Temporal Variations in Plate Convergence and Eruption Rates in the Western Cascades, Oregon

TECTONICS, VOL. 6, NO. 2, PAGES 197-209, APRIL 1987

TEMPORAL VARIATIONS IN PLATE CONVERGENCE AND ERUPTION RATES IN THE WESTERN CASCADES, OREGON

Emily P. Verplanck and Robert A. Duncan

College of Oceanography, Oregon State University, Corvallis

Abstract. New K-Ar age determinations on INTRODUCTION basalts, basaltic andesites, and ash flow tuffs from the central Western Cascades in The Cascades Volcanic Arc extends from Oregon range in age from 32 to 3 Ma. The northern California to southern British ages decrease from west to east and with Columbia, a distance of over 1000 km. Along increasing elevation. Volcanism has been this linear belt, volcanic activity began continuous throughout the evolution of the at approximately 42 Ma [Lux, 1981] with the Western Cascades, with some periods of eruption of basalt and basaltic andesire greater activity: 13-16 Ma, 22-26 Ma and lavas and dacitic ash flow tuffs. This 29-31 Ma. Relative volume estimates continental margin volcanic arc is indicate that eruption rates decreased by a associated with the subduction of the Juan factor of 6 from 35 Ma to the present. The de Fuca (Farallon) plate beneath the eruption rates during formation of the Pacific Northwest margin of North America Western Cascades Volcanic Arc were (Figure 1). The Cascades Volcanic Arc is influenced by changes in the direction and composed of two provinces; rocks of the rate of convergence between the Farallon older Western Cascades exposed along the and North American plates. We have western slope of the range are 42-10 Ma, developed a model of the Tertiary and rocks of the younger High Cascades on convergence of these two plates, based on the eastern margin are 10-0 Ma. the mantle-fixed hotspot reference frame, Geologic processes at $ubduction zones which indicates a factor of 5 decrease in are complex and difficult to separate. The convergence rate (16.0 to 3.2 cm/a). timing, spatial variation and volume of Clockwise rotation of the western margin of erupted volcanic material may depend on a the North American plate led to a decrease number of parameters, such a• the age of in the convergence angle since about 35 Ma. the subducting lithosphere and its volatile Apparently, slower, more oblique subduction content, the dip angle of mubduction, the resulted in a decrease in the volume of thickness of overlying continental erupted magmas. lithosphere, and the convergence rate. While it is impossible to isolate any one variable, the parameter of convergence rate Copyright 1987 can be examined in the evolution of by the American Geophysical Union. volcanism in the Cascades Volcanic Arc for the following reasons: Paper number 6T0722. (1) The age of the plate entering the 0278-7407/87/006T-0722510.00 subduction zone has remained young (between 198 Verplanck and Duncan: Plate Convergence and Eruption Rates

49 ø

OLYMPIC MTS.

JUAN DE FUCA PLATE

KLAMATH MTS.

42 ø 131 ø 121 ø Fig. 1. Physiographic boundaries of western Oregon and Washington and plate boundaries of the Pacific, Juan de Fuca and North American plates. Study area is indicated by stippled pattern.

10 and 20 Ma age) during the time of geochronological study of Western Cascades Western Cascades development [Engebretson, lavas and tuffs. 1982; Duncan and McElwee, 1984] . (2) The dip of the Benioff zone and depth of THE CENTRAL CASCADES VOLCANIC ARC melting are dependent on the age of the subducted slab [Abbott and Hoffman, 1984], Eruptive activity began approximately 42 which from (1) has been constant. (3) The Ma [Lux, 1981] in the westernmost exposed present crustal thickness beneath the part of the Cascades Volcanic Arc (Figure Cascades increases eastward from 22 km 1). Lava and ash flows of this early under the Willamette Valley to 37 km under eruptive sequence have been uplifted, the High Cascades axis [Canoe, 1983]. tilted slightly eastward, and subsequently Plutonism has added material beneath the eroded, resulting in a deeply dissected and volcanic arc during the last 40 m.y., but rugged topography [Peck et al., 1964]. crustal thickening is unlikely to have Eocene marine sedimentary rocks of the exceeded a factor of 2. (4) There has Spencer, Tyee, and Umpqua formations been, however, a fivefold decrease in the consist of sandstone, mudstone, and convergence rate between the Farallon and conglomerate which overlie and are North American plates during the formation interbedded with the westernmost Cascades of the older Western Cascades of 42 to 10 volcanic rocks [Peck et al., 1964]. Priest Ma age (Engebretson [1982] and below) . and Vogt [1983] proposed to divide the Thus we can expect that the volume and volcanic stratigraphy into an early Western character of volcanism can be related Cascades episode of 42 to 18 Ma and a late principally to the effects of slowing Western Cascades episode of 18 to 9 Ma on subduction with time in the Cascades the basis of an erosional unconformity Volcanic Arc. In this paper we examine which spans 19 to 17 Ma. The early Western this hypothetical relationship through a Cascades rocks consist of massive beds of Verplanck and Duncan: Plate Convergence and Eruption Rates 199

