ELSEVIER Earth and Planetary Science Letters 169 (1999) 227–244 www.elsevier.com/locate/epsl Cracking at the magma–hydrothermal transition: evidence from the Troodos Ophiolite, Cyprus K.M. Gillis a,Ł , M.D. Roberts b a School of Earth and Ocean Sciences, P.O. Box 3055, University of Victoria, Victoria, BC V8W 3P6, Canada b Economic Geology Research Unit, James Cook University, Townsville, Qld 4811, Australia Received 10 August 1998; revised version received 1 March 1999; accepted 12 March 1999 Abstract The nature of the magma–hydrothermal transition in oceanic hydrothermal systems is poorly understood, in part because the geological relations in this critical region have rarely been observed in modern ocean crust. Detailed mapping was conducted in the Troodos Ophiolite, Cyprus, where a gabbronorite sequence intrudes the sheeted dyke complex, which is truncated at its base by a thin contact aureole composed of massive hornfels. Geothermometric data for hornblende and pyroxene hornfels show that hydrated sheeted dykes were recrystallized at amphibolite to granulite facies conditions (778–986ºC). Quartz diorite veins and apophyses, and monomineralic amphibole veins cross-cut the contact aureole and show no preferred age relationships. Geothermometric data indicate that quartz diorite was injected at 817–919ºC and that fractures were filled with amphibole at 575–750ºC. Phase relations of quartz-hosted, halite-bearing fluid inclusions in quartz diorite veins constrain minimum entrapment temperatures of 225–520ºC (average 402ºC) and minimum pressures that span lithostatic and hydrostatic conditions. We believe that these characteristics are indicative of a conductive boundary layer that separates an active hydrothermal system from the heat source that drives it. Field and petrological data indicate that transient fracturing caused oscillations in temperature and pressure conditions within the conductive boundary layer, and mixing of hydrothermal and magmatic fluids at the magma–hydrothermal interface. Cross-cutting relations between magmatic and hydrothermal vein networks show that fracturing occurred prior to the cessation of magmatic activity. We explore plausible models for the causes and consequences of fracturing that consider the role of dyke injection, thermoelastic stresses, and volatile build-up. 1999 Elsevier Science B.V. All rights reserved. Keywords: Troodos Ophiolite; geothermal systems; contact metamorphism; aureoles 1. Introduction There is little direct evidence concerning the nature of the magma–hydrothermal transition in oceanic hydrothermal systems. Vent fluid compositions, in combination with experimental and theoŁ Corresponding [email protected] author. Fax: C1 250 721 6200; E-mail: retical studies, constrain conditions within the rootzones of hydrothermal cells [1,2]. In this region, hydrothermal fluid compositions become fixed by reaction with the surrounding rocks at low fluid=rock ratios, near supercritical temperatures, and hydrostatic pressures (500 bar), prior to their ascent to the seafloor (see review by [1]). The constancy of fluid chemistries, in particular soluble elements such as Li and B, on decadal time scales (i.e., the time pe- 0012-821X/99/$ – see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 1 2 - 8 2 1 X ( 9 9 ) 0 0 0 8 7 - 4 228 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 riod vents have been sampled) requires that circulating fluids are continuously in contact with fresh rock [1], either by the migration of hydrothermal rootzones or injection of fresh material by periodic dyking. Input of magmatic volatiles into hydrothermal systems is well documented [3], however, the link between the flux specific volatile species and magma chamber processes remains poorly constrained. Conceptual and mathematical models place the magma–hydrothermal transition at or near the sheeted dyke–plutonic boundary [4]. A critical aspect of these models is the presence of a conductive boundary layer that separates a vigorously convecting hydrothermal cell from the heat source that drives it [4–7]. These layers are modeled as dynamic features whose thickness and position in the crust varies during the life span of a hydrothermal system, depending on the balance between the heat supplied by the convecting magma and heat lost by conduction across the layer [6]. The thickness of the conductive layer varies with magmatic heat flux in that active hydrothermal systems with high thermal discharge (¾103 MW) have very thin layers (meters) whereas those with lower thermal discharge (¾10 MW) have layers on the order of a few hundred meters [7,8]. Models require that conductive boundary layers remain thin in order to maintain high heat flux hydrothermal systems on decadal time scales, by the downward migration of a cracking front into hot rock, or unknown magmatic or tectonic processes [7,8]. Ophiolites offer a complementary venue for studying oceanic hydrothermal processes. Indeed, Cyprus-type volcanogenic massive sulphide deposits are among the best known ancient analogues for modern, mid-ocean ridge sulphide deposits (e.g., TAG [9]). Although most key elements of ophiolite-hosted hydrothermal systems are similar to those at modern mid-ocean ridges, important differences exist in regard to the extent and conditions of alteration within sheeted dyke complexes. Ophiolite sheeted dyke complexes are more uniformly and pervasively altered, at higher fluid=rock ratios, than modern sheeted dyke complexes and contain large (up to 8 km2 ) zones of epidosite where dykes have been extensively metasomatized and recrystallized. Epidosites represent either the reaction zones or base of hydrothermal upflow zones [10,11] and are rare in mid-ocean ridge rock collections. In this paper we present new geological and petrological constraints about the nature of the magma– hydrothermal transition in oceanic hydrothermal systems. We selected the Troodos Ophiolite for study as a wealth of information has been accumulated since the 1960s concerning its magmatic, tectonic, and hydrothermal evolution (see [12]). Our mapping builds on the early work of Bear [13] and Malpas and Brace [14] in an area where intrusion of a sequence of gabbroic rocks into the base of the sheeted dyke complex produced a thin contact aureole. We interpret the contact aureole as a preserved conductive boundary layer which isolated an active hydrothermal cell from its magmatic heat source and document evidence for extreme oscillations in P –T conditions. 2. Geologic relations 2.1. Background The Troodos Ophiolite, located on the island of Cyprus in the eastern Mediterranean, lies at the northwestern end of a belt of Cretaceous ophiolites that are discontinuously exposed along the margin of the Arabian plate. The ophiolite is Turonian in age [15] and is thought to have formed in a supra-subduction zone setting within the Tethys Ocean [16]. Obduction of the Troodos oceanic crust involved a 90º anticlockwise rotation between the Maastrichtian and early Eocene and episodic uplift that culminated in the Early to Mid-Quaternary [12]. Emplacementrelated metamorphism was minimal and is restricted to late-stage fracture and fault-infilling [17]. Estimates for spreading rate of the Troodos ocean crust based on structural features and paleoseafloor topography span the range of possibilities (e.g., [18,19]). What is more relevant to this study is that field relationships suggest temporal changes in magma supply during crustal construction [20]. Field relations between the sheeted dyke complex and plutonic sequence, and between different plutonic bodies clearly show that the Troodos crust was built by multiple magma chambers [21,22]. The most compelling evidence of multiple intrusive relations is the common occurrence of xenoliths of one lithology in a host gabbro of slightly different composition [23]. K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 Complex relations between the sheeted dykes and gabbros also support models for polyphase magmatism. Sheeted dykes are commonly intruded by high level gabbroic plutons, which may be later intruded by dyke swarms, and, in a few locations, dykes arise directly from the pluton [21,24]. The relief along this transition is ½500 m as dyke swarms locally intrude the plutonics from depth and plutons locally intrude to high levels. 