Magma mixing and mingling in the Grayback pluton, Klamath Mountains, Oregon
Published:January 01, 2006
- PDF LinkChapter PDF
Kenneth Johnson, Calvin G. Barnes, 2006. "Magma mixing and mingling in the Grayback pluton, Klamath Mountains, Oregon", Geological Studies in the Klamath Mountains Province, California and Oregon: A volume in honor of William P. Irwin, Arthur W. Snoke, Calvin G. Barnes
Download citation file:
The 160-Ma Grayback pluton provides ample evidence for the role of mantle-derived magmas in the crystallization history of the magma chamber. Mafic magmas tapped several distinct mantle sources, and evidence for magma mixing is manifest in (1) widespread mafic enclave swarms in which basaltic magma mingled with basaltic to andesitic magma, (2) late-stage granodioritic dikes containing pillow-shaped enclaves of basalt and hybrid andesite, and (3) late-stage tonalitic dikes including microdioritic enclaves contained within a dark-colored hybrid matrix. These magma-mingling zones reflect large-scale magma mixing processes in the Grayback system. Much of the compositional variation among Grayback main-stage samples was apparently produced by magma mixing between mafic and intermediate magmas similar to those exposed in some of the mafic enclave swarms. Mixing between basaltic and granitic end-members at deeper levels in the Grayback magma chamber produced hybrid an-desite. Enclaves of basalt and hybrid andesite are observed in late-stage granodioritic dikes that cut the mafic enclave swarms. Mingling between partially crystalline basaltic and tonalitic mushes, resulting in a mega- and microscopically complex hybrid, is observed in the late-stage tonalitic dikes. Few Grayback main-stage samples were apparently produced in this manner.
The Grayback pluton is one of several similar-aged plutons in the Wooley Creek suite, in which mafic magma played a prominent role in petrogenesis. The Grayback system is unusual in that mafic magmas from several obviously different sources were present throughout its crystallization history. This circumstance probably resulted from its position in an extensional, back-arc setting.
Most Jurassic plutons in the Klamath Mountains province occupy “belts” or suites that are defined by relatively narrow age ranges, geographic settings, and exposed rock types (Hotz, 1971; Irwin, 1985; Irwin and Wooden, 1999). More recent work has shown that plutons within belts share broadly similar petrogenetic histories. This commonality is particularly the case for plutons within the Ironside Mountain batholith (Barnes et al., this volume, Chapter 10), the Wooley Creek suite (Barnes, 1983; Barnes et al., 1986, 1995; Gribble et al., 1990), which now includes several plutons from the Grayback belt of Irwin (1985), and for Late Jurassic to Early Cretaceous tonalitic and trondhjemitic intrusions in the eastern Klamath Mountains (Barnes et al., 1996; Allen and Barnes, this volume). Plutons within the Wooley Creek suite evolved largely as open systems, involving assimilation of crustal rocks, recharge of mantle-derived magma, and/or mixing between mantle and crustal melts. Detailed study of such open systems provides information about the source(s) of magmas, the nature of the crust through which they ascended, and the timing of magma mixing/mingling events.
Magma mixing is commonly invoked to explain compositional variations in plutonic and volcanic suites from a variety of tectonic environments. Evidence for the interaction of magmas at the outcrop scale is found in banded pumice and mafic magmatic enclaves. Mafic magmatic enclaves (hereafter referred to as “enclaves”) are common and often widespread features of both volcanic and plutonic rocks, and there is general consensus that they represent blobs of mafic magma that were rapidly un-dercooled when brought into contact with cooler, more felsic magma. Striking examples of basaltic “pillows” in felsic rocks have been presented that leave little doubt, at least at the outcrop scale, as to their magmatic origin (e.g., Wiebe, 1991; Wiebe et al., 2001). Often the petrogenetic significance of enclaves is uncertain, other than to indicate that mafic magma was present during the crystallization history of the pluton. If the presence of enclaves in mingling zones can be used to demonstrate that magma mixing was important in the compositional evolution of a magma, then such features may provide information about the physical and chemical processes involved in mixing.
The late Middle Jurassic Grayback pluton in southwestern Oregon is a large, reversely zoned member of the Wooley Creek plutonic suite. Magma-mingling zones are extensively exposed throughout the pluton. They feature a wide variety of intermingled rock types and provide evidence for mingling throughout the pluton's history. The observed mixing and mingling relationships indicate that mafic mantle-derived magmas were not only present, but played an important role in the evolution of the Grayback pluton. The Grayback pluton is not unique in this respect; all plutons in the Wooley Creek belt exhibit characteristics of open-system behavior and the involvement of mafic magmas during their crystallization histories (Barnes et al., 1986, 1987; Gribble et al., 1990). However, the Grayback pluton is particularly well suited for such study, because magma-mixing processes were arrested in situ in a variety of mixing stages and involved a range of end-member compositions. In this chapter, we describe the range of mingling zones in the Grayback pluton and demonstrate their utility as small-scale analogues to pluton-wide processes. We then discuss the tectonic conditions under which the Grayback system evolved, conditions that may have allowed for widespread mixing and mingling behavior. In the following text, “mixing” refers to the process of blending two or more magmas together to form reasonably uniform hybrid magma, and “mingling” refers to incomplete mixing in which end-members are physically distinct, typically resulting in enclave-bearing rocks or mingling zones.
The Klamath Mountains of northwestern California and southwestern Oregon comprise an arcuate assemblage of four lithotectonic terranes, which represent fragments of oceanic crust and island arcs that were accreted to the western margin of North America during Late Paleozoic through Middle Mesozoic time (Irwin, 1981). These units are, from east to west and from structurally highest to lowest, the eastern Klamath, central metamorphic, western Paleozoic and Triassic, and the western Jurassic terranes or belts (Fig. 1). The belts are separated by east-dipping thrust faults, so that structurally higher rocks were thrust over generally younger rocks to the west. Thrusting of the western Paleozoic and Triassic over the western Jurassic belt along the Orleans fault was responsible for some of the deformation associated with the Nevadan orogeny at ca. 152–150 Ma (Saleeby et al., 1982; Harper and Wright, 1984).