ash flow tuffs of andesitic to dacitic the result of a sampling bias; that is, a composition, with interbedded olivine relatively continuous regional volcanic basalt and basaltic andesite lava flows. history can be interpreted as being The late Western Cascades episode, of 18-9 episodic due to incomplete sampling, Ma age, is characterized by flows and flow particularly in regions between volcanic breccias of andesire, basaltic andesire, centers which may exhibit age gaps in the and basalt. The lavas are interbedded with volcanic stratigraphy. volcaniclastic and epiclastic continental For younger rocks, such as the High rocks. In the northernmost part of the Cascade volcanoes of Quaternary age, study area (Figure 1) the late Western reasonably accurate volume estimates can be Cascades strata are approximately made. For older Western Cascade rocks it is horizontal and unconformably overlie the impossible to know the original volume of early Western Cascades lavas, which volcanic material. An unknown volume of generally dip gently eastward [Priest and pyroclastic material has been transported Vogt, 1983; G. Walker, personal by winds to areas outside the volcanic arc, communication, 1984], whereas in the south and much material has been removed by the two units are conformable [N. MacLeod, erosion. In addition, the areal extent of personal communication, 1984]. older units covered by younger units is Small intrusive bodies, which occur as unknown. If we make the assumption that pipes, dikes, sills, and small stocks, are erosion rates were constant through time, thought to be eruptive centers for portions then relative volumes of the older erupted of the Western Cascades volcanism [e.g., material can be calculated by measuring Hammond, 1979]. These lie in a belt along thicknesses of volcanic layers of known age the western slope of the Cascades in the range in areas where the stratigraphic eastern half of the study area, which is relationships are well mapped. coincident with a zone of hydrothermal and propylitic alteration. Younger volcanic TIME-SPACE DISTRIBUTION OF THE VOLCANISM rocks of Pliocene and Pleistocene age, related to early High Cascades volcanism, The main objective of this are found in the eastern margin of the geochronological study is to estimate Western Cascades and have flowed west onto eruption rates in the Western Cascades the eroded surface as intracanyon flows. Volcanic Arc. Two other objectives are to McBirney et al. [1974] have estimated determine the spatial variation of ages in the volume of erupted material in the the Western Cascades and to determine Cascade Range from Miocene to Quaternary whether the volcanic activity has been time and identified a large pulse of episodic or continuous during the past 42 activity in the middle Miocene (16-14 Ma), followed by smaller episodes during the Samples for this study were collected in late Miocene (11-9 Ma), Pliocene (6-4 Ma), collaboration with George W. Walker and and Holocene (1.5-0 Ma). They proposed Norman S. MacLeod (U.S. Geological Survey), that pulses of volcanic activity have who are completing geologic mapping of the occurred in approximately 5 m.y. intervals, Salem and Roseburg 1:250,000 sheets. The which they correlated with synchronous present investigation was planned to activity in the circum-Pacific area. complement their mapping by providing K-Ar Variations in the rate of plate convergence age determinations on stratigraphically have been suggested as the cause of controlled lava flows and ash flow tuffs. episodicity in island-arc and continental We thus have a total of 55 new dated sites margin volcanic activity [Scheidegger and from the central Western Cascades in Kulm, 1975; Kennett and Thunell, 1975; Oregon. When combined with the age Kennett, 1982]. Correlations between East determinations of Lux [1981, 1982], Sutter Pacific Rise spreading rate, hotspot [1978], Hammond [1979], Priest and Vogt activity, ash accumulation, and circum- [1983], there are now over 150 K-Ar age Pacific volcanism during the last 4 m.y. determinations of volcanic rocks from the were reported by Rea and Schetdegger [1979]. Western Cascades of Oregon (summarized by The geochronological data presented by Fiebelkorn et al. [1983]). Sutter [1978] and Lux [1981, 1982] from the All of our sample sites are between 43 ø Cascades of Oregon appear to extend the 5 and 45 ø north latitude and 122 ø and 123 ø m.y. periodicity back to 40 Ma. However, west longitude. The types of outcrops Priest and Vogt [1983] contended that the sampled include roadcuts, active and periodicity observations could be merely abandoned quarries, stream cuts, and 200 Verplanck and Duncan' Plate Convergence and Eruption Rates

123 ø 122 ø 121 ø 45 ø I Salem • o

- - / A Mr.Jefferson

IV.Sister Mid. Sister S. Sister 44 ø

Age(M(]) This study Others

<15 15-20 ß 20-25 ß 25 -$0 ß 50-$5 ß >55 •Lux(1981,1982) Priest 8 Vogt (1983) Roseburg A Sutter Mr. Thie/sen F/'ebe/korn ef g/. 43 ø Smith el gl. (1982)

Crater Lake

Grants Pass McL oughlin o ^

Medford

42 ø 125 ø 122 ø Fig. 2. Spatial variations of age determinations in the Western Cascades, Oregon. The isochrons which are interpolated from dated lava flows and ash tuffs are oriented aproximately north-south and decrease from west to east. Isochrons are projected across drainages to simplify the diagram; in reality, they closely follow the topographic contours. The locations of five east-west cross sections are indicated by numbered line segments.