2.2. Platanistasa Window The Platanistasa Window is located along the northern flank of the ophiolite, between the villages of Alona, Platanistasa, and Polystipos (Fig. 1). A gabbronorite sequence intrudes the sheeted dyke complex, which is truncated at its base by a thin contact aureole composed of massive, fine grained hornfels. The relief along this contact is <50 m except where it is locally down-dropped <100 m to the east along high angle normal faults. 229 The sheeted dykes strike N to NW, dip steeply to the east (¾60º), and range in thickness from ¾0.5 to >4 m. There is no systematic increase in dyke thickness towards the plutonics, and dykes rooted directly in the plutonic sequence, as described by Allen [24], were not observed. The contact aureole, which separates the basal sheeted dyke complex from the subjacent gabbronorite sequence, is locally cut by a network of quartz diorite veins and apophyses, and monomineralic amphibole veins (Fig. 2). The plutonic sequence is dominantly composed of massive to layered gabbronorite, olivine gabbronorite, and wehrlite, with lesser amounts of gabbro, magnetite gabbro, diorite, quartz diorite, and plagiogranite. Layered gabbronorites extend up to the base of the sheeted dykes along most of the contact and are locally infiltrated by leucocratic vein networks or dykes (Fig. 3, left). The proportion of leucocratic veins in outcrop systematically increases from <10% to 30% as the contact is approached. Varitextured amphibole gabbro occurs within the Fig. 1. Geologic map of the Platanistasa Window. The contact aureole is centered along the igneous boundary between the sheeted dykes and layered series. Topographic relief exposes a vertical section (400–600 m) across the sheeted dyke–plutonic transition along two NE–SW ridges, such that from north to south, sheeted dykes ! plutonics ! sheeted dykes crop out. Inset shows the generalized geology of the Troodos Ophiolite. Contour interval is 200 feet. 230 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 Fig. 2. (A) Outcrop photograph and (B) sketch showing a sharp intrusive contact where gabbro intrudes into pyroxene hornfels. Note that the contact is cross-cut by an amphibole vein network. (C) Outcrop photograph and (D) sketch showing quartz diorite and amphibole veins cross-cutting very fine grained hornfels. layered sequence as massive outcrops, dykes, and veins, and has a wide range in texture and grain size on a deci-meter scale. Plagiogranite bodies locally intrude into the dykes at the same structural level as the contact aureole and have intrusive or gradational boundaries with the gabbronorite sequence (Fig. 1). Outcrops range in composition from diorite to quartz diorite to plagiogranite and contain partially resorbed xenoliths of hornfelsic basalt stoped from the base of the sheeted dyke complex. Elsewhere, gabbro, gabbronorite, magnetite gabbro, and varitextured gabbro with complex intrusive relationships form the roofzone assemblage. Isolated dykes which locally intrude the layered gabbronorite sequence strike NNE and are east-dipping (¾60º). Sheeted dykes and lavas exposed north of the Platanistasa Window are relatively unfaulted and un- tilted, and lie between the Solea and Mitsero grabens [18]. The prevalence of an intrusive sheeted dyke– plutonic boundary and absence of significant dyke rotation are indicative of formation during a phase of magmatic extension. By contrast, the extreme crustal attenuation in the Solea graben records a period of waning magmatism or amagmatic extension [20]. 3. Metamorphic evolution In this section mineral assemblages and compositions, and textural features of key lithologies are presented in order to determine the protolith of the hornfels and constrain the thermal conditions in the vicinity of the contact aureole. Mineral compositions for key lithologies are summarized in Table 1. K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 231 Fig. 3. (Left) Schematic diagram showing typical geological relationships along the sheeted dyke–plutonic transition. Sheeted dykes (vertical lines) are truncated at their base by the gabbronorite sequence (light gray). As the contact aureole is approached, the abundance of felsic melt impregnations (solid squares) within the gabbronorites increases (up to 30%). The contact aureole is cross-cut by quartz diorite (white lines C solid squares) and amphibole (white lines) veins; these veins show no preferred age relationships. (Right) Thermal conditions in the vicinity of the contact aureole. See text for details. Abbreviations: SD D sheeted dykes; HH D hornblende hornfels; PH D pyroxene hornfels; QD D quartz diorite; GS D gabbronorite series; gm D groundmass; 1 D mineral assemblage; 2 D plagioclase–amphibole geothermometer; 3 D orthopyroxene–clinopyroxene geothermometer; 4 D lava chemistry. Table 1 Summary of mineral compositions a Lithology Sheeted dykes Hornblende hornfels Pyroxene hornfels Quartz diorite Gabbronorite sequence CY4 sheeted dykes CY4 upper gabbros Plagioclase Clinopyroxene Orthopyroxene Amphibole Mg# b Mg# b (AlIV ) c Mg# b Primary Secondary An40–80 (avg. An65 ) An35–80 (peaks at An44–48 , An56–60 ) An65–95 An34–64 An85–90 ; An53–60 An49–90 (avg. An60–70 ) An80–95 An0–15 d – An5–15 d 0.64–0.78 An40–55 e ; An0–15 0.56–0.74 – An0–15 d 0.5–0.9 – 0.64–0.74 – 0.74–0.84 – – 0.56–0.74 – 0.70–0.85 0.59–0.66 0.68–0.78 0.11–1.04 0.27–1.15 0.14–1.30 0.28–1.25 0.11–1.52 – – 0.28–0.68 0.48–0.82 0.50–0.76 0.36–0.90 0.34–0.96 – – a Data were collected using JEOL 733 and 8900 microprobes at the Massachusetts Institute of Technology and University of Alberta, respectively; standard Bence–Albee corrections were used. b Mg# D Mg2C =(Mg2C C Fe2C ). c Amphiboles were recalculated on the basis of 23 anhydrous oxygens following the normalization of Robinson et al. d Adjacent to microfractures or patches within primary grain. e Narrow rims adjacent to amphibole. 232 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 3.1. Petrologic features The lowermost 200–300 m of the sheeted dyke complex is pervasively altered to greenschist facies mineral assemblages. Clinopyroxene is completely replaced by fibrous to granular amphibole, plagioclase is partially albitized, and interstitial zones are replaced by amphibole š quartz š chlorite š epidote. Where present, chlorite and quartz are minor phases (<5 modal%) and epidote occurs in trace amounts. Epidositized patches, where igneous minerals are completely replaced by epidote C quartz š chlorite š magnetite, were observed within a few isolated dykes. Larger scale epidosite bodies have been mapped to the NW [10] and NE of the Platanistasa Window. The contact aureole that separates the sheeted dyke complex and gabbronorite sequence is composed of massive, very fine grained outcrops of hornblende and=or pyroxene hornfels. Where both types of hornfels were identified, the hornblende hornfels formed further from the gabbronorite sequence than the pyroxene hornfels. Quartz diorite veins and apophyses, and amphibole veins cut the contact aureole. Amphibole vein networks with random orientations cross-cut some quartz diorite veins and are cut by other quartz diorite veins, suggesting that they formed during multiple fracturing events. Quartz diorite veins contain plagioclase C quartz C amphibole š clinopyroxene š Fe oxides š zircon š apatite and are typically coarser grained than the host hornfels. Hornblende hornfels are characterized by the assemblage granular amphibole C clinopyroxene C plagioclase C quartz C relict fibrous amphibole C relict clinopyroxene C magnetite C ilmenite. The extent of recrystallization varies from <20 to 100% over distances of mms to cms, with most samples showing <50% recrystallization. Zones with relict minerals have diabasic to intersertal textures that resemble the hydrothermally altered sheeted dykes, however, there is a notable absence of chlorite and epidote in the hornfels. Relict clinopyroxene is almost (>80%) completely replaced by fibrous amphibole whereas recrystallized clinopyroxene is finer grained, granoblastic, and unaltered. Granular amphibole has an interstitial to poikiolitic texture (Fig. 