Calc-alkaline and calcic magmatism occurred during Jurassic and Early Cretaceous times. Barnes et al. (1992a) showed that pre-Nevadan plutons are chemically and isotopically distinct from post-Nevadan ones, reflecting a change from largely mantle-derived magmas to those formed by crustal melting in response to underthrusting of ophiolitic rocks beneath older terranes to the east (Barnes et al., 1992b, 1996). A sample from the Grayback pluton main stage yielded a zircon U-Pb age of 160 ± 2 Ma (Yule et al., this volume), whereas late-stage samples yielded a zircon U-Pb age of 157 ± 2 Ma (Yule et al., this volume) and a hornblende 40Ar/39Ar age of 157.3 ± 1.4 Ma (Hacker et al., 1995). These ages place Grayback magmatism within the age range of plutons belonging to the 165- to 156-Ma “Wooley Creek suite” that includes the Wooley Creek batholith and Slinkard pluton (Barnes et al., 1986), the Russian Peak pluton (Cotkin and Medaris, 1993), the Ashland pluton (Gribble et al., 1990), and the English Peak pluton (Schmidt, 1994). To the west of the Wooley Creek belt are the 165- to 159-Ma Josephine ophiolite (Wyld and Wright, 1988; Harper et al., 1994) and 161- to 156-Ma Rogue–Chetco volcanic arc complex (Garcia, 1982; Yule et al., this volume).
The Grayback pluton intruded rocks of the Rattlesnake Creek and western Hayfork terranes, the westernmost of four terranes of the western Paleozoic and Triassic belt (Irwin, 1972; Donato et al., 1996; Ernst, 1998). The Grayback pluton is the largest of a linear, north–northeast-trending chain of four coeval plutons (ca. 161–160 Ma; Yule et al., this volume) that consists of the Weimer, Grayback, Sucker Creek, and Thompson Ridge plutons. The host rocks that separate these plutons are essentially dike swarms with as much as 50% dike rock (Barnes et al., 1995; Donato et al., 1996). The presence of mafic dike swarms along the southern edge of the Grayback pluton, evidence for contemporaneous normal faulting, and its setting inboard of the coeval Rogue–Chetco arc led Barnes et al. (1995) to conclude that it was emplaced in a back-arc setting.
SAMPLE PREPARATION AND ANALYSIS
Samples were crushed and powdered in tungsten carbide and alumina shatterboxes. All samples were analyzed by inductively coupled plasma atomic emission spectrometry (ICP-AES) at Texas Tech University following methods described in Shannon (1994). Rare earth elements (REEs) were separated using the cation exchange elution method of Shannon (1994) and then analyzed by ICP-AES. Rubidium abundances were determined by flame atomic emission.
Mineral compositions were determined with a JEOL JXA-733 electron probe microanalyzer at Southern Methodist University, Dallas, Texas. Operating conditions were 15-kV accelerating potential, 20-nA beam current, and a 10-μm spot, defocused to 20 μm for plagioclase. Online corrections were made with the Bence-Albee procedure, and a suite of natural and synthetic minerals and glasses was used as standards.
GEOLOGY OF THE MAGMA-MINGLING ZONES
The main stage of the Grayback pluton consists of tonalite, quartz diorite, gabbro, and norite that were intruded by syn-plutonic noritic and gabbroic magmas, some of which resulted in widely dispersed enclaves throughout the pluton (referred to as “main-stage enclaves”). Late-stage activity was characterized by the intrusion of tonalitic and granodioritic dikes, many of which were enclave bearing. Detailed discussions of the geology and petrology of the Grayback pluton were presented in Godchaux (1969) and Barnes et al. (1995).
This study focuses on three types of magma-mingling zones in the Grayback pluton: (1) large, widespread swarms of mafic enclaves in a mafic to intermediate host; (2) late-stage granodioritic dikes containing pillow-shaped basaltic and andesitic enclaves; and (3) late-stage tonalitic dikes containing microdioritic enclaves enclosed in an intermediate matrix.
Mafic Enclave Swarms
Field Observations. These swarms consist of microgabbroic enclaves in gabbroic to dioritic host rocks. They are exposed on Lake Mountain and the ridge extending north from Grayback Mountain (Fig. 1). Some can be traced laterally for several hundred meters. They are complex intrusive zones within the main stage of the pluton and are intruded by coarse-grained hornblende gabbro and enclave-bearing felsic dikes.
Individual enclaves range from <1.0 cm to >2.5 m in longest dimension. Enclaves are generally lens-shaped and oriented parallel or subparallel to foliation in the host (Fig. 2A). In some places, enclaves constitute >90% of outcrop area and are separated by a thin septum (<0.5 cm) of host rock. The enclave boundaries are sharp and well defined, but lack the fine-grained “chilled” margins suggestive of rapid undercooling of the enclave magma.
On Grayback Mountain, enclaves are generally flattened at the margins of the mingling zones with aspect ratios approaching 35:1, whereas in the center of the zone, aspect ratios are estimated to be <10:1. The enclaves are plastically deformed around brittle xenoliths, resulting in pressure shadows in which prismatic hornblende is randomly oriented. Elsewhere in the mingling zones, hornblende is parallel to the foliation.
Veins of the host rock, generally <0.5 cm in width, are common in most enclaves. However, most do not penetrate the entire enclave, but pinch out. These veins are generally parallel to the foliation of the host from which they originate and to the direction of enclave elongation (Fig. 2B). Aggregates of coarse-grained plagioclase and poikilitic hornblende, as much as 1.5 cm in length, are common throughout the finer-grained enclaves. The long axes of these aggregates are parallel to the foliation in the host and to the veins of host within the enclaves. They are commonly collinear with pinched out veins (Fig. 2C and D).
