or mixing with a lithospheric component. From Harpp (1995).

The table below illustrates the regular regional volcanologic and petrologic relationships. The volcanoes in each subprovince have similar elevations, caldera dimensions, and flank slopes. They also have similar ages. Lavas of the central subprovince tend to be primitive (e.g. MgO concentration), but there are wide variations in the extent of differentiation. Likewise, these lavas have wide variations in incompatible element ratios, as exemplified by the standard deviation of K/Ti ratios. Lavas from the western volcanoes have exceedingly small ranges in their incompatible element concentrations and some have small ranges in [MgO]. Other of the western volcanoes have either more primitive lavas (Ecuador, Cerro Azul) or more evolved differentiates (Pinzon, Alcedo). How much of this variation is attributable to the maturity of the volcano, and how much is due to other factors?

Volcano (We visit the ones in Bold)	Elevation (m)	Caldera Depth (m)	Age	MgO	100*K2O/TiO2 (Basalt)
Old Subprovince
Espanola	190	none	2.6 - 2.8 Ma
Santa Fe	220	none	2.5 - 2.8 Ma	6.8± 0.7	34 ± 14
Baltra	40	none	1.1 - 1.3 Ma
Central Subprovince
San Cristobal	820	none	0 - 2.3 Ma	9.1± 1.8	35± 12
Santa Cruz	786	none	0.05 - 0.59 Ma	7.7± 2.1	18 ± 8
Floreana	650	none	1.5 - 0.07 Ma	10.7± 2.0	62± 25
Santiago	915	none	0 - 770,000	6.2± 3.4	19 ± 15
Western Subprovince
Fernandina	1480	1100	0 - 6000	6.3± 0.4	14 ± 1
Cerro Azul	1620	480	0 - 82,000	7.5± 1.7	20 ± 3
Sierra Negra	1120	110	0 - 9000	5.5± 0.5	17± 2
Alcedo	1180	270	0 - 150,000	4.3± 2.2	17 ± 4
Darwin	1420	200
Wolf	1720	660	0 - 173,000	6.0± 0.5	17 ± 1
Ecuador	730	640	0 - 130,000	6.6± 1.7	23± 2
Pinzon	457	150	0.7 - 1.2 Ma	4.1± 1.7	18 ± 4
Roca Redonda	50	none		12.8± 5.4	26 ± 2
Northern Subprovince
Genovesa	70	65		7.4± 0.9	7 ± 2
Marchena	343	ca. 150	0 - 300,000	6.7± 0.3	15 ± 3
Pinta	785	none	0 - 700,000	6.3± 0.4	26 ± 4
Darwin I.	170	none
Wolf I.	255	none

Day 1: San Cristobal Island

We will not have much time to visit San Cristobal, but it is typical of the central subprovince.

Geologic overview

San Cristobal is the easternmost island in the Galápagos. It is the site of the only permanent stream in the archipelago and is also where Darwin first went ashore in 1835. San Cristobal has been colonized since before Darwin's time, and it was the site of an infamous penal colony at El Progresso. Since the mid-1980s, an airport has been open on the island which brings nearly half of the islands' 50,000 annual visitors.

San Cristobal is not the best wildlife site in the archipelago, although we will see some of the typical marine birds, marine iguanas perhaps, and sea lions. A large pod of whales was settled around Kicker Rock in late March.

The geology of San Cristobal is covered by Geist et al. (1986). San Cristobal island is made up of two coalesced volcanoes. The southwestern half is a symmetric shield volcano, with numerous satellite cinder cones. The southwestern shield became emergent around 2.35 Ma, and activity continued through the Brunhes/Matayuma transition at 0.7 Ma; the youngest dated lava from the SW shield is 0.66 ± .08 Ma. The northeastern half of the island is a more recently active system, dominated by eruptions from NE-trending fissures. The most recent flows are no more than several centuries old (Mark Kurz will talk about these).

Figure 4: Geologic map of San Cristobal (from Geist et al., 1986), showing the age distribution of the lavas: 1 = normally polarized (> 2.3 Ma); 2 = reversely polarized (2.3 to 0.7 Ma); 3 = normally polarized (10,000 (?) to 700,000 years old); 4 = several millennia (?) old; 5 = several centuries old.

The volcanic history and petrology of San Cristobal are similar to the other central and eastern Galápagos (especially Santiago and Santa Cruz). Activity has been more-or-less continual for over 2 million years, and no known major hiatuses are apparent despite fairly detailed study. The older southwestern shield is made by gently-dipping lavas and capped by a thick, deeply-weathered pyroclastic blanket. The northeastern half of the island is made of mostly younger lavas that erupted from a SW-NE trending fissure zone.

Most San Cristobal lavas are high-MgO basalts and picrites, although sparse hawaiities and icelandites erupted from the late-stage cinder cones on the older SW shield. A large number of San Cristobal lavas are near-primary, with Mg# = 60 to 72 and containing primitive olivine up to Fo₈₉. Most lavas have olivine, chromium-spinel, and plagioclase phenocrysts, many of which exist in snowflake-like glomerocrysts. Augite phenocrysts occur only in a few of the most evolved lavas. Troctolite and olivine gabbro xenoliths occur in some flows, and one norite clot is known. On San Cristobal, some flows have differentiated during flow, with a lower olivine-rich zone and an upper plagioclase-rich zone. Ever the brilliant observationalist, Darwin noted this and suggested that crystal-liquid fractionation could explain the differences between basalts, oceanites (picrites), and "trachytes" (referring to andesites he had seen the previous months in western South America).