4A) or rims clots of fibrous amphibole that pseudomorphically replace clinopyroxene (Fig. 4B), suggesting that it formed by the recrystallization of fibrous amphibole. Plagioclase commonly retains its igneous habit but may have sutured grain boundaries and subgrain development in areas where clinopyroxene is recrystallized. Pyroxene hornfels are characterized by the assemblage clinopyroxene C orthopyroxene C plagioclase C granular amphibole C magnetite C ilmenite š quartz š relict fibrous amphibole. The degree of recrystallization ranges from ¾20 to 100%, with most samples showing >50% recrystallization. Relict zones display diabasic textures and contain fibrous amphibole, plagioclase, and quartz; no igneous clinopyroxene is preserved. Recrystallized zones contain discrete grains of granoblastic clinopyroxene and orthopyroxene (Fig. 4C) or aggregates of grains whose form mimics relict pyroxene grains (Fig. 4D). Pyroxene aggregates contain >75% clinopyroxene, minor orthopyroxene (<10%), and trace quartz (<5%), and are locally mantled by poikiolitic brown, granular amphibole. Plagioclase typically retains its igneous habit and is coarser grained than associated granoblastic pyroxene (Fig. 4C). Less commonly, plagioclase has a granoblastic texture and equivalent grain size to pyroxene (Fig. 4D) or exhibits subgrain development. The gabbronorite sequence is variably altered to greenschist to amphibolite facies mineral assemblages. The extent of alteration is less uniform and pervasive than the sheeted dykes. Pervasive replacement of igneous phases is generally localized along vein margins and areas infiltrated by felsic magmas. 3.1.1. Protolith of the contact aureole Petrological features indicate that the contact aureole is dominantly composed of recrystallized sheeted dykes. The hornblende hornfels have diabasic to intersertal textures, and primary mineral compositions are most similar to the sheeted dykes. The pyroxene hornfels have features of both the sheeted dykes and gabbros, suggesting that they represent both recrystallized dykes and gabbros. 3.2. Temperature constraints Peak temperatures are constrained by mineral compositions and assemblages, in conjunction with K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 233 Fig. 4. Photomicrographs showing typical textures and mineral assemblages in the hornfels. (A) Hornblende hornfels with granular, brown amphibole (ga) poikiolitically enclosing plagioclase (pl) laths (sample KG93090). (B) Hornblende hornfels showing fibrous green amphibole (fa) replacing clinopyroxene (c) and rimmed with granular amphibole (ga) (sample KG93056). (C) Pyroxene hornfels with very fine grained, granular aggregates of clinopyroxene and orthopyroxene (c-o); note that plagioclase (pl) and the pyroxene aggregates have similar grain size (sample KG93181). (D) Pyroxene hornfels showing a very fined grained granoblastic assemblage of plagioclase, orthopyroxene, clinopyroxene, magnetite, and ilmenite (sample KG93171). Field of view is 1.5 mm. experimental studies [25–27] and using two-pyroxene [28] and plagioclase–amphibole [29] thermometers (Table 2). The long-lived thermal conditions prevalent in the vicinity of the contact aureole are recorded by the metamorphic mineralogy (Fig. 3B). Hydrothermal mineral assemblages within the basal sheeted dykes indicate temperatures between 400 and 550ºC whereas granular, groundmass amphibole formed at higher temperatures (600–770ºC). Why higher temperatures are recorded by groundmass amphibole rather than the other hydrothermal minerals is not clear. One possibility is that granular amphibole formed during peak metamorphic conditions and that the remaining metamorphic minerals record conditions during the waning stages of hydrothermal convection. Alternatively, these mineral assemblages may record temperature fluctuations during the life span of a hydrothermal system. Transient increases in temperature may lead to the formation of granular amphibole, by the breakdown of greenschist facies minerals such as chlorite and actinolite (see discussion below). Within the contact aureole, hornblende and pyroxene hornfels record recrystallization temperatures of 778–904ºC and 940–986ºC, respectively. Quartz diorite veins cross-cutting the aureole crystallized at magmatic temperatures (817–919ºC), whereas the amphibole veins sealed fractures at 575–750ºC. Temperatures abruptly increased across the upper boundary with the sheeted dykes, from either ¾450 or 700ºC to 830ºC. Thermal conditions across the lower boundary are more difficult to constrain as magmatic temperatures for the gabbronorite sequence have not 234 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 Table 2 Calculated metamorphic temperatures (ºC) a Sample # Groundmass amphibole b Sheeted dykes KG93011 KG93036 KG93038 KG93143 KG93188 KG93067 689–770 (2) 732 (1) 612–734 (4) 710–718 (2) 636 (1) – Hornblende hornfels KG93056 850–875 KG93090 549–904 KG93098 892 (1) KG93125 752–862 KG93215 779 (1) KG93217 774–802 (2) (2) (4) (2) Pyroxene hornfels KG92056 823 (1) KG92091 806–897 (5) KG93023 649–940 (2) KG93079 – KG93171 – KG93181 730–950 (2) Gabbronorites KG92074 KG93030 KG93204 KG93205 KG93209 KG93210 KG93213 673–835 691–845 652–937 775–819 653–778 717–838 778–839 (8) (4) (5) (7) (7) (3) (2) Quartz diorite vein amphibole b – – – – – – Vein amphibole b 493 (1) – – – – 615–775 (8) – – – – – – 556–929 (6) 864–898 (3) 823–865 (2) – – – 810–847 (14) – – – – 575–657 (3) – 659–698 (5) 609–661 (6) – – – – – – – – – – 773 (1) – – – – – Groundmass pyroxene c – – – – – – – – – – – – 969–972 (2) 872–909 (3) 953–1020 (4) 940–986 (2) 939–973 (3) 910–972 (4) – – – – – – – a Temperature range for each sample; number of analyses is shown in parentheses. Plagioclase–amphibole thermometers of Holland and Blundy [25]. c Two-pyroxene thermometer of Andersen et al. [24]. b been calculated. Glass and phenocryst compositions for the lava sequence indicate crystallization temperatures between 1225 and 1000ºC [30]. Temperatures clearly increased across this boundary from ¾960ºC to super-solidus conditions. 4. Microthermometry Microthermometric fluid inclusion data were collected for five quartz diorite veins which cross-cut the hornblende and pyroxene hornfels (Table 3). Variations in room-temperature phase ratios, homogenization temperatures, and salinity suggest multiple populations of fluid inclusions with variable chemical and thermal histories [31]. The following section focuses on quartz-hosted, halite-bearing fluid inclusions in order to constrain the evolution and trapping conditions of high-salinity fluids within the contact aureole. Phase relations for halite-bearing fluid inclusions, which largely homogenize by halite dissolution, indicate trapping of a magmatic fluid at lithostatic conditions and relatively low temperatures. 4.1. Methods Microthermometric determinations were carried out on a Fluid Inc. heating and freezing petrographic K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 235 Table 3 Summary of fluid inclusion data Sample Lith. a Type n Th b (min–max) AVG d Tm c (min–max) NaCl e (wt%) (min–max) KG92091 PH 1 2 14 10 170–220 179–395 194 249 – 275–520 3.4–20.2 38.6–62.5 KG93090 HH 1 2 23 8 167–325 220–350 230 270 – 225–440 2.5–5.3 33.2–52.0 KG93098 HH 1 15 221–345 302 KG93125 HH 2 1 19 6 190–376 230–330 245 295 – 215–403 1.5–7.1 32.7–47.8 KG93127 HH 1 2 21 13 206–367 198–373 307 – 372–478 2.3–3.3 44.5–56.8 – 2.6–4.9 a PH D pyroxene hornfels, HH D hornblende hornfels. Th D temperature of vapor bubble homogenization (ºC). c T D temperature of halite dissolution (ºC). m d AVG D average value of n inclusions. e Salinity calculated as wt% NaCl equivalent. b stage following methods outlined in Roedder [32]. Heating and freezing data were collected from individual fluid inclusions with a range in size from 10 to 20 µm. Consistency of phase ratios and homogenization temperatures among fluid inclusions within secondary arrays was used to ensure representative data. Measurement accuracy is estimated at š0.1ºC for final ice-melting determinations (š0.2 wt% NaCl equivalent) and š5ºC for liquid–vapor homogenization (Th ) and halite melting (Tm ) events (š1.0 wt% NaCl eq.). Salinity calculations were performed using the fluid inclusion software package MacFlincor (version 0.77), with the equations of Brown and Lamb [33] for the H2 O–NaCl–[KCl] system. Fluid inclusions are modeled in the H2 O–NaCl system and salinity is reported as wt% NaCl equivalent. 