Click for next page ( 73


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 72
Oxygen and Hydrogen Isotope Constraints on the Deep Circulation of Surface Waters into Zones of Hydrothermal Metamorphism and Melting HUGH P. TAYLOR, JR. California Institute of Technology INTRODUCTION The purpose of this paper is to marshal the evidence and try to build a case that (1) shallow (1 to 7 km) circu- lation of surface waters in the Earth's crust is an extremely widespread and common phenomenon in areas of igneous activity and (2) deep (10 to 15 km) circulation of surface waters can occur in certain favorable geological situations, particularly in rift zones and areas of extensional tecton- ics. It is shown that very large amounts of water may interact with the rocks in such zones and that this can take place at temperatures high enough for melting and meta- morphism to occur. Oxygen and hydrogen isotope studies have proven to be very useful in establishing the charac- teristics of such deeply circulating hydrothermal systems and in determining the origins of the aqueous fluids in- volved in producing granitic and rhyolitic magmas in such environments. This is mainly because oxygen-18 and deuterium are constituents of the H2O molecule itself, and thus stable isotope signatures are by far the best way to characterize hydrothermal fluids of different origins. These conclusions are most clear-cut when low-'8O, low-D mete- oric waters are involved in the isotopic exchange proc- esses, but ocean waters, sedimentary formation waters, metamorphic dehydration waters, and magmatic waters can also be distinguished from each other in favorable circumstances. 72 Because of aqueous-fluid interactions, rocks that ulti- mately undergo partial melting may exhibit isotopic signa- tures considerably different from those that they started with. Leaving aside effects ascribable to the intrinsic INTO and ED values of these different kinds of waters, it is proposed in this paper that stable isotope studies may be used to identify three broad classes of hydrothermal sys- tems, based mainly on the water/rock ratio (w/r), the temperature (T), and the length of time (hi chat fl,~iA_r^~lr interaction proceeds. ~^A^_ \ ~- ~-CAM A V-~ type 1. L;p~zonal systems with a wide variation in whole-rock 6~80 and extreme i8O/~60 disequilibrium among coexisting minerals (e.g., quartz and feldspar); these sys- tems typically have T = 200 to 600C, t < 106 yr, and form under hydrostatic pressure conditions. Type 2. Deeper-seated and/or longer-lived systems, also with a wide spectrum of whole-rock 6~80 but with equilibrated ~80/~60 ratios among coexisting minerals (T = 400 to 700C, t > 106 yr, and pressures transitional from hydrostatic to lithostatic. Type 3. Thoroughly homogenized and equilibrated systems with relatively uniform 6~80 in all lithologies; these probably form at large w/r, T = 500 to 800C, and t > 5 x 106 yr. The most common of these systems, and the type that is

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS most easily recognized through field and laboratory stud- ies, is Type 1. However, these three categories are not mutually exclusive; prior to melting, many Type 3 systems at an earlier stage may have been subjected to Type 1 or Type 2 conditions. The Type 3 systems very likely pass through an early hydrostatic phase, but at their peak tem- peratures they almost certainly end up at lithostatic pres- sures. The best available example of a Type 3 system is probably the Trois Seigneurs area in the Pyrenees; it is described and discussed in detail by Wickham and Taylor (chapter 6, this volume). Also note that even though the whole-rock 6~80 values (and 87Sr/86Sr values?) may be radically changed and homogenized in the Type 3 sys- tems, the whole-rock chemical compositions and other isotopic parameters (e.g., eNd) in many cases may be only slightly modified. Thus, under certain conditions, fluid-rock interaction can dramatically increase the i80/~60 heterogeneity of initially uniform terrane (e.g., a section of volcanic rocks in a Type 1 system), but on the other hand it also can smooth out original heterogeneities (e.g., a Type 3 re- gional metamorphic system). In either case most natural aqueous fluids usually produce an overall RIO depletion of the rocks (because equilibrium BIRO mineral-fluid values are smaller at high T than at lower T). However, at least for the relatively lower temperature Type 1 systems, fluids With 6180 values higher than seawater (i.e., >0) can produce ~80 enrichments in the rocks and associated melts (e.g., oceanic plagiogranites). OXYGEN ISOTOPE KINETICS IN THE QUARTZ FELDSPAR SYSTEM Figure 5.1 shows the characteristic patterns of 6~80 feldspar versus 6~80 quartz in some hydrothermally al- tered granitic rocks from British Columbia (mostly Type 1 systems). The 180/~60 envelopes displayed in Figure 5.1 cut across the 45 equilibrium lines at a steep angle as a result of the much faster 180/~60 exchange rate of feldspar relative to quartz. Analogous effects are observed for feldspar and pyroxene in hydrothermally altered gabbros (Taylor and Forester, 1979; Gregory and Taylor, 1981~. The kinetics of these steep trajectories have recently been quantitatively studied by Cnss et al. (1987) and Gregory et al. (1988~. They have shown that if a low-~8O aqueous fluid such as meteoric H2O or ocean H2O is involved in the exchange process, the slopes of the mineral-pair data points on a graph of BIRO feldspar versus BIRO quartz essentially constitute an "oxygen isotope clock," as indicated sche- matically in Figures 5.2 and 5.3. Each of the disequilibrium arrays on such diagrams can be regarded as an isochron in which the time increases as the slopes of the isochrons become less steep and approach 73 1 1 1 1 '' ' ~ SQUAMISH AREA Girl i,,/ THUYA BATHOLITH RAFT BATHOLITH .~ +8 it, +( in c, +4 J o a) +2 60 o -2 /~' ~\~ X,// TERTIARY ~ I - PORPHYRIES~ _; :: 1 1 1 ~1 1 1 1 1 O +2 +4 +6 +S +10 +12 +14 8180 QUARTZ 0/) FIGURE 5.1 A plot of 6~80 feldspar versus 6~80 quartz for various granitic plutons from southern British Columbia. Coast Batholith and Squamish area represent portions of the western part of the Coast Plutonic Complex. In most cases marked isotopic disequilibrium is observed among the coexisting miner- als of these hydrothermally altered granitic rocks. The steep trend lines can be extrapolated back to the 45 equilibrium line to obtain the 6~80 values of the original magmas or the unaltered rocks (from Magantz and Taylor, 1986~. 45 (Cries et al., 1987~. These "times" represent the dura- tion (e.g., in years) of a particular hydrothermal exchange event. These models are not yet~perfectly quantified, but for moderate hydrothermal temperatures (300 to 500C) the actual times indicated in Figure 5.2 are closely con- strained by theoretical studies of the "lifetimes" of hy- drothermal convective systems (e.g., Norton and Knight, 1977), as well as by available laboratory hydrothermal diffusion measurements (Giletti et al., 1978; Yund and Anderson, 1978; Giletti and Yund, 1984; Giletti, 1986; also see review by Cole and Ohmoto, 1986~. Although the relative times in Figure 5.2 are well estab- lished, where they are normalized to a value of unity for the rate-constant of the slow-exchanging mineral (Cries et al., 1987), the actual times indicated on the "oxygen iso- tope clock" in Figure 5.3 are obviously dependent on several variables such as temperature, permeability, grain size, etc. However, these effects can all be treated together and described by a single phenomenological parameter, namely the rate constant for hydrothermal isotopic exchange be

OCR for page 72
74 tween the slow-exchanging mineral and the aqueous solu- tion (Gregory et al., 1988~. For example, if the rate con- stant for mineral 2 (k2) is assumed to be 10-~4 S-], which is a very plausible value for quartz, then the values of t on the isochrons in Figure 5.2 (or on the inside of the clock rim in Figure 5.3) should all be multiplied by 10~4 S (=3.17 x 106 yr) in order to get the actual "lifetimes" of the hydrothermal systems. A near-vertical slope in Figure 5.2 is characteristic of epizonal hydrothermal systems, which we know typically last for at most only a few hundred thousand years (Norton and Knight, 1977; Norton and Taylor, 1979~. The time constraints described above thus require that the mineral- pair 6~8O values of all of these kinds of systems must attain near-equilibnum slopes (i.e., 45) on a time scale of about 4 to 6 million years (m.y.) (Gregory et al., 1988~. At higher temperatures (>600C) the times required for equili- bration and attainment of a 45 slope will be much smaller, almost certainly less than 1 to 2 m.y. In other words, if fluid-rock interactions persist for only about 50,000 to 300,000 yr (the characteristic dura 6 _ 4 _ ~1 2 _ o (I) fluid-dominated system Ad,, ......... ~ "closed" system '~ - ~ /~- t=2.0 1=1.0 l l l l l ' ' l l -2 -1 0 1 2 3 4 5 6 ~2 FIGURE 5.2 Graph of 6~ versus 62, showing sets of isochrons at various normalized times, t = 0.25, 0.5, 1.0, 2.0, and or, for two coexisting minerals that undergo isotopic exchange with water in (1) a fluid dominated system, (2) a "closed" system, and (3) a true open system (modified after Criss et al., 1987; Gregory et al., 1988~. The triangle indicates the initial values of 6: and 62. These calculated curves are taken from Figures 6, 7, and 8, and Eq. (12) of Criss et al. (1987), and they all include a uniform ratio of rate constants k~/k2 = 5. Note that the isochrons for the three different cases are nearly coincident. In fact, for t = 0.25 and ~ they are essentially identical. The times, t, are normalized to a unit value for k2. Thus, if in an actual case, k2 = 10-~4 Sol, (a plausible value for quartz), the values of t on the isochrons would be a factor of 10~4 longer than indicated. HUGH P. TAYLOR, JR. l2 ~ GIN ISo~ ~ t9 ,~ 3~ ~ -~-~ /~y . , FIGURE 5.3 A schematic "oxygen isotope clock" showing how the slopes of the disequilibrium isochrons (Cries et al., 1987; Gregory et al., 1988) change with time on a plot of the bi8O of a slow-exchanging mineral like quartz (on the abscissa) versus the SILO of a fast-exchanging mineral like feldspar (on the ordinate). As an array of data points (e.g., as in Figure 5.1) change in slope from near vertical to about 45, with increasing time the array will successively sweep past the positions occupied by the vari- ous isochrons during the "lifetime" of a typical epizonal hydro- thermal system (one having temperatures of about 250 to 500C). The times in such a case might change from about 300,000 yr to about 6 m.y., as shown; however, these times are only approxi- mate, as they are arbitrarily predicated on an assignment of 300,000 yr as the "lifetime" of the hydrothermal system appro- priate for the 0.09 isochron. The numbers from 0.09 to 1.84 along the inside of the "clock" rim are values of the dimension- less quantity k2 (see Gregory et al., 1988), where t is the length of time that the hydrothermal system is operating and k2 is the kinetic exchange rate constant for the slow-exchanging mineral (these may equally well be considered to be normalized values of the "lifetime" of the hydrothermal system; see Cnss et al., 1987~. lion of a modest-sized, epizonal hydrothermal system), we obtain a near-vertical slope in Figures 5.2 and 5.3 (i.e., the clock's hand points downward toward approximately 6:15~. This would constitute a Type 1 rock-fluid system as de- fined above and illustrated in Figure 5.1. If the fluid-rock interactions persist at similar temperatures for more than S or 6 m.y., this is sufficient time for isotopic equilibrium among the minerals to be essentially established, resulting in an approximately 45 slope (i.e., the clock's hand will point toward 7:15 or 7:30~. This would be termed a Type 2 rock-fluid system. If the fluid-rock interaction continues for even longer times or at higher temperatures and higher fluid-rock ratios, we might envision a Type 3 system in