San Cristobal lavas plot on the depleted end of the Galápagos array in Nd-Sr-Pb isotopic space; it seems to point to Floreana at the enriched end (Figure 1).

By far the most intriguing aspect of the San Cristobal lavas is the extreme diversity of trace element compositions, as exemplified by the rare-earth elements. Basically, this volcano has erupted everything from LREE-depleted, MORB-like lavas to lavas with REE patterns similar to those of Hawaiian tholeiite. Although an entire spectrum of compositional types exist, the lavas have been broken into three types: low-K, MORB-like tholeiites; ocean-island tholeiites; and alkali-olivine basalts.

Figure 5: Major element compositions of San Cristobal lavas. This is an enormous compositional range, basically extending from typical MORB to compositions typical of Hawaiian AOB.

In stark contrast to Hawaiian volcanoes, there is no clear petrologic evolutionary trend displayed by San Cristobal lavas. This is especially true on the northeastern half of the island, where all three lava types are randomly intercalated. The MORB-like lavas are rare on the southwestern shield, however.

Modeling of the trace elements indicates that the different magma series can be generated by a roughly 8-fold difference in the extent of partial melting. By this model, the alkali-olivine basalts are due to small degrees of melting and the MORB-like tholeiites larger degrees of melting. Many of the magmas were modified by cooling and crystallization during ascent, but most are substantially unmodified since derivation in the mantle. Relatively low Sm/Yb ratios and iron concentrations and high silica concentrations suggest relatively shallow melting, in the spinel stability field (White et al., 1993). Decreasing Yb concentrations and crossing patterns suggest melting in the presence of garnet, however.

Field Trip Itinerary

Our visit to San Cristobal will be a short one, but we will see many similar volcanic and petrologic features on Santiago and Santa Cruz. The airport is built on a platform of normally-polarized lavas that are 2.3 Ma. These primitive lavas belong to the alkaline series (K₂O = 0.53%; 1.8% normative nepheline; Mg# = 69.3) and contain 1 % plagioclase and 5% olivine phenocrysts. To the east of the airport, an excavated cinder cone shows that this part of the island has been emergent for millions of years and is not just a fault block. The cinder cone is made of differentiated alkali-olivine basalt that contain 3% olivine and 15% plagioclase phenocrysts.

We may also be able to stroll around the village of Puerto Baquerizo Moreno, which is built on lavas that are similar to those underlying the airport. Note that this area is not in the national park; it is one of the few areas in the archipelago where one can head off trail. Parasitic cinder cones in the area include Cerro Patricia, a couple kilometers east of town, and Cerro Chivo two kilometers north of town. These lavas are notably more differentiated than the flat-lying lavas and contain abundant plagioclase crystals.

If you get the opportunity to walk to these cones, note that the soil-forming process creates a curious patterned ground, with resistant cantaloupe-sized basalt chunks in a polygonal pattern surrounded by red-clay rich soil. This patterning characterizes the surface of ancient flows in the arid climatic zone (< 200 m in elevation) throughout the archipelago. We have no idea how it forms.

Many boats cruise through Kicker Rock, the remnant of a palagonite tuff cone 5 km to the north of San Cristobal (McBirney and Williams, 1969). This was the flank of a very large cone that rises 250 meters from the adjacent seafloor. As with must tuff cones, it is made up of alternating surge and fallout layers of mostly ash- and lapilli-sized palagonite and scoria. Many of these cones display spectacular soft-sediment deformation features.

	23	2	9	5	59	108
	Oldest flow	Cone near	Cerro	Cerro Chivo	Low-K	Isla Lobos
	SW tip	airport	Patricia		tholeiite	cliffs
SIO2	46.99	47.89	47.14	48.12	48.01	47.45
TIO2	1.46	2.29	2.06	2.33	1.02	1.49
AL2O3	15.73	16.99	16.11	17.06	17.47	15.50
FE2O3	2.16	3.79	2.39	4.28	4.50	4.10
FEO	7.19	6.87	7.14	7.65	4.34	5.90
MNO	0.19	0.21	0.18	0.22	0.17	0.18
MGO	11.53	6.69	11.38	5.95	9.33	10.98
CAO	10.79	10.17	9.93	9.58	12.65	10.62
NA2O	2.65	3.35	3.04	4.09	2.27	2.89
K2O	0.53	0.81	0.74	1.01	0.15	0.60
P2O5	0.23	0.40	0.28	0.46	0.10	0.27
BA	122	225	158	261	39	138
CR	432	144	543	55	260	502
NI	247	91	265	54	142	259
RB	8.8	14	8.8	19	5.9	9.3
SR	399	455	447	454	178	388
LA	11.36		12.45		2.85
SM	3.49		4.11		1.94
YB	2.08		2.68		2.34
EU	1.26		1.57		0.8
TB	0.71		0.87		1.67
LU	0.34		0.39		0.44
MG#	69.3	53.8	68.6	48.0	66.5	67.2
AGE	1	2	3	3	3	3

Day 2: Espanola Island

Espanola island is one of the smallest of the Galápagos islands, measuring 7 x 14 km. Espanola has the oldest lavas yet found in the Galápagos, around 3 Ma. Although it was originally interpreted as a horst of submarine lavas, it is now known to have been emergent during eruption (Hall, 1983). Espanola is the pioneering success story in the ecological rehabilitation of the islands. In the early 1970s, scientists from the Darwin Station and rangers and hunters from the National Park eradicated the feral goats and rejuvenated the tortoise population through breeding at the Station and repatriation.