4.2. Fluid inclusion types Fluid inclusions hosted within quartz are categorized into two types: (1) primary and secondary, lowsalinity, liquid-dominated inclusions and (2) primary and secondary, halite-bearing, liquid-dominated inclusions. Distinction between primary versus secondary inclusion genesis was equivocal in some samples due to the abundance of inclusions. Type 1 low-salinity fluid inclusions range between 1.5 and 8.6 wt% NaCl eq., and exhibit homogeniza- tion to the liquid at temperatures between 167 and 376ºC. Salinity values fall within the observed salinity range of fluids emanating from active hydrothermal vents (70% below to 200% above seawater values) [2]. Several rare liquid-dominated inclusions yielded salinities of ¾20 wt% NaCl eq. Type 2 halite-bearing, liquid-dominated inclusions appear to be primary and secondary in nature and are identified by the presence of one or more cubic halite daughter minerals. Halite-bearing inclusions yield three modes of homogenization behavior whereby the vapor bubble disappears at (1) temperatures less than halite melting temperatures (Th < Tm ), (2) temperatures equal to halite melting temperatures (Th D Tm ), and (3) temperatures greater than halite melting temperatures (Th > Tm ). Most inclusions display mode (1) behavior (92%) with vapor bubble homogenization temperatures of 179–395ºC and fluid salinities of 36–63 wt% NaCl (Tm D 277–520ºC) (Fig. 5). It should be noted that inclusions of 25–35 wt% NaCl may be under-represented due to the common failure of fluid inclusions within this salinity range to nucleate a halite crystal [34]. Many halite-bearing inclusions contain a reddish and=or black opaque daughter mineral that are tentatively identified as hematite and a sulphide phase, respectively. Quartz-hosted fluid inclusions from fault gouge 236 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 Fig. 5. Temperatures of halite dissolution (Tm ) and vapor bubble disappearance (Th ) for halite-bearing fluid inclusions. Black diamonds represent fluid inclusions hosted within quartz diorite veins which cross-cut the pyroxene and hornblende hornfels; clear circles represent fluid inclusions hosted within evolved plutonic rocks at the base of the contact aureole; the shaded field represents previously published data for fluid inclusions hosted within bodies of plagiogranite, epidosite and pegmatitic gabbro located proximal to the sheeted dyke–plutonic transition within the Troodos Ophiolite (from [42]). samples and evolved plutonic rocks from the field area display similar populations with several key differences [31]. Fluid inclusions within fault gouge samples yield low salinities and are mostly liquiddominated. In contrast, fluid inclusions from evolved plutonic rocks contain both low-salinity and halitebearing inclusions, as well as vapor-rich inclusions. 4.3. Conditions of fluid entrapment The minimum temperature of trapping for halitebearing inclusions which homogenize by halite dissolution (Th < Tm ) or by simultaneous vapor bubble disappearance and halite dissolution (Th D Tm ) is given by the temperature of halite dissolution [32]. As such, the trapping of high-salinity fluid (36– 63 wt% NaCl) within the quartz diorite veins occurred at minimum temperatures of 225–520ºC (avg. 402ºC). The lack of fluid inclusions which homogenize to the vapor phase, and the majority of halitebearing inclusions which homogenize by halite dissolution (Th < Tm ) indicate that the brines were not trapped as an immiscible fluid [32]. This ob- Fig. 6. (A) Temperature-pressure phase diagram for 40 wt% NaCl fluids in the H2 O–NaCl system. The curve L C V C H represents the pressure and temperature conditions whereby liquid, vapor and halite coexist in equilibrium ([58] and references therein); the critical curve separates the one-phase liquid field from the two-phase liquid–vapor field (subdivided into salinitydependent sub-fields) [59]; the dashed line represents the halite liquidus for a 40 wt% NaCl fluid [34]. The hashed region and the dot-dashed line represent possible P –T condition of trapping for fluid inclusions of 40 wt% NaCl which exhibit mode 1 (Th < Tm ) and mode 2 (Th D Tm ) homogenization behavior, respectively. Trapping conditions for inclusions which display mode 3 homogenization behavior (Th > Tm ) exist at temperatures greater than the line Th D Tm . Depth is modelled as a function of lithostatic pressure beneath the seafloor (see text). servation, in conjunction with consistent room temperature phase ratios, suggests that a single phase, homogeneous fluid was trapped. Isochoric projections were constructed to further constrain possible P –T conditions of fluid entrapment (Fig. 6). Using the equations of Bodnar and Vityk [34] isochores for the halite-bearing fluid inclusions hosted within the quartz diorite veins were projected from the three-phase curve (L C V C H) at their temperatures of vapor bubble homogenization. It should be noted that these equations are only valid for fluid inclusions of 0–40 wt% NaCl with homogenization temperatures between 50 and 700ºC at pressures below 6 kbar [34]. Isochores for halite-bearing inclusions with salinities greater that 40 wt% NaCl must be approximated using the slopes for inclusions of 40 wt% NaCl (Bodnar, pers. commun., 1998) because experimental data for fluids K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 with higher salinities are lacking. In addition, the true slope of isochores in the liquid C halite field have a slightly more positive slope than isochores in the one-phase liquid field [35]. Therefore, these projections only represent a first approximation of P –T trapping conditions for the halite-bearing inclusions. Based on the H2 O–NaCl model system, halitebearing fluid inclusions which homogenized by halite dissolution (Th < Tm ) have salinities that range from 36–63 wt.% NaCl. These inclusions were trapped in P –T space between their respective liquidus and the line which represents trapping conditions for inclusions which exhibit simultaneous vapor bubble disappearance and halite dissolution (Th D Tm ) (Fig. 6) [35]. Intersection of an isochore for a halite-bearing fluid inclusion that homogenized by halite dissolution (Th < Tm ) and its respective liquidus represents the minimum pressure of trapping for that inclusion [32]. Homogenization temperature data for halite-bearing fluid inclusions, modeled as 40 wt% NaCl fluids (see above), describe isochores which intersect the halite liquidus at pressures between 900 and 1200 bar (Fig. 6). Approximate isochores for inclusions with salinities greater than 40 wt% NaCl support minimum trapping pressures of ¾1000 bar. Although all isochores for inclusions which homogenize by halite dissolution (Th < Tm ) indicate a minimum pressure of trapping, inclusions for which the temperature of vapor bubble disappearance is significantly lower than halite dissolution record higher minimum pressures of trapping [36]. To place these minimum pressures in a geologic context, a model of hydrostatic and lithostatic pressure gradients with depth was constructed. Assuming Cretaceous ocean depths of 2000 m, a rock density of 2950 kg=m3 , and a water density of 1000 kg=m3 , lithostatic pressures of 900–1200 bar correspond to depths of 2275–3175 m below the seafloor, whereas the same pressures under hydrostatic conditions correspond to depths >7200 m below the seafloor. Field relations indicate that the paleo-depth of the contact aureole was 2000–2500 m below the seafloor. Based on available constraints these minimum trapping pressures suggest conditions of fluid entrapment at lithostatic pressures or greater, whereas trapping under hydrostatic conditions places the contact aureole at unrealistic depths. 