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS which not only are the mineral 6~80 values equilibrated but the whole-rock samples themselves all attain relatively homogeneous BIRO values. In an idealized Type 3 situ- ation the 45 array of data points could conceivably shrink down to a single point. MAJOR FOSSIL METEORIC HYDROTHERMAL SYSTEMS OF WESTERN NORTH AMERICA Cordilleran Batholiths of Southern British Columbia Magaritz and Taylor (1986) measured hydrogen and oxygen isotope ratios of 500 samples, mainly from granitic plutons (Figures 5.1 and 5.4) along a 700-km E-W traverse across the "accreted terranes" of southern British Colum- bia (latitudes 49 to 52N). Despite the geological com- plexity and range of intrusive ages (Late Triassic to Terti- ary), and even though there are "steps" in the isotopic values at some geologic boundaries (e.g., across the Strait of Georgia between the mainland and Vancouver Island), a clear-cut isotopic pattern was found: the l~o/~60 and DM ratios of the waters involved in hydrothermal interactions ~ i: - -of An 75 with the granitic rocks show a regular eastward trend of depletion in D and RIO (Figure 5.5), indicating clearly that these waters were surface (meteoric) in origin (although seawater may have been important in the extreme western- most terrane in Vancouver Island). Two groups of samples are unique in their high ED values (Figure 5.5~. The first group is represented by two geographically isolated batholiths (Guichon and Thuya) that were not affected by the Tertiary meteoric hydrother- mal systems and that have preserved a set of Early Jurassic to Triassic K/Ar ages. The second group is represented by the Jurassic plutons of Vancouver Island; there the hy- drothermal fluids were both D-rich and i8O-rich (DO > 0), as evidenced by the fact that feldspars in the altered gran- ites are enriched in i8O relative to coexisting quartz (see Figure 5.1~. Both "anomalies" can be explained if these terranes were located closer to the equator and/or in a maritime environment at the time of intrusive and hy- drothermal activity, in agreement with available paleo- magnetic data that indicate a considerable northward drift of these terranes prior to their accretion to the western margin of North America (see Magaritz and Taylor, 1986~. rem& ~m 110' Cr, ~ 1 1 1.' ::KO ~ CR -~.'Xt-~~-^IPi_'-',-LUp~I`~/~,~-,-~; ~ ~i- ~ ~ 1 ll sue FIGURE 5.4 Generalized geologic map of the area studied by Magaritz and Taylor (1986) showing the major granitic batho- liths, the various tectonostratigraphic terrane boundaries, and the locations of the samples analyzed for l8O/l60 and D/H. WR, so. O SO 100 lSO 200 Ktit 1 ' , ' ' 1 0 50 100 Btit~S Wrangellia; NK, Nooksack; QN, Quesnel; MT, Methow- Tyaughton; P. Pacific rim; C, Crescent; KO, Kootenay; SK, Skagit; HZ, Hozameen; TA, Tracy Arm; S. Stikine; CC, Cache Creek; CR, Craton.

OCR for page 72
76 -40 = -60 I -80 Cot to z -100 J he to ~ -120 to a 3-440 - C~ 60 FIGURE 5.5 Plot of ED of all samples-~60 _ studied by Magaritz and Taylor (1986)_ versus distance eastward from the Pacific_ Coast of Vancouver Island.W Excluding these anomalous areas, two distinct ages of meteoric hydrothermal activity can be identified along the 700-km traverse, namely Cretaceous in the west and early to mid-Tertiary in the east. The isotopic trends in the rocks are similar to the present-day patterns of meteoric waters in the region, with one primary difference: the paleowaters are enriched in D by about 20 per mil, com- patible with a northward translation of these terranes, a climatic change, or both. The similarities of the patterns suggest a topography similar to that of the present day (a mountain chain along the coast) during the early Tertiary. Although the freshest, least altered samples in each terrane in British Columbia typically have ED values close to or within the presently accepted "primary magmatic" ED range of-65 to -85, the subset of heavily altered granitic rocks exhibits steadily decreasing ED values from west to east (Figure 5.51. If we confine the discussion just to samples that show hydrothermal }8o/~6O effects (i.e., those that must have experienced very high water-rock ratios), the measured ED values change systematically, as shown in Table 5.1. The west-to-east ED changes in Table 5.1 are similar to what is observed in present-day meteoric waters in British Columbia (see Figure 8 in Magaritz and Taylor, 1986~. Neither of the two older batholiths that have preserved early Mesozoic K/Ar ages (the Guichon and Thuya batho- liths) are listed in Table 5.1 because neither shows any evidence of interaction with such low-6D meteoric waters. All of the Guichon and Thuya samples have essentially "normal" ED and INTO values (Figures 5.5 and 5.6~. The most plausible explanation of this phenomenon is that it is HUGH P. TAYLOR, JR. c eiotite or Hornblen'e /11 P stone {a ccrIty Chloritiz~ Volcanic Rocks ~ Dikes x Xenoliths + Metosedimentary Rocks - ~fSr~R~ COAST O~4NAGAN ~ PRISON _ 1 1 1 1 1 , 1 1 0 100 200 300 400 coo 600 700 DISTANCE IN KILOMETERS E the absence of late Mesozoic and Cenozoic meteoric hy- drothermal activity that preserved the "old" K/Ar ages of about 200 m.y. ago (Ma) in these granites. This is very feasible for these two batholiths, because both are rela- tively small and they crop out at considerable distances from any of the younger batholiths of southern British Columbia (Figure 5.4~. Whereas most of the batholiths shown in Figure 5.4 are composite in that they contain plutons with ages ranging from early Mesozoic to mid- Tertiary, the Guichon and Thuya batholiths exclusively exhibit only early Mesozoic ages. If we examine the geographic variations of both ED and INTO throughout southern British Columbia, we obtain a series of L-shaped patterns,-each one characteristic of a TABLE 5.1 D/H Ratios of the Most Altered Granites in Various Areas of Southern British Columbia (after Magantz and Taylor, 1986) ED Range, per mil Vancouver Island (except Kennedy Lake) Kennedy Lake area Western Coast Batholith Central Coast Batholith Eastern Coast Batholith Princeton area Raft Batholith Okanagan Batholith Nelson Batholith -46 to -63 -80 to -94 -65 to -95 -80 to -90 .-114 to-130 -107 to -129 -125 to-136 -130 to -15 1 -147 to-163

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS -80 -90 ~ -100 co -110 -120 - 130 - 140 -150 -40 , 1 , , ~1 ' ' ' /~\1 1 , , 1 ~o ~ ~^ As. . ~OKANAGAN .~ I ~_ 11 , -160 ~ . . NELSON l I I l I i ~ I ~I I I I I I -4 -2 o 2 4 6 8 10 12 8 180 FELDSPAR (/~) specific geographic area. These effects show up nicely when ED is plotted against NO feldspar (Figures 5.6 and 5.7~. In both diagrams the horizontal arms of each L represent the samples that have been subjected to the high- est water-rock ratios; in each case the horizontal arm dis- plays an approximately constant ED value characteristic of equilibrium with the meteoric hydrothermal waters of that particular geographic area. Making certain assumptions about the temperature and other parameters (see Taylor, 1977), we may calculate the ED value of the H2O that coexisted with the hydrothermal chlorites and biotites in the heavily altered samples, as shown in Figure 5.7. Such a ED value will represent the original value of the surface waters involved in the hy +40 ~ , ~ ~ ~ ' ' ' ~ ~ ' ' ' ' ' +20 O _ -20 _ -40 -60 C] To -80 -100 -120 -140 -160 1 1 , 1 1 1 1 1 1 1 , 1 1 1 _, '0-18 -16 -t4 -12 -10 -8 -6 -4 -2 0 +2 +4 +6 +8 +10 +12 8180 (Jo) Meteor/c doter l ice ~ EASTERN COPSE =-85' O~ANAGAN~- _ =-95 _~ HANCO~:SMOW =-15~ SHIRT / - - COASTAL =-50' ~ 77 FIGURE 5.6 Plot of ED versus 6180 feld- spar for samples of the various granitic batholiths studied by Magaritz and Taylor (1986). The stippled area labeled 200 m.y. indicates samples from the relatively old Thuya and Guichon batholiths (see text). drothermal convective systems at the time of alteration, so by plotting this value on the meteoric water line (Craig, 1961), we can obtain a complete picture of the original isotopic compositions of these waters prior to their en- trance into each hydrothermal system; such a calculation was carried out for the various geographical areas deline- ated in Figure 5.6, with the results as shown in Figure 5.7. A suite of samples was collected from the Okanagan Batholith along virtually the entire length of Okanagan Lake (Figure 5.41. This suite of samples differs dramati- cally from those in the main part of the Okanagan Batho- lith in that every sample is markedly depleted in deu- terium, typically -130 to -154. The Okanagan Lake samples are also typically depleted in LEO and intensely chlontized; FIGURE 5.7 Plot of ED versus 6'8O feld- spar, showing the various batholith fields from Figure 5.6, the meteoric water line of Craig (1961), and the calculated ED and 6'8O values of the pristine, unexchanged meteoric waters that were the source of the hydrothermal fluids that produced the various "inverted-L" patterns in each batho- lith in southern British Columbia (see Taylor, 1977, for details of the calcula- tions).

OCR for page 72
78 they lie within the horizontal arm of the inverted-L pattern shown for the Okanagan Batholith in Figure 5.6. Along the Okanagan Lake traverse, the Mesozoic plu- tons are intruded by several much younger porphyry stocks and dikes (the Eocene Coryell intrusions and associated Princeton volcanics). The dikes and porphyries consis- tently have lower INTO values than the granitic country rocks that they intrude, even though the Mesozoic plutons have locally been thoroughly depleted in i8O; some samples have whole-rock INTO as low as +0.2 and +1.1 and feldspar INTO as low as -2.8 and-0.2. The dramatically larger RIO depletions observed in the vicinity of Okanagan Lake are almost certainly due in large part to the abundant Tertiary dikes and stocks that occur near this lineament. However, Okanagan Lake also probably occupies a major fracture zone, providing for enhanced permeability that would al- low much greater circulation of surface waters to great depths. In fact, this zone of weakness (a rift zone?) may also have provided the access routes followed by the Ter- tiary magmas that were the immediate cause of the large- scale hydrothermal alteration that is so prominent in the vicinity of this topographic depression. The Nelson Batholith (Figure 5.4) is also a composite batholith, dominantly Mesozoic in age, whose southwest- ern portion is intruded by a group of Tertiary plutons (Coryell intrusions). All of the strikingly i8O-depleted samples from the Nelson Batholith were collected either (1) in close proximity to these Tertiary porphyry intru- sions or (2) along the edge of one of the large, narrow, north-trending lakes that are so prominent in southern British Columbia. Analogous to the situation described above for the more extensive sample set from Okanagan Lake, it is probable that both Slocan Lake and Arrow Lake occupy fracture zones that represented major hydrother- mal conduits for heated Tertiary meteoric waters as well as access routes for the Coryell intrusions. Meteoric Hydrothermal Effects of Eocene Magmatism, Southern Idaho Batholith A series of isotopic studies by Taylor and Magaritz (1978), Criss et al. (1982), Criss and Taylor (1983), and Criss and Fleck (1987) showed that widespread meteoric hydrothermal systems formed in the Idaho Batholith about 40 to 45 Ma, associated with the emplacement of several large epizonal Eocene plutons. The Eocene plutons were intruded at rather shallow depths, probably less than 7 km, and some intrude coeval volcanic rocks of the Challis vol- canic field. The map in Figure 5.8 illustrates the general- ized geology of the southern two-thirds of Idaho Batho- lith, particularly focusing on the ~8O/~60 effects produced by these hydrothermal systems. The surrounding Meso- zoic plutons underwent striking PRO depletions over more than 8000 kITl2 (Taylor, 1977; Taylor and Magaritz, 1978; HUGH P. TAYLOR, JR. Cress et al., 1982, 1984; Cnss and Taylor, 1983~. Deu- ter~um depletions in these rocks are observed across an even wider zone, probably at least 25,000 km2 (practically the entire area of the batholith shown in Figure 5.8 has ED < -120~. The very extensive D/H effects are a result of the fact that only tiny amounts of H2O are required to "reset" the D/H ratios of the hydroxyl-beanng minerals in a gran- ite (Taylor, 19771. The hatchured contours in Figure 5.8 indicate the ex- tent of He zones of PRO depletions and disequilibrium quartz- feldspar ~8O/~60 fractionations, analogous to those shown for southern British Columbia in Figure 5.1. So far as can be established with the presently available data, most of the Eocene hydrothermal systems in Idaho are Type 1. Only very local development of deeper-seated, hotter, Type 2 hydrothermal systems has been found, and these are only t~) 11~ __~ Us - A DA HO ~ BATHOLITH;` r: _ o to ~ o ~ C. ~ q FIGURE 5.8 Generalized geologic map of south-central Idaho showing the Idaho batholith (blank), Eocene epizonal plutons (solid black), and the Challis Volcanics (stippled). The hachured line denotes the perimeter of identified zones of intense meteoric hydrothermal alteration and 18O depletion, all of which are re- lated to these Eocene plutons. The CRZ is a giant zone of alteration associated with the Casto pluton, and the SRZ is the Sawtooth Ring Zone, both of which are thought to be remnants of giant Eocene calderas (after Criss and Taylor, 1983; Criss et al., 1984).