It is unlikely that we will see any tortoises, but Puerta Suarez is the nesting site of the waved albatross. Also, the marine iguanas at Espanola have a unique red pigment. Bahia Gardner is probably the nicest beach in the archipelago and is the home of a very prolific sea lion colony.

Along with Baltra and Santa Fe, Espanola is part of the "old" subprovince. These volcanoes had relatively short-lived activity (perhaps 500,000 years) and went extinct over one million years ago. They are mostly made of alkali-olivine basalts, similar to those erupted more recently from Santa Cruz, Santiago, and San Cristobal islands.

Lava flows dip centripetally about the center of the island, which presumably was the summit of the old shield. The island has been extensively block faulted along E-W faults, including one that forms the southern seacliffs. A number of cinder cones are exposed around the volcano. Weathering and soil development is more extensive on Espanola than any other site at low altitude in the Galápagos.

Figure 7: Map of Espanola (from Hall, 1983) showing topography, several attitudes, and the distribution of cinder cones. We will be visiting Puerta Suarez and Gardner Bay.

Only a couple samples from Espanola have been studied, but we are currently working on a stratigraphically-constrained suite from the island. The lava flows are alkali-olivine basalts, with phenocrysts of olivine and plagioclase in a groundmass containing distinctly-purple augite. Rare earth patterns are LREE enriched, with a tendency for the HREEs to be flat. Espanola lavas lie in the center of the Galápagos isotopic array. In a geographic sense, Espanola is a bit of an isotopic anomaly: when projected back to its eruption site at 3 Ma, it more depleted than the lavas that have erupted there more recently (Floreana). This is one of the few exceptions to the isotopic horseshoe.

Field Trip

We will be visiting two sites on Espanola, Punta Suarez and Bahia Gardner, both fantastic wildlife sites. At Suarez, we will walk over lava flows that have been dated at 2.8 Ma to the cliffs that form the south coast of the island. The Suarez lava flows are primitive alkali-olivine basalts (Mg# = 65; K₂O = 1.11%; 1.8% normative nepheline). These cliffs are used to advantage by the giant albatrosses and probably follow one of the horst-bounding faults. We will not be able to see an old cinder cone that is engulfed by younger flows in the cliff.

Gardner Bay is a beautiful white-sand beach with a prolific sea lion colony. The old lavas behind the bay are equivalent in age (2.6 ± .1 Ma) to those at Suarez, and the slight north dip is apparent from here. The Gardner Bay flows are slightly more evolved than those at Suarez (Mg# = 61; K₂O = 0.79%; 2.6% normative nepheline). Gardner Island and islets inshore of it are mostly the remnants of cinder cones.

The volcanic deposits and petrology and geochemistry of Espanola's lavas suggest that it was similar to the volcanoes of the central subprovince, but shorter-lived.

Day 3: Sierra Negra Volcano

Introduction

Owing to a permanent human population in Puerta Villamil and in the highlands, Sierra Negra is the most accessible volcano in the western Galápagos. The western Galápagos volcanoes are the type locality of the "Galápagos-type shield", characterized by gently dipping lower flanks, steep upper flanks, broad summit plateaus and large, complex calderas, the whole of which resembles an "overturned soup bowl". Also, they have a characteristic distribution of vents, with circumferential fissures near the summit and radial fissures on the lower flanks. The western volcanoes are especially active and thus provide an excellent setting for the study of the dynamic processes peculiar to ocean island volcanism.

Physical volcanology

Sierra Negra occupies a substantial part of the south side of Isabela island, which is itself constructed of six major shield volcanoes (Fig. 9). The volcano is 60 km long and 40 km wide and rises 1100 m above sea level. It is by far the most voluminous of the western Galápagos shields and exhibits many unusual morphologic features including an extensive lower flank apron, a large complex caldera, and a volcano-wide system of ENE-trending radial and circumferential fissures (Reynolds et al., 1995). The recent geologic history of Sierra Negra is characterized by voluminous eruptions of a'a from a complex of radial and circumferential fissures which together define a volcano-wide east-northeast trending rift system that extends from the lower flanks and deflects around the large summit caldera. Eruption locations along the rift system have varied considerably with time. Age determinations of surface lava flows by ¹⁴C and cosmogenic ³He are broadly distributed and Figure 9: Simplified geologic map of Sierra Negra volcano. Circled numbers indicate field trip stops described in text.

range from 400 to 6900 yr. In comparison to other western Galápagos volcanoes, Sierra Negra's somewhat subdued profile is principally due to its large discharge rate and voluminous eruptions, which cause most lava to accumulate on the lower flanks resulting in the development of a large coastal apron. Sierra Negra's flanks show no evidence of gravitational slumping; rather the upper flanks appear to be buttressed by the extensive lower flank apron.

Ten historical eruptions have been reported at Sierra Negra. On the basis of the past four historic eruptions, estimated discharge rates range from 40 to 250 m³/sec with an average historic growth rate of about 12 x 10⁶ m³/yr. The long term growth rate based upon the geologic age determinations is estimated to be approximately 1 x 10⁶ m³/yr. In total, the volcano has undergone in excess of 90 % resurfacing in the past 5000 yr. It has not erupted since 1979, thus is probably due for activity.