ridge-capping lavas. Intrusive rocks were whole rock, three were plagioclase avoided because of their uncertain separates, and one was a glass separate. stratigraphic relation with the lavas and Most dated specimens are from basalt, ash flow tuffs. With the guidance of the basaltic andesitc, or andesitc lava flows, field geologists, we were able to sample except for a few ash flow tuff samples of fresh-appearing rocks from locally dacitic or rhyodacitic composition. Almost continuous, mapped units. In some places, all of the rocks are plagioclase-phyric, age determinations assisted in identifying and some also contain clinopyroxene stratigraphic horizons for correlation phenocrysts. A few of the samples have between drainages. olivine phenocrysts, most of which have Conventional K-Ar techniques [Dalrymple been partly altered to iddingsite or clay and Lanphere, 1969] were employed to minerals. Ages on these samples range from determine ages for the 55 samples from the 3.0 to 32.1 Ma, and most are between 12 and central Western Cascades in Oregon' 51 were 32 Ma. These new age determinations are Verplanck and Duncan' Plate Convergence and Eruption Rates 201

Histogram of Age Determinations 1 •10- E 8 . 6' o 4

• 2' E z 5 10 15 20 25 30 35 40 45 Age (Ma) Fig. 3. Histogram of age determinations (K-Ar and 40Ar/39Ar). Solid black' this study. Stippled pattern' Lux [1981, 1982], Priest and Vogt [1983], Sutter [1978] and Fiebelkorn et al. [1983]. All data are from the Western Cascades, Oregon. available from the AGU data file, on between dated locations. Several samples microfiche. • See Verplanck [1985] for are from intracanyon flows which are out of complete sample descriptions and further sequence; their ages are not used in this details on age determinations. representation of the volcanic stratigraphy. The isochrons are projected Spatial Variations acrossdrainages to simplify the diagram As notedpreviously, the regionaldip of and, in general, follow the topographic the Western Cascadesvolcanic strata is contours. Wider spacing between tsochrons approximately 3ø-10øE, except for some of suggests that thicker amounts of volcanic the younger flows (<18 Ma), which dip more material were deposited, assuming the gently. With this relatively continuous volcanic stratigraphy has a regionally regional structure, the oldest volcanic constant dip. rocks should be exposed on the western The new K-Ar age determinations plus all margin, with ages decreasing progressively previously reported geochronologtcal data for the volcanic rocks of the Western eastward. Not surprisingly, this is what is observed in the distribution of Cascades province are plotted on a histogram in Figure 3. The results do not crystallization ages in the Western appear to indicate a periodicity in the Cascades. Complicating factors such as volcanism. There are certainly no minor faults, folds, and intracanyon flows, significant periods of quiescence. The data which disrupt the continuous stratigraphic show more or less continuous volcanic relations, do occur, but overall, the ages of the Western Cascades volcanic units activity throughout the mid-to-late decrease eastward and with increasing Tertiary, with possibly greater activity between 13-16 Ma, 22-26 Ma, and 29-31 Ma. elevation. Hence, there is little if any support for a Spatial variations of the ages of the 5-m.y. eptsodtctty in volcanism in the volcanic units are illustrated in Figure 2. Western Cascades from these age data. The new K-Ar age determinations are However, longer-term changes have occurred combined with reported geochronological in eruption rates, which we discuss next. data from the Western Cascades [Sutter, 1978; Lux 1981, 1982; Priest and Vogt, ..Temporal Variations in Eruption Rates 1983; Fiebelkorn et al., 1983; Smith et al., 1982]. Isochrons, i.e., lines of Previous workers have attempted to constant age in 5-m.y. intervals from 35 to estimate volumes of erupted material in the 10 Ma, were constructed by interpolation younger, Quaternary-age, High Cascades volcanic rocks, where the exposures are good and the volcanic units have been well ISupplement (data, tables, calculations, mapped [McBirney et al., 1974]. Reasonably etc.) is available with entire article on accurate quantitative volumetric data have microfiche. Order from American been obtained; the only serious unknown Geophysical Union, 2000 Florida Avenue, factor is the amount of volcanic material N.W., Washington, DC 20009. Document removed via wind and water transport. The T87-002; $2.50. Payment must accompany Western Cascades are deeply dissected by order. erosion and extensively covered by 202 Verplanck and Duncan: Plate Convergence and Eruption Rates