Petrography. The host rocks to these enclave swarms are hornblende norite (on Grayback Mountain) and biotite-hornblende quartz gabbro (on Lake Mountain). Host rocks consist of euhedral to subhedral, reversely to normally zoned plagioclase (An48–63) + magnesio-hornblende (Mg/[Mg + Fe] = 0.52–0.62) + Fe-Ti oxides ± orthopyroxene (En53Fs47Wo0 to En46Fs52Wo2) ± biotite ± quartz (Johnson, 1991). Titanite and zircon are minor accessories, whereas apatite is common. Hornblende is prismatic or poikilitic, with inclusions of plagioclase, apatite, and oxides. Minor late-stage cummingtonite is associated with hornblende.
Enclaves are micronorite (Grayback Mountain) and biotite hornblende quartz microgabbro (Lake Mountain). They have the same mineral assemblages as their respective hosts. Plagioclase compositions in the enclaves reach An69, whereas hornblende and orthopyroxene compositions are similar to those in host rocks (Johnson, 1991). Plagioclase xenocrysts from the host are strongly zoned and are partially resorbed and embayed by quartz. Hornblende in the quartz microgabbro is prismatic, but it forms rims around pyroxene grains in the micronoritic enclaves. Aspect ratios of apatite in the enclaves rarely exceed 5:1 and are very similar to those in the host.
In thin section, enclaves typically exhibit a well-developed lamination defined by the parallel alignment of plagioclase laths (Fig. 2E), although enclaves with little or no lamination were observed. Lamination in the enclaves is generally parallel to the enclave/host interface, and may be deflected around mineral aggregates and xenocrysts from the host (Fig. 2F).
Late-Stage Granodioritic Dikes
Field Observations. Granodioritic dikes containing basaltic and andesitic “pillows” were studied at several localities including Lake Mountain summit, the U.S. Forest Service Boundary Trail south of Grayback Mountain, and along logging roads east of Oregon Caves National Monument in the southern portion of the pluton.
Basaltic enclaves are fine-grained, dark green to gray mafic pillowlike blobs that range from a few centimeters to slightly less than one meter in diameter. The basalt-granodiorite interface is generally very sharp and convex toward the host (Fig. 3A), and the enclaves lack a preferred orientation. The enclaves are commonly enveloped by trondhjemitic zones (Fig. 3B) ∼0.75 cm wide that lack K-feldspar and biotite. On fresh enclave surfaces, clots of randomly oriented plagioclase phenocrysts and clots of actinolitic hornblende 2–3 mm in length are visible.
Andesitic enclaves are less abundant than basaltic ones. They are fine-grained, pale-green, and intermediate in color between the basaltic pillows and granodioritic host. These enclaves are coarser grained than the basaltic enclaves, and plagioclase phenocrysts are not visible on fresh surfaces. In addition, their margins are more diffuse than those of the basaltic enclaves, and the granodiorite adjacent to the enclave/host contact is somewhat darker than elsewhere.
Petrography. Biotite granodiorite has medium-grained hypidiomorphic granular texture and consists of plagioclase + K-feldspar + quartz + biotite + Fe-Ti oxides. Apatite, titanite, al-lanite, and zircon are accessories. Inclusions of biotite, apatite, Fe-Ti oxides, and minor amphibole (hornblende?) are common in plagioclase. Quartz and K-feldspar have interstitial and poikilitic habits. Apatite forms euhedral acicular prisms with aspect ratios of 20:1 or greater. Amphibole occurs as small rounded blebs up to 0.1 mm in diameter.
Basaltic enclaves are porphyritic and consist of plagioclase + biotite + actinolitic hornblende + Fe oxides and accessory titanite and apatite. Euhedral to subhedral plagioclase phenocrysts reach 3.0 mm in length. Hornblende and biotite are abundant groundmass phases. Aggregates of actinolitic amphibole (up to 1.0 mm in diameter) are rimmed by biotite, and probably represent relict pyroxene (e.g., Castro et al., 1990). Apatite occurs as acicular prisms; some are hollow, and aspect ratios exceeding 30:1 are common. Enclave margins adjacent to trondhjemitic zones in the host are sharp (Fig. 3C) and locally enriched in biotite.
Andesitic enclaves contain the same mineral assemblage as the host granodiorite. Long prismatic crystals of plagioclase form radiating clusters up to 4.0 mm in diameter (Fig. 3D). Inclusions of biotite, actinolite, apatite, and Fe-Ti oxides in plagioclase are common. Large poikilitic grains of K-feldspar up to 4.0 mm in diameter enclose plagioclase, biotite, quartz, titanite, and actinolite. Quartz is also interstitial and titanite is poikilitic. Actinolite (0.01–0.30 mm in length) occurs as small stubby blebs, tabular grains, and needles.
Late-Stage Tonalitic Dikes
Field Observations. Late-stage tonalitic dikes commonly contain mingling zones of variably hybridized enclaves. The best-exposed occurrence is on the summit of Lake Mountain. Sizes of individual enclaves range from <1 cm to >20 cm in longest dimension. Enclaves are fine- to medium-grained and black to dark gray and set in a medium- to coarse-grained, dark-to medium-gray matrix that grades into the pale gray of the tonalitic host (Fig. 4A). Contacts between the matrix and enclaves are commonly diffuse, and veins of host material in the enclaves are rare. Enclaves are typically oblate with tapered ends and aspect ratios <2:1. The enclaves are crudely oriented parallel to the gradational contact between the hybrid matrix and the tonalitic dike (Fig. 4A).
Enclaves are generally isolated, separated from one another by as much as 10–15 cm of matrix. They exhibit a wide range of grain sizes, as does the matrix, thereby making it difficult to discern the outlines of smaller enclaves. In fact, the matrix is littered with small, barely recognizable enclaves, which made sampling for chemical analysis difficult (Fig. 4B).