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS VALLES CALDERA LAKE CITY-SILVERTON CALDERAS . ARRAIGN RED ~=NTAlN; Stan La" DOSES ~ Rl" ~86~3/ GONG RING FAULT COD ~517t" ~4 INT - CAL~ ~ ~ But ~Rl" r - LJ5 or MOUE* TUFF _ _ - ... . ._._. hi SPRIN" ~- 6 CHRIS . 6 ~C FIEF DENSER sas_;~! oo~E ~sI JO ~ ~ it_ YELLOWSTONE NATI ONAL PARK o lo , . . . 0 10 20 30 NOKId 1 1 ~ 20~l ~I DAHO BATHOLITH observed directly adjacent to some of the largest Eocene plutons, where fresh unchloritized biotites exhibit 40 to 45 Ma K/Ar ages (Cries et al., 19821. The largest low-~8O zone, the 4500-km2 Casto Ring Zone (CRZ), is centered on the 700-km2 Casto pluton and includes the remains of two giant cauldron complexes in the Challis Volcanics (Cries et al., 1984~. Another large (2500 km2) system is the Sawtooth Ring Zone (SRZ), an annular zone of low-~8O values that encompasses the Sawtooth batholith (Figure 5.9) and that probably repre- sents the subvolcanic part of the ring-fracture system of an Eocene caldera. Several other zones, mostly a few hundred square kilometers in extent, are associated with hydrother- mal metamorphism in the contact zones of smaller igneous bodies. Isotopic material balance w/r ratios of approxi- mately unity characterize most of these anomalous re- gions, regardless of their size. The principal zones of EGO depletion in the Idaho Batholith are thus associated with either (1) the highly fractured ring zones of caldera com- plexes or (2) smaller stocks, some of which are resurgent intrusions emplaced into the central portions of deeply eroded calderas. The Eocene plutons in the northern part of the Idaho batholith are just as numerous and just as large as in the 79 FIGURE 5.9 Comparison of the sizes and hydrothermal flow patterns of caldera-re- lated hydrothermal systems (modified from Taylor, 1974c; Cnss and Taylor, 1983; and Larson and Taylor, 1986c). Black areas shown in the Lake City-Silverton calderas and the Idaho Batholith indicate zones of very strong ~80 depletion, typically associ- ated with caldera ring-fracture zones or with central resurgent intrusions within the calderas. Thermal springs in the presently active Yellowstone National Park are also shown as black areas (see Christiansen, 1984~. Note the similarity between the present-day distribution of thermal springs at Yellowstone and the Eocene features in the fossil meteoric hydrothermal systems in the SRZ (Sawtooth Ring Zone). The map of the Valles caldera is shown for size comparison [this is another present- day, very active geothermal system that is strikingly similar in size and geologic set- ting to the Miocene (23 Ma) Lake City caldera studied by Larson and Taylor (1986a,c)~. south, but the i~o/~60 hydrothermal effects there are much diminished compared to those shown in Figure 5.8 (Cries and Fleck, 1987~. The sizes of the areas with SD values less than -120 are also very much diminished in the north (Cries and Fleck, 1987~. This suggests that the southern area was either more highly fractured or that to the north we may be simply looking deeper into the Eocene conti- nental crust, either of which would imply smaller permea- bilities in the north. If it is the latter, this would imply that the Type 1 systems simply die out with depth as the w/r ratios decrease. This in turn supports the idea that Eocene Type 2 systems were rare in Idaho, and in fact such Type 2 systems in general probably encompass much smaller volumes of rock than Type 1 systems; this is because Type 2 systems require either (1) the simultaneous development of both high w/r ratios and high temperatures or (2) that the hydrothermal circulation persists for an unusually long period of time. The fossil hydrothermal systems in the Idaho Batholith are among the largest associated with granitic plutons anywhere in the world (Cries and Taylor, 1986), and it is interesting in Figure 5.9 to compare the isotopic relation- ships observed in Idaho with the relationships found in the Quaternary Yellowstone Volcanic Field (described in more

OCR for page 72
80 detail below). Despite the massive size of the Eocene hydrothermal systems in Idaho and despite their similarity in size and general character to the Yellowstone caldera systems, in Idaho we have not yet observed any develop- ment of the types of low-~8O magmas that are so common at Yellowstone (or in the southwestern Nevada caldera complex, in Iceland, or in the Seychelles Islands, as dis- cussed below). Thus, a transition to a Type 2 system may be a requirement for the production of significant volumes of low-~8O magmas. On the other hand, the apparent absence of low-~8O magmas in Idaho may simply be due to a lack of detailed sampling of the Eocene Challis Volcan- ics or to the evidence having been eroded away. Miocene Meteoric Hydrothermal Systems, Western Cascade Range, Oregon The first ~8O/~60 study to demonstrate widespread mete- oric hydrothermal effects in North America was that of Taylor (1971), who demonstrated that anomalously low 6~80 values (-6 to +4) occur around several Miocene dio- nte and granodiorite plutons that intrude the Tertiary vol- canic rocks of the Western Cascade Range. The reason that these were the first fossil meteoric hydrothermal sys- tems to be defined in North America is that the petrogra- phic descriptions in a paper by Buddington and Callaghan (1936) were perceived to be astonishingly similar to fea- tures observed in the rocks of the Scottish Hebndes and the Skaergaard intrusion; the latter were the first identified fossil meteoric hydrothermal systems in the world (Tay- lor, 1968), and it therefore seemed likely that the Oregon localities might also be good candidates for such systems (Taylor, 197 1~. The low-~8O plutonic and volcanic rocks of the Western Cascades are typically strongly propylitized and are esti- mated by Taylor (1971) to comprise a total of 1200 km2 (8 percent) of this volcanic area (Figure 5.10~. Rocks col- lected more than three stock diameters from the intrusive contacts typically have "normal" INTO values of +5.8 to +8.2. These isotopic data were interpreted in terms of convective circulation of heated groundwaters during the crystallization and cooling of the central intrusions. A 75- km2 zone of low-~8O volcanic rock (average +1.1) occurs in the Bohemia mining distnct, which is typical of these plutonic centers (Figure 5.10~. Two small (<3 km) stocks of augite granodionte and several dikes and plugs intrude the andesites, luffs, and breccias in the central part of this concentric low-~8O zone, and a 0.3- to 0.6-km-wide con- tact metamorphic aureole of tourmaline hornfels surrounds the larger stocks and grades into a zone of propylitic al- teration at greater distance (Buddington and Callaghan, 1936~. ~. . . . . . i_ HUGH P. TAYLOR, JR. 46. 45o 44o 43o , EXPLANATION ~ intrusives ~ WASHINGTON ..... ..... _ altered rocks a , / - pmrOSpeocrt Portland ~,~~~ , 7 Hd W:~; O EGOS | ~ ~ Jo ff orson Eugene 0 ~ ,~: 1 23 ~1~ 2 10 0 10 20 MILES la ' , I l I r 0 10 20 30 km FIGURE 5.10 Map of part of western Oregon showing the distribution of Tertiary volcanic rocks of the Western Cascade Range (dark stippled pattern); also shown are the occurrences of Tertiary diontes and granodiontes as well as areas of propylitic alteration and mineralization commonly associated with these medium-grained igneous rocks (after Taylor, 1971~. The num- bers indicate the venous localities studied by Taylor (19711; Area (2) is the Bohemia Mining District referred to in the text. Other Localities The above discussion has focused only on the three most widespread regional studies of fossil meteoric hy- drothermal systems in the western United States and Canada. A large number of such systems have now been described in the literature, but most are relatively small, isolated systems associated with individual granitic plu- tons (typically small stocks). Many of these occurrences nave open ~aou~a~ea oy Has and Taylor (1986), and a few of the most notable systems are described below.

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS Peninsular Ranges Batholith At the present stage of erosion this giant composite batholith in southern and Baja California shows little evidence for interactions with heated surface waters over most of its outcrop area (Taylor and Silver, 1978; Silver et al., 1979; Taylor, 19861. However, all along its western edge in California, which is the shal- lowest part of this plutonic complex and the only portion where the roof of the batholith is exposed, extremely low GINO values and nonequilibrium quartz-feldspar fractiona- tions are observed. Interestingly, just south of the United States-Mexico border, the roof-zone of this batholith changes its l80/~60 characteristics, and much higher INTO values and "reversed" quartz-feldspar fractionations are observed; Taylor and Silver (1978) attribute these effects to exchange with higher-'8O waters, such as marine forma- tion waters or ocean water. Thus, the southern California portion of this batholith was apparently intruded in a subaerial environment, but farther south the surface envi- ronment above the plutons must have changed from subaerial to submarine during the time period of emplace- ment 130 to 105 Ma in the Lower Cretaceous. An analo- gous situation today might be observed where the Aleutian chain of calc-alkaline volcanoes (submarine) grades onto the Alaskan Peninsula (subaerial). Lake City and Silverton Calderas, San Juan Mountains, Colorado Most of the mid-Tertiary calderas of the San Juan Mountains appear to have been associated with mete- oric hydrothermal systems (Taylor, 1974c). Two of the most heavily investigated examples are adjacent calderas (Silverton and Lake City) that occupy the even older Uncompahgre caldera complex in the Western San Juans; these two calderas are joined by the highly fractured and altered Eureka graben (Taylor, 1974c; Casadevall and Ohmoto, 1977; Forester and Taylor, 1980; Larson and Taylor, 1986a,b,c). As observed in Idaho and Yellow- stone (Figure 5.9), the most intense alteration and the most striking EGO depletions in both of these caldera complexes are associated with either (1) the caldera ring-fracture zone or (2) the central resurgent intrusion. The ]80/~60 effects have been measured over vertical distances of at least 2 km, and they may be readily inferred to have extended to at least a depth of 5 km beneath the original land surface. Boulder Batholith, Montana In western Montana ex- tensive oxygen and hydrogen isotope evidence for mete- oric hydrothermal activity was found in the Boulder Batho- lith by Sheppard and Taylor (1974~. They showed that the ED values of various plutons in this composite batholith range from - 3 to -155 per mil, with the lowest ED values typically being associated with relatively low INTO values, epidote and chlorite alteration, and disequilibrium quartz 81 feldspar LEO fractionations. The lowest ED and INTO values occur in mineralized areas along the western side (roof) of the batholith in the Butte and Wickes mining districts or are associated with small stocks intruded into the Bighorn Mountains Volcanics. Alterations may have begun in the Late Cretaceous, but the dominant episode was in the early Tertiary; these Paleocene meteoric waters have calculated D values of -100 to -115 and initial GINO values of -14 to -16 (compare with Figure 5.7~. The mineralization in the main-stage veins in the giant Butte Cu-Zn-Pb-Mn deposit show clear-cut evidence of deposition from meteoric hy- drothermal fluids, and these low-~8O effects can be ob- served in the slopes, shafts, and tunnels of the mine itself to extend over a vertical distance of more than 2 km. This Paleocene (~60 Ma) meteoric hydrothermal system formed entirely within the extremely highly fractured Late Creta- ceous Butte quartz monzonite pluton; some of the main- stage, sulfide-rich veins are several meters wide, and indi- vidual veins can be followed laterally for 4 to 5 km and vertically for 1 to 2 km. The entire meteoric hydrothermal system must have originally extended over a vertical dis- tance of at least 5 km. Western Nevada Gold-Silver Ore Deposits All of the so-called epithermal, volcanic-hosted, Au-Ag deposits of the Basin and Range Province appear to have formed from heated meteoric groundwaters (Taylor, 1973, 1974a). At least 25 separate localities have been documented by Taylor (1973, 1974a) and O' Neil and Silberman (1974~. Some of the largest and most significant examples are Tonopah, Goldfield, Comstock Lode, Bodie, and Aurora. Many of these deposits are associated with the ring fractures of mid- to late-Tertiary rhyolitic caldera complexes. Central British Columbia Magaritz and Taylor (1976) showed that the entire eastern side-of the Coast Plutonic Complex at latitudes 55 to 56N displayed evidence of meteoric hydrothermal alteration. These effects extend well out into the sedimentary and volcanic country rocks, and they can also be observed in the vicinity of smaller plutons as much as 200 km eastward from the Coast Batho- lith. The D/H ratios have been lowered to values less than -120 over an area of at least 5000 km2. The alteration appears to be mainly associated with Eocene plutons, which are abundant throughout the area. HYDROTHERMAL SYSTEMS ASSOCIATED WITH LAYERED GABBRO BODIES Layered gabbro plutons that have established meteoric hydrothermal convective systems are the Skaergaard in- trusion in East Greenland, Jabal at Tirf in Saudi Arabia, Stony Mountain in Colorado, the Islands of Skye and Mull