vegetation, and it is much more difficult stratigraphy (e.g., Figure 4). We have to estimate volumes of erupted material. constructed five such cross sections Also, the volcanic material produced by the perpendicular to the Western Cascades in Western Cascades volcanoes is more the study area (locations shown in Figure extensive than the High Cascade lavas. 2). The new and previously dated samples Robinson et al. [1984] suggested that, are located by elevation and longitude. beginning about 37 Ms, voluminous dacitic Each cross section combines age data within to andesitic air-fall material of the John an east-west swath approximately 20 km Day Formation of north central Oregon was wide. derived from the volcanoes within the An average dip slope of 5 ø was used for Western Cascades Range. Therefore, there the early Western Cascades sequence (>18 is presumably Oligocene and Miocene age Ms). A 2 ø dip slope was used for the late Western Cascades volcanic material Western Cascades sequence (<18 Ma) in the underlying the High Cascades whose volume northern half of the study area and a 5 ø cannot be measured. dip slope, conformable with the earlier We can approach this problem in a sequence, in the southern half. We next different way. Instead of volume calculated the stratigraphic thickness calculations, relative thickness estimates between dated samples (lavas and ash can be used to compare rates of volcanic tuffs). Ages from closely spaced samples activity through time. To do this, however, were averaged to date some volcanic certain assumptions are necessary. First, horizons. Samples from intracanyon flows erosion rates are assumed to have been more and landslide blocks were not included in or less constant during the time period the calculations. 40-10 Ms. Presumably there have been minor The thicknesses were converted to temporal fluctuations in the erosion rate, eruption rates by dividing by the time but we do not know of a method to quantify interval between dated samples. These these changes. Second, the variation in eruption rate estimates, from the five distances from volcanic centers is assumed cross sections, are plotted in Figure 5 to be averaged by the density and breadth against age and plate convergence rate of sampling. In general, proximity to a (calculated in the next section). It is volcano will result in thicker lava and ash apparent that eruption rates have flow deposition. This effect can be consistently decreased with time throughout neglected if sufficient transects are the Western Cascades study area. taken, thereby averaging out the variations Another possible mechanism to explain in distances from eruption sites. Third, the decrease in thickness of volcanic the sedimentary layers are assumed to be material is the progressive eastward interbedded in approximately equal migration of volcanic centers. The basalt proportions between the volcanic units and basaltic andesitc volcanic centers of throughout the Western Cascades the early Western Cascades are located up stratigraphic sequence. This assumption is to 50 km west of the late Western Cascades only valid for an extended time frame. The centers [Peck et al., 1964]. However, the thickness of sediment deposited between early Western Cascades volcanic sequence is 30-20 Ma and 20-10 Ma is roughly the same exposed to the west of the late Western [G.W. Walker and N.S. MacLeod, personal Cascades volcanic sequence. The horizontal communication, 1984]. The sediment input distance between the 30 Ma and 10 Ma increases the eruption rate (equal to isochrons varies from about 25 to 55 km. accumulation rate) estimates, but the The distance from the ancestral volcanic erosion of material decreases the estimates centers to the sampling sites, then, has by a greater amount. Therefore the not changed significantly with time. The calculated eruption rates are probably decrease in thickness of volcanic material minimum estimates of the original with time, therefore, appears to reflect a thicknesses of volcanic material deposited true decrease in eruption rate. with time. Table 1 compares average eruption rates In areas where the structure is well for 5-m.y. intervals from 35 Ma to the known from mapping [Smith et al., 1982; present. We conclude that eruption rates G.W. Walker and R.A. Duncan, unpublished have decreased by at least a factor of 6, manuscript, 1987; N.S. MacLeod, personal from 400 m/Ma to 60 m/Ma, during the communication, 1984], it is possible to use evolution of the Western Cascades. In the the geochronological data to produce next sections we relate this decrease in east-west cross sections of the volcanic volcanic activity to slowing convergence Verplanck and Duncan' Plate Convergence and Eruption Rates 203

o o o o o o o o ko o o

e- c•

N t-

o o o o u• o (m) uol{eAe13 204 Verplanck and Duncan: Plate Convergence and Eruption Rates

700- plate convergence and the rate of magmatic activity in the associated island arc or

600- continental arc [Scheidegger and Kulm, ß 1975; Kennett and Thunell, 1975; Kennett, 1982; Wells et al. , 1984] . To examine this 500- relationship for the Cascades Volcanic Arc, we have developed a model for the •E400. convergence of the Farallon and North American plates during Tertiary time. The convergence model is based on two major c300- o assumptions: that hotspots in the Atlantic ._ and Pacific basins have remained fixed with

LU200- respect to each other and that symmetrical spreading occurred along the ocean ridge separating the Farallon and Pacific oceanic plates. The former assumption has been 100- tested by Minster and Jordan [1978], Morgan O'.•__•• t• 1• 1,5Age(Ma) 27 2•0 •5 [1981] , and Duncan [1981] , who conclude 0 I • ConvergenceRate(•m/a) 6 7 that hotspots move less than 0.5 cm/a with respect to one another. This is about 1 Fig. 5. Eruption rates versus age and order of magnitude less than the plate plate convergence rate during the evolu- velocities, and, therefore, inter-hotspot tion of the Western Cascades volcanic movement is small enough that hotspots arc. Triangle is based on volume and constitute a useful reference frame. area estimates of late Miocene to recent The second assumption is necessary volcanic activity [McBirney et al., because most of the Farallon plate, all 1974]. Eruption rate and convergence that is older than 11 Ma (polarity chron rate are highly correlated and propor- 5), has been subducted beneath North tional. Nominal best-fitting linear and America. Therefore the only record of the quadratic functions are shown. spreading history prior to late Miocene time is the magnetic anomaly pattern on the between the Farallon and North American Pacific plate. Symmetrical spreading is a plates. reasonable assumption because it has been observed in the last 11 Ma at the Juan de MODELLING CONVERGENCE BETWEEN THE FARALLON Fuca-Pacific spreading ridge south of the AND NORTH AMERICAN PLATES Soyanco Fracture Zone [Riddihough, 1977]. The convergence calculations can be Numerous authors have proposed that described in three steps. First, the there is a relationship between the rate of absolute motion of the Pacific and North