Petrography. The biotite tonalite consists of plagioclase (An28–37) + quartz + biotite + Fe oxides ± hornblende ± K-feldspar. Accessory phases are apatite, as inclusions in plagioclase, and zircon. Euhedral to subhedral plagioclase (up to 3.0 mm in length) exhibits weak normal zonation. Quartz and minor K-feldspar are interstitial. Minor hornblende occurs as inclusions in biotite.
Microdioritic enclaves consist of plagioclase (An45–49) +hornblende + biotite + Fe-Ti oxides + clinopyroxene + orthopy-roxene ± quartz. Apatite is common, with aspect ratios of 5:1–10:1. Enclaves commonly exhibit a strong foliation defined by the alignment of euhedral to subhedral hornblende and plagioclase. Biotite is poikilitic and quartz is interstitial. Pyroxene occurs as cores in hornblende and as large clots rimmed with hornblende.
Rocks from the dark “hybrid” matrix appear to contain components of both enclave and host, resulting in a strongly bimodal distribution of grain size (Fig. 4C). Strongly zoned plagioclase grains with resorbed rims and cores are common, although unresorbed grains are also present. Pyroxene is generally rimmed with hornblende. Apatite exhibits two habits: long, acicular needles and short, stubby prisms. Relatively siliceous hybrid compositions contain more biotite, plagioclase, and quartz, and less pyroxene and hornblende.
Mafic Enclave Swarms
Micronoritic enclaves from the Grayback Mountain enclave swarm most closely resemble high-Al island arc tholeiites, on the basis of their low K2O and high CaO/Al2O3 and FeO*/MgO (Table 1, Fig. 5). There is little difference between enclave and host compositions, with the exception of slightly higher K2O, Cr, Zr, and Rb, and slightly lower TiO2, Fe (total Fe as FeO), and Ni abundances in the host (Fig. 6). In addition, host material (GM-25-8) adjacent to a large enclave contains higher Na2O, V, Nb, Y, Zr, and Ba, and lower MgO than host material isolated from enclaves (Table 1).
Unlike enclave-host relationships exposed on Grayback Mountain, enclaves in the Lake Mountain enclave swarm are more chemically distinct from their host (Table 1, Fig. 6). Lake Mountain enclaves have lower SiO2, TiO2, Al2O3, Na2O, K2O, P2O5, Zr, Sr, Ba, Rb, and FeO*/MgO, and higher MgO, CaO, Sc, V, Cr, Ni, Y, and CaO/Al2O3 than their host rocks (Fig. 6). Enclave compositions are similar to calc-alkaline basalt and basaltic andesite on the basis of their higher K2O and lower FeO*/MgO, relative to tholeiitic compositions. Enclave and host samples have light REE (LREE)-enriched chondrite-normalized patterns (Fig. 7). The host sample (LM-11) has a straight REE pattern, whereas the enclave (LM-7–4) has higher ΣREE and exhibits a negative Eu anomaly (Eu/Eu* = 0.73). Both enclave and host have similar initial 87Sr/86Sr (0.7042) and ϵNd (+5.5 and +4.8, respectively; Barnes et al., 1995), which are comparable to rocks of the main stage of the Grayback pluton.
Late-Stage Granodioritic Dikes
Basaltic enclaves in the late-stage granodioritic dikes are high-Al and calc-alkaline in composition. They have similar SiO2 contents to micronoritic enclaves in the mafic enclave swarms, but have higher TiO2, Na2O, K2O, Zr, and Rb, and lower FeO, MnO, CaO, Sc, V, and CaO/Al2O3 (Table 1, Fig. 5). One enclave sample is characterized by a linear chondrite-normalized REE pattern that lacks a Eu anomaly, by high ϵNd (+5.5), and by low initial 87Sr/86Sr (0.7027; Barnes et al., 1995). Apparently unmodified host granodiorite has SiO2 >70 wt%, K2O to nearly 4 wt%, and Ba abundances of 800–1342 ppm (Fig. 5). In addition, a granodiorite sample has low heavy REE (HREE) abundances (YbN = 5.5), and a slight negative Eu anomaly (Eu/Eu* = 0.76; Fig. 7). Andesitic enclaves are compositionally intermediate between basaltic enclaves and granodioritic host samples (Table 1); indeed, andesitic samples plot along a straight line between the basaltic and granodioritic end-members for most elements (Fig. 8).
Additional samples from the margin of an andesitic enclave and from dark-colored host granodiorite adjacent to the enclave were analyzed (Table 1). Compared to unmodified granodiorite, the granodiorite adjacent to the enclave is higher in TiO2, FeO, MgO, CaO, Sc, V, Cr, Ni, and Y, and lower in SiO2 and Sr. The sample from the margin of the andesitic enclave is depleted in Na2O and Zr, but enriched in K2O, Y, Ba, and Rb, relative to the enclave core.
Late-Stage Tonalitic Dikes
Tonalitic rocks from late-stage dikes are distinct from the main-stage tonalites. Barnes et al. (1995) characterized the late-stage rocks as high-Al tonalites, whereas main-stage tonalites are low-Al (c.f. Barker, 1979). In addition to its higher Al2O3 contents, the late-stage tonalite is characterized by higher SiO2, FeO*/MgO, and Zr, and chondrite-normalized REE patterns that show greater fractionation of HREE from LREE and positive Eu anomalies.
Microdioritic enclaves in the late-stage tonalitic dikes have similar SiO2 contents to basaltic enclaves in the late-stage granodioritic dikes, but have higher FeO, MgO, CaO, Sc, V, Zn, Ba, Sr, and CaO/Al2O3, and lower TiO2, Al2O3, Na2O, K2O, P2O5, Nb, Zr, and Rb (Table 1, Figs. 5 and 8). Concentrations of most elements in these microdioritic enclaves are similar to or intermediate between micronoritic and microgabbroic enclaves in the mafic enclave swarms, but microdiorite has higher TiO2, Na2O, K2O, Ba, and Rb, and lower CaO, Ni, and Nb. A microdioritic enclave from a late-stage tonalitic dike has a LREE-enriched chondrite-normalized pattern, and lacks a Eu anomaly (Eu/Eu* = 0.99; Fig. 7). This same enclave yielded an initial 87Sr/86Sr value similar to those of a microgabbroic enclave and its host from the mafic enclave swarm on Lake Mountain (0.7042), but has a much higher ϵNd (+10.3; Barnes et al., 1995).