OCR for page 72
82 in Scotland, and Ardnamurchan in Scotland (Taylor and Forester, 1971, 1979; Forester and Taylor, 1976, 1977, 1980; Taylor, 1980, 1983~. Analogous examples involving ocean-water hydrothermal systems are observed in all ophiolite complexes, including Cyprus and the Samail ophiolite, Oman (Gregory and Taylor, 1981), and in dredged samples from the Indian Ocean (Ito and Clayton, 1983; Stakes et al., 1983~. In all these areas large portions of the layered gabbro complexes display marked i8O/~60 dise- quilibrium between coexisting pyroxene and plagioclase, exactly analogous to the effects shown for quartz and feldspar in Figure 5.1. Mineralogical Alteration Effects in ~8O-Depleted Layered Gabbros Although there is local development of chlorite, epi- dote, actinolite, talc, sphere, prehnite, and other low-tem- perature (greenschist facies) minerals in most layered gabbros, particularly in areas of diking, heavy fracturing and veining, or multiple intrusion, it is important to realize that ~8O-depleted gabbro bodies in general are astonish- ingly free of any of the petrographic or mineralogical features that geologists commonly utilize as indicative of hydrothermal alteration. This statement, of course, does not apply where an older gabbro body has been invaded and hydrothermally metamorphosed by a younger intru- sion (e.g., the heavily altered Broadford gabbro body on Skye; see Forester and Taylor, 1977), but it definitely applies to all layered gabbro bodies involved in a single cycle of meteoric hydrothermal activity (i.e., the convec- tion cells set up by that particular gabbro magma chamber itself). Because petrographic evidence for hydrothermal altera- tion is usually very rare in such ~8O-depleted gabbros (al- though it can often be perceived through a careful thin- section study), the ~8o/~6o "reversals" between plagioclase and pyroxene always represent the easiest and most defini- tive way to characterize such phenomena in gabbros. However, even in the absence Of TO studies, fossil hydrothermal systems should be suspected if any of the following fairly subtle petrographic features are observed: (1) clouding or turbidity of some of the plagioclase; (2) development of minor talc-magnetite rims on olivine; (3) coarsening of exsolution lamellae around microfractures in clinopyroxene grains; (4) macroscopic veins that are easily seen on slightly weathered outcrops in glaciated areas or arid regions (on careful examination these veins often prove to contain pyroxene, magnetite, and high- temperature amphiboles such as hornblende); and (5) development of minor amounts of actinolite, biotite, chlo- rite, and epidote in zones transitional to lower-temperature HUGH P. TAYLOR, IR. hydrothermal activity (e.g., late-stage veins). All of the above features are described in the references given above and/or in Norton et al. (1984~. In addition, Ferry (1985) demonstrated that secondary magnetite (relatively pure Fe3O4) is ubiquitous in the Skye gabbros and also that the calculated temperatures of formation of the talc-olivine- orthopyroxene assemblages in these gabbros are in the range of 525 to 545C. The early veins throughout prac- tically the entire section of layered gabbro in the Skaer- gaard intrusion contain hydrothermal clinopyroxenes with minimum solvus temperatures of 500 to 750C (Manning and Bird, 1986~. It is thus an inescapable conclusion that layered gabbro bodies typically undergo meteoric hydrothermal alteration at very high temperatures, in large part in the range 500 to 900C (Norton and Tavlor 1979 T~vlor ~nc1 F~r~t~r 1979; Taylor, 1987~. ~ ~ . ~ _~ _ A _^ ~ v ~_^ ~ This conclusion is firm, despite statements to the contrary by Cathles (1983) on this mat- ter. Only at such high temperatures could we simultane- ously obtain the clear i80/~60 evidence for intense hy- drothermal alteration combined with the virtually com- plete absence of low-temperature hydrous minerals. The temperatures (500 to 900C) and PH2O values (200 to 800 bars) are clearly not within the stability fields of chlorite, epidote, serpentine, actinolite, and clay minerals during the bulk of the hydrothermal activity. In fact, much of the mineralogical alteration that does occur in these gabbros takes place at 450 to 550C, considerably higher than the average temperature of alteration in the granitic plutons described above but still lower than the tempera- tures at which the bulk of the ~8O depletion occurred in these gabbros (Taylor end Forester, 1979; Ferry, 1985~. Why then do we mainly see only the effects of 350 to 400C H2O, for example, at the mid-ocean ridge spreading centers and on land in places like the Salton Sea? As Norton and Knight ~ 1977) and Norton ~ 1984) have empha- sized, convective systems are strongly controlled by the approximate coincidence of a viscosity minimum, together with a maximum in the isobaric coefficient of thermal expansion in critical to supercritical H2O (350C < T < 450C; 200 < P < 800 bars). These are also the conditions where the density of H2O undergoes the most rapid change as a function of temperature. Furthermore, the heat capac- ity of the fluid under these pressure and temperature con- ditions is quite large and is maximized at the critical point. Thus, in this general pressure-temperature range, the buoy- ancy and heat transport properties of the fluid are maxi- mized and the drag forces minimized (Norton and Knight, 1977~. In fact, it is very likely that these physical proper- ties of H2O control the observed upper limit of 350 to 400C observed during surface venting of modern hy- drothermal systems, not the "fact" that all rocks hotter than 400C are impermeable, as proposed by Cathles (1983~.

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS volumes of 2500, 280, and 1000 km3, respectively. Very large scale, Type 1 hydrothermal systems are present all along the ring fractures of the youngest caldera and adja- cent to the resurgent domes (Figure 5.91. Given the fact that volcanic and hydrothermal activity has been going on here for more than 2 m.y., and that a large magma body exists at the present time at a depth of only a few kilome- ters (Christiansen, 1984), it is virtually certain that Type 2 systems also locally occur at depth. Almost all of the Yellowstone rhyolites are somewhat depleted in '8O relative to the "normal" 6'8O values of +7 to +10 usually observed in silicic volcanic rocks through- out the world (Taylor, 1968, 1974a). However, the post- caldera rhyolites of the first and third caldera cycles in- clude some extraordinarily low i8O eruptive units (Figure 5.11~. The combined areal extent of the two first-cycle low-~8O flows (6'8O = +2.9 to +3.6) is ~66 km2; their original volume was at least 10 km3, and they were erupted 30,000 to 350,000 yr after the >2500-km3 Huckleberry Ridge Tuff. These low-'8O lavas cannot be chemically distinguished from the higher-'8O Yellowstone rhyolites; they all contain ~76 wt.% SiO2 and 15 to 20 percent phenocrysts of quartz, sanidine, plagioclase, clinopyrox- ene, fayalite, and Fe-Ti oxides. During the third cycle, enormous i8O depletions were observed (Figure 5.11) in lavas that vented in two separate areas ~45 km apart. These were erupted immediately after the eruption of the >1000-km3 Lava Creek Tuff, along the , , , ,_ , , Extracoldera 15 unit. ~ km' LCT~d I. 1l: ~ rS ~tD _ ~ ~ ~ Intracc Ide ra `/ 21 units ~ -01~' TSC ~ t CF UT ~ ~ HRT (22) - \~' . Flrst Co/40 km3 of magma, perhaps as much as 70 km3, all of it having [jl80 between about +0.6 and +1.2. The second area of very low 6~8O third-cycle rhyolites is in the south- westem ring-fracture zone, where scattered exposures in a 25-km2 area have 6~8O values between ~.1 and +0.6. Hildreth et al. (1984) list the following observations that are critical to any theory of origin of the Yellowstone magmas: 1. The narrow range of 6'8O in the three major ash- flow sheets (+5 to +7) contrasts sharply with the wide variation in the postcaldera rhyolitic lavas (Figure 5.11~. The biggest '8O depletions directly follow the two largest ash-flow eruptions. 2. The '8O depletions were geologically short-lived events (<300,000 to 500,000 yr) that followed caldera subsidence in some cases by less than 50,000 to 100,000 yr. Successively younger postcaldera lavas show partial recovery of the magma toward precaldera 6'8O values, as a result of mixing with deeper levels of the magma reser- vo~r. 3. The pattern of stepwise ~8O depletions is reflected in lavas erupted as far .apart as 115 km, indicating that they were part of an integrated magmatic system. More than 100 km3 of magma was depleted by about 5 per mil and more than 1000 km3 of magma was depleted by 1 to 2 per milt 4. Depletion of i8O occurred only in the subcaldera reservoir; contemporaneous rhyolites from outside that caldera have very high 6'8O values. 5. Sr and Pb isotopic ratios of the rhyolites display a zigzag pattern similar to that displayed by 6'8O in Figure 5.11, jumping to more radiogenic values just subsequent to caldera formation (Doe et al., 1982~. The caldera- forming events obviously introduced country-rock Pb and Sr into the magma system. Low-'8O Basaltic and Rhyolitic Magmas in Iceland The data of Muehlenbachs et al. (1974), Hattori and Muehlenbachs (1982), Condomines et al. (1983), and Gunnarsson et al. (1988) can be briefly summarized as follows (see Figure 5.12~. 1. The relatively rare alkali olivine basalts on Iceland (~8O = +5.3 to +5.7), which are found only on the periph- ery of the island at the edges of any deep meteoric hy- drothermal circulation systems, have distinctly higher and more uniform (i.e., more "normal") 6~8O values than the much more abundant tholeiites and transitional basalts. 2. The lowest and most heterogeneous 6~8O values in