TABLE 1. Estimated Eruption Rates During Formation of the Western Cascades, Oregon Average Eruption Plate C•nvergence Time Interval, Ma Ratea m/Ma Rate , cm/a

35-30 399 6.0

30-25 297 5.2

25-20 236 5.2

20-15 190 4.0

15-10 109 3.2

10-0 59c 1.3

a b From estimates in Figure 5.

c From Figure 7. From volume and area estimates of McBirney et al. [1974]. Verplanck and Duncan: Plate Convergence and Eruption Rates 205

TABLE 2. Stage Poles for Farallon-North American Plate Convergence, 70 m.y. to Present

TIME, Rotation Pole Ma Latitude(+øN) Longitude(+øE) Angle(+øClockwise)

0- 5 35.1 -121.5 4.9 5-10 55.9 133.7 -2.2 10-15 27.6 108.7 -2.1 15-20 40.1 119.6 -1.9 20-25 42.1 127.6 -2.2 25-30 58.9 139.4 -3.2 30-35 58.5 139.4 -3.1 35-40 51.2 107.5 -5.1 40-45 45.4 102.4 -6.6 45-50 49.2 115.3 -6.7 50-55 55.1 136.4 -6.8 55-60 39.5 131.9 -6.5 60-65 31.1 128.2 -6.5 65-70 24.8 141.1 -7.4

American plates is found from the reference present rate of 3.4 cm/a. The decrease frame of hotspots which are fixed with occurs relatively continuously with time respect to the mantle [Duncan and Clague, except for a more sudden decrease at 40 Ma. 1985; Duncan, 1984]. The second step is to This significant decrease in rate is partly establish the motion of the Farallon plate a result of the change in Pacific plate relative to the Pacific plate. The magnetic absolute motion recorded by the Emperor linearions on the Pacific and Juan de Fuca Seamounts and Hawaiian Ridge intersection plates [Atwater and Menard, 1970; Riddihough, 1977] are used in conjunction with the magnetic polarity time scale FARALLON-NORTH AMERICA CONVERGENCE [Harland et al., 1982] to determine 70 Ma to the PRESENT rotation poles for Farallon-Pacific relative motions, following Engebretson

[1982]. The third step is to add the 7o Farallon-Pacific relative motion to the Pacific absolute motion; their sum is the Farallon absolute motion. The North American absolute motion is added to the 55 Farallon absolute motion; their resultant 6•65 50 is the convergent motion of the Farallon 45 and North American plates. The stage poles for the Tertiary to present history of this plate convergence are given in Table 2. 35 40 Figure 6 illustrates the results of the 15230 25 convergence model in terms of vectors whose length indicates the rate and whose azimuth is the direction of convergence between the Farallon and North American plates at Cape W • ConvergenceRateScale(cm/a)E Blanco, Oregon. Each vector is an average over a 5-m.y. interval, and the number $ beside each vector indicates the beginning Fig. 6. Farallon-North American plate of the time interval. It is apparent that convergence model. The vectors indicate the convergence has always been to the the rate and direction of convergence northeast from beginning Tertiary to the between the Farallon and North American present, with a variable component of plates at Cape Blanco, Oregon, during the northward translation. There has been a Tertiary. The number beside each vector drastic decrease in convergence rate, indicates the beginning of a 5-m.y. time however, from 16.5 cm/a at 70 Ma to the interval. 206 Verplanck and Duncan: Plate Convergence and Eruption Rates

Coast Range and Western Cascades Range have 15- rotated clockwise [e.g., Simpson and Cox, 1977; Magill and Cox, 1980], thereby increasing the obliquity of convergence motion at the margin with time. If we e10- assume that the Cascades and Coast Range have rotated together since the accretion of the latter at about 42 Ma, then the orientation of the subduction zone associated with Cascades volcanism can be 5' r'---i __-i i ,, reconstructed. Figure 8 illustrates the changes in both the convergence motion and ! '. the subduction zone orientation with time, in 5-m.y. intervals, since 40 Ma. Prior to %0 4O 3O 2O 10 0 40 Ma and the Coast Range accretion, the Time (Ma) orientation of the northwestern margin of Fig. 7. The normal component of con- the United States is not well enough vergence between the Farallon and North constrained to make predictions of the American plates since 50 Ma: (1) This convergence angle. study. Model is based on the absolute In summary, the rate and orthogonal reference frame, and the site of con- component of convergence motion between the vergence is Cape Blanco, Oregon. The Farallon and North American plates has normal component of convergence is cal- decreased during the evolution of the culated assuming the Western Cascades Western Cascades Volcanic Arc. The have rotated clockwise since 40 Ma, as determined from paleomagnetic data (see text). (2) Engebretson [1982]. Model is NORTH AMERICA/FARALLON CONVERGENCE based on the absolute reference frame.