The dark matrix that envelops enclaves in the late-stage tonalitic dikes has a wide range of compositions, from ∼51 to 63 wt% SiO2 (Table 1). There is considerable overlap of matrix compositions and those of enclave and tonalitic samples, which reflects the difficulty in distinguishing the different lithologies (e.g., micro-enclaves in similarly dark matrix rock; Fig. 4B). Matrix compositions plot along straight lines between enclave and tonalitic end-members for most elements, with the marked exception of TiO2 and P2O5, which are generally higher in matrix compositions (Fig. 8).
INTERACTIONS BETWEEN MAGMAS IN THE MINGLING ZONES
Enclaves in the Grayback pluton mingling zones possess characteristics that indicate they were liquid or only partially crystallized when brought into contact with their host magmas. For example, their fine grain size, evidence for plastic deformation, and the acicular nature of apatite crystals preclude an autolithic (e.g., early cumulate) origin and suggest that the enclave magmas were undercooled by contact with the host magma. Below, we discuss the nature of the interactions between magmas in the various magma-mingling zones exposed in the Grayback pluton and the implications of these mingling zones for magma mixing processes in the petrogenesis of the Grayback system.
Mafic Enclave Swarms
The mafic enclave swarms on Grayback and Lake Mountains represent relatively early mingling episodes in the Grayback pluton. Moreover, they represent mingling between two or more mafic and/or intermediate magmas, as opposed to mingling between mafic and felsic compositions, as in the late-stage dikes. The mingling zone on Grayback Mountain is unusual in that there is little compositional difference between enclave and host rocks.
One of the puzzling features of the mafic enclave swarms is the aggregates of coarse-grained poikilitic hornblende and plagioclase within finer-grained pyroxene-rich enclaves (Fig. 2C and F). We interpret these as relict veins of host material caught between enclaves during emplacement. The residual liquid in the veins was either squeezed out or incorporated into the enclave magma during localized contraction. In this scenario, the coarser hornblende and plagioclase would be xenocrystic to the enclaves, akin to the quartz ocelli frequently described in mafic enclaves in quartz-rich host rocks (e.g., Vernon, 1983). The preservation of quartz ocelli in mafic enclaves is typically cited as evidence that the enclaves represent hybrid compositions, which is taken to imply that magma mixing occurred prior to commingling and enclave formation.
Some veins of the host rock contain fragments of enclave material that were apparently separated from the enclave during vein formation. Thus, field evidence indicates that hybridization of the magmas occurred in two ways: (1) incorporation of host magma into enclave magma by collapse of veins, and (2) fragmentation of enclave material into the host magma. This mechanism of hybridization might explain why there is less compositional distinction between enclave and host rocks on Grayback Mountain, where the degree of localized flattening and thus incorporation of host magma into the enclaves, was apparently greater than on Lake Mountain.
It could be argued that enclave and host rocks in the Lake Mountain mingling zone represent the mafic and more evolved components, respectively, of a single magma undergoing fractional crystallization. However, the host has higher TiO2, Al2O3, P2O5, Sr, and Zr, and lower ΣREE than the enclaves, which is not the expected result of fractional crystallization of the observed phases. Therefore, we conclude that enclave and host compositions in the Lake Mountain mafic enclave swarm represent magmas from two calc-alkaline sources.
Although their outcrops are similar in appearance, it is clear that rocks comprising the mafic enclave swarm on Grayback Mountain are very different in composition from those exposed on Lake Mountain (Fig. 5). In contrast to Grayback Mountain samples, those from the Lake Mountain mafic enclave swarm are similar in composition to rocks comprising the noncumulate main stage of the Grayback pluton.
Late-Stage Granodioritic Dikes
The presence of both basaltic and andesitic enclaves in the late-stage granodioritic dikes suggests either that the enclave magmas were derived from two distinct sources or that the andesitic enclaves represent a hybridized mixture of basaltic and granodioritic magmas. We tested the mixing scenario, using element ratios according to the method of Langmuir et al. (1978). If a suite of rocks is related by binary mixing, then the samples should define a hyperbolic curve in ratio-ratio plots, define straight lines in companion plots, and remain in the same relative order in both ratio-ratio and companion plots. This approach, using Sc, Cr, K, and Ba abundances (Fig. 9A), produced consistent results that suggest binary mixing of basaltic and granodioritic end-members formed the andesitic magma. Major element least-squares mass balance calculations indicate that a mixture of 58% basalt and 42% granodiorite would produce a hybrid composition similar to that of the andesite (Table 2). Mixing calculations using the REE abundances of the same basaltic and granodioritic end-members produced hybrid compositions with chondrite-normalized REE patterns similar to the andesitic enclaves (Fig. 9B).
The presence of basaltic and hybrid andesitic enclaves in the late-stage granodioritic dikes is suggestive of a multi-stage mixing/mingling process (Koyaguchi, 1986). In the first stage, basaltic magma intruded the silicic magma, producing a layer of hybrid andesite at their interface. In the second stage, dike formation and magma ascent disrupted the basaltic and andesitic layers, resulting in their incorporation as enclaves in the silicic host magma. A post-emplacement stage involved chemical exchange between the magmas. The abundance of biotite and actinolite in the basaltic enclaves and the biotite- and K-feldspar–poor trondhjemitic halo that envelops several enclaves is suggestive of migration of H2O and K+ from the granodioritic to basaltic magmas (Watson and Jurewicz, 1984).