OCR for page 72
transit Alk Bas. ~ 4.0 to4.9 ~ O Alkali Basalt ' | ~ SiliacCertres ~ 3.0 to3.9 ~ 5.0 to5.7 ~ t8 0 50 km FIGURE 5.12 Map of Iceland showing the 6~80 values of tholeiitic and transitional basalts within the recently active volcanic zones (modified from Sheppard, 1986~. The location of silicic volcanic centers is also shown. Data are from Muehlenbachs et al. (1974) and Condomines et al. (1983~. Also shown are the locations of the three drill sites (1 = Reydarfjordur, 2 = Krafla, 3 = Reykjavik) where Hattori and Muehlenbachs (1982) obtained samples for TO analysis (see Figure 5.13~. HUGH P. TAYLOR, JR. rations between 6~80 and chemical composition; increas- ing SiO2 and K2O tend to be accompanied by decreasing INTO for each petrologic class of basalts. In particular, the tholeiite with by far the most extreme 6~80 value (+1.8) also has by far the highest K2O content. 6. The rhyolites and obsidians all over Iceland tend to have lower INTO values than the basalts, and the rhyolites with the lowest and most variable SILO values all come from the eastern rift zone. 7. Deep drill holes at several localities show a striking RIO depletion in the lavas as a function of depth, particu- larly in the eastern rift zone (Figure 5.13~. 8. There is a weak correlation between INTO and 3He/ 4He, with the intermediate and silicic volcanic glasses typically having 3He/4He ratios close to the atmospheric value. The isotopic relationships described above strongly support the idea that assimilation and/or partial melting of hydrothermally altered country rocks in the deeper parts of the rift zone is the most likely mode of formation of the low-~8O magmas from Iceland. This explanation is com- patible with the Pb isotope data of Welke et al. (1968) and the 87Sr/86~r and rare-earth data of O'Nions and Gronvold (1973), and it readily explains why the most contaminated (i.e., most RIO depleted) magmas are either the rhyolites or Ol the Icelandic basalts, +1.8 to +5.4, are all confined to the soo quartz tholeiites of the eastern rift zone, particularly those from the large Krafla volcano. 3. The olivine tholeiites from the western rift zone have intermediate INTO values, +4.0 to +5.7, with the high est (i.e., most "normal") values closest to the coast on the southwest part of Reykjanes Peninsula. 4. In flank-zone volcanoes (e.g., Torfajokull and Hekla), where volcanism is transitional alkalic to alkalic in com position and is superimposed on older crust formed during an earlier period of rifting, the basaltic to andesitic lavas tend to have intermediate INTO values; these lavas are consistently only slightly depleted in i8O (by about 0.5 to 1.0 per mill compared to "normal" basalts, while the spa tially related or contemporaneous silicic lavas have simi lar or much lower INTO values (e.g., Torfajokull: +3.6 to +4.4~. At Torfajokull, parallel volcanic fissures, 1 to 2.5 km apart, have erupted lavas of variable chemical compo sition, from basaltic to silicic; these are associated with small, fissure-dependent, and very systematic variations in 6~80 that correlate very well with major and trace ele ments, suggesting eruption from distinctly different, well mixed magma chambers beneath the central volcano. 5. On a regional scale there are also some crude corre , I I _ 1 1 ' ~ t7 ~ \ ~ b ~ 1000 - - I 15= En 2000 3000 D~/LL-COR~ SAMPLES - , - /CELA/~/D , , , , , 1 , , -10 -8 - 6 - 4 - 2 o +2 +4 8180 WHOLE- ROCK (I/) . CO ~_ ., +6 FIGURE 5.13 Plot of depth versus 6180 (modified after Hattori and Muehlenbachs, 1982) for samples of hydrothermally altered basalt from three drill holes in Iceland (see Figure 5.12 for locations). The vertical dotted line indicates the average TO value of most terrestrial tholeiitic basalts (including mid-ocean ridge samples) and lunar basalts. This figure demonstrates that essentially all of the volcanic rocks in Iceland are depleted in i8O down to at least 3 km, with the RIO depletions becoming more pronounced with depth and with proximity to a major active central volcano in the Eastern Rift Zone (e.g., Krafla). The solid lines connect samples from a single drill core.

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS those basalts that have probably most strongly interacted with roof rocks or rhyolite melts, namely the K-rich and Fe-rich tholeiites. Direct exchange with meteoric waters would not be expected to produce these relationships. The overall smaller ~8O-depletion effects in the lavas from flank- zone volcanoes compared to those of active rift-zone vol- canoes can be attributed to lower temperatures, lower water- rock ratios, and shallower hydrothermal circulation sys- tems within the flank-zone environment (Gunnarson et al., 1988). It is probably significant that all of the extremely low RIO basaltic and rhyolitic magmas are confined to the eastern rift zone, which has been active only during the past 3 m.y. to 4 m.y. (Saemundsson, 19741. The magmas in this rift zone are penetrating upward through the lower parts of volcanic and plutonic rocks, which presumably were intensively hydrothermally altered in an earlier epi- sode of magmatic activity at the time of their original formation in the western rift zone. Thus, the magmas coming up through the eastern rift zone would be interact- ing with country rocks that had already suffered heterogeneous RIO depletions and which, through subsi- dence~ have been brought down into a much higher tem- perature regime (15 to 20 km depths. Something on the order of 200 km3 of low-~8O tholeiite has been erupted in the eastern rift zone in the past 12,000 yr (Jacobsson, 19721. It was the scale of this process that most bothered Muehlenbachs et al. (1974) when they re- jected the meteoric hydrothermal explanation for the ori- gin of these types of magmas. This mechanism for the origin of the Icelandic volcanic rocks has, however, been strongly favored over the past few years by Taylor ~ 1974a, 1977,1979), and recently Hattori and Muehlenbachs (1982), Condomines e! al. (1983), and Gunnarson et al. (1988) have provided strong new support for this mechanism. The model of Condomines et al. (1983), based on combined He, O. Sr, and Nd isotopic relationships, is shown in Figure 5.14. It is similar to the models of Taylor (1977, 1979) and Gunnarsson et al. (19881. These work- ers propose that the primary l80/~60 ratios of mantle-de- rived magmas were changed in a deep magma reservoir by exchange or contamination between the magma and the surrounding meteoric hydrothermally altered basaltic crust. Such processes do not appear to have introduced much water into the magma because Icelandic volcanism, except for subglacial eruptions, is usually not explosive, and hydrous minerals in platonic ejecta are rare. The rhyolitic magmas are thus assumed to have been produced at rela- tively shallow crustal levels, either by melting of hydroth- ermally altered rocks or as the deeper magmas moved upward and underwent further contamination processes. These processes caused introduction of atmospherically derived helium into the magmas. 87 6,'80 basat ~__~.5~:;) Volcano Ohm. ~ I-rat \SOc~-- __ ,'0~3 ~ _ _ '1 cG\'9 ,80 deleted meteoric /~~~ `~.~3~ Hydrothermal alteration ~ 1 Roof -rock m elfin 9 and km-- - --:;'+~' Deep crustal . reservoir 'i' ~ ~80 ~ 5 5 ? - 'n cP ,' lo FIGURE 5.14 Schematic section summarizing current models for the evolution of the Icelandic crust (after Condomines et al., 1983~. Radiogenic helium (HeR) is generated in the subsided lavas according to their age and U and Th contents, taken to be 0.3 and 1 ppm, respectively. Assimilation, magma mixing, and/ or exchange processes between the magmas and the hydrother- mally altered volcanic pile, some of which has now subsided to great depths, are assumed to be the main processes that deter- mine the isotopic and chemical compositions of the magmas. Seychelles Batholith, Indian Ocean The Seychelles Islands are unique among oceanic areas in that they are almost wholly composed of granite, a remnant of the breakup of the continent of Gondwanaland left isolated between Africa and India in the middle of the Indian Ocean. The main island of Mahe (Figure 5.15) is made up of a remarkably homogeneous, leucocratic, coarse- grained hornblende microperthite granite (Baker, 1963~. This granite is latest Precambrian in age (650 Ma to 670 Ma; Miller and Mudie, 1961; Wasserburg et al., 1964; Michot and Deutsch, 1976~. No remnants of older country rocks have been found anywhere on the island, except for a few small xenoliths in the granite. Except for two granite porphyry intrusions on the western side of the island, practically all of Mahe is made up of a single granite pluton that is just as homogeneous in l80/~60 and D/H as it is in terms of texture and mineralogy (Taylor, 1974b, 1977~. The data in Figure 5.15 leave no doubt that the main Mahe pluton crystallized from a very homogeneous, low- ~8O granitic magma having a INTO = +3.3 + 0.3. This is proved because (1) the striking l80/~60 homogeneity in the Mahe pluton could not have been produced by later sub- solidus hydrothermal exchange and (2) the ALSO quartz- feldspar values are extremely uniform at 1.25 to 1.50, identical to the equilibrium values in "normal" granites. The existence of a meteoric hydrothermal event prior to emplacement of the pluton is confirmed by the occurrence

OCR for page 72
88 ED ~-82 tO -90 ~ BOO ~ 63 SO 6S J_ - So 81eOK ' 4.1 tO 5.0 ~ ,~ .. ~ ED ~-92 0-99 / ~..~..2: \ ^~8OO-K >> ~ 5 / it::::::::) B18OO *-Of IO 33 ...... 2 22 ~ B1BOK S-2.1 to 0~ ~.".22a22222222- ~ ~ 'm2:22.~ \ _.; at ~_~::::::: ~.,,,. MAIN PLUTON D-102 to-lO9 A180~: K ~1.2 to 1.5 818Oo *4.0 to 4.5 ~- 818OK = 2.6 to 32 O SCALE 5 FIGURE 5.15 Generalized geological map of Mahe, the princi- pal island of the Seychelles Group (after Taylor 1974b, 1977), showing the variation of 6~80 and ED (Q = quartz, K = alkali feldspar). Most of the island is composed of a coarse-grained hornblende microperthite granite with an extremely uniform isotopic composition, as shown in the figure. The only excep- tions are some locally distinct facies distributed along the coastal areas: (a) a higher-~8O granite border zone, remnants of which locally remain along the northwestern and eastern coasts; (b) a lower-~80 facies with very large 6~80 quartz-feldspar values, recognized only along the NE coast. The granite porphyry bod- ies (shown in solid black), although texturally very distinct, are isotonically similar to the higher-~8O granite facies (~80 quartz = 5.3 to 6.7; 6~80 feldspar = 3.8 to 4.7; ED = -82 to -951. The dotted contacts shown on the map are only approximate, but they suggest that the present outline of the island probably approxi- mates the original dimensions of the main Mahe pluton. of amphibolite xenoliths in some of the coastal granite samples that are even lower in 180 than the granite host rock. The extreme RIO depletions and the very large ALSO quartz-feldspar values (up to 3.5) in the northern coastal samples (see Figure 5.15) are evidence that the marginal portions of the Seychelles granitic complex underwent subsolidus meteoric hydrothermal alteration in a Type 1 system. We infer that a Type 2 system probably existed at even greater depths and that melting of such rocks pro- duced this large (>300 km3), remarkably homogeneous mass of low-~8O granitic magma that makes up the island of Mahe. Meteoric hydrothermal effects are not confined to Mahe. A granite sample from Praslin, the second largest island of the Seychelles Group (50 km northeast of Mahe), contains HUGH P. TAYLOR, JR. very turbid alkali feldspar with a KILO = -1.6 and has a very large nonequilibrium TO quartz-feldspar (+8.6~. All these data clearly document a widespread hydrothermal event in the Seychelles terrane about 670 Ma, involving waters with an initial 6180 at least as low as -5. Origin of Low-'80 Granitic Magmas (~80 < +6) One of the most important points to be made about low- ~8O magmas is simply the immense difficulty of identify- ing them in the geologic record. This, of course, is not a problem for very young, recently erupted lavas such as those found in Iceland. However, for older volcanic rocks the KILO values of even the freshest-appearing samples are notoriously unreliable as a result of incipient weathering and hydrothermal effects (Taylor, 1968~. Nevertheless, the KILO values of phenocrysts can often be used to calculate the KILO of the coexisting magma (see Taylor and Shep- pard, 1986~. This was in fact the way the Yellowstone RIO/ 160 analyses of Hildreth et al. (1984) were carried out (Figure 5.11~. However, occasionally the phenocrysts of such volcanic rocks will not be in equilibrium with their coexisting magmas. Also, if the rocks are deeply buried or if they suffer even a low-grade metamorphism or hydro- thermal alteration, it may be difficult (or impossible) to determine the KILO values of the original magmas. The above problems are particularly severe in the case of low-~8O granitic plutons. Without doing an extremely detailed geological and geochemical study, there is usu- ally no simple way to distinguish a pluton that formed from a low-~8O granitic magma from one that started out originally as a normal-~8O or high-~8O pluton but that later interacted with a Type 2 or Type 3 hydrothermal system. As an example of these ambiguities, look again at the data in Figure 5.1. Disequilibrium I8O/160 effects are common in this 700-km traverse from Vancouver Island eastward across southern British Columbia (Magaritz and Taylor, 1986~. However, whereas- the easternmost batholiths (Okanagan and Nelson) display the characteristic steep slopes of Type 1 systems, to the west in the deeper-seated Coast Plutonic Complex these KILO arrays define shal- lower slopes and swing over closer to the 45 line. In general, the samples that lie near the 45 line in Figure 5.1 probably represent an approach to Type 2 hydrothermal conditions, but this cannot be proved with the available data; many of the samples might represent low-~8O mag- mas. Some of these same phenomena are observed in the Hercynian-age high-grade gneisses, migmatites, and posttectonic granites of the southern part of the Black Forest (Schwarzwald) in West Germany. This classic migmatite area (e.g., Mehnert, 1968) is closely analogous in age and geologic setting to the Hercynian metamorphic