Shown is the normal component of Farallon convergence__ •,. orthogonalrate plate motion near present day San Fran- 10' angle;• 8.5cm/a cisco, relative to a fixed North America. OuJ 0' 3RTHOGONALCONVERGENCE Z

uJ -10. z 8•5o5'8 o at 43 Ma [Clague et al., 1975]. o -J -20- Coincidentally, this is also the time of z o •5.5 the Coast Range accretion [Duncan, 1982; •5.5 Heller and Ryberg, 1983] and the initiation of Western Cascades volcanic activity [Lux, •D 4.0 1982]. u_ 3.4 Engebretson [1982] has modeled the (D -50- convergence of the Farallon and North American plates during Tertiary time, also • -60- based on the absolute reference frame (mantle-fixed hotspots). His results are -7O- 3

compared to our new calculations in Figure 0 5 10 15 20 25 30 35 40 7. The normal component of convergence near TIME (Ma) present-day San Francisco (Engebretson's Fig. 8. The change in convergence model) and near present-day Cape Blanco motion between the Farallon and North (this study) are quite comparable. Both American plates since 40 Ma. Vectors models indicate a decrease in convergence indicate direction of convergence, and rate with time, the most significant line segments illustrate the orientation decrease occurring between 45 and 35 Ma. of the continental margin as deduced from The direction of convergence between the paleomagnetic data from the Coast Range Farallon and North American plates during and Western Cascades. Numbers on the Tertiary time had an average azimuth of right are the orthogonal rates of con- N45øE. Since 40 Ma, however, the 5-m.y. vergence (cm/a), and degrees indicate the interval convergence vectors have rotated convergence angle. Convergence has progressively counterclockwise from N53øE become increasingly oblique with time to N20øE. At the same time, the Oregon since about 35 Ma. Verplanck and Duncan: Plate Convergence and Eruption Rates 207

orthogonal rate has decreased by a factor Aleutian volcanic centers decrease in size of 5 from 8.5 to 1.6 cm/a, and the angle of from east to west, coincident with a convergence has decreased by two thirds, decrease in the velocity and angle of from 90 ø to 30 ø , between 40 and 10 Ma. The convergence of the Pacific plate. Karig next step is to compare this quantitative [1982] suggested that intensity of arc convergence model with temporal variations volcanism is influenced by the rate of in volcanic activity in the Western subduction. Karig and Kay [1981] compiled Cascades. data on rates of magmatism and of plate convergence for numerous active arc CORRELATION OF ERUPTION RATES WITH systems. A positive correlation between CONVERGENCE HISTORY rates of subduction and eruption was shown within the Lesser Antilles, Mariana, Central America and Cascades volcanic arcs. This study provides evidence that the temporal variation in eruption rate is a Orogenic periods of increased plutonism in the Andean Volcanic Arc coincided with result of the change in subduction rate of the Farallon plate beneath North America. periods of relatively higher plate The depth of partial melting is dependent convergence [Frutos, 1981; Bussell, 1983]. on the age and subduction rate of the As shown in Table 1 and Figure 5, oceanic lithosphere [Abbott and Hoffman, subduction (orthogonal convergence) rates 1984]. In the CascadesVolcanic Arc, the have decreasedfrom about 6.3 cm/a to 1.3 portion of the Farallon plate which entered cm/a at the sametime that eruption rates the subduction zone has remained young in the Cascades Volcanic Arc dropped from (10-20 Ma, Engebretson [1982]) for the past about 400 to 60 m/Ma. Therefore it appears 40 m.y., and therefore the age of the that subduction and eruption rates show descending plate has not changed similar trends and are roughly proportional significantly. The convergence rate, to one another during the evolution of this however, is a significant variable and volcanic arc. This is an important result decreased drastically by a factor of 5 because it may be used to correlate plate during the evolution of the Western motions and magma production in volcanic Cascades. arcs around the world. Faster convergence rates increase the The convergence rates were depth at which the subducting plate attains significantly greater (11 to 16 cm/a) prior thermal equilibrium with the surrounding to 40 Ma. It will be an interesting future mantle [Abbott and Hoffman, 1984] and, study to comparethe eruption rates of therefore, dehydrationreactions take place volcanic material which predates the at a deeperlevel. This conceivablycauses WesternCascades and test whetherthe more melting in the region 80-200 km and subduction rate and eruption rate are consequently more magma to be produced. indeed proportional to one another in this Volcanism is not necessarily proportional earlier period. The subduction zone is to magma generation, however. Additional thought to have been located further east factors are the geochemical composition of in the past, and volcanism was located the magma, the stress and strain regime, along the Challis axis in Washington and and the thicknessof the overlying crust. Idaho. The emplacementof the CoastRange Intuitively, however,one would expect that by approximately42 Ma cloggedthe previous more subducted material would result in a subduction zone and thereby caused the greater amount of volcanism in the subduction to jump seaward [Simpson and overlying arc. Cox, 1977]. We would expect the eruption Variations in volcanic activity in rates of volcanic material in the ancestral island-arc and continental-margin areas volcanic arc to greatly exceed those of the have been proposed to result from changes Western Cascades if all other factors in the rate of plate convergence remained equal. [Scheidegger and Kulm, 1975; Kennett and Thunell, 1975; Kennett, 1982]. Wells et SUMMARY al. [1984] suggested that the decrease in Farallon-North American convergence may be Geochronological data correlate time responsible for shifting the location of variations in eruptive history of Western the volcanic arc to the west, decreasing Cascades volcanism with changes in the the horizontal compressive stress on the convergence pattern between the Farallon continent and increasing ash flow tuff and North American plates. Our results, volcanism. Marsh [1982] observed that the combined with those for previously reported 208 Verplanck and Duncan: Plate Convergence and Eruption Rates ages, indicate that the oldest volcanic Clague, D. A., G. B. Dalrymple and R. material is exposed on the western margin Moberly, Petrography and K-Ar ages of the arc and the ages decrease with of dredged volcanic rocks from the increasing elevation and distance eastward, western Hawaiian Ridge and the southern which agrees with the volcanic Emperor seamount chain, Geol. Soc. Am. stratigraphic relations. Volcanism has Bull., 86, 991-998, 1975. been continuous throughout the evolution of Dalrymple, G. B. and M. A. Lanphere,