Petrology

Sierra Negra has erupted Fe-rich, hypersthene-normative tholeiitic basalt of very limited compositional range (Fig. 3; Reynolds and Geist, 1995). The major and trace element data indicate a comagmatic relationship by fractional crystallization of the observed phenocryst phases. Projections into pseudoternary phase diagrams suggest that the magmas cooled and crystallized at pressures between 1 and 3 kbar. High Sm/Yb ratios, light rare earth element (LREE) enrichment, and a steep REE slope are consistent with an origin by moderate extents (5-15 %) of partial melting of a garnet-lherzolite source with REE characteristics that are between chondritic and depleted mantle sources. Magmatic ³He/⁴He isotopic ratios from Sierra Negra are approximately 15 times the atmospheric ratio. In addition, Sierra Negra lavas have the most radiogenic lead and strontium isotopic ratios in the western Galápagos, indicating that the magmas have a relatively large contribution of plume material and have been only minimally contaminated by entrainment of MORB-producing mantle.Figure 10: (left) Alkali-silica variation diagram of 29 Sierra Negra lavas. Symbols correspond to stratigraphic age units. Diagonal line divides alkaline and tholeiite basalt. Error bars represent 1 sigma variation based upon duplicate analyses. (right) Silica-magnesia variation diagram of 29 Sierra Negra lavas. Symbols correspond to stratigraphic age units. A computer calculated fractionation model using the most primitive Sierra Negra composition (1200 - 1100 C and QFM buffer) is shown with a dashed arrow. The predicted trend is toward Si depletion.

Field Trip Guide.

Four stops are selected to highlight some of the more interesting features of the volcano (Fig. 9).

Stop 1: Villamil (El Estero).

The area surrounding the village of Villamil on the southern flank of Sierra Negra contains some of the oldest lavas on the volcano as well as evidence of recent uplift and examples of coastal phreatomagmatic activity (Fig.11). Numerous monogenetic cones dot the landscape. Large-scale surface features such as tumuli, pressure ridges, and boiler-plate pahoehoe are well preserved in the lavas, but delicate surface textures are long since eroded. The orientation of pressure ridges and flow surface cooling joints indicate that the flows originated from eruptive centers on the upper east flank near the village of Santo Tomás.

This stop shows an age 2 lava that flowed into water, resulting in the formation of a littoral cone, now partly excavated (Fig. 11). The town's freshwater supply is located in the saturated tephra of this littoral cone. The littoral cone has an exposed basal diameter of 300 m and a height of 20 m; it is constructed of 10-15 cm- thick beds of normally graded scoria,

intercalated with coral and shell fragments. The uppermost surface is partly overlain by spatter and unconsolidated beach sediments. The ¹⁴C age of shell fragments obtained from the littoral cone provides an age of 3195 +/- 100 yr. Other nearby deposits provide evidence for long term coastal uplift. A small, 5 m thick, age 4 plagioclase-phyric a'a flow erupted from a fissure centered at Cerro Pelado northwest of Villamil. In addition, an isolated lagoon and two raised beach deposits 1-2 km north of Villamil (Fig. 11) lie at 5-14 m elevation are interpreted to represent former shoreline deposits, suggesting that the area has been uplifted since the eruption of age 2 lavas. The long term rate of uplift of this part of the coast ranges from 1 to 17 mm/yr. and is based upon the relative position of the beach deposits between two dated lava flows.

Stop 2: Volcan Chico eruptive fissures.

This stop provides a traverse across the 1979 eruptive site. Circumferential fissures extend along the north and northwest upper flanks and deflect around the caldera immediately down slope from the summit. During the two month-long 1979 eruption, two areas of the circumferential fissure system 3 km apart were simultaneously active. Both areas are surmounted by a line of spatter ramparts and cinder cones that are oriented diagonally across the upper flank and curve around the summit caldera as part of the larger circumferential eruptive zone.

Age 5 flows comprise all historical eruptions and are restricted to the north flank circumferential eruptive fissures. Near-vent deposits consist of thin-crusted, platy pahoehoe flows and channels, and ramparts of unconsolidated scoria. Chains of pit craters linked by channels delineate parts of the fissure system. Basaltic pumice from the 1979 is 10-25 cm thick immediately south of the fissures, indicating prevailing winds from the north during eruption. Delicate pumice fragments of up to 12 cm in diameter are scattered about the center of the caldera. Ash and scoria were also reported as far away as Villamil, 26 km south of the vent. The lava discharge rate for this eruption approximately 120 m³/sec and covered an area of about 280 km².

Voluminous a'a flows make up the bulk of the age 5 lava fields. The flows are, aphyric to plagioclase-phyric. Most flows start from individual cones that surmount the north flank eruptive fissures. Rafted basalt boulders are common within open channels. Unlike the older east and southeast flanks, lava tubes are not observed in any of the young north flank flows. The historical lavas emanated from the summit circumferential fissures and flowed north, and some ultimately reached the ocean at Bahia Elizabeth. These north flank historical flow fields are constructed of numerous long narrow flows which, in turn, are partly overlapping and superimposed upon by shorter flows. A subdued flank profile has resulted from these long, fluid flows extending far out on the flanks.

On the north flank circumferential fissure system, degassing occurs at Chico peak and farther down slope at a partially collapsed cinder cone. Surface gas temperatures measured in 1992 were 57-69^oC and much of the surrounding ground consists of white, friable hydrothermally-altered basalt.

Stop 3: North caldera rim.