237 5. Discussion The prevalence of an intrusive sheeted dyke– plutonic boundary in the Platanistasa Window is indicative of formation during a phase of magmatic extension. Abrupt truncation of the basal dykes at the contact aureole implies that the pluton intruded and assimilated the roots of the basal dykes. Age relationships for veins cross-cutting the aureole show that it was multiply fractured and subjected to significant oscillations in pressure and temperature (see below). We believe that the dynamic nature of the contact aureole is best explained by formation during a phase of active hydrothermal circulation and magmatic crustal construction. Many features of the contact aureole exposed in the Platanistasa Window are indicative of a conductive boundary layer. It is situated at the top of a magma chamber, at the base of the sheeted dykes, in an area that has undergone very little tectonic extension. The contact aureole is thin (¾10–30 m) and records steep thermal gradients across its boundaries. The overlying sheeted dyke complex is more uniformly and pervasively altered than the subjacent gabbronorite sequence. In the following section we explore the evolution of fluid compositions and P – T conditions in the vicinity of the contact aureole and develop a dynamic model that requires transient fracturing and mixing of hydrothermal and magmatic fluids at the magma–hydrothermal interface. 5.1. Evolution of fluid compositions Numerous fluid inclusion studies have documented the presence of high-salinity fluids within gabbroic rocks recovered from the modern ocean crust and its ophiolitic analogues (see [37]). Phase separation of seawater has been cited as a possible mechanism for the salinity and phase variations observed within hydrothermal vent fluids and deep-seated fluid inclusions (see [38]). However, few oceanic hydrothermal fluid inclusion studies have unequivocally identified fluids trapped as immiscible liquid=vapor pairs [37,39,40]. Vapor–brine density contrasts and rock fracture configurations have been suggested as mechanisms to explain high-salinity inclusions in the absence of vapors [40,41]. In contrast, it has been argued that the physical prop- 238 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 erties of seawater coupled with rapid healing rates of quartz micro-fractures would inhibit the migration of seawater further into the two-phase field, thereby limiting an end-member liquid phase to less than 30 wt% NaCl and hindering segregation of cogenetic liquids and vapors [42]. Consequently, Kelley [37] has argued that halite-bearing fluid inclusions from evolved plutonic rocks which homogenize by halite dissolution indicate trapping of high-salinity fluids exsolved from late stage melts in the absence of a vapor phase. Halite-bearing fluid inclusions which homogenize by halite dissolution are trapped in the absence of a vapor phase [32]. This condition requires that a highsalinity fluid of seawater or magmatic origin must be cooled substantially prior to trapping (Fig. 6) [35]. Although high-salinity, magmatic fluids can be exsolved into either the one- or two-phase field, phase separation of high-salinity fluids generates a low mass fraction of vapor to liquid [36]. The pervasive absence of cogenetic, low-salinity vapor condensates and the microthermometric constraints provided by the halite-bearing fluid inclusions trapped within the quartz diorite veins strongly suggests a magmatic origin for the high-salinity fluids. Although no temporal relationships could be determined, the halitebearing fluid inclusions are hosted among numerous populations of secondary, near-seawater salinity fluid inclusions. Mixing of hydrothermal fluids with higher temperature magmatic fluids is a possible mechanism for cooling the high-salinity magmatic fluids. The upper minimum pressure of trapping suggests that the high-salinity fluids were trapped, after cooling, near to or greater than lithostatic conditions. We conclude that taken together, the origin and condition under which the high-salinity fluids were trapped, reflect extreme fluctuations in temperature and pressure within the conductive boundary layer. 5.2. Oscillations in P –T conditions Cracking events occur within contact aureoles in response to large physical and chemical gradients which vary in time and space (see [43]). In the Platanistasa Window, the steady-state thermal conditions within the contact aureole are constrained by the hornfels which document high temperatures (778–986ºC). Fluid inclusion data and features of the aureole indicate that lithostatic conditions dominated. Phase relations for halite-bearing fluid inclusions constrain minimum trapping pressures of 0.9– 1.2 kbar at lithostatic conditions. Recrystallization temperatures for the hornfels are higher than plausible temperatures for the brittle–plastic transition in oceanic crust, placing the aureole at lithostatic pressures [44]. Constraints for the brittle–plastic transition are based on dry diabase flow laws (see discussion in Hirth et al. [44]) and petrologic evidence that incipient cracking within oceanic gabbros occurs at ½700ºC (e.g., [45]). Constraints from the vein networks show that steady-state temperature and pressure conditions were periodically perturbed within the conductive boundary layer. Amphibole vein networks and fluid inclusions trapped along microfractures hosted in quartz diorite veins record fracture sealing at 575– 750ºC and 225–520ºC, respectively. Quartz diorite veins and apophyses intruded at magmatic temperatures (817–919ºC). The macroscopic vein networks suggest that fracturing reduced pressures from lithostatic to hydrostatic conditions within the aureole. As these vein types show no preferred age relationships, we conclude that P –T conditions oscillated during the life span of the conductive boundary layer. Theoretical studies and field observations at modern mid-ocean ridges provide a basis for understanding the mechanisms that created the fracture networks within the Platanistasa conductive boundary layer. Dyke intrusion is an extensional event which brings magma to the seafloor from a magma chamber residing beneath the conductive boundary layer. Dyke injection occurs within a narrow zone (tens of meters) (see [46]), at intervals of years to centuries. Theoretical constraints suggest that high temperatures within the conductive boundary layer cause a build-up in thermoelastic stresses, compressive along the cool side of the boundary, tensile along the hot [47]. As a dyke penetrates this compressive layer, fractures will rapidly propagate into the underlying layer that is under tension. Theoretical constraints also predict that mechanical stresses within a sheeted dyke complex caused by dyke injection would close fractures parallel to the plane of a dyke and open fractures oriented perpendicular to the dyke [48]. These horizontal tensile and compressive stresses would extend well beyond the K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 dyke injection zone, affecting the entire axial rise at a fast-spreading ridge (i.e., up to 1.5 km from dyke margins) [48]. Over several years, thermal contraction fractures would open and mechanical stresses would decrease to ambient levels. Thus, dyke injection may trigger a fracturing event within the conductive boundary layer, and will close and open fractures within the overlying sheeted dyke complex over a time span of several years. Thermoelastic stresses may also create permeability at the boundary between hot rock and cool seawater [5,6]. This is the basis of Lister’s conceptual model which invokes a cracking front that migrates down into the crust, extracting heat from the hot rock through which it is penetrating. Cracking occurs at temperatures below the rigidus, which for basaltic rocks is probably ½700ºC [44,45]. The cracking front is commonly thought of in terms of the waning stage of a magma body as it moves out of the zone of dyke injection and off-axis. It may also develop during the life span of a magma body in response to fluctuations in magma supply [49]. Other mechanisms for fracturing involve magmatic and=or metamorphic fluids. Fluids exsolved by the crystallization of a melt act to increase the total pressure within the magma chamber until the maximum tensile strength of the overlying layer is exceeded. In shallow environments, water saturation and volatile exsolution is generally not achieved until late in the crystallization history (>90%) [50]. This process can be episodic, with subsequent cracking events requiring less overpressure to fracture the overlying rock or to reopen pre-existing fractures. The early prograde path of contact metamorphism is characterized by thermally induced pore fluid expansion and metamorphic devolatilization reactions which can produce pore fluid pressures in excess of the ambient lithostatic load (¾3–25 MPa), which are sufficient to hydrofracture the host rocks [51]. Other processes, however, such as crack healing and pressure solution result in an overall permeability structure dominated by intergranular fluid flow [43]. 