Major element mass balance calculations indicate that the dark-colored granodioritic host adjacent to some andesitic enclaves represents a mixture of ∼18% andesitic and 82% granodioritic magmas (Table 2). The andesitic magma would have had a lower temperature and a resultant higher viscosity, relative to the basaltic magma, which would have prevented complete mechanical mixing of the andesitic and granodioritic magmas. We suggest instead that the dark-colored granodiorite formed by late-stage diffusive exchange between andesitic and granodioritic magmas during and/or after emplacement of the dike. This post-emplacement exchange was more pronounced than that between basaltic enclaves and host, because the basaltic magma was supercooled to a greater degree, which would have hindered chemical diffusion.
Late-Stage Tonalitic Dikes
As in the case of the late-stage granodioritic dikes, ratio-ratio and companion diagrams using enclave, host, and intermediate samples from late-stage tonalitic dikes (in this case, using Sc, Cr, Zr, and Ba abundances) suggest that magma mixing played a dominant role in the formation of intermediate compositions (Fig. 10). Minor deviations from linearity in the companion diagrams may result from fractional crystallization of augite in the enclave magma (see also Smith et al., 1990; Wiebe et al., 2002), resulting in enclave compositions with lower Sc/Zr values than those predicted for mixing alone (Fig. 10).
Major element mass-balance calculations suggest that binary mixing between dioritic and tonalitic magmas can produce almost the entire range of intermediate compositions observed in the late-stage tonalitic dikes. Most calculations yielded acceptable results (sum of the squares of residuals [Σr 2]) less than unity; Table 2). However, high negative Al2O3 residuals and high positive MgO and CaO residuals in some calculations (Table 2) suggest accumulation of plagioclase in, and possibly removal of augite from, the hybrid magma.
Anomalously high TiO2 and P2O5 abundances in hybrid rocks have been noted in mingling zones elsewhere (Wood, 1978; Vogel, 1982; Marshall and Sparks, 1984; D'Lemos, 1996; Kuritani et al., 2003) and have been attributed to pre-mixing fractional crystallization in the mafic magma, mixing between compositionally heterogeneous magmas, post-mixing TiO2 and P2O5 enrichment in the hybrid, and anomalous diffusion of Ti and P complexes. Enrichment of TiO2 in the matrix compositions is inconsistent with the removal of augite from the hybrid magma (see above), and mixing between heterogeneous mag-mas is not suggested by the uniform TiO2 and P2O5 in enclave and host samples. Alternatively, nonlinear relationships might reflect differences in diffusion coefficients between the elements of interest. However, differences in diffusion coefficients are unlikely to affect element correlations at the scale of hand samples used in this study (Perugini and Poli, 2004). In light of the fractional crystallization scenario described above, we suggest that the anomalously high TiO2 and P2O5 contents in the matrix samples (Fig. 8) are a reflection of pre-mixing removal of augite (and apatite) from the enclave magma.
Discussion of Rheologic Models
Introduction. Petrographic observations and whole-rock chemical compositions can be used to estimate the rheologic conditions under which magma mixing and mingling are possible (Sparks and Marshall, 1986; Frost and Mahood, 1987). Complete mixing between two magmas of contrasting compositions requires the equilibrium temperature of the mixture to be greater than the solidus temperature of the mafic end-member. The equilibrium temperature depends on the initial temperatures and mass fractions of the end-member magmas, the heat capacities of liquid and solid components, and the heats of fusion of the solids (Frost and Mahood, 1987).
Viscosity (η)-composition relationships from the mingling zones in the Grayback pluton were determined using the algorithms compiled by Frost and Lindsay (1988). Resultant viscosity curves are shown in Figures 11–14, with respect to the mass fraction of the mafic end member in each mixture (XM). Inter-sections of viscosity curves of coexisting magmas are circled, and represent the conditions at which both magmas possess equal viscosities ( , the “equilibrium” viscosity), and are thus most likely to mix. However, it should be noted that hybridization can occur between magmas with viscosity differences as great as 102 poise (Frost and Mahood, 1987) and at XM greater than that at (Sparks and Marshall, 1986). Terminations of the mafic end-member curves represent the η and XM values where the mafic magma is >60% crystallized, and thus essentially acts as a solid (Frost and Mahood, 1987).
Lake Mountain Mafic Enclave Swarm. One of the puzzling aspects of the mafic enclave swarms is the preservation of enclave and host, despite their compositional similarities. If these rocks are reflective of magma mingling in the main stage of the pluton, then it can be assumed that main-stage mixing occurred earlier in the Grayback system and that conditions changed enough to prohibit thorough mixing of these magmas within the enclave swarm. The rheological properties of enclave and host compositions were investigated in an effort to determine why the magmas did not mechanically mix within the mafic enclave swarm at the present level of exposure.
The hornblende-rich nature of enclaves and host rocks in the Lake Mountain enclave swarm suggests that both enclave and host magmas had initial H2O contents >3.0 wt% (e.g., Sisson and Grove, 1993). For an enclave magma temperature of 1050 °C, mixing to produce an intersection of mafic and felsic curves at requires a host magma temperature <950 °C. Temperatures >950 °C result in enclave magma with a lower viscosity than the host magma at all values of XM. A relatively small decrease in host magma temperature (50 °C) results in a large increase in XM (▵XM > 0.2) required for mixing at (Fig. 11). For mixing at equal viscosities, an increase in enclave magma H2O content results in a slight increase in and a slight decrease in XM(Fig. 11), whereas increasing the host magma H2O content results in slightly decreasing and increasing XM.
Complete mixing between the enclave and host magmas would have required at least 15% of the enclave end-member. That complete mixing did not occur within the mafic enclave swarm may reflect a mass decrease in the enclave end member to below 15%, or a shift of to higher values of XM as a result of lower magma temperatures or loss of H2O from the enclave magma (Fig. 11).