This stop provides a view south into the summit caldera. The summit of Sierra Negra is entirely occupied by a shallow, elliptical (7 x 10 km) caldera (Fig 12). Near-vertical ring faults circumscribe the summit on three sides and expose 100 to 140 m of lava flows. The north rim is outlined by an arc of spatter cones. Subparallel sets of caldera growth faults cut the western and southern caldera walls and are not associated with eruptive activity. Sierra Negra's caldera differs substantially from those of Fernandina and Cerro Azul, which have deep (480-1100 m), smaller diameter (4-6 km) calderas and wide rims. The inclination of remnant magnetism and the attitudes of the flows exposed in the caldera walls of sierra Negra indicate that the lavas have not been appreciably deformed.

The central caldera floor is covered by age 3 a'a that dips gently to the east. The structure of the north section of the caldera is controlled by an east-trending sequence of normal faults resulting in a series of small horst and graben blocks that gradually step down from the north caldera wall to the central caldera floor (Fig. 12). Displacement along the individual faults ranges from 3 -10 m. The west caldera floor contains a set of much larger down-dropped blocks, the sum of which forms a N-S trending moat. The moat is punctuated by cinder cones and sulfur fumaroles and separates the caldera walls from a high sinuous ridge. Several small outward-curving benches are positioned in front of the west-facing scarp of the sinuous ridge. In contrast, Cerro Azul, Alcedo, Wolf, Ecuador and Fernandina all have large benches that face inward from the caldera walls and are interpreted to represent stranded remnants of caldera.

The most remarkable structural component of Sierra Negra's caldera is the sinuous ridge Figs. 11 and 12). It consists of a 14 km-long complex set of overlapping, tilted blocks of caldera floor, the whole of which forms a curving C-shaped ridge that is open to the east. The entire structure is formed of outward curving faults. Fault scarps dip steeply (60-90^o), whereas the ridge flow tops initially dip inward at 20-600, then bend sharply to merge with the gently dipping (1-3 ^o) central caldera floor. Vertical displacement along the ridge varies from 1 m in the east part of the caldera to over 100 m in the west. In places, the sinuous ridge is higher than the caldera rim. None of the other Galápagos calderas exhibit such an unusual and complex caldera floor morphology.

Stop 4: West caldera rim.

This stop provides a view east into the caldera. Vigorous fumarolic degassing occurs within the western part of the caldera. The largest fumarole, known as Azufre, is located midway along the base of a north-trending section of the ridge. Large flows of elemental sulfur, boiling sand pits and extensively altered basalt are peculiar features of this fumarole. Near surface temperatures exceeded 200 ^o C in 1992. This fumarole has 3He/4He ratios identical to those of melt inclusions in the olivines (Fraser Goff, personal communication), indicating that it is almost purely magmatic gas.

The uppermost part of the western caldera wall is partly capped by a horizontal lens of palagonitized accretionary lapilli. The layer is up to 5 m thick and is exposed laterally for 1/2 km. This phreatomagmatic deposit likely erupted through a former caldera lake, perhaps similar to the lakes that presently occupy the calderas of Cerro Azul and Fernandina.

Day 4 Morning (Alcedo Volcano)

Alcedo volcano is unique within the archipelago, because it has erupted abundant rhyolite. These are the most evolved rocks in the archipelago, although nearby Pinzon volcano has erupted rhyodacite, and trachytes have erupted from Santiago and Rabida volcanoes. At Alcedo, at least 3 separate rhyolite eruptions have been documented, but most of the volume of rhyolite was erupted during a plinian eruption followed by effusion of rhyolitic lava flows (Geist et al., 1994). We will have the rare opportunity of visiting Alcedo, where the dispersal axis of the fallout tephra crosses the coast.

Despite the unusual occurrence of rhyolite at Alcedo, it is a fundamentally basaltic volcano: <1% of the exposed volume is rhyolite, and intermediate rocks are even sparser. A couple icelandite flows are exposed near the caldera wall, icelandite tephra sporadically occurs just below the rhyolite fallout south of the field trip stop, and dacite was erupted with the rhyolite during the later part of the plinian phase of the eruption. The age of the rhyolite is poorly known. Imprecise K-Ar ages indicate eruption at about 90,000 years ago. There is an extremely active hydrothermal system around the rhyolite vents, and the rhyolite is very lightly weathered, so the true age may be substantially younger. There have been about a dozen basaltic eruptions since the most recent rhyolitic eruption, the most recent in the 1950s, from a flank vent on the south apron of the volcano. There have been reports of seismic activity and strengthened hydrothermal output in 1998. The hydrothermal system is extremely transient; 2

Figure 13: Geologic map of Alcedo volcano (from Geist et al., 1994).

major new steam vents exploded in 1991 and are still so vigorous that no one has volunteered to measure their temperatures or sample their gases. Common subterranean explosions make field work in the caldera unsettling.

Figure 14: Chemical variation of Alcedo rocks. Except for the rhyolites, which constitute <1% of the volume of the volcano, the number of analyses are comparable to their volumetric abundance.

Thorough petrologic and geochemical study of the Alcedo suite indicates that the rhyolites are related to the basalts by fractional crystallization; about 85% crystallization of the most primitive olivine tholeiite is required to produce the rhyolite (Geist et al., 1995), although this interpretation is not universally agreed upon (McBirney, 1993). The rhyolites have the same Sr, Nd, and Pb isotopic ratios as the basalts, and they are about 0.7 per mil richer in ¹⁸O than the basalts. Major and trace element data show no evidence of anatexis or assimilation. Fractionation involved olivine+plagioclase+augite at early stages (to MgO about 4.5%), followed by crystallization of titanomagnetite and apatite.