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS systems in the Pyrenees described by Wickham and Taylor (Chapter 6, this volume). One of the main differences is that whereas the waters in the Pyrenees are apparently marine in origin, the Black Forest terrane was subaerial and involved meteoric waters; the latter area locally exhib- its remarkably low 6~80 values (0 to +4), and, as Magaritz and Taylor (1981) and Hoefs and Emmerman (1983) have shown, the ]80/~60 fractionations among coexisting miner- als range from nearly equilibrated values (e.g., Type 2) to strongly disequilibrium values (Type 1~. These are the lowest-~8O values yet found in such deep-seated mig- matites, and even though the adjacent host-rock schists are also depleted in RIO, the two kinds of rocks are in general not equilibrated (i.e., Type 3 conditions were either not attained or they were later overprinted by a retrograde Type 1 system). In other areas of the Black Forest the schists and gneisses have BIRO values that range upward to +7 to +12. A set of complex Type 1 and Type 2 meteoric hydrothermal systems was clearly produced in this area about 300 Ma, particularly within and along an elongate, down-dropped block of Upper Paleozoic sedimentary and volcanic rocks. The granitic rocks and migmatites devel- oped within this apparent rift zone are strikingly depleted in i8O relative to the rocks on either side (Taylor et al., 1989), and it is virtually certain that at least locally low- i8O granitic magmas were developed. More detailed stud- ies are continuing in this fascinating area to try to unravel these problems. In the light of the above discussion it might appear that the rarity of low-~8O magmas in the geologic record is more apparent than real. Nevertheless, a number of work- ers have been actively searching for more examples of such magma systems for about 20 yr, with only limited success. Two features stand out in this two-decade-long search: (1) such magmas are much less abundant than was originally thought to be the case in the 1970s and (2) no new giant occurrences comparable to Yellowstone or Ice- land have been found among Late Cenozoic lavas any- where in the world. Other than the two examples of low- ~8O rhyolitic and granitic magmas discussed above, the only other documented major occurrences of low-~8O magmas are in Iceland (e.g., Muehlenbachs, 1973; Hattori and Muehlenbachs, 1982; Condomines et al., 1983), in the southwestern Nevada caldera complex (Lipman, 1971; Lipman and Friedman, 1975), and a few rare examples in the Tertiary intrusive complex of Skye (Forester and Taylor, 19771. Hildreth et al. (1984) explain the development of low- ~80 magmas by direct influx of large amounts of H2O into the magma. On the other hand, Taylor (1974a, 1977, 1980, 1983, 1986, 1987) argues that such magmas form either by (1) partial melting of hydrothermally altered rhyolitic country rocks and/or (2) foundering of such ~8O 89 depleted roof rocks into the magma chamber, accompa- nied by melting, assimilation, or exchange of this foreign material with the magma reservoir. This conclusion is based mainly on the geological factors outlined below and on constraints imposed by the physics and chemistry of H2O transport through ductile rocks and silicate melts. The most abundant country rocks above and along the margins of the Yellowstone magma chamber prior to each ash-flow eruption were certainly hydrothermally altered, ~8O-depleted, earlier-cycle rhyolites. These young vol- canic rocks cannot provide the Pb and Sr isotope signa- tures that abruptly appear in the postcaldera rhyolites (Doe et al., 1982~; therefore, some melting of pre-Tertiary coun- try rocks is required, and, if this occurred, the much more abundant rhyolitic roof rocks would have been very exten- sively melted. However, the only chemical effect of this process on the magma reservoir would be: (1) lower 6~8O values, (2) lower ~D values, and possibly (3) higher H2O concentrations. This is because these rhyolitic country rocks are chemically and isotopically almost identical to their original parent magmas, except for the subsolidus hydrothermal changes they have undergone, which to a first approximation only involve hydration and depletion in ~8O and D. An important feature of the Hildreth et al. (1984) study is the abrupt interval over which the isotopic changes occur (Figure 5.11). Similar, short time-scale, catastrophic processes involving caldera collapse had been invoked previously by Taylor ~1974a, 1977), but up until the Hildreth et al. (1984) study there was no proof that such extreme 6~8O changes do in fact occur on such a rapid time scale. Even though diffusive transport of enormous amounts of H2O directly into the magma a priori seemed unlikely, there was always the possibility that given enough time it perhaps could occur. The data of Hildreth et al. ~ 1984) are extremely important because they-show that 6~8O changes in the magma chamber do not occur by such a long-term process. In fact, the long-term changes are in the opposite direction, toward recovery of the original magmatic 6~8O value. On such a short time scale there is no physical way to separate such a huge volume of pore water from its rock matrix. Hildreth et al. (1984) raised several objections against the partial melting or bulk assimilation process, among which are the material-balance problem and the fact that there is no evidence of the type of massive cooling or increase in phenocryst content that might be expected in the contaminated magmas, nor is there any obvious change in major or trace element composition. The first objection is invalid because mixing or exchange with large volumes of H2O would have an even more profound cooling effect. The second argument is also refuted by the fact that the dominant rock types being melted are older-cycle rhy

OCR for page 72
9o elites with essentially identical chemical compositions to the younger magmas. Such rhyolites are full of hydrous alteration minerals, and their latent heats of fusion are very low (<30 cal/g), both of which make them very sus- ceptible to melting. Hydrothermally altered rocks with INTO =-8 to -9 are common at Yellowstone (Hildreth et al., 1984), and at depths of several kilometers their INTO values could be expected to be as low as -12 or lower, judging by the data from the deep Krafla drill hole on Iceland (Figure 5.13~. Young volcanic rocks are also very porous, so the foundered roof rocks that fall into the magma chamber or that subside into a deep melting zone during such a catastrophic process are not just rock but H2O- saturated rock, with perhaps 20 to 35 percent pore space. Therefore, the water-saturated rocks could easily have an average 6~80 =-13 to-15, markedly easing the material- balance difficulties. Why are these low RIO magmas formed in enormous volumes in some volcanic fields, but not others? Why do some eruptive units in the southwest Nevada and Yellow- stone volcanic fields show an order of magnitude greater TO depletion than some other units or any of those from the central Nevada, San Juan, and Superstition volcanic fields? This is the most striking problem raised by the work of Larson and Taylor (1986b), who showed that magmas with remarkably uniform 6~80 values (overall variations of <1 per mil) were generated over time periods of more than 3 to 6 m.y. in two different caldera com- plexes in Nevada and Colorado. Larson and Taylor (1986b) were able to discern only one major factor that separates the low-~8O occurrences from the normal-~8O occurrences, namely emplacement into a rift-zone tectonic setting. Emplacement into a Rift-Zone Tectonic Setting The only two caldera complexes in North America that are known to have erupted large volumes of low RIO mag- mas (Yellowstone and Southwest Nevada) are both younger than 15 m.y. The period between 15 Ma and 20 Ma corresponds to the time of transition into the brittle, frac- ture-dominated extensional tectonic regime of the Basin- Range province (see, for example, Stewart, 1978; Zoback et al., 1981; Eaton, 1982~. The Yellowstone caldera, in fact, lies on the eastern end of the currently active Snake River Plain rift system, and the Southwest Nevada caldera complex lies right in the midst of abundant Basin-Range extensional features. Such region-wide extension must produce fractures that penetrate deeply into the crust. These fractures could allow meteoric water to circulate very deeply, as they clearly have in Iceland (Muehlenbachs et al., 1974), which is also a well-defined rift-zone spreading HUGH P. TAYLOR, JR. center and where low-~8O silicic volcanic rocks are also abundant. In retrospect, the connection between a rift-zone tec- tonic setting and low-~8O magmas may seem to be a fairly obvious one. After all, where on Earth is there a better chance of bringing into close juxtaposition the two geo- logical materials that are essential to make very high temperature hydrothermal systems and low-~8O magmas? Only in rift zones and spreading centers do we find the large-scale extensions and brittle fracturing that are neces- sary to allow massive amounts of magma to come upward into the crust, as well as providing the greatly increased fracture permeability that allows surface waters to pene- trate to depths of at least 10 or 15 km. Such environments represent the best way to attain the required combination of very high temperatures together with large quantities of low-~8O, hydrothermally altered rocks and meteoric pore waters. CONCLUSIONS The simple fact of the existence of low-~8O magmas, together with the gigantic scale at which meteoric hy- drothermal and seawater hydrothermal convective systems have been shown to operate on Earth, constitutes the most definitive proof that surface waters can in large amounts locally penetrate to great depths in the Earth's crust. The existence of low-~8O magmas also proves that these sur- face waters can circulate downward into very high tem- perature environments, essentially into the zone of melting itself. This in turn implies that some of these complex processes take place under lithostatic pressures. Deep circulation of surface-derived aqueous fluids under hydrostatic conditions appears to be ubiquitous in areas of igneous activity, and in some- fossil hydrothermal systems where the deep-seated rocks are exposed by erosion the effects can be shown by direct observation to have ex- tended to at least 8- to 10-km depth. Significant convec- tive circulation and isotopic exchange between rocks and aqueous fluids thus should always occur around igneous intrusions, except in certain cases and under the following conditions. 1. Intrusion into relatively impermeable country rocks- for example, into either (a) limestones and evaporites, which are susceptible to ductile deformation at pressure- temperature conditions where most other rocks undergo brittle fracture, or (b) ordinary silicate rocks at sufficient depth in the crust that the fracture permeability is less than 10ri5 cm2 (1 - millidarcy), or where the temperatures and pressures are high enough for recrystallization and ductile

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS metamorphism to occur, particularly in the absence of extensional tectonics (rifting), rapid strain rates, and brittle deformation (see Norton and Taylor, 1979; Taylor, 1987~. 2. In the immediate presence of the silicate melt, which also will be essentially impermeable to aqueous fluids under hydrostatic pressures, as long as the strain rates are low enough and the percentage of melt high enough so that an interconnected fracture network cannot develop. This is very important in layered gabbro plutons because the late-stage magma sheet at the roof of the intrusion can provide a barrier between two decoupled hydrothermal systems: a very hot (>500C), lower hydrothermal system in the layered cumulates and a cooler system at 250 to 450C and higher water-rock ratios in the roof rocks. 3. If magmatic aqueous fluids are being exsolved from late-crystallizing portions of a (granitic) pluton, this will produce a magmatic H2O envelope under lithostatic pres- sure that fills all available fractures outward from the crystallization front. This can keep the low-~8O meteoric waters outside the pluton until the magmatic fluid enve- lope has dissipated, by which time the temperatures will have fallen into the range of stability (e.g., sericite and chlorite). Thus, two decoupled hydrothermal systems may also commonly occur around H2O-rich granitic magma chambers (e.g., porphyry Cu deposits; see Taylor, 1974a), but in such cases the two types of aqueous fluids are genetically very different and are under different pres sures. We have shown that layered gabbro cumulates typi- cally undergo hydrothermal interaction with externally derived aqueous fluids (e.g., meteoric water, seawater) at much higher temperatures than in the case of granitic plutons (typically 500 to 900C versus 200 to 450C). This is manifested in the absence or rarity of hydrous alteration minerals (e.g., amphibole, chlorite, epidote) in gabbros that show clear-cut INTO signatures of having inter- acted with very large quantities of external H2O. In such rocks the GINO effects may be the only obvious indications of intense hydrothermal exchange. Although a number of characteristics contribute to the higher alteration tempera- tures of the gabbros, such as higher solidus temperature, greater latent heat of crystallization, higher melt-crystal ratio, and characteristically lower fracture density, the two most important factors appear to be (1) the different ge- ometry of crystallization of a layered gabbro pluton com- pared to a granite pluton and (2) the higher magmatic H2O contents of granitic plutons. Granitic magmas produced in rift zones and areas of extensional tectonics commonly (invariably?) appear to involve anatexis of rocks that were hydrothermally altered at high water-rock ratios. This conclusion also probably applies to local "pull-apart" basins associated with major 91 strike-slip faults. Such extensional zones provide favor- able access routes for upward rise of magmas as well as for downward penetration of surface waters; the magma bodies then act as giant "heat engines" that promote con- vective circulation of the waters. These aqueous fluids can produce large-scale l80/~60 changes in the source rocks of certain kinds of granitic magmas, particularly if the water is isotopically distinctive (as when low-~8O meteoric groundwaters are involved). The styles of water-rock interaction also change with increasing depth, because with increasing temperature and time of alteration the isotopically heterogeneous, unequilibrated Type 1 systems may locally give way to heterogeneous, equilibrated Type 2 systems, and finally at the highest T. t, and w/r we may in some cases obtain Type 3 systems that are both equili- brated and homogenized. Prior to or during the formation of these kinds of hy- drothermal-anatectic granitic magmas, there must be a transition from hydrostatic to lithostatic conditions; the nature and timing of this process are not yet well under- stood. Also, these hydrothermal-anatectic granitic mag- mas will be readily recognized only if they are abnormally depleted in ~8O, which specifically requires deep circula- tion of enormous volumes of very low ~8O groundwaters into the Type 2 and Type 3 hydrothermal regimes; such processes occur only in the most highly fractured, most permeable parts of the Earth's crust. In the absence of other compelling geochemical or geophysical data such as that prOesented by Wickham and Taylor (1985, 1987; Chap- ter 6, this volume), Wickham (1987), and Bickle et al. (1988), in general it will not be possible to prove "hy- drothermal" origin of such magmas. The }80/~60 signa- ture of the original source of such an aqueous fluid may be completely obscured, particularly if the w/r ratios are rela- tively small. The D/H signature will be preserved but commonly will not be useful, because except in the case of seawaters (very high SD) or high-latitude meteoric waters (very low SD) there will in general not be sufficient iso- topic contrast compared to typical magmatic and meta- morphic fluids (which have ~D = - 0 to -90~. Therefore, "disguised" magmas formed by the general type of hydrothermal-anatectic process outlined in this chapter and in Chapter 6 by Wickham and Taylor con- ceivably could be extremely common but very difficult to recognize. They might even be the dominant silicic magma type in most areas of rift-zone tectonics. The testing of this hypothesis will be an important problem to be ad- dressed in future studies of the petrology and geochemis- try of granites. In this connection Wickham et al. (1987) discovered a number of isotopic features in northeast Nevada that are similar to those in the Pyrenees, with one major difference: the fluids were low-~8O meteoric wa