the Western Cascades, although some Potassium-Armon_ Datinm:_ Principles._ intervals between 13-16 Ma, 22-26 Ma, and Techniques, and Applications to 29-31 Ma appear to be have been more Geochrono•og•y, 258 pp., W. H. Freeman, active. Most important, the New York, 1969. geochronological data presented here are Duncan, R. A., Hotspots in the southern used to estimate eruption rates during oceans -- An absolute frame of reference Western Cascades volcanic activity. We find for motion of the Gondwana continents, that eruption rates have decreased by at Tectonophysics, 74, 29-42, 1981. least a factor of 6 from 35 Ma to the Duncan, R. A., A captured island chain in present. the Coast Range of Oregon and Washington, Fromour modelcalculations of the J. Geophys.Res., 87, 10827-10837,1982, history of convergencebetween the Farallon Duncan,R. A., Age progressive volcanismin and North Americanplates we find that the the NewEngland seamounts and openingof

convergence direction has become more the central Atlantic Ocean, J. GeoDh¾s._ _ oblique and the convergence rate has Res,, 89, 9980-9990, 1984. decreased by a factor of 5 during the Duncan, R. A., and D. A. Clague, Pacific volcanic evolution of the Western Cacades. plate motion recorded by linear volcanic The changes in direction and rate of chains, in A. E. M. Nairn and F. G. convergence directly influence the volume Stehli, The Ocean Basins and Margins.,. 7A, of volcanism in the associated volcanic The Pacific, 89-121, 1985. arc. The decrease in convergence rate Duncan, R., and K. McElwee, Cenozoic plate results in less material being subducted in reconstructions, in Western North a given time and a corresponding decrease American Continental Maygin and Adjacent in the amount of magma produced and Ocean Floor off Oremon and Washinmton. _ _ erupted. We suggest that a slower Ocean Margin Drilling Program. Regional convergence rate results in a shallower Atlas Ser,, Atlas 1, L. D. Kulm et al., loss of volatiles and therefore a sheet 12, Marine Science International, decreasing degree of partial melting at Woods Hole, Mass, 1984. depth. Engebretson, D., Relative motions between oceanic and continental plates in the Acknowledgments. We thank George Pacific Basin, Ph.D. thesis, 211 pp., Walker and Norm MacLeod for their guidance Stanford Univ., Stanford, Calif., 1982. in sampling. The research was supported by Fiebelkorn, R. B., G. W. Walker, N. S. the U.S. Geological Survey and NSF grants MacLeod, E. H. McKee and J. G. Smith, EAR-8416946 and OCE-8219189. The Index to K-Ar age determinations for the manuscript was greatly improved by the state of Oregon, Isochron/West, 37, 3-60, comments of Martin Fisk, Craig White, and 1983. two anonymous reviewers. Frutos, J., Andean tectonics as a consequence of sea-floor spreading, REFERENCES Tectonophysic•, 72, 21-32, 1981. Canoe, S., Investigation of Pn propagation Abbott, D. H., and S. E. Hoffman, Archaean in Oregon, 97 pp., M.S. Thesis, Oregon plate tectonics revisited, 1, Heat flow, State Univ., Corvallis, 1983.

spreading rate and the age of subducting Gill, J. B., Orozenic_ Andesires and Pla• oceanic lithosphere and their effects on Tectonics, 390 pp. Springer-Verlag, New the origin and evolution of continents, York, 1981. Tectonics, •, 429-448, 1984. Hammond, P. E., A tectonic model for the Atwater, T., and H. W. Menard, Magnetic evolution of the Cascade Range, in 1tneattons in the north-east Pacific, Cenozoic Palcogeography of the W•teNp Earth Planet. Sci. Lett., 1, 445-450, United States, edited by J. M. 1970. Armentrout, M. R. Cole, and H. TerBest, Bussell, M. A., Timing of tectonic events Jr., pp. 219-237, Society of Economic in the Central Andes of Peru, J. Geol. Paleontologists and Mineralogists, Tulsa, Soc. London, !40, 279-286, 1983. Okla., 1979. Verplanck and Duncan: Plate Convergence and Eruption Rates 209