The plinian deposit is zoned, with scarce icelandite scoria having erupted first, followed by the rhyolitic pumice (>95% of the deposit), followed by the eruption of rhyolite+dacite+mixed pumices. A rhyolitic lava flow within the caldera followed the plinian phase.

The rhyolite coincides with an extreme decline in the Alcedo's growth rate: up until the rhyolite eruption, the average eruption rate was 1.8 x 10⁶ m³/y, and after it has been 0.1 x 10⁶ m³/y. We believe that the rhyolite formed as the volcano's subcaldera magma chamber stopped receiving supply of hot basalt. Thus, a large magma chamber evolved by closed-system crystal fractionation. Further evidence for crystal fractionation includes small, mostly disaggregated xenoliths of oxide-rich gabbro (note the absence of the word "cumulate" here and elsewhere within this guide), including some with inverted pigeonite.

Field Trip Guide

This locality is very close to the main dispersal axis of the fallout tephra, which is about 3 meters thick here. There are two conspicuous features of this deposit. The first is the extraordinary size of some of the pumices. These pumices have vesicularities of up to 94% and the vesicles are exceedingly elongate, with aspect ratios of over 100. Second, there are two layers. The lower "A" horizon contains only rhyolitic pumice, and the upper "B" horizon contains mostly rhyolitic pumice but also distinctly black dacitic pumice (ca. 65% silica) and marble-cake pumice.

The vent for the lower part of the deposit is almost certainly the crater at the head of the intracaldera rhyolite flow. The vent for the upper part of the deposit is believed to be a ring-fracture on the east caldera rim, where we have mixed rhyolite-dacite lavas.

The total eruptive volume of the Alcedo tephra is difficult to estimate with confidence, because most of the tephra was deposited into the sea. But a conservative estimate is 3.4 km³ (0.5 km³ DRE). The dispersal of the tephra suggest that the eruptive plume was 23 km high, and the winds were from the SW at 35 m/s. The estimated eruption intensity is 1.2 x 10⁸ kg/s, which is typical of plinian eruptions.

A large a'a flow overlies the rhyolite here. This flow, and the other younger basalts, are indistinct from the basalts older than the rhyolites. There is an active sea lion colony here, and very large land iguanas are rarely seen in this area. Alcedo has the world's largest population of the giant tortoise. But it has also been the site of an absolutely devastating goat infestation that have destroyed the environment here. Geist has been coming to Alcedo since 1983. In 1983, we did not see any goats. In a six-week field campaign in 1989, we saw 3 goats (two of which did not survive). In a week in 1991, roughly 100 were seen in and around the caldera. In 1995, estimates are that over 10,000 inhabited the caldera and upper flanks. An intensive hunting campaign is currently underway.

Analyses of Alcedo lavas at field trip stop:

APUM DG6 DG18 DG51 E149

White pumice Dark pumice Icelandite Overlying Underlying

Scoria A'a Pahoehoe

SiO2 71.8 64.88 53.2 48.49 47.98

TIO2 0.39 0.99 2.01 3.23 1.95

AL2O3 13.19 13.85 15.18 13.6 17.51

FE2O3 4.71 8.17 10.55 14.16 11.64

MNO 0.12 0.17 0.16 0.1 0.16

MGO 0.28 1.68 3.96 6.29 6.21

CAO 1.42 3.43 7.88 11.72 11.59

NA2O 4.85 5.27 3.87 2.67 2.64

K2O 3.18 2.23 1.12 0.39 0.32

P2O5 0.06 0.24 0.29 0.21 0.22

LA 59.1 48.49 24.3 14.9

CE 135 109 54.2 38.2

ND 65.3 59.1 30.9 23.6

SM 14.81 13.8 7.62 6.59

EU 2.91 3.92 2.2 2.2

GD 14.88 14.53 8.09 7.12

TB 2.39 2.19 1.25 1.12

YB 8.81 7.2 3.87 2.98

LU 1.33 1.12 0.59 0.4

TH 9.5 6.43 2.94 1.26

NB 82 65.5 33.36

Y 97 84.8 47.9 28

HF 20 16.1 7.34 5.07

TA 4.7 3.7 1.93 1.53

U 3 2 0.9 0.38

ZR 782 670 352 185

NI 6 30 31 142 23

Rb 81 44 31 4 8

SR 106 174 246 320 290

BA 418 305 135 95 136

Day 4 Afternoon: Santiago Volcano

We will be landing on the west coast of Santiago, which is a 35-km long island. Darwin landed just to the east of this site and spent more time on Santiago than any other island, and hiked to the summit region while the Beagle desperately searched for fresh water. The older core of the island is a flat-topped shield that reaches an elevation of 900 m (Swanson et al., 1974). This shield is surrounded by a broad apron of younger parasitic cones and fissures. Several high tuff cones mark the eastern and western ends of the island.

Figure 16: Geologic map of Santiago volcano.

All flows at Santiago are normally polarized, and the oldest K-Ar age is 770 ± 12 Ka. Most of the volcano is draped by younger flows, and there have been historical eruptions.

In most respects, Santiago is similar to Santa Cruz and San Cristobal. It is morphologically similar, has a linear fissure system, and has erupted lavas with an enormous range of trace element concentrations. These lavas are divided into a low-K, Morb-like series and alkali-olivine basalts (Baitis and Swanson, 1974). All of Santiago's lavas have depleted, Morb-like isotopic ratios, although they have slightly lower ¹⁴³Nd/¹⁴⁴Nd than lavas of Santa Cruz (White et al., 1993). The two suites have the same Nd, Sr, and Pb isotopic ratios. It is believed that the alkali-olivine basalts result from small extents of partial melting and the low-K basalts result from greater extents of melting.