OCR for page 72
92 ters rather than high-D marine pore fluids. This compan son is important because it is certain that Nevada was subaerial in the mid- to late-Tertiary and that major crustal extension took place over a broad area at that time. ACKNOWLEDGMENTS This work was supported by the National Science Foundation, grant no. EAR83-13106. I am particularly indebted to Robert E. Cnss, Robert T. Gregory, Robert I. Hill, Peter B. Larson, Mordekai Magantz, Denis Norton, Leon T. Silver, G. Cleve Solomon, and Stephen M. Wick- ham for their help and collaboration on the problems dis- cussed in this paper and for numerous discussions over the years on hydrothermal systems and the origin of granites. This is contribution no. 4590, Publications of the Division of Geological and Planetary Sciences, California Institute of Technology. REFERENCES Baker, B. H. (1963~. Geology and mineral resources of the Seychelles Archipelago, Memoir, Geological Survey of Kenya 3, 1-140. Bickle, M. J., S. M. Wickharn, H. J. Chapman, and H. P. Taylor, Jr. (1988~. A strontium, neodynium, and oxygen isotope study of the hydrothermal metamorphism and crustal anatexis in the Trois Seigneurs Massif, Pyrenees, France, Contributions to Mineralogy and Petrology 100, 399-417. Buddington, A. F., and E. Callaghan (1936~. Dioritic intrusive rocks and contact metamorphism in the Cascade Range in Oregon, American Journal of Science 31, 421-449. Burnham, C. W. (1979~. Magmas and hydrothermal fluids, in Geochemistry of Hydrothermal Ore Deposits, 2nd ea., H. L. Barnes, ea., John Wiley and Sons, New York, pp. 71-136. Casadevall, T., and H. Ohmoto (1977~. Sunnyside mine, Eureka mining district, San Juan County, Colorado: Geochemistry of gold and base metal ore deposition in a volcanic environment, Economic Geology 72, 1285-1320. Cathles, L. M. (1983~. An analysis of the hydrothermal system responsible for massive sulfide deposition in the Hokuroku Basin of Japan, in The Kuroko and Related Volcanogenic Massive Sulfide Deposits, Economic Geology Monograph 5, Society of Economic Geologists, Golden, Colo., pp. 439-487. Christiansen, R. L. (1984~. Yellowstone magmatic evolution: Its bearing on understanding large-volume explosive volcanism, in Explosive Volcanism, Geophysics Study Committee, Na- tional Academy Press, Washington, D.C., pp. 84-95. Cole, D. R., and H. Ohmoto (1986~. Kinetics of isotopic ex- change at elevated temperatures and pressures, in Stable Iso- topes in High Temperature Geological Processes, J. W. Val- ley, H. P. Taylor, Jr., and J. R. O'Neil, eds., Reviews in Mineralogy, vol. l 6, Mineralogical Society of America, Wash- ~ngton, D.C., pp. 41-90. Condomines, M., K. Gronvold, P. J. Hooker, K. Muehlenbachs, R. K. O'Nions, N. Oskarsson, and E. R. Oxburgh (19831. HUGH P. TAYLOR, JR. Helium, oxygen, strontium and neodymium isotopic relation- ships in Icelandic volcanics, Earth and Planetary Science Let- ters 66, 125-136. Craig, H. ~ 1961). Isotopic variations in meteoric waters, Science 133, 1702-1703. Criss, R. E., and R. J. Fleck (1987~. Petrogenesis, geochronol- ogy, and hydrothermal systems in the northern Idaho batholith and adjacent areas based on TO, D/H, 87Sr/86Sr, K-Ar, and 40Ar/39Ar studies, U.S. Geological Survey Professional Paper 1436. Criss, R. E., and H. P. Taylor, Jr. (1983~. An TO and D/H study of Tertiary hydrothermal systems in the southern half of the Idaho batholith, Geological Society of America Bulletin 94, 640-663. Criss, R. E., and H. P. Taylor, Jr. (1986~. Meteoric-hydrothermal systems, in Stable Isotopes in High Temperature Geological Processes, J. W. Valley, H. P. Taylor, Jr., and J. R. O'Neil, eds., Reviews in Mineralogy, vol. 16, Mineralogical Society of America, Washington, D.C., pp. 373-424. Criss, R. E., M. A. Lanphere, and H. P. Taylor, Jr. (19821. Effects of regional uplift, deformation, and meteoric-hydro- thermal metamorphism on K-Ar ages of biotites in the south- ern half of the Idaho batholith, Journal of Geophysical Re- search 87, 7029-7046. Criss, R. E., E. B. Ekren, and R. F. Hardyman (1984~. Casto ring zone: A 4500 k2 fossil hydrothermal system in the Challis volcanic field, central Idaho, Geology 12, 331-334. Criss, R. E., R. T. Gregory, and H. P. Taylor, Jr. (1987~. Kinetic theory of oxygen isotope exchange between minerals and water, Geochimica et Cosmochimica Acta 51, 952-960. Doe, B. R., W. P. Leeman, R. L. Christiansen, and C. E. Hedge (1982~. Lead and strontium isotopes and related trace ele- ments as genetic tracers in the Upper Cenozoic rhyolite-basalt association of the Yellowstone Plateau volcanic field, Journal of Geophysical Research 87, 4785-4806. Eaton, G. P. (1982~. The Basin and Range Province: Origin and tectonic significance, Annual Reviews of Earth and Planetary Sciences 10, 409-440. Ferry, J. M. (1985~. Hydrothermal alteration of Tertiary igneous rocks from the Isle of Skye, northwest Scotland, 1. Gabbros, Contributions to Mineralogy and Petrology 91, 264-282. Fleck, R. J., and R. E. Crisp (1985~. Strontium and oxygen isotopic variations in Mesozoic and Tertiary plutons of central Idaho, Contributions to Mineralogy and Petrology 90, 291- 308. Forester, R. LOW., and H. P. Taylor, Jr. (1976~. ~8O-depleted igne- ous rocks from the Tertiary complex of the Isle of Mull, Scotland, Earth and Planetary Science Letters 32, 11-17. Forester, R. W., and H. P. Taylor, Jr. (1977~. TO, D/H and ~3C/~2C studies of the Tertiary igneous complex of Skye, Scotland, American Journal of Science 277, 136-177. Forester, R. W., and H. P. Taylor, Jr. (1980~. Oxygen, hydrogen, and carbon isotope studies of the Stony Mountain complex, western San Juan Mountains, Colorado, Economic Geology 75, 363-383. Friedman, I., P. Lipman, J. D. Obradovich, J. D. Gleason, and R. L. Christiansen (1974~. Meteoric water in magmas, Science 184, 1069-1072.

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS Giletti, B. J. (1986~. Diffusion effects on oxygen isotope tem- peratures of slowly cooled igneous and metamorphic rocks, Earth and Planetary Science Letters 77, 218-228. Giletti, B. J., and R. A. Yund (1984~. Oxygen diffusion in quartz, Journal of Geophysical Research 89, 4039-4046. Giletti, B. J., M. P. Semet, and R. A. Yund (1978~. Studies in diffusion: III. Oxygen in feldspars: An ion microprobe determination, Geochimica et Cosmochimica Acta 42, 45-57. Gregory, R. T., and H. P. Taylor, Jr. (1981~. An oxygen isotope profile in a section of Cretaceous oceanic crust, Samail Ophio- lite, Oman: Evidence for ~8O-buffering of the oceans by deep (>5 km) seawater-hydrothermal circulation at mid-ocean ridges, Journal of Geophysical Research 86, 2737-2755. Gregory, R. T., R. E. Criss, and H. P. Taylor, Jr. (1988~. Oxygen isotope exchange kinetics of mineral pairs in closed and open systems: Applications to problems of hydrothermal alteration of igneous rocks and Precambrian iron formations, Chemical Geology, in press. Gunnarsson, B. J., H. P. Taylor, Jr., and B. D. Marsh (1988~. Origin of oxygen isotope anomalies in Icelandic lavas: Part I. Flank-zone volcanoes, Geological Society of America Abstracts with Programs 20. Hattori, K., and K. Muehlenbachs (1982~. Oxygen isotope ratios of the Icelandic crust, Journal of Geophysical Research 87, 6559-6565. Hildreth, W., R. L. Christiansen, and J. R. O 'Neil (1984~. Cata- strophic isotopic modification of rhyolitic magma at times of caldera subsidence, Yellowstone Plateau volcanic field, Jour- nal of Geophysical Research 89, 8339-8369. Hoers, J., and R. Emmerman (1983~. The oxygen isotopic composition of Hercynian granites and pre-Hercynian gneisses from the Schwarzwald, SW Germany, Contributions to Miner- alogy and Petrology 83, 320-329. Ito, E., and R. N. Clayton (1983~. Submarine metamorphism of gabbros from the Mid-Cayman Rise: An oxygen isotopic study, Geochimica et Cosmochimica Acta 47, 535-546. Jacobsson, S. P. (1972~. Chemistry and distribution pattern of recent basaltic rocks in Iceland, Lithos 5, 365-386. Larson, P. B., and H. P. Taylor, Jr. (1986a). An oxygen isotope study of water/rock interaction in the granite of Cataract Gulch, western San Juan Mountains, Colorado, Geological Society of America Bulletin 97, 505-515. Larson, P. B., and H. P. Taylor, Jr. (1986b). TO ratios in ash- flow luffs and lavas erupted from the central Nevada caldera complex and the central San Juan caldera complex, Colorado, Contributions to Mineralogy and Petrology 92, 146-156. Larson, P. B., and H. P. Taylor, Jr. (1986c). An oxygen isotope study of hydrothermal alteration in the Lake City caldera, San Juan Mountains, Colorado, Journal of Volcanology and Geo- thermal Research 30, 47-82. Lipman, P. W. (1971~. Iron-titanium oxide phenocrysts in compositionally zoned ash-flow sheets from southern Nevada, Journal of Geology 79, 438-456. Lipman, P. W., and I. Friedman (1975~. Interaction of meteoric water with magma: An oxygen-isotope study of ash-flow sheets from southern Nevada, Geological Society of America Bulletin 86, 695-702. Magaritz, M., and H. P. Taylor, Jr. (1976~. TO and D/H 93 studies along a 500 km traverse across the Coast Range batho- lith and its country rocks, central British Columbia, Canadian Journal of Earth Sciences 13 (11), 1514-1536. Magaritz, M., and H. P. Taylor, Jr. (1981~. Low t8O migmatites and schists from the tectonic contact zone between Hercynian (=Variscan) granites and the older gneissic core complex of the Black Forest (Schwarzwald), West Germany, Geological Society of America Abstracts with Programs 13, 501. Magaritz, M., and H. P. Taylor, Jr. (1986~. Oxygen-18/oxygen- 16 and D/H studies of platonic granitic and metamorphic rocks across the Cordilleran batholiths of southern British Columbia, Journal of Geophysical Research 91, 2193-2217. Manning, C. E., and D. Bird (1986~. Hydrothermal clinopyrox- enes from the Skaergaard intrusion, Contributions to Mineral- ogy and Petrology 92, 437-447. Mehnert, K. R. (1968~. Migmatites, Elsevier, Amsterdam, 393 PP. Michot, J., and S. Deutsch (1976~. Seychelles, Microcontinent or not? (abs.), 4th European Colloquim of Geochronology, Cosmochronology, and isotope Geology, 68. Miller, J. A., and J. D. Mudie (19611. K-Ar age determinations on granite from the Island of Mahe in the Seychelles Archipelago, Nature 93, 1174-1175. Muehlenbachs, K. (1973~. The oxygen isotope geochemistry of siliceous volcanic rocks from Iceland, Carnegie institution of Washington Yearbook 72, 593-597. Muehlenbachs, K., A. T. Anderson, and G. E. Sigvaldason (1974~. Low-i8O basalts from Iceland, Geochimica et Cosmochimica Acta 38, 577-588. Norton, D. (1984~. Theory of hydrothermal systems, Annual Reviews of Earth and Planetary Sciences 12, 55-177. Norton, D., and J. Knight (1977~. Transport phenomena in hydrothermal systems: Cooling plutons, American Journal of Science 277, 937-981. Norton, D., and H. P. Taylor, Jr. (1979~. Quantitative simulation of the hydrothermal systems of crystallizing magmas on the basis of transport theory and oxygen isotope data: An analysis of the Skaergaard intrusion, Journal of Petrology 20, 421-486. Norton, D., H. P. Taylor, Jr., and D. K. Bird (1984~. The geometry and high-temperature brittle deformation of the Skaergaard intrusion, Journal of Geophysical Research 89, 10,178-10,192. O'Neil, J. R., and M. L. Silberman (1974~. Stable isotope rela- tions in epithermal gold-silver deposits, Economic Geology 69, 902-909. O'Nions, R. K., and K. Gronvold (1973~. Petrogenetic relation- ship of acid and basic rocks in Iceland: Sr isotopes and rare earth elements in late post-glacial volcanics, Earth and Plane- tary Science Letters 19, 397-409. Saemundsson, K. (1974~. Evolution of the axial rifting zone in northern Iceland and Tjornes fracture zone, Geological Soci- ety of America Bulletin 85, 495-504. Sheppard, S. M. F. (1986~. Igneous rocks: III. Isotopic case studies of magmatism in Africa, Eurasia, and oceanic islands, in Stable Isotopes in High Temperature Geological Processes, J. W. Valley, H. P. Taylor, Jr., and J. R. O'Neil, eds., Reviews in Mineralogy, vol. 16, Mineralogical Society of America, Washington, D.C., pp. 319-371.