Harland, W. B., A. V. Cox, P. G. Llewellyn, of the Western Cascade Range in Oregon, C. A. G. Pickton, A. G. Smith, and R. U•S. Geol, Su•v, Prof. Pap,, 449, 1-56, Walters, • Geological T•me Sc•, 131 1964. pp., Cambridge University Press, New Priest, G. R., and B. F. Vogt, Geology and York, 1982. geothermal resources of the central Heller, P. L. and P. T. Ryberg, Sedimentary Oregon Cascades, Spec. Pap. Oreg. Dep, record of subduction to forearc Geol. Miner. Ind., 15, 1-123, 1983. transition in the rotated Eocene basin of Rea, D. K. and K. F. Scheidegger, Eastern western Oregon, Geology, 11, 380-383, Pacific spreading rate fluctations and 1983. its relation to Pacific area volcanic Karig, D. E., Temporal relationships episodes, J. VolcanoS, Geotherm, Res,, between back-arc basin formation and arc •, pp. 135-148, 1979. volcanism with special reference to the Riddihough, R. P., A model for recent plate Phillipine Sea, in The Tectonic and interactions off Canada's west coast, Geologic Evolution of Southeast Asian Can. J. Earth S½•,, 14, pp. 384-396, Seas and Islands, Part 2, Geophys. 1977. Monogr. 27, pp. 318-325, edited by D.E. Robinson, P. T., G. F. Brem, and E. H. Hayes, AGU, Washington, D.C., 1982. McKee, John Day Formation of Oregon' A Karig, D. E. and R. W. Kay, Fate of distal record of early Cascade volcanism, sediments on the descending plate at Geology, 12, pp. 229-232, 1984. convergent margins, Philos. Trans. R, Schetdegger, K. F., and L. D. Kulm, Late Soc. London Se•. A, 30•, 233-251, Cenozoic volcanism in the Aleutian arc' 1981. Information from ash layers in the Kennett, J.P., Marine Geology, 813 pp., northeastern Gulf of Alaska, Geol. Soc. Prentice-Hall, Englewood Cliffs, N.J., Am. Bull., 86, pp. 1407-1411, 1975. 1982. Simpson, R. W., and A. V. Cox, Kennett, J.P., and R. C. Thunell, Global Paleomagnettc evidence for tectonic increase in Quaternary explosive rotation of the Oregon Coast Range, volcanism, Science, 187, 497-502, 1975. Geolozv, i, 585-589, 1977. Lux, D. R., Geochronology, geochemistry and Smith, J. G., N.J. Page, M. G. Johnson, petrogenests of basaltic rocks from the B.C. Mortng, and F. Gray, Preliminary Western Cascades, Oregon, Ph.D. geologic map of the Medford 1 ø by 2 ø dissertation, 171 pp., Ohio State Univ., quadrangle, Oregon and California, U.S. Columbus, 1981. Geo!, Surv, Open Fi•e Rep,, 8•-0955, Lux, D. R., K-At and 4øAr-S•Ar ages of 1982. mid-Tertiary volcanic rocks from the Sutter, J. F., K/At ages of Cenozoic Western Cascade Range, Oregon, volcanic rocks from the Oregon Cascades Isoch•o•/Wes•, 33, 27-32, 1982. west of 121•30 • , Isochron/We$•, 21, Magtll, J., and A. Cox, Tectonic rotation 15-21, 1978. of the Oregon Western Cascades, Spec. Verplanck, E. P., Temporal variations in Pap, Oreg, Dep. Geol. Miner. Ind., volume and geochemistry of volcanism in 10,1-67, 1980. the Western Cascades, Oregon, M.S. Marsh, B., The Aleutians, in Andesites' thesis, 115 pp., Oregon State Univ., Orogen•c Andesires and Related Rocks, 724 Corvallis, 1985. pp., edited by R. S. Thorpe, John Wiley, Walker, G. W. and R. A. Duncan, Geology of New York, 1982. the Salem 2-degree Quadrangle, Oregon, McBirney, A. R., J. F. Sutter, H. R. U.S, Geol. Surv, publ,, in press, 1987. Naslund, K. G. Sutton and C. M. White, Wells, R. E., D.C. Engebretson, P. D. Eptsodtc volcanism in the central Oregon Snavely, Jr., and R. S. Coe, Cenozoic Cascade Range, Geolom_v, %, 585-589, 1974. plate motions and the volcanotectonic Minster, J. B., and T. H. Jordan, Present evolution of Western Oregon and day plate motions, J. Geophys. Res,, Washington, Tectonics, •, 275-294, 1984. 83, 5331-5354, 1978. Morgan, W. J., Hotspot tracks and the R. A. Duncan and E. P. Verplanck, opening of the Atlantic and Indian College of Oceanography, Oregon State Oceans, in ..The Se•, vol. 7, edited by C. University, Corvallis, OR 97331. Emiliani, Wtley-Intersctence , New York, pp. 443-475, 1981. Peck, D. L., A. B. Grigg, H. G. (Received September 15, 1986; Schlicker, F. G. Wells and H. M. Dole, revised December 4, 1986; Geology of the central and northern parts accepted December 4, 1986.)