5.3. Evolution of the Platanistasa Window conductive boundary layer A model that integrates geological and petrological constraints from the Platanistasa area is illus- 239 Fig. 7. Stages in the evolution of the Platanistasa conductive boundary layer. (A) During steady-state conditions a contact aureole (dark shading) separates sheeted dykes from a magma chamber below. A sharp thermal gradient exists at the top of the contact aureole where the conductive boundary layer (dashed line) separates hydrothermal fluids circulating under hydrostatic pressures from pore fluids at or above lithostatic conditions. (B) Episodic dyking generates fractures within the contact aureole, allowing penetration of hydrothermal fluids to deeper levels above a thinned conductive boundary layer. Sudden release in pressure promotes the expulsion of magmatic volatiles and allows for mixing of magmatic and hydrothenual fluids. (C) Fractures are subsequently sealed by mineral precipitation which effectively shifts the conductive boundary layer upwards. The new thermal structure results in recrystallization and assimilation of the injected dyke. (D) Cessation of magmatic activity is marked by the downward migration of thermal contraction fractures such that the conductive boundary layer subsides into the gabbronorite sequence. trated in Fig. 7. A steady-state hydrothermal system has been established above a magma chamber. The conductive boundary layer coincides with the contact aureole which is composed of previously hydrated sheeted dykes that are undergoing recrystallization to hornblende and pyroxene hornfels. Lithostatic pressures (¾1 kbar) and high temperatures (>750ºC) are prevalent within the conductive boundary layer. There is a steep thermal gradient between the conductive boundary layer and overlying hydrothermal system where hydrothermal fluids are convecting at temperatures ½350ºC. The subjacent magma cham- 240 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 ber is at super-solidus conditions and is undergoing complex crystallization involving the assimilation of previously stoped, hydrated dyke roots and magmatic differentiation. Fracturing disrupts these steady-state conditions and facilitates the migration of magmatic and hydrothermal fluids into the conductive boundary layer from below and above, respectively. As outlined in the previous section, we must rely on theoretical studies and observations from modern mid-ocean ridges to incorporate fracturing into our model for the Platanistasa conductive boundary layer. Field and petrological data document oscillations in P –T conditions. Cross-cutting relations between the magmatic and hydrothermal vein networks show that fracturing occurred prior to the cessation of magmatic activity. We cannot constrain, however, if the preserved features of the conductive boundary layer formed within an axial zone of spreading or off-axis during the waning stages of magmatic activity. If we consider that the conductive boundary layer evolved within an axial zone, steady-state conditions (Fig. 7A) are perturbed by the injection of a dyke into the brittle crust creating or re-opening fractures within the conductive boundary layer due to thermoelastic and mechanical stresses (Fig. 7B). Enhanced permeability causes the overlying hydrothermal system to deepen and the conductive boundary layer to thin which in turn increases the heat flux of the hydrothermal system [7,52]. Within this newly fractured zone, temperatures decrease and pressures oscillate from lithostatic to hydrostatic conditions (¾1 kbar to ¾400 bar). Sudden release in pressure promotes the expulsion of magmatic fluids into the overlying hydrothermal system, as well as metamorphic pore fluids generated by dehydration reactions within the contact aureole, causing the complex mixing of magmatic, metamorphic, and hydrothermal fluids (Fig. 7B) [43]. Magmatic and metamorphic fluids cool by mixing with cooler, hydrothermal fluids. A rapid drop in pressure, from lithostatic to hydrostatic conditions, promotes crystallization of the quartz diorite veins [50]. At the same time, hydrothermal fluids from the overlying system infiltrate the newly fractured zone and are heated by mixing with hotter magmatic and metamorphic fluids and=or by conduction from the wallrock. Temperature increase beyond the quartz solubility maxima leads to quartz precipitation [53] and the sealing of microfractures within the quartz diorite veins at pressures between hydrostatic and lithostatic. The abundance of amphibole veins within the aureole suggests that increasing hydrothermal fluid temperatures also promotes the precipitation of amphibole (e.g., [27]); however, we are not familiar with any theoretical calculations of amphibole solubility. As the permeability is reduced by mineral precipitation, the conductive boundary layer thickens upward until the system returns to steady-state conditions whereby the magmatic heat input is balanced by hydrothermal heat extraction (Fig. 7C). By this time, lithostatic pressures and high temperatures are restored within the conductive layer, causing the root of the recently injected dyke to be recrystallized within the contact aureole and assimilated within the roof-zone of the magma chamber. This completes a typical cycle (Fig. 7B,C). The time scale of a typical cycle is not well constrained but is likely to be similar to the frequency of dike intrusion (<10–100 years). Seismic monitoring of mid-ocean ridges suggests that dyke propagation lasts in the order of days [54]. Thermal effects in the wall rock adjacent to an intruding dyke last hours to days [55]. Mechanical stresses are restored to ambient values within several years [48]. Thus, fracture propagation associated with dyking is short-lived. Rates of fracture sealing by mineral precipitation are assumed to be rapid at high temperatures, with the time scale being dependent on parameters such as fracture geometry, fluid compositions, reaction kinetics, and flow velocities [56]. At the P –T conditions documented for the Platanistasa conductive boundary layer, microfracture sealing rates for quartz in equilibrium with saline solutions have time scales of hours to days [57]. At conditions above the quartz saturation maxima, quartz precipitation would reduce the porosity by a few volume percent on a time scale of decades [56]. Thus, it is probable that the permeability created by dyking is significantly reduced prior to the next dyking event. The thickness of the conductive boundary layer is, in part, linked to the time interval between dyking events. During extended periods of quiescence, the conductive boundary layer may thicken up into the sheeted dyke complex [49], beyond the upper boundary of the contact aureole, promoting dehydration re- K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 actions that lead to the formation of amphibole. The extent of recrystallization would be limited by the permeability of the basal dykes and the longevity of the thermal excursion. Within the Platanistasa Window, heterogeneous distribution of high temperature amphibole within sheeted dykes altered to greenschist facies mineral assemblages is consistent with transient upward migration of higher temperature conditions. In summary, steady-state conditions within the conductive boundary layer developed above a magma chamber within the axial zone of spreading (Fig. 7C). Dyke injection causes the conductive boundary layer to fracture, which facilitates the migration and mixing of magmatic and hydrothermal fluids into the conductive boundary layer and the sealing of fractures by mineral precipitation (Fig. 7B). As steady-state