OCR for page 72
94 Sheppard, S. M. F., and H. P. Taylor, Jr. (1974). Hydrogen and oxygen isotope evidence for the origins of water in the Boul- der Batholith and the Butte ore deposits, Montana, Economic Geology 69, 926-964. Sheppard, S. M. F., R. L. Nielsen, and H. P. Taylor, Jr. (19711. Hydrogen and oxygen isotope ratios in minerals from por- phyry copper deposits, Economic Geology 66, 515-542. Sheppard, S. M. F., P. E. Brown, and A. D. Chambers (1977~. The Lilloise intrusion, East Greenland: Hydrogen isotope evi- dence for the efflux of magmatic water into the contact meta- morphic aureole, Contributions to Mineralogy and Petrology 68, 129-147. Silver, L. T., H. P. Taylor, Jr., and B. W. Chappell (1979). Some petrological, geochemical, and geochronological observations of the Peninsular Ranges batholith near the international bor- der of the U.S.A. and Mexico, in Mesozoic Crystalline Rocks, P. L. Abbott and V. R. Todd, eds., Annual Meeting Guide- book, Geological Society of America, Boulder, Colo., pp. 83- 110. Stakes, D. S., H. P. Taylor, Jr., and R. C. Fisher (1983). Oxygen isotope and geochemical characterization of hydrothermal al- teration in ophiolite complexes and modern oceanic crust, in Ophiolites and Oceanic Lithosphere, I. D. Gass, S. J. Lippard, and A. W. Shelton, eds., Special Publication of the Geological Society of London, Blackwell Scientific Publishers, London, pp.199-214. Stewart, J. H. (1978~. Basin-Range structure in western North America, Geological Society of America Memoir 152, 1-131. Taylor, H. P., Jr. (19681. The oxygen isotope geochemistry of igneous rocks, Contributions to Mineralogy and Petrology 19, 1-71. Taylor, H. P., Jr. (1971~. Oxygen isotope evidence for large- scale interaction between meteoric ground waters and Tertiary granodiorite intrusions, Western Cascade Range, Oregon, Journal of Geophysical Research 76,7855-7874. Taylor, H. P., Jr. (1973~. ~8o/~6o evidence for meteoric-hydro- thermal alteration and ore deposition in the Tonopah, Com- stock Lode, and Goldf~eld mining districts, Nevada, Economic Geology 68,747-764. Taylor, H. P., Jr. (1974a). The application of oxygen and hydro- gen isotope studies to problems of hydrothermal alteration and ore deposition, Economic Geology 69,843-883. Taylor, H. P., Jr. (1974b). A low-~8O, Late Precambrian granite batholith in the Seychelles Islands, Indian Ocean: Evidence for formation of i8O-depleted magmas and interaction with ancient meteoric ground waters, Geological Society of Amer- ica Abstracts with Programs 6,981-982. Taylor, H. P., Jr. (1974c). Oxygen and hydrogen isotope evi- dence for large-scale circulation and interaction between ground waters and igneous intrusions, with particular reference to the San Juan volcanic field, Colorado, in Geochemical Transport and Kinetics, A. W. Hofmann, B. J. Giletti, H. S. Yoder, Jr., and R. A. Yund, eds., Carnegie Institution of Washington, Washington, D.C., pp. 299-324. Taylor, H. P., Jr. (1977~. Water/rock interactions and the origin of H2O in granitic batholiths, Journal of the Geological Soci- ety of London 133, 509-558. HUGH P. TAYLOR, JR. Taylor, H. P., Jr. (1980). Stable isotope studies of spreading centers and their bearing on the origin of granophyres and plagiogranites, in Orogenic Mafic-Ultramafic Association, C. Allegre and J. Aubouin, eds., Colloques Internationaux du C.N.R.S., No. 272, pp.149-165. Taylor, H. P., Jr. (1983~. Oxygen and hydrogen isotope studies of hydrothermal interactions at submarine and subaerial spread- ing centers, in Hydrothermal Processes at Seafloor Spreading Centers, P. A. Rona, K. Bostrum, L. Laubier, and K. L. Smith, Jr., eds., NATO Symposium Volume, Plenum Press, New York, pp. 83-104. Taylor, H. P., Jr. (1986~. Igneous rocks: II. Isotopic case studies of circumpacific magmatism, in Stable Isotopes in High Tem- perature Geological Processes, J. W. Valley, H. P. Taylor, Jr., and J. R. O'Neil, eds., Reviews in Mineralogy, vol. 16, Min- eralogical Society of America, Washington, D.C., pp. 273- 317. Taylor, H. P., Jr. (1987~. Comparison of hydrothermal systems in layered gabbros and granites, and the origin of low-~8O magmas, in Magmatic Processes: Physicochemical Principles, B. O. Mysen, ea., Special Publication 1, Geochemical Society, Washington, D.C., pp. 337-357. Taylor, H. P. Jr., and R. W. Forester (1971~. Low-~8O igneous rocks from the intrusive complexes of Skye, Mull, Journal of Petrology 12, 465-497. Taylor H. P., Jr., and R. W. Forester (1979~. An oxygen and hydrogen isotope study of the Skaergaard intrusion and its country rocks: A description of a 55-m.y. old fossil hydrother- mal system, Journal of Petrology 20, 355-419. Taylor, H. P., Jr., and M. Magaritz (i978~. Oxygen and hydro- gen isotope studies of the Cordilleran batholiths of western North America, in Stable Isotopes in the Earth Sciences, B. W. Robinson, ea., DSIR Bull. 220, New Zealand Department of Scientific and Industrial Research, Wellington, New Zealand, pp. 151-173 Taylor, H. P., Jr., and S. M. F. Sheppard (1986~. Igneous rocks: I. Processes of isotopic fractionation and isotope systematics, in Stable Isotopes in High Temperature Geological Processes, J. W. Valley, H. P. Taylor, Jr., and J. R. O'Neil, eds., Reviews in Mineralogy, vol. 16, Mineralogical Society of America, Washington, D.C., pp. 227-271. Taylor, H. P., Jr., and L. T. Silver (1978~. Oxygen isotope relationships in plutonic igneous rocks of the Peninsular Ranges batholith, southern and Baja California, in Short Papers of 4th international Conference on Geochronology, Cosmochronol- ogy, and Isotope Geology, R. E. Zartman, ea., U.S. Geological Survey Open-File Report 78-701, pp. 423-426. Taylor, H. P., Jr., M. Magaritz, and S. M. Wickham (1989~. Application of stable isotopes in identifying a major Her- cynian rift zone and its associated meteoric-hydrothermal ac- tivity, Southern Schwarzwald, West Germany, in Epstein 70th Birthday Symposium Volume, California Institute of Technol- ogy, Pasadena, pp. 86-91. Wasserburg, G. J., H. Craig, H. W. Menard, A. E. J. Engel, and C. G. Engel (1964~. Age and composition of a Bounty Island granite and age of a Seychelles Islands granite, Journal of Geology 71, 785-789.

OCR for page 72
OXYGEN AND HYDROGEN ISOTOPE CONSTRAINTS ON THE DEEP CIRCULATION OF SURFACE WATERS WeLke, H., S. Moorbath, G. L. Cumming, and H. Sigurdsson (1968~. Lead isotope studies on igneous rocks from Iceland, Earth and Planetary Science Letters 4, 221-231. Wickham, S. M. (19871. Crustal anatexis and granite petrogene- sis during low pressure regional metamorphism: The Trois Seigneurs Massif, Pyrenees, France, Journal of Petrology 28, 127-169. Wickham, S. M., and E. R. Oxburgh (1985~. Continental rifts as a setting for regional metamorphism, Nature 318, 330-333. Wickham, S. M., and H. P. Taylor, Jr. (1985~. Stable isotope evidence for large-scale seawater infiltration in a regional metamorphic terrace: The Trois Seigneurs Massif, Pyrenees, France, Contributions to Mineralogy and Petrology 91, 122- 137. Wickham, S. M., and H. P. Taylor, Jr. (1987~. Stable isotope constraints on the origin and depth of penetration of hydro- thermal fluids associated with Hercynian regional metamor 95 phi sm and crustal anatexis in the Pyrenees , Contributions to Mineralogy and Petrology 95, 255-268. Wickham, S. M., H. P. Taylor, Jr., and A. W. Snoke (19871. Fluid-rock-melt interaction in metamorphic core complexes- a stable isotope study of the Ruby Mountains-East Humboldt Range, Nevada, Geological Society of America Abstracts with Programs 19, 463. Yund, R. A., and T. F. Anderson (1978~. The effect of fluid pressure on oxygen isotope exchange between feldspar and water, Geochimica et Cosmochimica Acta 42, 235-239. Zartman, R. E. (1974~. Lead isotopic provinces in the Cordillera of the western United States and their geologic significance, Economic Geology 69, 792-434. Zoback, M. L., R. E. Anderson, and G. A. Thompson (1981~. Cenozoic evolution of the state of stress and style of tectonism of the Basin and Range province of the western United States, Philosophical Transactions of the Royal Society of London, Ser. A 300, 407-434.