Unlike San Cristobal and Santa Cruz, there is a clear stratigraphic relationship between the lava types. The old shield is largely constituted of the evolved alkaline rocks, whereas the younger apron is made up of stratigraphically-interfingered alkali-olivine basalt and low-K tholeiite. The alkaline suite has differentiated to form rocks as evolved as trachyte. This close correspondence between the compositions of the evolved rocks and the basalts (c.f. Alcedo) strongly supports the hypothesis that the evolved rocks are created by crystal-fractionation.

Field Trip Stop

We will be stopping at Puerto Egas, which is on the south part of James Bay. This will give us the opportunity to observe phreatomagmatic deposits overlying basalt flows, subaerial parasitic vents (Sugarloaf volcano), and the historical lava flows.

The James Bay flows cover about 13 km² and originated from spatter cones and shields about 3.5 km inland. Most of the field is pahoehoe. The flow is estimated to have erupted in 1759, on the basis of a unique geochronological tool. Jam jars (quince marmalade, to be specific) from a pirate's stash are found in the flow. Archaelogical study of the jars reveals their age, and further study of the jars by the rhenium-osmium method should confirm this.

Santiago is a vivid example of the ecological destruction inflicted on this archipelago by introduced mammals. The Savannah vegetation is completely artificial and due to overgrazing by the goats. Land iguanas are extinct due to competition from the goats and predation by feral pigs. Massive hunting campaigns keep the feral animals' numbers reduced, but it is all but impossible to rid these large islands of the introduced pests.

Analyses of Santiago lavas at field trip stop:

Days 5-9: Santa Cruz Island

Santa Cruz is the second largest of the Galápagos Islands, the site of our host institution, the Charles Darwin Research Station, and the site of Puerto Ayora, the largest town in the islands.

The island comprises two geologic parts: the northeastern end is termed the "Platform

Series" and the main shield the "Shield Series" (Bow, 1979). The Platform Series is distinctly older and is made up of submarine flows and limestones overlain by subaerial pahoehoe and a'a. It is strongly faulted and is separated from Baltra and N. Seymour islets by two large grabens. Precise K-Ar ages of the Platform Series cluster around 1.3 Ma, although imprecise ages are as great as 3.8±1.8 Ma.

The Shield Series erupted from a series of vents oriented WNW. The shield is entirely normally polarized, and K-Ar ages range from imprecise dates as old as 590±270 Ka to more precise ages as young as 24±11 Ka. Several prominent fault scarps oriented ENE cut Shield Series lavas in the town of Puerto Ayora.

The Shield Series lavas are compositionally indistinguishable from the bimodal alkali-olivine/low-K tholeiite series of Santiago and San Cristobal. As on those two islands, the magmas are isotopically indistinguishable, thus are thought to result from different extents of partial melting. Also, picrites and high-Mg basalts are the most abundant lavas. Some of the alkaline magmas are as evolved as hawaiite. Santa Cruz is in the center of the isotopic horseshoe and, with the exception of Genovesa, has erupted the most depleted magmas in the archipelago (White et al., 1993).

The Platform Series lavas are intermediate in every respect to the alkali-olivine basalts and the low-K tholeiites. Thus, if there is an evolutionary trend, it is different from Santiago, where consistently alkaline lavas gave way to the bimodal suite. The Platform Series tends to be more evolved than the Shield Series, and ferrobasalts and plagioclase-ultraphyric lavas are quite abundant in the sea cliffs.Figure 19: Compositional variation of Santa Cruz lavas, illustrating bimodality of suite.

Analyses of Santa Cruz Lavas Observed on Field Trip:

	E-1	Sc-46	Sc-106	SC-155	SC-64
	Darwin Station	Cliff behind Station	Ithabaca Channel Platform Series	Highland cone	Low-K Tholeiite
SIO2	46.14	46.04	48.73	48.12	46.31
TIO2	2.01	1.32	1.73	2.30	1.01
AL2O3	16.10	15.47	15.12	17.29	16.33
MGO	10.43	11.37	9.72	6.37	11.91
FE2O3	2.26	5.28	4.48	3.52	2.18
FEO	9.07	6.45	6.60	8.12	8.19
MNO	0.19	11.37	0.18	0.18	0.18
CAO	9.19	9.86	9.86	10.14	11.24
NA2O	3.64	2.65	3.10	3.22	2.53
K2O	0.28	0.23	0.39	0.31	0.05
P2O5	0.24	0.19	0.23	0.43	0.12
CR	340	+	+	134.00	442.00
NI	400	349	268.00	87.00	320.00
RB	3.1	2	-	3.00	2.00
SR	409	350	271.00	411.00	236.00
BA	42	+	+	0.00	0.00
Zr	188	134	147.00	207.00	78.00
LA	9.8	+	+	14.80	2.80
Ce	28.2	+	+	28.00	7.40
SM	5.17	+	+	5.60	2.30
Eu	1.85	+	+	0.50	0.3
Gd	5.71
Yb	2.85			3.00	2.0

References Cited

Baitis HW, Lindstrom MM (1980) Geology, petrography and petrology of Pinzon Island, Gal·pagos Archipelago. Contrib Mineral Petrol 72: 367-386

Baitis HW, Swanson FJ (1974) Ocean rise-like basalts within the Gal·pagos Nature 259: 195-197

Bow, CS (1979) The geology and petrogeneses of lavas of Floreana and Santa Cruz Islands: Gal·pagos Archipelago, Ecuador. Unpub. PhD thesis, University of Oregon.