conditions are restored (Fig. 7C), a typical cycle is completed. The conductive boundary layer undergoes multiple cycles prior to moving out of the axial zone. The time scale of each cycle is comparable to the frequency of dyking. An equally plausible model is that the Platanistasa conductive boundary layer solely records conditions during the waning stages of magmatic activity. In this off-axis setting, exsolution of magmatic fluids within the roof-zone of the magma chamber increases the total pressure beyond the maximum tensile strength of the thinned boundary layer and induces hydrofracturing [50]. As outlined above, fracturing would cause a change from lithostatic to hydrostatic conditions, and mixing of magmatic, metamorphic, and seawater-derived hydrothermal fluids within the conductive boundary layer. More than one fracturing event is required to produce the distribution of magmatic veins cutting hydrothermal veins, and vice versa. The final event for both our proposed models, which consider the features formed either within an axial zone or off-axis, is the solidification of the magma chamber (Fig. 7D). At this time, the conductive boundary layer migrates down into the gabbronorite sequence, followed by a zone of incipient cracking — the cracking front of Lister [5,6]. Cracking is initiated at conditions equivalent to the brittle–ductile transition, and sealed at temperatures of 575–750ºC. In this section, we have attempted to reconstruct 241 the causes and consequences of fracturing within the conductive boundary layer exposed in the Platanistasa Window. We have outlined two plausible models that are consistent with our geological and petrological data, theoretical considerations, and observations from modern mid-ocean ridges. Clearly, other models that consider the timing of dyking, volatile build-up, and cracking fronts in different ways are equally plausible. 6. Summary The prevalence of an intrusive sheeted dyke– plutonic boundary in the Platanistasa Window is indicative of formation during a phase of magmatic extension. Abrupt truncation of the basal dykes at a contact aureole implies that the pluton intruded and assimilated the roots of the basal dykes. Many features of the contact aureole are indicative of a conductive boundary layer that separates an active hydrothermal system from the heat source that drives it. The contact aureole formed within the basal sheeted dykes and uppermost gabbros due to the intrusion of a layered gabbronorite sequence. Mineralogical characteristics of the hornfels indicate that the basal dykes were hydrated prior to recrystallization, by previous or on-going hydrothermal processes, and that recrystallization occurred at amphibolite to granulite facies conditions (778–986ºC) and lithostatic pressures. Field and petrological data document oscillations in P –T conditions. Cross-cutting relations between magmatic and hydrothermal vein networks show that fracturing occurred prior to the cessation of magmatic activity. We speculate that the fracture networks formed either within the axial zone of spreading or during the waning stages of magmatic activity. As a starting point, the hydrothermal system is considered to be at steady-state conditions (i.e., between fracturing events) and the conductive boundary layer is located at or near the top of the contact aureole. If the preserved features of the conductive boundary layer form within the axial zone, dyke injection causes fracturing within the conductive boundary layer, causing it to thin and migrate toward the base of the contact aureole. Fracture propagation promotes the mixing of magmatic, metamorphic, and 242 K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 hydrothermal fluids, and leads to the deposition of quartz diorite and amphibole veins. Subsequent fracture sealing results in the resumption of steady-state conditions. The conductive boundary layer is subjected to multiple fracturing events prior to moving out of the axial zone. If the preserved features of the conductive boundary layer record conditions during the waning stages of magmatic activity, volatile build-up induces hydrofracturing, which again facilitates the mixing of magmatic, metamorphic, and hydrothermal fluids and the formation of quartz diorite and amphibole veins. Once a magma chamber solidifies off-axis, oscillations in temperature and pressure cease, and the conductive boundary layer migrates downward into the plutonic sequence. We recognize that we present two of many plausible models for the causes and consequences of fracturing within the conductive boundary layer. [3] [4] [5] [6] [7] [8] [9] Acknowledgements We are grateful to C. Xenophontos and the Cyprus Geological Survey Department for their on-going support of geological research programs. We thank the Hellenic Mining Company for logistical support in the field. The focus of this project evolved through many lively discussions in the field with J. Cann, J. Malpas, C. Xenophontos, K. Deveau, S. Agar, L. Marquez, and K. Klitgord. We acknowledge N. Chatterjee and R. Luth for their assistance with microprobe analyses and K. Sapp for her assistance with many aspects of the project during its early stages. We also thank W. Wilcox and an anonymous reviewer for their thorough reviews. This research was supported by NSF (EAR-9118791) and NSERC research grants. [CL] [10] [11] [12] [13] [14] [15] References [16] [1] P.J. Saccocia, K. Ding, M.E. Berndt, J.S. Seewald, W.E. Seyfried, Experimental and theoretical perspectives on crustal alteration at mid-ocean ridges, in: Lentz, D.R. (Ed.), Alteration and Alteration Processes Associated with Ore-forming Systems, Vol. 11, Geol. Assoc. Canada Short Course Notes, St. John’s, 1994, pp. 403–432. [2] K. Von Damm, Controls on the chemistry and temporal variability of seafloor hydrothermal fluids, in: Humphris, [17] [18] [19] S.E., Zierenberg, R.A., Mullineaux, L.S., Thomson, R.E. (Eds.), Seafloor Hydrothermal Systems, Geophys. Monogr. Ser. 91, Am. Geophys. Union, Washington, DC, 1995, pp. 222–247. D. Kadko, J. Baross, J. Alt, The magnitude and global implications of hydrothermal flux, in: Humphris, S.E., Zierenberg, R.A., Mullineaux, L.S., Thomson, R.E. (Eds.), Seafloor Hydrothermal Systems, Geophys. Monogr. Ser. 91, Am. Geophys. Union, Washington, DC, 1995, pp. 446–466. R.P. Lowell, P.A. Rona, R.P. Von Herzen, Seafloor hydrothermal systems, J. Geophys. Res. 100 (1995) 327–352. C.R.B. Lister, On the penetration of water into hot rock, Geophys. J. R. Astron. Soc. 39 (1974) 465–509. C.R.B. Lister, On the intermittency and crystallization mechanisms of sub-seafloor magma chambers, Geophys. J. R. Astron. Soc. 73 (1983) 351–366. C.R.B. Lister, Heat transfer between magmas and hydrothermal systems, or, six lemmas in search of a theorem, Geophys. J. Int. 120 (1995) 45–59. R.P. Lowell, L.N. Germanovich, Dike injection and the formation of megaplumes at ocean ridges, Science 267 (1995) 1804–1807. S.E. Humphris, P.M. Herzig, D.J. Miller et al., The internal structure of an active sea-floor massive sulphide deposit, Nature 377 (1995) 713–716. C.J. Richardson, J.R. Cann, H.G. Richards, J.G. Cowan, Metal-depleted root zones of the Troodos ore-forming hydrothermal systems, Cyprus, Earth Planet. Sci. Lett. 84 (1987) 243–253. P. Schiffman, B.M. Smith, R.J. Varga, E.M. Moores, Geometry, conditions and timing of off-axis hydrothermal metamorphism and ore-deposition in the Solea graben, Nature 325 (1987) 423–425. A. Robertson, C. Xenophontos, Development of concepts concerning the Troodos ophiolite and adjacent units in Cyprus, in: Prichard, H.M., Alabaster, T., Harris, N.B.W., Neary, C.R. (Eds.), Magmatic Processes and Plate Tectonics, Geol. Soc. Spec. Publ. 76 (1993) 85–119. L.M. Bear, The geology and mineral resources of the Akaki–Lythrodondha area, Government of Cyprus, Nicosia, 1960, 122 pp. J. Malpas, T. Brace, The geology of Pano Amiandos– Palekhori area, Cyprus, Geological Survey Department, Nicosia, 1987. S.B. Mukasa, J.N. Ludden, Uranium-lead isotopic ages of plagiogranites from the Troodos ophiolite, Cyprus and their tectonic significance, Geology 15 (1987) 825–828. J.A. Pearce, S.J. Lippard, S. Roberts, Characteristics and tectonic significance of supra-subduction zone ophiolites, Geol. Soc. London 