Late-Stage Granodioritic Dikes.Figure 12 shows viscosity-composition curves that describe mixing between basaltic and granodioritic magmas from the late-stage granodioritic dikes. The presence of plagioclase phenocrysts and the lack of early hornblende suggest that the basaltic magma had less than ∼3.0 wt% H2O (e.g., Eggler, 1972). Rare hornblende inclusions in plagioclase and the lack of evidence for vapor saturation constrain the H2O content in the granodioritic magma to between ∼4.0 and 6.5 wt% (Naney, 1983). Mixing of anhydrous basaltic magma with granodioritic magma having 4.0 wt% H2O would result in a relatively narrow range of hybrid compositions (∼53–57 wt% SiO2), requiring XM >0.7 (Fig. 12A). In contrast, mixing involving hydrous basaltic magma (2.0 wt% H2O) results in a wider spectrum of hybrid compositions (∼56–66 wt% SiO2) requiring proportionally less mafic magma (XM = 0.25–0.73; Fig. 12B–D), depending on the initial temperature and H2O content of the host magma. The equilibrium viscosity decreases from 105 to 104 poise with increasing H2O content in the host magma (Fig. 12B–D). At XM = 0.58, the proportion of enclave magma estimated by major and trace element mass balance calculations (see discussion above), mixing at requires host magma temperatures of 775–800 °C for initial H2O contents of 4–6 wt%.
Additional calculations were performed to evaluate the conditions and extent of mixing between the hybrid andesitic and host granodioritic magmas, the results of which are shown in Figure 13. The initial temperature and H2O content of the andesitic end-member were taken from results of the basalt-granodiorite mixing calculations (see above). Intersections between curves representing the andesitic and granodioritic end-members ( ) occur at higher viscosities (104.6–105.8 poise; Fig. 13) than those between basaltic and granodioritic magmas (<105 poise; Fig. 12). In addition, mixing between andesitic and granodioritic magmas at could only occur at XM >0.60, which is significantly greater than the XM∼ 0.18 predicted by major element mass balance. These results indicate that conditions suitable for complete binary mixing of andesitic and granodioritic magmas in the late-stage granodioritic dikes were never achieved and that the dark-colored granodiorite adjacent to the andesitic enclaves was instead produced by diffusion of chemical components across the enclave-host contact.
Late-Stage Tonalitic Dikes. Mixing results for the late-stage tonalitic dikes are more difficult to assess, because the hybrid compositions span a wide range in compositions. The hornblende-rich nature of the microdioritic enclaves suggests relatively high H2O contents of 3–6 wt% (Sisson and Grove, 1993); the low An content of plagioclase suggests, however, that the H2O content was at the low end of the range (Sisson and Grove, 1993). The lack of orthopyroxene and the abundance of early biotite in the tonalitic rocks suggest that the felsic magma was H2O-rich, but not H2O-saturated (5–7 wt% H2O; Scaillet and Evans, 1999). Calculations were performed using a range of host magma H2O contents (5.0–7.0 wt%) and initial temperatures (700–950 °C) and using an enclave magma H2O content of 4.0 wt% (Fig. 14). An increase in host magma H2O content from 5.0 to 7.0 wt% causes a decrease in from 104.2 to 103.5 poise; however, an increase in the enclave magma H2O content from 4.0 to 6.0 wt% over the same range of host magma H2O contents (5.0–7.0 wt%) and initial temperatures (750–900 °C) does not affect (104.3–103.5 poise). Ideal mixing of enclave and host magmas at cannot account for the range in hybrid compositions observed in the late-stage tonalitic dikes, particularly those having XM >0.7 (SiO2 <55 wt%), which is probably a result of accumulation (see above).
The coarser grain size of the microdioritic enclaves, compared to the basaltic enclaves in the granodioritic dikes, suggests that the dioritic magma did not experience the same degree of undercooling or that it was largely crystalline when it came into contact with the tonalitic magma. The former suggests a smaller temperature contrast between mafic and felsic magmas, whereas the latter might indicate a high contrast in initial viscosities. A decrease in the temperature contrast results in mixing at progressively lower XM (Fig. 14). However, a decrease in the initial viscosity contrast between the two magmas results in ideal mixing at slightly greater values for XM, but still well below the values suggested by mass-balance calculations (Fig. 14). The bimodal grain size of the hybrid rocks is characteristic of mixing between crystal-rich mushes (e.g., D'Lemos, 1996); therefore, it is unlikely that the hybrid compositions were produced by mixing of liquid magmas having equal viscosities (see Fig. 14).
Barnes et al. (1995) documented a great number of mafic rock types in the Grayback pluton. The lithologic heterogeneity observed, even at the hand sample scale, attests to the presence of coexisting magmas throughout main-stage magmatism. The main stage of the pluton comprises gabbro, norite, quartz diorite, and lesser amounts of pyroxenite and tonalite. Coeval calc-alkaline and evolved tholeiitic basaltic magmas point to a heterogeneous mantle source. Barnes et al. (1995) demonstrated that fractional crystallization of plagioclase + magnetite + olivine ± clinopyroxene ± orthopyroxene ± hornblende ± apatite from calc-alkaline basaltic magma of the main stage could produce felsic compositions similar to those observed in the late-stage granodioritic and granitic dikes. Such a process could explain the abundance of cumulate pyroxenites and melagabbros with Mg/(Mg + Fe) >0.58 (see Barnes et al., 1995).