Chadwick WW, Howard KA (1991) The pattern of circumferential and radial eruptive fissures on the volcanoes of Fernandina and Isabela islands. Bull Volcanol 53: 259-275

Chadwick, WW and Dieterich, JH (1991) Mechanical modeling of circumferential and radial dike intrusion on Galápagos volcanoes. J Volcanol Geotherm Res 66, 37-52, 1995.

Christie DM, Duncan RA, McBirney RA, Richards MA, White WM, Harpp KS, Fox CG (1992) Drowned islands downstream from the Gal·pagos hotspot imply extended speciation times. Nature 355: 246-248

Colony, WE and Nordlie, BE, (1973) Liquid sulfur at Volcan Azufre, Galápagos Islands, Econ Geol 68, 371-380.

Darwin CR (1844) Geological observations on volcanic islands. Smith, Elder Co., London

Delaney JR, Colony WE, Gerlach TE, Nordlie BE (1973) Geology of the Volc·n Chico Area on Sierra Negra volcano, Gal·pagos Islands. Geol Soc Am Bull 84: 2455-2470

Feighner MA, Richards MA (1994) Lithospheric structures and compensation mechanisms of the Gal·pagos Archipelago. J Geophyical Res 99: 6711-6729

Geist, D, Howard, KA, and Larson, P (1995) The generation of oceanic rhyolites by crystal fractionation: the basalt-rhyolite association at Volcan Alcedo, Galápagos Archipelago. J Petrol 36, 965-982.

Geist, D (1996) On the emergence and submergence of the Galápagos Islands, Noticias de Galápagos 56, 5-9.

Geist DJ, White WM, McBirney AR (1988) Plume asthenosphere mixing beneath the Gal·pagos Archipeligo. Nature 333: 657-660

Geist DJ, Howard KA, Jellinek AM, Rayder S (1994) The eruptive history of Volc·n Alcedo, Gal·pagos Archipeligo: a case study of rhyolitic oceanic volcanism. Bull Volcanol 56: 243-260

Geist DJ, McBirney AR, Duncan RA (1986) Geology and petrology of lavas from San Cristobal Island, Gal·pagos Archipelago. Geol Soc Am Bull 97: 555-556

Geist DJ, McBirney AR, Duncan RA (1985) Geology of Santa Fe Island: the oldest Gal·pagos volcano. J Volcanol Geotherm Res 52: 65-82

Graham, DW, Christie, DM, Harpp, KS, and Lupton, JE (1993) Mantle plume helium in submarine basalts from the Galápagos platform, Science 262, 2023-2026.

Hall, ML (1983) Origin of Espanola Island and the age of terrestrial life on the Galápagos Islands. Science 221, 545-547.

Hey R (1977) Tectonic evolution of the Cocos-Nazca spreading center. Geol Soc Am Bull 88, 1401-1406

Hickman, CS and Lipps, JH (1985) Geologic youth of Galápagos islands confirmed by marine stratigraphy and paleontology, Science 227, 1578-1580.

Ito, G., and J. Lin, 1995, Mantle temperature anomalies along the present and paleo-axes of the Galápagos spreading center as inferred from gravity analyses, J. Geophys. Res., 100, 3733-3745.

McBirney AR, Williams H (1969) Geology and petrology of the Gal·pagos Islands. Geol Soc Am Mem 118, pp 1-197

McBirney, AR (1993) Differentiated rocks of the Galápagos hotspot. In: Pritchard, HM, Alabaster, T, Harris, NBW, and Neary, CR (eds) Magmatic processes and plate tectonics, Geol Soc Spec Pub 76, 61-69.

Reynolds, RW and Geist, DJ (1995) Petrology of lavas from Sierra Negra volcano, Isabela Island, Galápagos archipelago, J Geophys Res 100, 24537-24553.

Reynolds, RW, Geist, D, and Kurz, MD (1995) Physical volcanology and structural development of Sierra Negra volcano, Isabela Island, Galápagos archipelago, Geol Soc Amer 107, 1398-1410.

Standish, J, Geist, D, Harpp, KS, and Kurz, MD (1998) The emergence of a Galápagos shield volcano: Roca Redonda, Contrib. Mineral. Petrol., in press.

Swanson FJ, Baitis HW, Lexa J, Dymond J (1974) Geology of Santiago, Rabida, and Pinzon Islands, Gal·pagos. Geol Soc Am Bull 85: 1803-1810

Verma SP, Schilling JG, Waggoner DG (1983) Neodymium isotopic evidence for Gal·pagos hotspot-spreading center system evolution. Nature 306: 654-657

Vicenzi EP, McBirney AR, White WM, Hamilton M (1990) The geology and geochemistry of Isla Marchena, Gal·pagos Archipeligo: an ocean island adjacent to a mid-ocean ridge. J Volcanol Geotherm Res 40: 291-315

White WM, McBirney AR, Duncan RA (1993) Petrology and geochemistry of the Gal·pagos Islands: portrait of a pathological mantle plume. J Geophyiscal Res 98: 19533-19564

Wilson, DS, Hey, RN (1995) History of rift propagation and magnetization intensity for the Cocos-Nazca spreading center. J Geophys Res 100: 10,041-10,056.