16 (1984) 77–94. K.M. Gillis, P.T. Robinson, Patterns and processes of alteration in the lavas and dykes of the Troodos Ophiolite, Cyprus, J. Geophys. Res. 95 (1990) 21523–21548. R.J. Varga, E.M. Moores, Spreading structure of the Troodos ophiolite, Cyprus, Geology 13 (1985) 846–850. S. Allerton, F.J. Vine, Spreading structure of the Troodos K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] ophiolite, Cyprus: some paleomagnetic constraints, Geology 15 (1987) 593–597. Varga, E.M. Moores, Intermittent magmatic spreading and tectonic extension in the Troodos Ophiolite: implications for exploration of black smoker-type ore deposits, in: Malpas, J., Moores, E.M., Panayiotou, A., Xenophontos, C. (Eds.), Ophiolites: Oceanic Crustal Analogues, Cyprus Geological Survey Department, Nicosia, 1990, pp. 53–64. J. Malpas, T. Brace, S.M. Dunsworth, Structural and petrologic relationships of the CY-4 drill hole of the Cyprus Crustal Study Group, in: Gibson, I.L., Malpas, J., Robinson, P.T., Xenophontos, C. (Eds.), Cyprus Crustal Study Project: Initial Report, Hole CY-4, Geological Survey of Canada Paper 88-9, Geological Survey of Canada, Ottawa, 1989, pp. 39–68. E.M. Moores, F.J. Vine, The Troodos Massif, Cyprus and other ophiolites as oceanic crust: evaluation and implications, Phil. Trans. R. Soc. London A 268 (1971) 443–466. J. Malpas, Crustal accretionary processes in the Troodos ophiolite, Cyprus: evidence from field mapping and deep crustal drilling, in: Malpas, J., Moores, E.M., Panayiotou, A., Xenophontos, C. (Eds.), Ophiolites: Ocean Crustal Analogues, Geological Survey Department, Nicosia, 1990, pp. 65–74. C.R. Allen, The petrology of a portion of the Troodos plutonic complex, Cyprus, Ph.D. thesis, Cambridge University, 1975. J.G. Liou, S. Kuniyoshi, K. Ito, Experimental studies of the phase relations between greenschist and amphiobolite in a basaltic system, Am. J. Sci. 274 (1974) 613–632. M.J. Apted, J.G. Liou, Phase relations among greenschist, epidote–amphibolite, and amphibolite in a basaltic system, Am. J. Sci. 283A (1983) 328–354. F.S. Spear, An experimental study of hornblende stability and compositional variability in amphibolite, Am. J. Sci. 281 (1981) 697–734. D.L. Andersen, D.H. Lindsley, P.M. Davidson, Quilf: A Pascal program to assess equilibria among Fe–Mg–Mn–Ti oxides, pyroxenes, olivine and quartz, Comput. Geosci. 19 (1993) 1333–1350. T. Holland, J. Blundy, Non-ideal interactions in calcic amphiboles and their bearing on amphibole–plagioclase thermometry, Contrib. Mineral. Petrol. 116 (1994) 433–447. P. Thy, C. Xenophontos, Crystallization orders and phase chemistry of glassy lavas from the pillow sequences, Troodos Ophiolite, Cyprus, J. Petrol. 32 (1991) 403–428. M.D. Roberts, The role of magmatic fluids and phase separation at the sheeted dyke–plutonic transition: a fluid inclusion study from the Troodos Ophiolite, Cyprus, B.Sc. thesis, University of Victoria, 1998. E. Roedder, Fluid inclusions, Mineralogical Society of America, Washington, DC, 1984, 644 pp. P.E. Brown, W.M. Lamb, P –V –T properties of fluids in the system H2 O š CO2 š NaCl: new graphical presentations and implications for fluid inclusions studies, Geochim. Cosmochim. Acta 53 (1989) 1209–1221. R.J. Bodnar, M.O. Vityk, Interpretation of microthermo- [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] 243 metric data for H2 O–NaCl fluid inclusions, in: De Vivo, B., Frezzotti, M.L. (Eds.), Fluid Inclusions in Minerals: Methods and Applications, Virginia Polytechnic Institute and State University, Pontignana–Siena, 1994, pp. 117–130. R.J. Bodnar, Synthetic fluid inclusions: XII. The system H2 O–NaCl. Experimental determination of the halite liquidus and isochores for a 40 wt% NaCl solution, Geochim. Cosmochim. Acta 58 (1994) 1053–1063. J.S. Cline, D.A. Vanko, Magmatically generated saline brines related to molybdenum at Questa, New Mexico, USA, in: Thompson, J.F.H. (Ed.), Magmas, Fluids, and Ore Deposits, Mineral. Assoc. Canada Short Course Ser. 23, Mineralogical Association of Canada, Nepean, 1995, pp. 153–174. D.S. Kelley, Fluid evolution in slow-spreading environments, in: Karson, J.A., Cannat, M., Miller, D.J., Elthon, D. (Eds.), Proc. ODP, Sci. Results, Vol. 153, Ocean Drill. Program, College Station, TX, 1997, pp. 399–415. P. Nehlig, Interactions between magma chambers and hydrothermal systems: oceanic and ophiolitic constraints, J. Geophys. Res. 98 (1993) 19621–19633. D.S. Kelley, J.R. Delaney, Two-phase separation and fracturing in mid-ocean ridge gabbros at temperatures greater than 700ºC, Earth Planet. Sci. Lett. 83 (1987) 53–66. D.A. Vanko, R.J. Bodnar, S.M. Sterner, Synthetic fluid inclusions: VIII. Vapor-saturated halite solubility in part of the system NaCl–CaCl2 –H2 O, with application to fluid inclusions from oceanic hydrothermal systems, Geochim. Cosmochim. Acta 52 (1988) 2451–2456. M.S. Goldfarb, J.R. Delaney, Response of two-phase fluids to fracture configurations within submarine hydrothermal systems, J. Geophys. Res. 93 (1988) 4585–4594. D.S. Kelley, P.T. Robinson, J.G. Malpas, Processes of brine generation and circulation in the oceanic crust: fluid inclusion evidence from the Troodos Ophiolite, Cyprus, J. Geophys. Res. 97 (1992) 9307–9322. R.B. Hanson, The hydrodynamics of contact metamorphism, Geol. Soc. Am. Bull. 107 (1995) 595–611. G. Hirth, J. Escartin, J. Lin, The rheology of the lower oceanic crust: implications for lithospheric deformation at mid-ocean ridges, in: Buck, W.R., Delaney, P.T., Karson, J.A., Lagabrielle, Y. (Eds.), Faulting and Magmatism at Mid-Ocean Ridges, Geophys. Monogr. Ser. 106, American Geological Union, Washington, DC, 1998. C. Manning, P.E. Weston, K.I. Mahon, Rapid high temperature metamorphism of the East Pacific Rise gabbros from Hess Deep, Earth Planet. Sci. lett. 144 (1996) 123–132. E. Hooft, H. Schouten, R.S. Detrick, Constraining crustal emplacement processes from the variation in seismic layer 2A thickness at the East Pacific Rise, Earth Planet. Sci. Lett. 142 (1996) 289–309. C.R.B. Lister, Differential stress in the earth, Geophys. J. R. Astron. Soc. 86 (1986) 319–330. D. Curewitz, J.A. Karson, Geological consequences of dike intrusion at mid-ocean ridges, in: Buck, W.R., Delaney, P.T., Karson, J.A., Lagabrielle, Y. (Eds.), Faulting and Mag- 244 [49] [50] [51] [52] [53] K.M. Gillis, M.D. Roberts / Earth and Planetary Science Letters 169 (1999) 227–244 matism at Mid-Ocean Ridges, Geophys. Monogr. Ser. 106, American Geophysical Union, Washington, DC, 1998. W.S.D. Wilcock, J.R. Delaney, Mid-ocean ridge sulfide deposits: Evidence for heat extraction from magma chambers or cracking fronts?, Earth Planet. Sci. Lett. 145 (1996) 49– 64. C.W. Burnham, Magmas and hydrothermal fluids, in: Barnes, H.L. (Ed.), Geochemistry of Hydrothermal Ore Deposits, John Wiley, New York, NY, 1997, pp. 63–124. B. Dutrow, D. Norton, Evolution of fluid pressure and fracture propagation during contact metamorphism, J. Metamorph. Geol. 13 (1995) 677–686. R.P. Lowell, L.N. Germanovich, On the temporal evolution of high-temperature hydrothermal systems at ocean ridge crests, J. Geophys. Res. 99 (1994) 565–575. C.J. Bruton, H.C. Helgeson, Calculation of the chemical and thermodynamic consequences of differences between fluid and geostatic pressure in hydrothermal systems, Am. J. Sci. 283A (1983) 540–588. [54] R.P. Dziak, C.G. Fox, A.E. Schreiner, The June–July 1993 seismo-acoustic event at CoAxial segment, Juan de Fuca Ridge: evidence for a lateral dike injection, Geophys. Res. Lett. 22 (1995) 135–138. [55] P.T. Delaney, D.D. Pollard, Solidification of basaltic magma during flow in a dike, Am. J. Sci. 282 (1982) 856–885. [56] J.T. Wells, M.S. Ghiorso, Coupled fluid flow and reaction in mid-ocean ridge hydrothermal systems: The behavior of silica, Geochim. Cosmochim. Acta 55 (1991) 2467–2481. [57] S.L. Brantley, The effect of fluid chemistry on quartz microcrack lifetimes, Earth Planet. Sci. Lett. 113 (1992) 145– 156. [58] J.L. Bischoff, Densities of liquids and vapors in boiling NaCl–H2 O solutions: a PVTX summary from 300º to 500ºC, Am. J. Sci. 291 (1991) 309–338. [59] S. Sourirajan, G.C. Kennedy, The system H2 O–NaCl at elevated temperatures and pressures, Am. J. Sci. 260 (1962) 115–141.
© Copyright 2026 Paperzz