Fractional crystallization resulted in an initial increase in P2O5, followed by an abrupt decrease at the onset of apatite crystallization (Fig. 5C). In addition, TiO2 (Fig. 5A), Al2O3, and Sr abundances steadily decreased owing to the removal of Fe-Ti oxides, clinopyroxene, and plagioclase. Few main-stage samples follow the fractional crystallization trend observed in Figure 5, but instead exhibit increasing TiO2, Al2O3, P2O5, and Sr with decreasing Mg/(Mg + Fe). Likewise, few samples follow the mixing trends exhibited by enclave-host pairs in the late-stage granodioritic and tonalitic dikes. It is difficult to discern a clear liquid line of descent with respect to most elements, but most of the noncumulate main-stage samples plot near the field of enclave and host compositions represented by the Lake Mountain mafic enclave swarm (Fig. 5). The compositional similarities between the Lake Mountain mafic enclave swarm and Grayback main-stage samples suggest that mixing between calc-alkaline basaltic and andesitic magmas occurred throughout much of the Grayback history, probably at a greater depth than represented by the current level of exposure. The enclave and host magmas observed in the Lake Mountain mafic enclave swarm were prevented from mixing, possibly as a result of devolatilization or a temperature decrease (and hence, a higher crystallinity). A similar scenario could be envisioned for tholeiitic compositions in the mafic enclave swarm on Grayback Mountain, which tapped a distinctly different mantle source (Barnes et al., 1995).
Some basaltic magma evolved to granitic compositions through fractional crystallization. Minor assimilation of meta-sedimentary rocks gave rise to the elevated 87Sr/86Sr and δ18O contents of the granites (Barnes et al., 1992a). Continued influx of basaltic magma and subsequent mixing with the silicic magma produced hybrid andesite, presumably at a deeper level in the Grayback system. Intrusion of the granitic magma to the present level of exposure entrained small volumes of basaltic and andesitic magmas that quickly solidified in their new environment, producing the mafic and intermediate enclaves. Although the late-stage granodioritic dikes lack evidence for extensive mixing between mafic and felsic magmas in situ, the presence of hybrid andesitic enclaves indicates that such mixing did occur in the Grayback system, requiring close to 60% of the basaltic end-member for complete mixing to take place. That the hybrid andesite compositions are unlike any main-stage samples suggests either that the andesite produced in this manner was volumetrically minor or that large volumes did not ascend to the present level of exposure.
Heat from mafic magma partially melted mafic meta-volcanic rocks residing deep in the Klamath crust, generating small volumes of high-Al tonalitic magma. Major and trace element signatures of the tonalitic rocks (high Al2O3 and Na2O, and low K2O, Y, and HREE; Drummond and Defant, 1990) suggest that partial melting occurred in equilibrium with a garnet-and hornblende-bearing residual assemblage at pressures exceeding 10 kb (Wolf and Wyllie, 1994; Rapp and Watson, 1995). Injection of calc-alkaline basaltic magma into the high-Al tonalitic magma resulted in limited mixing and the magma mingling relationships observed in the late-stage tonalitic dikes. Results of this study indicate that this basalt-tonalite mixing did not play a significant role in the compositional variation of the Grayback main stage.
Other plutons in the Wooley Creek belt experienced open-system behavior similar to the Grayback pluton. In the Wooley Creek batholith, basalt mixed with fractionated dacite in the upper part of the magma chamber to produce hybrid andesite (Barnes et al., 1986, 1987). Open-system behavior in the Ashland pluton (Gribble et al., 1990) led to mixing between alkalic basalt and tonalite and resultant quartz monzodioritic magma, which evolved by coupled fractionation and assimilation to more felsic compositions. Schmidt (1994) determined that mixing between andesitic magma and anatectic granodiorite and granite was important during later stages of the composite English Peak pluton. In addition, mafic enclaves throughout the Russian Peak pluton indicate not only the presence of mafic magma in the system but, coupled with ranges in Sr and O isotope abundances (Cotkin and Medaris, 1993), hint at the possible role of magma mixing in its history. The Grayback pluton, however, is unusual amongst these plutons in that mafic magmas from multiple mantle sources were prevalent throughout its history. It is likely that ascent of mantle-derived magmas into the Grayback system was facilitated by extension in the back-arc of the contemporaneous Rogue–Chetco arc complex (Harper et al., 1994). This setting, directly inboard of the arc, may have experienced greater extension than further inboard or to the south, positions currently occupied by the other Wooley Creek belt plutons.
A diverse range of magma-mingling zones in the Grayback pluton provides valuable information about the physicochemical interactions of magmas, and documents the coexistence of several mantle-derived magmas throughout the evolution of the Grayback system. Comparisons between rock compositions preserved in the magma-mingling zones and those from the main stage of the pluton suggest that most of the compositional variation within the Grayback pluton resulted from mixing of calc-alkaline basaltic and andesitic magmas. Late-stage activity resulted in the emplacement of enclave-bearing granodioritic and tonalitic dikes, which exhibit different styles of interaction between mafic and felsic end-members.
Grayback magmatism was coeval with other plutons of the Wooley Creek belt, which all show evidence for a significant role for mafic magmas in their evolution. Increased interaction of mantle-derived magmas and crustal rocks in Wooley Creek belt plutonic systems, compared with earlier episodes of Jurassic magmatism, was facilitated by extension in the back-arc region of a contemporaneous volcanic arc. Emplacement of these mafic magmas into already-warmed crust allowed for assimilation of crustal rocks and/or mixing with anatectic melts.
We are pleased to have the opportunity to contribute to this volume honoring Porter Irwin's pioneering and influential work in the Klamath Mountains. We thank Bennetta Schmidt and Melanie Barnes for help in the field, and Bill Shannon and Melanie for their invaluable assistance in the laboratory. We appreciate the thorough and constructive reviews of Nancy McMillan and Daniel Lux. Funding was provided by a Penrose Bequest of the Geological Society of America (KJ) and National Science Foundation grants EAR-8720141 and EAR-9117103 (CGB).
Figures & Tables
Geological Studies in the Klamath Mountains Province, California and Oregon: A volume in honor of William P. Irwin
- back-arc basins
- extension tectonics
- igneous rocks
- Klamath Mountains
- mafic composition
- magma chambers
- major elements
- plutonic rocks
- trace elements
- United States
- volcanic rocks
- Grayback Pluton